Medicinski Glasnik Official Publication of the Medical Association of Zenica-Doboj Canton Bosnia and Herzegovina

Published and copyright by: Medical Assotiation of Zenica-Doboj Canton; Address: Zenica, 72000, Bulevar kralja Tvrtka I 4, Bosnia and Herzegovina; tel./fax: +387 32 444 270; Email: [email protected], [email protected], web site: http//www.ljkzedo.ba For ordering information please contact: Jasenko Žilo, [email protected]; Access to this journal is available free online trough: www.ljkzedo.ba The Journal is indexed by MEDLINE, EMBASE (Exerpta Medica), Scopus, EBSCO; ISSN 1840-0132 DTP by: Graphic and web design studio “B Panel” Zenica, Zmaja od Bosne bb, www.bpanel.ba, e-mail: [email protected], tel. +387 32 441 290, 441 291; Printed by: ______Medicinski Glasnik Official Publication of the Medical Association of Zenica-Doboj Canton Bosnia and Herzegovina

EDITOR-IN-CHIEF Selma Uzunović, Zenica, Bosnia and Herzegovina

DEPUTY EDITOR Besim Prnjavorac, Tešanj, Bosnia and Herzegovina

RESEARCH INTEGRITY EDITOR Larisa Gavran, Zenica, Bosnia and Herzegovina

MANAGING EDITOR Tarik Kapidžić, Zenica, Bosnia and Herzegovina

EDITORS Solmaz Abdolrahimzadeh, Rome, Italy Luiz Ronaldo Alberti, Belo Horizonte, Brazil Mutay Aslan, Antalya, Turkey Adem Balić, Tuzla, Bosnia and Herzegovina Dubraka Bartolek, Zagreb, Croatia Branka Bedenić, Zagreb, Croatia Iva Christova, Sofia, Bulgaria Asja Čelebić, Zagreb, Croatia Josip Čulig, Zagreb, Croatia Filip Čulo, Zagreb, Croatia Jordan Dimanovski, Zagreb, Croatia Branko Dmitrović, Osijek, Croatia Davorin Đanić, Slavonski Brod, Croatia Ines Drenjančević, Osijek, Croatia Harun Drljević, Zenica, Bosnia and Herzegovina Mukaddes Esrefoglu, Istanbul, Turkey Ivan Fistonić, Zagreb, Croatia Roberta Granese, Messina, Italy Simona Gurzu, Tȋrgu Mureş, Romania Diane Medved Harper, Louisville, United State Lejla Ibrahimagić-Šeper, Zenica, Bosnia and Herzegovina Tatjana Ille, Ajman, United Arab Emirates Slobodan M. Janković, Kragujevac, Serbia Vjekoslav Jerolimov, Zagreb, Croatia Ioan Jung, Tȋrgu Mureş, Romania David Kovacevic, New Haven, United States Sven Kurbel, Osijek, Croatia Snježana Pejičić, Banja Luka, Bosnia and Herzegovina Belma Pojskić, Zenica, Bosnia and Herzegovina Asja Prohić, Sarajevo, Bosna Hercegovina Velimir Profozić, Zagreb, Croatia Radivoje Radić, Osijek, Croatia Amira Redžić, Sarajevo, Bosnia and Herzegovina Halima Resic, Sarajevo, Bosnia and Herzegovina Suad Sivić, Zenica, Bosnia and Herzegovina Sonja Smole-Možina, Ljubljana, Slovenia Vladimir Šimunović, Mostar, Bosnia and Herzegovina Ekaterine Tskitishvili, Liege, Belgium Aylin Türel Ermertcan, Manisa, Turkey Adrijana Vince, Zagreb, Croatia Jasmina Vraneš, Zagreb, Croatia

EDITORIAL ASSISTANT Hakija Bečulić, Zenica, Bosnia and Herzegovina

Secretary: Jasenko Žilo Proofreaders: Glorija Alić (English) MEDICINSKI GLASNIK Official Publication of the Medical Association of Zenica-Doboj Canton, Bosnia and Herzegovina Volume 17, Number 2, August 2020 Free full-text online at: www.ljkzedo.com.ba, and www.doaj.org (DOAJ, Directory of Open Access Journals)

Review 246 Interdisciplinary aspects of possible negative effects of dogs on humans in Bosnia and Herzegovina Muhamed Katica, Zarema Obradović, Nasreldin Hassan Ahmed, Enra Mehmedika-Suljić, Žana Stanić, Rowida Seifeldin Abdalaziz Mohamed, Emina Dervišević 252 Securing the base of the appendix during laparoscopic appendectomy Samir Delibegović, Reuf Karabeg, Milan Simatović 256 Epidemiology of injuries connected with dance: a critical review on epidemiology Giuseppe Rinonapoli, Marta Graziani, Paolo Ceccarini, Cristina Razzano, Francesco Manfreda, Auro Caraffa Original 265 Spatio-temporal data visualization for monitoring of control measures in the prevention of the spread of COVID-19 in Bosnia and Herzegovina article Mirza Ponjavić, Almir Karabegović, Elvir Ferhatbegović, Emin Tahirović, Selma Uzunović, Maja Travar, Aida Pilav, Maida Mulić, Sead Karakaš, Nermin Avdić, Zarina Mulabdić, Goran Pavić, Medina Bičo, Ivan Vasilj, Diana Mamić, Mirsada Hukić 275 Minimal impact of COVID-19 outbreak on the postoperative morbidity and mortality following emergency general surgery procedures: results from a 3-month observational period Charalampos Seretis, Lucy Archer, Lida Lalou, Shuker Yahia, Christian Katz, Iram Parwaiz, Altaf Haji, Lourdusamy Selvam 279 Model "P" in gender prediction based on the mastoid process Aida Sarač-Hadžihalilović, Edin Hojkurić, Miralem Musić, Ilvana Hasanbegović, Zurifa Ajanović, Lejla Dervišević, Selmira Brkić 285 The preemptive effect of tramadol and metamizole on the intensity of postoperative pain Ismet Suljević, Muamer Hadžiavdić, Ismana Šurković, Omer Suljević, Maida Turan, Ehlimana Mušija 290 Characteristics of autologous peripheral blood stem cells collection over a one-year period Gorana Ahmetović-Karić, Elma Ćatović-Baralija, Alma Sofo-Hafizović 297 ABO groups as a risk factor for myocardial infarction in the population of Bosnia and Herzegovina Ermina Mujičić, Nermir Granov, Armin Šljivo, Amina Selimović 301 A novel flavivirus strain detected in phlebotomine sandflies in Bosnia and Herzegovina Mirsada Hukić, Halida Avdihodžić, Ivan-Christian Kurolt, Alemka Markotić, Jasna Hanjalić, Lejla Kapur- Pojskić, Irma Mahmuljin, Monia Avdić, Irma Salimović-Bešić, Lejla Smajlović-Skenderagić 308 Phenotypic and genotypic detection of ESBL-producing E. coli isolates from chicken skin in Bosnia and Herzegovina Vildana Hadžić-Hasanović, Anesa Jerković-Mujkić, Ermin Hasanović, Azra Bačić, Mirsada Hukić 316 Association of endothelial nitric oxide synthase gene (G894T) polymorphism and hypertension in diabetic Bataknese patients Jelita Siregar, Ratna Akbari Ganie, Dharma Lindarto, Erna Mutiara, Delfitri Munir 321 The effect of bay leaf extract (Syzygium polyanthum) on vascular endothelial growth factor (VEGF) and CD31 (PECAM-1) expression in acute coronary syndrome Refli Hasan, Gontar Alamsyah Siregar, Dharma Lindarto 328 Neutrophil gelatinase-associated lipocalin is a predictor of complications in the early phase of ST- elevation myocardial infarction Nirvana Šabanović-Bajramović, Enisa Hodžić, Amer Iglica, Edin Begić, Nerma Resić, Kenana Aganović, Mirela Halilčević, Senad Bajramović 335 Cardiac biomarkers and left ventricular systolic function in acute myocardial infarction with ST- segment elevation in diabetes mellitus type 2 patients Mirela Halilčević, Edin Begić, Alen Džubur, Nirvana Šabanović-Bajramović, Mevludin Mekić, Malik Ejubović, Amela Džubur , Ena Štimjanin 341 Analysis of two single-nucleotide polymorphisms (rs2241766 and rs1501299) of the adiponectin gene in patients with coronary artery disease and coronary slow flow Muhammad Diah, Aznan Lelo, Zulfikri Muhktar, Dharma Lindarto 346 Interleukin 6 concentration elevation as a risk of carotid intima-media thickness in chronic kidney disease patients with dialysis Riri Andri Muzasti, Herman Hariman, Elvita Rahmi Daulay 352 Pectoralis (PecS) nerve block 1 for port-a-cath removal and central venous catheter (CVC) replacement Massimo Renzini, Umberto Ripani, Luisa Golia, Fulvio Nisi, Fabio Gori

356 Treatment effectiveness and outcome in patients with a relapse and newly diagnosed multidrug- resistant pulmonary tuberculosis Dmytro Butov, Valeriy Myasoedov, Mykola Gumeniuk, Galyna Gumeniuk, Oleksandra Choporova, Anton Tkachenko, Oleksandra Akymenko, Olena Borysova, Olena Goptsii, Yevhenii Vorobiov, Tetiana Butova

363 Preoperative tumour size as a predictor of the presence of lymphovascular invasion in lung adenocarcinoma Kemal Grbić1, Bakir Mehić, Dalma Udovičić-Gagula, Amina Valjevac, Adem Ćemerlić, Ferid Krupić

369 ABO blood group genotypes and ventilatory dysfunction in patients with allergic and nonallergic asthma Marina Lampalo, Irena Jukić, Jasna Bingulac-Popović, Hana Safić Stanić, Nikola Ferara, Sanja Popović-Grle

375 Epidemiology of neonatal sepsis caused by multidrug resistant pathogens in a neonatal intensive care unit level 3 Devleta Hadžić, Fahrija Skokić, Selmira Brkić, Amina Saračević, Delila Softić, Dženana Softić

383 Interleukin-8 heterozygous polymorphism (-251 T/A and +781 C/T) increases the risk of Helicobacter pylori-infection gastritis in children: a case control study Supriatmo, Gontar Alamsyah Siregar, Iqbal Pahlevi Adeputra Nasution, Oke Rina Ramayani 389 Endothelin-1 level as a predictor of hepatopulmonary syndrome in liver cirrhosis Darmadi, Riska Habriel Ruslie

395 The effect of roselle flower petals extract (Hibiscus sabdariffa Linn.) on reducing inflammation in dextran sodium sulfate-induced colitis Masrul Lubis, Gontar Alamsyah Siregar, Hakim Bangun, Syafruddin Ilyas 402 Association between CD133 expression and clinicopathological profile in colorectal cancer Imelda Rey, Agung Putra, Dharma Lindarto, Fauzi Yusuf 408 In vitro regulation of IL-6 and TGF-ß by mesenchymal stem cells in systemic lupus erythematosus patients Dewi Masyithah Darlan, Delfitri Munir, Nelva Karmila Jusuf, Agung Putra, Riyadh Ikhsan, Iffan Alif

414 An ethanol extract of Senduduk fruit (Melastoma malabathricum L) inhibits the expression of vascular endothelial growth factor and tumour necrosis factor alpha in HeLa cells Deri Edianto, Aznan Lelo, Syafruddin Ilyas, Marline Nainggolan 420 Snakehead fish extract as an enhancer of vascular endothelial growth factor and nitric oxide levels in cerebral angiogenesis: an insight of stroke therapy Iskandar Nasution, Hasan Sjahrir, Syafruddin Ilyas, Muhammad Ichwan 425 Combination of vitamin A and D supplementation for ischemic stroke: effects on interleukin-1ß and clinical outcome Alfansuri Kadri, Hasan Sjahrir, Rosita Juwita Sembiring, Muhammad Ichwan

433 Association between working memory impairment and activities of daily living in post-stroke patients Fasihah Irfani Fitri, Aida Fithrie, Aldy S Rambe 439 The impact of pain intensity on quality of life of postherpetic neuralgia patients Wizar Putri Mellaratna, Nelva K. Jusuf, Ariyati Yosi 445 The level of tumour necrosis factor-alpha and its relationship to the cognitive function of Malayan- Mongoloid patients with schizophrenia Mustafa M. Amin, Abdul Rasyid, Elmeida Effendy, Nurmiati Amir, Dwi Anita Suryandari 451 Poguntano effect on cell viability in cholesteatoma keratinocyte of chronic suppurative otitis media Harry A. Asroel, Aznan Lelo, Delfitri Munir, Jenny Bashiruddin, Restu Syamsul Hadi 457 Pregnancy outcome in women who survived genital or extragenital cancer Anis Cerovac, Dženita Ljuca, Gordana Bogdanović, Gordana Grgić, Haris Zukić 465 Preterm and term birth in twin pregnancies during a seven-year period: a call for obstetricians to declare about amnionicity and chorionicity Anis Cerovac, Gordana Grgić, Dženana Softić, Dženita Ljuca, Sergije Marković, Amer Mandžić 472 A therapeutic effect of Nigella sativa extract on female Wistar rats vulvovaginal candidiasis model Muhammad Rusda, Muhammad Fidel Ganis Siregar, Aznan Lelo, Syafruddin Ilyas, Ratna Akbari Ganie, Yusuf Effendi, Poppy Anjelisa Zaitun Hasibuan, Raden Roro Lia Iswara 477 Endovascular or open surgical treatment of high-risk patients with infrainguinal peripheral arterial disease and critical limb ischemia Dragan Totić, Vesna Ðurović Sarajlić, Haris Vranić, Amel Hadžimehmedagić, Nedžad Rustempašić, Muhamed Djedović, Haris Vukas, Alen Ahmetašević 485 The effect of Aloe vera ethanol extract on the growth inhibition of Candida albicans Via Karina Nabila, Imam Budi Putra 490 Strut graft vs. traditional plating in the management of periprosthetic humeral fractures: a multicentric cohort study Giuseppe Rollo, Michele Biserni, Gazi Huri, Christian Carulli, Mario Ronga, Michele Bisaccia, David Gomez-Garrido, Nezih Ziroglu, Enrico Maria Bonura, Andrea Alberto Ruberti, Andrea Schiavone, Luigi Meccariello 498 Surgical treatment of multifragmentary segmental femur shaft fractures with ORIF and bone graft versus MIPO: a prospective control-group study Giuseppe Rollo, Mario Ronga, Enrico Maria Bonura, Rocco Erasmo, Michele Bisaccia, Paolo Pichierri, Antonio Marsilio, Andrea Pasquino, David Gomez Garrido, Raffaele Franzese, Andrea Schiavone, Luigi Meccariello 509 Tibial diaphyseal fractures in children: indications and limitations of the treatment with monolateral and hybrid external fi xator Michele Bisaccia, Giuseppe Rinonapoli, Salvatore Di Giacinto, Andrea Schiavone, Simone Lazzeri, David Gomez-Garrido, Juan Antonio Herrera-Molpeceres, Umberto Ripani, Cristina Ibáñez-Vicente, Giuseppe Rollo, Enrico Maria Bonura, Raffaele Franzese, Luigi Meccariello, Auro Caraffa 517 Ten-year risk assessment for type 2 diabetes mellitus using the Finnish Diabetes Risk Score in family medicine Suzana Savić, Saša Stanivuković, Biljana Lakić 523 Glycaemia, renal function and body mass in men and women with type 2 diabetes Marko Pirić, Dunja Šojat, Valentina Mišković, Matej Šapina, Zvonimir Popović, Tatjana Bačun 530 Lifestyle risk factors and comorbidities of cancer patients in a country with limited resources Nejra Mlačo, Armin Šljivo, Ahmed Mulać, Amina Kurtović-Kozarić, Anes Pašić, Semir Bešlija, Šejla Cerić, Timur Cerić 538 Factors associated with overweight and obesity in preschool children Amira Kurspahić-Mujčić, Amra Mujčić

Medicinski Glasnik is indexed by MEDLINE, EMBASE (Exerpta Medica), EBSCO and Scopus #SigurnaProfesija Prvo i jedino Osiguranje zdravstvenih radnika i u doba pandemije!

Zaštitite i Vi svoju karijeru od profesionalne pogreške.

Zašto ugovoriti osiguranje Osiguranje za sve zdravstvene radnike “Sigurna profesija”? „Sigurna profesija“ je proizvod koji je namijenjen ne samo zdravstvenim radnicima koji su u radnom odnosu ili obavljaju Medicinsko osoblje je oduvijek, a posebno u vrijeme pandemije, samostalnu djelatnost, nego i zdravstvenim ustanovama, koje jako izloženo nizu neugodnih situacija povezanih s profesional- žele obezbjediti zaštitu za svoje uposlenike i obezbjediti kontinui- nim rizicima. S obzirom na složenost medicinskog poziva i tet poslovanja firme. Osigurati se mogu doktori medicine, dokto- činjenicu da se u doba koronavirusa neminovno povećava svijest ri stomatologije, medicinsko osoblje, farmaceuti i pripadnici o odgovornosti, ne čudi povećano interesovanje ljekara i drugih drugih srodnih zanimanja. zdravstvenih radnika za rješenja kako zaštititi sebe i svoje poslova- nje u doba krize. Sigurna profesija u nesigurnim vremenima Proizvod je posebno koristan u neizvjesnim vremenima jer pruža osiguranje profesionalne odgovornosti i pravne zaštite u Pravi izbor slučajevima, na primjer, pokretanja postupka zbog profesionalne pogreške doktora ili zdravstvenog osoblja koji mogu rezulti- i u vrijeme pandemije, za rati velikim finansijskim gubicima. sve pripadnike medicinske profesije, je “Sigurna profesija” - program osiguravajuće zaštite Prve štete već isplaćene zdravstvenih radnika, koji na Ljekari i stomatolozi, koji su prije proglašenja pandemije sklopili bh. tržištu nudi jedino sa UNIQA osiguranjem ugovor o osiguranju koji uključuje rizik UNIQA osiguranje. finansijskog gubitka u slučaju prekida rada, već su imali priliku da se uvjere u opravdanost i benefite osiguranja. Odštete su isplaćene privatnim doktorima i vlasnicima specijalističkih ordinacija u Banja Luci, Sarajevu i Živinicama koji su pretrpili Zahvaljujući „Sigurnoj profesiji“ svi zdravstveni radnici za gubitak prihoda uslijed prekida rada zbog pandemije koronavi- relativno simboličan godišnji iznos mogu obezbjediti komplet- rusa. Zahvaljujući UNIQA osiguranju, pomenuti doktori mogu nu zaštitu od profesionalne odgovornosti, pravnu zaštitu u lakše prebroditi trenutnu situaciju i umanjiti negativne posljedi- krivičnom i prekršajnom postupku kao i osiguranje finansijsk- ce pandemije u narednom periodu. og gubitka.

Iskustva naših klijenata pogledajte na www.uniqa.ba Šta osiguranje Više informacija putem dežurnog telefona 061 723 842 ili pokriva? e-maila [email protected] - materijalne i nematerijalne štete počinjene trećim licima prilikom obavljanja ljekarske djelatnosti

- pokriće za troškove pravnog savjetovanja, advokata i vještače- nja u disciplinskom, prekršaj- nom ili krivičnom postupku REVIEW

Interdisciplinary aspects of possible negative effects of dogs on humans in Bosnia and Herzegovina

Muhamed Katica1, Zarema Obradović2, Nasreldin Hassan Ahmed3, Enra Mehmedika-Suljić4, Žana Stanić5,6, Rowida Seifeldin Abdalaziz Mohamed7, Emina Dervišević8

1Department of Pathological Physiology, Veterinary Faculty; University of Sarajevo, Bosnia and Herzegovina, 2Faculty for Health Stud- ies; University of Sarajevo, Bosnia and Herzegovina, 3Emergency Medicine Clinic, 4Neurology Clinic; Clinical Centre of the University of Sarajevo, Bosnia and Herzegovina, 5Department of Gynaecology and Obstetrics, Zabok General Hospital and Croatian Veterans Hospital, 6Institute for Integrative Gynaecology, Obstetrics and Minimally Invasive Gynaecological Surgery, Faculty of Dental Medicine and Health, University of Josip Juraj Strossmayer, Osijek; Croatia, 7Pharmacy Vita Sarajevo, 8Department of Forensic Medicine, School of Medicine, University of Sarajevo; Bosnia and Herzegovina

ABSTRACT

The coexistence of humans and dogs, in addition to all positive effects, can result in negative effects on human health. A particular risk is posed by a population of stray dogs, that is, dogs without owners and veterinary supervision. A contact with dogs in addition to bites, carries the risk of viral, bacterial and parasitic zoonoses, and can also cause psychological trauma. Children, the elderly and Corresponding author: pregnant women are the categories most susceptible to the nega- Muhamed Katica tive effects of dogs. The aim of the paper was to make an inter- Department of Pathological Physiology, disciplinary analysis of the negative effects of dogs on humans. Dog bites cause wounds and dysfunction of damaged tissue, and Veterinary Faculty, often lead to various infections. The risks of rabies and tetanus are University of Sarajevo particularly significant if proper and timely treatment is not per- Zmaja od Bosne 90, 71000 Sarajevo, formed. Ongoing training for dog owners can significantly reduce Bosnia and Herzegovina the number of bites inflicted by owned dogs, but stray dogs remain Phone: +387 33 729 155; a serious social problem and pose potential health risks of some Fax: +387 33 617 850; zoonosis. Timely and adequate management of bite wounds and the use of rabies-post-exposure prophylaxis as well as psycho-the- E-mail: [email protected] rapy, where indicated, significantly reduce possible adverse health ORCID ID: https://orcid.org/orcid. effects for patients who have been bitten by dogs. org/0000-0002-8184-0065 Key words: canids, bites, risks, epidemiology, zoonosis

Original submission: 27 April 2020; Revised submission: 04 May 2020; Accepted: 08 June 2020 doi: 10.17392/1187-20

Med Glas (Zenica) 2020; 17(2): 246-251

246 Katica et al. Negative effects of dogs on humans

INTRODUCTION In addition to the risks of infection, dog bites also cause psychological trauma (18). Dogs without A dog is an animal most commonly found with owners do not have secure alimentation, and so humans; it is used to such a lifestyle, and it is are constantly on the lookout for food. They are imperative that people take care of them and their often infected by the food they eat, as is the case health. In developed countries of the West dogs with the consumption of raw meat at slaughter- are common pets that live in flats with their te- houses and other places of slaughter of animals nants, and they are given full care and attention. (19). Because of that, Taenia echinococcus can In developing and underdeveloped countries, and be introduced into the dogs' bodies and through especially in areas with an armed conflict, dogs excretion of its eggs through the faeces, contami- are often abandoned by their caregivers. They nate the environment and be introduced into the are known by the name "stray dogs", and they human body and cause Echinococcosis (20,21). are unwanted, often unhealthy and unvaccinated Different types of microorganisms are excreted (1-3). In contact with humans, when looking for through dog faeces, which, if not properly dispo- food, they often attack in self-defence or defence sed of, contaminate the surfaces, posing a risk of of their pups. Most often children, the elderly and them being introduced into the human body and pregnant women are attacked (4-6). of a set of diseases (22,23). Dogs are reservoirs People are attacked and injured by dogs with of leishmaniasis that can be transmitted to humans owners, as well as by non-owner dogs, in relati- by phlebotomus causing leishmaniasis (24). Dogs vely equal measure (3,7,8). In certain cantons of can transmit a range of diseases (e.g. parvovirus, Bosnia and Herzegovina (B&H), between 1996 distemper, leptospirosis, rabies etc.) to other dogs, and 2005, there were as many as 6.9% more dog domestic and wild animals (15, 25-28). bites by dogs with owners than by stray dogs The aim of this paper is to point out the negative (9). Almost half of all registered dog bites occur effects of dogs on humans by analysing interdisci- within the family in which the dog resides (10). plinary aspects: public health, emergency and fo- Fewer owned dogs attack people in public areas, rensic medicine, paediatrics, psychiatry and infec- and most often because of the owners' irrespon- tiology, and to emphasize the importance of timely sible behaviour (11). However, dogs of known treatment in the prevention of particular zoonoses. owners are more or less under the veterinary control; they are most often vaccinated against rabies, EPIDEMIOLOGICAL ASPECT OF CONTAMINA- which is not the case with dogs without owners. TION OF PUBLIC AND ARABLE LAND BY CANINE Owned dogs are fed with adequate dog food, and FAECES AND URINE non-owner dogs tend to seek food among the waste, assuming that stray dogs have a higher quan- Contamination of public areas by dog faeces, uri- titative and qualitative representation of microor- ne and hair pose a risk to human health (22,23). ganisms in the mouth. Because of this, they are The basic parameter for assessing human health considered to be significantly greater carriers of risk is the amount of faeces excreted in an area zoonoses to humans and some other diseases to (8,29). The average daily amount of excreted domestic and wild animals, compared to owned faeces of an adult dog ranges from 340-1000 dogs (12). Dog bites introduce different microor- grams, while the average daily amount of excre- ganisms into the tissue of a bitten person, which ted urine is about 40 mL per kilogram of body in turn can lead to infection (3). The situation is weight (8,30). Dogs, weighing 15 kg, for exam- especially serious if there is a risk of tetanus or ple, excrete about 0.6 litres of urine a day (8,30). rabies (13). This is especially important in coun- The estimates correspond to the study of Ma et al. tries where these diseases are endemic (14,15). (2020) (31), which states that the daily excretion of faeces in dogs ranges from 21 g to 1074 g, de- In countries where there is a high population den- pending on the body weight and breed. The body sity of stray dogs in big cities, such weight range of the tested dogs was from about as in India, rabies poses a huge problem (15,16). 1.72 kg (miniature poodles) to up to 90.7 kg of The risk of bites in urban areas from aggressive body weight in large breeds (shepherd dogs) (31). dogs is high (17). According to Katica et al. (8), in four cantons of

247 Medicinski Glasnik, Volume 17, Number 2, August 2020

Bosnia and Herzegovina, a population of 21.800 Pregnant women can also become victims of dog stray dogs was recorded, meaning that they con- bites and suffer considerable emotional trauma, taminated public areas daily with about 10.900 reliving for years the memories of fear and pain kilograms of faeces and 13.080 litres of urine. caused by an aggressive dog attack (40,41). Alt- In the four cantons surveyed in Bosnia and Her- hough there is no reliable scientific evidence, it zegovina, there is no reliable information on re- is believed that a dog bite trauma can endanger gistered dogs with known owners, and approxi- pregnancy and assist in the onset of miscarriage mate estimates are that there are around 23.000 or preterm birth (4). A number of legal norms re- dogs (8). There are serious assumptions that this gulate these issues (42,43). It is known that envi- number is much higher than estimated and that a ronmental factors affect emotional, behavioural significant number of these registered dogs have and cognitive components of foetal development, irregular veterinary health surveillance, or have and that pregnancy stress can have long-lasting never had one. Therefore, 23.000 registered dogs effects on offspring neurodevelopment (44). The- daily contaminate with about 11.500 kilograms se consequences include a wide range of disor- of faeces and 13.800 litres of urine (8). ders ranging from emotional distress to cognitive impairment (45). Even the foetal fingerprint Since eggs of Taenia echinococcus may be found pattern is known to change under the influence in the excreted faeces of humans, soil contamina- of prenatal stress resulting from changes in foetal tion caused by dog excrement is directly related brain development (46). Undoubtedly, the stress to the occurrence of human echinococcosis (32). hormone cortisol plays an important role in the Human echinococcosis is usually a consequen- development of these changes (47). ce of soil contamination, contamination of low- growing vegetables and fruits eaten raw, or direct SURGICAL THERAPEUTIC TREATMENT OF A BIT- contact with dogs (20). To humans, echinococcus TEN PREGNANT WOMAN eggs are introduced through mouth, contaminated food or contaminated hands, but due to a long Surgical treatment of a bitten pregnant woman does incubation period, health effects of this zoonosis not deviate significantly from the classic treatment. can be expected only after a few years (33). With additional psychological support, special attention should be paid to the possible development PSYCHOLOGICAL TRAUMA IN BITTEN PERSONS of local and systemic infection from the bite wound Dog bites can also include phenomena such as the due to the specific immune status of the pregnant appearance of intense fear, helplessness or horror, woman (48). In addition to classic, well-known pat- which ultimately often leads to many symptoms of hogens, various exotic infections originating from post-traumatic stress disorder (PTSD) (34-36). In canine oral cavity from visceral leishmaniasis (49), paediatric care, this risk to child health is undere- trypanosomiasis (50), to toxoplasmosis (51) have stimated, since they are the most vulnerable group been demonstrated in pregnant women. in this respect (37). The reason of PTSD occurring A special therapeutic dilemma for many physici- in some children is unknown. It is evident that it ans who do not have clinical experience with this is closely related to the intensity, character and issue is the use of post-exposure antiretroviral aggressiveness of the animal attack (36). Children prophylaxis in pregnant women, in fear of possible who are victims of dog bites often require medical adverse effects of the vaccine and immunoglobulin attention, but psychological support is rarely offe- on the foetus (52). Numerous papers have been ci- red (37). In such circumstances, rapid intervention ted in the literature on the outcome of pregnancies is recommended in order to prevent PTSD. Chil- in pregnant women bitten by infected dogs, with dren who have had violent and/or multiple dog bi- and without subsequent anti-rabic prophylaxis tes often have a heterogeneous developmental dis- (53-55). Evidence-based medicine today takes order characterized by distraction, impulsiveness, the view that this concern is unwarranted and that irritability and hyperactivity, Attention Deficit any pregnant woman bitten by a stray dog, whose Hyperactivity Disorder (ADHD), and they are the vaccination status cannot be confirmed, should be patients with particular priority for engagement in treated with an anti-rabies vaccine according to the appropriate psychotherapy (38,39). prescribed protocol (56).

248 Katica et al. Negative effects of dogs on humans

FORENSIC MEDICAL ASPECTS OF DOG BITES countries these are pets, and in underdeveloped and developing countries, stray dogs. Dogs pose Dog attacks on humans can sometimes result in a potential health risk due to the transmission death (57). Over ten deaths a year in the US are of zoonoses, especially stray dogs. In preven- related to dog bites (10,58). Bite wounds (vulnus ting the onset of the risk of zoonoses originating morsum) in humans by dogs are phenomena enco- from dogs, it is important to adequately remove untered by forensic experts as well. The forensic the faeces and urine of dogs to prevent contami- importance of bite wounds is that they can easily nation of public areas. In case of bites, timely be mistaken for some mechanical injuries, espe- and adequate management of vulnus morsum by cially puncture and contusion wounds (59). Dog physicians of appropriate specialties, the imple- bites can cause a variety of injuries, ranging from mentation of rabies-post-exposure prophylaxis, individual injuries to body dismemberment and and, if necessary, the inclusion of psycho-therapy death, which is extremely rare. Death occurs more are important. This can significantly reduce the often from complications and infections that deve- adverse health risks for patients who have suffe- lop after the bite than from a direct bite (60). red dog bites. Neck bites represent a particular risk, that is, a possible interruption of the continuity of the ca- ACKNOWLEDGMENT rotid artery, which will cause an inevitable death The authors thank dr. Nirves Bulaja, Director (61). About 78% of dog bites in children occur on of the Cantonal Veterinary Stations of Sarajevo the head and neck, which is particularly risky due Canton, for technical support and collecting in- to the localization of the bite (62,63). formation for registered dogs. The authors are Forensic experts often carry out expert asse- grateful to MA iur. Lejla Suljević for the help ssments of injuries inflicted on people from bite provided to perform this work. wounds. The occurrence of the injury and the circumstances of the attacks and bites of dogs are FUNDING proved in court, using medical records, and the No specific funding was received for this study. examination of witnesses, eyewitnesses and possibly a police record, if it exists. TRANSPARENCY DECLARATION In conclusion, dogs and humans live in direct Competing interests: None to declare. contact. Differently developed countries are do- minated by different types of dogs; in developed

REFERENCES 1. Ortega-Pacheco A, Rodriguez-Buenfil JC, Bolio- 6. Schalamon J, Ainoedhofer H, Singer G, Petnehazy Gonzalez ME, Sauri-Arceo CH, Jiménez-Coello M, T, Mayr J, Kiss K, Höllwarth ME. Analysis of dog Forsberg CL. A Survey of dog populations in urban bites in children who are younger than 17 years. Pe- and rural areas of Yucatan, Mexico. Anthrozoös diatrics 2006; 117:e374-79. 2007; 20:261-74. 7. Vučinić M, Đorđević M, Radenković-Damnjanović 2. Cortez-Aguirre GR, Jiménez-Coello M, Gutiérrez- B, Janković LJ, Mirilović M. Bites to humans cau- Blanco E, Ortega-Pacheco A. Stray dog population sed by stray and owned dogs in Belgrade. Acta Vete- in a city of Southern Mexico and Its Impact on the rinaria (Belgrade) 2008; 58:563-71. contamination of public areas. Vet Med Int 2018; 8. Katica M, Obradović Z, Gradaščević N, Hadžimusić 2018:2381583. N, Mujkanović R, Meštrić E, Bašić F, Toplaović B, 3. Katica M, Obradović Z, Ahmed NH, Dervišević E, Smajkić R, Dupovac M, Čelebičić M. Assessment of Delibegović S. Dog bites in the Federation of Bosnia the effect of stray dogs as a risk factor for the health and Herzegovina: a review. Cyprus J Med Sci 2019; of population in certain areas of Bosnia and Herze- 4:136-40. govina. EJBPS 2017; 4:107-11. 4. Speroff L, Fritz MA. Clinical Gynecologic Endocri- 9. Katica M, Mujkanović R, Imamović Dž, Šaljić E, nology and Infertility. Philadelphia: Lippincott- Kadić M, Kartal S. Problems with stray dogs in Ze- Williams and Wilkins, 2005. nica-Doboj Canton and Sarajevo Canton. Proceedin- 5. Cavalcanti AL, Porto E, Santos BF, Cavalcanti CL, gs of the 4th Symposium of Agriculture, Veterinary, Cavalcanti AFC. Facial dog bite injuries inchildren: Forestry and Biotechnology with International Parti- a case report. Int J Surgery Case Rep 2017; 41:57– cipation, September, 21 – 23, 2006, Zenica, Bosnia 60. and Herzegovina.

249 Medicinski Glasnik, Volume 17, Number 2, August 2020

10. Presutti RJ. Prevention and treatment of dog bites. 26. Vojinović D, Bogićević N, Vasić A, Manić M, Ele- Am Fam Physician 2001; 63:1567-72. zović- Radovanović M , Rogožarski D, Marić J , 11. Katica, M, Gradaščević N, Hadžimusić N, Obrado- Valčić M. Seroepidemiological survey of leptospiral vić Z, Mujkanović R, Meštrić E, Čoloman S, Du- infection in stray dogs in Serbia. Turk J Vet Anim Sci povac M. Widespread of stray dogs: methods for 2015; 39:719-23. solving the problem in certain regions of Bosnia and 27. Özkan B. Ultrasonographic findings in two dogs Herzegovina. IJRG 2017a; 5:414-22. with canine parvoviral enteritis. JIVS 2017; 1:11-5. 12. Dilegge SK, Edgcomb VP, Leadbetter ER. Presence 28. Rendon-Marin S, da Fontoura Budaszewski R, Ca- of the oral bacterium capnocytophaga canimorsus nal CW, Ruiz-Saenz J. Tropism and molecular pat- in the tooth plaque of canines. Vet Microbiol 2011; hogenesis of canine distemper virus. Virol J 2019; 149: 437-45. 16:30. 13. Hurt JB, Maday KR. Management and treatment of 29. Nijsse R, Mughini-Gras L, Wagenaar JA, Franssen animal bites. JAAPA 2018; 31: 27-31. F, Ploeger HW. Environmental contamination with 14. Gamble L, Gibson AD, Shervell K, Lohr F, Otter I, Toxocara eggs: a quantitative approach to estimate Mellanby RJ. The problem of stray dogs. Rev Sci the relative contributions of dogs, cats and foxes, Tech 2018; 37:543-50. and to assess the efficacy of advised interventions in 15. Uzunović S, Skomorac M, Bašić F, Mijač-Musić I. dogs. Parasites Vectors 2015; 8:397. Epidemiological features of human cases after bites/ 30. Kadić M. Practicum of Pathological Physiology of scratches from rabies-suspected animals in Zenica- Veterinary Medicine. Sarajevo: University of Sara- Doboj Canton, Bosnia and Herzegovina. J Prev Med jevo, 1998. Public Health 2019; 52:170-8. 31. Ma J, Zhu H, Kannan K. Fecal Excretion of Perfluo- 16. Menzes R. Rabies in India. CMAJ 2008; 178:564-6. roalkyl and Polyfluoroalkyl Substances in Pets from 17. Amiri S, Maleki Z, Nikbakht HA, Hassanipour S, New York State, United States. Environ Sci Technol Salehiniya H, Ghayour AR, Kazemi H, Ghaem H. Lett 2020; 7: 135–42. Epidemiological patterns of animal bites in the Naja- 32. Umhang G, Raton V, Comte S, Hormaz V, Boucher fabad, center of Iran (2012-2017). Ann Glob Health JM, Combes B, Boué F. Echinococcus multilocularis 2020; 86:38. in dogs from two French endemic areas: no evidence 18. Ferreira S, Ayres Quaresma LE, Timóteo CA, da of infection but hazardous deworming practices. Vet Silva Fabris AL, Faverani LP, Francisconi GB, So- Parasitol 2012; 188:301-5. uza FA, Júnior IR. The primary closure approach of 33. Deplazes P. Ecology and epidemiology of Echino- dog bite injuries of the nose. Craniofac Surg 2014; coccus multilocularis in Europe. Parassitologia 25:e216-8. 2006; 48:37-9. 19. Van Bree FPJ, Bokken GCAM, Mineur R, Franssen 34. Anyfantakis D, Botzakis E, Mplevrakis E, Symvo- F, Opsteegh M, van der Giessen JWB, Lipman LJA, ulakis EK, Arbiros I. Selective mutism due to a dog Overgaauw PAM. Zoonotic bacteria and parasites bite trauma in a 4-year-old girl: a case report. J Med found in raw meat-based diets for cats and dogs. Vet Case Rep 2009; 3:100. Rec 2018;182:50. 35. Ziegler MF, Greenwald MH, De Guzman MA, Si- 20. Obradović Z. Ehinokokoza (Echinococcosis) [In mon HK. Post-traumatic stress responses in chil- Bosnian]. Sarajevo: Faculty of Health Studies, Uni- dren: awareness and practice among a sample of versity of Sarajevo, 2009. pediatric emergency care providers. Pediatrics 2005; 21. Obradović Z. Public health aspect of animal waste 115:1261-7. (animal corpses and slaughter house confiscates). 36. American Psychiatric Association: Diagnostic and Scientific-professional Conference with Internatio- Statistical Manual of Mental Disorders, DSM-IV. nal Participation. Management of Animal Waste and 4thdn. Washington: APA, 1994. Sustainable Development, Special Editions CXLVII, 37. Peters V, Sottiaux M, Appelboom J, Kahn A. Posttra- Proceedings, Sarajevo, Bosnia and Herzegovina umatic stress disorder after dog bites inchildren. J 2012; 21:75-82. Pediatrics 2004; 144:121-2. 22. Medina-Pinto RA, Rodríguez-Vivas RI, Bolio- 38. Şahin B, Bozkurt A, Karabekiroğlu K. Sleep pro- González ME. Zoonotic intestinal nematodes in blems in children with attention deficit hyperactivity dogs from public parks in Yucatán, México. Biome- disorder. Duzce Med J 2018; 20:81-6. dica 2018; 38:105-10. 39. Yehuda R. Current status of cortisol findings in post- 23. Runesvärd E, Wikström C, Fernström LL, Hansson traumatic stress disorder. Psychiatr Clin North Am I. Presence of pathogenic bacteria in faeces from 2002; 25:341-68. dogs fed raw meat-based diets or dry kibble Vet Re- 40. Davidson JRT. Surviving disaster: what comes after cord 2020; 1-6. the trauma? Br J Psychiatr 2002; 181:366-8. 24. Apostolopoulos N, Mitropoulou A, Thom N, Moritz 41. Ahmed AS. Post-traumatic stress disorder, resilien- A. Update on therapy and prevention of canine le- ce and vulnerability. Advances in Psychiatric Tre- ishmaniasis. Aktuelle Kenntnisse zu Therapie und atment 2007; 13:369-75. Prävention der kaninen Leishmaniose. Tierarztl Prax 42. Aaron L. Dog Bite Injury Law. Expert Law. https:// Ausg K Kleintiere Heimtiere 2018; 46: 315-22. www.expertlaw.com/library/personal_injury/dog_ 25. FAO. Dog population management. Report of the bite_injuries.html (27 May 2020) FAO/WSPA/IZSAM expert meeting - Banna. Ani- mal Production and Health Report. No. 6. Rome, Italy, 2014.

250 Katica et al. Negative effects of dogs on humans

43. Krivični zakon Federacije Bosne i Hercegovine. 55. Nguyen HT, Tran CH, Dang AD, Tran HG, Vu Službene novine FBiH br. 36/2003,21/2004 (Crimi- TD, Pham TN, Nguyen HV, Nguyen AN, Pieracci nal Code of the Federation of Bosnia and Herzego- EG, Tran DN. Rabies vaccine hesitancy and deaths vina) Official Gazette of FBiH no. 36 / 2003, 21 / among pregnant and breastfeeding women — Viet- 2004.[In Bosnian]. https://advokat- prnjavorac.com/ nam, 2015–2016. MMWR Morb Mortal Wkly Rep zakoni/Krivicni_zakon_FBiH.pdf (04 June 2020) 2018; 67:250–2. 44. Weinstock M. The potential influence of maternal 56. Manning SE, Rupprecht CE, Fishbein D, Hanlon CA, stress hormones on development and mental health of Lumlertdacha B, Guerra M, Meltzer MI, Dhankhar the offspring. Brain Behav Immun 2005; 19:296-308. P, Vaidya SA, Jenkins SR, Sun B, Hull HF. Human 45. Seng JS, Graham-Bermann SA, Clark MK, McCar- rabies prevention- United States, 2008: recommen- thy AM, Ronis DL. Posttraumatic stress disorder and dations of the Advisory Committee on Immunization physical comorbidity among female children and Practices MMWR Recomm Rep 2008; 57:1-28 adolescents: results form service-use data. Pediatrics 57. Ciampolini R, Cecchi F, Spinetti , Rocchi I, Bisca- 2005; 116:767-76. rini F. The use of genetic markers to estimate relati- 46. King S, Mancini-Marïe A, Brunet A, Walker E, Mea- onships between dogs in the course of criminal inve- ney MJ, Laplante DP. Prenatal maternal stress from a stigations. BMC Res Notes 2017; 10:414. natural disaster predicts dermatoglyphic asymmetry 58. Sacks JJ, Lockwood R, Hornreich J, Sattin RW. Fatal in humans. Dev Psychopathol 2009; 21:343-53. dog attacks, 1989-1994. Pediatrics 1996; 97:891-5. 47. O'Donnell K, O'Connor TG, Glover V. Prenatal stre- 59. Zečević D, Brkić H, Busljeta M, Crnić I, Čadež Z, ss and neurodevelopment of the child:focus on the Ferenčić S, Gusić S, Kovačić Z, Kregar A, Križa- HPA axis and role of the placenta. Int J Dev Neurosci nac Š, Kubat M, Mayer D, Mikuličić V, Palimović 2009; 31:285-92. V, Petrovečki V, Završki P, Strinović D, Škavić J, 48. Raj RS, Bonney EA, Phillippe M. Influenza, immune Šlaus M. Sudska medicina i deontologija (Forensic system, and pregnancy. Reprod Sci 2014; 21:1434-51. medicine and deontology) [In Croatian] Zagreb: Me- 49. Figueiró-Filho EA, Duarte G, El-Beitune P, Quinta- dicinska naklada, 2018. na SM, Maia TL. Visceral leishmaniasis (kala-azar) 60. Simeunović S. Urgentna stanja u medicini. (Emer- and pregnancy. Infect Dis Obstet Gynecol 2004; gency conditions in medicine) [In Serbian] Beograd: 12:31-40. Naša knjiga, 2008. 50. De Raadt P. Congenital trypanosomiasis and leish- 61. Lacey L, Dabbas N, Saker R, Blakeley C. Dissection maniasis. Arch Fr Pediatr 1985; 42:925-7. of the carotid artery as a cause of fatal airway ob- 51. Stanić Ž, Fureš R. Toxoplasmosis: a global zoonosis. struction. Emerg Med J 2007; 24:367–8. Veterinaria 2020; 69:32-42. 62. Sacks JJ, Sattin RW, Bonzo SE. Dog bite-related 52. Windrim C, McGeer AJ, Murphy KE. Exposure to fatalities from 1979 through 1988. J American Med rabies during pregnancy CMAJ 2018; 190: E1281. Assoc 1989; 262:1489-92. 53. Fayaz A, Simani S, Fallahian V, Eslamifar A, Hazrati 63. Chand-Meena M, Kumar-Naik SH, Mittal S, Band M, Farahtaj F, Howaizi N, Biglari P. Rabies antibody R. Fatal dog bite injury – a case report. Int J Med levels in pregnant women and their newborns after Toxicology Forensic Med 2015; 5:164-7. rabies post-exposure prophylaxis. Iran J Reprod Med 2012; 10:161-3. 54. Qu ZY, Li GW, Chen QG, Jiang P, Liu C, Lam A. Survival of a newborn from pregnant women with a abies infection. J Venom Anim Toxins incl Trop Dis 2016; 22:14-6.

251 REVIEW

Securing the base of the appendix during laparoscopic appendectomy

Samir Delibegović1,2, Reuf Karabeg3, Milan Simatović4

1Clinic for Surgery, University Clinical Center Tuzla, Trnovac bb, Tuzla; 2Faculty of Medicine, University of Tuzla, Tuzla, Bosnia and Herzegovina; 3Surgical polyclinic ,,Karabeg'' Sarajevo, Bosnia and Herzegovina; 4Clinic for Surgery, Clinical Center of Republika Srpska, Banja Luka, Bosnia and Herzegovina

ABSTRACT

In contrast to classical appendectomy where the appendiceal stump is secured by a single or double ligature, in laparoscopic appendectomy various ways of securing the stump are mentioned. Each of these methods has advantages and disadvantages. Since different possibilities exist for closing the stump, it is very important to find the optimum method for closure of the appendiceal stump, bearing in mind their simplicity, biocompatibility and price. The aim of this review article has been to present the problem of securing the Corresponding author: base of the appendix during laparoscopic appendectomy. Samir Delibegović University Clinical Centre Tuzla, Key words: appendix, laparoscopic appendectomy, linear stapler, Clinic for Surgery titanium Trnovac bb, 75000 Tuzla, Bosnia and Herzegovina Phone +387 35 303 500; Fax: +387 35 250 474; E-mail: [email protected] ORCID ID: https://orcid.org/0000-0003- 0525-3288

Original submission: 08 January 2019; Revised submission: 26 February 2020; Accepted: 16 March 2020 doi: 10.17392/1131-20

Med Glas (Zenica) 2020; 17(2): 252-255

252 Delibegović et al. Securing the appendix base

INTRODUCTION is all that is needed to resect the appendix and secure the stump, which reduces the duration of the In contrast to classical appendectomy where the surgery significantly. Staplers are simple to use appendiceal stump is secured by a single or do- and are easy for any surgeon to master. uble ligature (1), in laparoscopic appendectomy various ways of securing the stump are described There is no difference between the Hem-o-lok (2). Each of these methods has advantages and and titanium DS clip in relation to the time nee- disadvantages. ded or the length of the operation (6). However, one disadvantage of DS clips is that it is impo- Closure of the appendiceal stump is the most cri- ssible to remove and replace them if they are not tical part of laparoscopic appendectomy. placed correctly on the appendiceal stump (6). The appendiceal stump may be closed using an A strength of Hem-o-lok plastic clips is that it is endoloop, which is the standard method (2), or possible to remove them, even though it is nece- a stapler which is a more expensive method, but ssary to insert another instrument to remove clips clinical evidence favours this method (3). The that are in the wrong position for some reason. use of plastic (4) or titanium clips (5) has been Another advantage of Hem-o-lok is the noise mentioned recently for closure of the appendiceal made by the clips as they are used, whereas DS stump. clips do not produce any sound. The opening of Since different possibilities exist for closing the Hem-o-lok clips is also wider so they are easier stump (Figure 1), it is very important to find the to place on a large, inflamed appendix (6). optimum method for closure of the appendiceal In all situations the size of the appendix signifi- stump, bearing in mind their simplicity, biocom- cantly affects the simplicity of use. patibility and price. Size of the appendix It is often forgotten that the size of the appendix can affect the choice of method used to close the appendiceal stump. When the appendix is enlarged, it is sometimes very difficult to place a standard endoloop ligature. In that case, a stapler is the only option (6). The size of the stump is particularly important. The average size of a phlegmonous appendiceal stump was 10.29 ± 3.13 mm, the size of a gangre- Figure 1. A) The base of the appendix secured by a single en- nous stump 12.41 ± 3.56 mm, and of a perforated doloop; B) the base of the appendix secured by an endosta- stump 12.42 ± 3.64 mm (7). As a result, in most pler; C) the base of the appendix secured by a single Hem-o- cases plastic clips, Hem-o-lok XL and titanium lok, XL size; D) the base of the appendix secured by a DS clip (Delibegović S, 2011) DS clips were able to encompass the appendix (6). Sometimes, however, the stump may be as large HANDLING/SIMPLICITY OF APPLYING as 22.75 mm. In this case, only an endoloop or a stapler are able to close the appendiceal stump. Endoloop is the standard method, but a period of training is needed to be able to apply an endolo- The size of the widest part of the appendix may op correctly. In the limited space of the ileocecal be up to 45.10 mm (recorded in acute perforated fossa, in a slim person with a small abdomen, appendicitis), and in this situation it is not easy to placing an endoloop, especially when the appen- place an endoloop. The tip of the appendix was dix is bulging, is not easy. up to 28.68 mm, which should not affect the closure of the appendiceal stump (6). It must be emphasized that the time needed for application and the duration of the operation Thus, the safety of the method used to close the were much shorter using a stapler than the other appendiceal stump depends on the size and shape method (6). This is because the instruments only of the endoloop, stapler and the length of the need to be inserted once, and a single movement arms of the clip.

253 Medicinski Glasnik, Volume 17, Number 2, August 2020

Size of the endoloop, length of stapler and clip Price legs Loop ligatures have advantages for reducing me- The endoloop is elliptical in shape, and a dical costs because they are 6–12 times cheaper polydyoxanone (PDS) endoloop has the largest than stapling devices (11). Using a stapler is a width of 24.0 mm. It can be used to close all ap- safe method, and staplers rarely malfunction. pendiceal stumps, but it is difficult to place it in a Their only weakness is their high price in rela- case of acute perforated appendix, since the stump tion to other ways of securing the appendiceal may be as large as 45.0 mm. An endoloop is 62.44 stump. Bearing in mind the fact that in U.S. hos- mm in size, but due to the lack of space in the cae- pitals operating room charges average $62/min cal fossa, in the case of an extremely enlarged ap- (range $22–$133/min) (12), the use of staplers pendix, this size is only theoretical (7). could reduce these costs in terms of time needed, Although the opening of a stapler is large, the lar- but since they are a more expensive method, this gest being an Echelon 60, and it is easier to place, advantage is lost. However, Hem-o-lok and DS the cost is much higher and it should only be used clips are cheaper than loops and staplers, and also for forms of acute appendicitis where other met- have other advantages. hods are impossible (7). DISCUSSION The internal length of the legs of DS clips are longer than Hem-o-lok clips, however, the ope- The method used for closure is important in terms ning of a Hem-o-lock clip is larger (9.4 mm) so it of the simplicity of application, biocompatibility, is much easier to place than a DS clip (Table 1). If price and other characteristics of the instruments we consider the maximum diameter of the appen- themselves. dix (23.13) recorded in the perforated form, the It is sometimes difficult to insert a loop ligatu- internal diameter of the clip should be greater (6). re in the area bordered by the iliac fossa, and a certain level of training is needed (6), but Hem- Table 1. Characteristics of each method of closure of the o-lok and DS clips may be applied by a surgeon appendiceal base with minimal previous training. Using a stapler Vycril PDS en- Stapler Hem-o-lok Characteristics DS clip is simple and safe, and its only weakness is its endoloop doloop echelon 45 XL clip External length high price in comparison with other methods of 97.7* 97.7* 17.33 16.2 of legs (mm) securing the appendiceal stump. Staplers are said Internal length 62.24† 62.24† 42 13.58 14.2 to be safer than laparoscopic loop ligatures regar- of legs (mm) Opening (mm) 24.0‡ 29.5‡ 12.35 9.4 5.5. ding the formation of intraabdominal abscesses Weight (mg) 21.00 14.00 72.00 121.00 176.00 and readmission (11). The biocompatibility of Price (Euro) 38.10 36.99 167.69 7.5 20 staplers is best in comparison to other forms of * † ‡ PDS, polydioxanone; DS, titanium clip; volume; diameter; the closure of the base of the appendix (10). greatest width of the endoloop when open and in the form of an extended ellipse The malfunction rate of staplers is very low, e. g. 0.003%, although some believe that the inci- Biocompatibility dence is higher (13). There is no case mentioned The biocompatibility of the different materials in the literature of either Hem-o-lok or DS clips used in securing the base of the appendix varies. slipping or falling out of a closed stump. The PDS causes a milder inflammatory reaction In rare cases, when the appendiceal base was ne- in comparison to Vicryl (polyglactin 910), so an crotic or gangrenous, a postoperative leak may endoloop made from a PDS ligature should be occur following laparoscopic appendectomy preferred (8,9). However, titanium stapler clips (14). However, in most studies comparing endo- cause the mildest inflammatory reaction and cre- loop and staplers, or endoloop and clips (15,16), ate the least adhesions, when we compare sta- no stump leaks occurred in any group. It is in- plers with hem-o-lok and endoloops (10). After teresting that in a study comparing laparoscopic staplers, the mildest reaction is caused by plastic versus open surgery for complicated appendici- clips (10). Therefore, in terms of biocompatibi- tis, leakage was 9.5 % in the laparoscopic group, lity staplers and titanium clips are most prefera- and 12.8 % in the open appendectomy group. In ble, followed by plastic clips.

254 Delibegović et al. Securing the appendix base

the laparoscopic group, securing the appendiceal ch is needed to improve the mechanical features stump was performed using a stapler (17). of the available materials, in order to attain the In view of the price, the size of the opening, radi- optimum method for securing the appendiceal ological advantage (18) and biocompatibility (10), stump. Bearing all this in mind, it may be said Hem-o-lok clips are the most effective, although that all these forms of closure of the appendix their internal diameter should be increased. The base are acceptable, but plastic and titanium clips DS clips are also effective, but the size of the ope- have the best potential for further development, ning sometimes makes application difficult, and and will probably become the method of choice any increase in the size of the opening would make in securing the base of the appendix. this clip ideal, also due to the excellent biocompa- FUNDING tibility of the titanium clips (10). Staplers have the best characteristics, but their price means they are No specific funding was received for this study. only an option for forms where it is not possible to close the stump using other methods. TRANSPARENCY DECLARATION It may therefore be concluded that none of these Competing interests: none to declare. methods has reached the ideal and further resear-

REFERENCES 1. Brunicardi CF, Andersen DK, Billiar TR, Dunn 9. Delibegović S, Katica M, Ferid Latić, Jakić-Razu- DL, Kao, Hunter JG, Matthews JB, Pollock RE. mović J, Koluh A, Njoum M. Biocompatibility and Schwartz’s Principles of Surgery. 11ed. McGraw- adhesion formation of different endoloop ligature Hill Education, 2019. in securing the base of the appendix. JSLS 2013; 2. Neugebauer EAM, Troidl H, Kum CK, Eypasch 17:543-8. E, Miserez M, Paul A, The EAES clinical practice 10. Delibegovic S, Iljazovic E, Katica M, Koluh A. Ti- guidelines on laparoscopic holecystectomy, appen- ssue reaction to absorbable endoloop, nonabsorbable dectomy, and hernia repair. In: Neugebauer EAM, titanium staples, and poly-mer Hem-o-lok clip after Sauerland S, Fingerhut A, Millat B, Buess G, eds. laparoscopic appendectomy. JSLS 2011; 15:70-76. EAES Guidelines for Endoscopic Surgery. Berlin: 11. Beldi G, Vorburger S, Bruegger L, Kocher T, Inder- Springer; 2006: 265-89. bitzin D, Candinas D. Analysis of stapling vs endo- 3. Wagner M, Aronsky D, Tschudi J, Metzger A, Klai- loops in appendiceal stump closure. Br J Surg 2006; ber C. Laparoscopic stapler appendectomy. A pros- 93:1390-3. pective study of 267 consecutive cases. Surg Endosc 12. Shippert RD. A study of time-dependent operating 1996; 10:895-99. room fees and how to save $100 000 by using time- 4. Delibegovic S, Matovic E. Hem-o-lok plastic clips in saving products. AJCS 2005; 22:25-34. securing of the base of the appendix during laparos- 13. Kwaznenski D, Six C, Stahfeld K. The unacknowled- copic appendectomy. Surg Endosc 2009; 23:2851-4. ged incidence of laparoscopic stapler malfunction. 5. Rickert A, Krüger CM, Runkel N, Kuthe A, Könin- Surg Endosc 2013; 27:86-89. ger J, Jansen-Winkeln B, Gutt CN, Marcus DR, 14. Sharma B, Teodor P, Grantcharov TP. Laparoscopic Hoey B, Wente MN, Kienle P. The TICAP-Study Appendectomy (03. July 2005) www.eaes-eur.org/ (titanium clips for appendicular stump closure): a getmedia/…344d…/Grantcharov_Appendectomy.pdf prospective multicentre observational study on ap- 15. Safavi A, Langer M, Skarsgard ED. Endoloop ver- pendicular stump closure with an innovative titani- sus endostapler closure of the appendiceal stump in um clip. BMC Surg 2015; 17;15:85. pediatric laparoscopic appendectomy. Can J Surg 6. Delibegović S, Mehmedovic Z. The influence of the 2012; 55:37-40. different forms of appendix base closure on patient 16. Colak E, Kement M, Ozlem N, Mutlu T, Yildirim K, outcome in laparoscopic appendectomy: randomi- Gurer A, Aktimur RA. Comparison of nonabsorba- zed trial. Surg Endosc 2018; 32:2295-9. ble polymeric clips and endoloop ligatures for the 7. Delibegovic S, Mehmedovic Z. The Influence of the closure of the appendicular stump in laparoscopic appendiceal base diameter on appendix stump clo- appendectomy: a prospective, randomized study. sure in laparoscopic appendectomy. World J Surg Surg Laparosc Endosc Percutaneous Tech 2013; 2016; 40:2342-7. 23:255-258. 8. Kosan M, Gonulalan U, Ozturk B, Kulacoglu S, 17. Taguchi Y, Komatsu S, Sakamoto E, Norimizu S, Erguder I, Akdemir O, Cetinkaya M. Tissue reacti- Shingu Y, Hasegawa H. Laparoscopic versus open ons of suture materials (polyglactine 910, chromed surgery for complicated appendicitis in adults: a ran- catgut and polydioxa-none) on rat bladder wall and domized. Surg Endosc 2016; 30:1705-12. their role in bladder stone formation. Urol Res 2008; 18. Delibegovic S. Radiologic advantages of potential 36:43-9. use of polymer plastic clips in neurosurgery. World Neurosurg 2014; 3/4:549-51.

255 REVIEW

Epidemiology of injuries connected with dance: a critical review on epidemiology Giuseppe Rinonapoli1, Marta Graziani2, Paolo Ceccarini1, Cristina Razzano3, Francesco Manfreda1, Auro Caraffa1

1Orthopaedic and Traumatology Department, University of Perugia, Perugia, 2Intensive Rehabilitation Centre, USLUmbria 2, Cascia (Perugia), 3Health Direction Unit, Physiatry and Rehabilitation AIAS Centre, Lauria; Italy

ABSTRACT

The aim of this review was to identify all types of injuries connected to the gestures of dancers and understand the associated biomechanical patterns. This is the first step in the definition of a prevention program that lacks in this kind of athletic activity. A search of Medline/PubMed, EMBASE, and the Cochrane database from 1990 to 2019 using the search terms ‘‘dance and injuries’’ and ‘‘dance and injuries and epidemiology’’ initially resulted in 601 citations. A total of 16 articles were eligible for a review. All health problems that lead to stop the activity of a dancer are classified as "dancer's injuries". They were divided in acute and overuse Corresponding author: injuries, the first being traumatic and the latter ones microtrauma- Giuseppe Rinonapoli tic. The anatomical region most affected by injuries in dance was Orthopaedic and Traumatology clearly the ankle and foot. It can be inferred that professional and Department, University of Perugia pre-professional dancers had a higher prevalence of back injuries in comparison to amateur dancers, while amateurs suffered more Via del Lino 22, 06125 Perugia, Italy frequently from hip/groin/thigh injuries. Doctors, teachers, sport Phone: +39 347 8670873; trainers and dancers themselves, all those who contribute to the E-mail: [email protected] dancer's performance, should know the most prevalent dancers’ ORCID: https://orcid.org/0000-0002- injuries. Moreover, they should know the prevention procedures, 4244-166X in order to minimize the risk of injury and recurrences. Key words: dancer, joint, back, ankle, knee

Original submission: 12 May 2020; Revised submission: 28 May 2020; Accepted: 12 June 2020 doi: 10.17392/1201-20

Med Glas (Zenica) 2020; 17(2): 256-264

256 Rinonapoli et al. Dance and injuries

INTRODUCTION A search of Medline/PubMed, EMBASE, and the Cochrane database from 1990 to the end of Dance requires a deep concentration and invol- 2019 was done, using the search terms ‘‘dance ves the correct understanding of the technique and injuries’’ and ‘‘dance and injuries and epi- in order to take maximum advantage of the own demiology’’. Among the 601 initially obtained body and minimize injury risks. Well identified published papers, we selected those with larger risk factors connected to the gestures of the dan- case collections (minimum 50 dancers), in which cers are addressed in literature: anatomical postu- the epidemiology of the injuries connected with re, poor training, technical mistakes, execution dance was reported. The studies in which the type speed, psycho-physical condition of the dancer of injury was not classified, and the affected joint at the moment of the injury, and environmental was not precisely specified, were discarded. features, such as ground, type of shoe, lightening, temperature (1-4). All of them could expose the We have classified more common injuries in dan- dancers to several injuries that should be distin- ce, coupling them with more common generating guished on the basis of age, or when the subject biomechanical event: ankle sprain, knee sprain, started to practice, and other important factors, stress fractures, traumatic mechanism and the like the technique used and the psycho-physical part of the body were combined, so that the most constitution of the dancer (1-3, 5-10). reported traumatic injuries were ankle sprains, knee sprains and muscular injuries, while the Being an activity with high performance, dance most reported overuse injuries were stress fractu- presents a risk for many types of injuries sustai- res, shoulder impingement syndrome, tendino- ned by different pathogenesis (1-3, 5-10). pathies of the inferior limb. DANCER'S INJURIES DEFINITION A total of 16 articles out of 601 were eligible for the analysis (5-7, 9,14, 16-26) All health problems that lead to stop the activity of a dancer are classified as "dancer's injuries" OVERALL PREVALENCE OF DANCE INJURIES (5,6, 9-11). They should be divided in acute and overuse injuries, the first being traumatic and the Independently from the age and type of dance, latter microtraumatic (1,10,12). Acute injuries are back injuries had a prevalence from 7% (12) to usually associated with a single episode during 62% (14). Pre-professional and professional dan- which the tissue is submitted to a sudden tension cers’ back pain was found in 17.6%, in amateur (1,10,12). This overtakes the capacity of the tissue dancers 9.8%, on average. to maintain the physiological and functional inte- Injuries of upper limb had a low prevalence (27- grity, with a consequent macro-injury (1,10,12). 38). A typical anatomical region injured in dancers One of the potential reasons for the acute injuries was hip/groin/thigh, ranging from 4% (5) to 26.8% in dance is the incorrect execution of a movement, (6) with an average of 13.1% (8.1% in professio- which can occur subsequently due to other fac- nal and professional dancers, 17.6% in amateurs). tors, such as tiredness, muscular fatigue, or loss Knee injuries range from 6% (7) to 40.4% (8), of balance (13). Overuse injuries are caused by an with an average of 16.2% (12.5% in professional excessive use (1,4,6,8, 13-15). They are caused by and professional dancers, 17.4% in amateurs). repeated microtrauma in bones, joints or soft ti- In regard to leg injuries, the data showed a percen- ssues; the anatomical structures and their capacity tage of 17.8% (from 6.1% to 24%). A very high in- of response can be insufficient when the traumas cidence of injuries involved the ankle and the foot, are repeated and predisposing factors are not eli- from 14% (9) to 54% (5). Comparing professional minated. The causal variables can be considered as and pre-professional dancers versus amateur dan- physiological, technical, or due to environmental cers, the first had a prevalence of 36% of ankle and factors (1-3, 5-10). foot injuries, the latter ones 33.5%. The aim of this review was to identify all types of In three of 16 selected papers (5,9,16), the stu- injuries and to understand associated biomecha- dy population was made of professional dancers, nical patterns as the first step in the definition of in one paper (22) data are exclusively relative to a prevention program that lacks in this kind of pre-professional dancers. Five studies (6, 17, 18, athletic activity.

257 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 1. The table reports the data of the main articles on dance injuries. The data reported for each paper include the study population and the percentage of location of the injuries in the specific body regions. Where the data were missing, it is due to the fact that the authors did not report the specific data. The injuries of ankle and foot were separately reported by some authors, together by other authors. The last column indicates the percentage of ankle and foot together. In brackets the sum of the percentages of foot and ankle injuries is presented, when separately mentioned in the two previous columns Year of Prevalence (%) of the injury out of total injuries according to location Authors Dancers population publication Back Shoulder/arm Hip/Groin/Thigh Knee Leg Ankle Foot Ankle and foot Garrick and Requa 1993 Professional 23 0.6 6.8 6.8 6.1 13.3 23.9 37.2 Nilsson et al. 2001 Professional 17.9 4 11 13 54 Coplan et al. 2002 Amateur 13.6 3 4.5 36 22.7 13.6 4.5 18.1 Byhring and Bo 2002 Professional dancers 8 4 14 16 24 3 11 14 Negus et al. 2005 Amateur 9.8 26.8 7.3 19.5 25.6 11 36.6 Gamboa et al. 2008 Amateur 9 22 16 Liederbach et al. 2008 All 9 6 57 Campoy et al. 2011 All 22.1 24 28.7 Leanderson et al. 2011 Amateur 11 21 Drezewska et al. 2013 All 62 Steinberg et al. 2013 8-16 year old females 19.2 40.4 Ekegren 2014 Pre-professional ballet 8.9 10.3 17.1 24.3 14.8 39.1 Steinberg et al. 2014 10-18 year old 19.3 Bowerman et al. 2014 Amateur 7 24 7 16 15 31 46 Trentacosta et al. 2017 All 19.8 Sobrino et al. 2017 All 20.5 2.2 14.5 18 3.8 19.9 14.8 34.7

20, 24) were carried out on amateur dancers, whi- bility and possible residual proprioceptive altera- le five papers (7,19,14,25,26) concerned all dan- tions, predisposing to recurrent sprains and pero- cers, without distinguishing different categories. neal tendinopathy (41-42). A paper (21) limited its research to 8-16 year old The knee sprain occurs when the femur rotates female dancers, while another paper by Steinberg with the foot fixed to the ground and a subsequ- et al (23) concerned 10-18 year old dancers. ent rotation asynchronicity between femur and As you can see in table I, only a few studies tibia, leading to a temporary loss of the connec- (9,16,17) reported the injuries of the superior tions between the two articular surfaces of the limb, while almost all the authors reported tho- knee and possible ligamentous or cartilagineous se located in the hip/groin/thigh region, althou- injuries (43-44). In dance, this can occur during gh with different percentage of prevalence. The a relapse from a jump, during the execution of a knee, also, was considered by almost all studies, pirouette with a sudden change of direction and and most studies reported the data about the regi- hyperextension of the knee (43-44). Obviously, on of foot and ankle (Table 1). depending on the mechanism of the sprain, the pathological findings will be different. For exam- DANCE INJURIES AND EPIDEMIOLOGY ple, if the injury occurs in valgus stress combined Several papers treat dance injuries and their with flexion, an injury of the medial collateral epidemiology (1-3, 5,7,10,11,13,39,40). The can occur, with possible ACL or medial meniscus high interest for this discipline is attested by the lesion. If the injury occurs in varus stress, a tear existence of a dedicated scientific journal, the of the collateral ligament, ACL or lateral menis- Journal of Dance and Medicine & Science. cus can occur (10, 45-53). Stress fractures are caused by an excessive stress Ankle and knee sprain on a bone which is normally elastic (which has Ankle sprain represented the most frequent injury preserved his physiological rigidity). They can in dance. The usual mechanisms which cause this be subdivided in insufficiency fractures, which damage are due to improper jumps and to the occur in case of weak bone (e.g. osteoporosis, slipping of the lateral side of the foot when it is tumors, osteogenesis imperfecta) and fatigue in demi-pointe position (41-42). In both cases the fractures, typical in sports activities. In the latter, ankle is in plantar flexion, and from this position it seems that the recurring load on the bone cau- it is more probable that the dancer undergoes a ses a repeated bone deformation with consequ- medial sprain with a persistent functional insta- ent local hypoperfusion and a local bone remo-

258 Rinonapoli et al. Dance and injuries

delling, which leads to local osteoporosis with a en pointe), and, if the tendon weakens, it causes fracture (54). In dancers, the most frequent fati- a pronation of the foot with a tendon sprain and a gue fractures occurs in the tibia. consequent suffering of the tendon itself. The hallux valgus is very frequent in dancers. It Muscular injuries is caused by the continuous practice of the rota- Dancers can frequently be subjected to muscular tion and of the pointe. It is characterized by the injuries, e.g. muscular contractions, elongations deformity of the hallux, which leads to a lateral with an excessive distension of the fibres, distracti- deviation of the phalanx, with a dislocation of the ons with tear of muscular bundles or fibres or, even sesamoids (8,55). Steinberg et al (8) screened a worst, rupture of the muscle in all its thickness. In group of 1336 young female dancers aged 8-16 some cases, a low force indirect injury mechani- years, and 226 control participants of the same sm can cause a muscular damage. This can be due age cohort for the presence of hallux valgus, body to an intrinsic weakness of the muscle itself or an physique characteristics, joint range of motion, imbalance of the locomotor system. Usually these and anatomical anomalies. Among the dancers, forces are provoked by a neuromuscular dysfunc- 40% had bilateral hallux valgus and 7.3% unila- tion as a sudden passive elongation due to traction teral. Among the controls, 32.3% had bilateral and force, applied during the contraction phase, or to a 1.8% unilateral hallux valgus. Following logistic too fast contraction of the muscle from a state of regression analysis, age, genu varum and scoliosis complete relaxation (15). were found to be significant predicting factors for hallux valgus in the dancer group, whereas in the Shoulder impingement syndrome control group, the predicting factors were age and The etiopathogenesis of the shoulder impingement ankle plantar flexion range of motion. syndrome is connected with the gestures executed The tendinopathy of the Achilles tendon is a by dancers with the raised arms, when in the sho- pathology which includes acute and chronic ten- ulder the subacromial space is reduced and it cre- dinosis. The pain is generally localized near the ates friction of the tendons of the rotator cuff with calcaneal insertion (at a distance of 2-8 cm), the- tendinitis and bursitis (acute or chronic) with pain refore it is vulnerable to degenerations and tears and functional limitation of the shoulder (10). (10,12). As for other athletes, the pathological forms of the Achilles tendon are more frequ- Foot and ankle injuries ently associated with training mistakes or with The dancer's foot and ankle are subjected to high an improper technique. These forms can deve- forces and unusual stresses in training and perfor- lop during the rotation, because they lead to an mance. Injuries are common in dancers, and the increase in the pronation of the metatarsus and foot and ankle are particularly vulnerable. Ankle the backfoot. During landing from a jump, if the sprains, ankle impingement syndromes, flexor heel touches the ground incorrectly, the risk of hallucis longus tendonitis, cuboid subluxation, injuries increases. stress fractures, midfoot injuries, heel pain, and The ankle posterior impingement syndrome first metatarsophalangeal joint problems -inclu occurs in the most commonly used positions such ding hallux valgus, hallux rigidus, and sesamoid as the en-pointe position (maximum plantar flexi- injuries are typical of this activity (39). on) and the demi-pointe position; these positions The Sever-Blenke disease is an osteochondro- submit to an excessive load the posterior distal sis (11) of the calcaneus, which in younger dan- margin of the tibia, the posterior side of the talus cers (between 10 and 14 years) and other athle- and the superior surface of the calcaneus with a tes can cause an insertional tendinopathy of the consequent anterior impingement (56), that can Achilles tendon, with pain prevailing during the be osseous, fibrous or osteofibrous. jump execution. The ankle anterior impingement syndrome The tendinopathy of the posterior tibial ten- occurs due to the excessive rotation of the medial don is connected with the creation of a rigid lever ankle, which can compress the anterior and the arm of the subtalar complex during the execution lateral part of the ankle, or to the execution of the of the relevé (plantar flexion or in demi-pointe or plié. In dance, this is a common gesture, where

259 Medicinski Glasnik, Volume 17, Number 2, August 2020

the rotation of the foot can increase the lateral Lumbar facet sprain and sacroiliac joint sprain compression forces and predisposes, for this rea- Lumbar facet sprain and sacroiliac (SI) jo- son, to an increased risk of anterior impingement. int sprain are causes of back pain in dancers Like the posterior one, the anterior impingement that are often overlooked and usually a result is secondary to hypertrophy of the soft tissues of of hyperextension of the lumbar spine causing the region of the ankle or to the proliferation of compression of the zygapophyseal and SI joint. the osteophytes which limit the space between Discogenic back pain accounts for approximately the anterior margin of the tibia and the backside 40% of mechanical back pain and is often seen of the neck of the talus. This happens frequently in male dancers. This pain is often aggravated in athletes, which are submitted to a forced plan- by lifting partners, such as in the pas de deux, a tar flexion, as in the case of the dance (57). duet between two principal dancers that usually is showcased at a pivotal part of the production Iliotibial band friction syndrome (40). Occasionally with advanced disc degenera- The iliotibial band friction syndrome is charac- tion, the dancer may have symptoms of sciatica terized by a “click” of the hip, that is to say a fee- or other neurological symptoms (61). ling of jerk which takes place during the flexion of the thigh. The lateral click (external) corresponds Paraspinal muscle spasm to the movement of the iliotibial band (ITB) on the It is usually attributed to extrinsic factors such greater trochanter. This lateral jerk affects more as overuse, muscle imbalance from poor tech- commonly the support leg during the rotation mo- nique, and occasionally choreography that forces vements, during the landing from jumps, or when a dancer to assume positions of imbalance, and the hip is extended from a flexion position (58-59). to intrinsic factors such as ‘‘adolescent growth The patellofemoral syndrome and the tendino- spurts, asynchronous growth, and muscular and pathy of the patellar tendon occur during the ligamentous growth lag’’(61-62). The dancer will execution of the plié and the grand-plié, in which complain of sharp, cramp-like pain located in the there is a combined flexion of the knee and the paraspinal regions (63-67). hip in external rotation position. This position creates a friction between femur and patella that, Piriformis syndrome if repeated, can induce a chondropathy and an The piriformis muscle is responsible for exter- overload of the patellar tendon. The consequen- nal rotation, weak abduction and flexion of the ces are altered patellar “tracking” or insertional hip, and stabilization of the pelvic girdle during tendinopathy of the patellar tendon. Clinically, weight bearing or movement, and it is engaged dancers suffer from an anterior knee pain (12). constantly during dancing (68-69). This con- Winslow and Yoder (60) report 29% of patellofe- stant activation causes spasm, and depending on moral pain in female dancers. the course of the sciatic nerve, irritation and subsequent pain. Hyperlordosis Snapping hip syndrome, sometimes called Hyperlordosis is typical in classical dance, whe- dancer's hip, is a condition in which you hear a re the aesthetic “perfection” requested from the snapping sound or feel a snapping sensation in canons of this genre, comes from the contraction your hip when you walk, run, get up from a cha- of the adductors and from the external rotation ir, or swing your leg around. For most people, of the hips and of the legs with an ideal 180° the condition is little more than an annoyance angle (execution en-dehor); this is a very impor- and the only symptom is the snapping sound tant element during which a lot of dancers force or sensation itself. But for dancers or athletes, this position beyond their normal limits causing snapping hip syndrome symptoms may also a lumbar pain associated with an anteversion of include pain that interferes with performan- the pelvis. In this way, an imbalance of the physi- ce. In most cases, snapping is caused by the ological curves of the vertebral column occurs, movement of a muscle or tendon over a bony leading to hyperlordosis and giving dancers the structure in the hip. The most common site is so called “duck butt” posture (40). on the greater trochanter, where the iliotibial

260 Rinonapoli et al. Dance and injuries

band passes over a part of the thigh bone that Prevention juts out. The iliopsoas tendon, which connects The prevention is the cornerstone to avoid inju- to the inner part of the upper thigh, can also ries in dance. The prevention strategies must be snap with hip movement. Another site of snap- addressed to the risk factors, which have been ping is the femoral head. The snapping occurs previously identified. They request a specific when the tendon of the rectus femoris, which training regimen, during which it is necessary to runs from inside the thighbone up through the take under control both the intensity and the du- pelvis, moves back and forth across the ball ration, which have to be progressive, and the type when the hip is bent and straightened. Winston of activity (for example, it is possible that this le- et al (70) reported on a sample of 87 professi- ads to a new dance style). A clinical screening of onal and amateur dancers with 91% prevalence the dance technique is important to build a new of snapping hip, of which 58% were painful. prevention strategy, with an evaluation of the fa- Other hip pathologies like post-traumatic sequ- ulty movements and the general health problems, elae (71), or arthritis that requires total repla- especially the musculoskeletal ones. The final cement (72), could cause problems in young objective is to avoid muscular skeletal problems, amateur dancers, or in older ones, respectively. but also to make the dancers and the trainers awa- CONCLUDING REMARKS re of the individual risk factors, in order to reduce the incidence of acute and chronic injuries. Papers regarding injuries in dance are numerous. The motor rehabilitation can help the dancers to We selected those with larger case collections, in become aware of the ineffective movement stra- which the epidemiology of the injuries connected tegies and to guide them toward alternatives in with dance was expressed. The distribution of dan- order to reduce the excessive use and the muscu- ce injuries varies from one study to another, ma- lar imbalance. king the data sometimes controversial. The yearly percentage of injured dancers goes from 42% to For example, the “transition dance” represents 97% among the different genres and skills. It has an alternative to the regular dance practice, be- been demonstrated that the most injured part of the cause it is executed under the supervision and body is the lower limb (9), especially in classical with the collaboration of a dance teacher and a dance owing to pointe shoes or to the exercise exe- health expert in order to prevent functional pro- cuted en-dehors (extra rotation of the lower limbs). blems caused by injuries. This type of dance is In modern dance, except for the lower limb, other recommended, when possible, because it allows parts of the body which are frequently injured are the dancers to continue their training in a con- the upper limb and the vertebral column, owing to trolled environment. However, this resource the high stress for the muscles in these body parts is not always available for all, but only for the and the execution of acrobatic elements during the biggest ones and the most important dance com- performances (5,9). panies and schools. Analysing the data collected from the selected In conclusion, doctors, teachers, sport trainers papers, the most anatomical region affected by and dancers themselves, that is to say all those injuries in dance is clearly the ankle and foot. who contribute to the dancer performance, sho- From the data, it can be inferred that professional uld know about the most prevalent dancers’ inju- and pre-professional dancers have a higher preva- ries. Moreover, they should know the preventi- lence of back injuries in comparison to amateur on procedures, in order to minimize the risk of dancers, while amateurs suffer more frequently injury and recurrences. from hip/groin/thigh injuries. Knee and ankle/ FUNDING foot regions injuries make no significant difference between professional and pre-professional No specific funding was received for this study. dancers versus amateurs. TRANSPARENCY DECLARATION Conflict of interest: None to declare.

261 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES 1. Lee L, Reid D, Cadwell J, Palmer P. Injury inciden- 19. Campoy FAS, de Oliveira Coelho LR, Nascimento ce, dance exposure and the use of the movement Bastos F, Netto JJ, Marques Vanderlei LC, Monteiro competency screen (mcs) to identify variables asso- HL, Padovani CR, Pastre CM. Investigation of risk ciated with injury in full-time pre-professional dan- factors and characteristics of dance injuries. Clin J cers. Int J Sports Phys Ther 2017; 12:352-70. Sport Med 2011;21:493–98 2. Van Seters C, van Rijn RM, van Middelkoop M, 20. Leanderson C, Leanderson J, Wykman A, Strender Stubbe JH. Risk factors for lower-extremity injuries LE, Johansson SE, Sundquist K. Musculoskeletal among contemporary dance students. Clin J Sport injuries in young ballet dancers. Knee Surg Sports Med 2017; 30:60-6. Traumatol Arthrosc 2011; 19:1531–5. 3. Kenny SJ, Palacios-Derflingher L, Shi Q, Whittaker 21. Steinberg N, Siev-Ner I, Peleg S, Dar G, Masharawi JL, Emery CA. Association between previous injury Y, Zeev A, Hershkovitz I. Injuries in female dancers and risk factors for future injury in preprofessional aged 8 to 16 years. J Athl Train 2013; 48:118-23. ballet and contemporary dancers. Clin J Sport Med 22. Ekegren CL , Quested R , Brodrick A. Injuries in 2017; 29:209-17 pre-professional ballet dancers: incidence, charac- 4. Fortina M, Maniscalco P, Carulli C, Meccariello L, teristics and consequences. J Sci Med Sport 2014; Colasanti GB, Carta S. Jockey injuries during the Si- 17:271-5. ena "Palio". A 72-year analysis of the oldest horse 23. Steinberg N, Aujla I , Zeev A, Redding E. Injuries race in Italy. Injury 2019; 50(Suppl 4):56-9 among talented young dancers: findings from the 5. Nilsson C, Leanderson J, Wykman A, Strender LE. U.K. Centres for Advanced Training. Int J Sports The injury panorama in a Swedish professional Med 2014; 35:238-44. ballet company. Knee Surg Sports Traumatol Ar- 24. Bowerman EA, Whatman C, Harris N, Bradshaw throsc 2001; 9:242-6. E, Karin J. Are maturation, growth and lower extre- 6. Negus V, Hopper D, Briffa NK. Associations betwe- mity alignment associated with overuse injury in eli- en turnout and lower extremity injuries in classical te adolescent ballet dancers? Phys Ther Sport. 2014; ballet dancers. J Orthop Sports Phys Ther 2005; 15:234-41. 35:307-18. 25. Trentacosta N, Sugimoto D, Micheli LJ. Hip and 7. Liederbach M, Dilgen FE, Rose DJ. Incidence of groin injuries in dancers: a systematic review. Sports anterior cruciate ligament among elite ballet and Health 2017; 9:422-27. modern dancers: a 5-year prospective study. Am J 26. Sobrino FJ, Guillé P: Overuse injuries in professio- Sports Med 2008; 36:1779-88. nal ballet. Influence of age and years of professional 8. Steinberg N, Siev-Ner I, Zeev A, Dar G. The asso- practice. Orthop J Sports Med 2017; 5:1-11. ciation between hallux valgus and proximal joint ali- 27. Bisaccia M, Rinonapoli G, Meccariello L, Bisaccia gnment in young female dancers. Int J Sports Med O, Ceccarini P, Rollo G, Ibáñez-Vicente C, Cerve- 2015; 36:67-74. ra-Irimia J, Sánchez-Sánchez F, Ribes-Iborra A, 9. Byhring S, Bø K. Musculoskeletal injuries in the Gomez-Garrido D, Caraffa A. Validity and reliabi- Norwegian National Ballet: a prospective cohort lity of mini-invasive surgery assisted by ultrasound study. Scand J Med Sci Sports 2002; 12:365-70. in achilles tendon rupture. Acta Inform Med 2019; 10. Brown TD. Foot and Ankle Injuries in Dance. Am J 27:40-4. Orthop (Belle Mead NJ) 2004; 33:303-9. 28. Rollo G, Meccariello L, Rotini R, Pichierri P, Bi- 11. Nunes N, Haddad JJ, Bartlett DJ, Obright K.D. Mus- saccia M, Fortina M. Efficacy of the “Salento tech- culoskeletal Injuries among young, recreational, nique”, a modified two-incision approach in distal female dancers before and after dancing in pointe biceps brachii tendon repair. Surgical description shoes. Pediatric Physical Therapy 2002; 14:100-6. and outcomes analysis. J Clin Orthop Trauma 2019; 12. Brotzman SB, Wilk KE. Clinical Orthopaedic Reha- 10:959-64. bilitation. 2nd Ed. Philadelphia: Mosby 2003. 29. Rollo G, Prkić A, Pichierri P, Eygendaal D, Bisaccia 13. Macintyre J., Joy E. Foot and ankle injuries in dance. M, Marco Filipponi M, Giaracuni M,Hitov P, Tanov- Clin Sports Med 2000; 19:351-68. skie K, Meccariello L. Plate-and-bone-strut fixation 14. Drezewska M, Sliwinski Z. Lumbosacral pain in of distal third humeral shaft aseptic non-unions: A ballet school students. Pilot study. Orthop Traumatol consecutive case series. J Clin Orthop Trauma 2019; Rehabil 2013; 15:149-58. 10(Suppl 1):127-32 15. Järvinen M, Orava S. Groin pain (Adductor Syndro- 30. Rollo G, Prkic A, Bisaccia M, Eygendaal D, Pichierri me). Oper Techn Sport Med 1997; 5:133-7. P, Marsilio A, Giaracuni M, Meccariello L. Grafting 16. Garrick JG, Requa RK: Ballet injuries: an analysis and fixation after aseptic non-union of the humeral of epidemiology and financial outcome. Am J Sports shaft: A case series. J Clin Orthop Trauma 2020; Med 1993; 21:586–90. 11(Suppl 1):51-5. 17. Coplan J. Ballet dancer’s turnout and its relationship 31. Rollo G, Porcellini G, Rotini R, Bisaccia M, to self-reported injury. J Orthop Sports Phys Ther Pichierri P, Paladini P, Guerra E, De Cruto E, Fran- 2002; 32:579-84. zese R, Grubor P, Pace V, Meccariello L. A new plate 18. Gamboa JM, Roberts LA, Maring J, Fergus A. Injury design to treat displaced 3-4 parts proximal humeral patterns in elite preprofessional ballet dancers and fractures in comparison to the most tested and used the utility of screening programs to identify risk plate: clinical and radiographic study. Med Glas (Ze- characteristics. J Orthop Sports Phys Ther 2008; nica) 2019; 16(2) [Epub ahead of print] 38:126–36.

262 Rinonapoli et al. Dance and injuries

32. Rollo G, Rotini R, Eygendaal D, Pichierri P, Bi- 46. Rollo G, Falzarano G, Ronga M, Bisaccia M, Grubor saccia M, Prkic A, Stasi A, Meccariello L. Effect P, Erasmo R, Rocca G, Tomé-Bermejo F, Gómez- of trochleocapitellar index on adult patient-reported Garrido D, Pichierri P, Rinonapoli G, Meccariello outcomes after noncomminuted intra-articular dis- L. Challenges in the management of floating knee tal humeral fractures. J Shoulder Elbow Surg 2018; injuries: results of treatment and outcomes of 224 27:1326-32. consecutive cases in 10 years. Injury 2019; 50(Suppl 33. Rollo G, Pichierri P, Marsilio A, Filipponi M, Bi- 4):30-8. saccia M, Meccariello L.The challenge of nonunion 47. Bisaccia M, Caraffa A, Meccariello L, Ripani U, after osteosynthesis of the clavicle: is it a biomecha- Bisaccia O, Gomez-Garrido D, Carrado-Gomez M, nical or infection problem? Clin Cases Miner Bone Pace V, Rollo G, Giaracuni M, Rinonapoli G. Dis- Metab 2017; 14:372-8. placed patella fractures: percutaneous cerclage wi- 34. Rollo G, Rotini R, Pichierri P, Giaracuni M, Stasi A, ring and second arthroscopic look. Clin Cases Miner Macchiarola L, Bisaccia M, Meccariello L. Grafting Bone Metab 2019; 16:48-52. and fixation of proximal humeral aseptic non union: 48. Rollo G, Bisaccia M, Irimia JC, Rinonapoli G, a prospective case series. Clin Cases Miner Bone Pasquino A, Tomarchio A, Roca L, Pace V, Pichierri Metab 2017; 14:298-304. P, Giaracuni M, Meccariello L. The advantages of 35. Rollo G, Bisaccia M, Franzese R, Pichierri P, Fi- type III Scaphoid Nonunion Advanced Collapse lipponi M, Giaracuni M, Gomez-Garrido D, Ripani (SNAC) treatment with partial carpal athrodesis in U, De Cruto E, Pieretti G, Meccariello L. The bio- the dominant hand: results of 5-year follow-up. Med mechanical potential of the bone graft in the proxi- Arch 2018; 72:253-6. mal ulna non-union surgery. Clin Cases Miner Bone 49. Gómez-Garrido D, Triviño-Mayoral V, Delgado-Al- Metab 2019; 16:53-61. cala V, Cervera-Irimia J, Medina-Lorca M, Sánchez- 36. Medici A, Meccariello L, Rollo G, De Nigris G, Sánchez F, Ibáñez-Vicente C, Pérez-Gurbindo I, Mccabe SJ, Grubor P, Falzarano G.Does routine Meccariello L, Rollo G, Pica G, Tomarchio A, carpal tunnel release during fixation of distal radius Pasquino A, Bisaccia M. Five year long term results fractures improve outcomes? Injury 2017; 48( Suppl of total joint arthroplasties in the treatment of tra- 3):30-3. peziometacarpal osteoarthritis. Acta Biomed 2019; 37. Manni M, Bisaccia M, Rinonapoli G, Schiavone A, 90:451-6. Meccariello L, Mccabe SJ, Bisaccia O, Vicente CI, 50. Falzarano G, Pica G, Medici A, Rollo G, Bisaccia M, Cappiello A, Caraffa A. Reliability, feasibility and Cioffi R, Pavone M, Meccariello L. Foot loading and value of echography in clinical-functional results in gait analysis evaluation of nonarticular tibial pilon patients affected by carpal tunnel syndrome: is there fracture: a comparison of three surgical techniques. J a correlation? Acta Inform Med 2017; 25:44-8. Foot Ankle Surg 2018; 57:894-8. 38. Rollo G, Prkić A, Pichierri P, Eygendaal D, Bisaccia 51. Petruccelli R, Bisaccia M, Rinonapoli G, Rollo G, M, Filipponi M, Giaracuni M, Hitov P, Tanovski K, Meccariello L, Falzarano G, Ceccarini P, Bisaccia Meccariello L. Plate-and-bone-strut fixation of distal O, Giaracuni M, Caraffa A. Tubular vs profile plate third humeral shaft aseptic non-unions: a consecuti- in peroneal or bimalleolar fractures: is there a real ve case series. J Clin Orthop Trauma 2019; 10(Suppl difference in skin complication? A retrospective stu- 1):127-32. dy in three level I trauma center. Med Arch 2017; 39. Kadel N. Foot and ankle problems in dancers. Phys 71:265-9. Med Rehabil Clin N Am 2014; 25:829-44. 52. Meccariello L, Cioffi S, Franzese R, Olivieri M, Mu- 40. Gottschlich LM, Young CC: Spine injuries in dan- gnaini M. Percutaneous reduction with k-wires of a cers. Curr Sports Med Rep 2011; 10:40-4. bimalleolar Salter Harris III: a case report. EBMJ 41. Russell JA, McEwan IM, Koutedakis Y, Wyon MA. 2012; 7 84-7. Clinical anatomy and biomechanics of the ankle in 53. Rinonapoli G, Carraro A, Delcogliano A. The clini- dance. J Dance Med Sci 2008;12:75–82 cal diagnosis of meniscal tear is not easy. Reliability 42. Ahonen J. Biomechanics of the foot in dance: a li- of two clinical meniscal tests and magnetic resonan- terature review. J Dance Med Sci 2008; 12:99–108. ce imaging. Int J Immunopathol Pharmacol 2011; 43. Whiting WC, Zernicke RF. Biomecânica da lesão 24(Suppl 2):39-44. músculo-esquelética (Biomechanics of musculo- 54. Stretanski MF, Weber JG. Medical and rehabilitation skeletal lesions) [Portuguese] Rio de Janeiro, Brazil: issues in classical ballet. Am J Phys Med Rehabil Guanabara Koogan, 2001. 2002; 81:383-91. 44. Zatsiorsky VM. Biomecânica no esporte. Performan- 55. Ceccarini P, Rinonapoli G, Nardi A, Bisaccia M, Di ce do desempenho e prevenção de lesão (Compari- Giacomo LM, Caraffa A. Bunionette: minimally in- son of the kinematics of serve in category volleyball vasive and percutaneous techniques: a topical review youth among female and male athletes) [Portuguese] of the literature. Foot Ankle Spec 2017; 10:157-61. Rio de Janeiro, Brazil: Guanabara Koogan, 2004. 56. Hamilton LH, Hamilton WG, Meltzer JD, Marshall 45. Fortina M, Maniscalco P, Carulli C, Meccariello L, P, Molnar M. Personality, stress, and injuries in pro- Colasanti GB, Carta S. Jockey injuries during the Si- fessional ballet dancers. Am J Sports Med 1989; ena "Palio". A 72-year analysis of the oldest horse 17:263-7. race in Italy. Injury 2019; 50 (Suppl 4):56-9. 57. Tol JL, Slim E, van Soest AJ, van Dijk CN. The relationship of the Kicking Action in Soccer and Anteri- or Ankle Impingement Syndrome; Am J Sports Med 2002; 30:45-50.

263 Medicinski Glasnik, Volume 17, Number 2, August 2020

58. Foch E, Aubol K, Milner CE. Relationship between 66. Bonura EM, Morales DJO, Fenga D, Rollo G, iliotibial band syndrome and hip neuromechanics in Meccariello L, Leonetti D, Traina F, Centofanti F, women runners. Gait Posture 2020; 77:64-8. Rosa MA. Conservative treatment of spondylodisci- 59. Clearmann RR. Physical Medicine and Rehabilita- tis: possible therapeutic solution in case of failure of tion. Philadelphia: Lippincott Williams & Wilkins, standard therapy. Med Arch 2019; 73:39-43. 2005. 67. Cervera-Irimia J, González-Miranda Á, Riquel- 60. Winslow J, Yoder E. Patellofemoral pain in female me-García Ó, Burgos-Flores J, Barrios-Pitarque C, ballet dancers: correlation with iliotibial band ti- García-Barreno P, García-Martín A, Hevia-Sierra E, ghtness and tibial external rotation. J Orthop Sports Rollo G, Meccariello L, Caruso L, Bisaccia M. Sco- Phys Ther 1995; 22:18-21. liosis induced by costotransversectomy in minipigs 61. Khan K, Brown J, Way S, et al. Overuse injuries in model. Med Glas (Zenica) 2019; 16(2) [Epub ahead classical ballet. Sports Med 1995; 19:341-57. of print] 62. Baker R, Patel D. Lower back pain in the athlete: 68. Boyajian-O’Neill L, McClain R, Coleman M, Tho- common conditions and treatment. Prim Care 2005; mas P. Diagnosis and management of piriformis 32:201-29. syndrome: an osteopathic approach. JAOA 2008; 63. Medici A, Meccariello L, Falzarano G. Non-operati- 108:657-64. ve vs. percutaneous stabilization in Magerl's A1 or 69. Kirschner J, Foye P, Cole J. Piriformis syndrome, di- A2 thoracolumbar spine fracture in adults: is it re- agnosis and treatment. Muscle Nerve 2009; 40:10-8. ally advantageous for a good alignment of the spine? 70. Winston P, Awan R, Cassidy JD, Bleakney RK. Cli- Preliminary data from a prospective study. Eur Spine nical examination and ultrasound of self-reported J 2014; 23(Suppl 6):677-83. snapping hip syndrome in elite ballet dancers. Am J 64. Meccariello L, Muzii VF, Falzarano G, Medici A, Sports Med 2007; 35:118-26. Carta S, Fortina M, Ferrata P. Dynamic corset ver- 71. Fortina M, Carta S, Crainz E, Urgelli S, Del Vecchio sus three-point brace in the treatment of osteoporotic E, Ferrata P. Management of displaced intracapsular compression fractures of the thoracic and lumbar femoral neck fracture in young adult: Why compli- spine: a prospective, comparative study. Aging Clin cations are still so high? Case report of posttraumatic Exp Res 2017; 29:443-9. avascular necrosis in a 30-year-old man and a brief 65. Maiettini D, Orgera G, Bisaccia M, Piscitelli L, review. J Trauma 2009; 67:163-6. Laurino F, Meccariello L, Rebonato S, Schiaroli E, 72. Ferrata P, Carta S, Fortina M, Scipio D, Riva A, Di Rossi M, Rebonato A. Percutaneous vertebroplasty Giacinto A. Painful hip arthroplasty: definition. Clin improves pain control and quality of life in patients Cases Miner Bone Metab 2011; 8:33-6. suffering from back pain: a single center experience. Iran J Radiol 2017 :e41746 (In Press)

264 ORIGINAL ARTICLE

Spatio-temporal data visualization for monitoring of control measures in the prevention of the spread of COVID-19 in Bosnia and Herzegovina Mirza Ponjavić1, Almir Karabegović2, Elvir Ferhatbegović3, Emin Tahirović4, Selma Uzunović5, Maja Travar6, Aida Pilav7, Maida Mulić8, Sead Karakaš9, Nermin Avdić10, Zarina Mulabdić11, Goran Pavić12, Medina Bičo13, Ivan Vasilj14, Diana Mamić15, Mirsada Hukić16

1International Burch University, Sarajevo, 2Electrical Engineering Faculty, University of Sarajevo, 3GAUSS Centre for Geospatial Research Sarajevo, 4International University of Sarajevo; Sarajevo, 5Department of Clinical Microbiology, Institute for Health and Food Safety Ze- nica, Zenica, 6University Clinical Centre of the Republic of Srpska, Banja Luka, 7Institute for Public Health of Sarajevo Canton, Sarajevo, 8Institute for Public Health of Tuzla Canton, Tuzla, 9Institute for Public Health of Central Bosnia Canton, Travnik, 10Institute for Public Health of Herzegovina-Neretva Canton, Mostar, 11Institute for Public Health of Una-Sana Canton, Bihać, 12Institute for Public Health of Posavina Canton, Orašje, 13Institute for Public Health of Bosnian-Podrinje Canton, Goražde, 14Institute for Public Health of West Her- zegovina Canton, Grude, 15Institute for Public Health of Herzeg-Bosnian Canton, Livno, 16Academy of Sciences and Arts of Bosnia and Herzegovina, Sarajevo; Bosnia and Herzegovina

ABSTRACT

Aim The damage caused by the COVID-19 pandemic has made the prevention of its further spread at the top of the list of priorities of many governments and state institutions responsible for health and civil protection around the world. This prevention implies an effective system of epidemiological surveillance and the application of timely and effective control measures. This research focuses on the application of techniques for modelling and geovisualizati- Corresponding author: on of epidemic data with the aim of simple and fast communicati- Mirza Ponjavić on of analytical results via geoportal. International Burch University Methods The paper describes the approach applied through the Francuske revolucije bb, project of establishing the epidemiological location-intelligence 71210 Ilidža, Sarajevo, system for monitoring the effectiveness of control measures in Bosnia and Herzegovina preventing the spread of COVID-19 in Bosnia and Herzegovina. Phone: +387 33 944 400; Results Epidemic data were processed and the results related to Email: [email protected] spatio-temporal analysis of the infection spread were presented by ORCID ID: https://orcid.org/0000-0003- compartmental epidemic model, reproduction number R, epi-cu- 0124-6603 rve diagrams as well as choropleth maps for different levels of administrative units. Geovisualization of epidemic data enabled the release of numerous information from described models and indicators, providing easier visual communication of the spread of the disease and better recognition of its trend.

Original submission: Conclusion The approach involves the simultaneous application 11 June 2020; of epidemic models and epidemic data geovisualization, which Revised submission: allows a simple and rapid evaluation of the epidemic situation and the effects of control measures. This contributes to more infor- 14 June 2020; mative decision-making related to control measures by suggesting Accepted: their selective application at the local level. 20 June 2020 doi: 10.17392/1215-20 Key words: COVID 19 pandemic, decision making, epidemiological techniques, geocoding, health information systems, reproduction number Med Glas (Zenica) 2020; 17(2): 265-274

265 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION mics of the disease spread and spatio-temporal visualization of data with a clear presentation of Following reports on the first patients with viral the geographical area and time interval related to pneumonia caused by COVID-19 in December these indicators, which allows easier comparison 2019, in the city of Wuhan in China, the disease of the effects of control measures. spread rapidly throughout the world (1). Due to the major global problems caused by this new corona- These functions are built into the ELIS geoportal, virus, prevention of its spread has become the top which is a presentation tool for spatio-temporal priority for most governments, medical, economic, data visualization. and political communities around the world (2). The aim of this research was to improve the In order to respond quickly to the rapid deve- application of epidemiological techniques with lopment of the epidemic situation related to the visualization of spatio-temporal epidemic data spread of COVID-19 in Bosnia and Herzegovi- and the functionality of the geoportal in terms of na (B&H), the Committee for Microbiology and simpler and faster epidemic analysis and commu- Related Disciplines of the Department of Medical nication of the results to make more informati- Sciences at the Academy of Sciences and Arts of ve decisions related to the application of control Bosnia and Herzegovina (ANUB&H) launched measures. The paper describes the approach used the Project of Epidemic Location and Intelligen- in Bosnia and Herzegovina through the Project ce System (ELIS) for monitoring the spread of of the Epidemic Location Intelligence System, COVID-19. The system was tasked with provi- which will serve for the implementation of epi- ding research and collecting epidemic data on demic analysis functionality important for emer- COVID-19, communicating with epidemiological gency response to the epidemic threat. teams, exchanging information on infection con- MATERIALS AND METHODS trol, analysing the situation and control measures efficiency, reporting to public health institutions Study setting and design on the evaluation of epidemic surveillance system, informing government institutions and the general The public health system in B&H is organized in public about the epidemic situation. accordance with its administrative and political structure as a country consisting of Brčko Dis- The ELIS functionality includes: data structure trict (BD) and two entities: Republic of Srpska analysis and modelling (3), geocoding and map- (RS) and the Federation of Bosnia and Herze- ping of tested, confirmed and active cases (4) govina (FB&H) with 10 cantons. using different address models (5), visualization of infection spread over time, mapping of areas For the needs of the Epidemic Location and In- with vulnerable age groups, mapping of health telligence System (ELIS) implementation, due to capacities with available medical resources to the heterogeneity of health information systems respond effectively to the current epidemic si- and protocols for the use of epidemiological data tuation (6), epidemic modelling and prediction, in B&H, a special network for data exchange was and communication and cooperation tools for the established allowing direct access to epidemic epidemic surveillance professionals. records and databases of public health institutes (as the main data sources at the local level), and Due to its importance for the whole country and providing continuous downloading of data pu- its role in the academic community, this system blished via the official websites of B&H instituti- has been accepted as a common platform for ons. Epidemic data collected from these sources further research and a source of information for were processed applying the adopted analytical monitoring the effectiveness of control measures methodology, and the results were presented by in preventing the disease spread. tables, diagrams, interactive maps and animati- Of particular importance for the monitoring and ons, arranged in the geoportal (Figure 1). evaluation of control measures are the functions of the system related to modelling and communi- Methods cation, because they enable presentation of key To analyse the data and to identify the trend of indicators and critical information on the dyna- infection spread, several different methods were

266 Ponjavić et al. Spatio-temporal epidemic data visualization

cal data for cantons and municipalities, were taken from databases and available records of public health institutions (10). These data sets were supplemented by other available clinical and laboratory information necessary to develop the model for prediction of the spatio-temporal spread of infection. Analytical models for monitoring the spread of COVID-19 were selected in accordance with the availability of epidemiological data sets. The Figure 1. Epidemic Location and Intelligence System (ELIS) geoportal: presentation of interactive map of the infection dynamics of infectious growth was modelled spread by municipalities with a related report and epidemic using the SIR (Susceptible; Infectious; Recove- models red) model (Figures 2, 3) applied for the two possible scenarios (Figure 4), i.e. for the pessimistic and optimistic scenario of the infection spread. Also, epi-curves (Figure 5) and reproduction number R (Figures 6, 7, 8) were used to monitor Figure 2. Susceptible – Infectious – Recovered (Resistant) the effectiveness of control measures. Based on model with parameters ß (probability of contact between in- these models, reports with epidemiological para- fected and susceptible persons) and g (probability of recovery of the infected person) meters related to the dynamics and prediction of the disease spread were generated and published applied including compartmental epidemic mo- via the ELIS geoportal. del, reproduction number R, epi-curve diagrams, For modelling of the growth dynamics of dise- descriptive statistics (with different types of ased cases, the SIR model (Figures 2, 3) was graphs), as well as choropleth maps for different applied, which was fitted by calculating the pa- levels of administrative units. Each of the met- rameters β (probability of contact between in- hods is useful for indicating spatial and temporal fected and susceptible persons), γ (probability changes related to the spread of the disease, but of recovery of the infected person, which is infor the correct interpretation of the results their verse to the average recovery time) (11), and the simultaneous use is often required. total number of susceptible persons, Nopt for the

Modelling of epidemiological data for monito- optimistic scenario and Npes for the pessimistic ring and predicting the spread of COVID-19. scenario (1). The genetic algorithm was used as Basic data sets including officially daily reported a mechanism to calculate the optimal values of values on cumulative number of laboratory confir- the parameters β and Nopt. med cases of COVID-19 for B&H entities, Brčko He optimistic scenario (Figure 4b) refers to the District and the whole country, and epidemiologi- assumption that due to the action of the control

Figure 3. The Susceptible – Infectious – Resistant (SIR) model for Bosnia and Herzegovina based on an optimistic scenario and data collected from March 4 to May 18, 2020. The model is updated by continuous data entry on new confirmed cases

267 Medicinski Glasnik, Volume 17, Number 2, August 2020

Figure 4. Growth of the number of infected cases from March 4 to May 8, 2020 (66 days from the beginning of the epidemic monitoring) in Bosnia and Herzegovina: A) current and modelled optimistic and pessimistic growth of infection and B) current and modelled optimistic growth of infection (detailed view)

measures taken, the spread of COVID-19 will ber of secondary cases of the disease caused rapidly slow down and stop. The pessimistic sce- by a single infected individual over his/or her nario (Figure 4a) refers to the assumption that the infectious period. The R0 (basic reproducti- proportion of the total infected cases will be sion number) represents a starting value (at the milar to those countries that took similar control beginning of the epidemic), assuming that the measures, but had a significantly higher rate of the whole population is susceptible to the infection disease. These two scenario models are based on and no restrictive societal measures have been continuously updated data enabling continuous undertaken so far. Although R0 is useful for prediction of the disease dynamics in time close judging of general severity of the epidemic at to the current moment. This approach was used to its own start, it is of limited use for assessing report on current dynamics and prediction of the a subsequent change caused by population be- growth of the infection at all levels of spatial units haviour changes. Calculating instantaneous in B&H, i.e. for settlements, municipalities, can- reproduction number Rt allows an evaluation tons, entities, district and the country as a whole. of restrictive policies imposed that it results in Reproduction number R. The reproducti- higher alertness of the population, which sho- on number (R) represents the average num- uld lead to a decrease of Rt.

Figure 5. Epi-curve diagram based on daily confirmed cases in Bosnia and Herzegovina from March 5 to April 10 2020

268 Ponjavić et al. Spatio-temporal epidemic data visualization

Figure 6. Instantaneous reproduction number Rt for Bosnia and Herzegovina (B&H) with control measures introduced in Republic of Srpska (RS) (full vertical lines) and the Federation of B&H (dashed lines), and relaxing of the introduced measures (dotted lines) Mapping of epidemic data. In general, epidemi- cumulative number of infected cases and their ological data, such as, for example, quantitative geographical distribution, a time lining animati- data on the representation of tested or infected on technique was applied (Figure 10). persons, qualitative data on the manufacturer and type of test, or territorial jurisdiction of dispen- RESULTS AND DISCUSSION saries and hospitals, relating to specific spatial Key indicators on the dynamics of the disease units (settlements, municipalities, cantons) can spread for certain geographical areas were obtai- be classified according to the appropriate ranks ned from the SIR epidemic model, for the whole and shown with the applied classification sche- country (Figure 3), entities, cantons or cities. Also, me of colours on the map (8). As a special met- epi-curves were used to show the growth rate of hod for spatial-temporal visualization of epide- diseased cases (Figure 5), and diagrams showing mic data, the technique with series of maps was the R reproduction number (7) were used as a spe- used (Figure 9). Also, to show the growth of the cial mechanism for monitoring the effectiveness

Figure 7. Instantaneous reproduction number Rt for Federation of B&H: inception of the restrictive measures (dashed vertical lines) and relaxing of the measures (dotted lines)

269 Medicinski Glasnik, Volume 17, Number 2, August 2020

Figure 8. Instantaneous reproduction number Rt for Republic of Srpska: inception of the restrictive measures (full vertical lines) and relaxing of the measures (dotted lines)

of control measures and making decisions on the- number application for monitoring of the control ir intensification or relaxation. These techniques measures in the prevention of the spread of CO- are very useful for temporal presentation (12); VID-19 in B&H. however, they are not suitable or sufficient per se Being able to calculate the instantaneous repro- for comparing different indicators and epidemic ductive number Rt allows an evaluation of effec- information by geographical regions. For exam- tiveness of the introduced societal measures and ple, a diagram with the R reproduction number to recognize good time for them to be intensified represents the relative change in the growth rate or relaxed. The Rt and the rate of spread (expo- of infected persons in a given time interval for nential model) of the epidemic are functionally the observed region, which can be useful for re- related, and the relationship depends on the assu- cognizing the effect of current control measures. med epidemiological model (12). The Rt value It cannot be used for quantitative comparison of that exceeds 1 implies that the infection is sprea- the effects of the same control measures between ding at an exponential rate. In practice, using the different regions. Furthermore, tabular and graphi- real data from the field, the Rt is calculated using cal representations of epidemic parameters and in- time delayed data: since Rt captures the potential formation for a limited number of spatial units are of the infectious spread, reporting data need to descriptive and understandable, but not suitable be adjusted for the length of symptomatic peri- for comparison between a large number of units at od before reporting day, as well as, the length of different administrative levels, such as settlements the incubation period. A difference between re- or municipalities. However, geovisualization of porting date of the first known COVID 19 case epidemic data in conjunction with diagrams and and imputed date of an infection (Figures 6, 7, 8) tabular data can achieve spatio-temporal visual illustrates this time difference. communication and thinking (14) about the distri- Figure 6 shows the change of the Rt for B&H, bution of a disease magnitude (9). illustrating the changes in the disease trend that It is necessary to emphasize the importance of were affected by all control measures applied in geovisualization and its synthesis with epidemic FB&H and RS. modelling techniques in recognizing the trend of Figures 7 and 8 show the change of the Rt for the disease spread, correct inference and more FB&H and RS, respectively, with the inception informative decision-making related to control of control measures and dates of their relaxation. measures. This synthesis proved to be a parti- From the diagram of the reproduction number R cularly useful approach with the reproduction

270 Ponjavić et al. Spatio-temporal epidemic data visualization

Figure 9. Daily average of confirmed cases per 100,000 inhabitants (top raw) and total number of cases per ten-day interval (bottom raw) in municipalities over three ten-day periods from April 09 to May 08 2020 for the entity of RS (Figure 8), where the last re- infection throughout the area of interest. In addi- ading was below 0.5, it could be concluded that tion to the application of reproduction number R, the current situation is generally better and that this implies geovisualization of relevant epidemic control measures work better than in the entity of data and their proper simultaneous interpretation. the FB&H (Figure 7), where the last reading was Visualization of spatio-temporal epidemic data 0.5. However, this is not the case, because these diagrams represent the relative change in the rate Epidemic data can be aggregated at different ad- of infection spread and cannot be compared with ministrative levels and presented through various each other in absolute terms. In order to be able to chart types. This presentation is clear and under- make a quantitative comparison of the effects of standable, and the values related to individual control measures between entities, i.e. geographi- spatial units can be easily compared with each cal regions in general, it is necessary to have a other. Figure 11 shows a bar chart with the num- broader understanding of changes in the rate of the ber of confirmed cases per 10,000 inhabitants for the various administrative levels, i.e. for the three most affected cities, for B&H, RS, FB&H, BD and 10 cantons. In this way, the number of confirmed cases can be compared among spatial units at administrative levels (for example for municipalities), cantons and entities in B&H. However, if it is necessary to expand the comparison by certain time periods, this type of presentation is not sufficiently readable and clear, and it is necessary to introduce other techniques of presentation, e.g. visualization of data. One of the visualization possibilities is the application of time lining tool for geovisualization, e.g. animation in the ELIS geoportal, which ena- Figure 10. Spatial distribution of laboratory-confirmed cases bles temporal analysis of the epidemic data by gi- of COVID-19 in Bosnia and Herzegovina for May 18, 2020. The figure shows the cities and municipalities with higher intensity ving a broader overview of the spatial distributi- of infectious spread on and rate of spread of the infection (Figure 10).

271 Medicinski Glasnik, Volume 17, Number 2, August 2020

Figure 11. Number of confirmed cases per 10,000 inhabitants for the three the most affected cities in Bosnia and Herzegovina (B&H) with two entities, the district, and for ten cantons in the Federation of B&H This tool can be used simultaneously to sequen- provide a dynamic comparison and broader un- tially view changes for an epidemic phenomenon derstanding of changes in the intensity of the on a map (number of infected and resistant), and infection between spatial units at a given admi- to temporally animate these changes in a space nistrative level. (13). Combined with interactive symbols, anno- tations and thematic maps, it represents a very Epidemic analysis and communication of the useful and powerful mechanism for monitoring findings the development and identification of spatio-tem- Interactive maps and control panels were used to portal patterns of behaviour of an epidemiolo- communicate the findings via the ELIS geoportal gical phenomenon, such as the rate of spread of (Figure 1) so that the epidemic surveillance pro- the disease and its spatial distribution (14). This fessionals can quickly communicate the current mechanism can also be applied to the thematic situation and be informed about its expected deve- presentation of data with a choropleth map (14). lopment. Interactive dashboards are a particularly As an example, Figure 9 shows the daily average useful tool because they display key indicators and of confirmed cases per 100,000 inhabitants and provide summary critical information (3). Geopor- the total number for the period in B&H munici- tal users, scientists, health and other professionals palities for three ten-day periods from April 9 to responsible for epidemic surveillance, have access May 8, 2020. Such maps, in combination with to cartographic layers, diagrams, tables, animati- other thematic representations, such as socio-de- ons, series of maps, reports, various records and mographic data (risk age groups, sensitive social other documents related to specific spatial units categories), meteorological data, contamination, used to analyse the epidemic situation (Figure 1). and with background contextual layers (orthorec- This analysis should offer answers to questions tified images, street maps) represent a very useful related to the assessment of the current situation, set of information for epidemiological analysis. the effectiveness of control measures, the growth Their role in the context of assessing epidemic trend of infected and resistant cases, spatial spre- situation and monitoring control measures is to ad of the infection throughout and in parts of the

272 Ponjavić et al. Spatio-temporal epidemic data visualization

country etc. It is especially important to continuo- es and municipalities (Figures 9, 0). In terms of usly monitor the situation in the regions that have the number of infected persons, Banja Luka was recorded an intensive spread of the infection and the most severely affected city in B&H (Figures rapid observation of spatial patterns to indicate 9, 10) and had a risk for the infection to spread new outbreaks (13). to other areas. This suggests the application of The following epidemic indicators are used for selective control measures at the local level (by monitoring the situation: cumulative growth of individual cantons, municipalities or settlements) confirmed cases, number of daily confirmed, that would be balanced between economic and recovered, tested cases and deaths, expected social justification in terms of sustainability of number of new cases and their ratios by age and economic activities and minimal impact on the gender groups, reproduction number R and other vulnerable population (15). relevant information. They are mainly presented Such an epidemic analysis and description of the in the form of individual values, tables and dia- situation throughout the country would not be grams. However, only when they are applied si- simple without the simultaneous application of multaneously with mapped data that give them a epidemic models, geovisualization of the epide- geographical context, this information can provi- mic data and the communication functionality of de a complete overview of the epidemic situation the geoportal that arranges all this. in the space (13). In conclusion, epidemiological techniques for Based on epidemic models (SIR) and diagrams monitoring of the spread of infection and the (epi-curves and R), spatial-temporal animation effectiveness of control measures are very use- and series of maps of the infection spread, the ful for temporal presentation, but are often not spread of COVID-19 virus can be monitored and sufficient to compare different indicators and predicted, and the effects of control measures in epidemic information by geographical regions. individual regions can be evaluated. Based on the Geovisualization of epidemic data enables the SIR model it can be concluded that on May 18 release of numerous information from described 2020, the trend in the number of confirmed cases models and indicators, providing an easier visual in B&H (Figure 4) was declining. Reproduction communication of the spread of the disease. Its number R (Figure 6) was below 1, indicating that synthesis with epidemiological modelling tech- the situation was generally good and that the con- niques provides better recognition of the trend of trol measures taken were effective and on time. the infection spread and a more transparent asse- However, the situation differed at lower admini- ssment of the epidemic situation. strative levels. In the RS entity, the number of infected persons was significantly higher (Figure ACKNOWLDGEMENT 11), but declined faster than in the Federation of Would like to thank to President of the Presi- B&H (Figures 7, 8). The trend of infection spre- dency of Bosnia and Herzegovina for supporting ading in the entity of the Federation of B&H has the Project of the Epidemic Location Intelligence been somewhat more favourable in the period of System, to the Academy of Sciences and Arts of last twenty days (Figure 9) compared to the who- Bosnia and Herzegovina (ANUB&H) for the male previous period, what is also confirmed by the terial provided for the preparation of this paper, declination in the reproductive number R (Figure and to the Cantonal Institutes for Public Health 7). Furthermore, the situation is different at the for epidemic data provided. lower level in some entities. In the city of Ba- nja Luka (RS), the situation has deteriorated over FUNDING the last twenty days (Figure 9), while in most No specific funding was received for this study. cantons the spread of the infection has stopped. Potential outbreaks may be associated with the TRANSPARENCY DECLARATION cities of Mostar (Herzegovina-Neretva Canton) Conflicts of interest: None to declare. and Maglaj (Zenica-Doboj Canton), as they had a slower declining trend compared to other citi-

273 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES 1. Huang Y, Yang L, Dai H, Tian F, Chen K. Epidemic 9. Murad A, Khashoggi BF. Using GIS for disease situation and forecasting of COVID-19 in and outsi- mapping and clustering in Jeddah, Saudi Arabia. de China. Bull World Health Organ 2020; [E-pub ISPRS Int J Geo-Inf 2020; 9:328. ahead of print] 10. Ministry of Civil Affairs of Bosnia and Herzegovina. 2. World Health Organization (‎2020) COVID-19 Stra- Epidemic Situation (Daily Review) 2020 http://mcp. tegy Update https://www.who.int/publications-deta- gov.ba/publication/read/epidemioloska-slika-covid- il-redirect/covid-19-strategy-update (14 April 2020) 19?pageId=3 (13 June 2020) 3. Karabegovic A, Ponjavic M. Geoportal as Interfa- 11. Althaus CL, Low N, Musa E O, Shuaib F, Gsteiger ce for Data Warehouse and Business Intelligence S. Ebola virus disease outbreak in Nigeria: transmi- Information System. In: Mach-Król M, Pełech-Pi- ssion dynamics and rapid control. Epidemics 2015; lichowski T, eds. Advances in Business ICT. Ad- 11:80–4. vances in Intelligent Systems and Computing. Vol 12. Wallinga J, Lipsitch M. How generation intervals 257. Springer, Cham 2014: 27–40. shape the relationship between growth rates and re- 4. Herbreteau V, Révillion C, Trimaille E. GeoHealth productive numbers. Proc R Soc B 2007; 274:599– and QuickOSM, two QGIS plugins for health appli- 604. cations. Nicolas Baghdadi; In: Mallet C, Zribi M, 13. Sarwar S, Waheed R, Sarwar S, Khan A. COVID-19 eds. Earth Systems – Environmental Sciences: QGIS challenges to Pakistan: is GIS analysis useful to draw in Remote Sensing Set, 1. ISTE, 2018: 257-86; solutions? Sci Total Environ 2020; 730:139089. 5. Coetzee S, Ivánová I, Mitasova H, Brovelli M. Open 14. Ponjavic M, Karabegovic A, Ferhatbegovic E. Ge- Geospatial Software and Data: a review of the cu- oinformation Management: Concepts and Methods rrent state and a perspective into the future. ISPRS (textbook). Tuzla: University of Tuzla, Bosnia and Int J Geo-Inf 2020; 9:90. Herzegovina 2020: 129–36. 6. Zhou C, Su F, Pei T, Zhang A, Du Y, Luo B, Cao Z, 15. World Health Organization. Considerations for Wang, J, Yuan, W, Zhu Y, Song C, Chen J, Xu J, Li quarantine of individuals in the context of conta- F, Ma T, Jiang L, Yan F, Yi J, Hu Y, Liao Y, Xiao H. inment for coronavirus disease (‎‎‎COVID-19)‎‎‎: inte- COVID-19: challenges to GIS with big data. Geo- rim guidance, 19 March 2020 https://www.who.int/ graphy and sustainability 2020; 1:77–87. publications/i/item/considerations-for-quarantine- 7. Goldstein E, Paur K, Fraser C, Kenah E, Wallinga J, of-individuals-in-the-context-of-containment-for- Lipsitch M. Reproductive numbers, epidemic spread coronavirus-disease-(covid-19) (13 June 2020) and control in a community of households. Math Bi- osci 2009; 22:11–25. 8. Farkas D, Hilton B, Pick J, Ramakrishna H, Sarkar A, Shin N. A Tutorial on Geographic Information Systems: a ten-year update. JAIS 2016; 38, Article 9.

274 ORIGINAL ARTICLE

Minimal impact of COVID-19 outbreak on the postoperative morbidity and mortality following emergency general surgery procedures: results from a 3-month observational period

Charalampos Seretis, Lucy Archer, Lida Lalou, Shuker Yahia, Christian Katz, Iram Parwaiz, Altaf Haji, Lourdusamy Selvam

Department of General Surgery, George Eliot Hospital NHS Trust, Nuneaton, Warwickshire, United Kingdom

ABSTRACT

Aim The outbreak of the COVID-19 pandemic has had a major impact on the delivery of elective, as well as emergency surgery on a world-wide scale. Up to date few studies have actually assessed the impact of COVID-19 on the postoperative morbidity and mortality following emergency gastrointestinal surgery. Herein, we present our relevant experience over a 3-month period of unin- terrupted provision of emergency general surgery services in Ge- orge Eliot Hospital NHS Trust, the United Kingdom.

Methods We performed a retrospective analysis of a prospective institutional database, which included the operation types, paracli- nical investigations and postoperative complications of all patients Corresponding author: undergoing emergency general surgery operations between March Charalampos Seretis –May 2020. Specialty Registrar in General Surgery, Results The occurrence of a 5% overall respiratory complication Department of General Surgery, rate postoperatively, with 3% infection rate for COVID-19 was George Eliot Hospital NHS Trust found; no patient had unplanned return to intensive care for ven- College Street, CV10 7DJ, Nuneaton, tilator support and there was no mortality related to COVID-19 Warwickshire, West Midlands, infection. United Kingdom Conclusion When indicated, emergency surgery should not be de- Phone: +44 24 7635 1351; layed in favour of expectant/conservative management in fear of Email: [email protected] COVID-19-related morbidity or mortality risks. ORCID ID: https://orcid.org/0000-0002- Key words: emergency, SARS-CoV-2, surgery 7762-5733

Original submission: 03 July 2020; Accepted: 09 July 2020 doi: 10.17392/1229-20

Med Glas (Zenica) 2020; 17(2): 275-278

275 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION performed surgery, the overall in hospital length of stay, the duration of hospitalisation and possi- Following the initial infection wave form CO- ble readmission to intensive care, as well as the VID-19 in China in 2019, the European health occurred postoperative complications and asse- services started to become affected during the ssed the existence of potential correlation betwe- first months of 2020, altering in many aspects en those parameters and suggested infection from the delivery of emergency medical and surgical COVID-19. Our study did not influence the pati- care (1). With respect to the provision of acute ent care, hence no approval from our institutional care surgery, the options of delaying the defini- ethical committee was required. tive operative management can lead to dramatic increase of the morbidity and mortality, in con- Methods trast to elective operations, most of which can be postponed with relatively low risk for the pati- Screening for COVID-19 infection on the acute ents (2). During these last months, most of the surgical admissions depended on the presence of main scientific bodies and surgical societies have concurrent symptomatology, as well as the current issued relevant guidance regarding the aspects of institutional guidance at the time of patients’ ad- provision of acute care surgery, focusing on pati- missions, which was updated at least twice per ent and personnel safety (3,4). With the scientific month. Our policy to assess the COVID-19 pati- community still in the process of understanding ent status included a combination of oropharyn- the disease process and its system-specific im- geal swabs, which was analysed with real-time plications, acute care surgery services inevitably polymerase chain reaction (RT-PCR), and uncon- have to continue to function; therefore, it is of trasted computed tomography (CT) of the chest. paramount importance to assess the true impact All patients that clinically were likely to undergo of the COVID-19 pandemic on the emergency a laparotomy and require postoperative admission surgery postoperative outcomes. Up to date, the- to our Intensive Care Unit, had a completion CT re are only few clinical studies and case series chest along with the performance of abdominopel- addressing the impact of the COVID-19 infection vic CT that was requested during the diagnostic on the emergency surgery postoperative outco- workup for their presenting acute surgical patho- mes with variable results (5,6). logy. Those additional thoracic CT scans were per- Aim of this study was to present our experience formed and reported by a Consultant Radiologist relating to COVID-19-associated postoperative within our institution and the analysis of RT-PCR morbidity and mortality after emergency gastrointe- specimens was performed in our institution as stinal surgery over a 3-month period, between Mar- well, with average time of results release fluctua- ch-May 2020, during which our centre provided ting between 48 – 72 hours from admission. continuous emergency general surgery services. RESULTS PATIENTS AND METHODS A total of 100 patients were analysed, of which 56 (56%) were females and 44 (44 %) males, Patients and study design with an age span between 17-88 years (mean age We reviewed retrospectively all general surgery 55.6 years). With respect to the type of surgery emergency operations performed between 01 performed, out of 100 patients, 32 underwent a March– 31 May 2020 in George Eliot Hospital laparoscopic procedure, 65 had upfront open NHS Trust using the relevant software from our surgery and 3 were laparoscopic converted to dedicated emergency theatre. open cases. In detail, the following procedures were undertaken: 26 appendicectomies (24 lapa- A total of 103 patients were identified, of which roscopic, 2 laparoscopic converted to open), 22 three patients were excluded from the outcome emergency laparotomies, 14 open hernia repairs analysis, as they were still inpatients at the time (9 inguinal, 2 umbilical, 1 obturator, 1 femoral, of the analysis (all three tested negative via RT- 1 Spigelian), 10 cholecystectomies (9 laparosco- PCR on admission for COVID-19, with no res- pic, 1 open), 25 abscess requiring drainage under piratory complications to date). Our final sample general anaesthesia (13 anorectal, 12 on trunk), consisted therefore of a total of 100 patients. 2 examinations of the anorectum under general We reviewed the patients’ demographics, type of anaesthesia with no additional procedure and fi-

276 Seretis et al. COVID-19 morbidity/mortality in emergency surgery

Table 1. Characteristics of the 11 patients with non-respiratory postoperative complications*

Overall ITU Primary procedure Type of complication Management LOS LOS (days) (days) Laparoscopic converted to open appendicectomy Intra-abdominal collection Antibiotics 24 0 Laparotomy and adhesiolysis (no resection) Urinary tract infection Antibiotics 12 0 Laparoscopic converted to open appendicectomy Intra-abdominal collection Antibiotics 5 0 Laparotomy and right hemicolectomy Pulmonary embolism Therapeutic anticoagulation 10 2 Open suture repair of obturator hernia Central line and wound infection Antibiotics 14 2 Laparoscopic appendicectomy Urinary tract infection Antibiotics 2 0 Open suture repair of inguinal hernia Acute coronary syndrome Therapeutic anticoagulation 8 1 Laparoscopic cholecystectomy Intra-abdominal collection Antibiotics 4 0 Open nesh repair of Spigelian hernia Pulmonary embolism Therapeutic anticoagulation 3 0 Laparotomy & Hartmann’s procedure Wound infection / Intra-abdominal collection Antibiotics 21 3 Laparotomy and defunctioning ostomy Cardiac arrythmia / Intra-abdominal collection Medical therapy, antibiotics 14 0 *Mean overall LOS 10.6 days, mean ITU LOS 0.6 days LOS, length of stay; ITU, Intensive Therapy Unit nally 1 diagnostic laparoscopy, which was con- Regarding the postoperative outcomes, mortality verted to open right hemicolectomy. rate was 1%, with this single death occurring due to Overall, 35% of the admitted patients had pre- decompensated organ failure on the first postope- operatively a chest CT scan, with none sugge- rative day after emergency laparotomy for intesti- stive of typical COVID-19 related changes. In nal obstruction, with the patient having substantial the postoperative period, chest CT scans were cardiac comorbidities and frailty. It has to be noted requested for four patients, with only one demon- that the patient was not tested for COVID-19 either strating changes in consistency with COVID-19 with swab or CT chest, as this was not the protocol infection (the particular patient had an initially at the time of her admission (first week of March). negative swab for COVID-19 and during the hos- Our overall morbidity rate was 16%, with 11 (out pitalisation, was re-tested positive). of 16) patients who had non-respiratory postopera- With respect to swabs for RT-PCR, pre-ope- tive complications (Table 1). Concerning the five rative swabs were taken in 49% of emergency patients who developed respiratory postoperative admissions and all were reported as negative. complications, two (out of five) had swabs positive Postoperative swabs were taken in seven (7%) for COVID-19 in the post-operative period (ad- patients upon development of suspicious respi- missions swabs negative) and the one patient who ratory symptomatology, with only three (out of had a positive swab postoperatively was minimally seven) being positive (two patients with initially symptomatic from the respiratory aspect (no pre- negative swab re-tested positive and one patient operative swab taken) (Tables 2, 3). who was not tested with swabs on admission, su- No re-operation was required for any of the patients bsequently tested positive). and there was no unplanned re-admission to our in-

Table 2. Characteristics of the five patients with respiratory postoperative complications*

Overall ITU Type of compli- Primary procedure Management Swab for COVID Chest imaging LOS LOS cation (days) (days)

Incision and drainage of Planned non-inva- back abscess Respiratory Not tested pre/post-ope- sive ventilation on No essay performed 6 0 Known COPD patient / failure ratively ward & antibiotics NIV at home Pre-op swab negative, Pre-op CT chest: non-specific changes Laparoscopic Respiratory tract Antibiotics post-op swab after symp- Post-op CXR: B/L infiltrates, in consi- 22 0 cholecystectomy infection toms positive stency with COVID-19 Respiratory tract Antibiotics / Pre-op swab negative, Pre-op CXR: non-specific changes Laparoscopic infection / Pulmo- Therapeutic post-op swab after symp- Post-op CT chest: B/L infiltrates, in 26 0 cholecystectomy nary embolism anticoagulation toms positive consistency with COVID-19 Pre-op swab negative, Laparotomy& Hartmann’s Respiratory tract Pre-op CT chest: negative Antibiotics post-op swab after symp- 16 4 procedure infection Post-op CXR: non-specific changes toms negative

Laparotomy and repair of Respiratory tract Not tested pre/post-ope- Pre-op CXR: negative Antibiotics 18 2 perforated duodenal ulcer infection ratively Post-op CXR: non-specific changes

Mean overall LOS 17.6 days, mean ITU LOS 1.2 days LOS, length of stay; ITU, Intensive Therapy Unit; COPD, chronic obstructive pulmonary disease; CXR, chest X-ray; B/L, bilateral; NIV, non-invasive ventilation

277 Medicinski Glasnik, Volume 17, Number 2, August 2020

tensive care unit for ventilator support. Although Our unit during the 3-month study period pro- we had only 3% patients who were tested positive vided acute care surgery services in an unin- for COVID-19, these patients had a longer duration terrupted manner to our catchment area, with our of hospitalisation compared to the entire cohort as medical and intensive care units admitting and well as when compared to the subgroup of patients treating patients with suspicion or confirmation with postoperative morbidity (Tables 1, 2). of COVID-19 infection, in a similarly continuous service during the study period. Evaluating DISCUSSION postoperative morbidity and mortality outcomes The impact of COVID-19 pandemic has dra- after emergency gastrointestinal surgery during stically led to a substantial change of the priori- a 3-month observational period, we reported tisation with respect to elective general surgery absence of major respiratory complications des- procedures, mainly through application of local pite the COVID-19 pandemic outbreak. protocols allowing only the performance of ope- To the best of our knowledge, our study is the rations that need by default to be performed on first in the international literature to address the expedited basis, such as for cases of gastrointesti- overall prevalence of COVID-19 infection and nal malignancies. In the elective context though, its potential impact on the postoperative mor- since the procedure and perioperative care can be bidity and mortality in an unselected cohort of planned in advance, the patients can follow self- patients undergoing emergency general surgery isolation protocols and screening for COVID-19 operations over an extended time period. Based can be performed at convenient stage, allowing the on our experience, we advise acute care gastroin- surgical teams to amend the admission and care testinal surgeons that, whenever indicated, emer- plans if the patients test positive for COVID-19 gency surgery should not be delayed in favour and minimise associated postoperative risks (7). of non-operative management in anticipation of On the contrary, when it comes to emergency COVID-19-related morbidity or mortality risks. general surgery, even if the patients get screened for COVID-19 during their admission time, the- FUNDING se results usually get released 48-72 hours later, No specific funding was received for this study fact which obliges the acute care surgeons to make judgment calls about operative versus conserva- TRANSPARENCY DECLARATION tive treatment of these patients without evidence Conflicts of interest: None to declare. regarding their COVID-19 infection status (8).

REFERENCES 1. De Simone B, Chouillard E, Di Saverio S, Pagani Pietrabissa A, Asbun H, Pryor A. SAGES and EAES L, Sartelli M, Biffl WL, Coccolini F, Pieri A, Khan recommendations for minimally invasive surgery M, Borzellino G, Campanile FC, Ansaloni L, Ca- during COVID-19 pandemic. Surg Endosc 2020; tena F. Emergency surgery during the COVID-19 34:2327-31. pandemic: what you need to know for practice. Ann 5. Gao Y, Xi H, Chen L. Emergency surgery in suspec- R Coll Surg Engl 2020; 102:323-32. ted COVID-19 patients with acute abdomen: case 2. Wexner SD, Cortés-Guiral D, Gilshtein H, Kent I, series and perspectives. Ann Surg 2020; 272:e38- Reymond MA. COVID-19: impact on colorectal e39. surgery. Colorectal Dis 2020; 22:635-40. 6. COVIDSurg Collaborative. Mortality and pulmo- 3. Coimbra R, Edwards S, Kurihara H, Bass GA, Ba- nary complications in patients undergoing surgery logh ZJ, Tilsed J, Faccincani R, Carlucci M, Mar- with perioperative SARS-CoV-2 infection: an tínez Casas I, Gaarder C, Tabuenca A, Coimbra international cohort study. Lancet 2020; S0140- BC, Marzi I. European Society of Trauma and Emer- 6736(20)31350-7. gency Surgery (ESTES) recommendations for trau- 7. COVIDSurg Collaborative. Elective surgery cance- ma and emergency surgery preparation during times llations due to the COVID-19 pandemic: global pre- of COVID-19 infection. Eur J Trauma Emerg Surg dictive modelling to inform surgical recovery plans. 2020; 46:505-10. Br J Surg 2020; Epub ahead of print. 4. Francis N, Dort J, Cho E, Feldman L, Keller D, 8. Fransvea P, Di Grezia M, La Greca A, Cozza V, Lim R, Mikami D, Phillips E, Spaniolas K, Tsuda Sganga G. Are emergency surgical patients “Colla- S, Wasco K, Arulampalam T, Sheraz M, Morales S, teral Victims” of COVID-19 outbreak? Injury 2020; S0020-1383(20)30545-3. Epub ahead of print.

278 ORIGINAL ARTICLE

Model "P" in gender prediction based on the mastoid process

Aida Sarač-Hadžihalilović1, Edin Hojkurić2, Miralem Musić3, Ilvana Hasanbegović1, Zurifa Ajanović1, Lejla Dervišević1, Selmira Brkić4

1Department od Anatomy, School of Medicine, University of Sarajevo, Sarajevo, 2Community Health Center, Tešanj, 3Department of Pathophysiology, School of Medicine, University of Sarajevo, Sarajevo, 4Department of Pathophysiology, School of Medicine, University of Tuzla, Tuzla; Bosnia and Herzegovina

ABSTRACT

Aim To determine the degree of accuracy in determining the sex of the skull based on classical morphometric analysis of the mastoid process. Morphometric analysis excluded a subjective approach expressed in osteoscopic analysis.

Methods The study was conducted on a sample of 100 macerated skulls of known gender and age from the second half of the 20th century, including the Bosnian population. Of the 100 skulls, 50 (50%) were male and 50 (50%) were female. Male skulls were on average 60 (47-71) years old and female 57 (43-76) years old. At each mastoid process, 3 measurements were made: mastoid length, mastoid width (medio-lateral diameter) ML) and anteroposte- Corresponding author: rior diameter (AP) of the mastoid process. Aida Sarač-Hadžihalilović Department of Anatomy, Results Using the univariate method, we found a significant difference between length, width, AP diameter and size of mastoid pro- School of Medicine, University of Sarajevo cesses as well as between gender (p <0.05). Multivariate binary Čekaluša 90, 71 000 Sarajevo, logistic regression showed statistically significant differences in Bosnia and Herzegovina AP diameter of the mastoid process (p<0.05). Phone: +387 33 665 949; Fax: +387 33 203 670; Conclusion The created model "P" ("P"=exp [X]/1+exp [X]) for sex determination based on mastoid process showed sensitivity of E-mail: [email protected]; 82% correct prediction for female skulls and 65% accurate pre- [email protected] diction for male skulls. This discourse with respect of population ORCID ID: https://orcid.org/https://orcid.org/ standards grants most effective anthropological proof and as such may be suggested for forensic expertise based on human skull.

Kay words: differentiation, quantitative analysis, skull, sex Original submission: 19 February 2020; Revised submission: 20 March 2020; Accepted: 21 April 2020 doi: 10.17392/1145-20

Med Glas (Zenica) 2020; 17(2): 279-284

279 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION actual with no less significance than before, regardless of the wide use of DNA analyses (9). Skeletal sex determination is the process of deter- Even all univariately analysed diameters may mining whether a skeleton or parts of a skeleton show bigger or smaller statistic significance, just are from a male or from a female. The skeleton by using multivariate binary logistic regression is used because after death bones are preserved some of them stand out with predictive power for the longest period of time, and there are no creating a model for gender predilection (9). significant changes that would lead to erroneous findings (1). Almost all bones show some degree We decided to perform this osteometric study on of sexual dimorphism (2). For secure identifica- osteological material of the Department of Ana- tion of gender and other identity indicators it is tomy, School of Medicine, University of Sarajevo, ideal if there is an intact, complete skeleton (3). because we had data about gender and age for each However, due to a variety of circumstances, both skull, and this was crucial for our analysis and our natural and artificial, often only parts of the ske- predictive model. This discourse with respect to leton are found (4). The accuracy of gender deter- the population standards, which we have in our mination is highest in pelvic analysis, however, study, is a guarantee for the most effective anthro- pelvis itself is not always available for analysis pological proof and as such may be suggested for (5). The skull is therefore considered the second forensic expertise based on human skull. best option for sex determination (6) The aim of our study was to determine a degree The anatomical and morphological structures of of accuracy in determining the sex based on ma- the skull used for the purpose of sex determina- stoid process by using multivariate binary logi- tion are numerous: the frontal bone (position of stic regression. squamous part, the appearance of the supraciliary MATERIALS AND METHODS arch, the sharpness and shape of the orbit), the zygomatic bone (presence of marginal tubercle on Materials and study design the frontal process), the temporal bone (size and shape of the mastoid process, width of the zygo- The study was conducted at the Department of matic processes), the occipital bone and mandible Anatomy, School of Medicine, University of Sa- (angle between body and mandible ramus – angle rajevo, in the period February to June 2019. of mandible), shape of nasal root, muscular inser- The study was performed on a sample of 100 tions on bones, tooth size and face shape (7). Oste- macerated and degressed skulls of known gender oscopy and classical skull morphometric analysis and age from the second half of the 20th centu- can determine the sex with an accuracy of 92%, ry, including the Bosnian population, and which and if only the mandible is analysed the accuracy belong to an osteological collection of the De- is estimated at 90% (8,9). Over time, many studies partment of Anatomy, School of Medicine, Uni- have shown that gender can be determined based versity of Sarajevo. The average age of the skull on the human skeleton, especially by examining was 58.4 years. Of the 100 skulls, 50 (50%) were the pelvis and skull (9). Thus, different methods male with an average age of 60 (47-71) years, have been refined over time, and today, in addition while 50 (50%) were female with an average age to visual identification of sexual characteristics, of 57 (43-76) years. A sample of 100 whole hu- various univariate and multivariate statistical man skulls (50 males and 50 females) were ran- analyses are used, leading to discriminant functi- domly selected from a total sample of 211 skulls onal analyses (10,11). History of the development (139 males and 72 females) of skeleton-based sex determination shows its developmental dynamics from osteoscopic deter- Methods minations to osteometric ones, which additionally It was an osteometric study, where 3 diameters of inherits the application of sophisticated mathema- the mastoid process were measured on each skull tical-statistical methods (9). using a slider (Schubler; GPM Swiss Made) on The application of classical methodology both sides. The size of the mastoid process was approach, such as classical anthropometric asse- calculated by a given formula. ssment based on human skeleton is still very

280 Sarač-Hadžihalilović et al. Model "P" and gender prediction

Length of the mastoid process. The length of the Statistical analysis mastoid process was measured from the tip of the The collected data were analysed with descrip- external acoustic meatus (Porion) vertically down tive statistics by measures of central tendency to the tip of the mastoid process. The skull was (arithmetic mean and median) and measures of laterally positioned so that one side was always fa- variability (standard deviation, interquartile ran- cing the observer. The scale of the sliding calliper ge 25th, 75th percentile). Univariate binary lo- was laid behind the mastoid process, so that the gistic regression examined the individual influ- fixed part of the calliper was tangential with the ence of independent variables on the binary (0 upper edge of the external acoustic meatus. The or 1) dependent variable sex of the skull (male/ movable part of the sliding calliper was moved to female). Multivariate binary regression analysis the top of the mastoid process and the measure- examined the influence of independent predictors ment was read off from the scale of the slider (12). (model 1), which univariately showed a signifi- Width of the mastoid process (medio-lateral cant effect on the dependent variable “skull gen- diameter). The width was measured from the der”. The value of the model was tested by the highest part of the medial side within the fossa Hosmer and Lemeshow Test. digastrica to the highest laterally positioned point We defined a model for female prediction, using of the mastoid extension in the same plane (12). variable-predictors with statistical significance, Anteroposterior (A-P) diameter of the mastoid i.e. their standardized coefficients for calcula- process. It was measured from the lowest point, ting the constant of the equation: X = Constant where the tympanic part of the temporal bone (model) + B1Y1 + B1Y2 + BnYn, where B1-n contacts the anterior surface of the mastoid pro- is standardized coefficients for each independent cess to the posterior border of the mastoid proce- variable, and Y1-n is an independent variable ss in the same plane (12). with a statistically significant prediction to the Size of the mastoid process. The size of mastoid dependent variable sex of the skull. The model process was measured with the formula: is represented by the equation: Model “P” = exp (X) / (1+ exp (X). We got the model from logit or natural logarithm of chances (log odds): logit (p) = ln (p / 1-p). By exposing the logarithm i.e. by Before the measurement, the skull was placed in EXP (logit) we got the odds or odds e logit (p) = the “Frankfurt horizontal” position, that is, we EXP [ln (p / 1-p)]. put the line that connects the upper edge of ope- All analyses were estimated at the level of the ning of the external acoustic meatus (porion) and statistical significance of p <0.05. the lower edge of the left orbit (orbitale) in the horizontal position (12). RESULTS The slider (Schubler; GPM Swiss Made) was All measured parameters showed that there was used to measure smaller distances on the body. a statistically significant difference between the It comes in several different variants, such as a genders on the basis of the mastoid process. All Martin slider that has a 20 cm scale, while a no- parameters were higher in male skulls than in fe- nius slider has a range of 15 cm. In both cases male ones (Table 1). the scale was calibrated to 0.1 cm. The result is read on a line that coincides with the inner edge Table 1. Average values of measurement parameters of 100 of the moving part of the slider. It measures to an mastoid processes accuracy of 0.1 cm. Variable Mean (±SD) Min. Max. p Length of mastoid M 31.50 (±3.68) 25.00 43.00 With method Backward Wald, in step 1 of the 0.001 process (mm) F 29.70 (±3.15) 23.00 39.00 multivariete regression analyses all variables Width of mastoid M 12.75 (±1.90) 8.00 17.00 0.0001 that univariately showed statistical significance process (mm) F 11.78 (±2.02) 8.00 20.00 Anteroposterior M 15.34 (±2.22) 11.00 22.00 for gender prediction were included. In the next 0.0001 diameter (mm) F 12.94 (±2.18) 7.00 18.00 steps (2,3,4) one by one the variable that had the Size of mastoid M 62.73 (±19.67) 24.60 132.00 0.0001 weakest impact fell off. The last step (4) pointed process (mm3) F 46.68 (±17.88) 18.40 122.40 to a variable, which was the best discriminator. M, male; F, female; Min., minimum; Max., maximum

281 Medicinski Glasnik, Volume 17, Number 2, August 2020

The length, width and anteroposterior diameter probability in the model we used the given equ- of the mastoid process could distinguish a female ation (Table 2). from a male skull p <0.05 (Figure 1). “P” values exceeding 0.455 were the skulls of the female sex. Model P recognized 117 female mastoid processes, of which 82 were real female, positive predictive value (PPV)= 82/117 (70%). The model recognized 83 male mastoid processes, of which 65 were real male, negative predictive value (NPV) = 65/83 (78.3%). Of the 100 female mastoid processes, the model accurately identified 82, with a sensitivity of 82/100 (82%) (Table 3). Of the 100 male mastoid processes, the model accurately identified 65, with a sensitivity of 65/100 (65%) (Table 3).

Table 3. Sensitivity and specificity of predictive value of model “P” Mastoid process of Model “P“ known skull gender Total Figure 1. ROC curve of length, width and anteroposterior diam- Female Male eter of mastoid process, as a marker for distinguishing gender Gender probability based on Female 82 35 117 of the skull mastoid continuation Male 18 65 83 Total 100 100 200 Multivariate binary logistic regression used to test the influence of independent predictors of DISCUSSION the mastoid process (which univariately showed A study conducted in an Indian population significant influence) on skull gender differen- concluded that the mean length of the right tiation. By the 4 step Backward Wald method, mastoid process for male skulls was 34.82 mm the A-P diameter of the mastoid process had a (±4.4), and the mean length for the left mastoid statistically significant (p<0.05) effect on all va- process for the male skull was 35.64 mm (±2.87); riables (Wald = 36.45, with B = -0.515, model the mean length of the right mastoid process in constant of 7.274) (Table 2) female skulls was 24.36 mm (±3.6), and the Table 2. Common predictive values of mastoid process mean length of the right mastoid process in the parameters on skull gender differentiation female skull was 26.73 mm (±3.1), which corre- Model 1 B S.E. Wald p Exp(B) lates with our results (13). Length (mm) -.165 .113 2.122 .145 .848 Width (mm) -.502 .263 3.640 .056 .606 A study on North American skulls suggested that Step 1 A-P (mm) -.935 .254 13.510 .0001 .393 the size of the mastoid process may have more Size (mm3) .105 .056 3.559 .059 1.111 influence than the shape in gender classification Constant 18.723 6.530 8.221 .004 135343596.190 Width (mm) -.207 .169 1.499 .221 .813 in Caucasians of American origin. Most variati- A-P (mm) -.669 .173 14.936 .0001 .512 ons in mastoid processes were influenced by ma- Step 2 Size (mm3) .034 .028 1.400 .237 1.034 stoid size, explaining 87.3% of the total mastoid Constant 10.166 2.663 14.577 .0001 25993.440 Width (mm) -.040 .093 .191 .662 .960 variation (14). Step 3 A-P (mm) -.500 .092 29.843 .0001 .606 In a study on skulls of German origin, the use of Constant 7.556 1.383 29.841 .0001 1911.689 discriminant function did not exceed 65%. In the A-P (mm) -.515 .085 36.450 .0001 .597 Step 4 Constant 7.274 1.213 35.932 .0001 1442.396 German forensic sample, accuracy was generally B, standardized coefficient; S. E., standard error of B ; Wald, determined by the number of correctly identified unstandardized coefficient; Exp(B), odds ratio; A-P, anteroposterior diameter; male skulls (61% vs. 52% women) (15). Model "P" for prediction of female skull based A study on skulls of Malaysian origin (16) conclu- on A-P diameter of the mastoid process was des that the best parameter selected by the multi- performed: X = 7.274 - 0.515. To calculate the variate discriminant analysis is the mastoid trian-

282 Sarač-Hadžihalilović et al. Model "P" and gender prediction

gle parameter. Cross-validation accuracy for men, dy. The authors also stated that the length of the women, and combination was 82.3%, 88.5%, and mastoid process is the best discriminator, medio- 84.4%, respectively. Prediction accuracy in multi- lateral diameter is the second, and height is the variate discriminant function was based on asteri- third best discriminator in determining gender on-mastoid parameter and mastoid width, which from fragmented debris (19). This finding is in are considered the best parameters with 87% accu- contrast to our finding where the anteroposterior racy. It is in contrast to our study, but also to other diameter is the best discriminator in determining studies in the Asian population where mastoid len- gender, then width, and ultimately the height of gth was often found as the best parameter (16). the mastoid process. The discriminant function was performed on Based on our results, the following conclusions skulls of Indian origin including all variables and were drawn: accurately classified 76.4% of skulls (80% male The results of our study have shown that male and 73.3% female); by multivariate analysis, ma- skulls had on average greater values of length, stoid length was found to be the best indicator for width and anteroposterior diameter of the masto- gender differentiation, although the classification id process than female skulls, and higher values rate dropped from 76.7% to 66.7% (12). of the mastoid process than female skulls; incre- When analysing the discriminant function, Galda- asing the value of length, width, anteroposterior mes et al. (17) found that the group of analysed diameter, and size of the mastoid process incre- linear dimensions (Porion-Mastoid, Porion- ases the probability of classifying the skull as Asterion) represented low discriminant capacity male; the anteroposterior diameter of the mastoid (Lambda of Vilks = 0.960, canonical correlation = process stood out as statistically significant for 0.199). Porion-Mastoid was a variable that allowed the differentiation of skull sex. the classification of male groups from women with In conclusion, the created model "P" ("P" = exp an overall accuracy of 64.2% but with high sensiti- [X] / 1 + exp [X]) for gender determination based vity for correctly classifying men (93%) and very on the mastoid process of the skull is recommen- low sensitivity for women (17.7%). The results of ded for the use within the Bosnian populati- Galdames et al. match our results for males (65%), on, with the expectation of a greater degree of but not females (85%) (17). matching in females. We recommend that combi- Nagaoka et al. (18) showed the usefulness of the ned qualitative and quantitative anatomical-ant- mastoid process for gender assessment: gender hropological studies be performed, with respect classification accuracy was more than 80% with for the population standards in order to maximize only one variable and reaches 82-92% with a the prediction of sexual dimorphism. combination of height and width. In our study, the accuracy of gender estimates based on one ACKNOWLEDGMENTS variable is above 80% based on height and an- The authors would like to express their gratitude teroposterior diameter, and based on width it is to Gojak Refet, MD PhD, Clinical Centre of the 68%. Based on the discriminant function accor- University of Sarajevo, Bosnia and Herzegovina ding to our model, 82% of female skulls and 65% for help with statistical analysis. of male skulls were correctly classified. Sukre et al. (19) conducted a study where univa- FUNDING riate analysis showed that the mean of mastoid No specific funding was received for this study. variables such as mastoid length (25.32), mean lateral diameter (10.71), anteroposterior diame- TRANSPARENCY DECLARATION ter (21.60) was greater in males than in females Conflict of interest: None to declare. (p <0.005), which is in correlation with our stu-

283 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES 1. Stévant I, Papaioannou MD, Nef S. A brief history of 11. Ajanović Z, Sarač – Hadžihalilović A. Multyvariante sex determination. Mol Cell Endocrinol 2018; 468:3– analysis of cranioscopic and craniometric parameters 10. in gender determination of skulls. Health MED 2018; 2. Sunil P, Gupta S, Choudhary P, Verma KR. Sex deter- 12:75-84. mination using mastoid process measurements of dry 12. Sumati, Patnaik V, Phatak A. Determination of sex skull bone: a descriptive analysis. Int J Multispecia- from mastoid process by discriminant function lity Health 2016; 2:17-21. analysis. J Anat Soc India 2010; 59:222–8. 3. Durić M, Rakocević Z, Donić D. The reliability of sex 13. Mittal SK, Jaleswararao S, Goyal J, Mittal L, Goyal determination of skeletons from forensic context in G. Sex determination using mastoid process of dry the Balkans. Forensic Sci Int 2005; 147:159-64. skull. IJBAMR 2018; 7:404-8. 4. Bidmos MA, Gibbon VE, Štrkalj G. Recent advances 14. Jung H, Woo EJ. Evaluation of mastoid process as sex in sex identification of human skeletal remains in indicator in modern white Americans using geometric South Africa. S Afr J Sci 2010; 106:1-6. morphometrics. J Forensic Sci 2016; 61:1029–33. 5. Walker, Phillip L. Greater sciatic notch morphology: 15. Kemkes A, Göbel T. Metric assessment of the “Ma- sex, age, and population differences. Am J Phys Ant- stoid triangle” for sex determination: A validation hropol 2005; 127:385-91. study J Forensic Sci 2006; 51:985–9. 6. Gupta AD, Banerjee A, Kumar A, Rao SR, Jose J. 16. Ibrahim A, Alias A, Shafie MS, Das S, Nor FM. Oste- Discriminant function analysis of mastoid measure- ometric estimation of sex from mastoid triangle in ments in sex determination. Life Sci 2012; 4:1–5. Malaysian population. Asian J Pharm Clin Res 2018; 7. Bubuc I. Sexsualne oznake na lobanji čovjeka (Sex- 11:303-7. ual marks on a human skull) [Bosniana] Folia Ana- 17. Suazo Galdames IC, Zavando Matamala DA, Smith tomica 1973; 2:17-28. RL. Sex determination using mastoid process mea- 8. Kozerska M, Skrzat J, Sczepanek A. Application of surements in Brazilian skulls. Int J Moprhol 2008; the temporal bone for sex determination from the ske- 26:941-4. letal remains. Folia Med Cracov 2015; 55:33-9. 18. Nagaoka T, Shizushima A, Sawada J, Tomo S, Ho- 9. Sarač – Hadžihalilović A. Anatomically – Anthro- shino K, Sato H, Kazuaki H. Sex determination using pological Significance of the Skull. English edition mastoid process measurements: standards for Japane- (faculty textbook). Sarajevo: University of Sarajevo, se human skeletons of the medieval and early modern 2017. periods. Anthropol Sci 2008; 116:13-105. 10. Ajanović Z, Sarač – Hadžihalilović A, Gojak R. De- 19. Sukre SB, Chavan PR, Shewale SN. Morphometric termination of sex by discriminant function analysis analysis of mastoid process for sex determination of linear diameters in Bosnian human skulls. In: Bad- among Marathwada population. MIJOANT 2017; njevic A (ed.) CMBEBIH 2017. IFMBE Proceedings 1:27-32. 2017; 62:88-94, Singapore: Springer.

284 ORIGINAL ARTICLE

The preemptive effect of tramadol and metamizole on the intensity of postoperative pain

Ismet Suljević1, Muamer Hadžiavdić2, Ismana Šurković3, Omer Suljević4, Maida Turan5, Ehlimana Mušija6

1Clinic for Anaesthesia and Resuscitation, 2Clinic for Bone Surgery, 3Department for Endocrinology, Diabetes, and Diseases of Metabo- lism; Clinical Centre of University of Sarajevo, 4Sarajevo Health Centre, 5Healthcare Group Acibadem, Representative Office in Sarajevo, 6Clinic for Heart Diseases and Rheumatism, Clinical Centre of University of Sarajevo; Sarajevo, Bosnia and Herzegovina

ABSTRACT

Aim To demonstrate the analgesic effect of preemptively administered tramadol and metamizole on the postoperative pain severity, after an elective operative hysterectomy with adnexectomy.

Methods There were three groups with 30 patients in each group. Patients included in the study were between 45 to 67 years old. They were all in the ASA group II. Randomization was performed in random order according to the regular elective operating program. Patients in Group I received i. m. tramadol 1mg/kg, and in Group II 30mg/kg of metamizole, five minutes before anaesthesia induction. Patients did not receive preemptive analgesia in Group III (control). All patients underwent the same induction anaesthesiology procedure with propofol, fentanyl, tracrium, supplemented Corresponding author: with O2, N2O, and sevoflurane at an appropriate dose until MAC Ismet Suljević 1 was reached. Surgeries lasted for 80-120 minutes. Every patient Clinic for Anaesthesiology and performed a resting pain assessment 30 minutes after an extubati- Resuscitation, on by Numerical Pain Scale (NPS). Clinical Centre of University of Sarajevo, Results We found out that tramadol had a better effect in preemp- Bolnička 25, 71000 Sarajevo, tive analgesia and that the average pain score for Group I was 6.10 Bosnia and Herzegovina (p=0.043). In Group II, it was 7.93 (p=0.022). There is significant Phone: +387 33 297 148; difference in pain intensity between patients in the control group, +387 33 297 841; (pain intensity was 9.16), and those who received tramadol and E-mail: [email protected] metamizole. There was no significant difference in the intensity of pain when using these two analgesics (p=0.733). ORCID ID: https://orcid.org/0000-0002- 4357-6654 Conclusion The effect of preemptively administered tramadol pri- or to the introduction of general anaesthesia in postoperative pain Original submission: is significantly more favourable than the effect of metamizole. 30 December 2019; Key words: analgesia, general anaesthesia, hysterectomy, pain Revised submission: assessment 30 January 2020; Accepted: 14 April 2020 doi: 10.17392/1118-20

Med Glas (Zenica) 2020; 17(2): 285-289

285 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION sts) group II (8), 45-67 years old, with a body mass index (BMI) of 18.5-24.9. Exclusion criteria were: By the definition of the International Associati- a BMI below 18.5 or above normal body weight on for Study of Pain (IASP), pain is described as (24.9), a history of allergic manifestations to one “unwanted sensory, motor, and emotional experi- of the analgesics used, and the patients whose sur- ence connected with an actual or threatening tissue gery lasted longer than 120 minutes. Randomiza- damage, or described from patient perspective tion was performed consecutively according to the with terms that indicates that kind of damage” (1). regular elective operative program. It is an obligation for every doctor in the Intensive Care Unit (ICU) to eradicate pain, or to lower its The patients were explained the procedure for intensity to a bearable level. Taking into considera- participation in the study and their consent was tion that pain is an individual feeling, it is difficult obtained. All patients agreed to the trial with the to estimate it, and the commonly used option is to prior knowledge of the study design and the po- ask a patient to self-asses the pain level by using ssibility of withdrawing from the study whenever the pain scales which are used in every day prac- they wished. Post-operative pain assessment of tice (2). Mostly used scale for pain assessment of patients was performed 30 minutes after extuba- older patients is a numerical pain scale (NPS) (3). ting and being placed in the Intensive Care at bed These days analgesic offer gives us a wide range rest. We did not determine the intensity of pain in in cupping postoperative nociceptive pain. In that the movement because our patients were encou- sense there are many guides that are directing the raged to move 2 hours after surgery. schemes in postoperative treatment of pain, all to An approval of the Scientific Committee of the the extent of preemptive or preoperative applicati- Clinical Centre in Sarajevo was obtained for this on of analgesics (4). In many studies it was proven research. that administration of analgesics before anaesthe- Methods sia induction allowed more favourable postoperative analgesic score (5). In preemptive approach There were 30 patients in the Group I who re- to postoperative analgesia, it is possible to use ceived, 1 mg/kg tramadol i. m., 5 minutes before different analgesics and techniques, but in many induction of anaesthesia. In the Group II patients studies tramadol was confirmed as a primary and received i. m. metamizole at a dose of 30 mg/kg, 5 most adequate analgesic (6,7). minutes before the induction of anaesthesia. In the Because we often use metamizole and tramadol Group III (control group) there were 30 patients in our clinical practice, we wanted to compare who underwent the same operative procedure but their preemptive effect on pain in our patients. without preemptively administered analgesia. The aim of this study was to investigate the effect All patients underwent the same induction ana- of tramadol and metamizole in preemptive appli- esthesiology procedure with pre-induction with

cation in terms of decreasing postoperative pain spontaneous ventilation of 100% O2 of 5 L/min, after hysterectomy and adnexectomy. and i. v. application of 1% propofol 2mg/kg, fentanyl 3 µg/kg, and atracurium besilate 0.5 mg/kg. PATIENTS AND METHOD After intubation, anaesthesia was managed with

supplementation of O2, N2O (50/50%), sevoflurane Patients and study design in the appropriate volume percentage for the achie- This prospective, randomized, comparative, blin- ved minimum alveolar concentration (MAC 1). ded (the patients did not know in which group The operative course went neatly in all patients. they would be) study was conducted at the Clinic All surgeries took 80-120 minutes. After awake- for Gynaecology and Obstetrics at the Clinical ning from anaesthesia and intensive care place- Centre in Sarajevo (tertiary-level hospital serves ment that happened within 30 minutes of extu- a population of half a million) over a six-month bation, all patients underwent pain evaluation period. It included three groups of patients, 30 in and were instructed to report pain on the 11-point each group, who were prepared for elective opera- numerical pain scale (NPS) with zero equalling tive hysterectomy with adnexectomy. The patients “no pain” and 10 indicating “the worst imagina- were ASA (American Society of Anaesthesiologi- ble pain“ (6).

286 Suljević et. al. Preemptive effect of tramadol and metamizole

Statistical analysis Table 1. Number of patients, average age and duration of surgery according to the patient groups Data were entered into previously prepared ta- Patient No of Pain Average age Average surgery bles with patients initials, age, length of surgery, group patients scores (years) duration (min.) and pain intensity recorded. The obtained results I (n=30) 7 5 58.5 110.1 were statistically analysed with ANOVA analysis 14 6 55.4 105.0 8 7 55.5 99.7 of variance of small samples (to test the equality 1 8 67.0 95.0 of multiple groups), Levene's test (to analyse va- 30 6.10 57.9 102.4 riance equality) and Student t- test (the confiden- II (n=30) 9 7 61.7 103.5 15 8 58.5 102.4 ce intervals and hypotheses about the equality/ 5 9 56.4 99.4 difference of arithmetic means of the compared 1 10 66.0 110.0 groups) were applied. Statistical significance was 30 7.93 59.4 103.8 III (n=30) 3 7 59.0 95.6 set up at p<0.05. 5 8 58.0 98.8 6 9 53.5 96.3 RESULTS 16 10 58.0 104.1 30 9.17 57.1 98.7 The average pain score in the Group I where tra- Group I, Tramadol; Group II, Metamizole; Group III, control madol was administered was 6.10. In the Group I 14 patients reported feeling pain with a score of (p=0.635) and metamizole (p=0.154) group com- 6. In the Group II with metamizole administrati- paring to the control group, as well as between on, the average pain score was 7.93; the highest the tramadol and metamizole groups (p=0.326). number of pain ratings being given number 8 There was no statistically significant difference occurred in 15 patients. In the control Group III in the length of the surgery when comparing the the average pain score was 9.17; a score of 10, duration of the tramadol (p=0.155) and meta- which reflects the most severe pain, was obtained mizole (p=0.722) group with the control group, in 16 patients of this group. There was a signifi- as well as between the tramadol and metamizole cant difference of the average values of pain in- groups (p=0.269). tensity in the individual groups (p<0.05). Levene's test for three groups showed equal By comparing the control group and the group (p=0.057), which is bigger than 0.05 (evaluati- which received tramadol preemptively, it was fo- on mistake); therefore, the variances of these und that the average pain score in tramadol group three samples did not differ statistically (if the was lower (<9.17 and 6.10, respectively). These Levene's test is significant, than the variances two groups did not have the same variance based are not equal). It was confirmed that variability on the Levene's test (p=0.043). between the groups was bigger than variability The average pain score of patients who received inside the groups (Table 1). metamizole was lower than the pain of patients in Based on the number of the patients included in the control group (7.93 and <9.17, respectively). the study, we calculated the possibility of 99% These two groups did not have the same variance, that the study will reveal the difference in the tre- based on the Lavene's test (p=0.022) (Table 1). atment on double-sided level of p=0.05 signifi- The pain intensity score in the patients who re- cance, considering that the real difference betwe- ceived metamizole or tramadol showed the same en the treatments was 1.833 times bigger than the variance with 95% significance (based on the standard deviation. Levene’s test) (p=0.733). It is also evident that DISCUSSION there was no statistically significant difference in pain intensity when using these two analgesics. Our study investigated the effect of metamizole The average age in the Group I was 57.9 years, and tramadol in preemptive analgesia in patients in the Group II 59.4 years and in the Group III from Sarajevo Canton. Patients received analgesics 57.1 years. The average duration of operations via intramuscular (i.m.) application on the opera- in the Group I was 102.4 min, in the Group II- ting table before the induction into anaesthesia. All 103.8 min, and in the Group III-98.7 min (Table patients underwent the same anaesthesia induction 1). There was no statistically significant diffe- procedure and had the same balanced anaesthesia. rence in the age of the patients in the tramadol The well-known fact is that movement of patients

287 Medicinski Glasnik, Volume 17, Number 2, August 2020

immediately postoperatively enhances the feeling postoperatively, as well as a lower need for extra of pain due to tightening of the abdominal muscles, dosage of analgesics postoperatively (13). so that was not our interest at that moment (8). We Also, in the study in which the preemptive diclo- did not perform additional measurements of pain fenac and metamizole were used in the gynae- intensity (12h or 24h) (8) because our goal was to cological operations of 108 patients divided into determine the effect of a single preemptively given two groups, analgesia was given immediately af- dose of tested analgesic, and not the prolonged ter the induction to general anaesthesia showing analgesic effect of the same analgesic, as well as the positive preemptive analgesic effect of diclo- the effect of repeated doses that are moving out of fenac and metamizole, with precedence of diclo- the scope of preemptive administration. Since the fenac effects, but without significant difference research was in accordance with the methodology 6 hours after the application of the analgesic foreseen, we did not need to develop a separate (p=0.13) (14). flow chart of the study. The NPS scale was used, Preemptive use of analgesics in surgery is a according to which all patients evaluated their known method of analgesia and it is used in va- pain. In the control group, postoperative pain was rious fields of surgery (15). Analgesics can be the strongest. In the patients who received meta- administered preemptively in a variety of ways mizole we found out that the effect of preemptive (15). Oral administration of analgesics also pro- analgesia was significant in 15 of them who eva- duces good results in reducing pain. A study in luated pain with a score of 8. Patients receiving which gabapentin and tramadol were used orally, preemptive tramadol showed significantly lower showed a satisfactory postoperative effect (16). intensity pain. Tramadol is a more pharmacologi- Application of i. v. dexketoprofen trametamol cally potent analgesic in the opioid group, which is also a favourable option for preemptive pain in our patients also showed a significant analgesic therapy in dental surgery (17). Some studies fa- effect in preemptive analgesia (9). The effects of vour the preventive multimodal approach to the tramadol are not well known. Its pharmacological treatment of postoperative pain over preemptive effect is achieved by different actions at multi- analgesia (18). In a study where tramadol was ple receptors; it stimulates „mu“ opioid receptors administered subcutaneously preemptively, the and releases serotonin (10). Tramadol also blocks need for opioids was decreased postoperatively NMDA, muscarinic and nicotinic acetylcholine (19). In the study by Sittl et al. the success of receptors (10). Its specific analgesic effect is also preemptive analgesia was shown in postoperati- confirmed by the partial antagonistic effect of na- ve period, but to prevent chronic pain, prevention loxone, which only partially eliminates the effect is proposed on a multimodal basis (20). Wang of tramadol (11). Unlike tramadol, metamizole is et al. in a study of preemptive i.v. application of a long-used analgesic for various types of pain, 100mg tramadol with the application of a bolus including postoperative pain (11). The mechanism dose of 0.5mg/kg of tramadol immediately after of action of metamizole is based on the inhibition surgery with a continuous infusion of 0.1mg/kg/h of prostaglandin synthesis in peripheral tissues and of tramadol for 48h period with a PCA applicati- in the central nervous system (CNS) (12). Its anal- on of morphine 0.02mg/kg, found a significantly gesic effect is rather weaker than the one of trama- more beneficial effect in the reduction of pain, dol (12), as shown by our study. reduction of side effects and patient satisfaction, We could not find the same design studies online, than in the placebo control group (21). In the stu- and therefore we compared similar studies which dy of Lu et al. a significantly better effect was fo- used metamizole and tramadol separately in com- und in the flurbiprofen axetil group comparing to parison with other analgesics in preemptive anal- the tramadol group (22). A study that investigated gesia in hysterectomy (12) or similar or other ope- the effect of i. v. administered ketorolac and me- rative procedures (12). In a study that covered 60 tamizole preemptively in children under various patients for elective operations, tramadol i. v. 2mg/ operations with inhaled anaesthesia, showed a sa- kg and paracetamol 15mg i. v were used 15 minu- tisfactory analgesic effect (23). A study analysing tes before the induction to general anaesthesia; si- the preemptive effect of ibuprofen and metamizo- gnificantly better pain reduction was displayed in le on diabetic retinopathy laser surgery showed a the paracetamol group, than in the tramadol group satisfactory effect of both analgesics during sur-

288 Suljević et. al. Preemptive effect of tramadol and metamizole

gery (24). Many studies confirm the preemptive postoperative pain. The effect of tramadol admi- analgesic effect of metamizole and tramadol, as nistered preemptively prior to the introduction of well as other analgesics in the postoperative pe- general anaesthesia in immediate postoperative riod in terms of reduced opioid use postoperati- analgesia is significantly more favourable than vely. These studies should be continued in order the effect of metamizole. to find out which analgesics, individually or in FUNDING combination, produce the best results in terms of analgesia and in terms of reducing side effects. No specific funding was received for this study. In conclusion, preemptive administration of TRANSPARENCY DECLARANTIONS analgesics has a beneficial analgesic effect on Competing interests: None to declare. REFERENCES 1. Aydede M. Does the IASP definition of pain need 13. Agrawal A, Panditrao MM, Joshi S. To compare intra- updating? Pain Rep 2019 4:e777. venous paracetamol with tramadol given pre-empti- 2. Haefeli M, Elfering A. Pain assessment. Eur Spine J vely for intraoperative and postoperative analgesia- a 2006; 15(Suppl 1): S17–S24. randomized controlled trial. Indian Journal of Clinical 3. Metry AA, Wahba RM, Nakhla GM, Abdelmalek FA, Anaesthesia 2018; 5:9-12. Ragaei MZ, Fahmy NG. Comparative study between 14. Paz-Estrada C, Céspedes-Cuenca Y. Diclofenac ver- preemptive intra-articular injection of levobupivaca- sus metamizol in preventive analgesia for gynecologi- ine and tramadol for control of postoperative pain. cal surgery Rev Mex Anest 2009; 32:209-13. Anesth Essays Res 2019; 13:84-90. 15. Dahl J.B, Moiniche S. Pre-emptive analgesia. British 4. Solmaz FA, Kovalak E. Comparison of tramadol/ace- Medical Bulletin 2004; 71:13-27 taminophen fixed-dose combination, tramadol, and 16. Farzi F, Naderi Nabi B, Mirmansouri A, Fakoor acetaminophen in patients undergoing ambulatory F, Atrkar Roshan Z, Biazar G, Zarei T. Postoperati- arthroscopic meniscectomy. Acta Orthop Traumatol ve pain after abdominal hysterectomy: a randomized, Turc 2018; 52:222-5. double-blind, controlled trial comparing the effects 5. Xu Z, Zhang H, Luo J, Zhou A, Zhang J. Preemptive of tramadol and gabapentin as premedication. Anesth analgesia by using celecoxib combined with trama- Pain Med 2016; 6:e32360. dol/APAP alleviates post-operative pain of patients 17. Cağıran E, Eyigör C, Sezer B, Uyar M. Preemptive undergoing total knee arthroplasty. Phys Sportsmed analgesic efficacy of dexketoprofen trometamol on 2017; 45:316-22. impacted third molar surgery. Agri 2014; 26:29-33. 6. Mathew P, Aggarwal N, Kumari K, Gupta A, Panda 18. Pogatzki-Zahn EM, Zahn PK. From preemptive to N, Bagga R. Quality of recovery and analgesia after preventive analgesia. Curr Opin Anaesthesiol 2006; total abdominal hysterectomy under general anesthe- 19:551-5. sia: a randomized controlled trial of TAP block vs 19. Jabalameli M, Hazegh P, Talakoub R. Preemptive epidural analgesia vs parenteral medications. J Ana- subcutaneous tramadol for post-operative pain in esth Clin Pharmac 2019; 35:170-5. lower abdomen surgeries: a randomized double blin- 7. Farzi F, Naderi Nabi B, Mirmansouri A, Fakoor ded placebo-control study. Adv Biomed Res 2013; F, Atrkar Roshan Z, Biazar G, Zarei T. Postoperati- 2:68. ve pain after abdominal hysterectomy: a randomized, 20. Sittl R, Irnich D, Lang PM. Update on preemptive double-blind, controlled trial comparing the effects analgesia: options and limits of preoperative pain the- of tramadol and gabapentin as premedication. Anesth rapy. Anaesthesist 2013; 62:789-96. Pain Med 2016; 6:e32360. 21. Wang F, Shen X, Xu S, Liu Y. Preoperative tramadol 8. Mayhew D, Mendonca V, Murthy BVS. A review of combined with postoperative small-dose tramadol in- ASA physical status – historical perspectives and mo- fusion after total abdominal hysterectomy: a double- dern developments. Anaesthesia 2019; 74:373-9. blind, randomized, controlled trial. Pharmacol Rep 9. Kocabas S, Karaman S, Uysallar E, Firat V. The use 2009; 61:1198-205. of tramadol and morphine for pain relief after abdom- 22. Lu JC, Zhang XF, Liu C. Effect of preemptive anal- inal hysterectomy. Clin Exp Obstet Gynecol 2005; gesia with flurbiprofen axetil on patient-controlled in- 32:45-8. travenous analgesia with tramadol in patients under- 10. Stoops WW, Lofwall MR, Nuzzo PA, Craig LB, Sie- going postburn plastic surgery. Nan Fang Yi Ke Da gel AJ, Walsh SL. Pharmacodynamic profile of trama- Xue Xue Bao 2009; 29:1255-6. dol in humans: influence of naltrexone pretreatment. 23. Peñuelas-Acuña J1, Oriol-López SA, Hernández- Psychopharmacology (Berl) 2012; 223:427-38. Bernal CE, Castelazo Arredondo JA. Ketorolac vs 11. Arbel R, Stanleigh J, Ioscovich A. Pain management metamizol preemptive analgesia in children. Cir Cir following abdominal hysterectomy: novel approaches 2003; 71:50-4. and review of the literature. J Clin Gynecol Obstet 24. Macaferri Del Santo A, Maluf Auge R, Amaral 2013; 2:51-5. Ferra C. Preemptive analgesia of metamizole versus 12. Lirk P, Thiry J, Bonnet MP, Joshi GP, Bonnet F; ibuprofen in retinal laser photocoagulation. Rev Bras PROSPECT Working Group. Pain management after Oftalmol 2016; 75: 14-7. laparoscopic hysterectomy: systematic review of literature and PROSPECT recommendations. Reg Ane- sth Pain Med 2019; 44:425-36.

289 ORIGINAL ARTICLE

Characteristics of autologous peripheral blood stem cells collection over a one-year period

Gorana Ahmetović-Karić1, Elma Ćatović-Baralija1, Alma Sofo-Hafizović2

1Department for Hemapheresis with Blood Bank, Blood Transfusion Institute of the Federation of Bosnia and Herzegovina, 2Clinic for Haematology, University Clinical Centre; Sarajevo, Bosnia and Herzegovina

ABSTRACT

Aim To present characteristics of collecting autologous peripheral blood stem cells over a one-year period with an emphasis on efficiency and safety.

Methods A retrograde analysis of 24 leukapheresis in 20 adult patients with malignant haematological diseases in the Blood Tran- sfusion Institute of the Federation of Bosnia and Herzegovina in Sarajevo, was done. Cell separators Amicus and Spectra Optia were used for collection procedures.

Results The patient’s age ranged from 27 to 65 years. Target cells were collected in one procedure in sixteen patients, while in four patients they were collected in two procedures. The mean Corresponding author: CD34+ collection efficiency was 57.7%. The median number of Gorana Ahmetović-Karić CD34+ cells and percentage of CD34+ cells in the products were Department for Hemapheresis with Blood 5.52x10e6/kg (range 3.28-9.00) and 1.57% (range 0.96-2.91). Bank, Blood Transfusion Institute of the A strong positive correlation was found between the number of CD34+ cells in peripheral blood on the apheresis day and the Federation of Bosnia and Herzegovina amount of CD34+ cells collected in the products (rs=0.73). A total Čekaluša 86, Sarajevo, of 95% of patients collected the amount of ≥3x10e6/kg and 55% Bosnia and Herzegovina of ≥5x10e6/kg CD34+ cells for a single transplant. A decrease in Phone: +387 33 567 327; platelet count, haemoglobin and haematocrit values after the pro- Fax: +387 33 567 333; cedure was not significant. Potassium decrease showed statistical E-mail: [email protected] significance (p<0.000). Adverse events occurred during one procedure (4.2%). ORCID ID: https://orcid.org/0000-0001- 9736-6196 Conclusion A low number of adverse events and good collection efficiency with adequate patient monitoring, indicate that leukapheresis is a safe procedure that is successfully used in the Original submission: autologous transplantation process in the treatment of malignant 20 March 2020; haematological diseases. Revised submission: Key words: CD34, hematopoietic stem cells, leukapheresis 04 May 2020; Accepted: 28 May 2020 doi: 10.17392/1171-20

Med Glas (Zenica) 2020; 17(2): 290-296

290 Ahmetović-Karić et al. Peripheral blood stem cell collection

INTRODUCTION our study, the focus was on the collection procedure, as well as on the patient during that proce- In the last twenty years, peripheral blood stem dure and on the parameters resulting from it, so it cells (PBSCs) collection by apheresis, as an seems important sharing experience with perfor- efficient and safe method, has replaced the bone ming this kind of the procedure in the autologous marrow as a source for autologous transplanta- transplantation process to the local and general tion in the treatment of patients with malignant professional community. haematological diseases (1-4). The PBSCs transplantation has been used as a treatment in The aim of this study is to present the charac- different haematological and non-haematolo- teristics of collecting autologous PBSCs over a gical diseases (3). Peripheral blood cells in the one-year period, with an emphasis on efficiency cell separator used for collection are stratified by and safety. differential centrifugation with respect to specific PATIENTS AND METHODS density, with consequent separation and collection of needed mononuclear cells (MNC) and the Patients and study design return of other blood cells to the patient (2). The PBSCs are present in a population of blood A retrospective analysis of 24 leukapheresis pro- cells that express the CD34 marker on the sur- cedures for the collection of autologous PBSCs face membrane, located within a mononuclear in 20 adult patients at the Blood Transfusion In- cells layer (1). stitute of the Federation of Bosnia and Herzego- vina in Sarajevo was done. Data over a one-year During the collection, a small number of other period, from December 2017 to December 2018 cells are also collected at the same time. Apheresis were analysed. Those were the patients from the procedure can be accompanied by certain symp- Clinic for Haematology, University Clinical Cen- toms related to withdrawal of blood into cell se- tre in Sarajevo with malignant haematological parator, vascular access or metabolic changes (2). diseases who are involved in the process of auto- Normally, a small number of PBSCs are found logous transplantation. During that process they in peripheral blood and their mobilization sti- undergo apheresis procedures for the collection mulates their exit from the bone marrow into of PBSCs and their cryopreservation and storage the peripheral blood (3). Common mobilization at the Department for Haemapheresis with Blood strategies include the use of granulocyte stimula- Bank at the Blood Transfusion Institute of the Fe- ting growth factor alone or in combination with deration of Bosnia and Herzegovina in Sarajevo. chemotherapy (3,4). An increase in the number There were eight (40%) male and 12 (60%) fe- of cells is greater with their combination (4). male patients, whose age ranged from 27 to 65 After the collection, autologous PBSCs undergo years (mean 48), body weight ranged from 45 to cryopreservation and adequate storage up to the 115 kg (mean 74) and body height ranged from time of reinfusion in the transplant process (5). 155 to 187 cm (mean 170). Transplantation of autologous PBSCs is a met- Patients with completed medical documentati- hod of treatment which provides reconstitution of on were taken for analysis. All patients signed a hematopoietic system of the patients whose dise- statement and a consent for the entire process of ase is primary eradicated by myeloablative the- autologous peripheral hematopoietic stem cell rapy with or without irradiation, and irreversible transplantation with the consent of a member of bone marrow damage is overcome by infusion the Ethics Committee of University Clinical Cen- of previously collected and PBSCs preserved by tre in Sarajevo that the ethical principles were met. freezing (1,4). Sufficient amount of hematopoie- All the analyses covered by this study are stan- tic stem cells is an important predictor for succe- dard analysis performed during the whole proce- ssful transplantation (3). dure of the collection of hematopoietic stem cells Studies on this topic in our country are insuffi- from peripheral blood in the process of autologous cient. There is a study assessing influence of transplantation. The research was approved by the peripheral CD34+ cells number on collected Expert Council of the Blood Transfusion Institute CD34+ cells number in one day collection (6). In of the Federation of Bosnia and Herzegovina.

291 Medicinski Glasnik, Volume 17, Number 2, August 2020

Methods cells counting started when peripheral blood leu- The mobilization process included the use of kocyte counts were >1x10e9/L. The leukaphe- chemotherapy followed by daily stimulation with resis process was initiated when the CD34+ cell granulocyte-stimulating growth factor (G-CSF) count in peripheral blood on the apheresis day - biosimilar filgrastim (Zarzio, Sandoz, GmbH). was >15x10e6/L and platelet count >30x10e9/L. Mobilization strategies for PBSC for myeloma All patients had an inserted central venous catheter patients were performed with cyclophosphamide to sustain an adequate inlet flow rate during aphe- 2-4 gr/m2 for two days with the addition of G- resis. If some problem occurred with central veno- CSF 5-10 mg/m2 starting from the third day until us catheter, peripheral apheresis/dialysis needles leukapheresis, during the first complete remissi- (17 GA, Fresenius Medical Care, Bad Homburg, on or very good partial remission, mostly after Germany) were used in peripheral veins. 4-5 prior chemotherapy cycles. In some myelo- Leukapheresis for PBSC collection was perfor- ma patients cyclophosphamide dose was adjusted med with a continuous flow blood cell separators due to renal impairment. Myeloma patients were - Amicus (Fresenius Kabi, Bad Homburg, Ger- planned for two transplantations. many; version 4.4) and Spectra Optia (Terumo In Hodgkin’s lymphoma patients, mobilization stra- BCT, Lakewood, CO, USA; version 7, 11) with tegies included dexamethasone, cytarabine, cispla- MNCs collection protocol. Extracorporeal blood tin (DHAP) with G-CSF 5-10 mg/m2 starting from volumes were 160 mL and 191 mL. Patients were the fifth day until leukapheresis, during the first connected to the cell separator by their centrally relapse and in one patient with refractory disease. or peripherally inserted venous catheter. One line Mobilization strategies for non-Hodgkin’s lymp- was used to withdraw blood out of the patient and homa patients included DHAP, cyclophosphami- into the separator, and the other line was used de, doxorubicin, vincristine, prednisone (CHOP), to return processed blood back to the patient. cyclophosphamide, doxorubicin, vincristine, Collections can occur on a daily basis until target etoposide, prednisone (CHOEP) and etoposi- CD34+ cells levels are achieved. For extracorpo- de, prednisone, vincristine, cyclophosphamide, real anticoagulation anticoagulant citrate dextro- doxorubicin (DA-EPOCH) with G-CSF 5-10 se solution formula A (ACD-A) was used at the mg/m2, also starting from the fifth day until le- ratio of 12:1 (whole blood: anticoagulant). Daily ukapheresis. Five mantle cell lymphoma pati- PBSCs collections were performed in order to ents were mobilized with one of the therapeutic reach a yield of ≥3x10e6/kg or an optimal yield cycles, when bone marrow was without lympho- of ≥5x10e6/kg CD34+ cells in the leukapheresis ma cells infiltration. Three diffuse large B cell product required for a single transplant. Periphe- lymphoma patients with high International Pro- ral blood electrolyte values (Na, K, Ca, Cl) were gnostic Index (IPI) score were mobilized during determined before and after leukapheresis. All the first line therapy and one anaplastic lympho- patients received prophylactic intravenous calci- ma kinase (ALK)-positive patient was mobilized um infusion before leukapheresis. Patient’s blo- during therapeutic cycles when bone marrow od pressure and heart rate were monitored during was without lymphoma cells infiltration. Prior procedures and all adverse events were recorded. chemotherapy cycles in non-Hodgkin’s lympho- Right after the collection, cell products were fro- ma patients included CHOEP, R-CHOP (rituxi- zen using the controlled rate freezer (Consarc- mab + CHOP)/DHAP, R-CHOEP (rituximab + tic BIOFREEZE BV45 Freezer, Schollkrippen, CHOEP)/DHAP, DA-EPOCH, R-DA-EPOCH GmbH). Cryopreserving solution contained 10% (rituximab + DA-EPOCH), while in Hodgkin’s dimethyl sulfoxide (DMSO) in autologous pla- lymphoma patients DHAP and bleomycin, eto- sma. A number of bags used for cryopreservation poside, doxorubicin, cyclophosphamide, vincri- (CryoMACS Freezing Bag 500, 750, Miltenyi stine, procarbazine, prednisone (BEACOPP). Biotec, GmbH) depended on the product volume Complete blood cell count was evaluated daily and cell numbers. Cryopreserved products were after mobilization with chemotherapy. The num- stored in a container with liquid nitrogen at tem- ber and percent of CD34+ cells in the peripheral perature from -150-196 °C (Consarctic BSF 350, blood and in the leukapheresis product, also their Schollkrippen, GmbH) until reinfusion. viability were counted by flow cytometry. CD34+

292 Ahmetović-Karić et al. Peripheral blood stem cell collection

Statistical analysis number of CD34+ cells count and mononuclear cells count in the product in all procedures were The distribution of data from the study was tested 5.52x10e6/kg (range 3.28-9.00) and 1.51x10e8/ for normality by Kolmogorov-Smirnov test. The kg (range 1.09-2.28). The median haemoglobin results are presented as arithmetic means ± stan- and haematocrit values in the collected product dard deviation (SD) or medians with a range, de- were 6.45 g/L (range, 4.68-18.00), and 2.17% pending on the data distribution. The significance (range 0.79-5.48), respectively, while the median of the difference in peripheral blood cell counts platelet count in the product was 478.5x10e9/L and electrolyte values before and after the proce- (range 332-677). The median viability of CD34+ dure were analysed by non-parametric Wilcoxon cells in the product after leukapheresis was matched pairs test. The association between pe- 98.75% (range 98-99.05) (Table 2). ripheral and collected CD34+ cells was assessed with non-parametric Spearman correlation. For Table 2. Leukapheresis procedure and product parameters all comparisons, the level of statistical signifi- Procedure parameters x ± SD Range (min-max) cance was p<0.05. Processed total body volume (mL) 12755±3959 4908-21539 AC volume (mL) 1160±345.1 453-1949 Inlet blood flow rate (mL/min) 50.17±10.33 32.60-75 RESULTS Procedure time (min) 293.5±48.43 193-395 Product volume (mL) 143.6±57.54 64-262 The total blood volume of analysed patients vari- Platelet reduction (%) 20±12.10 3.9-36 ed from 3097 to 6446 mL (mean 4586 mL). Out CD34+ collection efficiency CE (%) 57.7±17.48 26-95 of the total of 20 patients, five (25%) were pati- Product parameters Median Range ents with Hodgkin's lymphoma, nine (45%) with CD34+ cells x10e6/L 2728 2028-4612 CD34+ cells (%) 1.57 0.96-2.91 non-Hodgkin's lymphoma and six (30%) with Haemoglobin (g/L) 6.45 4.68-18.00 myeloma multiplex. Most of the patients, nine Haematocrit (%) 2.17 0.79-5.48 (45%), were A+ blood type (Table 1). Platelets x10e9/L 478.5 332-677 Mononuclear cells x10e9/L 83.20 60.20-107.2 Leukocytes x10e9/L 173 156.3-202.8 Table 1. Characteristics of 20 haematological patients who CD34+ cells x10e6/kg 5.52 3.28-9.00 underwent 24 leukapheresis procedures for the collection of Mononuclear cells x10e8/kg 1.51 1.09-2.28 autologous peripheral blood stem cells Viability of CD34+ cells (%) 98.75 98-99.05 Characteristics Value AC, anticoagulant; CE, collection efficiency; x, mean; SD, standard Diagnosis No (%) deviation Hodgkin's lymphoma 5 (25) Non-Hodgkin's lymphoma 9 (45) Myeloma multiplex 6 (30) Table 3. Characteristics of patients, number of collected cells Gender No (%) per patient and number of bags used for cryopreservation per Male 8 (40) patient Female 12 (60) Ordinal No. of CD34+ CD34+ No. of ABO blood type No (%) Diagnosis patient No procedures (x10e6/kg) (%) bags used O+ 3 (15) 1. NHL 1 5.40 1.85 3 A+ 9 (45) 2. NHL 1 10.6 3.42 2 B+ 3 (15) 3. NHL 1 7.66 1.32 4 O- 3 (15) 4. NHL 2 4.86 0.88 8 A- 2 (10) 5. NHL 1 5.17 2.70 2 Age (±SD/range) (min-max) (years) 48.40±12.61 (27-65) 6. NHL 1 3.18 1.44 2 Weight (±SD/range) (min-max) (kg) 74.30±18.60 (45-115) 7. MM 1 6.30 1.30 2 Height (±SD/range) (min-max) (cm) 170.5±9.105 (155-187) 8. NHL 1 5.64 1.70 2 Total blood volume (±SD/range) 4586±1011 (3097-6446) (min-max) (mL) 9. NHL 1 3.77 0.91 3 10. NHL 1 19.67 6.53 3 SD, standard deviation 11. HL 1 14.96 4.51 2 Processed total body blood volume during leu- 12. HL 1 6.07 1.43 4 13. HL 1 13.20 4.90 3 kapheresis procedures varied from 4908-21593 14. HL 1 8.68 3.80 2 mL (mean 12755). The leukapheresis procedures 15. HL 1 3.60 1.11 4 lasted from 193-395 min. (mean 293.5), which 16. MM 2 6.40 2.07 4 was an average of 5 hours. The CD34+ collecti- 17. MM 1 5.80 1.86 3 18. MM 2 12.32 5.63 4 on efficiency ranged from 26-95% (mean 57.7), 19. MM 1 9.11 2.08 2 and the percentage of platelet reduction ranged 20. MM 2 7.24 1.24 4 from 3.9-36% (mean 20) (Table 2). The median NHL, non-Hodgkin's lymphoma; HL, Hodgkin's lymphoma; MM, Myeloma multiplex

293 Medicinski Glasnik, Volume 17, Number 2, August 2020

In 16 (80%) patients, the collection of the target yield was performed with one procedure, while in four (20%) patients it was performed with two procedures on two consecutive days. The median CD34+ cells count and percentage of CD34+ cells in the leukapheresis product in all patients were 6.35x10e6/kg (range 5.34-9.48) and 1.86% (range 1.32-3.52), respectively. The mean number of bags used for cryopreservation after the Figure 1. Correlation of collected CD34+ cells counts with pe- completion of the collection in all patients was ripheral blood CD34+ cell count in all procedures r : Spearman’s correlation coefficient 3.15 (range 2-8) (Table 3). s A decrease in platelet count, haemoglobin and haematocrit values in peripheral blood after le- DISCUSSION ukapheresis was observed, without statistical si- The PBSCs are usually collected during the he- gnificance (p=0.139 and p=0.714, respectively). matologic recovery phase after mobilizing agents The median number of circulating CD34+ cells in have been administered and the time of onset of peripheral blood on apheresis day was 62x10e6/L the collection depends on the type of the disease (range 28.99-89.38). Calcium and potassium va- and patients characteristics (2-4). The mobili- lues in peripheral blood decreased after the pro- zation process may be accompanied by adverse cedure, where the decrease in potassium showed events (2). Administration of chemotherapy cau- statistical significance (p<0.000) (Table 4). ses bone marrow aplasia, so patients are usually Table 4. Peripheral blood and electrolytes parameters before cytopenic prior to collection, which applies spe- and after leukapheresis cifically to peripheral blood platelet counts (2). Median/range Median/range In our study, 20% received a platelet transfusion Parameter p (before) (after) prior to leukapheresis with the aim of raising the 9.06 17.55 Leukocytes x 10e9/L <0.001 platelet count to >30x10e9/L. (4.33-16.55) (8.87-29.28) 3.45 3.45 Erythrocytes x 10e9/L 0.748 In autologous collection, in case of inadequate (3.02-3.74) (3.09-3.77) peripheral veins, central venous catheters may be 77.20 69.05 Platelets x 10e9/L 0.422 (66.08-93.35) (59.60-98.48) used for the collection due to the need for suffi- 103 98.45 cient blood flow through the cell separator (7). Haemoglobin (g/L) 0.139 (90.88-109.3) (92.05-105.8) During one procedure, one of our patients expe- 31.10 31.05 Haematocrit (%) 0.714 (28.20-32.48) (28.13-32.78) rienced problems with venous flow through inlet 2.15 2.29 line of the central venous catheter and the proce- Mononuclear cells x 10e9/L 0.007 (1.23-2.71) (1.62-3.26) dure was continued using a peripheral vein for 62 CD34+ cells x10e6/L - - (28.99-89.38) the inlet line of blood flow. 2.25 2.22 Calcium (mmol/L) 0.324 In leukapheresis procedures, for extracorporeal (2.13-2.32) (2.13-2.39) anticoagulation citrate-based anticoagulation is 4.30 3.50 Potassium (mmmol/L) <0.000 (3.80-4.80) (3.02-3.80) preferred over heparin anticoagulant solution for the reasons of lower cost, safety, effectiveness and rapid systemic clearance (8). The most commonly A strong positive correlation was found betwe- used anticoagulant solution is ACD-A (5,8). Cell en the number of circulating CD34+ cells in separators use total blood volume to calculate the peripheral blood on the apheresis day and the anticoagulant infusion rate delivered to the patient amount of CD34+ cells collected in the products (8). Citrate anticoagulation causes metabolic chan-

(rs=0.73) (Figure 1). ges such as hypocalcemia, hypokalemia, hypoma- Adverse events occurred during one procedure gnesaemia, other electrolyte derangements and (4.2%) due to impaired blood flow through one metabolic alkalosis that may be accompanied by of the lines of central venous catheter. appropriate symptoms (8). The most common reactions during apheresis procedures are princi- pally related to the effect of hypocalcemia (2,8).

294 Ahmetović-Karić et al. Peripheral blood stem cell collection

Factors influencing symptoms development inclu- authors report an average reduction in the plate- de the rate of citrate infusion, the rate of decline in let count after leukapheresis by 43% and 40-45% ionized calcium levels and hepatic metabolism of depending on the type of the used cell separator the infused citrate (8). All patients in our study re- and the length of the procedure (2,10). ceived an intravenous calcium gluconate solution The mean collection efficiency of CD34+ cells to prevent the onset of symptoms of hypocalce- by cell separators in this study was 57.7%. Other mia, which could occur during the procedure due authors in their studies report an average collec- to the use of citrate anticoagulant solution. During tion efficiency of 42-43% and 50-55% depending the procedure patients also drank an oral solution on the type of the used cell separator (10,11). of effervescent calcium tablets. There is conside- A strong positive correlation in our study was rable individual variability in the development of found between peripheral blood CD34+ cell count symptoms and signs of citrate-induced hypocalce- and the final CD34+ cell collection, and on the ba- mia (8). Prophylactic oral calcium supplementa- sis of this and other studies this strong correlation tion or an infusion of intravenous calcium can be shows that the number of peripheral CD34+ cells effective in reducing severity and the incidence of was the best predictor of the amount of CD34+ citrate-induced symptoms during the procedures cells collected in the product per leukapheresis (1,8). In our study, the median calcium value af- (1,12). There is still no consensus on the amount ter the leukapheresis procedure was lower than the of blood stem cells needed to be infused to achieve values before, although this decrease was not sta- adequate hematopoietic recovery. Reportedly, the tistically significant, while the median potassium safe minimum dose of CD34+ cells in the leuka- value after the procedure was significantly lower. pheresis product for single transplant is still an One patient received oral solutions of effervescent amount of ≥2x10e6/kg (6, 12-19). Lower counts potassium tablet at the beginning of both procedu- will be the concomitant risk of delayed neutrophil res during two consecutive days, because of low and platelet engraftment (19). Recommended stem levels of potassium before them. Oral potassium cell collection target is 3-5x10e6/kg CD34+ cells supplementation is a sufficient measure for the in the product (5,12,16). Reinfusion of 5x10e6/kg correction of hypokalemia and it should be admi- CD34+ cells results in prompt engraftment and nistered only in patients with lower levels before should be a preferred target or an optimal target leukapheresis (1). None of our patients experien- for successful transplantation (19,20). Higher ced problems that could be associated with a de- doses of target cells can result in faster engraft- crease of electrolyte levels. Leukapheresis proce- ment, with the balance between targets and the dures may be accompanied by other side effects number of apheresis procedures required to attain such as hypotension, hypertension, nausea/vomi- them (16). The collection of PBSCs for autologous ting, headache, vertigo (9). However, any of these transplantation is a well-established process (21). side effects did not occur in our patients. The number and viability of CD34+ cells is a sur- Extracorporeal blood volume during leukaphere- rogate for predicting successful engraftment (21). sis depends on the type of the used cell separa- In conclusion, the leukapheresis procedure is tor, but when it comes to adult patients, mostly fully automated and involves the use of standar- it does not present a problem because of a small dized cell collection protocols. The results of this volume of blood outside circulation (2). study, such as low number of adverse events and During the PBSC collection, with mononuclear good collection efficiency with adequate patient cells a portion of other cells is always collected, monitoring, indicate that leukapheresis is a safe especially from the platelet layer, so the platelet procedure that is successfully used in the autolo- count always decreases (10,11). gous transplantation process in the treatment of In our study, there was a decrease in the platelet malignant haematological diseases. count, haemoglobin and haematocrit values after the procedure, although this decrease was not sta- FUNDING tistically significant. The mean platelet percenta- No specific funding was received for this study. ge decrease in this study was 20% and no patient TRANSPARENCY DECLARATION required transfusions after the procedure. Other Competing interest: none to declare.

295 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES

1. Bojanić I. Peripheral blood stem cell collection by 11. Setia RD, Arora S, Handoo A, Choudhary D, Shar- large volume leukapheresis procedure. University of ma SK, Dadu T, Doval D, Kapoor M, Bajaj S, Zagreb, Zagreb 2009; Ph. D. thesis. Bachchas V. Comparison of 'Amicus and COBE 2. Bojanić I, Golubić Cepulić B, Nemet D, Rajić LJ, Spectra' for autologous peripheral blood stem cell Batinić D, Labar B. Osobitosti sakupljanja autolo- harvest: An Indian experience. Glob J Transfus Med gnih krvotvornih matičnih stanica iz periferne krvi 2017; 2:102-8. u pedijatrijskih bolesnika (Autologous peripheral 12. Wallis WD, Qazilbash MH. Peripheral blood stem blood stem cell collection in pediatric patients) [in cell mobilization in multiple myeloma: Growth Croatian] Liječ Vjes 2006; 128:43-48. factors or chemotherapy? World J Transplant 2017; 3. Grubovic RM, Georgievski B, Cevreska L, Genadie- 7:250-9. va-Stavric S, Grubovic MR. Factors associated with 13. Hequet O. Hematopoietic stem and progenitor cell successful mobilization and collection of peripheral harvesting: technical advances and clinical utility. J blood hematopoietic stem cells in autologous and Blood Med 2015; 6:55-67. allogenic donors. SEE J Immunol 2017; 2017:20012. 14. Sakashita AM, Kondo AT, Ribeiro AAF, Cipolletta 4. Pivkova A, Genadieva-Stavrik S, Stojanoski Z, Ce- ANF, Colesanti MV, Hamerschlak N, Kutner JM. vreska L, Karanfilski O, Trajkova S, Krstevska Bal- Factors affecting autologous peripheral blood he- kanov S, Milenkov V, Georgievski B. Mobilization matopoietic stem cell collections by large-volume strategies for autologous collection and cryopreser- leukapheresis: a single center experience. Einstein vation of peripheral blood stem cells in patients with 2011; 9:196-200 lymphoproliferative diseases. Maced J Med Sci 15. Demirer T, Ilhan O, Ayli M, Arat M, Dagli M, Ozcan 2009; 2:209-16. M, Haznedar R, Genc Y, Fen T, Ayyildiz E, Dincer 5. Urošević I, Balint B, Popović S. Matične ćelije S, Arslan O, Gurman G, Konuk N, Dalva K, Uysal hematopoeze periferne krvi – biologija, aferezno A, Koc H, Ozet G, Akan H. Monitoring of peripheral prikupljanje i kriokonzervacija (Peripheral blood blood CD34+ Cell counts on the first day of aphe- hematopoietic stem cells – biology, apheresis collec- resis is highly predictive for efficient CD34+ cell tion and cryopreservation) [in Serbian] Med Pregl yield. Ther Apher 2002; 6:384-9. 2007; 60:42-7. 16. Giralt S, Costa L, Schriber J, DiPersio J, Maziarz R, 6. Sofo-Hafizovic A, Dizdarevic-Rekic A, Ibricevic- McCarty J, Shaughnessy P, Snyder E, Bensinger W, Balic L, Burazerovic L, Vila M, Suljovic-Had- Copelan E, Hosing C, Negrin R, Petersen FB, Ron- zimesic E, Skuric-Tomic M, Catovic-Baralija E, delli D, Soiffer R, Leather H, Pazzalia A, Devine S. Ahmetovic G, Sahovic A, Sicanica S, Omanovic A. Optimizing autologous stem cell mobilization stra- Successful collection of stem cells in one day in the tegies to improve patient outcomes: consensus gu- process of autologous stem cell transplantation. In: idelines and recommendations. Biol Blood Marrow Badnjevic A. (eds) CMBEBIH 2017. IFMBE Pro- Transplant 2014; 20:295-308. ceedings, Vol 62. Singapore: Springer, 2017. 17. Domanovič D. Napovedna vrednost števila celic 7. Vacca M, Perseghin P, Accorsi P, Pierelli L. Central CD34+ v krvi bolnika/darovalca pred zbiranjem kr- venous catheter insertion in peripheral blood hema- votvornih matičnih celic z levkaferezami (Predictive topoietic stem cell sibling donors: the SidEM (Itali- value of CD34+ cells in blood of patient/donor befo- an Society of Hemapheresis and Cell Manipulation) re hematopoietic stem cells collection by leukaphe- point of view. Transfus Apher Sci 2014; 50:200-206. resis) [in Slovenian] Zdrav Vestn 2004; 73 (Suppl 8. Lee G, Arepally GM. Anticoagulation techniques in 1):63-7. apheresis: from heparin to citrate and beyond. J Clin 18. Yuan S, Wang S. How do we mobilize and collect Apher 2012; 27:117-125. autologous peripheral blood stem cells? Transfusion 9. Henriksson MM, Newman E, Witt V, Derfler K, 2017; 57:13-23. Leitner G, Eloot S, Dhondt a, Rock G, Ptak J, Blaha 19. Mohammadi S, Mohammadi AM, Nikbakht M, No- M, Lanska M, Gasova Z, Hrdlickova R, Ramlow W, rooznezhad AH, Alimoghaddam K, Ghavamzadeh Prophet H, Liumbruno G, Mori E, Griskevicius A, A. Optimizing stem cell mobilization strategies to Audzionijene J, Vrielink H, Roumbout S, Aandahl ameliorate patient outcomes: a review of guide-lines A, Sikole A, Tomaz J, Lalic K, Mazic S, Strineholm and recommendations. Int J Hematol Oncol Stem V, Brink B, Berlin G, Dykes J, Toss F, Axselsson CG, Cell Res 2017; 11:78-85. Stegmayr B, Nilsson T, Norda R, Knutson F, Ram- 20. Jillella AP, Ustun C. What is the optimum number sauer B, Wahlstrom A. Adverse events in apheresis: of CD34+ peripheral blood stem cells for an auto- an update of WAA registry data. Transfus Apher Sci logous transplant. Stem Cell Dev 2004; 13:598-606. 2016; 54:2-15. 21. Deneys V, Fabry A, Van Hooydonk M, Sonet A, 10. Lee SN, Sohn JY, Kong JH, Eom HS, Lee H, Kong Andre M, Bourgeois M, Botson F. Efficiency of au- SY. Comparison of two apheresis systems of Cobe tologous stem cell collection: comparison of three and Optia for autologous peripheral blood stem cell different cell separators. Transfus Apher Sci 2017; collection. Ann Lab Med 2017; 37:327-30. 56:35-8.

296 ORIGINAL ARTICLE

ABO groups as a risk factor for myocardial infarction in the population of Bosnia and Herzegovina Ermina Mujičić1, Nermir Granov1, Armin Šljivo2, Amina Selimović3

1Department of Cardiovascular surgery, Clinical Centre of the University of Sarajevo, 2School of Medicine, University of Sarajevo, 3Department of Paediatrics, Clinical Centre of the University of Sarajevo; Sarajevo, Bosnia and Herzegovina

ABSTRACT

Aim To investigate the association of ABO blood types and appearance of myocardial infarction (MI) among the population in Bo- snia and Herzegovina.

Methods This was a cross sectional study conducted at the De- partment for Cardiovascular Surgery at the Clinical Centre of the University of Sarajevo from January 1st to December 31st 2019. Patients were divided into two groups, defined by their blood type, A (N=91) and non-A (N=109). ABO blood groups were determined using standard agglutination techniques. Information regarding history of MI was taken from their medical documentation during the preparation for surgery.

Corresponding author: Results In total 200 patients were involved in this study. The stu- Ermina Mujičić dy sample consisted of 151 (74.5%) males and 49 (25.5%) females. Mean age was 62.98 ± 7.73. Various risk factors associated Department of Cardiovascular Surgery, with myocardial infarction were tobacco 92 (46.0%) and alcohol Clinical Centre of the University of 54 (27.0%) consumption, obesity 77 (38.5%), diabetes mellitus Sarajevo 47 (23.5%) and hypertension 91 (45.3%). Being A blood type Bolnička 25, 71 000 Sarajevo, (OR=3.308; 95% Cl 1.317-8.311; p=0.011) and being male and Bosnia and Herzegovina having hypertension (OR=3.086; 95% Cl 1.262-7.545; p=0.013) Telephone number: +387 33 297 941; significantly increased the risk for the development of MI among young adults. Patients with A blood type were significantly yo- Fax: +387 33 298 522; unger [median 58.0 vs. 63.0; U = 2738.5; p =0.027] when they E-mail: [email protected] developed MI compared to non-A blood type. ORCID ID: https://orcid.org/0000-0002- 2995-4719 Conclusion The age of myocardial infarction occurrence in the population of Bosnia and Herzegovina is lower in patients with blood group A compared to non-A blood types.

Original submission: Key words: acute myocardial infarction, A blood type, ABO blo- 04 May 2020; od groups, coronary artery disease, risk factors Accepted: 12 June 2020 doi: 10.17392/1197-20

Med Glas (Zenica) 2020; 17(2): 297-300

297 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION dard agglutination techniques. Information regarding other risk factors and history of MI was The development of myocardial infarction (MI) is taken from their medical documentation during associated with various risk factors such as tobacco the preparation for surgery. and alcohol consumption, high lipid concentration, patient-reported hypertension, obesity, poor diet, Methods lack of physical activity and some psychological factors (1). Among these, a lot of studies (2-5) in Data collected from patients documentation con- the last decade have shown a significant, yet also sisted of the risk factors: age, gender, history or controversial association of ABO blood types with ongoing tobacco and alcohol consumption, body MI. Blood is an individual and unchangeable risk mass index (BMI), ongoing diabetes mellitus and factor for coronary heart disease and development hypertension. Young patients were considered be- of MI. Some studies (3,5,6) suggest that non- low the age of 55 years (9). Tobacco consumption O type blood groups, with the highest prevalence was defined as an adult who has smoked mini- among A blood type, are associated with the in- mum 100 cigarettes in his or her life and who cur- creased risk of MI. The coronary arteriosclerosis rently smokes cigarettes (10). Unhealthy alcohol plaques of O type blood group were exhibited to consumption was defined as drinking more than be more stable compared with non-O type blood seven standard glasses of wine or beer each week group. Moreover, the non-O type blood groups (11). Diabetes mellitus was defined as having fast- have more serious coronary artery stenosis than ing plasma glucose above 7 mmol/L on at least two O type blood group (7). The possible pathophysi- occasions or HbA1c above 6.5%, or history or on- ological mechanism of these findings lays in ge- going use of hypoglycemic medications (12). Hy- netic variations of rs495828 gene, which could pertension was defined as having blood pressure be responsible for higher platelet aggregation and greater than 140 mmHg/90 mmHg on several mea- high levels of arachidonic acid and collagen (8). sures, or history of diagnosed and/or treated with More studies are needed to investigate various ge- medication or non-pharmacological treatment (13). netic and pathophysiological mechanism for ABO Statistical analysis blood types contribution to the development of MI. To our knowledge, this is the first study assessing In order to summarize the data, descriptive sta- relation between ABO blood types and MI in Bo- tistics were run and results were displayed in snia and Herzegovina. This research is needed frequencies and percentages. In order to define in order to prove that not only lifestyle, but also independent predictors for the development of genetic predispositions which define ABO blood MI, binary logistic regression was conducted. types contribute to MI in our population. Mann Whitney U test was used to show significant association between different variables. The aim of this study was to investigate the association of ABO blood types and appearance of RESULTS MI among adults in Bosnia and Herzegovina. In total 200 patients who underwent CABG sur- PATIENTS AND METHODS gery in the Clinic for Cardio Surgery of the Cli- nical Centre, University in Sarajevo, were inclu- Patients and study design ded in the study. The sample consisted of mainly This cross sectional study was conducted at the males, 151 (74.5%). Mean age was 62.98 ±7.73 Clinic for Cardiovascular Surgery of the Clini- (range 41-81) years. cal Centre, University in Sarajevo, Bosnia and The patients were divided in two groups accor- Herzegovina, from January 1st to December 31st ding to their blood groups, 91 (45.5%) patients 2019. All patients were hospitalized at the Cli- were A blood type group and 109 (54.5%) pati- nic and were prepared for coronary artery bypass ents were non-A [60 (30.0%) O, 32 (16.0%) B, grafting (CABG) surgery. 17 (8.5%) AB]. Patients were divided into two groups, defined by Out of 200 patients, 92 (46.0%) reported to be cu- their blood type: A (N=91) and non-A (N=109). rrent smokers, of whom 43 (46.7 %) in A blood ABO blood groups were determined using stan- group and 49 (43.3 %) in non-A blood group; 108

298 Mujičić et al. ABO groups and myocardial infarction

(54.0%) were non-smokers. Among 92 smokers, Table 2. Distribution of myocardial infarction (MI) according 45 (48. 9%) patients had more than 10 years of to the age smoking history, e.g. 22 (48.9%) in A blood group Blood type Variable Non-A p and 23 (51.1%) in non-A (p>0.05). Alcohol con- A B AB O sumption was found among 54 (27.0%) patients No of patients 91 32 17 60 62.8 64.0 60.7 63.3 with drinking more than seven standard glasses of Mean age (±) (years) wine or beer each week, of which 29 (53.7%) in ±7.9 ±6.4 ±10.7 ±7.1 Age of the first MI (median) (years) 58 64 63 63 0.027 A blood group and 25 (46.3%) in non-A (p>0.05). The BMI ranged from 21.1 to 44.3 kg/m2, with Binary logistic regression of independent predic- a median value of 29.0 kg/m2. Most of the pati- tors for the development of MI showed no stati- ents were overweight, 97 (48.5%) of whom 43 stical significance (p=0.534). It explained 58.9% 2 (44.3%) in A and 54 (55.7%) in non-A blood (Nagelkerke R ) of the variance, and correctly group (BMI 25-29.9), followed by 77 (38.5%) classified 87.9% of cases. Being A blood type obese patients (BMI >30), 32 (41.5%) in A and (OR=3.308; 95% Cl 1.317-8.311; p=0.011), 45 (58.5%) in non-A blood group, and only 26 male, and having hypertension (OR=3.086; 95% (13.0%) patients had BMI within the normal ran- Cl 1.262-7.545; p=0.013) significantly increa- ge (18.5-24.9), 16 (61.5%) in A and 10 (38.5 %) sed the risk for the development of MI among in non-A blood group. (p>0.05). young adults. The model also explained that being younger (OR=0.755; 95% Cl 0.694-0.822; Diabetes mellitus was found in 47 (23.5%) pati- p<0.001) decreased the risk of the development ents, of whom the majority 10 (21.1%) used oral of MI among young adults. hypoglycemic medications. Hypertension was recorded among 91 (45.3%) patients. (Table 1). DISCUSSION Table 1. Risk factors associated with myocardial infarction To our knowledge this is the first study regarding according to the blood types the association between ABO blood types and the No (%) of patients in the blood type group development of MI in Bosnia and Herzegovina, Risk factor p Non-A as well as in this region. The majority of our pati- A B AB O ents had positive history of smoking and obesity. No of patients 91 32 17 60 >0.001 A blood type was identified as an independent 69 22 14 46 Sex (male) >0.001 (75.8) (20.2) (12.8) (42.2) predictor for MI and patients with the same blood 73 29 12 53 type were significantly younger when they deve- Age (>55 years) >0.001 (80.2) (26.6) (11.0) (48.6) loped MI compared to non-A blood types. 43 15 9 25 Smoking >0.001 (47.2) (13.8) (8.2) (22.9) Because of the lack of national screening pro- 29 7 8 10 Alcohol consumption >0.001 grams, we could not compare distribution of ABO (31.9) (6.4) (7.4) (9.2) blood types to the national data. Studies which 32 17 13 15 BMI > 30 >0.001 (35.2) (15.6) (11.9) (13.8) assessed the association between ABO blood types 25 9 5 8 Diabetes mellitus >0.001 and various entities conducted in Croatia (14-16) (27.5) (8.2) (4.9) (7.3) and studies regarding the association of ABO blo- 37 17 9 28 Hypertension >0.001 (40. 6) (15.6) (8.2) (25.7) od types and MI (3,6,17) in other countries showed BMI, body mass index; similar distribution of ABO blood types among the population. Our study sample consisted in majo- Out of 200 patients, 91 (45.5%) patients were A, rity of males (74.5%) similar to Turkish (80.3%) 60 (30.0%) O, 32 (16.0%) B and 17 (8.5%) pati- (6) and Canadian study (100%) (17), and it is ents were AB blood type. different from the Taiwanese study (49.8%) (3). When age of MI development was compared When comparing different risk factors associated between the patients who were A blood type and with MI, our results showed a higher prevalence those who were not, a statistical significance was of obesity, hypertension and diabetes mellitus than found (median 58.0 and 63.0 years (p =0.027). other studies (3,6,17) and a lower prevalence of Patients with A blood type were significantly yo- smokers than the Turkish study (6). A blood type unger when they developed MI compared to non- was identified as a predictor for the development A blood type (Table 2).

299 Medicinski Glasnik, Volume 17, Number 2, August 2020

of MI, which was shown in Taiwanese (3) and study limited to several risk factors. Future studies Canadian (14) studies, and it was significantly at should also include detailed genotype information younger age, which was also shown in the Turkish of the ABO alleles, as well as other risk factors and study (6). This research confirms that the patients comorbidities that contribute to MI. with A blood type, male and having hypertension In conclusion, the age of MI occurrence in the significantly increased the risk for the development population of Bosnia and Herzegovina was lower MI among young adults (less than 55 years). in patients with blood group A compared to non- Our study had several limitations. Firstly, the study A blood types. design was cross-sectional, which makes it difficult to infer causality between various factors. Secon- FUNDING dly, we analysed the patients who were supposed No funding was received for this study. to have coronary bypass grafting surgery, e.g. not all patients with IM from the population. Additio- TRANSPARENCY DECLARATION nally, limited data in patients’ histories made our Competing interests: None to declare

REFERENCES 1. Lisowska A, Makarewicz-Wujec M, Filipiak KJ. 10. Kim HN, Shin MA, Roh JH, Han MK, Won YM, Cho Risk factors, prognosis, and secondary prevention of IR, Park HJ, Lee TK, Park TK, Ha HM, Yang SW, myocardial infarction in young adults in Poland. Kar- Min SH, Lee SY, Lee SH, Kim JH, Kwon SJ, Lee diol Pol 2016; 74:1148-53. YS, Ko YW, Kim IH, Kwak JH, Jung TG, Jeon JW, 2. Hu X, Qiao S, Qiu H, Ye S, Feng L, Song L. Associa- Oh KR, Ha HS, kim MS, Kim YM, Kim MJ, Kim tion between ABO blood group and acute myocardi- TY, Chin JH. Association between cigarette smoking al infarction. Zhonghua Xin Xue Guan Bing Za Zhi frequency and health factors among Korean adults. 2015; 43:785-7. Iran J Public Health 2018; 47(Suppl 1):19-26. 3. Lee HF, Lin YC, Lin CP, Wang CL, Chang CJ, Hsu 11. Department of Health. Alcohol Guidelines Review LA. Association of blood group A with coronary ar- - Report from the Guidelines development group to tery disease in young adults in Taiwan. Intern Med the UK Chief Medical Officers. HM Government. 2012; 51:1815-20. 2016. https://www.gov.uk/government/consultations/ 4. Dentali F, Sironi AP, Ageno W, Crestani S, Franchini health-risks-from-alcohol-new-guidelines (28 April M. ABO blood group and vascular disease: an update. 2020) Semin Thromb Hemost 2014; 40:49-59. 12. Kharroubi AT, Darwish HM. Diabetes mellitus: the 5. Zhou B, Wu N, Zhu C, Gao Y, Guo Y, Qing P, Li X, epidemic of the century. World J Diabetes 2015; Wang Y, Dong Q, Li J. ABO blood group is a risk fac- 6:850-67 tor for coronary artery disease in patients with poor 13. Schwartz CL, Mcmanus RJ. What is the evidence blood pressure control. Clin Exp Hypertens 2017; base for diagnosing hypertension and for subsequent 39:366-70. blood pressure treatment targets in the prevention of 6. Sari I, Ozer O, Davutoglu V, Gorgulu S, Eren M, cardiovascular disease? BMC Med 2015;13:256. Aksoy M. ABO blood group distribution and ma- 14. Pisk SV, Vuk T, Ivezić E, Jukić I, Bingulac-Popović jor cardiovascular risk factors in patients with acu- J, Filipčić I. ABO blood groups and psychiatric disor- te myocardial infarction. Blood Coagul Fibrinolysis ders: Croatian study. Blood Transfus 2019; 17:66-71. 2008; 19:231-4. 15. Jukic I, Bingulac-Popovic J, Dogic V, Babic I, Culej 7. Huang X, Zou Y, Li L, Chen S, Hou J, Yu B. Relati- J, Tomicic M, Vuk T, Sarlija D, Balija M. ABO blood on of ABO blood groups to the plaque characteristic groups and genetic risk factors for thrombosis in Cro- of coronary atherosclerosis. Biomed Res Int 2017; atian population. Croat Med J 2009; 50:550-8-16. 2017:2674726. 16. Karabuva S, Carević V, Radić M, Fabijanić D. The 8. Christiansen MK, Larsen SB, Nyegaard M, Neer- association of ABO blood groups with extent of co- gaard-Peterson S, Würtz M, Grove EL, Hvas AM, ronary atherosclerosis in Croatian patients suffering Jensen HK, Kristensen SD. The ABO locus is asso- from chronic coronary artery disease. Biochem Med ciated with increased platelet aggregation in patients (Zagreb) 2013; 23:351-9. with stable coronary artery disease. Int J Cardiol 17. Capuso E, Bonfanti C, Fratitini F, Montorosi P, Tur- 2019; 286:152-8. do R, Previdi MG, Turriri E, Franchini M. The relati- 9. Awad-lkarim AA, Bagger JP, Albers CJ, Skinner JS, onship between ABO blood group and cardiovascular Adams PC, Hall RJ. A prospective study of long term disease: results from the Cardiorisk program. Ann- prognosis in young myocardial infarction survivors: Transi Med 2016; 4:189-93. the prognostic value of angiography and exercise testing. Heart 2003; 89:843-7.

300 ORIGINAL ARTICLE

A novel flavivirus strain detected in phlebotomine sandflies in Bosnia and Herzegovina

Mirsada Hukić1,2, Halida Avdihodžić2, Ivan-Christian Kurolt3, Alemka Markotić3, Jasna Hanjalić4, Lejla Kapur-Pojskić4, Irma Mahmuljin2, Monia Avdić5, Irma Salimović-Bešić6, Lejla Smajlović-Skenderagić5,7

1Department of Medical Sciences, Academy of Sciences and Arts of Bosnia and Herzegovina, 2Institute for Biomedical Diagnostics and Research “Nalaz”; Sarajevo, Bosnia and Herzegovina, 3Clinic for Infectious Diseases “Dr. Fran Mihaljević”, Zagreb, Croatia, 4Institute for Genetic Engineering and Biotechnology, University of Sarajevo, Bosnia and Herzegovina 5International Burch University, Sarajevo, Bosnia and Herzegovina, 6Unit for Clinical Microbiology, Clinical Center, University of Sarajevo; Sarajevo, Bosnia and Herzegovina, 7International Committee of the Red Cross, Tbilisi, Georgia

ABSTRACT

Aim Phlebotominae sandflies are primary vectors of phleboviruses, causing the sandfly fever disease. The aim of this study was to detect and report the presence of flaviviruses in Phlebotominae sandflies captured in Bosnia and Herzegovina.

Methods After a microscopic and morphometric analysis, the final identification of collected Phlebotomus specimens was confirmed by PCR, using a hemi-nested polymerase chain reaction on extracted and reversely transcribed RNA.

Corresponding author: Results We obtained a 155 nt long fragment of the viral non-structu- Lejla Smajlović-Skenderagić ral protein 5 (NS5) gene (GenBank accession no. MN090154). International Committee of the Red Cross The acquired nucleotide sequence, provisionally named as Drež- Ilo Mosashvili St.24, 0162 Tbilisi, nica, showed a maximum of 70-80% identity in 70-88% (110-137 nucleotides) of the query coverage with several Anopheles, Sa- Georgia; bethes, Calbertado and Culex flaviviruses. Maximum likelihood Fax: +995 322 35 55 20; phylogenetic analysis showed that the new flavivirus Drežnica E-mail: [email protected] clusters together with the flavivirus isolated from Culiseta annu- Hukić Mirsada ORCID ID: https://orcid. lata mosquitos. org/0000-0002-7870-4863 Conclusion We report the presence of flavivirus in Phlebotominae sandflies, captured in Drežnica, Herzegovina for the first time. The next phase of research will be directed towards virus cultivation, obtaining a longer or complete virus sequence and clarifying the medical and epidemiological importance of the Drežnica virus.

Key words: arboviruses, Anopheles, mosquitoes, Phlebotomus, Original submission: sandfly 28 May 2020; Revised submission: 08 June 2020; Accepted: 16 June 2020 doi: 10.17392/1212-20

Med Glas (Zenica) 2020; 17(2): 301-307

301 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION fever virus infections in the Balkans, dating back to the end of the 19th century. In the 1886 year, the Outbreaks of emerging and re-emerging neurotro- Austro-Hungarian military physician Alois Pick pic viral disease transmitted by arthropods marked described the pathology and treatment of disea- the beginning of the 21st century. There are more se that occurred in Herzegovina (24). Only after than 130 arthropod-borne viruses (arboviruses) Pick’s discoveries, scientists began to study the known to cause human disease, among which the disease in more detail, and found that it is preva- majority belong to RNA virus families: Togaviridae lent in the regions with a widespread of sandflies, (genus Alphavirus), Flaviviridae (genus Flavivirus), which afterwards led to the renaming of the infec- Bunyaviridae (genera Nairovirus, Orthobunyavi- tion from the "dog disease" to the "sandfly fever". rus, Phlebovirus, and Tospovirus), Rhabdoviridae During the second half of the 20th century, the (genus Vesiculovirus) and Orthomyxoviridae (ge- presence of antibodies against SFNV in patients nus Thogotovirus) (1-5). Arboviruses are distribu- from B&H was reported in two different publica- ted worldwide, causing mild and almost asympto- tions (25,26). In 2009, for the first time in B&H, matic to severe human infections (6-9). TOSV infection was confirmed in seven (10.3%) In the last two decades, among all arboviruses, of 68 patients with febrile illness of unknown ori- the greatest public attention attracted viruses from gin, that were tested for the presence of IgG and genus Flavivirus species Dengue virus (DENV), IgM antibodies specific for TOSV (19). In 2017 Yellow fever virus (YFV), Zika virus (ZIKV) and Ayhan et al. reported the presence of the Balkan West Nile virus (WNV), transmitted by mosquitoes virus (BALKV) that belongs to the Sandfly fever as well as genus Phlebovirus (especially Sandfly Naples virus species, in one pool of Phlebotomus fever Naples group and Sandfly fever Sicilian gro- neglectus captured in the village of Sovići in Her- up) transmitted by Phlebotomus sandflies. zegovina (27). The Balkan virus was first detec- Sandfly fever, also called “Pappataci fever” or “3- ted in Albania, but a former reports confirm the day fever”, is characterized with influenza-like presence of the virus in both B&H and Croatia symptoms that occur after a 3–6-day incubation (27,28). Although BALKV is the first genetically period (10). Symptoms include myalgia, physical identified phlebovirus in B&H, there is still no evi- sensitivity of the eyes to light exposure, headache, dence that this virus can cause human infections. abdominal discomfort, fatigue and eating disorder, In spite of the fact that it has been more than 130 stiffness of the neck and back, followed with spe- years since the sandfly fever was first described in

cific blood findings (11,12). Two main serocom- B&H, the disease causing phlebovirus has not yet plexes associated with sandfly fever are the Na- been isolated from Phlebotomus spp. or infected ples serocomplex and Sicilian serocomplex, from humans, nor genetically analysed in this region. The which Sicilian virus (SFSV), sandfly Cyprus virus increase in the number of reported patients with the (SFCV), sandfly Naples virus (SFNV) and Toscana unclear febrile condition, serous meningitis or en- virus (TOSV) circulate in the Mediterranean Basin cephalitis without identified aetiology, occurring in (13-16). In endemic areas, infections occur in the summer months in B&H, was noticed (19). summer, with a peak during the month of August, The aim of this study was to investigate the presen- which coincides with the peak activity of sandfli- ce of phleboviruses (Sandfly virus) in samples of es as vectors (17-21). Human infections with the Phlebotomus sandflies collected in B&H and the symptoms very similar to sandfly fever can be also possible presence of flavivirus strains in trapped caused by many Flaviviruses (e. g. DENV, WNV, sandflies, considering the report of flavivirus pre- YFV, Japanese encephalitis virus or ZIKV). Altho- sence in Phlebotomus perniciosus pools trapped in ugh the flaviviruses are usually detected in mosqu- Algeria and also in Portuguese sandflies. itoes, Moureau et al. for the first time isolated flavivirus RNA from two pools of male Phlebotomus MATERIALS AND METHODS perniciosus collected in the territory of Algeria (22). Also, Amaro, Ze-Ze and Alves submitted to the Ge- Materials and study design nBank a flavivirus RNA sequence (HM563684) In September 2018 Phlebotomus sandflies were isolated from Portuguese sandflies (23). collected from 44 micro localities in Herzegov- Bosnia and Herzegovina (B&H) is the country ina (southern and non-administrative region of with the longest history of publications on sandfly B&H) (Stolac, Drežnica, Ljubuški, Mostar and

302 Hukić et al. A novel flavivirus strain in sandflies

the village of Blatnica near Čitluk) and 12 micro (Roche, Mannheim, Germany) according to the localities in northeast Bosnia (Živinice and Lu- manufacturer’s protocol. kavac). Sandflies were trapped in custom-made Reverse transcription (RT) was performed using traps, set by twilight, and collected in the morn- random hexamers and 200 U/μl M-MLV reverse ing. They were grouped according to the location, transcriptase (Promega, Madison, USA) in a two- into pre-labelled tubes and immediately stored step protocol. In the first step, 2 µg of isolated on dry ice. Until sorting and final treatment, all RNA (or up to 16 µL) and 1 µL of 500 μg/mL samples were kept at -80 °C. The identification random hexamers in the total volume of 17 µL in of all collected Phlebotomus specimens, based ultrapure DEPC-pretreated water, were incubated on the microscopic and morphometric analysis, for 5 minutes at 70 °C, and immediately cooled was conducted at the Institute for Biomedical to 4 °C. In the second step, 8 µl of RT mix was Diagnostics and Research “Nalaz”; Sarajevo, prepared using 5 μL of 5X reaction buffer, 1 μL Bosnia and Herzegovina. For molecular analysis of 10 mM dNTPs, 1 μL of RNase inhibitor (40 the collected sandflies were further divided into U/mL) and 1 μL of M-MLV reverse transcriptase 30 pools according to the location with up to 46 (200 U/mL) and incubated for 1 minute at 37 °C. phlebotomus per pool. RT mix was then added to the annealed hexam- The Ethics Committee of the Department of Medi- ers/template, and incubated for 60 min at 37 °C. cal Sciences, Academy of Sciences and Arts of Bo- Enzyme inactivation was performed at 95 °C for snia and Herzegovina, approved the study protocol. 5 min, and then at 4 °C for 5 min, after which the cDNA was stored at -20 °C. Methods For the virus detection in sandflies, "nested" PCR Individual pools were added into the pre-prepared approaches were applied to detect all Phlebovirus tubes with approximately 0·5 mL of 1 mm borosi- genomes or specifically TOSV genomes. Flavi- licate beads (LabTIE B.V., Rosmalen, Netherlan- virus specific genomes were detected utilizing a ds) and 700 µL of 20% fetal bovine serum (FBS) hemi-nested PCR approach (29,30) (Table 1). in phosphate-buffered saline (PBS) buffer, and PCR was performed in the volume of 50 µl, using homogenized using bead-based homogenization 0·3 µL of Hot Start Taq polymerase (Jena Biosci- (MagNA Lyser, Roche, Mannheim, Germany) for ence, Jena, Germany), 5 µL of 10x PCR buffer, 1 two rounds at 6500 rpm for 30 seconds with a brief µL of 10 mM dNTPs, 0·4 µL of 50 µM primers intermitted incubation on ice. (Metabion international, Planegg, Germany) and The isolation of RNA was done from 200 µL ultrapure DEPC-pretreated water. As template 5 supernatant using Tri Pure Isolation Reagent µL cDNA were used in each case for the first PCR,

Table 1. Primers and assays used for the detection of viral sequences in Phlebotomus spp. Primer name Sequence (5’- 3’) Function Band size (bp) Reference Pan-Phlebovirus PCR NPhlebo1+ ATGGARGGITTTGTIWSICIICC outer forward 550 NPhlebo1- AARTTRCTIGWIGCYTTIARIGTIGC outer reverse (29) NPhlebo2+ WTICCIAAICCIYMSAARATG inner forward 250 NPhlebo2- TCYTCYTTRTTYTTRARRTARCC inner reverse TOSV S segment PCR TosN123 GAGTTTGCTTACCAAGGGTTTG outer forward 660 TosN829 AATCCTAATTCCCCTAACCCCC outer reverse (29) TosN234 AACCTTGTCAGGGGNAACAAGCC inner forward 520 TosN794 GCCAACCTTGGCGCGATACTTC inner reverse TOSV L segment PCR NPhlebo1+ ATGGARGGITTTGTIWSICIICC outer forward 550 NPhlebo1- AARTTRCTIGWIGCYTTIARIGTIGC outer reverse (29) NPhlebo2+ WTICCIAAICCIYMSAARATG inner forward 130 ATos2- RTGRAGCTGGAAKGGIGWIG inner reverse Pan-Flavivirus PCR MAMD AACATGATGGGRAARAGRGARAA outer forward 260 cFD2 GTGTCCCAGCCGGCGGTGTCATCAGC outer reverse (30) MAMD AACATGATGGGRAARAGRGARAA outer forward 210 FS778 AARGGHAGYMCDGCHATHTGGT inner reverse

303 Medicinski Glasnik, Volume 17, Number 2, August 2020

whereas 1 µL first-round amplicon was added into ethanol / EDTA / sodium acetate precipitation the nested respectively hemi-nested PCR. according to the manufacturer’s recommendation Amplification with NPhlebo1 primers was- car amplicons were analysed on an ABI Prism 310 ried out at 94 °C for 2 min, followed with 40 PCR Genetic Analyzer (Applied Biosystems), using cycles at 94 °C for 30 secs, 45 °C for 60 secs and compatible POP-6 Polymer and 50 cm well-to- 72 °C for 30 secs, extension at 72 °C for 5 min, read capillary of the same manufacturer. Visu- and held at 4 °C for the infinite time. Nested PCR alization of obtained nucleotide sequence was with NPhlebo2 FW/REV primers or NPhlebo2 determined using Sequencing Analysis Software and ATos2 primers were carried using 35 cycles v5.2 (Applied Biosystems), and settlement and and the same cycling conditions as used for comparative sequence analysis using the Bio Edit NPhlebo1 primers. Amplification with TosN123 Sequence v.7.2.5.0 (31). and TosN829 outer primers was carried out at 95 The obtained nucleotide sequences were compa- °C for 90 secs, followed with 40 PCR cycles at 95 red with sequences stored in the NCBI gene bank °C for 30 secs, 51 °C for 30 secs and 72 °C for 40 using BLASTN algorithm analysis (32). secs, extension at 72 °C for 5 min, and held at 4 °C The multiple sequence alignment was done in pro- for the infinite time. Amplification with TosN234 gram Mega X, using ClustalW and MUSCLE al- and TosN794 was carried out at 95 °C for 90 secs, gorithms for maximum likelihood (ML) analyses followed with 35 cycles at 95 °C for 30 secs, 56 °C (33). After the alignment, sequences were trimmed for 30 secs and 72 °C for 30 secs, final extension to the size of the 155 nt sequence obtained from at 72 °C for 5 min and 4 °C for the infinite time. Phlebotomine sandflies. Maximum Likelihood Amplification of Flavivirus RNA with MAMD phylogeny analysis was done using the Bootstrap and cFD2 primers was carried out at 94 °C for method with 1000 replications and Tamura-Nei 2 min, followed with 40 PCR cycles at 94 °C for model (34) with uniform rates among sites in 30 secs, 53 °C for 30 secs and 72 °C for 45 secs, the MEGA X software. The phylogeny was also one step of final extension at 72 °C for 5 min, and estimated using the Neighbor-Joining algorithm, held at 4 °C for infinite time. Heminested PCR Maximum Composite Likelihood method (35) with MAMD and FS 778 primers was carried and 1000 bootstrap replications. using 35 cycles and same cycling conditions as used for first PCR. RESULTS Of all performed PCRs, 10 µL were combined Phlebotomus spp. were trapped in Herzegovina with 2 µL of 6x Loading dye and analysed using (Stolac, Drežnica, Blatnica, Ljubuški and Mo- electrophoresis on 1% or 2% agarose gel, depend- star) as well as in small towns Živinice and ing on amplicon size, prepared in SBA buffer and Lukavac in northeast Bosnia. A total of 751 0·025% of Roti-GelStain (Carl Roth, Karsruhe, phlebotomine sandflies were collected and mor- Germany), run alongside 6·5 µL of DNA ladder phologically identified. Based on the location, (GeneRuler DNA Ladder Mix, ThermoFischer sandflies were divided into 30 pools, consist- Scientific, Waltham, USA). Electrophoresis was ing of 3 specimens only to the maximum of 46 conducted at 120 V for 22 min (Cleaver Scientif- sandflies specimens. ic LTD, Rugby, United Kingdom). Visualization After the amplification with specific Phleboviru- of bands was done under a UV transilluminator ses or TOSV primers, none of the thirty pooled (omniDOC Gel Documentation System, Cleaver samples gave a positive result, while fragments Scientific LTD, Rugby, United Kingdom). of the expected size were obtained for all PCR For sequencing, bands of according size were gel- positive controls. excised and purified by AccuPrep Gel Purification The amplification with the outer primers speci- Kit (Bioneer Corporation, Daejeon, South Korea) fic for Flaviviruses generated a large number of and protocol recommended by the manufacturer. non-specific fragments in tested samples. After Sequencing was done using Big Dye Terminator the hemi-nested PCR, a fragment of the expected v1.1 Cycle Sequencing kit or Big Dye Terminator size (210 bp) was detected in one pooled sample v.3.1 Cycle Sequencing kit (Applied Biosystems). from Drežnica and in PCR positive control. The After purification of extensions products by purified ~210 bp fragment was sequenced, resul-

304 Hukić et al. A novel flavivirus strain in sandflies

ting in a 155 nt sequence (GenBank accession no. 127 nt long alignment was observed with Fla- MN090154) (36). vivirus RNA isolated from Portuguese sandflies Since we were not able to culture the detected (GenBank: HM563684). In comparison, there unknown virus, basic local alignment analysis was only 71% of sequence identity between 106 and phylogenetic analysis were conducted on compared nucleotides of Drežnica sequence and the obtained 155 nt long sequence of the pooled Phlebotomus flavivirus from Algeria (GenBank: sample that is provisionally named as “Drežnica”. FJ817075). The nucleotide sequence of the sample Drežnica The obtained phylograms provided similar tree showed a maximum of 70-80% sequence identity topologies using two different techniques. Phylo- in 70-88% (110-137 nucleotides) of the query co- genetic analysis indicates that the novel sequence verage, mostly with several sequences of Anophe- Drežnica isolated from Phlebotomus sandflies is les, Sabethes, Calbertado and Culex flaviviruses. more closely related to insects-only flaviviruses For the phylogenetic analysis, nucleotide sequ- than to phleboviruses. ence of Drežnica was compared to a total of 18 DISCUSSION nucleotide sequences (16 flavivirus sequences with the highest sequence identity and two fla- According to the European Centre for Disease viviruses sequence isolated from phlebotomine Control (ECDC) until May 2018, four types of sandflies (GenBank: FJ817075 and HM563684). Phlebotomus sandflies were identified in B&H: Maximum likelihood phylogenetic analysis of P. mascittii, P. neglectus, P. tobbi and P. papatasi. chosen sequences showed that new virus Drež- It is known that the distribution of Phlebotomus nica clusters together with flavivirus (GenBank: sandflies is limited to areas that have a sufficient JF707857) isolated from Culiseta annulata mo- amount of moisture and temperature above 15·6 squitos (Figure 1), however, two sequences have °C for at least three months of the year. Due to cli- only 73% nucleotide identity in 137 nt long ali- mate change, currently many regions of B&H have gnment. The same sequence identity (73%) in summer and autumn temperatures that are suitable for Phlebotomus survival. All previous studies on Phlebotomus sandflies in B&H were conducted only in the area of Herzegovina and this was the first study that confirmed the presence of Phlebo- tomus spp. in the regions of northeast Bosnia. The number of captured Phlebotomus specimens in those regions is a good indicator that sandflies are adapted to the Western Balkan climate, and probably could be found in other parts of B&H. The main focus of this study was to isolate and identify Phlebovirus strains of the captured Phle- botomus sandflies. Using nested amplification strategy, we were not able to detect phlebovirus RNA in any of the 30 tested sandflies pools. However, 155 nt long sequence fragment of Figure 1. Phylogenetic tree based on 155 nucleotide sequence a flaviviral NS5 gene was detected in the pool of MN090154 Drežnica flavivirus strain. The Maximum likeli- of sandflies captured in Drežnica. According to hood phylogenetic analysis is based on a total of 155 nucleotides the composition, the nucleotide sequence of the fragment of 19 sequences. The tree with the highest log likelihood was displayed by using the program MegaX, Tamura-Nei Drežnica sample showed the highest similarity model with 1000 bootstrap replicates and Maximum Composite with nucleotide sequences of several Anopheles Likelihood (MCL) approach. The tree branch lengths measured and Sabethes flavivirus strains. Also, sequence in the number of substitutions per site. GenBank accession numbers for used sequences: MN090154 Drežnica, EU716415, alignment performed with the two flaviviruses FJ817075, HE997074, HF548540, JF707857, JQ409190, RNA sequences isolated from Phlebotomus spp. JX897904, KC464457, KM088037, KT192549, KU707220, showed only a medium level of identity (71-73% KX148547, KX669687, KX907452, MG970163, MH188006, in 106 and 127 nt long coverage). MH899446, and HM563684.

305 Medicinski Glasnik, Volume 17, Number 2, August 2020

Although phlebotomine sandflies are common phlebotomine sample from Drežnica in Herze- vectors of phleboviruses, in 2010 Moureau et al. govina, implies that we detected a new flavivirus reported for the first-time isolation of flavivirus or phlebotomine-related flavivirus, it is necessary RNA from Phlebotomus perniciosus (22). Authors to conduct a more comprehensive study. Future reported that the nucleotide sequences of those no- research should be directed towards isolation and vel Phlebotomine flaviviruses are genetically the culturing Drežnica virus in order to obtain lon- most similar to nucleotide sequences of flaviviru- ger or complete virus sequence that will provide ses isolated from Culex mosquitoes. In the same more information for the virus phylogeny analy- year, another flavivirus RNA sequence isolated sis. Clinical and epidemiological significance from phlebotomine sandflies captured in Portugal of isolated virus of the Drežnica sample is still was submitted to GenBank. This study was not unknown. As imperative, future research should followed with a related publication, and details on also attempt to clarify the pathogenicity of the Phlebotomus species, gender or place of capture Drežnica virus. are not available. Also, the novel flavivirus was for In conclusion, it is necessary to investigate the the first time isolated from New World’s sandflies biological distribution and taxonomic differen- (Psathyromyia abonnenci, previously named Lut- tiation of Phlebotomus spp. across the whole of

zomyia abonnenci) in Equador (37). Besides the Bosnia and Herzegovina, and to indicate areas presence of flaviviruses in Phlebotomus sandflies, with the increased risk of contracting the disease different authors also reported the presence of rhab- transmitted by these insects. doviruses and reoviruses (37-40). Those findings together with the results reported here suggest that FUNDING the presence of flaviviruses, as well as the presence This research was supported by the Ministry of of other viruses in sandflies has not been studied Education and Science of the Federation of Bo- comprehensibly. Zoonotic potential of different vi- snia and Herzegovina (Grant no. 09-3-09-2-350- ruses transmitted by Phlebotomus spp. should be 1/17) through the call titled “Competition for further investigated in the future. financing / co-financing scientific / scientific-re- Based on the results obtained, it seems feasible search and research-development projects in the that we detected a formerly unknown flaviviral Federation of Bosnia and Herzegovina in 2017”, species in phlebotomes with the closest relation within the framework of the project "Genetic di- to other flaviviruses found in Culiseta annulate versity and ecology Sandfly (pappataci) virus in mosquitoes or Phlebotomus perniciosus sandflies Herzegovina". across the Mediterranean basin, with so far unknown zoonotic pathology. TRANSPARENCY DECLARATION Although preliminary analysis conducted on 155 Conflicts of interest: None to declare. nt long virus sequence isolated from the pooled

REFERENCES 1. Karabatsos N. ed. International catalogue of ar- 5. Blitvich B. Arboviruses: Molecular Biology, Evolu- boviruses: including certain other viruses of ver- tion and Control. Nikos Vasilakis and Duane J. Gu- tebrates. 3rd ed. San Antonio, Tex: Published for bler. Am J Trop Med Hyg 2016; 95:488-9. the Subcommittee on Information Exchange of the 6. Alatoom A, Payne D. An overview of arboviruses American Committee on Arthropod-borne Viruses and bunyaviruses. Lab Med 2009; 40:237-40. by the American Society of Tropical Medicine and 7. Hollidge BS, González-Scarano F, Soldan SS. Ar- Hygiene, 1985. boviral encephalitides: transmission, emergence, 2. Weaver SC, Reisen WK. Present and future arbovi- and pathogenesis. J Neuroimmune Pharmacol 2010; ral threats. Antiviral Res 2010; 85:328-45. 5:428-42. 3. Dash AP, Bhatia R, Sunyoto T, Mourya DT. Emer- 8. Chastel C. Asymptomatic infections in man: A Tro- ging and re-emerging arboviral diseases in Southeast jan horse for the introduction and spread of mosqu- Asia. J Vector Borne Dis 2013; 50:77. ito-borne arboviruses in non-endemic areas? Bull 4. Go YY, Balasuriya UB, Lee CK. Zoonotic encepha- Soc Pathol Exot (1990) 2011; 104:213-9. litides caused by arboviruses: transmission and epi- 9. Wesula Olivia L, Obanda V, Bucht G, Mosomtai G, demiology of alphaviruses and flaviviruses. Clin Otieno V, Ahlm C, Evander M. Global emergence of Exp Vaccine Res 2014; 3:58-77. Alphaviruses that cause arthritis in humans. Infect Ecol Epidemiol 2015; 5:29853.

306 Hukić et al. A novel flavivirus strain in sandflies

10. Guler S, Guler E, Caglayik DY, Kokoglu OF, Uc- 27. Ayhan N, Alten B, Ivovic V, Dvořák V, Martinkovic mak H, Bayrakdar F, Uyar Y. A sandfly fever virus F, Omeragic J, Stefanovska J, Petric D, Vaselek S, outbreak in the East Mediterranean region of Turkey. Baymak D, Kasap OE. Direct evidence for an expan- Int J Infect Dis 2012; 16:e244-6. ded circulation area of the recently identified Balkan 11. Sabin AB, Philip CB, Paul JR. Phlebotomus (Pappa- virus (Sandfly fever Naples virus species) in several taci or Sandfly) fever: a disease of military impor- countries of the Balkan archipelago. Parasit Vectors tance summary of existing knowledge and prelimi- 2017; 10:402. nary report of original investigations. J Am Med 28. Ayhan N, Velo E, de Lamballerie X, Kota M, Kadriaj Assoc1944; 125:693-9. P, Ozbel Y, Charrel RN, Bino S. Detection of Leis- 12. Tufan ZK, Weidmann M, Bulut C, Kinikli S, Hufert hmania infantum and a novel phlebovirus (Balkan FT, Dobler G, Demiroz AP. Clinical and laboratory Virus) from sand flies in Albania. Vector Borne Zoo- findings of a sandfly fever Turkey Virus outbreak in notic Dis 2016; 16:802-6. Ankara. J Infect 2011; 63:375-81. 29. Sánchez-Secso MP, Echevarría JM, Hernández L, 13. Palacios G, Tesh RB, Savji N, da Rosa AP, Guzman Estévez D, Navarro-Marí JM, Tenorio A. Detection H, Bussetti AV, Desai A, Ladner J, Sanchez-Secso and identification of Toscana and other phleboviru- M, Lipkin WI. Characterization of the Sandfly fever ses by RT-nested-PCR assays with degenerated pri- Naples species complex and description of a new mers. J. Med. Virol 2003; 71:140-9. Karimabad species complex (genus Phlebovirus, fa- 30. Scaramozzino N, Crance JM, Jouan A, DeBriel DA, mily Bunyaviridae). J. Gen. Virol 2014; 95:292. Stoll F, Garin D. Comparison of flavivirus universal 14. Kocak Tufan Z, Tasyaran MA, Guven T. Sandfly primer pairs and development of a rapid, highly sen- Fever: A Mini Review. Virol Mycol 2013; 2:2161- sitive heminested reverse transcription-PCR assay 0517. for detection of flaviviruses targeted to a conserved 15. Ozbel Y. The infections transmitted by sand flies in region of the NS5 gene sequences. J Clin Microbiol Turkey. Vet Fak Derg 2013; 60:225-8. 2001; 39:1922-7. 16. Nicoletti L, Ciufolini MG, Verani P. Sandfly fever 31. Hall T, Biosciences I, Carlsbad C. BioEdit: an im- viruses in Italy. In: Imported Virus Infections. Vi- portant software for molecular biology. GERF Bull enna: Springer, 1996:41-7. Biosci 2011; 2:60-1. 17. Nicoletti L, Verani P, Caciolli S, Ciufolini MG, Ren- 32. Altschul SF, Madden TL, Schäffer AA, Zhang J, zi A, Bartolozzi D, Paci P, Leoncini F, Padovani P, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and Traini E, Baldereschi M. Central nervous system in- PSI-BLAST: a new generation of protein database volvement during infection by Phlebovirus toscana search programs. Nucleic Acids Res 1997; 25:3389- of residents in natural foci in central Italy (1977– 402. 1988). Am J Trop Med Hyg 1991; 45:429-34. 33. Kumar S, Stecher G, Li M, Knyaz C, Tamura K. 18. Dionisio D, Valassina M, Ciufolini MG, Vivarelli MEGA X: molecular evolutionary genetics analysis A, Esperti F, Cusi MG, Marchi A, Mazzoli F, Lupi across computing platforms. Mol Biol Evol 2018; C. Encephalitis without meningitis due to sandfly 35:1547-9. fever virus serotype Toscana. Clin Infect Dis 2001; 34. Tamura K, Nei M. Estimation of the number of 32:1241-3. nucleotide substitutions in the control region of mi- 19. Hukić M, Saumović-Bešić I. Sandfly-Pappataci fe- tochondrial DNA in humans and chimpanzees. Mol ver in Bosnia and Herzegovina: the new-old disease. Biol Evol 1993; 10:512-26. Bosn J Basic Med Sci 2009; 9:39. 35. Tamura K, Nei M, Kumar S. Prospects for inferring 20. Cusi MG, Savellini GG, Zanelli G. Toscana virus very large phylogenies by using the neighbor-joining epidemiology: from Italy to beyond. Open Virol J method. Proc Natl Acad Sci 2004; 101:11030-5. 2010; 4:22. 36. GenBank. https://www.ncbi.nlm.nih.gov/nuccore/ 21. Özkale Y, Özkale M, Kiper P, Cetinkaya B, Erol I. MN090154 (15 May 2020) Sadfly fever: two case reports. Turk Pediatri Ars 37. Alkan C, Zapata S, Bichaud L, Moureau G, Lemey 2016; 51:110. P, Firth AE, Gritsun TS, Gould EA, de Lamballerie 22. Moureau G, Ninove L, Izri A, Cook S, Lamballerie X, Depaquit J, Charrel RN. Ecuador Paraiso Es- XD, Charrel RN. Flavivirus RNA in phlebotomine condido virus, a new flavivirus isolated from New sandflies. Vector Borne Zoonotic Dis 2010; 10:195- World sand flies in Ecuador, is the first representa- 7. tive of a novel clade in the genus Flavivirus. J Virol 23. GenBank. https://www.ncbi.nlm.nih.gov/nuccore/ 2015; 89:11773-85. HM563684 (15 May 2020) 38. Tesh R, Saidi S, Javadian E, Loh P, Nadim AN. Is- 24. Pick A. Zur Pathologie und Therapie einer eigent- fahan virus, a new vesiculovirus infecting humans, hümlichen endemischen Krankheitsform. Wien gerbils, and sandflies in Iran. Am J Trop Med Hyg Med Wochenschr 1886; 33:1141-5. 1977; 26:299-306. 25. Terzin AL, Matuka S, Fornazaric MR, Hlaca DM. 39. Tesh RB, Modi GB. Growth and transovarial tran- Antibodies against some Arbor-Viruses and against smission of Chandipura virus (Rhabdoviridae: Vesi- the Bedsonia Antigen in Sera of Men, Sheep and culovirus) in Phlebotomus papatasi. Am J Trop Med Cattle in Bosnia and Herzegovina. Acta Med Iugosl Hyg 1983; 32:621-3. 1962; 16:301-17. 40. Vasilakis N, Widen S, Mayer SV, Seymour R, Wood 26. Vesenjak-Hirjan J, Porterfield JS, Arslanagić E, eds. TG, Popov V, Guzman H, da Rosa AP, Ghedin E, Arboviruses in the Mediterranean countries. Stutt- Holmes EC, Walker PJ. Niakha virus: a novel gart: Gustav Fischer Verlag, 1980. member of the family Rhabdoviridae isolated from phlebotomine sandflies in Senegal. Virology 2013; 444:80-9.

307 ORIGINAL ARTICLE

Phenotypic and genotypic detection of ESBL-producing E. coli isolates from chicken skin in Bosnia and Herzegovina

Vildana Hadžić-Hasanović1, Anesa Jerković-Mujkić2, Ermin Hasanović3, Azra Bačić4, Mirsada Hukić5,6

1Hospital for Respiratory Diseases and Tuberculosis, Travnik, 2University of Sarajevo, Faculty of Science, Sarajevo, 3Health Centre Travnik, 4Institute for Biomedical Diagnostic and Research „Genom“, Travnik, 5Academy of Sciences and Arts of Bosnia and Herze- govina, Sarajevo, 6Institute for Biomedical Diagnostics and Research “Nalaz”, Sarajevo; Bosnia and Herzegovina

ABSTRACT

Aim To identify E. coli from chicken meat, establish their antibiotic resistance profiles and to confirm ESBL isolates with real time PCR, as well as to identify risk factors and farming practice associated with the antimicrobial resistance E. coli.

Methods The study included 100 chicken skin samples collected randomly from retail supermarkets, butcheries and slaughterhouses. Disk susceptibility testing was performed using the Kirby-Ba- uer method. Detection of ESBL-producing isolates was performed with double disk synergy test. Molecular analysis of phenotypic ESBL-producing Escherichia coli strains was performed at 7500 Corresponding author: real time PCR System. Molecular-genetic analysis included detec- Vildana Hadžić-Hasanović tion of CTX-M 1, 2, and 9 gene families and mutations in the TEM and SHV encoding extended spectrum β-lactamases. Hospital for Respiratory Diseases and Tuberculosis Results Prevalence of the phenotypic ESBL-producing E. coli iso- Bašbunar 5, 72 270 Travnik, lates was 29%, and they exhibited remarkable sensitivity to carba- Bosnia and Herzegovina penems (100%) as well as to amikacin (93.10%). All ESBL-producing strains were multidrug resistant. Molecular analysis was Phone: +387 30 511 553; performed as the final confirmation of the production of extended Fax: +387 30 511 556; spectrum β – lactamases for 24 isolates out of 29 phenotypicaly E-mail: [email protected] ESBL-producing E. coli isolates. ORCID ID: https://orcid.org/0000-0002- 8504-7056 Conclusion It is important to pay attention to people's awareness of bacterial antimicrobial resistance in food chain, as well as to understand its effects on human health and the environment. Phenotypic and molecular analysis demonstrated the presence of ESBL-producing E. coli isolates from chicken skin samples. Original submission: 15 May 2020; Key words: antimicrobial resistance, food safety, veterinary me- Revised submission: dicine 01 June 2020; Accepted: 16 June 2020 doi: 10.17392/1206-20

Med Glas (Zenica) 2020; 17(2): 308-315

308 Hadžić-Hasanović et al. ESBL-producing E. coli from chicken

INTRODUCTION Furthermore, E. coli encoding for ESBL and ca- rabepenemase are resistant to more than one class Antimicrobial resistance is a complex and mul- of antimicrobials and hence they are multidrug tifaceted problem that threatens human, animal resistant (MDR), presenting a serious challenge health, the global economy, national and global in healthcare settings (17). Enterobacteriaceae security (1). The widespread use of antimicro- that produce ESBLs are a challenging problem in bials in farming practices has been attributed to human and veterinary medicine due to the limita- the presence of antimicrobial resistant bacteria in tion in the treatment options (18). food (2,3). Currently, few data are available regarding the contamination of retail foods with E. The aim of this study was to isolate and identify coli, especially those that are multi-resistant and E. coli from chicken meat, establish their antibio- pathogenic (4). tic resistance profiles and to confirm ESBL isolates with real time PCR, as well as to identify risk Little is known about the antibiotic-resistant factors and farming practice associated with the frequency of microorganisms in poultry that were antimicrobial resistance E. coli. raised by free-range farming, but it is known that poultry industries use antibiotics for therapeutic MATERIALS AND METHODS purposes and for growth promotion (5,6). An important issue in the context of increasing con- Materials and study design sumption and production of poultry meat is to ensure microbial safety (7). This investigation was prospective, experimental and applied. An experimental part was conducted In the community and environment poultry are in the period February 2018 to February 2019 at recognized as an important source for dissemina- the Institute of Biomedical Diagnostic and Re- tion of antimicrobial resistant E. coli (8). search ‘’Genom“ Travnik and the Institute for Both extended spectrum beta-lactamase (ESBL) Biomedical Diagnostics and Research NALAZ, and carbapenemase, which are produced mainly in Sarajevo. Escherichia coli are utilized as crucial resistance This study was approved by the Doctoral Studi- mechanism to cephalosporins and carbapenems (9). es Council of the University of Tuzla, Faculty of Pathogenic E. coli in poultry represent a direct Science and the Ethical Committee. threat to both poultry industry and human health as they may result in hard-to-treat infections (10). Methods Farm animals, especially broiler chickens, affec- Isolation of E. coli from meat samples ted by ESBL-producing E. coli, can be transmitted Collection and preparation of samples, isolation via food chain from animals to humans (11,12). of Escherichia coli from chicken skin (neck), as Direct contact with broilers has been identified well as phenotypic detection of ESBL strains were as a risk factor for carriage of extended-spectrum performed in the Laboratory for Microbiology β-lactamase (ESBL) and plasmidic AmpC of the Institute for Biomedical Diagnostics and (pAmpC) producing E. coli for humans on bro- Research, GENOM Travnik. The study included iler farms (13). In addition, resistant E. coli may 100 chicken skin samples (50 frozen samples and act as transporters for antimicrobial resistant 50 raw samples), collected randomly from retail genes to other pathogens (13). Other pathogens supermarkets, butcheries and slaughterhouses in may receive antimicrobial resistant genes from Travnik, Bosnia and Herzegovina (Table 1). The resistant E. coli who act as a transporter (14). samples were aseptically collected and then pac- Administration of antimicrobial agents in deve- kaged in sterile polythene zip bags and carried to loped countries is not only restricted for the tre- the laboratory in aseptic conditions in a cold box atment purpose, but they can also be used to en- within two hours from the time of purchase. Du- hance animal productivity, growth rate and feed plicate samples were obtained whenever possible. conversion rate in food producing animals (15). All samples were analysed within 2-4 hours after Favourable conditions for persistence and spread their arrival to the laboratory. A sharp sterile kni- strains with genetic traits that can resist to anti- fe was used to cut sample from surface in sterile microbials allow this type of farming (16). tray. To isolate bacteria, a 25-g portion of sample

309 Medicinski Glasnik, Volume 17, Number 2, August 2020

was placed into sterile 225 mL buffered peptone oculums were prepared in sterile saline solutions water. For the preparation of samples, the study from grown cultures. When 0.5 McFarland tur- used guidelines given in the Microbiology of Food bidity value was obtained for each bacterial inoc- and Feeding Standards - Preparation of test sam- ulum, they were inoculated onto Mueller Hinton ples, initial suspensions and decimal dilutions for Agar (Hi Media, India) by sterile cotton swab. microbiological tests (19). Then, amoxicillin/clavulanic acid was placed at After the preparation of the samples, the micro- the centre and the others around it that were far biological analysis was carried out according to from each other for 24 mm from centre to centre. the ISO standard (20). Subsequently, they were incubated at 35 +/- 2ºC for 16-18 h. A clear extension or protrusion (syn- Antibiotic susceptibility testing. Mueller-Hinton ergistic effect) of the edge of the inhibition zone agar (HiMedia, India) was used to perform antibi- of any of the antibiotics toward the disk contain- otic susceptibility testing of all E. coli isolates using amoxicillin/clavulanic acid was interpreted ing 14 antibiotic discs (Mast Group, UK) follow- as positive for ESBL production (Table 1). ing Clinical Laboratory Standards Institute (CLSI) standard (2017) (21): ampicillin 10 µg (AMP), The MAR index was determined for all of 29 cephalexin (30 µg) (CFX); ceftazidime (30 μg) - phenotypic ESBL-producing E. coli strains. The CAZ, imipenem (10 μg) (IMIP), aztreonam (30 results were read by measuring diameter of the in- μg) (ATM), ciprofloxacin (5 μg) (CIP), amikacin hibition zone according to CLSI (21). Therefore, (AK) (30 μg), amoxicillin (25 μg) ( AMX), ceftri- the isolates for each antibiotic were characterized axone (30 µg) (CRO), meropenem (MRP) (10 µg), as: sensitive (S), intermediate (I) and resistant (R). nalidic acid (30 µg), amoxicillin with clavulanic For each phenotypically confirmed ESBL isolate, acid (20 µg-10 µg) (AUG), cefpodoxime (10 µg) a multiple antibiotic resistance indexing (MAR) (CPD), cefotaxime (30 µg) (CTX). E. coli isolates was determined according to the formula: were inoculated in nutrient broth and incubated at The MAR index was calculated so that the nume- 35 +/- 2 ˚C for 16-18 h. rator contains the number of antibiotics to which After antibiotic susceptibility testing of all E. coli the isolate is resistant, and the denominator isolates, strains that had a phenotypical indicati- shows the total number of antibiotics used during on for ESBL production were selected, and the the research. ESBL confirmatory test according to CLSI (21) DNA isolation and identification. DNA isolati- was performed. on from bacterial cells was performed at the In- Extended spectrum β-lactamase (ESBL) constitute for Biomedical Diagnostics and Research firmatory test. The ESBL phenotype was con- "Nalaz" in the Molecular Diagnostic Laboratory firmed by assessing the ability of clavulanic acid in Sarajevo according to a protocol (Institute for (10 μg) to inhibit the activity of cefotaxime (30 Applied Laboratory Analysis Ltd, 2007) (Genial, μg) and ceftazidime (30 μg) in a standard combi- Germany) (22). After the isolation of DNA from nation disk diffusion test: if the addition of clavu- bacterial cells, an identification and amplification lanic acid increased the zone of inhibition by ≥5 procedure was immediately started. At the be- mm when compared to the drug alone, the isolate ginning of the analysis, the number of reactions was defined as an ESBL-producer (21). Kleb- was determined, a positive and a negative control siella pneumoniae ATCC 700603 (Liofilchem, (water) with an IC reagent were prepared, and so- Italy), and as an ESBL negative strain of E. coli lution A was stored at -20 ° C. According to the ATCC 25922 (Microbiologics, USA) were used parameters described in the User Manual Check as control strains. Point (2015) ligation reaction mix was prepared and ligation reaction process was performed (23). After ESBL confirmatory test a double disk synergy test was performed using the following anti- Real time amplification. A novel, rapid real-tibiotic disks: amoxicillin/clavulanic acid (30 μg), me PCR for the detection of ESBL genes in phe- aztreonam (30 μg), cefotaxime (30 μg), ceftazi- notypic ESBL-producing Escherichia coli strains dime (30 μg), ceftriaxone (30 μg) and cefpodox- was used (Check-MDR ESBL PCR; CheckPo- ime (10 μg) (Mast group, UK) (19). Briefly, in- ints) (23). The principle of the Check-Points Real-time PCR assay is based on specific reco-

310 Hadžić-Hasanović et al. ESBL-producing E. coli from chicken

Table 1. Distribution of E. coli and phenotypic ESBL-producing E.coli chicken skin isolates according to the source Frozen (50) samples Fresh (50) samples Chicken skin No (%) of No (%) of E. No (%) of phenotypic No (%) of phenotypic sample source No (%) of samples No (%) of E. coli samples coli ESBL-producing E. coli ESBL-producing E. coli Supermarket 22 (44) 12 (24) 5 (10) 24 (48) 20 (40) 8 (16) Butcheries 18 (36) 8 (16) 1(2) 17 (34) 15 (30) 11 (22) Slaughterhouse 10 (20) 4 (8) 0 11 (22) 5 (10) 4 (8) Total 50 24 6 50 40 23 gnition of target sequences by two adjacent DNA of used antibiotics, for phenotypically detected probes, followed by ligation of these probes. ESBL- and non-ESBL-producing E. coli isolates. Subsequently, the ligated probes were amplified by real-time PCR on an ABI 7500 PCR (Thermo RESULTS Fisher Scientific, USA) instrument using univer- A total of 100 chicken skin samples were sal primers and a molecular beacon. In a typical analysed, of which 50 were fresh and 50 were reaction, the PCR product was formed exponen- frozen meat. We determined 64 E. coli isolates, tially. Because it took several cycles to replicate 24 (48%) frozen meat samples and 40 (80%) fre- enough DNA copies, the curve of the ratio of the sh meat samples (Table 1). E. coli was mostly de- fluorescence intensity to the number of cycles tected in supermarket fresh meat, 12 (24%) and showed a sigmoidal shape. In late cycles, the frozen meat 8 (16%). substrates of the reaction (DNA, dinucleotides, Antimicrobial susceptibility test results enzyme) were depleted, the PCR product no longer doubles and the curves become flat - a "pla- Tested E. coli strains showed 100% resistance to teau". The point on the curve when the fluores- penicillines, nalidic acid and cephalexin (Table cence intensity increases sharply, usually several 2). With regard to the third-generation cephalos- standard deviations above the baseline, is called porins, 100% ceftriaxone resistance was found, the threshold cycle (Ct value). Analyses and data while 96.55% were cefpodoxime resistant. Resi- interpretations were performed according to the stance to aztreonam was found in 79.31% strains. manufacturer’s instructions. The phenotypic ESBL-producing E. coli isolates Molecular-genetic analysis included an examina- exhibited 100% sensitivity to carbapenem and tion of the presence of CTX-M 1, 2, and 9 gene 93.10% to amikacin. Amoxicillin with clavula- families and mutations in the TEM and SHV nic acid showed a sensitivity of 82.75% and in- encoding extended spectrum β-lactamases. For termediate-sensitivity of 6.89% in ESBL-positive samples with FAM CT> 36 or undetermined, and strains. Greater inhibition zone of at least five mm with IC Cy5> 36 or undetermined, the analysis around the disc of cefotaxime with clavulanic was not well performed and the DNA extraction acid (CTXCV) than around the disc of cefotaxime of the bacteria needed to be repeated. Real time (CTX) alone (93.1%), and ceftazidime disc with amplification results were read from the amplifi- clavulanic acid (CAZCV) relative to ceftazidime cation panel (23). (CV) alone (82.75%) was found, indicating the presence of extended β-lactamase (Table 3). Statistical analysis The average MAR index of phenotypic ESBL- Student's t-test was used for testing of parametric producing E. coli strains was 0.70, maximum significance. T test was used to confirm whether 0.78, and the minimum 0.57 (Table 3). there was a statistically significant difference Statistically significant difference between the at the p<0.05 between the inhibition zone sizes sizes of inhibition zones for antibiotics used

Table 2. Indication of the zone inhibition size of 29 phenotypically identified E. coli strains from chicken skin on antibiotics Antibiotic Zone inhibition (mm) IMI MRP AMX AMP AK NA CIP AUG ATM CTX CAZ CRO CPD CFX Average 25.38 26.90 0.41 1.17 20.24 1.14 6.03 18.69 12.76 6.66 5.21 4.79 3.45 0.00 Maximum 30 30 7 10 24 12 26 26 27 27 24 20 17 0 Minimum 23 23 0 0 8 0 0 10 0 0 0 0 0 0 IMI, imipenem; MRP, meropenem; AMP, ampicilin; AMX, amoxicilin; AK, amikacin; NA, nalidixic acid; CIP, ciprofloxacin; AUG, amoxicillin- clavulanic acid; ATM, aztreonam; CFX, cefalexin; CTX, cefotaxime; CAZ, ceftazidime; CRO, ceftriaxone; CPD, cefpodoxime

311 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 3. Antimicrobial profile of phenotypically confirmed ESBL E. coli isolates and multiple antibiotic resistance (MAR) index Antibiotic resistance profile Source of a Sample tag strain CTX CAZ ATM CRO CPD CAZ/CV CTX/CV Other antibiotics MAR INDEX

SM Ch1FR R S R R R - + AMP, AMX, CFX, NA, CIP 0.643 SM Ch2FR R S S R R - + AMP, AMX, CFX, NA, CIP 0.571 SM Ch3FR R R S R R - + AMP, AMX, CFX, NA, CIP 0.643 SM Ch4FR R R R R R + + AMP, AMX, CFX, NA, CIP 0.714 BT Ch5FR R R R R R + + AMP, AMX, CFX, NA, CIP 0.714 SM Ch6FR R R R R R + + AMP, AMX, CFX, NA, CIP 0.714 BT Ch7F R R S R R + + AMX, AMP, CFX, NA 0.571 BT Ch8F R R R R R + + AMX, AMP, CFX, NA,CIP 0.714 SM Ch9F R R S R R + + AMX, AMP, CFX, NA, CIP, AK, AUG 0.786 SM Ch10F R R R R S + + AMX, AMP, CFX, NA, CIP 0.643 SM Ch11F R R S R R + + AMX, AMP, CFX, NA, CIP, AUG 0.714 BT Ch12F R R R R R + + AMX, AMP, CFX, NA, CIP, AUG (I) 0.786 SM Ch13F R R R R R + + AMX, AMP, CFX, CIP, NA, AK 0.786 SL Ch14F R R R R R + + AMX, AMP, CFX, CIP, NA, 0.714 SL Ch15F R R R R R + + AMX, AMP, AK, CIP, CFX, NA 0.786 BT Ch16F R R R R R + + AMX, AMP, CIP,CFX, NA 0.714 SL Ch17F R S R R R - + AMX, AMP, CIP, CFX, NA 0.643 BT Ch18F S R R R R - + AMX, AMP, CIP, CFX, NA, AUG 0.786 SM Ch19F R R R R R + + AMX, AMP, CIP, CFX, NA 0.714 BT Ch20F S R S R R + - AMX, AMP, CIP, CFX, NA, AUG 0.643 BT Ch21F R R R R R + + AMX, AMP, CIP, CFX, NA, 0.714 BT Ch22F R R R R R + + AMX, AMP, CIP, CFX, NA, AUG 0.786 BT Ch23F R R R R R + + AMX, AMP, CIP, CFX, NA, AUG(I) 0.786 SM Ch25F R R R R R + + AMX, AMP. CIP, CFX, NA 0.714 BT Ch36F R R R R R + + AMX,AMP, CIP(I), CFX, NA 0.714 SM Ch26F R R R R R + + AMX, AMP, CIP, CFX, NA 0.714 SM Ch27F R R R R R + - AMX, AMP, CIP, CFX, NA 0.714 BT Ch28F R R R R R + + AMX, AMP, CIP, CFX, NA 0,714 SL Ch29F R R R R R + + AMX, AMP, CFX, NA 0.643 AMX, AMP, CFX, NA (100% R); Overall 82.75 93.1 CIP (89.65% R; 3.4% I); resistance 93.10 89.6 79.31 100 96.55 (+) (+) AUG (17.24% R; 6.89% I); rate AK (10.34% R) AMP, ampicilin; AMX, amoxicilin; CFX, cefalexin; AUG, amoxicillin-clavulanic acid; CAZ/CV, ceftazidim/clavulanic acic; CTX/CV, cefotaxim/ clavulanic acid; CAZ, ceftazidime; CTX, cefotaxime; CPD, cefpodoxime; CRO, ceftriaxone; IMI, imipenem; MRP, meropenem; ATM, aztreonam; CIP, ciprofloxacin; NA, nalidixic acid; AK, amikacin; MAR, multiple antibiotic resistance; SM, supermarket; BT, butcheries; SL, slaughterhouse; F, fresh; FR, frozen; Ch, chicken; S, Sensitive; R, Resistant; I, Intermediate; +, positive ESBL test; -, negative ESBL test

between phenotypically ESBL-producing E. as mutations in SHV and TEM in the total of coli isolates and non-ESBL isolates was found 24 strains. E. coli produces broad spectrum β – for ceftazidim (p<0.05), ampicillin (p<0.05) and lactamases (Table 4). amoxicillin (p<0.05), but not with nalidic acid (p=0.498) and imipenem (p=0.2). DISCUSSION The World Health Organization (WHO) has Molecular detection included antimicrobial resistance as one of the After the detection of phenotypic ESBL-produ- top ten threats to global health in 2019 (24). cing Escherichia coli isolates, the strains were There is no available scientific literature about subjected to molecular analysis. DNA isolation the prevalence of ESBL strains of E. coli in fo- was derived from a total of 29 phenotypic ESBL- ods of animal origin in Bosnia and Herzegovina. producing Escherichia coli strains. Considering recent studies (1-5, 11,12,14), it was Out of 29 analysed isolates, five did not have ge- necessary to carry out a study that would indicate nes for extended spectrum β - lactamase produc- the problem of antibacterial resistance as well as tion. It is clearly seen that the curves reached the the occurrence of ESBL isolates of animal origin amplification threshold of 0.05 over 40 cycles. in our area, especially in chicken sources. In our Molecular analysis showed the presence of ge- study we detected 64% E. coli isolates from chic- nes from the CTX family M-1, 2 and 9, as well ken skin samples. A similar prevalence of E. coli

312 Hadžić-Hasanović et al. ESBL-producing E. coli from chicken

Table 4. ESBL chicken skin E. coli strains confirmed by real obtained MAR index values in our study indica- time PCR analysis with Ct values te that all isolates were from environments that Sample tag C * t are highly contaminated with antibiotics. In our Ch5FR 22.82 Ch6FR 23.75 work, minimum MAR index values were greater Ch7F 25.16 than 0.4 for all isolates. In the study of Hussain Ch8F 23.79 et al. (32), 78.5% of ESBL-producing E. coli was Ch9F 30.22 Ch10F 23.58 found to be multi-drug resistant. Ch11F 26.24 Pacholewicz et al. (33) suggest that management Ch12F 27.36 factors like availability of adequate monitoring Ch13F 27.75 Ch14F 30.66 procedures and food handlers compliance with Ch15F 21.77 these procedures may influence the bacterial con- Ch16F 20.43 centrations on carcasses. They demonstrated that Ch17F 21.22 Ch18F 22.93 compliance with procedures differed between Ch19F 21.17 slaughterhouses, and might be associated with Ch23F 22.56 faecal contamination of carcasses and thus with Ch21F 23.29 Ch26F 26.34 higher bacterial concentrations. Ch1FR 22.89 Molecular analysis of real time PCR was perfor- Ch27F 23.51 Ch28F 28.51 med as the final confirmation of the production Ch29F 20.26 of extended spectrum β – lactamases for 24 of Ch4FR 23.49 29 E.coli isolates. Of 29 analysed isolates, five Ch3FR 21.13 did not detect genes for extended spectrum β - *Ct value up to 31 was considered as positive result Ch, chicken; F, fresh; FR, frozen; Ct value (cycle threshold), number of cycles lactamase production. It is clearly seen that the required for the fluorescent signal to cross the threshold; a relative curves do reach the amplification threshold of measure of the concentration of target in the PCR reaction. 0.05 over 40 cycles. Molecular analysis showed of 56.3% in poultry and beef meat samples was the presence of genes from the CTX family - noted by Eyi & Arslan (25). M-1, 2 and 9, as well as mutations in SHV and Overall prevalence of ESBL–producing E.coli TEM in the total of 24 strains. The difference in isolates was 29% , with the resistance rate of the number of phenotypically and genotypically 100% to penicillines, nalidic acid, cephalexin confirmed isolates can be explained by the limi- and ceftriaxon. None of the isolates were resi- ted PCR detection spectrum of the Check MDR stant to meropenem or imipenem. ESBL kit used in the study. There are much more ESBL gene groups, so we assumed that the used Our study results are similar to the study of Davis kit did not detect other genes also responsible for et al. (26) which proved 87.6% overall prevalen- ESBL production, explaining the difference in ce of resistant E. coli isolates from chicken meat the number of phenotypically and genotypically to 10 of 12 antibiotics tested. None of the isola- confirmed ESBL strains. tes were resistant to amikacin or imipenem. These results are similar to the study results of Rahman Poultry – derived ESBL-producing E. coli isola- et al. (27) who recovered 47.6% ESBL-producing tes in the study of Hussain et al (32) harboured E.coli isolates from chicken meat out of 42 sam- different variants of ESBL genes including ples. Also, a study conducted in Germany by Reich blaCTX-M-15 (40%), bla CTX-M-27 (4%), bla CTX-M-55, et al. (28) reports a high resistance of ESBL produ- bla CTX-M-55 (24%), bla TEM-1 (16%) and bla SHV- 122 cing Escherichia coli from poultry to cefuroxime, (14%). cefotaxime, ceftazidime and ceftriaxone. Skin isolates are exposed to the environment, The value of the multiple resistance index (MAR) thus to a high diversity of micro-organisms, and was assigned to determine the rate of antibiotic have full access to the genetic pool presented by contamination of the analysed meat. According them. It is clear that chances of receiving resi- to different authors (29-31), a MAR index value stance genes against antimicrobial agents are greater than 0.2 indicates that the isolate origi- higher for skin isolates (34). nates from a high-risk source, e.g. from an envi- The increasing resistance of bacteria to the effects ronment where the antibiotic use is frequent. The of antibiotic therapy is a major problem in both

313 Medicinski Glasnik, Volume 17, Number 2, August 2020

veterinary and human medicine. Its solution requ- inappropriate use of antimicrobials by farmers, as ires understanding the development and spread of well as in human and animal medicine. It should be resistance between humans and animals. The high considered that all components of the ecosystem rate of ESBL-producing E. coli is possibly due to are interconnected. Food safety and the fight aga- overuse or consistent usage of antibiotics during inst antibiotic resistance are particularly relevant poultry production. Based on the results of the in One health approach. Only such a synergistic research and the facts presented about antimicro- approach enables the complete collection of data bial resistance as a public health problem, we po- for the rapid and efficient operation and establis- int out the basic factors and strategies that would hment of a system for monitoring and combating help in solving this global problem. It is very im- antimicrobial resistance for the purpose of impro- portant to pay attention to people's awareness of ving public health. antimicrobial resistance as well as to understand its effects on human health and the environment, FUNDING then to continuously monitor and optimize the use No specific funding was received for this study. of antimicrobials in human and animal medicine. In addition, anti-microbial resistance control mea- TRANSPARENCY DECLARATION sures include legal frameworks that will limit the Competing interests: None to declare.

REFERENCES 1. White A, Hughes JM. Critical importance of a One 10. Kaper JB, Nataro JP, and Mobley HT. Pathogenic Es- Health approach to antimicrobial resistance. Eco He- cherichia coli. Nat Rev Microbiol 2004; 2:123–40. alth 2019; 16:404–9. 11. Lazarus B, Dpaterson L, Mollinger JL, Rogers 2. Landers TF, Cohen B, Wittum TE, and Larson EL. BA.“Do human extraintestinal Escherichia coli infec- A review of antibiotic use in food animals: perspec- tions resistant to expanded-spectrum cephalosporins tive, policy and potential. Public Health Rep 2012; originate from food producing animals? A systematic 127:4–22. review. Clin Infect Dis 2015; 60:439-52. 3. Qumar S, Majid M, Kumar N, Tiwari SK, Semmler T, 12. Vitas AI, Naik D, Pérez-Etayo L, González D. Incre- Devi S, Baddam R, Hussain A, Shaik S, Ahmed N. Ge- ased exposure to extended-spectrum –lactamase pro- nome dynamics and molecular infection epidemiology ducing multidrug-resistant Enterobacteriaceae throu- of multidrug-resistant Helicobacter pullorum isolates gh the consumption of chicken and sushi products. Int obtained from broiler and free-range chickens in India. J Food Microbiol 2018; 269:80–6. Appl Environ Microbiol 2017; 83: e2305–16. 13. Huijbers PMC, Graat EAM, Haenen APJ, Van Santen 4. Zhang J, Massow A, Stanley M, Papariella M, Chen MG, Van Essen-Zandbergen A, Mevius DJ, van Duij- X, Kraft B, Ebner P. Contamination rates and anti- keren E, van Hoek A. Extended-spectrum and AmpC microbial resistance in Enterococcus spp., Esche- β-lactamase-producing Escherichia coli in broilers richia coli, and Salmonella isolated from "no anti- and people living and/or working on broiler farms: biotics added"-labeled chicken products. Foodborne prevalence, risk factors and molecular characteristics. Pathog Dis 2011; 8:1147–52. J Antimicrob Chemother 2014; 69:2669–75. 5. Brower CH, Mandal S, Hayer S, Sran M, Zehra A, Pa- 14. Akond M. Antibiotic Resistance of Escherichia coli tel SJ, Kaur R, Chatterjee L, Mishra S, Das BR, Singh isolated from poultry and poultry environment of P, Singh R, Gill PS, Laxminarayan R. The prevalence Bangladesh. Am J Environ Sci 2009; 5:47–52. of extended-spectrum Beta lactamase-producing mul- 15. Snary EL, Kelly LA, Davison HC, Teale CJ, Wool- tidrug-resistant Escherichia coli in poultry chickens dridge M. Antimicrobial resistance: a microbial risk and variation according to farming practices in Pu- assessment perspective. J Antimicrob Chemother njab, India. Environ Health Perspect 2017; 125:1–10. 2004; 53:906–17. 6. Singer RS, Hofacre CL. Potential impacts of anti- 16. Castanon JIR. History of the use of antibiotic as biotic use in poultry production. Avian Dis 2006; growth promoters in European poultry feeds. Poult 50:161–72. Sci 2007; 86:2466–71. 7. Authority, EFS. The European Union summary report 17. Rahman SU, Ali T, Ali I, Khan NA, Han B, Gao J. on trends and sources of zoonoses, zoonotic agents and The growing genetic and functional diversity of food-borne outbreaks in 2015. EFSA J 2016; 14: 4634. extended spectrum beta-lactamases. BioMed Res Int 8. Van Den Bogaard AE, London N, Driessen C, Sto- 2018; 9519718. bberingh EE. Antibiotic resistance of faecal Esche- 18. Projahn M, Pacholewicz E, Becker E, Correia-Carre- richia coli in poultry, poultry farmers and poultry sla- ira G, Bandick N, Kaesbohrer A. Reviewing inter- ughterers. J Antimicrob Chemother 2001; 47:763–71. ventions against Enterobacteriaceae in broiler pro- 9. Meletis G. Carbapenem resistance: overview of the cessing: using old techniques for meeting the new problem and future perspectives. Therap Adv Infec challenges of ESBL E. coli? Biomed Res Int 2018; Dis 2016; 3:15-21. 2018: 7309346.

314 Hadžić-Hasanović et al. ESBL-producing E. coli from chicken

19. International Standard Organisation. Microbiology of 28. Reich F, Atanassova V, Klein G. Extended-Spectrum- food and animal feeding stuffs - Preparation of test β-Lactamase- and AmpC producing enterobacteria in samples, initial suspensions and decimal dilutions for healthy broiler chickens, Germany. Emerg Infect Dis microbiological tests - Part 1: General rules for the 2013; 19:1253-9. preparation of initial suspensions and decimal diluti- 29. Mishra M, Patel AK, Behera N. Prevalence of multi- ons. BAS EN ISO 6887-1:1999. drug resistant E.coli in the river Mahanadi of Sambal- 20. International Standard Organisation. Horizontal met- pur. Curr Res Microbiol Biotechnol 2013; 1:239-44. hod for glucuronidase positive Escherichia coli coun- 30. Thenmozhi S, Rajeswari P, Suresh Kumar T, Sai- ting - Part 2: BAS EN ISO 16649-2:2001. priyanga V, Kalpana M. Multi-drug resistant patterns 21. Clinical Laboratory Standards Institute. Performance of biofilm forming Aeromonas hydrophila from urine Standards for Antimicrobial Susceptibility Testing. samples. Int J Pharm Sci Res 2014; 5:2908-18. 27th ed. Tests for ESBLs.Wayne PA: CLSI, 2017. 31. Joseph AA, Odimayo MS, Olokoba LB, Olokoba AB, 22. Institute for applied laboratory analysis LTD. All-ti- Popoola GO. Multiple antibiotic resistance index of ssue DNA-extraction kit. Genial Germany 2007, 11. Escherichia coli isolates in a tertiary hospital in south- 23. Check-MDR ESBL PCR Check Points. User Manual. west Nigeria. Medical Journal of Zambia 2017; 44: Check Point The Netherlands 1.3, 2015. 225-32. 24. World Health Organization. Ten threats to global he- 32. Hussain A, Shaik S, Ranjan A, Suresh A, Sarker N, alth in 2019 https://www.World Health Organization. Semmler T, Wieler Lothar H, Alam M, Watanabe H, int/emergencies/ten-threats-to-global-h ealth-in-2019 Chakravortty D, Ahmed N. Genomic and functional (20 April 2020) characterization of poultry Escherichia coli from In- 25. Eyi A, Arslan S. Prevalence of Escherichia coli in re- dia revealed diverse extended-spectrum β-lactamase- tail poultry meat, ground beef and beef. Med Weter producing lineages with shared virulence profiles. 2012; 68. Front Microbiol 2019; 10. 26. Davis GS, Waits K, Nordstrom L. Grande H, Weaver 33. Pacholewicz E, Barus SAS, Swart A, Havelaar AH. B, Papp K, Horwinski J, Koch B, Hungate BA, Liu Influence of food handlers' compliance with procedu- CM, Price LB. Antibiotic-resistant Escherichia coli res of poultry carcasses contamination: a case study from retail poultry meat with different antibiotic use concerning evisceration in broiler slaughterhouses. claims. BMC Microbiol 2018; 18:174. Food Control 2016; 68: 367-78. 27. Rahman SU, Ahmad S and Khan I. Incidence of 34. Ghyselinck J. Antimicrobial resistance in human and ESBL-producing-Escherichia coli in poultry farm broiler chicken Escherichia coli isolates. Ghent: Uni- environment and retail poultry meat. Pak Vet J 2019; versiteit Gent, 2008; Ph. D. thesis. 39:116-20.

315 ORIGINAL ARTICLE

Association of endothelial nitric oxide synthase gene (G894T) polymorphism and hypertension in diabetic Bataknese patients

Jelita Siregar1, Ratna Akbari Ganie1, Dharma Lindarto2, Erna Mutiara3, Delfitri Munir4

1Department of Clinical Pathology, 2Department of Internal Medicine, 3Department of Population and Biostatistics, 4Department of Otorhinolaryngology, Head and Neck Surgery; School of Medicine, Universitas Sumatera Utara, Medan, Indonesia

ABSTRACT

Aim To assess endothelial nitric oxide synthase (eNOS) gene (G894T) polymorphism and nitric oxide (NO) level in hypertensive diabetic Bataknese patients.

Methods A hospital-based, case control study (hypertensive and normotensive diabetic patient) was conducted. Genotyping of eNOS gene (G894T) was done using polymerase chain reaction- restriction fragment length polymorphism (PCR-RFLP). Nitric oxide was quantified by sandwich enzyme-linked immunosorbent assay (Sandwich ELISA).

Corresponding author: Results GT polymorphism and T allele were higher in the hypertensive diabetic patients, 37.5% (p=0.6) and 54% (p=0.39), Jelita Siregar respectively. Nitric oxide level tended to be lower in the hyperten- Department of Clinical Pathology, sive diabetic patients (88.87µmol/L) comparing to the normoten- School of Medicine, sive (95.42 µmol/L (p=0.54), as well as GT and TT polymorphism Universitas Sumatera Utara type (p=0.75). dr. Mansyur 5, Medan, Indonesia Conclusion eNOS gene (G894T) polymorphism is not associated Phone: +62 61 821 1045; with NO level and hypertension in the diabetic Bataknese patients. Fax: +62 61 821 6264; E-mail: [email protected] Key words: diabetes mellitus, hypertension, nitric oxide, nitric oxide synthase type III, polymorphism ORCID ID: https://orcid.org/0000-0002- 6054-913X

Original submission: 11 October 2019; Revised submission: 20 January 2020; Accepted: 21 February 2019 doi: 10.17392/1088-20

Med Glas (Zenica) 2020; 17(2): 316-320

316 Siregar et al. eNOS in hypertensive diabetic Bataknese patients

INTRODUCTION PATIENTS AND METHODS Hypertension is one of the major public health pro- Patients and study design blems in the world. It is called “silent killer” since it may have no warning signs and symptoms (1). This was a hospital-based, case control study of Around 31% of all adults worldwide suffer from eNOS gene polymorphism in T2DM patients. hypertension. While prevalence decreases in high- Samples were all Bataknese T2DM patients who income countries, hypertension increases in low/ visited Internal Medicine Department of Univer- middle-income countries (2). In 2015, prevalence sitas Sumatera Utara Hospital, Medan, Indonesia of hypertension in Indonesia was 33.4% (3). Hyper- from June to July 2019. tension could cause numerous health problems to Patients with T2DM and hypertension were in- human body. One of the most deadly complications corporated in a case group, while patients with of hypertension is cardiovascular disease, which T2DM but without hypertension (normotensive) accounts for 35% of deaths in Indonesia (4). were included in a control group. Hypertension Hypertension is associated with several risk fac- was defined as having systolic blood pressu- tors. Type 2 diabetes mellitus (T2DM) is one of re ≥140 mmHg or diastolic blood pressure ≥90 the common underlying diseases of hypertensi- mmHg (17). Cases and controls were matched on. When hypertension and T2DM co-exist, the on sex, age, HbA1C and plasma glucose level. risk of stroke or coronary arterial disease (CAD) Inclusion criteria were T2DM patients and Ba- will increase 2-4 times (5,6). Pathogenesis of taknese ethnicity (pure ethnic for three gene- hypertension in T2DM is a complex process. ration). T2DM was defined as having fasting Endothelial dysfunction is the main cause of car- plasma glucose (FPG) ≥126 mg/dL and 2 hours diovascular complication in T2DM (7,8). Hyper- post-prandial plasma glucose (2h PP) ≥200 mg/ glycaemia, free fatty acid, and insulin resistance dL (18). Bataknese is an ethnic group originating reduce nitric oxide (NO) and increase endothe- from Medan (19). In this study, we included only lin-1 that cause prolonged vasoconstriction lea- pure descendants of Batak tribe whose parents ding to hypertension (9). and grandparents are of pure Bataknese ethnicity. Nitric oxide synthesis is catalysed by nitric oxide Patients with secondary hyperglycaemia (e.g. synthases (NOS). Endothelial NOS (eNOS) is corticosteroid medication) and kidney complica- an NOS isoform responsible for catalysing the tions were excluded. formation of NO from l-arginine in endothelial This research was approved by the Ethical Com- cells, and therefore plays a central role in vas- mittee of School of Medicine, Universitas Suma- cular function (10). The eNOS is expressed by tera Utara. the eNOS gene. Glu298Asp (aspartate instead of glutamate in codon 298) as a result of G894T (su- Methods bstitution of guanine by tymin in position 894 at Demographic data, glucose level, NO level, exon 7) is an eNOS gene polymorphism known eNOS gene polymorphism were collected in this to be responsible for altered NO production and study. Sex and age were obtained from the medi- to cause some cardiovascular problems, such as cal record. HbA1C and blood plasma level were hypertension and CAD (11-14) and in some the- examined by Automatic Cobas 6000 C (Roche ories also causes T2DM (15-16). Diagnostics, Mannheim, Germany) according A study on eNOS gene polymorphism in hyper- to manufacturer’s instruction. Nitric oxide was tensive diabetic patients has never been conduc- quantified by sandwich enzyme-linked immu- ted on Bataknese ethnicity, Indonesia. Knowled- nosorbent assay (Sandwich ELISA) on serum ge in this field is important to determine better sample. Six ml blood was taken in non-ethyl- management for this population. enediaminetetraacetic acid (EDTA) vacutainer The aim of this study was to assess eNOS gene and processed to obtain serum. Endothelial NOS (G894T) polymorphism and NO level in hyper- gene (G894T) polymorphism was examined by tensive and normotensive diabetic Bataknese pa- deoxyribonucleic acid (DNA) isolation, contin- tients in Indonesia. ued by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP).

317 Medicinski Glasnik, Volume 17, Number 2, August 2020

Isolation of DNA was conducted by using Wizard Table 1. Patient characteristic and nitric oxide (NO) level in Genomic DNA Purification Kit (Promega, Madi- the case and control group Case group Control group Variable p son, WI). PCR was done in 30 µl PCR mixture (Mean±SD) (Mean±SD) consisting of 15 µL GoTaq Green Master Mix Age (years) 59.32 ± 8.562 57.71 ± 10.302 0.371 , HbA1C (%) 8.305 ± 1.875 7.659± 1.725 0.060 (Promega) 11 µL ddH2O, 1 µl forward primer (5’-AACCCCCTCTGGCCCACTCCC-3’) FPG (mg/dL) 184.93 ± 73.322 165.29 ± 71.126 0.153 2hPP (mg/dL) 247.38 ± 92.994 228.41 ± 77.670 0.244 10 µM and 1µL reverse primer (5’-TC- NO (µmol/L) 88.87±54.359 95.42±57.133 0.536 CATCCCACCCAGTCAAT-3’) 10 µM and 2 µL FBG, fasting plasma glucose; 2hPP, 2 hours post-prandial plasma DNA (15). The mixture was incubated initially glucose at 95 °C for 8 minutes, followed by 40 cycles of µmol/L) (Table 1). More GT polymorphism denaturation at 94 °C for 1 minute, annealing at (37.5%) was seen in the case group than the con- 63 °C for 45 seconds, extension at 72 °C for 1 trol group (29.57%), while more GG polymor- minute and final extension at 72 °C for 7 minutes. phism (60.71%) was seen in the control group After that, RFLP was carried out by incubating than the case group (60.71%). Moreover, there 5 µL PCR product, 0.5 µL 10× Buffer R (with was no association between polymorphism type BSA) (Thermoscientific), 3.5 µl H O and 0.5 µL 2 and NO level (GG mean 95.02µmol/L, GT 87.09 Mbol restriction enzyme (Thermoscientific) in 37 µmol/L, TT 80.80 µmol/L, p=0.75), although o C for 1 hour (15). The electrophoresis of PCR- the level tended to be lower in GT and TT poly- RFLP products was performed in 3% agarose gel morphism. Genotypic distribution of the G894T containing ethidium bromide (0.5 μg/mL) and polymorphism was in Hardy-Weinberg Equilib- photographed using the gel documentation sys- rium (HWE) among both case and control group tem. Results were categorized into homozygote (p = 0.568 and p=0.265, respectively) (Table 2). GG, which showed single 206 bp band, homozygote TT 119 bp and 87 bp bands, and heterozy- Table 2. Polymorphism type in the case and control group gote 206 bp, 119 bp and 87 bp bands. eNOS gene No (%) of patients in the group Total p polymorphism Case Control Statistical analysis Genotype GG 34 (60.71) 39 (69.64) 73 GT 21 (37.50) 16 (28.57) 37 0.601 Bivariate analysis between age, HbA1C and pla- TT 1 (1.79) 1 (1.79) 2 sma glucose level in case and control group were Allele G 89 (79.46) 94 (83.93) 183 0.388 T 23 (20.54) 18 (16.07) 41 done using independent t-test. Analysis of nitric eNOS, endothelial nitric oxide synthase oxide level and eNOS gene polymorphism between case and controls were done using independent DISCUSSION t-test and chi-square test, respectively. eNOS gene To our knowledge, this is the first study assessing polymorphism was analysed with nitric oxide le- genetic background of hypertension in T2DM pa- vel by using ANOVA. Hardy-Weinberg equilibri- tients in Indonesia. Moreover, a study of eNOS um was performed in both case and control group gene polymorphism has never been performed in for the analysed polymorphism. a Bataknese patient who resides in the western part of Indonesia. Only diabetic patients were included RESULTS because the combination of T2DM and hyperten- A total of 112 patients were included in this study sion could lead to worse complications. Special with 56 patients in both case and control group. management is needed for this population. The case group consisted of 23 male and 33 fema- Nitric oxide is well-known for its capability to le patients, while the control group had 24 male reduce the risk of cardiovascular disease in hu- and 32 female patients. All patients’ characteristics mans, including hypertension. T2DM is a risk were not different statistically (Table 1) (p>0.05). factor of low NO level in humans. Low NO level No significant differences between the groups in T2DM could cause vasoconstriction resulting were observed both in NO level (p=0.536) and in hypertension (9). In this study, we found no eNOS gene polymorphism (p=0.601). In the case difference of NO level in hypertensive and nor- group, NO level (mean 88.87 µmol/L) tended motensive diabetic patients. However, mean NO to be lower than the control group (mean 95.42 level tended to decrease in hypertensive diabetic

318 Siregar et al. eNOS in hypertensive diabetic Bataknese patients

patients. Although NO is usually higher in both specific, therefore, the management should be diseases when they occur separately (20), the made based on the ethnic group. This study indi- study on NO level in co-existence of T2DM and cates that the management of hypertension might hypertension is scarce and conflicting. Manju not have to consider G894T polymorphism in et al. stated that NO level was lower in diabe- Bataknese diabetic patients. tic patients as compared to non-diabetic patients, Furthermore, the effect of G894T polymorphism but failed to demonstrate significant difference on NO level was studied. In this study no associa- when comparing hypertensive and normotensive tion was found. Single nucleotide polymorphism diabetic patients (21). Shahid et al. found lower (from guanin to tymin) happening in eNOS gene NO level in hypertensive diabetic patients than was known to cause reduction of NO level in hu- in normotensive diabetic patients (22). Given the mans. However, it was not observed in Asian eth- fact that our study and published studies are case nicity and CAD (28). In this study, patients inclu- control studies with a low number of patients and ded also had T2DM in addition to hypertension. a high number confounding factors, an exact cau- T2DM as an underlying disease could contribute sality could not be made. A cohort study should to lower NO level in both groups. Also, other fac- be conducted in the future to assess this problem. tors affecting NO level were not assessed. Causal The effect of eNOS gene polymorphism on hyper- relationship could not be drawn in this study. tensive and normotensive diabetic patients was This study has several limitations to be conside- also assessed. We found more GT polymorphism red for interpretation. Several confounding fac- types than GG polymorphism types in hyperten- tors were not assessed. Since hypertension and sive diabetic patients, although the difference was T2DM are of multifactorial etiology, the cause of not significant. T allele also tended to be higher in NO elevation may not be only a result of gene hypertensive diabetic patients. G894T polymorp- polymorphism. For example, as the mean age of hism was associated with hypertension in some patients in both groups was above 50 years, we ethnicities. Asian, especially Chinese and Indian could suspect that the environmental effect could ethnicities were associated with a higher risk for mask the effect of genetic background. In a next hypertension caused by this genetic predisposi- study, younger patients and several confounding tion (23,24). In some areas outside Asia, such as factors should be addressed to produce better re- Egypt, the association was also observed (13). sults. Also, the case-control nature of this study Allele T was associated with the higher risk for prevents us from making a conclusion on causal hypertension, especially in females (25). Howe- relationship of G894T polymorphism and hyper- ver, in several studies, for example in Sudan (26), tension in diabetic patients. among East Asian (27), and among Whites (23), no association was found. In Indonesia, one stu- FUNDING dy showed no association between eNOS gene No specific funding was received for this study. (G894T) polymorphism and essential hypertension in Javanese patients, who live on a separated TRANSPARENCY DECLARATION island with Bataknese. We have concluded that Conflicts of interest: None to declare. the effect of G894T polymorphism is ethnicity

REFERENCES 1. World Health Organization. Cardiovascular Dise- 4. World Health Organization. Indonesia. 2019. https:// ases. 2019. https://www.who.int/news-room/fact- www.who.int/nmh/countries/idn_en.pdf (01 July sheets/detail/hypertension (01 July 2019) 2019) 2. Bloch MJ. Worldwide prevalence of hypertensi- 5. Lonati C, Morganti A, Comarella L, Mancia G, Za- on exceeds 1.3 billion. J Am Soc Hypertens 2016; nchetti A, IPERDIA Study Group. Prevalence of 10:753-4. type 2 diabetes among patients with hypertension 3. Peltzer K, Pengpid S. The prevalence and social under the care of 30 Italian clinics of hypertension: determinants of hypertension among adults in In- results of the (Iper)tensione and (dia)bete study. J donesia: a cross-sectional population-based national Hypertens 2008; 26:1801-8. survey. Int J Hypertens 2018; 5610725.

319 Medicinski Glasnik, Volume 17, Number 2, August 2020

6. Cheung B. 2010. The hypertension-diabetes conti- 17. Chobanian AV, Bakris GL, Black HR, Cushman nuum. J Cardiovasc Pharmacol 2010; 55:333-9. WC, Green LA, Izzo Jr JL, Jones DW, Materson BJ, 7. Avogaro A, Albiero M, Menegazzo L, de Kreutze- Oparil S, Wright Jr JT, Roccella EJ. The seventh re- nberg S, Fadini GP. Endothelial dysfunction in dia- port of the joint national committee on prevention, betes: the role of reparatory mechanisms. Diabetes detection, evaluation, and treatment of high blood care 2011; 34(Supplement 2):S285-90. pressure: the JNC 7 report. J Am Med Assoc 2003; 8. Tan KC, Chow WS, Ai VH, Lam KS. Effects of 289:2560-71. angiotensin II receptor antagonist on endothelial 18. American Diabetes Association. Classification and vasomotor function and urinary albumin excretion diagnosis of diabetes. Diabetes Care 2015; 38:8-16. in type 2 diabetic patients with microalbuminuria. 19. Simanjuntak BA. Struktur sosial dan sistem politik Diabetes Metab Res Rev 2002; 18:71-6. Batak Toba hingga 1945 (Toba Batak social structu- 9. Luscher TF, Creager MA, Beckman JA, Cosentino re and political system until 1945) [Indonesian]. Ja- F. Diabetes and vascular disease: pathophysiology, karta: Yayasan Obor Indonesia, 2006. clinical consequences, and medical therapy: Part I. 20. Ayub T, Khan SN, Ayub SG, Dar R, Andrabi KI. Re- Circulation 2003; 108:1527-32. duced nitrate level in individuals with hypertension 10. Leineweber K, Moosmang S, Paulson D. Genetics of and diabetes. J. Cardiovasc Dis Res 2011; 2:172-6. NO deficiency. Am J Cardiol 2017; 120(8S):S80-8. 21. Manju M, Mishra S, Toora BD. Relationship betwe- 11. Men C, Tang K, Lin G, Li J, Zhan Y. ENOS-G894T en glycosylated hemoglobin, serum nitric oxide and polymorphism is a risk factor for essential hyper- mean arterial blood pressure. Int J Biomed Sci 2014; tension in China. Indian J Biochem Biophys 2011; 10:252-7. 48:154-7. 22. Shahid SM, Mahboob T. Diabetes and hypertensi- 12. Moraes MP, eSilva KSF, Lagares MH, Barbosa AM, on: correlation between glycosylated hemoglobin Martins JVM, Campedelli FI, da Costa IR, Rodri- (HbA1c) and serum nitric oxide (NO). Aust J Basic gues DA, Moura KKVO. Polymorphism of the ge- Appl Sci 2009; 3:1323-7. nes eNOS, GSTT1 and GSTM1 are significantly 23. Niu W, Qi Y. An updated meta-analysis of endothe- associated with atherosclerotic disease in hyperten- lial nitric oxide synthase gene: three well-characte- sive patient. Genet Mol Res 18:GMR18089. rized polymorphisms with hypertension. PloS One 13. Arafa S, Abdelsalam S, El-Gilany AH, Mosaad YM, 2011; 6:e24266. Abdel-Ghaffar A. Endothelial nitric oxide synthase 24. Shankarishan P, Borah PK, Ahmed G, Mahanta J. Glu 298 Asp (G894T) and apolipoprotein E gene Endothelial nitric oxide synthase gene polymorphi- polymorphism as possible risk factors for coronary sms and the risk of hypertension in an Indian popu- heart disease among Egyptians. Egypt Heart J 2018; lation. Biomed Res Int 2014; 2014:793040 70:393-401. 25. Li J, Cun Y, Tang WR, Wang Y, Li SN, Ouyang HR, 14. Zhu B, Si X, Gong Y, Yan G, Wang D, Qiao Y, Liu Wu YR, Yu HJ, Xiao CJ. Association of eNOS gene B, Hou J, Tang C. An association between the endo- polymorphisms with essential hypertension in the thelial nitric oxide synthase gene G894T polymorp- Han population in southwestern China. Genet Mol hism and premature coronary artery disease: a meta- Res 2011; 10:2202-12. analysis. Oncotarget 2017; 8:77990-8. 26. Gamil S, Erdmann J, Abdalrahman IB, Mohamed 15. Moguib O, Raslan HM, Rasheed IA, Effat L, Mo- AO. Association of NOS3 gene polymorphisms with hamed N, El Serougy S, Hussein G, Tawfeek S, Ab- essential hypertension in Sudanese patients: a case delRahman AH, Omar K. Endothelial nitric oxide control study. BMC Med Genet 2017; 18:128. synthase gene (T786C and G894T) polymorphisms 27. Zintzaras E, Kitsios G, Stefanidis I. Endothelial NO in Egyptian patients with type 2 diabetes. J Genet synthase gene polymorphisms and hypertension: a Eng Biotechnol 2017; 15:431-6. meta-analysis. Hypertension 2006; 48:700-10. 16. Angeline T, Krithiga HR, Isabel W, Asirvatham AJ, 28. Luo Z, Jia A, Lu Z, Muhammad I, Adenrele A, Song Poornima A. Endothelial nitric oxide synthase gene Y. Associations of the NOS3 rs1799983 polymorp- polymorphism (G894T) and diabetes mellitus (type hism with circulating nitric oxide and lipid levels: a II) among South Indians. Oxid Med Cell Longev systematic review and meta-analysis. Postgrad Med 2011; 2011:462607. J 2019; 865:361-71.

320 ORIGINAL ARTICLE

The effect of bay leaf extract (Syzygium polyanthum) on vascular endothelial growth factor (VEGF) and CD31 (PECAM-1) expression in acute coronary syndrome Refli Hasan1, Gontar Alamsyah Siregar1, Dharma Lindarto1

1Department of Internal Medicine, School of Medicine, Universitas Sumatera Utara, Medan, Indonesia

ABSTRACT

Aim To investigate effect of bay leaf extract in endothelial integrity, observed by vascular endothelial growth factor (VEGF) level, VEGF and CD31 expression.

Methods Thirty-two acute coronary syndrome surgery-induced Wistar Rats (Rattus novergicus) were divided into 16 bay leaf extract (treatment) groups and 16 control groups, sacrificed on day 1, 4, 7, and 14 after the induction. Serum VEGF level was determined by ELISA and expression of VEGFR-2 and CD31 were detected on immunohistochemistry. Corresponding author: Refli Hasan Results This study showed increased expression of serum VEGF level, and VEGFR-2 expression was found significantly on day Department of Internal Medicine, 7 and 14 in the treatment group compared to the control group. School of Medicine, CD31 expression was significantly different compared to the con- Universitas Sumatera Utara trol groups on day 4, 7, and 14 of administration. dr. Mansyur 5, Medan, Indonesia Conclusion The potential effect of bay leaf extract on angioge- Phone: +62 61 8211 045; nesis in acute coronary syndrome (ACS) as adjuvant for the tre- @ E-mail: refli_hasan yahoo.co.id atment. Bay leaf extract has been shown to support angiogenesis ORCID ID: https://orcid.org/0000-0003- and maintain endothelial integrity that leads to better prognosis for 3738-117X reperfusion on ischemic tissue.

Key words: ischemia, flavonoids, herbal medicine

Original submission: 19 May 2020; Accepted: 12 June 2020 doi: 10.17392/1208-20

Med Glas (Zenica) 2020; 17(2): 321-327

321 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION (CD31), is considered as one of the most important mechanosensors, which is highly expressed at en- Cardiovascular disease is projected to be the dothelial cell junctions. In thrombotic process that single leading cause of death. The World Health challenges the shear system, CD31 is responsive Organization (WHO) stated that by 2030, almost for direct transduction of mechanical forces, not 23.6 million people would suffer from CHDs, only in the macrophages but also in the endothe- mainly from heart disease and stroke (1). As an lial cells (10). CD31 signals to regulate all critical ever-growing developing country, Indonesia fa- aspects related to the pathogenesis of atherosclero- ces an increasing trend of non-communicable sis: from leukocyte detachment, T-cell activation, disease (NCD) cases (2). This explains the fact platelet activation, to angiogenesis (12). Corres- published by the Ministry of Health of the Re- ponding to VEGF level, high CD31 expression is public of Indonesia in 2018 that stated coronary related to its function to protect vascular endothe- heart disease (CHD) as one of the diseases with lial barrier integrity (13). A study by Xia has pro- the highest prevalence in Indonesia, leading to ven that low ESS promotes expression of CD31 the second highest number of deaths in Indone- in vivo, in atherosclerotic rat models (14). The sia (3). As the highest cause of mortality, more study was done by constricting abdominal aorta; studies are needed to investigate further on the and opportunities are open to assess whether the pathogenesis of CHD, specifically on interventi- result is analogous to occur on coronary arteries or ons needed for the prevention of CHD cases and/ acute coronary syndrome-induced models in vivo. or the progression of the disease. Serum VEGF levels, VEGFR-2 and CD31 expre- One of the important parts of CHD pathogenesis is ssions in heart tissue were chosen as inflammatory arteries’ supply blockage caused by atherosclero- markers in this study. sis (4). The progression of atherosclerosis begins Indonesia has great resources of plants biodiver- in childhood or adolescence to progress over time. sity, due to its geographic location along Equator Atherosclerosis would accumulate to the point of line. Indigenous people residing in Indonesia have shear stress as its essential feature (5). This wo- consumed traditional herbs for health. Empirically, uld correlate to the level of endothelial shear stress bay leaf (Syzygium polyanthum) or salam leaf (in (ESS), as low ESS induces inflammation (tribute Bahasa Indonesia) stew was used as home remedy to its potent proinflammatory profile) and signals for hypercholesterolemia, diabetes mellitus, and atherogenesis process (6). These processes are hypertension. Bay leaf contains flavonoid (15), associated with the development and progressi- which is a natural antioxidant polyphenol com- on of coronary atherosclerosis (7). In response, pound and proven to protect arterial endothelium this condition further upregulates the expression by preventing damage of the cells and reducing of potent vascular smooth muscle cells (VSMC) cholesterol deposition on the surface of the arte- mitogens, including growth factors such as vas- rial endothelium (16,17). In response to the shear cular endothelial growth factor (VEGF) (8). The stress, VEGF and CD31 are upregulated to secure expressions of these factors are amplified with endothelial integrity. A study published by Ba- the ongoing formation of reactive oxygen species ssino proved that flavonoid promoting VEGFR2 (ROS) and pro-inflammatory cytokines signalled phosphorylation was significantly increased (18); from low ESS (9). In the condition of shear stress, specifically, the study proved it occurs only by sti- mechanosensors play an important role. A study mulation with HSP (Hesperidin), one example of by Tzima et al. has identified a mechanosensory flavonoid compound. complex in atherosclerotic plague pathophysiology consists of CD31, vascular endothelial– The aim of this experimental study was to de- cadherin (VE-cadherin) and vascular endothelial monstrate protective effect of bay leaf extract on growth factor receptor-2 (VEGFR2) that respon- endothelial integrity on rat model with myocar- ded towards external shear stress (10), with regar- dial infarction. Hypothesis of the study accentu- ds to VEGF receptor 2 (VEGFR2) as the principal ates that the bay leaf administration would target receptor of VEGF in blood vessels (11). the atherogenesis by targeting mechanosensory pathway, which involves CD31 and VEGFR2 Platelet endothelial cell adhesion molecule (PE- expressions. CAM-1), known as a cluster of differentiation 31

322 Hasan et al. Bay leaf on VEGF and CD31 in ACS

MATERIALS AND METHODS Science, US). With sandwich-ELISA principle, the micro plate wells had been pre-coated with an anti- Sample and study design body specific to Rat VEGF-A. After samples were added to the wells and combined with the specific This study involved thirty-two Wistar rats (Rattus antibody, a biotinylated detection antibody specific novergicus) weighing around 200 grams, 10-12 for Rat VEGF-A and avidin-horseradish peroxida- weeks of age. Rats were divided into two groups se (HRP) conjugate were added to each well and consisting of sixteen rats each, namely treatment incubated. Free components were washed away, and control groups. Rats were housed in steel then substrate solution was added. Wells conta- cage given free access to self-feed and conditio- ining Rat VEGF-A, biotinylated detection anti- ned with 12 hours dark/light cycle at temperature body and Avidin-HRP conjugate appeared blue in of 20-25 °C. Samples resided in the Animal Ho- colour. The enzyme-substrate reaction was halted use, Research Laboratory, School of Medicine, by adding the stop solution when the colour of the Universitas Brawijaya, Malang, Indonesia. All solution in wells turned yellow. The optical density procedures were conducted in December 2019. (OD) value was measured at the wavelength of 450 The institutional Ethics Committee of Universi- nm. The OD value was proportional to the concen- tas Sumatera Utara has approved procedures and tration of VEGF level. Therefore, the concentration interventions conducted in this study. of VEGF level was calculated as a result of OD va- Methods lue of the samples compared to the standard curve. Surgical induction of myocardial infarction. Immunohistochemistry staining of VEGFR-2 To induce samples with acute coronary syndro- and CD31. The tissue specimens were embed- me, coronary arteries of rats were ligated, accor- ded by paraffin and cut into four to six micron of ding to a study by Wu et al. (19). Ketamine was thickness using microtome. Slides were deparaffi- given as an anaesthetic agent. Samples from both nised, rehydrated, and then heated on microwave groups underwent thoracotomy followed by liga- with EDTA antigen repair solution for 20 minutes. tion of anterior descending (LAD) artery, where Slides were left to cool by putting it in cold water, sign of infarct is proven by blanching in myocar- and rinsed three times using phosphate buffered sa- dium presented in distal of the ligation site. The line (PBS), pH 7.4, each 3 minutes long. Hydrogen surgical wound was later sutured. peroxide 3% solution was incubated for 15 minutes to halt endogen peroxidase activity, and then Administration of bay leaf (Syzygium the slides were rinsed with the respective method. polyanthum) extract. An extract of Syzygium Primary antibody was detected by biotinylated polyanthum was prepared by maceration in the detection antibody and avidin-conjugated horse- Research Unit laboratory, School of Medicine, radish peroxidase (HRP) from then rinsed. Anti- Universitas Brawijaya, Malang, Indonesia. The VEGFR-2 (sc-6251, Santa Cruz Biotechnology, suspension was given in a dose of 3.6 mg (0.72 USA) and Anti-CD31 (sc-376764, Santa Cruz Bio- mL in volume), administered via orogastric tube technology, USA) were added at room temperature and syringe. for 20 minutes, then rinsed. A signal was detected Sample collection. Samples were sacrificed on using diamino-benzidine (DAB) solution. day one, four, seven, and fourteen after admini- The slides were then counterstained with hema- stration. Blood samples were collected from the toxylin followed by histological slides exami- coronary artery of all rats and followed with sur- nation by two blinded pathologists. Final slides gical removal of the heart. Blood samples were score was calculated as a result of the positive subjected to measure serum VEGF level, heart cells score multiplied by intensity of staining. Po- specimens were fixated using formaldehyde bu- sitive cells were scored as (0) for less than 10% ffer solution of 10% for 24 hours, paraffin em- cells stained, (1) for 10 – 25% stained, (2) for bedded then cut into four to six microns thick. 25 – 50% stained, (3) for 50 – 75% stained, and VEGF ELISA. Blood samples were centrifuged at (4) for more than 75% cells stained. Intensity of 1500 g at room temperature. Collected supernatant the staining was categorized into (1) weak, (2) was assessed using Rat VEGF-A (Vascular Endo- moderate, and (3) strong intensity. thelial Cell Growth Factor A) ELISA Kit (Elab

323 Medicinski Glasnik, Volume 17, Number 2, August 2020

Statistical analysis Both slides of VEGFR-2 and CD31 expression showed clear contrasting results since the Samples fit to normal distribution were assessed first day of administration, presented in brown- using independent samples t-test, to compare se- stained cells. There was a significantly higher rum VEGF level and expressions of VEGFR-2 and immunohistochemistry score of VEGFR-2 in the CD-31 between the treatment and control groups. treatment group compared to the control group If data were not normally distributed, data would on day seven (p=0.010), and fourteen (p=0.011) be compared by using Mann-Whitney U-test. P (Figure 3). value under 0.05 indicated statistical significance.

RESULTS Thirty-two drug-naive rats, weighing in average 200 gr, were divided into two experimental groups. Serum VEGF level, immunohistochemistry scoring of VEGFR-2, and CD31 were compared between the treatment group to the control group on day one, four, seven and fourteen of bay leaf extract administration. Higher serum VEGF levels in the treatment group on day one, seven and fourteen days compared to the control group were found (Table 1) Table 1. Serum vascular endothelial growth factor (VEGF) level between the control and the treatment group Figure 3. Immunohistochemistry scoring of VEGFR-2 on acute Serum VEGF level in the group (mean, SD) coronary syndrome rat model after Syzygium polyanthum ex- Day Mean difference tract administration (*p < 0.05) Control Treatment df p (95% CI) Significantly higher CD31 expressions were 38.98 50.27 -11.29 -3.15 1 0.020 (6.53) (2.93) (-20.05, -2.52) (6) shown in the treatment group compared to the 42.81 53.93 -11.11 -2.30 control group on day four (p=0.002), seven 4 0.061 (8.88) (3.80) (-22.94, 0.71) (6) (p=0.006), and fourteen (p=0.006) days of tre- 49.39 59.91 -10.51 -2.57 7 0.042 (4.86) (6.54) (-20.49, -0.53) (6) atment (Figure 4). 55.34 65.32 -9.97 -4.86 14 0.003 (3.38) (2.32) (-15.15, -4.79) (6) SD, standard deviation; df, Degrees of freedom; Slides were examined for the expression of VEGFR-2 (Figure 1) and CD31 (Figure 2).

Figure 1. Immunohistological slides of VEGFR-2 expression on myocardial of surgical-induced acute coronary syndrome rats. Black arrow indicates expression of VEGFR-2 Figure 4. Immunohistochemistry scoring of CD31 on acute coronary syndrome rat model after Syzygium polyanthum extract administration (*p < 0.05)

DISCUSSION This study found higher serum VEGF levels in groups with one, seven and fourteen days of bay Figure 2. Immunohistological slides of CD31 expression on leaf extract administration. Serum VEGF level myocardial of surgical-induced acute coronary syndrome rats. is elevated in acute phase of ACS and therefore Black arrow indicates expression of CD31

324 Hasan et al. Bay leaf on VEGF and CD31 in ACS

has the potential to be the marker of myocardi- CD31 expression confirmed improved neovascu- al injury (13). Increased VEGF expression in larization within the infarct border zone, thereby myocardium after acute onset correlates to both contributing to protection of nonischemic areas of reperfusion in ischemic area and evolving (28). This study is concordant with a previous progression of the ACS (19). This pattern could study, which found increased CD31 concentration be explained as VEGF expresses differently to in patients with coronary artery stenosis (29). The infarcted myocardium, noninfarcted myocar- study by Serebruany et al. stated that CD31 level dium, and the border zone. As the ischemia in plasma and platelet were observed to increase to process undertakes, VEGF expression would peak of 3 hours after acute onset and followed by a reach its peak differently: 2 hours after onset at significant decrease later at 24 hours (30). Platelet infarcted myocardium, 12 hours after onset at CD31 was known to decline early and rise as the border zone, while at noninfarcted myocardi- second phase of ‘re-expression’ takes place and um it remains unchanged (20). This explained this may explain the significant difference in our the significant difference even in the first day study, as proven in our study by the result of day of administration. seven and fourteen. Co-expression of CD31 re- To date, there is no sufficient research to reach flects that myocardial reperfusion takes place and a consensus of certain VEGF level as a bench- the expression was concentrated from the affected mark (21). A study found high VEGF level and area. However, high expression may also result expression associated to well-developed coro- as the effect of angiogenesis itself, which is not nary collateral arteries in which improved survi- distinguished in this study. Angiogenesis related val in patients with coronary artery disease (22) to increased expression of CD31 has also been as increased VEGF expression has a potential shown in a previous study in rat model, proving role in cardiac repair following the acute onset of that CD31 expression correlates with hypoxia in- myocardial infarction (20). Prolonged expression ducible factor-1 alpha (HIF-1α) that responded to afterwards is resulted after HIF-alpha identifies ischemic tissue (31). hypoxia in border zone, as VEGF expression is Flavonoid compound, especially quercetin, has needed for angiogenesis meant to supply collate- proven to improve prognosis in rat models with ral vessels. If the area of ischemic is large enou- ischemic brain injury. The study was done in Ca- gh, it needs more time to build adequate density ragana sinica (32), which shares the same proof vascular collateral to supply affected area. perty of quercetin with Syzygium polyanthum. This co-expression may be consistently different In case of acute onset of infarction, nitric oxide up to 14 days as shown in this study, supported (NO) level is upregulated in response to ischemia, by previous studies (23,24), noting serum VEGF as well as being a starter for reperfusion. At certa- levels elevated gradually and reaching a peak on in NO concentration, it activates cNOS enzyme to day 14 after onset of acute myocardial infarction. start reperfusion. By modulating HIF-1 alpha in the VEGFR2 expressions showed consistent results pathway, quercetin upregulates VEGF expression with serum VEGF levels of increased expressi- in ischemic condition (33). Overcontentration of on after bay leaf administration, considering the NO responds to high ROS, which contributes to role of VEGF-A/VEGFR2 signal-dependent an- ischemia-reperfusion injury. Flavonoid, given as giogenesis pathway (25). Our results suggest that adjuvant, let reperfusion occur in myocardium but VEGF and its receptor pathways may be respon- avoid the injury, as it directly scavenge excessive sible for outcome range shown in this study. harmful reactive species (34). Quercetin also indu- To further understand ACS progression, it is im- ces upregulation of CD31 (32). This is concordant portant to measure both VEGF functions of as to the result of the study, which shows an increa- angiogenesis regulators and their anti-angiogenic sed CD31 expression after bay leaf extract admi- isoform that inhibits angiogenesis, feasibly thro- nistration. An earlier study published by Flego at ugh its receptors (26). Overexpression of VEGF al. found that an increase in CD31 expression had must be avoided as it may worsen the clinical protective effect against helper T cell dysregulati- manifestation rather than support the improve- on and lymphocyte overreaction observed during ment of ischemic tissues (27). acute onset (25).

325 Medicinski Glasnik, Volume 17, Number 2, August 2020

Effective dosage of bay leaf extract or flavono- In conclusion, this study shows the potential id active compound may alter changes in CD31 effect of bay leaf extract on angiogenesis in ACS expression and serum VEGF level; this is within as adjuvant for the treatment. Bay leaf extract has the limitation of this study. More samples may been shown to support angiogenesis and mainta- present clearer results within each group. Further in endothelial integrity that leads to a better pro- studies are needed to evaluate dose-dependent gnosis for reperfusion on ischemic tissue. and active flavonoid compound from bay leaf extract to seek beneficial effects of bay leaf FUNDING on endothelial integrity and angiogenesis post No specific funding was received for this study myocardial infarction. Expressions of VEGF-related angiogenesis may be done at different place TRANSPARENCY DECLARATION and time, considering that different areas of in- Conflict of interest: None to declare farction express VEGFR-2 and CD31 differently.

REFERENCES

1. World Health Organization. Cardiovascular diseases 11. Olszewska-Pazdrak B, Hein TW, Olszewska P, Car- fact sheet no. 317. https://www.who.int/en/news-ro- ney DH. Chronic hypoxia attenuates VEGF signa- om/fact-sheets/detail/cardiovascular-diseases-(cvds) ling and angiogenic responses by downregulation of (25 March 2020) KDR in human endothelial cells. Am J Physiol Cell 2. Purnamasari D. The emergence of non-communica- Physiol 2009; 296:C1162-70. ble disease in Indonesia. Acta Med Indones 2019; 12. DeLisser HM, Newman PJ, Albelda SM. Platelet en- 50:273-4. dothelial cell adhesion molecule (CD31). Curr Top 3. Ministry of Health Republic of Indonesia. Ri- Microbiol Immunol 1993; 184:37 set kesehatan dasar 2018 (Basic health research 13. Clement M, Fornasa G, Guedj K, Mkaddem SB, 2018) [in Indonesian] http://kesmas.kemkes.go.id/ Gaston AT, Khallou-Laschet J, Morvan M, Nicoletti assets/upload/dir_519d41d8cd98f00/files/Hasil-ri- A, Caligiuri G. CD31 is a key coinhibitory receptor skesdas-2018_1274.pdf (25 March 2020) in the development of immunogenic dendritic cells. 4. Stefanadis C, Antoniou CK, Tsiachris D, Pietri P. Proc Natl Acad Sci USA 2014; 111:E1101-10. Coronary atherosclerotic vulnerable plaque: current 14. Xia L, Zhang B, Sun Y, Chen B, Yu Z. Syk/PE- perspectives. Am Heart J 2017; 6:e005543. CAM-1 signaling pathway mediates low shear stre- 5. Heo KS, Fujiwara K, Abe JI. Shear stress and athe- ss-induced inflammatory in atherosclerosis. http:// rosclerosis. Mol Cells 2014; 37:434-40. dx.doi.org/10.2139/ssrn.3369761 (28 March 2020) 6. Ford TJ, Berry C, De Bruyne B, Yong AS, Barlis 15. Hartanti L, Yonas SM, Mustamu JJ, Wijaya S, Se- P, Fearon WF, Ng MK. Physiological predictors of tiawan HK, Soegianto L. Influence of extraction acute coronary syndromes: emerging insights from methods of bay leaves (Syzygium polyanthum) on the plaque to the vulnerable patient. JACC Cardio- antioxidant and HMG-CoA reductase inhibitory ac- vasc Interv 2017; 10:2539-47. tivity. Heliyon 2019; 5:e01485. 7. Siasos G, Sara JD, Zaromytidou M, Park KH, 16. Monsalve B, Concha-Meyer A, Palomo I, Fuentes Coskun AU, Lerman LO, Oikonomou E, Maynard E. Mechanisms of endothelial protection by natural CC, Fotiadis D, Stefanou K, Papafaklis M. Local bioactive compounds from fruit and vegetables. An low shear stress and endothelial dysfunction in pati- Acad Bras Cienc 2017; 89:615-33. ents with nonobstructive coronary atherosclerosis. J 17. Khurana S, Venkataraman K, Hollingsworth A, Am Coll Cardiol 2018; 71:2092-102. Piche M, Tai TC. Polyphenols: benefits to the car- 8. Chatzizisis YS, Coskun AU, Jonas M, Edelman ER, diovascular system in health and in aging. Nutrients Feldman CL, Stone PH. Role of endothelial shear 2013; 5:3779-827. stress in the natural history of coronary atheroscle- 18. Bassino E, Antoniotti S, Gasparri F, Munaron L. rosis and vascular remodeling: molecular, cellular, Effects of flavonoid derivatives on human microvas- and vascular behavior. J Am Coll Cardiol 2007; cular endothelial cells. Nat Prod Res 2016; 30:2831- 49:2379-93. 4. 9. Chistiakov DA, Orekhov AN, Bobryshev YV. 19. Wu Y, Yin X, Wijaya C, Huang MH, McConnell BK. Effects of shear stress on endothelial cells: go with Acute myocardial infarction in rats. J Vis Exp 2011; the flow. Acta Physiol 2017; 219:382-408. 48:2464. 10. Tzima E, Irani-Tehrani M, Kiosses WB, Dejana 20. Lee SH, Wolf PL, Escudero R, Deutsch R, Jamieson E, Schultz DA, Engelhardt B, Cao G, DeLisser H, SW, Thistlewaite PA. Early expression of angioge- Schwartz MA. A mechanosensory complex that nesis factors in acute myocardial ischemia and in- mediates the endothelial cell response to fluid shear farction. N Engl J Med 2000; 342:626-33. stress. Nature 2005; 437:426-31.

326 Hasan et al. Bay leaf on VEGF and CD31 in ACS

21. Briasoulis A, Androulakis E, Christophides T, Tou- 28. Privratsky JR, Paddock CM, Florey O, Newman soulis D. The role of inflammation and cell death in DK, Muller WA, Newman PJ. Relative contribu- the pathogenesis, progression and treatment of heart tion of PECAM-1 adhesion and signaling to the failure. Heart Fail Revs 2016; 21:169-76. maintenance of vascular integrity. J Cell Sci 2011; 22. Ogawa H, Suefuji H, Soejima H, Nishiyama K, Mi- 124:1477–85. sumi K, Takazoe K, Miyamoto S, Kajiwara I, Sumi- 29. Deutsch MA, Brunner S, Grabmaier U, David R, Ott da H, Sakamoto T, Yoshimura M. Increased blood I, Huber BC. Cardioprotective potential of human vascular endothelial growth factor levels in patients endothelial-colony forming cells from diabetic and with acute myocardial infarction. Cardiology 2000; nondiabetic donors. Cells 2020; 9:588. 93:93-9. 30. Fang L, Wei H, Chowdhury SH, Gong N, Song J, 23. Ramos C, Napoleão P, Selas M, Freixo C, Viegas Heng CK, Sethi S, Koh TH, Chatterjee S. Associa- Crespo AM, Mota Carmo M, Cruz Ferreira R, Pinhe- tion of Leu125Val polymorphism of platelet endo- iro. Prognostic value of VEGF in patients submitted thelial cell adhesion molecule-1 (PECAM-1) gene to percutaneous coronary intervention. Dis Markers & soluble level of PECAM-1 with coronary artery 2014; 2014. disease in Asian Indians. Indian J Med Res 2005; 24. Meier P, Gloekler S, Zbinden R, Beckh S, de Marchi 121:92-9. SF, Zbinden S. Beneficial effect of recruitable co- 31. Flego D, Severino A, Trotta F, Copponi G, Manchi llaterals: a 10-year follow-up study in patients with M, Pedicino D, Giglio AF, Crea F, Liuzzo G. Redu- stable coronary artery disease undergoing quanti- ced CD31 expression on CD14+ CD16+ monocyte tative collateral measurements. Circulation 2007; subset in acute coronary syndromes. Int J Cardiol 116:975–83. 2015; 197:101-4. 25. Flego D, Severino A, Trotta F, Previtero M, Ucci S, 32. Serebruany VL, Gurbel PA. Effect of thrombolytic Zara C, Pedicino D, Massaro G, Biasucci LM, Liuz- therapy on platelet expression and plasma concen- zo G, Crea F. Altered CD31 expression and activity tration of PECAM-1 (CD31) in patients with acute in helper T cells of acute coronary syndrome pati- myocardial infarction. Arterioscler Thromb Vasc ents. Basic Res Cardiol 2014; 109:448. Biol 1999; 19:153. 26. Zhao T, Zhao W, Chen Y, Ahokas RA, Sun Y. Vascu- 33. Wilson WJ, Poellinger L. The dietary flavonoid quer- lar endothelial growth factor (VEGF)-A: role on car- cetin modulates HIF-1n activity in endothelial cells. diac angiogenesis following myocardial infarction. Biochem Biophys Res Commun 2002; 293:446-50. Microvasc Res 2010; 80:188-94. 34. Wang CZ, Mehendale SR, Calway T, Yuan CS. Bo- 27. Yang Z, Wan J, Pan W, Zou J. Expression of vascu- tanical flavonoids on coronary heart disease. Am J lar endothelial growth factor in cardiac repair: Si- Chin Med 2011; 39:661-71. gnaling mechanisms mediating vascular protective effects. Int J Biol Macromol 2018; 113:179-85.

327 ORIGINAL ARTICLE

Neutrophil gelatinase-associated lipocalin is a predictor of complications in the early phase of ST-elevation myocardial infarction

Nirvana Šabanović-Bajramović1, Enisa Hodžić1, Amer Iglica1, Edin Begić2,3, Nerma Resić1, Kenana Aganović1, Mirela Halilčević1, Senad Bajramović4

1Intensive Care Unit, Clinic for Heart, Blood Vessel and Rheumatic Diseases, Clinical Center University of Sarajevo, 2Department of Cardiology, General Hospital ''Prim.dr. Abdulah Nakaš'', 3Department of Pharmacology, School of Medicine, Sarajevo School of Science and Technology, 4Urology Clinic, Clinical Centre of the University of Sarajevo; Sarajevo, Bosnia and Herzegovina

ABSTRACT

Aim To evaluate a correlation of serum level of neutrophil gelatinase-associated lipocalin (NGAL) to the risk of the occurrence of complications in patients with the early phase of ST-segment elevation myocardical infarction (STEMI) treated with fibrinolytic therapy prior to percutaneous coronary intervention (PCI).

Methods A total of 54 patients with the diagnosis of STEMI treated with fibrinolytic therapy (alteplase) prior to PCI were included. Patients were admitted to the Intensive Care Unit (ICU) of Clinic for Heart, Blood Vessel and Rheumatic Diseases in the period January to March 2018. All patients underwent coronary angiography and PCI within the maximum of 48 hours delay after Corresponding author: fibrinolysis, according to the hemodynamic and electrical stability and PCI availability. Blood samples were taken immediately Nirvana Šabanović-Bajramović after admission prior to fibrinolytic administration. Patients were Intensive Care Unit, Clinic for Heart, divided into two groups according to NGAL values (less or more Blood Vessel and Rheumatic Diseases, than 134.05 ng/mL). Clinical Centre University of Sarajevo Results Higher values of NGAL have effect on a higher mean Bolnička 25, 71000 Sarajevo, systolic and diastolic pressure (p=0.001 and p=0.003, respecti- Bosnia and Herzegovina vely). Patients with higher NGAL values also have higher values Phone: +387 33 297 521; of brain natriuretic peptide (p=0.0001) and highly sensitive tropo- Fax: +387 33 297 805; nin I (p=0.002). In that group relative risk (RR) for lethal outcome E-mail: [email protected] was 6.4 times significantly higher (p=0.002), for the development ORCID ID: https://orcid.org/0000-0003- of heart failure 2.88 times (p=0.0002), for post-myocardial infarction angina pectoris 2.24 times (p=0.0158), and for ventricular 3749-6073 rhythm disturbances (ventricular tachycardia, ventricular fibrillation) 1.96 times higher (p=0.0108).

Original submission: Conclusion Increased NGAL value is related to an unfavourable 07 February 2020; outcome of patients in the early phase of STEMI treated with fibri- Revised submission: nolytic therapy prior to PCI. 08 May 2020; Key words: acute coronary syndrome, lipocalin-2, prognosis Accepted: 15 June 2020 doi: 10.17392/1139-20

Med Glas (Zenica) 2020; 17(2): 328-334

328 Šabanović-Bajramović et al. NGAL as a predictor of STEMI complications

INTRODUCTION PATIENTS AND METHODS Neutrophil Gelatinase-Associated Lipocalin Patients and study design (NGAL) is a 25 kDa siderophore 178 amino acids binding protein, composed of 8 beta sheets that This prospective study included 54 patients with form a shaped structure, and represents an acute the diagnosis of STEMI treated with fibrinolytic renal injury marker (AKI) (1-7). Its advantage is therapy (alteplase) prior to PCI. The patients that it can detect AKI after two hours of injury, were admitted to the Intensive Care Unit (ICU) much earlier than a rise in creatinine concentration of Clinic for Heart, Blood Vessel and Rheumatic (3,4). It is synthetized in renal tubular, intestinal, Diseases, Clinical Centre University of Sarajevo hepatic, and pulmonary tissue (4). Circulating in the period January to March 2018. The pati- NGAL is filtered by the glomerulus and reabsor- ents were divided into two groups according to bed in the proximal tubule (NGAL is stored in NGAL value (less or more than 134.05 ng/mL). neutrophils within specific granules) so it can be Criteria for inclusion were diagnosis of STEMI, monitored in both urine and plasma, but also in treatment with fibrinolytic therapy prior to PCI whole blood and serum (5,7). It is associated with and accepted participation in the research. Exclu- many processes, such as inflammation, infection, sion criteria were diagnosis of sub-acute myocar- intoxication, ischemia, neoplastic transformation, dial infarction, diagnosis of non-ST segment ele- transport of pheromones and with the synthesis vation myocardial infarction (NSTEMI), those of prostaglandin (7,8). It is significant for the pre- who were not treated by fibrinolytic therapy prior diction of renal dysfunction in patients with chro- to PCI and patients over 85 years of age. nic heart failure (HF) (9). In acute HF, it can be a The research was conducted in accordance with predictor of worsening of the renal function and basic principles of the Declaration of Helsinki strongly predicts adverse clinical outcomes (10). (last revision in 2008) on the rights of patients in- The prognostic mortality factor is in patients with volved in biomedical research. An informed con- HF, both with and without chronic kidney disease, sent was obtained from all patients included in and it is superior to estimated glomerular filtration the study. An ethical approval was obtained from rate (eGFR) and cystatin C values (11). The value the Ethical Committee of the Clinical Centre of of NGAL urine in the early phase of acute myocar- the University of Sarajevo. dial infarction (MI) is associated with NT-pro BNP values, and it can be related with localization Methods of MI (12). It represents an important part of mi- The dose of alteplase was 0.9 mg/kg infused over neralocorticoid-stimulated vascular fibrosis (13). 60 minutes (not exceeding 90 mg). Patients were It is found in endothelial cells, smooth muscle followed up for 5 to 7 days (median follow-up time cells, and macrophages in atherosclerotic plaques, of patients was 6 days). All patients underwent cor- and could therefore be associated with the deve- onary angiography and PCI within maximum of 48 lopment, or progression of, atherosclerotic stable hours delay after fibrinolysis, according to the he- or unstable plaque (via endothelial dysfunction, modynamic and electrical stability and PCI avail- inflammation and matrix degradation and plaque ability. Blood samples were taken immediately instability), and it is in relation with anamnestic after the admission prior to fibrinolytic administra- data, risk factors and medication intake (14-17). tion. The NGAL value in plasma was measured at Serum NGAL level is strongly predictive for sur- the Institute for Clinical Biochemistry and Immu- vival to hospital discharge after cardiac arrest (16). nology, Clinical Centre of the University of Sara- Renal function is one of the most important para- jevo using Human NGAL Rapid ELISA Kit (Bio- meters that contribute to the outcome of a patient Porto Diagnostics, Hellerup, Denmark) test. NGAL with STEMI, so early diagnosis of renal dysfuncti- values were presented as third quartile of labora- on may be helpful in the therapeutic modality. The tory values (it was consider as pathological) (18). aim of research was to evaluate a correlation of se- Values of low density lipoproteins (LDL), high rum level of NGAL with the risk of complications density lipoproteins (HDL), glycolized haemoglo- occurrence in patients with the early phase of ST- bin (HbA1C), brain natriuretic peptide (BNP), high segment elevation myocardial infarction (STEMI) sensitive Troponin I (hsTnI), urea and creatinine treated with fibrinolytic therapy prior to PCI. were also taken at admission. Anamnestic data of

329 Medicinski Glasnik, Volume 17, Number 2, August 2020

comorbidities and risk factors (age, gender, diagno- Patients with high NGAL had significantly sis of hypertension, diabetes mellitus, dyslipidae- higher brain natriuretic peptide (BNP) le- mia, previous myocardial infarction, previous coro- vel (647.06±1001.37 vs. 2249.89±1336.43; nary artery bypass surgery) and habits (smoking) p=0.0001), high sensitive troponin I level were collected. Echocardiography was performed (14840.85±17965.39 vs. 37313.5±22699.86; on the fifth day of hospitalization, and ejection frac- p= 0.002), larger left atrial diameter (3.63±0.3 tion of left ventricle (EFLV) was assessed by the vs. 4.23±0.28; p=0.0001), significantly larger Simpson method. Complications of STEMI (acute left ventricular internal dimension at end-systo- heart failure during hospitalization, post-infarction le (LVIDs) (4.18±0.34 vs. 3.68±0.47; p= 0.006), angina pectoris, ventricular heart rhythm disorders, larger mitral peak velocity of early filling (E) to lethal outcome) were also monitored. mitral peak velocity of late filling (A) (E/A) ratio (0.81±0.25 vs. 1.48±0.89, p=0.001). Patients with Statistical analysis high NGAL had significantly reduced ejection The study used c2 test, Student's t test and Relati- fraction (45.16±5.22 vs. 37.3±5.06, p=0.00001). ve Risk. Receiver operating curve (ROC) analysis The significant difference was not proven in other was used to determine sensitivity and specificity. investigated factors (cholesterol, triglycerides, The values of monitored laboratory parameters, high density lipoproteins (HDL), low density lipo- systolic and diastolic function, as well as the com- proteins (LDL), urea, creatinine, glycolized hae- plications of STEMI were all analysed comparing moglobin and left ventricular internal dimension to plasma NGAL values. All analytical results at end-diastole (LVIDd)) (Table 2). with p<0.05 were considered statistically signifi- Table 2. Laboratory and echocardiography findings of patients cant. Identity and all personal data of patients are with ST-segment elevation myocardial infarction (STEMI) accord- permanently protected in accordance to regulati- ing to neutrophil gelatinase-associated lipocalin (NGAL) value Reference NGAL ons of protection of identification data. Identifica- Parameter p value <134.05 ng/mL >134.05 ng/mL tion number was assigned to every patient in order Cholesterol 3.1-5.2 5.84±0.83 6.16±0.34 0.244 to protect personal information and that number (mmol/L) Triglycerides was used in statistical analysis. 0.11-1.7 2.49±3.13 3.13±1.04 0.180 (mmol/L) RESULTS HDL (mmol/L) 1.06-1.94 1.6±0.69 2.14±0.82 0.049 LDL (mmol/L) 1.4-3.4 3.16±0.69 3.5±0.36 0.145 Patients with high NGAL (above 134.05 ng/mL) Urea (mmol/L) 2.0-7.8 7.46±1.69 9.45±2.29 0.005 Creatinine 63-109 86.88±22.37 102.6±23.16 0.060 had significantly higher mean systolic and diasto- (μmol/L) lic blood pressure compared to patients with low HbA1c (%) 4-6.5 6.65±1.55 7.82±2.34 0.075 14840.85 37313.5 NGAL (p=0.001 and p=0.003, respectively). The hsTnI (pg/mL) 5-30 0.002 ±17965.39 ±22699.86 significant difference was not proven in other in- BNP (pg/mL) <100 647.06±1001.37 2249.89±1336.430.0001 vestigated factors (Table 1). LVIDs (cm) 3.5-5.6 3.68±0.47 4.18±0.34 0.006 LVIDd (cm) 2.0-4.0 5.5±0.46 5.36±2.02 0.694 LAD (cm) 2.0-4.0 3.63±0.3 4.23±0.28 0.0001 Table 1. Characteristics of patients with ST-segment eleva- EF (%) >50 45.16±5.22 37.3±5.06 0.0001 tion myocardial infarction (STEMI) according to neutrophil E/A 1-2.2 0.81±0.25 1.48±0.89 0.001 gelatinase-associated lipocalin (NGAL) value LDL, low density lipoproteins; HbA1C, glycolized haemoglobin; NGAL BNP, brain natriuretic peptide; hsTnI, high sensitive Troponin I; Parameter p LVIDd, left ventricular internal dimension at end-diastole; LVIDs, <134.05 ng/mL >134.05 ng/mL left ventricular internal dimension at end-systole; LAD, left atrial Age (mean) (±SD) (years) 60.6±10.6 66.2±11.3 0.155 diameter; EF, ejection fraction; E/A, mitral peak velocity of early Gender (No, %) 18 (56.3) 8 (80.0) 0.165 filling (E) to mitral peak velocity of late filling (A) ratio Diabetes mellitus (No, %) 8 (25.0) 5 (50.0) 0.136 Hypertension (No, %) 24 (75.0) 7 (70.0) 0.524 Dyslipidaemia (No, %) 19 (59.4) 8 (80.0) 0.212 Patients with higher NGAL had a relative risk Smoking (No, %) 18 (56.3) 8 (80.0) 0.320 (RR) of lethal outcome 6.4 times significantly Previous myocardial 3 (9.4) 5 (50.0) 0.012 higher (CI 1.9466-21.0147; p=0.0022), while infarction (No, %) Previous CABG (No, %) 1 (3.1) 3 (30.0) 0.036 for development of HF 2.88 times significantly SBP (mean) (±SD) 133.5±12.9 152.1±6.7 0.0001 higher (CI = 1.6552-5.0112; p=0.0002). For post- DBP (mean) (±SD) 88.5±7.8 97.6±5.6 0.001 myocardial infarction angina pectoris it was 2.24 CABG, coronary artery bypass surgery; SBP, systolic blood pressure; times significantly higher (CI 1.1638-4.3114; DBP, diastolic blood pressure

330 Šabanović-Bajramović et al. NGAL as a predictor of STEMI complications

p=0.0158), while for ventricular rhythm distur- found elevated NGAL values in patients with re- bances (ventricular tachycardia, ventricular fi- novascular hypertension (19). In our study patients brillation) 1.96 times significantly higher (CI = with elevated NGAL value also had the elevated 1.1695-3.3158; p=0.0108) (Table 3). brain natriuretic peptide (BNP) value, which is followed by echocardiographic changes in the form Table 3. Occurrence of complications of the patients in the of a systolic and diastolic dysfunction. Kim et al. early phase of ST-segment elevation myocardial infarction (STEMI) established NGAL as independent predictor of left No (%) of patients with ventricular hypertrophy (LVH) and left ventricular NGAL diastolic dysfunction (LVDD) on 332 pre-dialysis Complication p <134.05 ng/ >134.05 chronic kidney disease patients (20). Increased mL ng/mL values of BNP are a sign of heart left ventricular Acute HF during hospitalization 10 (31.3) 9 (90.0) 0.002 dysfunction (20). Elevated NGAL serum levels on Post-infarction angina pectoris 10 (31.3) 7 (70.0) 0.036 Ventricular heart rhythm disorders 13 (40.6) 8 (80.0) 0.033 admission are associated with a worsening of re- (VT, VF) Lethal outcome 9 (9.4) 6 (60.0) 0.0001 nal function in patients with acute HF (10), which NGAL, neutrophil gelatinase-associated lipocalin; HF, heart failure; directly causes the development of complications. VT, ventricular tachycardia; VF, ventricular fibrillation Nakada et al. the verified elevated NGAL values in urine as the marker of the prediction of the The ROC curve estimates the best possible sen- occurrence of acute HF (21). Although the NGAL sitivity (90%) and specificity (71.9%) at a certain increase occurs before the rise of creatinine, the cut off value (Figure 1). For lethal outcome, op- Acute Kidney Injury Neutrophil Gelatinase-Asso- timal sensitivity and specificity for NGAL values ciated Lipocalin Evaluation of Symptomatic Heart occur at cut off values> 362.9 ng/mL, with si- Failure Study (AKINESIS) trial did not confirm gnificant (p=0.0022) area under curve (AUC) of the superiority of NGAL compared to creatinine in failure or systolic 0.85. 930 patients for predicting worsening renal functi-

1.0 on or in-hospital outcomes (22). Al-Afify in 52 patients with STEMI treated with fibrinolytic therapy (reteplase) concluded that death and in-hospital 0.8 complications were significantly higher in patients with high NGAL than those with low NGAL and that plasma NGAL was a significant predictor of 0.6 mortality and in hospital complications (23). Se- rum levels of NGAL are higher in patients with

Sensitivity 0.4 acute coronary syndrome than those with stable angina (24). Akcay et al. in a study of 100 consecutive patients with STEMI found that in-hospital 0.2 and 1-year major adverse cardiovascular event rates were significantly higher in the high-NGAL

0.0 group compared to the low NGAL group (25). Ne- 0.0 0.2 0.4 0.6 0.8 1.0 utrophil activation has been reported in unstable 1 - Speciﬁcity angina and acute myocardial infarction but not in Figure 1. Sensitivity and specificity of the neutrophil gelati- patients with stable angina (26). Plasma NGAL is nase-associated lipocalin (NGAL) value for the prediction of a significant predictor of mortality but is weakly lethal outcome associated with several traditional cardiovascular DISCUSSION risk factors including age, systolic blood pressure, hypertension and diabetes (27,28). That is in accor- The research showed higher levels of NGAL in dance to results obtained in our research. At cell patients with higher systolic and diastolic pressu- model of myocardial infarction after percutaneous re. Gharishvandi et al. have proven that NGAL coronary intervention, NGAL is overexpressed by concentrations are higher in hypertension, and the heart after MI and blocks cardiac dysfunction that they are elevated in hypertensive patients with and fibrosis in experimental MI, and may represent early stages of renal failure (17). Elrin et al. have therapeutic modality (29). Direct effects of NGAL

331 Medicinski Glasnik, Volume 17, Number 2, August 2020

expression on cardiomyocyte size and number 9 enzymatic activity (34). Our research showed could be related to cardiac hypertrophy and heart that increased NGAL concentration, increased free failure (30). Plasma level of lipocalin 2 gene is ele- NGAL, lead to MMP degradation, resulting with vated at 1-3 days in patients with ischemic stroke plaque rupture. As this plaque rupture is greater, (31) (ischemic stroke is related to the progression the values of NGAL will be higher, and MI will of atherosclerotic process). be more severe with the occurrence of a greater Although a precise mechanism of plaque rupture is number of complications. Hedmahl et al. proved poorly understood, it is generally accepted that the the same thing in mouse models (36). Helanova disorder occurs at the site of a fibrous capsule that et al. proved that increased levels of NGAL were is highly infiltrated by macrophages and T lymp- detected in patients with acute myocardial infarcti- hocytes where the underlying necrotic nucleus is on, heart failure or stroke, and were demonstrated usually large (30,31). Degradation of atheroscle- to be strong predictors of adverse prognosis (37). rotic plaque and its fibre cap is considered to occur The research presents an importance of NGAL through the extracellular matrix protein degradati- value in the early stage of STEMI, and shows that on of extracted matrix metalloproteinase (MMP), NGAL can be one additional marker that can be which is directly involved in the plaque rupture helpful in the prediction of occurrence of com- process (30,31). Plasma MMP-9 antigen was plications, and according to that helpful in mo- elevated in unstable plaques (30). An important dification of the therapeutic approach. A larger function of NGAL is the formation of a complex number of patients, primary PCI and analysis of with MMP-9, which slows down the inactivation patients in relation to the characteristics of pri- of MMP-9 by tissue inhibitors of matrix metallo- mary PCI could further establish NGAL as a use- proteinases (TIMP-1) and results in a longer effect ful maker in STEMI. of proteolytic activity of MMP-9 (30). NGAL may In conclusion, higher concentrations of NGAL exist as monomer, dimer and/ or NGAL/ MMP-9 correlate with the occurrence of complications in complex forms in humans (32). Furthermore, the the early phase of STEMI treated with fibrinolytic NGAL/ MMP-9 complex has also been detected therapy prior to PCI, and increase mortality risk, in a variety of tumour tissues and in acute cystitis correlate with BNP values and predict the origin and plays a role in many inflammatory processes, of acute HF. Also, NGAL is a predictor of post- including chronic inflammatory processes and ne- myocardial infarction angina pectoris as well as oplastic changes (33). The formation of a complex ventricular rhythm disturbances. Further research of NGAL and MMP-9 prevents the degradation should indicate the importance of NGAL values of MMP-9 and reinforces its proteolytic activity, in the prediction of major adverse cardiovascular which can take part in the formation of unstable events after STEMI. atherosclerotic plaque (31). Excessive MMP activity can weaken vessels and, more importantly, FUNDING destabilize plaques leading to increased risk of No specific funding was received for this study. rupture (34-36). Degradation of MMP-9 has been shown to be significantly inhibited in the presence TRANSPARENCY DECLARATION of NGAL and results in the preservation of MMP- Competing interests: None to declare. REFERENCES

1. Pedersen KR, Ravn HB, Hjortdal VE, Nørregaard 4. Bachorzewska-Gajewska H, Malyszko J, Sit- R, Povlsen JV. Neutrophil gelatinase-associated lipo- niewska E, Malyszko JS, Poniatowski B, Pawlak calin (NGAL): validation of commercially available K, Dobrzycki S. NGAL (neutrophil gelatinase-asso- ELISA. Scand J Clin Lab Invest 2010; 70:374-82. ciated lipocalin) and cystatin C: are they good pre- 2. Haase-Fielitz A, Haase M, Devarajan P. Neutrophil dictors of contrast nephropathy after percutaneous gelatinase-associated lipocalin as a biomarker of coronary interventions in patients with stable angina acute kidney injury: a critical evaluation of current and normal serum creatinine? Int J Cardiol 2008; status. Ann Clin Biochem 2014; 51:335–51. 127:290–1. 3. McIlroy DR, Wagener G, Lee HT. Neutrophil gela- 5. Barreto R, Elia C, Solà E, Moreira R, Ariza X, Ro- tinase-associated lipocalin and acute kidney injury dríguez E, Graupera I, Alfaro I, Morales-Ruiz M, after cardiac surgery: the effect of baseline renal Poch E, Guevara M1 Fernández J, Jiménez W, function on diagnostic performance. Clin J Am Soc Arroyo V, Ginès P. Urinary neutrophil gelatinase- Nephrol 2010; 5:211–9.

332 Šabanović-Bajramović et al. NGAL as a predictor of STEMI complications

associated lipocalin predicts kidney outcome and de- 19. Eirin A, Gloviczki ML, Tang H, Rule AD, Woollard ath in patients with cirrhosis and bacterial infections. JR, Lerman A, Textor SC, Lerman LO. Chronic re- J Hepatol 2014; 61:35–42. novascular hypertension is associated with elevated 6. Verna EC, Brown RS, Farrand E, Pichardo EM, For- levels of neutrophil gelatinase-associated lipoca- ster CS, Sola-Del Valle DA, Adkins SH, Sise ME, lin. Nephrol Dial Transplant 2012; 27:4153-61. Oliver JA, Radhakrishnan J, Barasch JM, Nickolas 20. Kim IY, Kim JH, Kim MJ, Lee DW, Hwang CG, Han TL.Urinary neutrophil gelatinase-associated lipo- M, Rhee H, Song SH, Seong EY, Lee SB. Plasma ne- calin predicts mortality and identifies acute kidney utrophil gelatinase-associated lipocalin is indepen- injury in cirrhosis. Dig Dis Sci 2012; 57:2362-70. dently associated with left ventricular hypertrophy 7. Noto A, Cibecchini F, Fanos V, Mussap M. NGAL and diastolic dysfunction in patients with chronic and metabolomics: the single biomarker to reveal kidney disease. PLoS One 2018; 13:e0205848. the metabolome alterations in kidney injury. Biomed 21. Nakada Y, Kawakami R, Matsui M, Ueda T, Nakano Res Int 2013; 2013:612032. T, Takitsume A, Nakagawa H, Nishida T, Onoue K, 8. Mishra J, Qing MA, Prada A, Mitsnefes M, Zahedi Soeda T, Okayama S, Watanabe M, Kawata H, Oku- K, Yang J, Barasch J, Devarajan P.. Identification of ra H, Saito Y. Prognostic value of urinary neutrophil neutrophil gelatinase-associated lipocalin as a novel gelatinase-associated lipocalin on the first day of early urinary biomarker for ischemic renal injury. J admission for adverse events in patients with acu- Am Soc Nephrol 2003; 14:2534–43. te decompensated heart failure. J Am Heart Assoc 9. Palazzuoli A, Beltrami M, Pellegrini M, Nuti R. Na- 2017; 6:e004582 triuretic peptides and NGAL in heart failure: does a 22. Maisel AS, Wettersten N, van Veldhuisen DJ, Mu- link exist? Clin Chim Acta 2012; 413:1832-8. eller C, Filippatos G, Nowak R, Hogan C, Kontos 10. Aghel A, Shrestha K, Mullens W, Borowski A, Tang MC, Cannon CM, Müller GA, Birkhahn R, Clopton WH. Serum neutrophil gelatinase-associated lipoca- P, Taub P, Vilke GM, McDonald K, Mahon N, Nuñez lin (NGAL) in predicting worsening renal function J, Briguori C, Passino C, Murray PT. Neutrophil ge- in acute decompensated heart failure. J Card Fail latinase-associated lipocalin for acute kidney injury 2010; 16:49-54. during acute heart failure hospitalizations: the AKI- 11. van Deursen VM, Damman K, Voors AA, van der NESIS Study. J Am Coll Cardiol 2016; 68:1420-31. Wal MH, Jaarsma T, van Veldhuisen DJ, Hillege 23. Al-Afify AA. Prognostic value of neutrophil - gela HL. Prognostic value of plasma neutrophil gelatina- tinase-associated lipocalin in predicting in-hospital se-associated lipocalin for mortality in patients with complications in patients with st-segment elevation heart failure. Circ Heart Fail 2014; 7:35-42. myocardial infarction. Research Journal of Cardio- 12. Kirbis S, Gorenjak M, Sinkovic A. The role of urine logy 2013; 6:10-8. neutrophil gelatinase--associated lipocalin (NGAL) 24. Kafkas N, Demponeras C, Zoubouloglou F, Spanou in acute heart failure in patients with ST--elevati- D, Babalis D, Makris K. Serum levels of gelatinase on myocardial infarction. BMC Cardiovasc Disord associated lipocalin as indicator of the inflammatory 2015; 15:49. status in coronary artery disease. Int J Inflammation 13. Leopold JA. The Central Role of Neutrophil Gelati- 2012; 189797 nase-Associated Lipocalin in Cardiovascular Fibro- 25. Akcay AB, Ozlu MF, Sen N, Cay S, Ozturk OH. Pro- sis. Hypertension 2015; 66:20–2. gnostic significance of neutrophil gelatinase-asso- 14. Lindberg S, Pedersen SH, Mogelvang R, Jensen ciated lipocalin in ST-segment elevation myocardial JS, Flyvbjerg A, Galatius S, Magnusson NE. Pro- infarction. J Invset Med 2012; 60:508-13. gnostic utility of neutrophil gelatinase-associated 26. Sarndahl E, Bergstrom I, Brodin VP, Nijm J, Sette- lipocalin in predicting mortality and cardiovascu- rud HL, Jonasson L. Neutrophil activation status lar events in patients with ST-segment elevation in stable coronary artery disease. PLoS One 2007; myocardial infarction treated with primary percuta- 2:e1056. neous coronary intervention. J Am Coll Cardiol 27. Yndestad A, Landrø L, Ueland T, Dahl CP, Flo TH, 2012; 60:339-45. Vinge LE, Espevik T, Frøland SS, Husberg C, Chri- 15. Mozos I. Laboratory markers of ventricular stensen G, Dickstein K, Kjekshus J, Øie E, Gullestad arrhythmia risk in renal failure. Biomed Res Int L, Aukrust P. Increased systemic and myocardial 2014; 2014:509204. expression of neutrophil gelatinase-associated lipo- 16. Elmer J, Jeong K, Abebe KZ, Guyette FX, Murugan calin in clinical and experimental heart failure. Eur R, Callaway CW, Rittenberger JC; Pittsburgh Post- Heart J 2009; 30:1229–36. Cardiac Arrest Service. Serum Neutrophil Gelati- 28. Daniels LB, Barrett-Connor E, Clopton P, Laughlin nase-Associated Lipocalin Predicts Survival After GA, Ix JH, Maisel AS. Plasma neutrophil gelatinase- Resuscitation From Cardiac Arrest. Crit Care Med associated lipocalin is independently associated with 2016; 44:111-9. cardiovascular disease and mortality in community- 17. Gharishvandi F, Kazerouni F, Ghanei E, Rahimipour dwelling older adults: The Rancho Bernardo Stu- A, Nasiri M. Comparative assessment of neutrophil dy. J Am Coll Cardiol 2012; 59:1101–9. gelatinase-associated lipocalin (NGAL) and cystatin 29. Martínez-Martínez E, Buonafine M, Boukhalfa I, C as early biomarkers for early detection of renal Ibarrola J, Fernández-Celis A, Kolkhof P, Rossignol failure in patients with hypertension. Iran Biomed J P, Girerd N, Mulder P, López-Andrés N, Ouvrard- 2015; 19:76–81. Pascaud A, Jaisser F. Aldosterone target NGAL 18. Singer E, Schrezenmeier EV, Elger A, Seelow ER, (neutrophil gelatinase-associated lipocalin) is invol- Krannich A, Luft FC, Schmidt-Ott KM. Urinary ved in cardiac remodeling after myocardial infarc- NGAL-Positive Acute Kidney Injury and Poor tion through NFκB pathway. Hypertension 2017; Long-term Outcomes in Hospitalized Patients. Kid- 70:1148-56. ney Int Rep 2016; 1:114-24.

333 Medicinski Glasnik, Volume 17, Number 2, August 2020

30. Marques FZ, Prestes PR, Byars SG, Ritchie SC, 34. Bouchet S, Bauvois B. Neutrophil gelatinase -asso- Würtz P, Patel SK, Booth SA, Rana I, Minoda Y, ciated lipocalin (NGAL), pro-matrix metalloprotei- Berzins SP, Curl CL, Bell JR, Wai B, Srivastava PM, nase-9 (pro-MMP-9) and their complex pro-MMP-9/ Kangas AJ, Soininen P, Ruohonen S, Kähönen M, NGAL in leukaemias. Cancers 2014; 6:796–812. Lehtimäki T, Raitoharju E, Havulinna A, Perola M, 35. Lippi G, Cervellin G. Neutrophil gelatinase-asso- Raitakari O, Salomaa V, Ala-Korpela M, Kettunen ciated lipocalin: a more specific assay is needed J, McGlynn M, Kelly J, Wlodek ME, Lewandowski for diagnosing renal injury. Clin Chim Acta 2012; PA, Delbridge LM, Burrell LM, Inouye M, Harrap 413:1160–1. SB, Charchar FJ. Experimental and human eviden- 36. Leclercq A, Houard X, Philippe M, Ollivier V, ce for lipocalin-2 (neutrophil gelatinase-associated Sebbag U, Meilhac O, Michel J-B. Involvement of lipocalin [NGAL]) in the development of cardiac intraplaque hemorrhage in atherothrombosis evolu- hypertrophy and heart failure. J Am Heart Assoc tion via neutrophil protease enrichment. J Leukoc 2017; 6:e005971. Biol 2007; 82:1420-9. 31. Chou WH, Wang G, Kumar V, Weng YC. Lipoca- 37. Hemdahl AL, Gabrielsen A, Zhu C, Eriksson P, He- lin-2 in Stroke. Neuro 2015;2:38–41. din U, Kastrup J, Thorén P, Hansson GK. Expression 32. Hsu WL, Chiou HC, Tung KC, Belot G, Virilli A, of neutrophil gelatinase-associated lipocalin in athe- Wong ML, Lin FY, Lee YJ. The different molecular rosclerosis and myocardial infarction. Arterioscler forms of urine neutrophil gelatinase-associated lipo- Thromb Vasc Biol 2006; 26:136-42. calin present in dogs with urinary diseases. BMC Vet 38. Helanova K, Spinar J, Parenica J. Diagnostic and Res 2014; 10:202. prognostic utility of neutrophil gelatinase-associated 33. Cai L, Rubin J, Han W, Venge P, Xu S. The ori- lipocalin (NGAL) in patients with cardiovascular gin of multiple molecular forms in urine of HNL/ diseases - review. Kidney Blood Press Res 2014; NGAL. Clin J Am Soc Nephrol 2010; 5:2229–35 39:623-9.

334 ORIGINAL ARTICLE

Cardiac biomarkers and left ventricular systolic function in acute myocardial infarction with ST- segment elevation in diabetes mellitus type 2 patients Mirela Halilčević1, Edin Begić2,3, Alen Džubur1, Nirvana Šabanović-Bajramović1, Mevludin Mekić4, Malik Ejubović5, Amela Džubur6 , Ena Štimjanin5

1Department of Cardiology, Clinic for Heart, Blood Vessel and Rheumatic Diseases, Clinical Center University of Sarajevo; 2Department of Cardiology, General Hospital ‘’Prim.dr. Abdulah Nakaš’’, Sarajevo; 3Department of Pharmacology, School of Medicine, Sarajevo School of Science and Technology, Sarajevo; 4Department of Rheumatology, Clinic for Heart, Blood Vessels and Rheumatic Diseases, Clinical Center University of Sarajevo; 5Department of Internal Medicine, Cantonal Hospital Zenica, Zenica; 6Department for Social Medicine, School of Medicine, Univesity of Sarajevo, Sarajevo.

ABSTRACT

Aim To determine a status of systolic function in patients with diabetes mellitus (DM) type 2 with ST-segment elevation acute myocardial infarction (STEMI), to determine values of cardiac biomarkers in patients with DM type 2 with STEMI and correlate the parameters with ejection fraction of left ventricle (EFLV).

Methods A total of 80 patients were divided into two groups, the study group (group I) consisting of 40 patients admitted with the diagnosis of DM type 2 and STEMI, and a control group (group Corresponding author: II) with 40 patients with STEMI without diagnosed DM type 2. Malik Ejubović Cardiac biomarkers - creatine kinase MB fraction (CKMB), and Department of Internal Medicine with troponin I were monitored. The EFLV was evaluated echocardio- Haemodialysis, graphically (using Simpson method) five days after primary percutaneous coronary intervention (pPCI). Cantonal Hospital Zenica Crkvice 67, 72000 Zenica, Results In the group I the EFLV five days after pPCI was signifi- Bosnia and Herzegovina cantly correlated with troponin values (with a minimum r = -0.47; Phone: +387 32 447 143; p=0.002, a maximum r = -0.339; p = 0.032, as well as with an average value of r = -0.389; p=0.013), and with an average CK Fax: +387 61 414 680; value (r = -0.319; p=0.045). In the group II there was a significant E-mail: [email protected] negative correlation of EFLV with the maximum value of troponin Mirela Halilčević ORCID ID: https://orcid. (r = -0.309; p=0.05). org/0000-0003-1583-6567 Conclusion Troponin values have an effect on the EFLV after STEMI, and thus on the left ventricular status, as well as on the Original submission: pharmacological modality itself. 21 October 2019; Revised submission: Keywords: prognosis, ST elevation myocardial infarction, troponin 09 January 2020; Accepted: 27 January 2020 doi: 10.17392/1094-20

Med Glas (Zenica) 2020; 17(2): 335-340

335 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION ssment for major adverse cardiovascular events (MACE) after primary percutaneous intervention Acute coronary syndrome (ACS) covers unstable (pPCI) is therefore focused on systolic function, angina, non-ST elevation myocardial infarction and accordingly, there are numerous studies that (NSTEMI) and ST elevation myocardial infarc- have shown that the ejection fraction of left vention (STEMI) (1). Atherosclerosis accounts for tricle (EFLV) or other closely related parameters about 80% of cardiovascular diseases, and given are powerful guides for selecting therapy and pre- the large percentage of potentially fatal complica- dicting the risk of future events (1,9). tions of atherothrombosis (the process progresses through life, before finally manifesting as an acu- Cardiovascular risk factors that can be influenced te ischemic event), a proper understanding of the by lifestyle and thus eliminated as possible causes pathogenesis of this disease is of great importance of coronary disease are called preventable-varia- for determining optimal modalities of prophylaxis ble risk factors. Variable risk factors include nico- and therapy (1,2). Biomarkers are measurable tinism, dyslipidaemia, obesity, diabetes mellitus and quantitative biological parameters that serve (DM), metabolic syndrome, arterial hypertension, as indicators of health and physiology of asse- increased heart rate and insufficient physical ac- ssment, and in 1979, the World Health Organi- tivity (1). DM in early stage of diagnosis has an zation (WHO) recommended a panel of creatine effect on metabolic imbalance (9-12). kinase (CK), aspartate aminotransferase (AST), Presence of DM as a comorbidity additionally and lactate dehydrogenase (LDH) for the diagno- complicates clinical outcome of patients with the sis of acute myocardial infarction (AMI), while diagnosis of ACS hence it is of critical importan- later, troponin was given its place in the diagnosis ce to stratify those patients adequately (12). The itself after the introduction of immunoassays into combination of DM and cardiovascular disease clinical practice in the 1990s (3). The CK enzyme also has the impact on the gender-independent is responsible for the transfer of the phosphate gro- prevalence of mortality and morbidity (12). It has up from ATP to creatine, and consists of two M been clear for 40 years that DM is a risk factor for and or B subunits, and accordingly distinguishes AMI (13). The existence of silent myocardial is- three forms of CK- MM, CK-MB, and CK-MM. chemia in diabetics, and its early detection, is one The CKMB is considered as sensitive and specific of the imperatives of modern cardiology (13,14). marker for AMI; it begins to grow 3-4 hours af- The aim of the study was to determine a status ter the development of AMI, reaches its peak after of systolic function in patients with DM type 2 10-24 hours, and returns to normal after 72 hours with STEMI, to determine values of serum mar- (3,4). The concentration of troponin in the blood kers in patients with DM type 2 with STEMI, and begins to rise 4-6 hours after the onset of symp- to correlate parameters of systolic function and toms, meaning that blood sample must be taken values of the serum markers in patients with DM on admission and again 6-9 hours later. Maximum and AMI. The aim was also to evaluate factors values of troponin appear 18-24 hours after the that influence patient stratification, and to answer onset of the problem (4-6). The only advantage the question whether such stratification is already of CK-MB over troponins is fast clearance that possible on initial patient admission. Similar stu- helps to detect reinfarction, and therefore serum dies have not been conducted in Bosnia and Her- troponin levels along with the CK-MB fraction le- zegovina. Also, studies that correlate biomarkers vel are sensitive and significant for the diagnosis of cardiac necrosis and systolic function after of myocardial infarction (5). Troponin is released pPCI have not been performed in overall. from dead cells of the heart muscle and the level of troponin in the blood is related to the size of PATIENTS AND METHODS the affected myocardium, or the size of necrosis (6-8). The AMI is characterized by loss of contrac- Patients and study design tile tissue and changes in left ventricular geometry. This prospective study which included 80 pati- This leads to alteration of systolic as well as dia- ents attended to the Department of Cardiology, stolic function, each of which may separately alter Clinic for Heart, Blood Vessel and Rheumatic further clinical course of the disease (9). Risk asse- Diseases, Clinical Center of the University of

336 Halilčević et al. Diabetes mellitus effects in STEMI

Sarajevo between January 2018 and June 2018. for unequal variance where needed. The frequen- The patients were divided into two groups, a cies of the qualitative variables were compared study group (group I) consisting of 40 patients using a chi-square test with continuity correction with DM type 2 admitted and treated under the for 2x2 tables. Relationships between the varia- diagnosis of AMI, and a control group (group II) bles were tested using the parametric Pearson comprising of 40 patients diagnosed with AMI, correlation. All tests were performed with the without diagnosed DM type 2. Inclusion crite- accuracy level of 95% (p<0.05). ria were: chest pain, electrocardiographic signs according to the STEMI criteria (1), increase in RESULTS serum CKMB more than double of normal va- A total of 80 patients with a mean age (SD) of lues, increase in troponin I greater than 0.5 ng/ 64 ± 10 years (ranging from 41 to 89 years) were mL, patients with DM type 2. Exclusion criteria analyzed. No statistical difference was observed were previously diagnosed myocardial infarction regarding the mean age in the study and control or already diagnosed cardiomyopathy, diagnosed group (66±10 and 63±10 years, respectively; with DM type 2 during hospitalization, patients p=0.19). In the overall sample, there were 51 who left the clinic before completing the tre- (63.7%) male and 29 (36.3%) female patients (the atment or patients who exited during hospitaliza- ratio of 1.76:1). In group I there were 20 (50%) tion (intrahospital death). male patients versus group II where there were Ethical approval was obtained from the Ethics 31 (77.5%) males (p=0. 02). In group I of patients Committee of the Clinical Center of the Univer- with DM type 2 in 90% of cases diagnosis was sity of Sarajevo. established within five years. In group I, a total of 27 (67.5%) patients were on oral antidiabetic Methods drugs, 12 (30%) were on insulin, and one patient Patients on admission had creatine kinase, (2.5%) was on combined treatment. In this group CKMB, and troponin I values (reference values: of patients, the average measured value of gluco- 22 to 198 U/L, 0-25 IU/L, and up to 0.5 ng/mL, se was 7.05±1.0 mmol/L and the average HbA1c respectively) analyzed. These analyses were re- level was 11.71±3.45%. When it comes to risk peated after 8, 16 and 24 hours. factors, both hypertension and dyslipidemia were significantly higher in the group I (p=0.04 and Five days after admission, patients underwent p=0.00, respectively). No statistically significant echocardiographic evaluation of EFLV. The difference was found in the smoking habit betwe- EFLV was assesed by Simpson method (1). en the study groups (p=0.50 ) as well as family All patients underwent pPCI and were previously history of heart diseases (p=1.0). There was no treated with antiaggregation and anticoagulation significant difference in the frequency of STEMI therapy. The existence of the diagnosis of arterial localisation between the study groups (p=0.68). hypertension in medical history was analyzed, as No statistically significant difference of average well as the existence of the diagnosis of dysli- values of EFLV between the study groups was pidemia (elevated cholesterol over 5.8 millimo- found (44.85±8.82 vs. 45.60±8.62; p=0.70). No les per liter (mmol/L) or triglyceride values over statistically significant difference of average val- 1.85 mmol/L) and smoking habit. Patients in the ues of troponin I between the study groups was first group were divided into those with HbA1C found for all measured values (p=0.64). Average values below and above 6.5%. CK values were higher in the control group of Statistical analysis patients (p=0.06). Significantly higher values of CKMB in group I was found (48.63 ± 30.35vs. All variables were tested for normal distribution 35.15 ±21.85; p=0.03) (Table 1). using the Kolmogorof-Smirnov test. All variables There was no statistically significant difference are presented descriptively using appropriate me- (p=0.92) in comparing the level of troponin valu- asures of central tendency (arithmetic mean and es with the localization of infarction. There was median) and dispersion (standard deviation and no statistically significant difference (p=0.10) interquartile range). Quantitative variables were in comparing the level of troponin values with compared using Student's t-test with correction

337 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 1. Ejection fraction of left ventricle (EFLV) and cardiac to prognosis, patients with coronary artery dise- enzymes in patients with (group I) and without (group II) ase and DM usually have worse outcome than diabetes mellitus (DM) type 2 patients without DM (16). On the other hand, Variable Group Mean ±SD) I 44.85±8.82 atherosclerosis itself has vascular manifestations EFLV (%) II 45.60±8.62 on three levels: coronary arteries, carotid arteries I 8.14±10.49 Troponin I minimal (ng/mL) and lower limbs arteries. Abnormal metabolism II 7.22±6.82 I 31.88±33.75 connected with diabetes is a common cause of Troponin I maximal (ng/mL) II 21.25±25.53 arterial disfunction (16,17). I 20.01±20.21 Troponin I mean (ng/mL) II 14.24±15.25 Hyperglicemia, which is common in diabetic pa- I 331.18±229.34 tients, inhibits production of nitric oxide (NO) CK (mean) (IU/L) II 251.73±135.63 by blocking synthesis of endothelial nitric oxide I 48.63±30.35 CKMB (mean) (IU/L) synthethasis (eNOS) and increases production of II 35.15±21.85 SD, standard deviation; CK, creatine kinase; CKMB, creatine kinase reactive oxygen species, above all, superoxide MB fraction; anion in endothelial and vascular smoothe muscle the EFLV to both groups. There was no signifi- cells (18). In the Multiple Risk Intervention Trial cant difference in EFLV values between the two (MRFIT) 9.7% (out of 5163 male patients) died subgroups of group of patients with DM type 2 due to cardiovascular diseases within the 12-year (p=0.18). The type of antidiabetic therapy had no period (19). A study conducted by Haffner et al. effect on EFLV (p=0.47). In group I, the EFLV (1373 non-diabetic and 1059 patients previously five days after pPCI significantly correlated diagnosed with DM type 2) during a 7-year period with all troponin I values (minimum r = -0.47; found an increase in incidence of AMI in diabe- p=0.002, a maximum r = -0.339; p=0.032, and an tic patients regardless of their medical history of average value of r = -0.389; p=0.013), and with previous myocardial infarction (20). Furthermore, an average of CK (r = -0.319; p=0.045). In group Haffner et al. showed that DM increased the inci- II of the patients there was a marginal significant dence of both, early and late complications in pati- negative correlation of EFLV with the maximum ents with acute coronary syndrome (20). value of troponin (r = -0.309; p=0.05) (Table 2). According to a trial by the Organisation to Assess Strategies for Ischemic Syndromes (OASIS), a Table 2. Correlation of left ventricular systolic function (LVSF) and cardiac enzymes in patients with (group I) and 57%-increase of mortality rate was found in pa- without (group II) diabetes mellitus five days after primary tients with unstable angina and AMI due to DM. percutaneous intervention (pPCI) A poorer long-term prognosis after the myocardial CK CKMB Patient’s Troponin I (ng/mL) infarction, including reccurent myocardial infarc- (mean) (mean) group Miminimal Maximal Mean IU/L IU/L tion, heart failure and sudden cardiac death, has Correlation I -0.470 -0.339 -0.389 -0.319 -0.177 been found among patients with DM (21). Feng et coefficient al. have found that glycolised haemoglobin con- p 0.002 0.032 0.013 0.045 0.275 Correlation centration affected major adverse cardiovascular II 0.125 -0.309 -0.226 0.098 0.121 coefficient events (MACE) in patients with AMI (22). In ad- p 0.444 0.052 0.162 0.549 0.459 dition, Lipsic et al. have found that lower levels of CK, creatine kinase; CKMB, creatine kinase MB fraction haemoglobin are associated with higher short-term DISCUSSION mortality in patients with acute MI (23). Our results have failed to show the connection between Average time between onset of symptoms of the glycolised haemoglobin and EFLV five days Type 1 DM and the diagnosis is 15 days, while after STEMI. However, taking in consideration the this time in Type 2 DM is from 6 to 10 years (15). number of patients in this research, these results The Atherosclerosis Risk in Communities Study can lead to new studies, which could confirm that (ARIC) trial proved that values of HbA1c close unregulated DM is likely to affect EFLV. More reli- to upper normal limit can still be one of the risk able results could be achieved by following the de- factors for the development of cardiovascular di- velopment of MACE, as well as analysing the cor- seases, and that HbA1c is a reliable risk factor onarography findings. In group 1, the EFLV five of all-cause and cardiovascular mortality in both days after pPCI was significantly correlated with diabetics and non-diabetics (15). When it comes

338 Halilčević et al. Diabetes mellitus effects in STEMI

minimum troponin values, which are essentially They could be used as a better predictor of EFLV the initial values of tropinin I on admission. Maxi- as compared to the maximum values. A limitation mum values also correlate significantly with EFLV. of the study is a relatively smaller sample, and it In patients without the diagnosis of DM, only max- should be considered as an initial one. Nonethele- imal values of troponin I were significantly corre- ss, there is no large number of studies that compa- lated. Cubbon et al. have found that the mortality re the minimum and maximum values of troponin risk rapidly increased among male patients with I as a predictor factors of EFLV. Furthermore, to- the growth of glycaemia as compared to female tal CK values are much better predictor of EFLV patients (24). Furthermore, a study by Deedwania for patients with DM as compared to the CKMB et al. has found that hyperglycaemia increases the values. They could indicate the importance of pa- mortality rate among patients with acute coronary tient observation and anticipation of new values syndrome (25). Monterio et al. have confirmed that of troponin after three to four hours in emergency hyperglycaemia is a bad prognostic marker and centres, even when troponin itself has borderline strong predictor of intrahospital mortality (26). All values and ECG signs are not convincing. The patients in their research were treated with metfor- increase in troponin should be a determinant for min and sulphonylurea (SU) with or without insu- the diagnosis of ACS, with mandatory interpreta- lin therapy (26). It is well known that the derivate tion and correlation with total CK values, which of SU, meglitinides and thiazolidinediones have is very important in emergency centers, when higher risk of new-onset acute coronary syndrome the MB fraction is not available. It is necessary as compared with insulin (27). Contemporary ther- to emphasize that the only advantage of CK-MB apy of DM recommends the usage of inhibitors of over troponin is in detecting reinfarction after sur- dipeptidyl peptidase-4, Sodium-glucose co-trans- gical revascularization, while in all other cases porter-2 and glucagon-like peptide-1 agonists in troponin should be used as the main determinant. patients with a high cardiovascular risk (28). Some authors have also suggested using CK-MB Our results have shown the relationship between to CK ratio to improve specificity, however this values of cardiac troponin and systolic function of approach significantly reduces sensitivity (31). left ventricle five days after STEMI. A weak ne- In the conclusion, we have failed to show the gative statistical correlation between EFLV and connection between the values of glycolised hae- maximal value of cardiac troponin has been found moglobin and the modality of treatment of DM in non-diabetic patients. On the other hand, a si- and the EFLV after the diagnosis of MI. Values gnificant correlation between EFLV and all values of troponin affect EFLV after STEMI. Optimi- of troponin, as well as average value of creatinine zed pharmacological treatment is an imperative kinase has been found in diabetes patients. Khan et in preventing future cardiovascular incidents and al. have failed to confirm that the increase in valu- better control of DM and its complications. More es of cardiac troponin would lead to lower EFLV extensive research is needed. (29). Furthermore, Shah et al. have found that value of cardiac troponin could lead to remodelling FUNDING of left ventricle (30). These results are likely to be No specific funding was received for this study. connected to timing of reperfusion therapy, as well as the patient’s profile itself. TRANSPARENCY DECLARATION Our research has shown the importance of the mi- Competing interests: None to declare. nimum values of troponin I in patients with DM.

REFERENCES 1. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax 2. Rafieian-Kopaei M, Setorki M, Doudi M, -Barada JJ, Morrow DA, White HD; Executive Group on ran A. Nasri H. Atherosclerosis: process, indicators, behalf of the Joint European Society of Cardiology risk factors and new hopes. Int J Prev Med 2014; (ESC)/American College of Cardiology (ACC)/ 5:927-46. American Heart Association (AHA)/World Heart 3. Garg, P, Morris P, Fazlanie AL, Vijayan S, Dancso B, Federation (WHF) Task Force for the Universal Dastidar A, Dastidar AG, Plein S, Mueller C, Haaf Definition of Myocardial Infarction. Fourth Univer- P. Cardiac biomarkers of acute coronary syndrome: sal Definition of Myocardial Infarction. Glob He- from history to high-sensitivity cardiac troponin. In- art 2018; 13:305-38. tern Emerg Med 2017; 12:147-55.

339 Medicinski Glasnik, Volume 17, Number 2, August 2020

4. Sharma S, Jackson PG, Makan J. Cardiac tropo- 20. Haffner SM, Letho S, Romnema T, Pyörälä K, Laak- nins. J Clin Pathol 2004; 57:1025–6. so M. Morality from coronary heart disease inpati- 5. Daubert MA, Jeremias A. The utility of troponin me- ents with typ 2 diabetes and in nondiabetes patients asurement to detect myocardial infarction: review of with and without prior myocardial infaction. N Engl the current findings. Vasc Health Risk Manag 2010; J Med 1998; 339:229-224. 6:691–9. 21. Malmberg K, Yusuf S, Gerstein HC. Impact of dia- 6. Jaffe AS. Chasing troponin: how low can you go betes on long-term prognosis in patients with unsta- if you can see the rise? J Am Coll Cardiol 2006; ble angina and non-Q-wave myocardial infarction: 48:1763–4. results of the OASIS (Organization to Assess Strate- 7. Higgins JP, Higgins JA. Elevation of cardiac tropo- gies for Ischemic Syndromes) Registry. Circulation nin I indicates more than myocardial ischemia. Clin 2000; 102:1014-19. Invest Med 2003; 26:133–47. 22. Feng QZ, Zhao YS, Li YF. Effect of haemoglobin 8. Hodzic E, Durek A, Begic E, Sabanovic-Bajramovic concentration on the clinical outcomes in patients N, Durak-Nalbantic A. Effect of myocardial infarc- with acute myocardial infarction and the factors re- tion on the occurrence of erectile dysfunction. Med lated to haemoglobin. BMC Res Notes 2011; 4:142. Glas (Zenica) 2019; 16:5-39. 23. Lipsic E, van der Horst IC, Voors AA, van der Meer 9. Cleland JG, Torabi A, Khan NK. Epidemiology and P, Nijsten MW, van Gilst WH, van Veldhuisen management of heart failure and left ventricular DJ, Zijlstra F. Hemoglobin levels and 30-day mor- systolic dysfunction in the aftermath of a myocardial tality in patients after myocardial infarction. Int J infarction. Heart 2005; 91:ii7–ii48. Cardiol 2005; 100:289-92. 10. Woods KL, Samanta A, Burden AC. Diabetes melli- 24. Cubbon R, Rajwani A, Abbas A, Gale C, Grant P, tus as a risk factor for acute myocardial infarction in Wheatcroft S, EMMACE-2 (Evaluation of Methods Asians and Europeans. Br Heart J 1989; 62:118–22. Management of Acute Coronary Events) Study Gro- 11. Einarson TR, Acs A, Ludwig C, Panton UH. Preva- up. Hyperglycaemia, in relation to sex, and mortality lence of cardiovascular disease in type 2 diabetes: after acute coronary syndrome. Eur J Prev Cardiol a systematic literature review of scientific evidence 2007; 14:666-71. from across the world in 2007-2017. Cardiovasc 25. Deedwania P, Kosiborod M, Barrett E, Ceriello A, Diabetol 2018; 17:83. Isley W, Mazzone T, Raskin P. Hyperglycemia and 12. Leon BM, Maddox TM. Diabetes and cardiovascular acute coronary syndrome. Anesthesiology 2008; disease: Epidemiology, biological mechanisms, tre- 109:14-24. atment recommendations and future research. World 26. Monteiro S, Monteiro P, Gonçalves F, Freitas M, J Diabetes 2015; 6:1246–58. Providência L. Hyperglycaemia at admission in 13. Fuller JH, Shipley MJ, Rose G, Jarrett RJ, Keen H. acute coronary syndrome patients: prognostic value Mortality from coronary heart disease and stroke in in diabetics and non-diabetics. Eur J Prev Cardiol relation to degree of glycaemia: the Whitehall stu- 2010; 17:155-9. dy. Br Med J (Clin Res Ed) 1983; 287:867–70. 27. Lin TT, Wu CC, Yang YH, Lin LY, Lin JL, Chen PC, 14. Matheus AS, Tannus LR, Cobas RA, Palma CC, Ne- Hwang JJ. Anti-hyperglycemic agents and new-on- grato CA, Gomes MB. Impact of diabetes on cardi- set acute myocardial infarction in diabetic patients ovascular disease: an update. Int J Hypertens. 2013; with end-stage renal disease undergoing dialysis. 2013:653789. PLoS One 2016; 11:e0160436. 15. Selven E, Steffes MW, Zhu H. Glycated Hemoglo- 28. Schmidt AM. Diabetes mellitus and cardiovascular bin, diabetes, and Cardiovascular risk in nondiabetic disease emerging therapeutic approaches. Arterios- Adults. N Engl J Med 2010; 362:800-811. cler Thromb Vasc Biol 2019; 39:558-68. 16. Beckman JA, Creager MA, Libby P. Diabetes and 29. Khan MH, Islam MN, Aditya GP, Islam Atherosclerosis. JAMA 2002; 287: 2570-81. MZ, Bhuiyan AS, Saha B, Bhowmick K, Hassan 17. Agiolillo DJ. Antiplatelet therapy in type 2 diabetes M, Paul PK, Chowdhury S, Rahman R. Correlati- mellitus. Curr Opin Endocrinol Diabetes Obes 2007; on of troponin-IlLevel with left ventricular ejection 14:124–31 fraction and in-hospital outcomes after first attack 18. De Vriese AS, Verbeuren TJ, Van de Voorde J. Endo- of non-ST segment elevation myocardial infarction. thelial dysfunction in diabetes. Br J Pharmaco 2000; Mymensingh Med J 2017; 26:721-31. 130:963-74. 30. Shah RV, Chen-Tournoux AA, Picard MH, Januzzi 19. Stemler J, Vaccaro O, Neaton D, Wentworth D. Dia- JL. Association between troponin T and impaired betes and other risk factors and 12- yr crdiovascular left ventricular relaxation in patients with acute de- mortality for men screend in Multiple risk interven- compensated heart failure with preserved systolic tion Trial. Diabetes Care 1993; 16:434-44. function. Eur J Echocardiogr 2009; 10:765–8. 31. Mythili S, Malathi N. Diagnostic markers of acute myocardial infarction. Biomed Rep 2015; 3:743–8.

340 ORIGINAL ARTICLE

Analysis of two single-nucleotide polymorphisms (rs2241766 and rs1501299) of the adiponectin gene in patients with coronary artery disease and coronary slow flow

Muhammad Diah1,2, Aznan Lelo3, Zulfikri Muhktar4, Dharma Lindarto5

1Division of Cardiology, Department of Internal Medicine, School of Medicine, Syiah Kuala University/Dr. Zainoel Abidin General Centre Hospital, 2School of Medicine, Universitas Sumatera Utara, 3Department of Pharmacology, 4Department of Cardiology, 5Division of Endocrine Metabolic; Department of Internal Medicine, Universitas Sumatera Utara

ABSTRACT

Aim To investigate the association of two single-nucleotide polymorphisms (SNPs) of the adiponectin gene (+45 T>G [rs2241766] and +276 G>T [rs1501299]) in patients with coronary artery disease (CAD), coronary slow flow (CSF) and in healthy subjects.

Methods The study investigated 90 patients: 30 patients with coronary artery disease (the CAD group), 30 with only coronary slow flow (the CSF group) and 30 healthy patients. Genotyping was carried out for two of these SNPs in the adiponectin gene using the Corresponding author: TaqMan polymerase chain reaction (PCR) method. Muhammad Diah Results There were no significant differences in the frequencies of School of Medicine, polymorphism +45T>G (rs2241766) genotype (p=0.464) and the Universitas Sumatera Utara allele (p=0.362). There were also no significant differences betwe- Dr. Mansyur Street No. 5, Padang Bulan, en the +276 G>T (rs1501299) genotype and the allele (p=0.624 Medan, Indonesia and p=0.281, respectively). Phone: +62 811 682 2238; Conclusion Our study indicated that in patients with CAD, CSF, E-mail: [email protected] and in healthy subjects, the SNP +276G>T and +45T>G alleles of ORCHID ID: https://orcid.org/0000-0002- the adiponectin gene were not associated. 1183-8621 Key words: genetic association, cardiovascular disease, coronary angiography

Original submission: 27 November 2019; Revised submission: 28 January 2019; Accepted: 31 January 2019 doi: 10.17392/1102-20

Med Glas (Zenica) 2020; 17(2): 341-345

341 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION PATIENTS AND METHODS Coronary artery disease (CAD) is one of the most Patients and study design common cardiovascular diseases. The World He- alth Organization has issued several reports in- The study was conducted between December dicating that since 2008, cardiovascular disease 2017 and February 2018 in the Division of Car- (CVD) has become the principle cause of death diology, Department of Internal Medicine, Scho- worldwide. One report estimated that mortality ol of Medicine, University Syiah Kuala and the caused by CVD would rise from 17.1 million in Cardiac Catheterization Laboratory at Dr. Zai- 2004 to 23.4 million in 2030 (1). noel Abidin Regional General Hospital, Banda CAD involves genetic and environmental factors Aceh, Indonesia. and their interactions. Traditional risk factors Patients involved were aged 26 years and above account for more than half of the prevalence of and were consecutively recruited. Patients who CAD, and despite attempts conducted to esta- had experienced clinical symptoms of chest pain blish the molecular and genetic determinants that with typical angina problems were divided into may be accountable for variations in CAD, the CAD and CSF groups, based on the results of etiology and complex multigenic bases of athe- coronary angiography. Patients who did not have rosclerosis are still not completely understood chest discomfort as confirmed by electrocardi- (2, 3). Adiponectin is one of the most intensi- ography and had not had coronary angiography vely discussed secretion products of white fat were healthy group. Patients with total stenosis cells that has been implicated increasingly in the of coronary arteries and coronary artery bypass pathogenesis of atherosclerosis and in insulin re- grafts were excluded. Patients with significant sistance, and data on the prospective impacts of comorbidities, including hypertension, diabe- adiponectin plasma-concentration determination tes mellitus, the use of anti-inflammatory drugs in cardiovascular diseases in humans are evol- other than aspirin, renal or hepatic dysfunction, ving. Several clinical studies have demonstrated body mass index (BMI) of >22.9 kg/m2 or <18.5 strong correlations between low plasma adipo- kg/m2, smoking, and alcohol consumption, were nectin levels and CAD (4). excluded from the healthy group. It is therefore possible that functional genetic The protocol of this cross-sectional study was polymorphisms that may alter the expression level approved by the Ethical Review Committee, of adiponectin may also affect individual suscepti- School of Medicine, Universitas Sumatera Utara, bility to CAD. Two of the most commonly studied Medan, Indonesia. Prior to the study, all patients SNPs are the silent T to G substitution in exon 2 provided written informed consents. (+45T>G) and the G to T substitution in intron 2 Methods (+276G>T). However, inconsistent findings regarding the association of these two SNPs, either in- Venous blood samples were taken and tested for dependently or as a haplotype have been reported hemoglobin, hematocrit, white blood cells, pla- (5, 6). The two SNPs +45T>G19–22 and +276G>T telets, erythrocyte sedimentation rate, total cho- 23–25 have been repeatedly found to correlate lesterol, low-density lipoprotein (LDL) choleste- with CAD, albeit with some controversial studies rol, high-density lipoprotein (HDL) cholesterol, concluding that there was no association with CAD triglycerides, urea, creatinine, uric acid, fasting and related complications, findings that may be due glucose, and 2-hour post prandial glucose, using to differences in ethnic populations studied (7). To standard methods. clarify these conflicting results, the data regarding Coronary angiography and thrombolysis in the association between adiponectin polymorphi- myocardial infarction (TIMI). The standard sms and CAD require further investigation. Judkins technique was employed to perform co- Aim of this study was to investigate the associa- ronary angiography of the participants (8). The tion between two single-nucleotide polymorp- angiography was recorded at the left anterior hisms (SNPs), +45 T>G (rs2241766) and +276 oblique, cranial, right anterior oblique, caudal, G>T (rs1501299) of the adiponectin gene in pa- and horizontal positions. These examinations tients with CAD, CSF, and in healthy subjects. were carried out by two cardiologists who had

342 Diah et al. Adiponectine gene

not had any information about the clinical charac- parisons were conducted using suitable testing. teristics of the patients. In addition, the coronary Paired-sample t-testing was used to compare artery flow of the patients was assessed using the dependent groups. Genotype and allele frequ- TIMI frame-count method (9) and the differences encies for +45 T>G (rs2241766) and +276 G>T between the first and the last frames in the count (rs1501299) polymorphisms of the adiponectin were calculated. The cut-off values for the len- gene were determined by direct counting. Sta- gth for normal visualization of coronary arteries tistical comparisons between the three groups were 36.2±2.6 frames for the left anterior descen- were performed by applying the chi-squared test ding artery (LAD), 22.2±4.1 frames for the left and Fisher’s exact test with p<0.05 being consi- circumflex artery (LCx), and 20.4±3 frames for dered as statistically significant. the right coronary artery (RCA). The corrected cut-off value for the LAD was 21.1±1.5 frames. RESULTS The mean TIMI frame count for each subject was The study comprised 90 patients: 30 in the CAD, calculated by dividing the sum of the TIMI frame 30 in the CSF and 30 in the healthy group (Table 1). counts for LAD, LCx, and RCA by three. Based on collected clinical data, there was a sta- Determination of adiponectin genotyping. Ge- tistically significant difference between the CAD, nomic DNA was extracted from 5 mL of who- CSF, and the healthy groups in terms of age, sex, le blood using a commercially available DNA and systolic blood pressure (SBP) (p<0.001). extraction kit (Genomic DNA Mini Kit, Ge- Total cholesterol, serum triglycerides, and crea- neaid, USA) according to the manufacturer’s tinine were significantly higher in patients with instructions. Researchers screened two SNPs CAD compared to the patients with CSF and the in rs2241766 and rs1501299 of the adiponectin healthy ones (p<0.001). gene in all patients with assays using quantitati- There was no significant difference in the TIMI ve real-time TaqMan PCR (Applied Biosystems, frame counts for RCA, LCx, LAD, corrected Foster City, CA, USA). TIMI frame count (cTFC) LAD, and cTFC betwe- Statistical analysis en the groups. The most common target vessel in the CSF patients was RCA (93.3%) (Table 2). Conformity to normal distribution of the conti- The distribution of +45 T>G and +276 G>T alle- nuous variables was examined using the Shapi- les and genotypes was comparable between the ro–Wilk test. Parametric or non-parametric tests patients with CAD, CSF, and the healthy group. In were performed based on the data distribution 90 patients, there were no significant differences and the number of groups, and post hoc com- in the frequencies of polymorphisms of +45T>G

Table 1. Demographic and clinical characteristics of coronary artery disease (CAD), coronary slow flow (CSF) and healthy patients Variable CAD (n=30) CSF (n=30) Healthy (n=30) p Age (years) (No, %) 55 (37-65) 52 (37-65) 35 ( 27-51) <0.001 Male female (No, %) 29 (96%) /1 (4%) 10 (33%) / 20 (67%) 17 (56%)/13 (44%) <0.001 BMI (kg/m2) (mean±SD, median minimum-maximum) 23.7 (21.6-29.0) 22.7 (18.1-31.2) 23.2 (18.5-25.1) 0.08 SBP (mmHg) 130 (112-152) 126.50 (109 -147) 116.50 (107-128) <0.001 DBP (mmHg) 84.30±7.13 84.93±9.63 78.10±8.84 0.004 Hemoglobin (g/dL) 14.2 (10.1-17.1) 13.2 (8.7–16.6) 13.2 (11.6-17.5) 0.009 Hematocrit ( %) 42.17±4.93 39.90±4.85 39.93±3.44 0.085 WBC (uL) 8350 (5900-14900) 8700 (4600-16600) 7700 (4500-10100) 0.039 Platelet (103/uL) 263.4±59.4 270.7±74.2 278.1±65.2 0.697 Neutrophil-lymphocyte count ratio 1.6 (0.8-19) 1.3 (0.8-5.1) 1.3 (0.8 -1.7) 0.002 Total cholesterol (mg/dL) 198 (112-461) 194 (127-336) 149 (90-198) <0.001 LDL cholesterol (mg/dL) 50.4±18.31 49±17.7 64.8±14.5 0.001 HDL cholesterol (mg/dL) 121.5 (30-225) 129 (50-252) 111 (62-156) 0.623 Triglyceride (mg/dL) 143.5 (88-231) 106 (55-373) 94 (21 -129) <0.001 Urea (mg/dL) 30.9±11.03 23,8±6,62 23.3±5.20 0.004 Creatinine ( mg/dL) 1.0 (0.6-1.6) 0.8 (0.6 -1.3) 0.7 (0,5 – 1,0) <0.001 Fasting glucose (mg/dL) 101 (71-218) 109 (89 – 154) 98 (82 – 117) 0.013 2 hours post prandial glucose (mg/dL) 129 (100 – 226) 133 (103 -226) 126 (97 -176) 0.214 Data were presented as mean±SD, median (minimum-maximum) or n (%); BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; ESR, erythrocyte sedimentation rate; LDL, low-density lipoprotein; HDL, high-density lipoprotein

343 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 2. Thrombolysis in myocardial infarction (TIMI) frame ted with CSF. In our study, age, sex, BMI and counts of coronary artery disease (CAD) and coronary slow SBP were significantly different. There were also flow (CSF) patients statistically significantly differences in the para- Variable CAD (n=30) CSF (n=30) p TIMI frame (mean±SD) meters for lipids and creatinine. RCA 28.07±8.82 26.83±3.87 0.486 Results of the present study showed that the most LCX 34.4±10.34 28.37±2.76 0.003 LAD 47.27±10.11 42.37±7.77 0.04 common target vessel in the CSF group was cTFCLAD 27.8±5.9 24.67±4.8 0.029 RCA. These results seem to be in accordance cTFC 30.13±7.94 26.83±3.553 0.042 with previous reports on angiographic characte- Target vessel (No, %) RCA 22 (73.3) 28 (93.3) 0.038 ristics in patients with CSF, and RCA has been LCx 4 (66.6) 25 (83.3) 0.136 identified as the predominant vessel involved LAD 26 (86.6) 25 (83.3) 0.718 in CSF (11). However, another study found that RCA, right coronary artery; LCx, left circumflex artery; LAD, left the most common artery involved was LAD, anterior descending artery; cTFC, corrected thrombolysis in myocardial infarction frame count followed by LCX, and RCA (12). Adiponectin is a 244-amino-acid protein secre- (rs2241766) genotype (p=0.464) and allele ted exclusively by adipocytes. This protein has (p=0.362) in all three groups of patients (Table 3). been proposed as protecting against CVD via its Table 3. Adiponectin +45T>G (rs2241766) genotype and metabolic anti-inflammatory effects mediated by allele distribution in patients with coronary artery disease crosstalk between the cAMP-PKA and the NF (CAD), coronary slow flow (CSF) and healthy patients Kappa B signaling pathways (13). As found in No (%) of patients SNP Genotype p the study by Matsubara et al, adiponectin also sti- CAD (n=30) CSF (n=30) Healthy (n=30) TT 19 (63.3) 21 (70) 22 (73) mulated fatty-acid oxidation, decreased plasma +45T>G TG 8 (26.6) 6 (20) 8 (26) 0.464 triglycerides (TG), and improved glucose meta- GG 3 (10) 3 (10) 0 (0) bolism by increasing insulin sensitivity. Howe- T 46 (76.6) 48 (80) 52 (86.6) Allele 0.362 G 14 (23.3) 12 (20) 8 (13.3) ver, the lack of consistent data on the association SNP, single-nucleotide polymorphism between adiponectin and its genetic variants and the risk of CAD means that the precise role of In terms of the frequency of +276 G>T this protein in CAD and CVD in general is still (rs1501299) genotype and allele, there was no unclear (14). significant difference (p=0.624 and p=0.281, respectively) (Table 4). Several studies have been conducted into the association of adiponectin polymorphisms with Table 4. Adiponectin +276 G>T (rs1501299) genotype and CAD, such as the study by Ohashi et al. who fo- allele distribution in patients with coronary artery disease (CAD), coronary slow flow (CSF) and healthy patients und no association of SNP276 and SNP94 in 383 No (%) of patients Japanese patients with confirmed coronary heart SNP Genotype p CAD (n=30) CSF (n=30) Healthy (n=30) disease and 368 healthy controls (15). Lacque- GG 16 (53.3) 17 (56.6) 12 (40) mant et al. reported that among 162 Caucasians +276G>T GT 10 (33.3) 11 (36.6) 13 (43.3) 0.624 with type 2 diabetes, SNP+45 showed significant TT 4 (13.3) 2 (6.7) 5 (16.7) G 42(70%) 45 (45) 37 (61.6%) association with increased CAD risk, whereas Allele 0.281 T 18 (30) 15 (25) 23 (38.3) SNP+276 failed to show any association (16). SNP, single-nucleotide polymorphism On the other hand, in a study recently reported by DISCUSSION Filippi et al. among 325 CAD patients and 270 members of the control group, SNP+276 showed CSF has demonstrated to be more common in significant association with CAD (17). males, smokers, and individuals with hyperli- Bacci et al. reported that among 376 Caucasians pidemia, metabolic syndrome, and/or obesity. with type 2 diabetes, SNP+276 showed signifi- There have been many studies conducted by cant association with CAD. This association was examining clinical features of CSF, such as the independent of serum adiponectin level (18). A one conducted by Yilmaz et al. in a Turkish po- recent study by Jung et al. did not find any corre- pulation (10). In this study, BMI, glucose levels, lation between SNP+45 and SNP+276 and the lipid derangements, and metabolic equivalents presence of CAD (19). (METs) were identified as significantly associa-

344 Diah et al. Adiponectine gene

In our study, there were no significant associati- +45 T>G and +276 G>T in patients with CAD, ons between SNP+45 and SNP+276 in patients CSF, and healthy subjects. The researchers su- with CAD, CSF, or the healthy ones. The diffe- ggest that further studies be conducted with lar- rence between our study and the other one is a ger study populations. relatively small number of patients. However, there are no other studies recently conducted or FUNDING published on polymorphism and adiponectin in No specific funding was received for this study. patients with CAD and CSF in Indonesia. TRANSPARENCY DECLARATION In summary, this study found no significant associations between adiponectin polymorphisms Competing interests: None to declare.

REFERENCES 1. Global Burden of Disease Study 2013 Collaborators. 11. Dai YX, Li CG, Huang ZY, Zhong X, Qian JY, Liu Global, regional, and national incidence, prevalen- XB, Gel L, Fan B, Wang QB, Zhang F, Huang D, ce, and years lived with disability for 301 acute and Yao K, Ma JY, Zhu MH, Ge JB. Clinical and an- chronic diseases and injuries in 188 countries, 1990– giographic characteristics of patients with slow co- 2013: a systematic analysis for the global burden of ronary flow. Zhonghua Xin Xue Guan Bing Za Zhi disease study 2013. Lancet 2015; 386:743–800. 2011; 39:642-6. 2. Zhong C, Zhen D, Qian Q, Genshan M. A lack of 12. Mukhopadhyay S, Kumar M, Yusuf J, Gupta VK, association between adiponectin polymorphisms Tyagi S. Risk factors and angiographic profile of co- and coronary artery disease in a Chinese populati- ronary slow flow (CSF) phenomenon in North Indi- on. Genet Mol Biol 2010; 33:428–33. an population: an observational study. Indian Heart 3. Zdravkovic S, Wienke A, Pedersen NL, Marenberg J 2018; 70:405–9. ME, Yashin AI, De Faire U. Heritability of death from 13. Ouchi N, Kihara S, Arita Y, Okamoto Y, Maeda K, coronary heart disease: a 36-year follow-up of 20,966 Kuriyama H, Hotta K, Nishida M, Takahashi M, Swedish twins. J Intern Med 2002; 252:247–54. Muraguchi M, Ohmoto Y, Nakamura T, Yamashita 4. Schnabel R, Messow CM, Lubos E, Klein CE, S, Funahashi T, Matsuzuwa Y. Adiponectin, an adi- Rupprecht HJ, Bickel C, Sinning C, Tzikas S, Keller pocyte-derived plasma protein, inhibits endothelial T, Genth-Zotz S, Lackner KJ, Münzel TF, Blenke- NF-kappaB signaling through a cAMPdependent nberg S. Association of adiponectin with adverse pathway. Circulation 2000; 102:1296–301. outcome in coronary artery disease patients: re- 14. Matsubara M, Maruoka S, Katayose S. Decrea- sults from the Athero Gene study. Eur Hear J 2008; sed plasma adiponectin concentrations in women 29:649–57. with dyslipidemia. J Clin Endocrinol Metab 2002; 5. Menzaghi C, Trischitta V, Doria A. Genetic influ- 87:2764–9. ences of adiponectin on insulin resistance, type 2 15. Ohashi K, Ouchi N, Kihara S, Funahashi T, Naka- diabetes, and cardiovascular disease. Diabetes 2007; mura T, Sumitsuji S. Adiponectin I164T mutati- 56:1198–209. on is associated with the metabolic syndrome and 6. Zhou D, Jin Y, Yao F, Duan Z, Wang Q, Liu J. Asso- coronary artery disease. J Am Coll Cardiol 2004; ciation between the adiponectin +45 T>G genotype 43:1195–200. and risk of cardiovascular disease: a meta-analysis. 16. Lacquemant C, Froguel P, Lobbens S, Izzo P, Dina Heart Lung Circ 2014; 23:159–65. C, Ruiz J. The adiponectin gene SNP+45 is associa- 7. Ghazouani L, Elmufti A, Baaziz I, Chaabane I, Ben ted with coronary artery disease in type 2 (non-insu- Mansour H. Contribution of adiponectin polymor- lin-dependent) diabetes mellitus. Diabet Med 2004; phisms to the risk of coronary artery disease in a 21:776–81. North-African Tunisian population. J Clin Lab Anal 17. Filippi E, Sentinelli F, Romeo S, Arca M, Berni A, 2018; 32:e22446. Tiberti C. The adiponectin gene SNP+276G>T asso- 8. Judkins MP. Percutaneous transfemoral selective ciates with early-onset coronary artery disease and coronary angiography. Radiol Clin North Am 1968; with lower levels of adiponectin in younger coro- 6:467-92. nary artery disease patients (age

345 ORIGINAL ARTICLE

Interleukin 6 concentration elevation as a risk of carotid intima- media thickness in chronic kidney disease patients with dialysis Riri Andri Muzasti1, Herman Hariman2, Elvita Rahmi Daulay3

1Division of Nephrology and Hypertension, Department of Internal Medicine, 2Department of Clinical Pathology, 3Department of Radiol- ogy; Universitas Sumatera Utara, Medan, Indonesia

ABSTRACT

Aim To investigate the relationship between IL-6 concentration and the risk of carotid artery calcification in chronic kidney disease patients with dialysis (CKD-5D).

Methods This analytic observational cross-sectional study includes 95 clinically stable patients who underwent regular haemodialysis for at least three months at Rasyida Renal Hospital Medan, Indone- sia. Serum IL-6 level was measured using the enzyme-linked immunosorbent assay (ELISA). Carotid artery calcification was determined by measuring Carotid Intima-Media Thickness (CIMT) using Corresponding author: Real-Time B-mode ultrasound. Riri Andri Muzasti Results There were 53 males (55.8%) of the total samples with Division of Nephrology and Hypertension. the mean duration of haemodialysis of 81.28±67.40 months. Ul- Department of Internal Medicine, trasound examination showed that 28 samples (29.5%) had carotid Faculty of Medicine, artery calcification. Statistical test significantly showed that patients with IL-6 ≥81.1 pg/mL were more likely to have carotid Universitas Sumatera Utara artery calcification with an increased risk of 12.92 times (95% CI: Dr. Mansyur 5 Medan, Indonesia 5.54-30.12) compared to the group of patients who had IL-6 level Phone: +62 81 260 556 872; <81.1 pg/mL (p <0.001). Fax: +62 8161 821 6264; Conclusion This study proves that a high level of IL-6 can increa- E-mail: [email protected] se the risk of carotid artery calcification in CKD-5D patients. ORCID ID: https://orcid.org/ 0000-0001- 7834-0740 Key words: chronic kidney disease, interleukin-6, carotid intima- media thickness

Original submission: 20 March 2020; Revised submission: 04 June 2020; Accepted: 27 June 2020 doi: 10.17392/1172-20

Med Glas (Zenica) 2020; 17(2): 346-351

346 Muzasti et al. IL-6 and vascular calcification

INTRODUCTION was performed by measuring serum level based on the ELISA technique. Ultrasonography (USG) Cardiovascular disease is a primary cause of mor- examination was performed by a radiologist who bidity and mortality in patients with chronic kid- did not know the patient’s clinical condition. The ney diseases (CKD), and it accounts for approxi- presence of carotid artery calcification was defi- mately 39% of deaths among those on dialysis (1). ned if the thickness or distance between the inti- One of the causes is vascular calcification that can ma-media tunica as measured by real-time B-mo- occur both in tunica intima and media, which later de ultrasound was >1 mm. Levels of calcium and contributes to the occurrence of myocardial isch- phosphate were measured using the multiplication emia, arrhythmia, and stroke (2,3). This vascular of serum level of phosphate and calcium (Ca x P). calcification has become a strong independent predictor of mortality in the general population (4) Statistical analysis and dialysis population (5-7). The characteristic of regular haemodialysis pa- Chronic kidney disease is associated with a chro- tient data was expressed as a percentage for cat- nic inflammatory state characterized by increased egorical data. Numerical data were expressed as levels of proinflammatory cytokines (8). As one mean ± standard deviation (SD) if normally dis- of the proinflammatory cytokines, IL-6 is known tributed and expressed as median (min-max) if the as a central regulator of the inflammatory process data distribution were not normally distributed. and plays a crucial role in the induction of immu- The normality test was done using Kolmogorov ne effectors and acute phase responses. IL-6 is Smirnov. The χ2 or Fisher's Exact test was used also strongly associated with morbidity and a to compare proportions between the two groups strong predictor of cardiovascular mortality in with variable data categories. The logistic regres- patients undergoing haemodialysis (9,10). sion test was used to explain the association be- The aim of this study is to investigate the rela- tween risk factors and carotid calcification. The tionship between IL-6 concentration and carotid p <0.05 was considered significant. To determine artery calcification in CKD-5D patients. the cut-off value of the IL-6 level as a predictor of the calcification carotid artery, the receiver PATIENTS AND METHODS operating characteristics (ROC) curve was used Patients and study design to obtain the area under the curve (AUC) value.

This study was an analytic observational with RESULTS a cross-sectional design. Ninety-five clinically stable patients who underwent regular haemodi- There were 53 (55.8%) males of the total sam- alysis for at least three months at Rasyida Renal ples, the mean duration of haemodialysis (HD) Hospital (Medan, Indonesia) and were willing to was 81.28±67.40 months. The median level of conduct laboratory and ultrasonography exami- calcium, phosphate and calcium phosphate was 2 2 nation, which was proven by signing an informed 9.80 mg/dL, 5.50 mg/dL and 53.76 mg / dL . consent, were included in this study. The patients Through the ELISA method, the mean IL-6 level with incomplete medical record data were exclu- was 97.95±117.93 pg/mL. ded from the study. The results of the USG examination showed that The investigation was approved by the Health most patients did not experience carotid artery Research Ethical Committee of the Medical calcification, 67 (70.5%) (Table 1). School of Universitas Sumatera Utara/H. Adam Total area under the curve was 95.9% with Malik General Hospital p<0.001 (95% CI: 91.2% - 100.0%), meaning that the IL-6 level had an excellent diagnostic va- Methods lue (> 90%) (Figure 1). The patients were interviewed to determine hi- Based on the analysis with the ROC curve, the story of previous illness and how long they had cut-off value of IL-6 level can also be determined, been undergoing haemodialysis. which has the highest combination of sensitivity Standard laboratory methods examined serum and specificity. The curve showed that the IL-6 le- phosphate and calcium level. The IL-6 analysis vel with a value of 81.1 had the highest combinati-

347 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 1. Characteristics of the patients with chronic kidney Table 2. Diagnostic values of IL-6 disease and dialysis (CKD-5D) Sensi- Speci- Cut-off PLR NLR PPV NPV Accuracy No (%) of tivity ficity Variable Mean ± SD Median (min-max) patients IL-6 96.4% 92.5% 12.9 0.04 84.4% 98.4% 98.6% Gender (≥81.1 pg/mL) Male 53 (55.8) level for carotid artery calcification PLR, positive likelihood Female 42 (44.2) ratio (LR+); NLR, negative likelihood ratio (LR-); PPV, positive Duration of HD (months) 81.28 ± 67.40 62.00 (43-676) predictive value; NPV, negative predictive value; <60 38 (40.0) ≥60 57 (60.00) Frequency of HD (hours/week) 10 72 (75.8) 12 23 (24.2) Calcium (mg/dL) 9.72 ± 0.74 9.80 (8.0-10.9) ≤9.5 38 (40.0) >9.5 57 (60.0) Phosphate (mg/dL) 5.43 ± 0.64 5.50 (4.0-6.8) ≤5.5 50 (52.6) >5.5 45 (47.4) Ca x P (mg2/dL2) 52.98 ± 9.08 53.76 (33.20-69.36) <55 53 (55.8) ≥55 42 (44.2) Figure 2. The cutting point value of IL-6 level as a predictor of IL-6 (pg/mL) 97.95 ± 117.93 70.10 (25.4-898.0) carotid artery calcification based on the ROC curve Calcification of carotid artery Absence 67 (70.5) Table 3. Characteristics of HD patients based on carotid Presence 28 (29.5) artery calcification Carotid artery calcification Variable PR 95% CI p Presence Absence (n=28) (n=67) IL-6 (pg/mL) 12.92 5.54-30.12 <0.001 < 81.1 1 (1.6) 62 (98.4) ≥ 81.1 27 (84.4) 5 (15.6) Calcium (mg/dL) 1.10 0.66-1.86 0.713 ≤9.5 12 (31.6) 26 (68.4) >9.5 16 (28.1) 41 (71.9) Phosphate (mg/dL) 1.03 0.68-1.55 0.906 15 (30.0) 35 (70.0) ≤5.5 >5.5 13 (28.9) 32 (71.1) Duration of HD (months) 1.02 0.71-1.45 0.927 <60 11 (28.9) 27 (71.7) ≥60 17 (29.8) 40 (70.2) PR, Prevalence Ratio, CI, Confidence Interval

DISCUSSION Figure 1. ROC curve of IL-6 level as a predictor of calcification There are many risk factors that can cause kidney carotid artery function to decline, such as genetic components, on of sensitivity (96.4%) and specificity (92.5%). sex, age, diabetes, and hypertension (11). Sex has The higher IL-6 level above the cut-off value, the traditionally been seen as an essential factor influ- higher the risk of calcification. The cut-off point encing the development of kidney disease (12). value of IL-6 ≥81.1 had an excellent diagnostic Regardless of its etiology, females tend to progress value and accuracy (Table 2, Figure 2). more slowly to end-stage kidney disease (13). Our The patients who had IL-6 level ≥81.1 pg/mL study shows that the prevalence of patients under- were experiencing more vascular calcification going haemodialysis was commonly found in ma- compared to those with IL-6 level ≤81.1pg/mL les (55.8%). These results are in line with the study (p<0.001) (Table 3). The risk of carotid artery of Lumtergul et al. in CKD patients. They found calcification increased 12.92 times (95% CI: that most of the dialysis and non-dialysis patients in 5.54-30.12) if the patient had IL-6 level ≥81.1 Thailand were male with the percentage of 55.4% pg/mL compared to the group of patients who and 53.5% respectively (14). The same results were had IL-6 level <81.1 pg/mL (p <0.001). obtained by Nakayama et al. (15) and Ok et al. (16).

348 Muzasti et al. IL-6 and vascular calcification

Carotid artery calcification has been recognized routine haemodialysis (29). The exact cause is un- as a risk factor for cardiovascular events in pati- known. However, it is suspected that every time ents with end-stage renal disease (15). Traditional a patient undergoes haemodialysis, the formation risk factors alone, such as hypertension and dysli- of reactive oxygen species (ROS) occurs, which pidaemia, do not sufficiently contribute to the plays an essential role in endothelial dysfunction high calcification burden in the dialysis patient and atherogenesis, which is modulated by IL-6 (17,18). Hyperphosphatemia and increased level (31). A study conducted by Beberashvili et al. of calcium x phosphate products, which is caused showed chronic inflammation characterized by by reduced renal phosphate excretion, were also elevated levels of IL-6 correlated with all-cause identified as an essential cause of accelerated ar- mortality in stable chronic haemodialysis patients terial calcification in this group of patients (4, 19- (32). A study conducted by Maddhumathi et al. 22). Our study and the study conducted by Barre- showed that a median plasma IL-6 level in 206 to et al. (23) found the mean phosphate levels and patients undergoing haemodialysis was 7.9 pg/mL calcium-phosphate multiplication still within the (ranging from 0.1 to 90.2 pg/mL), and was found target range for dialysis patients determined by to be higher in patients with vascular disease (9). the Kidney Disease Outcomes Quality Initiative In this study, from the 95 patients, we found that (KDOQI), which are 8.4-9.5 mg/L for calcium, the proportion of patients with IL-6 ≥81.1 pg/mL 3.5-5.5 mg/dL for phosphate and <55 mg2/dL2 for was more likely to have carotid artery calcification calcium-phosphate multiplication (24). Its target with an increased risk of 12.92 times compared to range was achieved probably because almost all the group of patients who had IL-6 levels <81.1 patients have taken the phosphate binders drugs. pg/mL. Similar results were also obtained by Kato A study conducted by Barreto et al. took 2.97 et al. (25) and Krasniak et al. (33). pg/mL as the cut-off IL-6 (23), while the study In this study, we found that the mean of the dura- conducted by Kato et al. dividing the IL-6 cut- tion of haemodialysis was 81.28 ± 67.40 months off into three groups consisting of ≤1.1 pg/mL, and showed no significant relationship between 1-1-2pg/mL and >2 pg/mL (25). The IL-6 cut- the duration of haemodialysis and the occurrence off by Honda et al. was 8.1 pg/mL (sensitivity of carotid artery calcification. It may be because 63.3%, specificity 73.3%) (26). In this study, the the process of vascular calcification can occur at IL-6 cut-off obtained was 81.1 pg/mL (sensitivity any time and started at a younger age (34). Even 96.4%, specificity 92.5%). This cut-off differen- the study conducted by Nitta et al. with the mean ce can be due to the characteristics of the sample duration of haemodialysis about 7.7 ± 5.8 years and the cut-off method used. also showed no significant results between the The prevalence of vascular calcification varies duration of haemodialysis and the occurrence of greatly from 60-100% depending on the location vascular calcification (35). of the examination and the diagnostic method used In conclusion, this study proves that high level of and, in the area where the study was conducted IL-6 can increase the risk of carotid artery calci- (27). London et al. reported that 68% of 202 pati- fication in CKD-5D patients. ents undergoing haemodialysis in France had arterial calcifications determined by radiography and ACKNOWLEDGEMENTS echocardiography (28). A study conducted by Na- Universitas Sumatera Utara financially supported kayama et al. in 135 patients undergoing routine this work under Talenta research implementation haemodialysis at hospitals in Japan, the prevalence contracts 2019. of carotid artery calcification was 71% (15). This FUNDING study found that the prevalence of carotid artery This work was supported by Universitas Suma- calcification was 29.5%. This difference in preva- tera Utara under Talenta research implementa- lence can be due to sample characteristics such as tion contracts 2019, number: 4167/UN5.1.R/ genetic, demographic, lifestyle differences, as well PPM/2019 as the distribution and number of samples. Systemic inflammation is commonly found in TRANSPARENCY DECLARATION patients with chronic kidney disease undergoing Conflict of interest: None to declare.

349 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES 15. Masaru N, Ura Y, Nagata M, Okada Y, Sumida Y, Nishida K, Ikeda H, Kaizu Y. Carotid artery cal- 1. Bethesda, MD. The United States Renal Data cification at the initiation of hemodialysis is a risk System. 2018 USRDS annual data report: Epide- factor for cardiovascular events in patients with end- miology of kidney disease in the United States. stage renal disease: a cohort study. BMC Nephrol Bethesda, MD: National Institutes of Health: Nati- 2011:12:56. onal Institute of Diabetes and Digestive and Kidney 16. Ok E, Asci G, Toz H, Ok ES, Kircelli F, Yilmaz M, Diseases, 2018. Hur E, Demirci MS, Demirci C, Duman S, Basci A, 2. Benz K, Hilgers KF, Daniel C, Amann K. Vascular Adam SM, Isik IO, Zengin M, Suleymanlar G, Yil- calcification in chronic kidney disease: the role of maz ME, Ozkahya M, Turkish Online Haemodiafil- inflammation. Intl J Nephrol 2018; 2018:4310379. tration Study. Mortality and cardiovascular events in 3. Floege, Jürgen; Ketteler, Markus. Vascular calcifi- online haemodiafiltration (OL-HDF) compared with cation in patients with end-stage renal disease. Nep- high-flux dialysis: results from the Turkish OL-HDF hrol Dial Transplant 2004; 19(Suppl 5):V59-66. Study. Nephrol Dial Transplant 2013; 28:192-202. 4. Wilson PW, Kauppila LI, O’Donnell CJ, Kiel DP, 17. Longenecker JC, Coresh J, Powe NR, Levey AS, Hannan M, Polak JM, Cupples LA. Abdominal Fink NE, Martin A, Klag MJ. Traditional cardiovas- aortic calcific deposits are an important predictor cular disease risk factors in dialysis patients compa- of vascular morbidity and mortality. Circulation red with the general population: the CHOICE Study. 2001;103: 1529–34. J Am Soc Nephrol 2002; 13:1918–27. 5. Blacher J, Guerin AP, Pannier B, Marchais SJ, Lon- 18. Sarnak MJ, Coronado BE, Greene T, Wang SR, Ku- don GM. Arterial calcifications, arterial stiffness, sek JW, Beck GJ, Levey AS. Cardiovascular disease and cardiovascular risk in end-stage renal disease. risk factors in chronic renal insufficiency. Clin Nep- Hypertension 2001; 38:938–42. hrol 2002; 57:327–35. 6. London GM, Guerin AP, Marchais SJ, Metivier F, 19. Goodman WG, Goldin J, Kuizon BD, Yoon C, Gales Pannier B, Adda H. Arterial media calcification in B, Sider D, Wang Y, Chung J, Emerick A, Greaser end-stage renal disease: impact on all-cause and L, Elashoff RM, Salusky IB. Coronary artery calci- cardiovascular mortality. Nephrol Dial Transplant fication in young adults with end-stage renal disease 2003; 18:1731–40. who are undergoing dialysis. N Engl J Med 2000; 7. Okuno S, Ishimura E, Kitatani K, Fujino Y, Kohno 342:1478–83. K, Maeno Y, Maekawa K, Yamakawa T, Imanishi Y, 20. Chertow GM, Raggi P, Chasan-Taber S, Bommer J, Inaba M, Nishizawa Y. The presence of abdominal Holzer H, Burke SK. Determinants of progressive aortic calcification is significantly associated with vascular calcification in hemodialysis patients. Nep- all-cause and cardiovascular mortality in maintenan- hrol Dial Transplant 2004; 19:1489–96. ce hemodialysis patients. Am J Kidney Dis 2007; 21. Raggi P, Boulay A, Chasan-Taber S, Amin N, Dillon 49:417–25. M, Burke SK, Chertow GM. Cardiac calcification in 8. Stenvinkel P, Alvestrand A. Inflammation in end- adult hemodialysis patients. A link between end-sta- stage renal disease: sources, consequences, and the- ge renal disease and cardiovascular disease? J Am rapy. Semin Dial 2002; 15:329-37. Coll Cardiol 2002; 20:39:695–701. 9. Madhumathi Rao et al. Plasma interleukin-6 predicts 22. Asmus HG, Braun J, Krause R, Brunkhorst R, Hol- cardiovascular mortality in hemodialysis patients. zer H, Schulz W, Neumayer HH, Raggi P, Bommer Am Kidney Dis 2005; 45.2:324-33. J. Two-year comparison of sevelamer and calcium 10. Pecoits-Filho R, Lindholm B, Axelsson J, Stenvin- carbonate effects on cardiovascular calcification kel P. Update on Interleukin -6 and its role in chro- and bone density. Nephrol Dial Transplant 2005; nic renal failure. Nephrol Dial Transplant 2003; 20:1653–61. 18:1042-5. 23. Barreto DV, Barreto FC, Liabeuf S, Temmar M, 11. Kazancioglu R. Risk factors for chronic kidney dise- Lemke H-D, Tribouilloy C, Choukroun G, Vanhol- ase: an update. Kidney Int Suppl 2013; 368-71. der R, A Massy ZA, European Uremic Toxin Work 12. Baylis, Chris. Sexual dimorphism in the aging kid- Group (EUTox). Plasma interleukin-6 is indepen- ney: differences in the nitric oxide system. Nat Re- dently associated with mortality in both hemodi- views Nephrol 2009; 5: 384-96. alysis and pre-dialysis patients with chronic kidney 13. Neugarten J, Acharya Anjali, Silbiger, Sharon R. disease. Kidney Int 2010; 77:550-6. Effect of gender on the progression of nondiabetic 24. NKF KDOQI Guidelines. KDOQI Clinical Practi- renal disease: a meta-analysis. J Am Soc Nephrol ce Guidelines for Bone Metabolism and Disease in 2000; 11:319-29. Chronic Kidney Disease. 2003 http://kidneyfounda- 14. Lumlertgul D, Kantachuvesiri S, Apichaiyingyurd tion.cachefly.net/professionals/KDOQI/guidelines_ S, Treamtrakanpon W, Rattanasompattikul M, Go- bone/guide6.htm (28 January 2020) jaseni P, Thanakitcharu P, Trakarnvanich T, Poon- 25. Akihiko K, Odamaki M, Takita T, Maruyama Y, vivatchaikarn U, Vareesangthip K, Impact-CKD Kumagai H, Hishida A. Association between inter- investigators. Prevalence of and predictive factor for leukin-6 and carotid atherosclerosis in hemodialysis abdominal aortic calcification in Thai chronic kidney patients. Kidney Int 2002; 61:1143-52. disease patients. Ther Apher Dial 2017; 21:611-9. 26. Hirokazu H, Qureshi AR, Heimbürger O, Barany P, Wang K, Pecoits-Filho R, Stenvinkel P, Lindholm B. Serum albumin, C-reactive protein, interleukin 6, and fetuin as predictors of malnutrition, cardiovascular disease, and mortality in patients with ESRD. Am J Kidney Dis 2006, 47:139-48.

350 Muzasti et al. IL-6 and vascular calcification

27. Guillermo R-D, Griselda B, Graciela F, Adriana P, 32. Beberashvili I, Sinuani I, Azar A, Yasur H, Shapi- Fabian O, Miriam L, Martin O, Gustavo L, Mariana ro G, Feldman L, Averbukh Z, Weissgarten J. IL-6 V-D, Soledad C, Mariano F, Carlos M, Marcelo O, levels, nutritional status, and mortality in prevalent Ricardo H, Carlos D, Rcardo G-M, Oscar V, Hernan hemodialysis patients. Clin J Am Soc Nephrol 2011; T M. Prevalence of factor related to vascular calci- 6:2253-63. fication in patients with chronic kidney disease on 33. Krasniak A, Drozdz M, Pasowicz M, Chmiel G, dialysis. MedICINA (BAires) 2017; 77:3. Michalek M, Szumilak D, Podolec P, Klimeczek P, 28. London GM, Guerin AP, Marchais SJ, Metivier F, Konieczynska M, Muniak EW, Tracz W, Khoa TN, Pannier B, Adda H. Arterial media calcification in Souberbielle JC, Drueke TB, Sulowicz W. Factors end-stage renal disease: impact on all-cause and involved in vascular calcification and atherosclerosis cardiovascular mortality. Nephrol Dial Transplant in maintenance hemodialysis patients. Nephrol Dial 2003, 18.9:1731- 40. Transplant 2007; 22:515-21. 29. Lau WL, Kalantar-Zadeh K, Vaziri NDJN. The gut 34. Jayalath RW, Mangan SH, Golledge J. Aortic calci- as a source of inflammation in chronic kidney disea- fication. Eur J Vasc Endovasc Surg 2005, 30:476-88. se. Nephron 2015; 130:92-8. 35. Nitta K, Akiba T, Uchida K, Kawashima A, Yumura 30. Spittle MA, Hoenich NA, Handelman GJ, Adhikar- W, Kabaya T, Nihei H. The progression of vascular la R, Homel P, Levin NW. Oxidative stress and in- calcification and serum osteoprotegerin levels in pa- flammation in hemodialysis patients. Am kidney dis tients on long-term hemodialysis. Am J Kidney Dis 2001; 38:1408-13. 2003; 42:303-9. 31. Yudkin JS, Kumari M, Mohamed-Ali V. Inflamma- tion, obesity, stress, and coronary heart disease: is interleukin-6 the link. Atherosclerosis 2000; 148.2:209-14.

351 ORIGINAL ARTICLE

Pectoralis (PecS) nerve block 1 for port-a-cath removal and central venous catheter (CVC) replacement

Massimo Renzini1, Umberto Ripani2, Luisa Golia3, Fulvio Nisi3, Fabio Gori1

1Department of Pain Medicine Hospital S.M.M, Perugia, 2Division of Clinic Anaesthesia, Department of Emergency Hospital Riuniti, Ancona, 3Department of Anaesthesia, Intensive Care and Pain Medicine, Hospital S.M.M., Perugia; Italy

ABSTRACT

Aim The use of PecS block 1 as perioperative analgesia for a central catheter removal -reimplantation combined procedure.

Methods A 55-year-old woman suffering from peritoneal metastases from gastric cancer needed to have a port-a-cath implanted for infection removed and to have a central venous catheter (CVC) implanted in the homolateral axillary vein due to patient’s history of deep vein thrombosis of the right upper limb. We used PECS 1 block for perioperative analgesia.

Results Compared to the traditional catheter implantation tech- Corresponding author: nique, reduction in the doses of local anaesthetics, shortening in the execution time, less intra-procedural bleeding, better patient’s Umberto Ripani compliance, and no need for a rescue dose of local anaesthetic Division of Clinic Anaesthesia, were observed. Department of Emergency Hospital Riuniti Conca Street 71, Conclusion The PEC1 block was effectively and safely used to remove an infected port-a-cath and to place a CVC on the same 60126 Ancona AN, Italy side. We hypothesize that it may be useful also for simple port-a- Phone: +39 071 596 3809; cath positioning. Fax: +39 071 596 3871; E-mail: [email protected] Key words: complication, local anaesthesia, outcomes, pectoralis Massimo Renzini ORCID: https://orcid. block, port-a-cath org/0000-0001-7386-1520

Original submission: 06 March 2020; Revised submission: 10 March 2020; Accepted: 24 March 2020 doi: 10.17392/1158-20

Med Glas (Zenica) 2020; 17(2): 352-355

352 Renzini et al. PecS block 1 for port-a-cath

INTRODUCTION teroides vulgatus. In addition to this, the lack of peripheral venous access and the need to continue Pectoran nerve (PecS) block 1 is a simple anal- antibiotic therapy and parenteral nutrition required gesic block performed by means of a local ana- the placement of a new central venous catheter esthetic administration in the layer that separates (CVC) after the removal of the port-a-cath. two pectoral muscles, the pectoralis major and the pectoralis minor, at the level of the third rib (1). Methods The interfascial administration of 20 mL of local anaesthetic is suitable to block the lateral pecto- According to the RaCeVA protocol (7), we per- ral nerve and most of the branches of the medial formed systematic ultrasound examination of the pectoral nerve (2). The block is performed with a venous system and excluded implanting the new linear probe positioned 3-4 cm distal to the coraco- CVC in the right axillary vein due to the patient’s id process with a transverse view of the pectoralis history of deep vein thrombosis of the right upper minor muscle (3). It has been successfully used for limb and because the vessel did not show any pa- perioperative analgesia in adult breast surgery (1- tency while imaging (lack of collapse if compre- 4) and minimally invasive cardiac surgery (3). ssed). Thus, we decided to remove the left axillary port-a-cath and to position the new central The traditional approach to the local anaesthesia catheter in the homolateral axillary vein. With for port-a-cath implantation requires the admi- an ultrasound-guided and sterile technique, PecS nistration of 25-35 mL of local anaesthetic (4). block 1 was performed (Figure 1) with mepiva- It also needs two or three site punctures for best caine 200 mg and levobupivacaine 50 mg. accomplishment, this being especially risky in patients on anticoagulant therapy (1-4). Moreo- ver, it does not adequately reduce pain during the creation of the pocket for the porth-a-cath cham- ber placement (1-4). We hypothesize that PecS block-1 may reduce both the dose of the local anaesthetic and the relative risk of toxicity, the number of site punctures required and, of course, ongoing pain during the creation of the pocket; this approach may work not only in the plant, but also in the removal of the port-a-cath. The aim of this study was to describe this approach in one of our patients (5,6), who needed to have the port-a-cath for infection removed, and a new central venous catheter implanted.

PATIENT AND METHODS Figure 1. The PecS block 1 performed with an ultrasound- Patient and study design guided technique. The needle tip reaches and injects the local anaesthetic mixture in the layer between the pectoralis nd We presented a case of a 55-year-old woman suffe- major and pectoralis minor muscles (left); the 2 rib acts as a protecting fence against unintentional pleural puncture (Ri- ring from peritoneal metastases from gastric can- pani U, 2018) PM, pectoralis major muscle; Pm, pectoralis minor cer previously treated with radiotherapy and surgi- muscle; LA, local anesthetic mixture injected cal removal. She had no other relevant diseases in her past medical history. She had no coagulation RESULTS deficiency but she was on anticoagulant treatment with low molecular weight heparin for prevention The patient presented mild pain at the incision of of venous thromboembolism and chemotherapy the pocket (numeric rating scale, NRS: 4) (8), whi- for neoplasm treatment. Our attention was drawn le all other phases of the procedure (port-a-cath re- to the septic state due to infection of the implan- moval, pocket closure, ultrasound-guided puncture ted port-a-cath (in the left axillary vein) by Bac- of the axillary vein, CVC implantation) were pain

353 Medicinski Glasnik, Volume 17, Number 2, August 2020

free (NRS: 0). There were no immediate compli- Recently, a clinical experience on children un- cations, such as bleeding. The procedure was jud- dergoing implantation and removal of port-a-cath ged easy to perform by the operator and it requi- has been published by the Stanford hospital (2). red about 30 minutes for completion (Figure 2). It Authors have retrospectively analysed the impact was observed that the dose and volume of the used of PecS 1 block versus the traditional technique anaesthetic was reduced comparing to the classic on the need for rescue analgesia, long acting opi- method. Within one month after the procedure the oids use, postoperative pain scores and post-ana- patient did not show any late complication. esthesia care unit length of stay; all parameters analysed were in favour of the PecS 1 group but without showing any statistical significance, perhaps due to small sample. In our experience we routinely performed the same comparison on adult patients and we observed a reduction in the doses of local anaesthetics, a shortening in the execution time, less intra-procedural bleeding, better patient’s compliance, and no need for rescue dose of local anaesthetic. Thus, we hypothesize that PecS 1 approach may be used successfully to implant a port-a-cat, re- placing the traditional approach. The PECS block has already been used succe- Figure 2. The result at the end of the procedure. Patient’s left shoulder is showed. The new implanted catheter is clearly vis- ssfully in breast surgery, pacemaker placement ible ahead, while the entrance point of the old one has been and upper limb surgery (1,3,9-26). We have sutured caudally (Ripani U, 2018) effectively and safely used the PecS 1 block to DISCUSSION remove an infected port-a-cath and to place a CVC on the same side with a reduction of pain In this case we performed an ultrasound-guided and doses of local anaesthetic used and without PecS block 1 with 20 mL of local anaesthe- any early or late complication. tic (mepivacaine 1% 10 ml and levobupivacai- We hypothesize that it may be useful also for port- ne 0.5% 10 ml) to provide analgesia for a cen- a-cath positioning in terms of reduction of local tral catheter removal-reimplantation combined anaesthetic doses and relative risk of toxicity, re- procedure. A skin incision was made in the left duction of the risk of periprocedural complications port-a-cath site. The catheter was removed and (such as bleeding). However, as there are currently its pocket was closed. A two-way CVC was then no randomized trials in literature confirming these placed in the same left axillary vein with an US- hypotheses and clinical observations, large sample guided technique. There are no similar procedu- clinical trials are needed yet. res described in the literature. FUNNDING Moreover, we did not find any published study comparing port-a-cath placement and removal No specific funding was received for this study. under the local anaesthesia with the classic tech- TRANSPARENCY DECLARATION nique and PecS -1 block. Conflict of interest: None to declare.

REFERENCES 1. Kulhari S, Bharti N, Bala I, Arora S, Singh G. Effi- 3. Yalamuri S, Klinger R, Bullock W, Glower DD, cacy of pectoral nerve block versus thoracic para- Bottiger BA, Gadsden JC . Pectoral fascial (PECS) vertebral block for postoperative analgesia after ra- I and II blocks as rescue analgesia in a patient un- dical mastectomy: a randomized controlled trial. Br dergoing minimally invasive cardiac surgery. Reg J Anaesth 2016; 117:382–6. Anesth Pain Med 2017; 42:764–6. 2. Farrukh Munshey F, Ramamurthi RJ, Tsui B. Early 4. Blanco R. The ‘pecs block’: a novel technique for experience with PECS 1 block for Port-a-Cath inser- providing analgesia after breast surgery. Anaesthesia tion or removal in children at a single institution. J 2011; 66:847–8. Clin Aesth 2018; 49: 63–4.

354 Renzini et al. PecS block 1 for port-a-cath

5. Falzarano G, Piscopo A, Grubor P, Rollo G, Medi- type III Scaphoid Nonunion Advanced Collapse ci A, Pipola V, Bisaccia M, Caraffa A, Barron EM, (SNAC) treatment with partial carpal arthrodesis in Nobile F, Cioffi R, Meccariello L. Use of common the dominant hand: results of 5-year follow-up. Med inflammatory markers in the long-term screening of Arch 2018; 72:253-6. total hip arthroprosthesis infections: our experience. 16. Rollo G, Rotini R, Eygendaal D, Pichierri P, Bi- Adv Orthop 2017; 2017:9679470. saccia M, Prkic A, Stasi A, Meccariello L. Effect 6. Franzese, R, Conte M, Gagliardo N, Pieretti, G. of trochleocapitellar index on adult patient-reported Children vs elderly in orthopedic surgery site of in- outcomes after noncomminuted intra-articular dis- fection. Are there difference? Experience of a high tal humeral fractures. J Shoulder Elbow Surg 2018; volume plastic surgeon consultant. Acta Med Sal 27:1326-32. 2109; 49:24–9. 17. Rollo G, Pichierri P, Marsilio A, Filipponi M, Bi- 7. Spencer TR, Pittiruti M. Rapid Central Vein Asse- saccia M, Meccariello L. The challenge of nonunion ssment (RaCeVA): a systematic, standardized after osteosynthesis of the clavicle: is it a biomecha- approach for ultrasound assessment before central ve- nical or infection problem? Clin Cases Miner Bone nous catheterization. J Vasc Access 2019; 20:239-49. Metab 2017; 14:372-8. 8. Hawker GA, Mian S, Kendzerska T, French M. Me- 18. Rollo G, Rotini R, Pichierri P, Giaracuni M, Stasi A, asures of adult pain: Visual Analog Scale for Pain Macchiarola L, Bisaccia M, Meccariello L. Grafting (VAS Pain), Numeric Rating Scale for Pain (NRS and fixation of proximal humeral aseptic nonunion: Pain), McGill Pain Questionnaire (MPQ), Short- a prospective case series. Clin Cases Miner Bone Form McGill Pain Questionnaire (SF-MPQ), Chro- Metab 2017; 14:298-304. nic Pain Grade Scale (CPGS), Short Form-36 Bodily 19. Medici A, Meccariello L, Rollo G, De Nigris G, Pain Scale (SF-36 BPS), and Measure of Intermittent Mccabe SJ, Grubor P, Falzarano G. Does routine and Constant Osteoarthritis Pain (ICOAP). Arthritis carpal tunnel release during fixation of distal radi- Care Res (Hoboken). 2011; 63 Suppl 11:S240-52. us fractures improve outcomes? Injury. 2017 Oct; 9. Rollo G, Prkic A, Bisaccia M, Eygendaal D, Pichierri 48(Suppl 3):S30-3. P, Marsilio A, Giaracuni M, Meccariello L. Grafting 20. Bisaccia M, Meccariello L, Rinonapoli G, Rollo and fixation after aseptic non-union of the humeral G, Pellegrino M, Schiavone A, Vicente CI, Ferra- shaft: a case series. J Clin Orthop Trauma 2020; ra P, Filipponi M, Caraffa A. Comparison of pla- 11(Suppl 1):S51-5. te, nail and external fixation in the management 10. Gómez-Garrido D, Triviño-Mayoral V, Delgado-Al- of diaphyseal fractures of the humerus. Med Arch cala V, Cervera-Irimia J, Medina-Lorca M, Sánchez- 2017; 71:97-102. Sánchez F, Ibáñez-Vicente C, Pérez-Gurbindo I, 21. Manni M, Bisaccia M, Rinonapoli G, Schiavone A, Meccariello L, Rollo G, Pica G, Tomarchio A, Meccariello L, Mccabe SJ, Bisaccia O, Vicente CI, Pasquino A, Bisaccia M. Five year long term results Cappiello A, Caraffa A. Reliability, feasibility and of total joint arthroplasties in the treatment of tra- value of ecography in clinical-functional results in peziometacarpal osteoarthritis. Acta Biomed 2019; patients affected by carpal tunnel syndrome: is there 90:451-6. a correlation? Acta Inform Med 2017; 25:44-8. 11. Rollo G, Prkić A, Pichierri P, Eygendaal D, Bisaccia 22. Bisaccia M, Rinonapoli G, Falzarano G, Medici A, M, Filipponi M, Giaracuni M, Hitov P, Tanovski K, Meccariello L, Rosati R, Pellegrino M, Ibáñez Vi- Meccariello L. Plate-and-bone-strut fixation of distal cente C, Piscitelli L, Caraffa A. Clinical and radiolo- third humeral shaft aseptic non-unions: a consecuti- gical outcomes of distal radius fractures treated with ve case series. J Clin Orthop Trauma 2019; 10(Suppl orif with volar plates. EMBJ 2016; 11:9-14. 1):S127-32. 23. Bisaccia M, Rinonapoli G, Bisaccia O, Meccariello 12. Rinonapoli G, Bisaccia M, Meccariello L, Mancini L, Ibáñez Vicente C, Ceccarini P, Colleluori G, An- GB, Marrani F, Grubor P, Rollo G, Caraffa A. Tran- drea Schiavone A, Caraffa A. Articular fractures of sient osteomyelitis of the distal radius in a three-ye- distal radius: comparison of treatment and clinical ar-old patient. Med Glas (Zenica) 2020; 17:178-82. and radiological outcomes with volar plate versus 13. Rollo G, Meccariello L, Rotini R, Pichierri P, Bi- hoffmann bridging external fixator. EMBJ 2017; saccia M, Fortina M. Efficacy of the "Salento tech- 12:18–23. nique", a modified two-incision approach in distal 24. Rollo G, Bisaccia M, Franzese R, Pichierri P, Fi- biceps brachii tendon repair. Surgical description lipponi M, Giaracuni M, Gomez-Garrido D, Ripani and outcomes analysis. J Clin Orthop Trauma 2019; U, De Cruto E, Pieretti G, Meccariello L. The bio- 10:959-64. mechanical potential of the bone graft in the proxi- 14. Rollo G, Porcellini G, Rotini R, Bisaccia M, mal ulna non-union surgery. Clin Cases Miner Bone Pichierri P, Paladini P, Guerra E, De Cruto E, Fran- Metab 2019; 16:53-61. zese R, Grubor P, Pace V, Meccariello L. A new plate 25. Di Giacinto S, Meccariello L. Supracondylar fractu- design to treat displaced 3-4 parts proximal humeral res of the humerus, gartland 3: pediatric urgency. fractures in comparison to the most tested and used Case report and literature review. EMBJ 2016; plate: clinical and radiographic study. Med Glas (Ze- 11(S1):6. nica) 2019; 16:284-92. 26. Falzarano G, Medici A, Meccariello L. Challenges 15. Rollo G, Bisaccia M, Irimia JC, Rinonapoli G, of the sub-amputated humerus: to amputate or not? Pasquino A, Tomarchio A, Roca L, Pace V, Pichierri A case study. EMBJ 2016; 11(S1):4-5. P, Giaracuni M, Meccariello L. The advantages of

355 ORIGINAL ARTICLE

Treatment effectiveness and outcome in patients with a relapse and newly diagnosed multidrug-resistant pulmonary tuberculosis Dmytro Butov1, Valeriy Myasoedov1, Mykola Gumeniuk2, Galyna Gumeniuk2, Oleksandra Choporova1, Anton Tkachenko3, Oleksandra Akymenko4, Olena Borysova1, Olena Goptsii5, Yevhenii Vorobiov6, Tetiana Butova6

1Department of Phthisiology and Pulmonology, Kharkiv National Medical University, Kharkiv; 2 Department of Technologies of Treat- ment of Nonspecific Lung Diseases, National Institute of Phthisiology and Pulmonology named after FG Yanovskyi NAMS of Ukraine, Kyiv; 3Department of Biochemistry, Kharkiv National Medical University, Kharkiv; 4Medical Division #3, Regional Anti-TB Dispensary #1, Kharkiv; 5Department of Internal Medicine #1, Kharkiv National Medical University, Kharkiv; 6Department of Internal Medicine, VN Karazin Kharkiv National University, Kharkiv; Ukraine

ABSTRACT

Aim To investigate the treatment effectiveness and outcome in patients with pulmonary tuberculosis relapse and newly diagnosed multidrug resistant pulmonary tuberculosis (MDR-TB).

Methods A total of 240 pulmonary MDR-TB patients, including 114 ones with tuberculosis relapse and 126 cases of newly diagnosed pulmonary tuberculosis, were examined. Effectiveness of the basic antimycobacterial therapy course was evaluated based on the time to normalization of tuberculosis clinical manifestation, sputum culture and acid-fast bacilli stain conversion, cavity closure, Corresponding author: disappearance of infiltrative and focal changes in the pulmonary tissue. Treatment outcomes were evaluated as cured, treatment Dmytro Butov completed, treatment failed, died and lost to follow-up according Departments of Phthisiology and to the World Health Organization guidelines. Pulmonology, Kharkiv National Medical University Results When assessing the treatment effectiveness in patients Nauky Avenue 4, 61022, Kharkiv, Ukraine with MDR-TB, a worse clinical and chest radiograph dynamics was observed in tuberculosis relapse against the background of Phone:+38095 015 90 58; high parameters of treatment failure (18.4 %) and low cured (34.2 E-mail: [email protected] %) compared with newly diagnosed pulmonary tuberculosis (7.1% ORCID ID: https://orcid.org/0000-0002- and 58.7 %, respectively) (p=0.008 and p<0.001, respectively). 8792-901X Conclusion Standard treatment effectiveness in patients with newly diagnosed MDR-TB manifested by faster improvement and stabilization of health, earlier sputum culture and smear conversion, higher frequency of cavity closure and achievement of certain clinical and radiographic improvement against the background of Original submission: fewer cases of treatment failure and a higher number of cured pa- 10 April 2020; tients compared with MDR-TB relapse. Revised submission: Key words: Mycobacterium, X-ray, culture, microscopy 27 April 2020; Accepted: 11 May 2020 doi: 10.17392/1179-20

Med Glas (Zenica) 2020; 17(2): 356-362

356 Butov et al. Treatment in patients with MDR tuberculosis

INTRODUCTION Thus, the aim of this research was to investigate the effectiveness and outcomes of treatment in Tuberculosis is a leading cause of poor health patients with tuberculosis relapse and newly di- and one of the most common causes of death agnosed pulmonary MDR-TB. worldwide. Furthermore, tuberculosis accounts for numerous deaths from HIV/AIDS (1). The PATIENTS AND METHODS emergence of tuberculosis forms that are resistant to anti-tuberculosis drugs is a huge chall- Patients and study design enge to the End TB strategy of the World Health Organization (WHO), which aims at reducing A retrospective observational cohort study was tuberculosis incidence by 90% and decreasing performed on registry data obtained from the Re- tuberculosis mortality by 20% by 2035 compared gional Anti-tuberculosis Dispensary #1, Kharkiv, to 2015 (2). Thus, resistant forms of tuberculo- Ukraine. A total of 240 patients with pulmonary sis, in particular multidrug-resistant tuberculo- MDR-TB, including 114 cases of tuberculosis sis (MDR-TB), are a global threat to humanity. relapse (group 1) and 126 patients with newly di- MDR-TB is defined as a tuberculosis in which agnosed pulmonary tuberculosis (group 2) aged M. tuberculosis (MTB) is resistant to rifampicin between 20 years and 70 years, were enrolled. and isoniazid, which are the most effective dru- The period of inclusion lasted from 2012 to 2017. gs for tuberculosis treatment in accordance with The randomization allowed providing equal dis- the history of their administration in 127 countri- tribution of baseline characteristics, namely age, es (3,4). Moreover, chemoresistant tuberculosis gender, height, and body weight (Table 1). more frequently develops especially in patients To be eligible for participation in this study, pa- with tuberculosis relapse than among those with tients must have met the following inclusion cri- newly diagnosed pulmonary tuberculosis (5,6). teria: male or female sex, age of 18 to 60 years, It is worth noting that the issue of tuberculosis multi-drug resistant pulmonary tuberculosis, tu- recurrence remains of huge practical importance berculosis relapse, newly diagnosed pulmonary nowadays, since the incidence of tuberculosis tuberculosis, full completion of the inpatient pha- reactivation in patients who suffered from active se of care. Exclusion criteria included pregnancy, tuberculosis has remained high for decades (7). lactation, tuberculosis relapse two or more ti- Thus, high rates of resistant tuberculosis may be mes, incomplete in-patient phase of care, grou- observed due to an increase in the incidence of ps with the compromised immunological status cases with tuberculosis relapse (8). (HIV, diabetes mellitus, malnutrition, patients To succeed in reducing the prevalence of chemo receiving immunosuppressive therapy, including resistant tuberculosis, it is necessary to under- cytostatic, corticosteroid, radiation therapy and stand the complexity of tuberculosis treatment TNF-α, and other conditions that comply with and evaluation of its outcome (8). Proper under- WHO guidelines) (9). standing and resolution of the issue of tuberculo- Active pulmonary tuberculosis was defined in sis relapse could prevent the spread and increase accordance with the medical history and clinical in the amount of tuberculosis patients, especially findings, which were compatible with tuberculo- those with chemo resistant forms, by restricting sis, chest X-ray examination demonstrating lung the major type of tuberculosis transmission, na- involvement, and smear positivity for acid-fast mely the airborne spread of M. tuberculosis. bacilli, as well as positive cultures for MTB. We have decided to conduct this study, since the The study was approved by the Ethics and Bio- prevalence of relapse MDR-TB has been increa- ethics Committee of Kharkiv National Medical sing for years in the world. Furthermore, in our University, Kharkiv, Ukraine). clinical practice, clinical features of relapse tuberculosis form have drawn attention compared Methods to newly diagnosed pulmonary tuberculosis. Un- Upon hospitalization to the anti-tuberculosis derstanding of treatment effectiveness and outco- dispensary, isoniazid, rifampicin, pyrazinami- mes may provide novel insights into the preven- de, and ethambutol were prescribed to patients tion of relapses exactly for this cohort of patients. assuming that the causative agent was suscepti-

357 Medicinski Glasnik, Volume 17, Number 2, August 2020

ble to these drugs. Mutations in an rpoB MTB ducted monthly starting from the beginning of gene were determined using Xpert MTB/RIF the treatment. Isolates of MTB were tested for (10) within several days to confirm the diagno- susceptibility to the first- and second-line anti- sis of rifampicin-resistant tuberculosis (RR-TB). tuberculosis drugs using commercially available Then, the standard chemotherapy was administe- kits (Tulip Diagnostics Pvt Ltd., Goa, India). Si- red: pyrazinamide, prothionamide (ethionamide), milarly, liquid cultures (Mycobacterial Growth second-line injectable drugs, fluoroquinolone, Indicator Tube – MGIT-960, Becton Dickinson, cycloserine, and p-aminosalicylic acid. This tre- Franklin Lakes, NJ, USA) were used for diagno- atment lasted until the outcome of bacteriologi- stic purposes. M. tuberculosis resistance to the cal resistance tests was obtained. Treatment was first-line anti-tuberculosis drugs was determined prescribed based on the MTB susceptibly profile when reagents were available. on the basis of the outcome of MTB resistance Sputum culture was carried out for all patients. culture tests. Doses of anti-tuberculosis drugs Sputum smear microscopy was not performed in were selected with respect to the body weight. A four and five patients in the group 1 and group 2, treatment of MDR-TB included two phases: in- respectively, because of sputum missing. In these tensive (administration of drug injections during cases, bronchoalveolar lavage fluid (BALF) was at least 8 months) and continuation (administrati- performed instead. on of injectable forms was cancelled). The latter The bacterial load was evaluated by a Ziehl– lasted for at least 12 months. Neelsen method: smear-negative (acid-fast ba- Effectiveness of the basic course of anti-tubercu- cillus were not found in 100 fields), smear single losis chemotherapy was evaluated according to the positive (+) (10-99 acid-fast bacillus per 100 fi- following parameters: the time to normalization of elds), smear double positive (++) (1-9 acid-fast tuberculosis clinical manifestations, smear con- bacillus per fields), smear triple positive (+++) version, cavity closure, disappearance of infiltrati- (over 10 acid-fast bacillus in each fields) (8,12). ve and focal changes in the lung tissue, as well as The bacterial load was assessed at a Lowenstein– total clinical treatment effectiveness (8). Jensen medium also: solitary colonies of sputum/ Time to normalization of the most important tu- BALF MTB: single positive (1+) (20-100 colo- berculosis clinical manifestations and conven- nies), double positive (2+) (100-200 colonies), tional diagnostic methods under the influence triple positive (3+) (200-500 colonies - almost of standard chemotherapy was evaluated by the totally covered), quadruple positive (4+) (over disappearance of intoxication and chest-related 500 colonies - totally covered) (8,12). symptoms. Discontinuation of the intoxication Pathological features, i.e. the severity of the dise- syndrome was confirmed by the presence of ase and changes in the localization of the process subjective signs (increased appetite, disappea- in lungs, were evaluated by X-ray examination. rance of general weakness, fatigue, sweating) in patients, normalization of body temperature Monitoring was performed at the beginning of the and body weight, and changes in urinalysis and treatment every 4 months during the intensive pha- other signs, which were considered symptoms se, and every 6 months at the continuation phase. of tuberculosis. Disappearance or reduction Statistical analysis of shortness of breath, cough, chest pain, hae- moptysis, and pulmonary haemorrhages was Standard parametric Student’s t-test (13) was also assessed. chosen based on the outcomes of Shapiro-Wilk and Kolmogorov-Smirnov normality tests. The The treatment outcome was assessed according to difference was considered to be statistically si- the WHO guidelines: cured, treatment completed, gnificant at p<0.05. treatment failed, died, and lost to follow-up (11). Standard microbiological examinations of a RESULTS sputum smear stained in accordance with Zie- A total of 240 patients with pulmonary MDR- hl–Neelsen, and culture Lowenstein–Jensen TB, including 114 cases of tuberculosis relapse methods (12) were performed prior to the en- (group 1) and 126 patients with newly diagnosed rolment of patients. Furthermore, they were con-

358 Butov et al. Treatment in patients with MDR tuberculosis

pulmonary tuberculosis (group 2) were included. months of the treatment; 14.3%) and 24 (out of There was no statistically significant difference 40; 60.0%) patients with relapse (p=0.030) (Ta- among the groups relating to the age, gender, ble 2). Chest-related symptoms were consistent height and weight (Table 1). with changes in the intoxication intensity. Thus, the clinical effectiveness of standard chemothe- Table 1. Baseline characteristics of patients with pulmonary tuberculosis rapy in patients with tuberculosis relapse was Characteristic Group 1 (n = 114) Group 2 (n = 126) p significantly lower than in patients with newly Mean age (±) (years) 41.9±1.2 43.8±1.1 0.256 diagnosed pulmonary tuberculosis. Gender (No, %) Male 87 (76.3) 100 (79.4) 0.563 Post-treatment results of bacteriological exami- Female 27 (23.7) 26 (20.6) Height (cm) 171.6±0.8 172.2±0.8 0.620 nation Weight (kg) 63.1±1.1 62.8±1.1 0.891 Group 1, pulmonary tuberculosis relapse; Group 2, newly diagnosed An analysis of bacterial loads in the patients with pulmonary tuberculosis tuberculosis revealed higher MTB loads in patients with tuberculosis relapse that in the 2 gro- An analysis of 114 patients with tuberculosis re- up. The patients from the group 2 were mainly lapse revealed that among patients with newly characterized by either sputum acid-fast bacilli diagnosed pulmonary tuberculosis 37 (32.4%) smear negativity or “+” positivity, while in the were diagnosed until 2014, while 77 (67.6%) af- patients from group 1 higher bacterial loads were ter 2014. observed (“+++”). The same trend was revealed when culture methods were used. In particular, Post-treatment changes in cardinal symptoms in group 1 MTB triple and quadruple positivity Upon admission to the hospital intoxication syn- was mainly detected bacteriologically, while this drome was observed in 96 (out of 114; 84.2%) parameter was either “1+” or “2+” in the 2 group patients with TB relapse and in 83 (out of 126; (Table 3). 65.9%) patients in the group 2 (p=0.001). During the intensive phase of chemotherapy, we obser- Table 3. Sputum and bronchoalveolar lavage fluid (BALF) smear and sputum acid-fast bacilli /MTB positivity in two ved an earlier stoppage of intoxication syndro- groups of patients me in the patients of group 2 than in the patients No (%) of patients Method p with relapse of tuberculosis (Table 2). After the Group 1 (n = 114) Group 2 (n = 126) first month of the treatment, there were still 51 Microscopy: smear (61.4%) and 78 (81.2%) patients who had intoxi- Negative 6 (5.2) 21 (16.6) 0.005 Single positive (+) 27 (23.7) 51 (40.5) 0.006 cation syndrome in the patients with newly dia- Double positive (++) 28 (24.6) 32 (25.4) 0.886 gnosed pulmonary tuberculosis and in the group Triple positive (+++) 53 (46.5) 22 (17.5) <0.001 with relapse, respectively (p=0.003). Bacteriology: Sputum and BALF colony number Solitary 2 (1.7) 14 (11.2) 0.003 After four months of the treatment, intoxicati- Single positive (1+) 21 (18.4) 44 (34.9) 0.004 on was almost completely eliminated in 44 (out Double positive (2+) 36 (31.6) 47 (37.3) 0.354 Triple positive (3+) 27 (23.7) 11 (8.7) 0.001 of remaining 51 after the first month of the tre- Quadruple positive (4+) 28 (24.6) 10 (7.9) <0.001 atment; 86.3%) patients in the group with newly Group 1, pulmonary tuberculosis relapse; Group 2, newly diagnosed diagnosed pulmonary tuberculosis and 38 (out of pulmonary tuberculosis 78; 48.7%) patients with relapse (p<0.001). Af- ter eight months of chemotherapy, intoxication As for the outcomes of sputum smear microsco- ceased in one (out of remaining seven after four py, acid-fast bacilli stain conversion in group 1 occurred statistically significantly (p<0.001) later Table 2. Cessation of intoxication syndrome in two groups of patients than in patients with newly diagnosed pulmonary Cessation after a No (%) of patients tuberculosis. In particular, the mean time to smear p month Group 1 (n = 96) Group 2 (n = 83) conversion was 96.6±6.5 days in 82 patients with One (the first) 18 (18.7) 32 (38.6) 0.003 tuberculosis relapse (out of 114, 11 patients died Four 38 (39.6) 44 (53.0) 0.074 Eight 24 (25.0)* 1 (1.2)* <0.001 and 21 had the treatment failed) and 63.8±4.5 *The remaining patients experienced treatment failure or they died days in 111 patients from group 2 (out of 126, six Group 1, pulmonary tuberculosis relapse; Group 2, newly diagnosed patients died and nine had the treatment failed). pulmonary tuberculosis

359 Medicinski Glasnik, Volume 17, Number 2, August 2020

Almost the same dynamics was observed when conversely, when pulmonary tissue infiltration evaluating the outcomes of bacteriological tests. persisted for a while. Sputum culture conversion in the patients from group 2 was more likely than in patients from Treatment outcome group 1. Thus, mean time to sputum culture con- Despite statistically insignificant results of the version was 85.5±5.4 days in patients with tuber- comparison of death rates and other treatment culosis relapse (n=82) and 56.2±3.6 days in gro- outcomes, such as lost to follow-up and the up 2 (n=111), respectively (p<0.001). There was treatment completed, a statistically significant no statistically significant difference between the difference between the groups studied in terms outcomes of bacteriological culture and micros- of such outcomes as cured and the treatment copic tests. Furthermore, some patients experi- failed was observed. Therefore, tuberculosis re- enced sputum bacteriological culture conversion lapse was characterized by lower rates of cured prior to smear conversion. It can be assumed that patients (39; 34.2%) and higher rates of the pa- this is due to the presence of nonviable MTB du- tients with treatment failure (21; 18.4%) comparing microscopy. red with patients from group 2 (74; 58.7% and 9; 7.1%, respectively) (Table 5). Post-treatment chest X-ray examination Table 5. Multidrug resistant tuberculosis (MDR-TB) treatment It is important to note that the most important cri- outcomes teria characterizing the treatment effectiveness in No (%) of patients Treatment outcome p patients with pulmonary tuberculosis are cavity Group 1 (n = 114) Group 2 (n = 126) closure and disappearance of infiltrative changes Died 11 (9.6) 6 (4.8) 0.149 (Table 4). Treatment failed 21 (18.4) 9 (7.1) 0.008 Lost to follow-up 24 (21.1) 20 (15.9) 0.299 Treatment completed 19 (16.7) 17 (13.5) 0.489 Table 4. Time to cavity closure for two groups of patients Cured 39 (34.2) 74 (58.7) <0.001 No (%) of patients with Group 1, pulmonary tuberculosis relapse; Group 2, newly diagnosed Months after the beginning of cavity pulmonary tuberculosis p treatment Group 1 Group 2 (n=102) (n=104) DISCUSSION Cavity closure after four months 31 (30.4) 56 (53.8) <0.001 Cavity closure after eight months 37 (36.3) 29 (27.9) 0.198 The analysis of patients with tuberculosis relapse Cavity was presented after the 34 (33.3) 19 (18.3) 0.014 revealed that among patients with newly diagno- eighth month sed pulmonary tuberculosis 32.4% were diagnosed Group 1, pulmonary tuberculosis relapse; Group 2, newly diagnosed pulmonary tuberculosis until 2014, while 67.6% started suffering from the disease after 2014. After analysing various factors A total of 206 patients with tuberculosis experi- that could affect this parameter, we found that un- enced a destructive process in the lungs. Cavities til 2014 these patients with newly diagnosed pul- were observed in 102 (89.5%) patients with tu- monary tuberculosis had received an eight-month berculosis relapse and in 104 (82.5%) with newly course of chemotherapy for the first time, whereas diagnosed pulmonary tuberculosis (p=0.121). after 2014 all patients had undergone a six-month- Cavern closure more likely occurred in the pa- long tuberculosis treatment. In this case, the dura- tients with newly diagnosed pulmonary tuber- tion of the treatment is supposed to be a factor that culosis than in group 1 (81.7% versus 66.7%), can prevent tuberculosis relapse. This conclusion e.g. destruction was present after eight months of is consistent with findings reflected in other rese- the treatment in 19 (out of 104; 18.3%) patients arches (5,8), albeit the negative impact of a long- with newly diagnosed tuberculosis compared to lasting course of chemotherapy on the occurrence 34 (out of 102; 33.3%) patients with TB relapse of tuberculosis relapse has been emphasized in (p=0.014) (Table 4). some studies (14). Thus, it is worth considering As for the time to disappearance of infiltrative the occurrence of new cases of tuberculosis relapse and focal changes, its values were almost indi- in patients with resistant forms of tuberculosis in stinguishable from the time necessary for cavity response to the implementation of new short-term closure with an exception of solitary cases when treatment regimens recommended by the WHO infiltration disappeared before cavern closure, or, (15). Moreover, the most effective treatment re-

360 Butov et al. Treatment in patients with MDR tuberculosis

gimen to reduce the probability of relapse in the including results presented in this study, that the at-risk group of patients has not been determined effectiveness of anti-tuberculosis therapy in pati- by clinical trials (16). Recent studies have shown ents with tuberculosis relapse is much lower than that long-term treatment reduces the relapse rate in in patients with newly diagnosed pulmonary tu- patients with newly diagnosed pulmonary tuber- berculosis. However, we have noticed that in all culosis (17). Other studies have demonstrated that studies there was no evaluation of the treatment short-term chemotherapy regimens increase the ineffectiveness and outcomes in patients with cidence of tuberculosis relapse (18). Implementa- MDR-TB relapse. tion of novel innovative prolonged chemotherapy In conclusion, standard treatment effectiveness regimens for patients with tuberculosis relapse for the patients with MDR newly diagnosed pul- seems to be of vital importance for reaching more monary tuberculosis usually manifests by a faster favourable treatment outcomes (19,20). improvement and stabilization of health, less in- When evaluating the effectiveness of treatment in tense bacterial secretion in sputum, earlier smear the patients with tuberculosis relapse based on cli- conversion, higher frequency of cavity closu- nical and morphological changes, more pronoun- re and achievement of clinical and radiological ced symptoms of intoxication were observed com- improvements, more cases of cured patients and pared with patients with the primary disease or less common treatment failures compared with newly diagnosed pulmonary tuberculosis (8,21). multidrug-resistant tuberculosis relapse. Our findings corroborate this conclusion. Moreo- ver, it has been reported that the prevalence of the ACKNOWLEDGMENTS pathological process in patients with tuberculosis We acknowledge the wholehearted support of all relapse is higher with the presence of more ca- clinicians, nurses, and lab staff who contributed verns in the lungs, evidenced by radiographic exa- to this research and made this study possible. Our mination, than in patients with newly diagnosed sincere gratitude is expressed to the experts in the pulmonary tuberculosis (8,21). However, authors tuberculosis field who kindly shared their opini- did not take into account MTB resistance to an- ons and suggestions with us. ti-tuberculosis drugs. Nevertheless, some authors (22,23) argue that there is no statistically signifi- FUNDING cant difference in tuberculosis symptoms in pati- This research was supported by the Ministry of ents with relapse compared to those with newly Health of Ukraine using the funds provided by diagnosed pulmonary tuberculosis. the state budget as a fragment of research titled As for the evaluation of sputum bacterial loads and ‘’To develop a strategy for the prevention and treatment effectiveness by the presence of MTB in treatment of liver toxicity in patients with multi- sputum, most authors (23,24) indicate that more drug-resistant tuberculosis in chemotherapy and massive bacterial loads and later sputum culture smoking’’ (No 0120U102042) conversion were observed in patients with tuber- No commercial interest was involved. culosis relapse than newly diagnosed pulmonary tuberculosis, which is consistent with our findings. TRANSPARENCY DECLARATION Furthermore, there is some evidence (25,26), Competing interests: None to declare.

REFERENCES 1. World Health Organization. Global Tuberculosis Re- 3. Butov DO, Kuzhko MM, Makeeva NI, Butova TS, port 2019: WHO Report 2019. https://apps.who.int/iris/ Stepanenko HL, Dudnyk AB. Association of inter- bitstream/handle/10665/329368/9789241565714- leukins genes polymorphisms with multi-drug resist- eng.pdf?ua=1 (01 December 2019) ant tuberculosis in Ukrainian population. Pneumonol 2. World Health Organisation. The End TB Strategy. Alergol Pol 2016; 84:168–73. Global strategy and targets for tuberculosis preven- 4. Dudnyk A, Butov D, Crudu V, Lange C, Chesov D. tion, care and control after 2015. https://www.who. MDR-TB in Eastern Europe in the era of the TB int/tb/post2015_TBstrategy.pdf (01 December 2019) elimination action framework. Int J Tuberc Lung Dis 2017; 21:2–3.

361 Medicinski Glasnik, Volume 17, Number 2, August 2020

5. Cox H, Kebede Y, Allamuratova S, Ismailov G, 14. Berkani M, Chaulet P, Darbyshire JH, Nunn Davletmuratova Z, Byrnes G, Stone C, Niemann A, Fox W. Results of a therapeutic trial comparing S,Rüsch-Gerdes S, Blok L, Doshetov D. Tubercu- a 6-month regimen to a 12-month regimen in the losis recurrence and mortality after successful tre- treatment of pulmonary tuberculosis in the algerian atment: Impact of drug resistance. PLoS Med 2006; sahara. Final report: results 3 years after the onset of 3:e384. treatment. Rev Mal Respir 1986; 3:73-5. 6. Gelmsnova IY, Ahmad Khan F, Becerra MC, Ze- 15. World Health Organisation consolidated guide- mlyanaya NA, Unakova IA, Andreev YG, Berezina lines on drug-resistant tuberculosis treat- VI, Pavlova VE, Shin S, Yedilbayev AB, Krasnov ment. https://apps.who.int/iris/bitstream/hand VA, Keshavjee S. Low rates of recurrence after le/10665/311389/9789241550529-eng.pdf?ua=1(30 successful treatment of multidrugresistant tubercu- December 2019) losis in Tomsk, Russia. Int J Tuberc Lung Dis 2015; 16. Mirsaeidi M, Sadikot RT. Patients at high risk of 19:399-405. tuberculosis recurrence. Int J Mycobacteriol 2018; 7. Chin AT, Rylance J, Makumbirofa S, Meffert S, Vu 7:1-6. T, Clayton J, Mason P, Woodruff P, Metcalfe J. 17. Chang KC, Leung CC, Yew WW, Ho SC, Tam CM. Chronic lung disease in adult recurrent tuberculosis A nested case-control study on treatment-related risk survivors in Zimbabwe: a cohort study. Int J Tuberc factors for early relapse of tuberculosis. Am J Respir Lung Dis 2019; 23:203-11. Crit Care Med 2004; 170:1124-30. 8. Butov D, Gumenuik M, Gumeniuk G, Tkachenko 18. Qin F, Barry PM, Pascopella L. Factors associated A, Kikinchuk V, Stepaniuk R, Peshenko A,Butova with extended treatment among tuberculosis patients T. Effectiveness of anti-tuberculosis chemotherapy at risk of relapse in California. Int J Tuberc Lung Dis in patients with tuberculosis relapse compared with 2016; 20:363-9. newly diagnosed patients. Int J Mycobacteriol 2019; 19. Rogozhina NA, Gur’ianov VN, Babin MM. The 8:341-6. social-clinical aspects of recurrences of pulmonary 9. World Health Organisation. Tuberculosis pre- tuberculosis. Probl Tuberk 1993; 1:54-55. vention, care and control: a practical directory 20. Butov D, Feshchenko Y, Kuzhko M, Gumenuik of new advances. https://apps.who.int/iris/bit- M, Yurko K, Grygorova A, Tkachenko A, Nekrasova stream/handle/10665/44756/9789241502658_eng. N, Tlustova T, Kikinchuk V, Peshenko A, Butova T. pdf?sequence=1&isAllowed=y (01 December 2019) Effectiveness of Intravenous Isoniazid and Etham- 10. World Health Organisation. Xpert MTB/RIF imple- butol Administration in Patients with Tuberculosis mentation manual: technical and operational ‘how- Meningoencephalitis and HIV Infection. Tuberc to’; practical considerations. https://apps.who.int/iris/ Respir Dis 2020; 83:96-03. bitstream/handle/10665/112469/9789241506700_ 21. Feshchenko Y, Butov D, Kuzhko M, Gumeniuk eng.pdf?sequence=1 (01 December 2019) M, Butova T. Efficacy and safety of intravenous 11. World Health Organisation. Definitions and re- chemotherapy during intensive treatment phase in porting framework for tuberculosis – 2013 re- patients with newly diagnosed pulmonary tubercu- vision. https://apps.who.int/iris/bitstream/ losis. Adv Respir Med 2018; 86:159–67. handle/10665/79199/9789241505345_eng. 22. Kissina TE, Freidlin IS, Knoring BE, Basek TS, El- pdf?sequence=1 (01 December 2019) kin AB. Features of specific immune response in the 12. Stinson K, Eisenach K, Matsumoto M, Siddiqi S, patients with fibrous/cavernous tuberculosis of lungs Nakashima S, Hashizume H, Timm J, Morrissey A, Med Immunol 2006; 8:501-10. Mendoza M, Mathai P, eds. Mycobacteriology Lab- 23. Krasnov VA, Potashova VA, Zyrianova TV, Nary- oratory Manual. Geneva: WHO, 2014. https://www. shkina SL. Clinical aspects and outcome of recurrent who.int/tb/laboratory/mycobacteriology-laboratory- pulmonary tuberculosis. Probl Tuberk 1993; 5:14-6. manual.pdf (01 December 2019) 24. Riekstinia V, Torp L, Leĭmane V. Risk factors for 13. Lapach SN, Chubenko AV, Babich PN. Statistical early relapse of tuberculosis in Latvia. Probl Tuberk methods in biomedical studies using excel. Kyiv: Bolezn Legk 2005; 1:43-7. Morion; 2000. 25. Zakoska M. Risk factors for relapses of tuberculosis. Eur Resp J 2003; 22:2172-9. 26. Datiko DG, Lindtjørn B. Tuberculosis recurrence in smear-positive patients cured under DOTS in South- ern Ethiopia: retrospective cohort study. BMC Pub- lic Health 2009; 9:348.

362 ORIGINAL ARTICLE

Preoperative tumour size as a predictor of the presence of lymphovascular invasion in lung adenocarcinoma

Kemal Grbić1, Bakir Mehić2, Dalma Udovičić-Gagula3, Amina Valjevac4, Adem Ćemerlić5, Ferid Krupić6

1Clinic for Thoracic Surgery, 2Clinic for Lung Disease, 3Department of Pathology; University Clinical Centre Sarajevo, 4Department of Human Physiology, School of Medicine, University of Sarajevo, 5School of medicine, University of Sarajevo, Sarajevo, Bosnia and Herzegovina, 6Department of Anaesthesiology, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Sweden

ABSTRACT

Aim To examine whether preoperative tumour size may serve as a biomarker for the occurrence of lymphovascular invasion (LVI) in centrally and peripherally located lung adenocarcinoma.

Method The study included 261 patients surgically treated for diagnosed lung adenocarcinoma. A ROC curve was used to determine the biomarker potential of tumour size relative to the occurrence of LVI. Binary logistic regression was used to show changes of tumour size impact on the status of LVI.

Result Tumour prevalence according to localization had no statistical significance (p=0.464), while the presence of LVI in central, Corresponding author: as well as peripheral positions, was statistically significantly dif- Grbic Kemal ferent (p<0.001). The area under the curve of 0.978 highlights the Clinic of Thoracic Surgery, fact that tumour size is an excellent marker of the presence of LVI Clinical Centre University of Sarajevo in centrally located adenocarcinomas of the lung. A similar find- Bolnička 25, 71000 Sarajevo, ing was confirmed in peripherally located lung adenocarcinomas Bosnia and Herzegovina with an area below the curve of 0.943. Binary logistical regression showed that in centrally localized adenocarcinomas of the Phone: +387 33 297 238; lung, each additional centimetre of tumour growth represents an Fax: +387 33 297 937. increase in the likelihood of LVI+ by 17.14 times. In peripherally E-mail: [email protected] located adenocarcinomas of the lung, this increase in likelihood of ORCID ID: https://orcid.org/0000-0003- LVI for each centimetre of growth was 5.46 times. 3721-6369 Conclusion With a high degree of sensitivity and specificity, preoperative tumour size may serve as an important biomarker and positive predictor of the presence of LVI in lung adenocarcinoma of any location. Original submission: 06 May 2020; Key words: binary logistical regression, CT scanning, histopathological examination, surgical resection, T-descriptor Revised submission: 11 June 2020; Accepted: 23 June 2020 doi: 10.17392/1198-20

Med Glas (Zenica) 2020; 17(2): 363-368

363 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION tion (central and peripheral position), and according to LVI status, as tumours with the presence of The presence of lymphovascular invasion (LVI) LVI (LVI+) and tumours without the presence of in a tumour represents an independent negative LVI (LVI‒). individual prognostic factor in invasive lung adenocarcinoma (1). Majority of studies that tracked The approval for this study was obtained from the relation of LVI and the disease course showed the Ethics Committee of the University of Sara- a direct correlation with the appearance of local jevo Clinical Centre. and distant metastases, as well as modest results Methods in regards to survival without progression of the disease, disease-free survival, and the overall Preoperative tumour size was determined during five-year survival rates, which for all histologi- standard preoperative CT scanning of the tho- cal pictures of lung adenocarcinoma amounts to racic cage. The examination was performed with a modest 13-18% (2‒6). a GE LightSpeed VCT multi-slice CT (MSCT) Lung adenocarcinoma is most commonly char- machine, with 64 rows of detectors (General acterized by peripheral localization in the lung Electric Company, Fairfield, Connecticut, USA) parenchyma, zones of cavitation and necrosis, in the native and contrast (Ultravist 370, Scher- slow growth compared to other malignant lung ing, Germany) series in slices of 0.625 mm. tumours, masked clinical findings, and metasta- Transverse sections and 3D reconstruction was sis in the early course of the disease (6‒8). The used for the analysis. Among other radiological most common secondary metastases arising from characteristics, tumour size was measured by the primary disease are discovered in the central the thoracic radiologist in three planes; antero- nervous and musculoskeletal systems, as well as posterior, laterolateral and craniocaudal diam- the liver and lungs (6‒9). eters were measured, with the largest diameter in this study being expressed in millimetres (mm). Preoperative tumour size represents a signifi- Tumours that infiltrated lobar and proximal por- cant parameter of the therapeutic and prognos- tions of the segmental bronchi were classified as tic course, and is thus included in the current central, while those more distally localized were Tumor-Node-Metastases (TNM) classification of labelled as peripheral tumours. the disease (10). Previous research demonstrated a direct association of the clinical T descriptor All patients underwent preoperative broncho- and the appearance of LVI (11‒13), however not scopic examination under local anaesthesia with a enough research has been conducted on the deli- flexible Olympus series bronchoscope. During the cate cutoff tumour size at which the presence of bronchoscopy, fluid and tissue biopsies were taken LVI may be expected, as well as the reliability of for cytological and/or histological analysis. Tis- such a finding. sue biopsies included sampling of endobronchial tumour lesions, as well as transtracheal and trans- The aim of this study was to examine whether bronchial needle biopsies. In peripherally local- preoperative tumour size may serve as a bio- ized tumours, transthoracic needle biopsies were marker for the occurrence of lymphovascular in- performed under the control of CT or ultrasound. vasion (LVI) in centrally and peripherally located lung adenocarcinoma. The histopathological analysis of the resected pulmonary tissue was performed at the Depart- PATIENTS AND METHODS ment of Clinical Cytology and Pathology of USCC. After the primary sectioning and a 24- Study design and patients hour fixation of the resected tumour specimen in This cross-sectional study included 261 surgi- 10% buffered formalin, sliced tissue sections of cally treated patients at the Clinic for Thoracic 4-5 mm thickness were created and automatically Surgery, University of Sarajevo Clinical Centre processed into the tissue histoprocessor “Logos (USCC) with previously diagnosed lung ad- One Milestone” according to the manufacturer’s enocarcinoma, from January 2017 to December protocol. These specimens were then moulded 2018. Patients who underwent complete tumour into paraffin blocks and cut into slices of 4-5 µ resection were stratified according to the localiza- in thickness and were stained using the standard

364 Grbic and al. Lung adenocarcinoma and LVI

Hematoxylin-Eosin (H&E) method. Utilizing mean age of males was 62.36 ± 6.45 years and it microscopic examination of the stained sections, was statistically significantly higher than the fe- the type of tumour according to the current World male patients, 60.12 ±9.11 years (p=0.022). Health Organization (WHO) classification was The finding of tumours in the central and periph- determined (14). Without question, the presence eral positions was almost equal, 136 (52.11%) of lymphovascular invasion was denoted as lym- and 125 (47.89%), respectively (p=0.464), while phovascular invasion positive (LVI+), or absent the presence of lymphovascular invasion (LVI+) - lymphovascular invasion negative (LVI‒). was more frequently found than absent (LVI‒) in both central and peripheral tumour localizations Statistical analysis (p<0.001) (Table 1). Continuous variables with normal distribution The sensitivity of a finding of the presence of were expressed as mean ± standard deviation and LVI in a centrally positioned tumour >4.5 cm categorical variables were expressed as a number in size amounted to 89.7%, with a specificity of (percentage). A two sample t-test was used to com- 100%; the positive predictive value was 100%, pare continuous variables with normal distribution while the negative predictive value was found to and χ2 or Fisher’s exact test (with Yates correction be 61.3%. The AUC for tumour size amounted to when needed) to compare categorical variables. 0.98 (p<0.001) (Figure 1, Table 2). Sensitivity and specificity of tumour size for pre- The sensitivity of a finding of the presence of LVI dicting lymphovascular invasion in patients with in a peripherally positioned tumour >4.5 cm in central and peripheral tumour localization were size amounted to 92.7%, while specificity was determined, and receiver operating characteristic 90.1%; the positive predictive value was found (ROC) curves were constructed by plotting sensitivity against (1- specificity). The area under the curve (AUC) was calculated and analysed with a one-tail test. Cutoff points were obtained by calculating the Youden index. Binary logistical regression was performed to show that a change in tumour size may impact the status of LVI. The assigned level of statistical significance was p<0.05.

RESULTS In the observed sample of diseased patients (n=261) a statistically significant difference in the incidence of the disease in relation to gender was found (male:female = 1.75:1) (p<0.001). The

Table 1. Relationship of tumour localization and status of lymphovascular invasion (LVI) Location of the No (%) of patients p tumour LVI + LVI - Total Figure1. Receiver Operating Characteristics (ROC) curve Central 117 (44.83) 19 (7.29) 136 (52.11) of the tumour size as a potential marker of the appearance Peripheral 82 (31.41) 43 (16.47) 125 (47.89) < 0.001 of lymphovascular invasion in centrally positioned tumours; Total 199 (76.24) 62 (23.76) 261 (100) AUC, area under the curve; CI, confidence interval Table 2. Sensitivity and specificity of tumour size as a potential marker for the appearance of lymphovascular invasion in centrally positioned tumours Variable AUC Sensitivity (%) Specificity (%) PPV (%) NPV (%) 95% CI p Tumour size (cm) (cutoff – 4.5) 0.978 89.7 100 100 61.3 0.957 - 0.999 <0.001 AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value; CI, confidence interval

Table 3. Sensitivity and specificity of the tumour size as a potential marker for the appearance of lymphovascular invasion in peripherally positioned tumours Variable AUC Sensitivity (%) Specificity (%) PPV (%) NPV (%) 95% CI p Tumour size (cm) (cutoff – 4.5) 0.943 92.7 90.1 92.5 86.7 0.894 - 0.992 <0.001 AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value; CI, confidence interval

365 Medicinski Glasnik, Volume 17, Number 2, August 2020

of the disease in those with LVI present compared to those tumours without LVI (96.2% and 70.7%, respectively) (17). The prevalence of centrally localized tumours in our study was not statistically significantly more frequent compared to peripherally localized tumours, even though other studies have shown that adenocarcinoma of the lung is more frequently positioned peripherally (7,8,18,19). Our results are not consistent with the report by the American Cancer Society, which unequivocally states that adenocarcinoma of the lung is a peripheral tumour (6). Yang et al. (20), in their large meta-analysis, which included 397,189 diseased patients with Figure 2. Receiver Operating Characteristics (ROC) curve of lung adenocarcinoma, found peripheral tumours the tumour size as a potential marker of the appearance of in 97.5%, and only 2.5% centrally localized tu- lymphovascular invasion in peripherally positioned tumours; AUC, area under the curve; CI, confidence interval mours; the authors found more aggressive tumours among centrally positioned adenocarci- to be 92.5%, while the negative predictive value noma, meaning more frequent local and distant amounted to 86.7%. The AUC for tumour size metastases, and shorter overall survival conclud- was 0.94 (p<0.001) (Figure 2, Table 3). ing that adenocarcinoma location in the main Binary logistic regression showed that the tu- bronchus may be a predictor of metastasis and mour size was an independent positive predictor poor prognosis of this histologic tumour type. of the occurrence of LVI in lung adenocarcinoma Moon et al. (18), in patients with diagnosed lung with central (OR=17.14; 95%CI (4.32‒68.04); adenocarcinoma with present LVI, found more (p<0.001), but also peripheral tumour locations frequent nodal and distant metastases in those (OR=5.46; 95%CI (3.053‒9.78) (p<0.001). with centrally located tumours compared to those with peripheral positions. Sun et al. (19), in fol- DISCUSSION lowing patients with adenocarcinoma, found In this study, LVI+ was present in 76.24% resec- more frequent nodal metastases and shorter pe- tions, which is significantly higher than the re- riods without the disease relapse in those with sults of research by other authors (3-5). LVI as centrally positioned tumours. an independent prognostic factor of poor disease The results of this study showed that the size of outcome has been established by Funai et al.; by lung adenocarcinoma is an excellent marker of following surgically treated patients due to lung the occurrence of LVI with a sensitivity of 89.7% adenocarcinoma, up to 3 cm in size, they found and specificity of 100% for centrally positioned that those patients that had LVI in tumours had tumours, and with 97.2% sensitivity and 90.1% a worse 5-year survival rate than those in whom specificity for peripherally positioned tumours. tumours LVI was not found (70.9% vs. 94.5%) Studies by other researchers have shown that (15). Fan et al. (16) followed patients with op- with an increasing tumour size, the proportion erated adenocarcinoma with tumour size up to of lymphovascular invasion increases linearly 2 cm and found that diseased patients with the (15,21,22). Igai et al. (21) found in lobectomy re- presence of LVI had more numerous relapses of sectate at lung adenocarcinoma up to 1 cm in size the disease within the five-year follow-up period the presence of LVI in only 9% of cases. Funai et than the patients without LVI (87.5% and 72.1%, al. (15) found LVI present in 15% of the surgi- respectively). A group of Japanese authors (Nor- cally treated patients with lung adenocarcinoma ifumi et al.) followed patients with lung adeno- up to 2 cm. Higgins et al. (22), in non-small cell carcinoma of all subtypes in whom tumours were lung cancer (NSC resectate, of which there were smaller than 3 cm and found a higher proportion 734 patients operated on due to adenocarcinoma

366 Grbic and al. Lung adenocarcinoma and LVI

up to 3 cm in size, found involvement of vascular authors, in 208 patients with NSCLC resectate, and lymphatic vessels in 22% of patients. of which 57.4% were found to have lung ade- In our study, logistic regression showed that the nocarcinoma, concluded that central positions of tumour size was an independent positive predic- tumours with a diameter larger than 3 cm, dia- tor of the occurrence of LVI in lung adenocar- gnosed preoperatively, are significant predicti- cinoma with central, but also peripheral tumour ve factors of invasion of lymphatic vessels and positioning meaning that in centrally located lymph node drainage networks (24). lung adenocarcinomas, the growth of the tu- In conclusion, preoperative tumour size repre- mour size by each additional centimetre results sents an excellent biomarker and is a trustworthy in an increase in the likelihood of LVI+ by 17.14 positive predictor of the occurrence of lympho- times. In peripherally located lung adenocarci- vascular invasion in adenocarcinoma of the lung, noma this increase in likelihood was lower, and for both central as well as peripheral positions. for each additional centimetre the risk increases From a practical standpoint, it is important for by 5.46 times. Similarly, a group of Korean aut- clinicians to be cognizant of the fact that a preo- hors, following surgically treated patients with perative finding of lung adenocarcinoma greater lung adenocarcinoma and utilizing regression than 4.5 cm confidently implies that LVI is pre- analysis, showed that preoperative size, central sent. Metastatic lesions which may not be iden- localization, and the presence of LVI in a tumour tified with imagining techniques may be encoun- represent independent negative predictors of the tered intraoperatively, and also that the indicated disease recurrence (23). In lung adenocarcino- tumour size may help guide recommendations ma patients, Sun et al. showed with multivariate for therapeutic treatment by multidisciplinary analysis that the tumour size of 1-4 cm in the cen- teams of physicians. tral lung position is an independent prognostic factor of LVI status, and a period without disease FUNDING progression (19). Igai et al. (21) analysed several No specific founding was received for this study. parameters in patients with adenocarcinoma of the lung showing that lymphovascular invasion TRANSPARENCY DECLARATION was a negative prognostic factor for the course Competing interest: None to declare. of the disease. Conversely, a group of German

REFERENCES 1. Mollberg NM , Bennette C , Howell E , Backhus L , 5. Russell PA, Wainer Z, Wright GM, Daniels M, Con- Devine B , Ferguson MK. Lymphovascular invasion ron M, Williams RA. Does Lung Adenocarcinoma as a prognostic indicator in stage I non-small cell subtype predict patient survival? J Thorac Oncol lung cancer: a systematic review and meta-analysis. 2011; 6:1496‒504. Ann Thorac Surg 2014; 97:965‒71. 6. American Cancer Society, Cancer Facts & Figures 2. Sung SY, Kwak YK, Lee SW, Jo IY, Park JK, Kim 2018. http://www.cancer.org/acs/groups/content/@ KS, Lee KY, Kim YS. Lymphovascular invasion research/documents/webcontent/acspc-042151.pdf increases the risk of nodal and distant recurrence in (19 February 2020). node-negative stage I-IIA non-small-cell lung can- 7. Lin Y, Shidan W, David G, Yaniun Z, Feng X, Yui- cer. Oncology 2018; 95:156‒62. wei L, Hao L, Guanghua X, Qinghua Z, Adi G, Yang 3. Okiror L, Harling L, Toufektzian L, King J, Routled- X. Main bronchus location is a predictor for metas- ge T, Harrison-Phipps K, Pilling J, Veres L, Lal R, tasis and prognosis in lung adenocarcinoma; A large Bille A. Prognostic factors including lymphovas- cohort analysis. Lung Cancer 2018; 120:22‒6. cular invasion on survival for resected non-small 8. Popper HH. Progression and metastasis of lung can- cell lung cancer. J Thorac Cardiovasc Surg 2018; cer. Cancer Metastasis Rev 2016; 35:75‒9. 156:785‒93. 9. Riihimäki M, Thomsen H, Sundquist K, Sundquist J, 4. Ichikawa T, Aokage K, Sugano M, Miyoshi T, Koji- Hemminki K. Clinical landscape of cancer metasta- ma M, Fujii S, Kuwata T, Ochiai A, Suzuki K, Tsu- ses. Cancer Med 2018; 7:5534‒42. obi M, Ishii G. The ratio of cancer cells to stroma 10. Goldstraw P, Chansky K, Crowley J, Rami-Porta R, within the invasive area is a histologic prognostic Asamura H, Eberhard W, Nicholson AG, Grome P, parameter of lung adenocarcinoma. Lung Cancer Mitchell A, Bolejack V. The IASLC Lung Cancer 2018; 118:30‒5. Staging Project: Proposals for revision of the TNM stage groupings in the forthcoming (Eighth) edition of the TNM classification for lung cancer. J Thorac Oncol 2016; 1:39‒51.

367 Medicinski Glasnik, Volume 17, Number 2, August 2020

11. Wang S, Zhang B, Qian J, Qiao R, Xu J, Zhang L, 18. Moon J, Kyo Lee Y, Sook WS, Jae KP. Differing his- Zhao Y, Zhang Y, Wang R, Zhao R, Han B. Propo- topathology and prognosis in pulmonary adenocar- sal on incorporating lymphovascular invasion as a cinoma at central and peripheral locations. J Thorac T-descriptor for stage I lung cancer. Lung Cancer Dis 2016; 8:169‒77. 2018; 125:245‒52. 19. Sun W, Yang X, Liu Y, Yuan Y, Lin D. Primary tumor 12. Chanyeong P, In JL, Seung HJ, Jae WL. Factors af- location is a useful predictor for lymph node metas- fecting tumor recurrence after curative surgery for tasis and prognosis in lung adenocarcinoma. Clin NSCLC: impacts of lymphovascular invasion on ear- Lung Cancer 2017; 18:49‒55. ly tumor recurrence. J Thorac Dis 2014; 6:1420‒28. 20. Yang L, Wang S, Gerber DE, Zhou Y, Xu F, Liu J, 13. Grbic K, Mehic B. Characteristics of lymphovascu- Liang H, Xiao G, Zhou Q, Gazdar A, Xi Y. Main lar metastatic spread in lung adenocarcinoma accor- bronchus location is a predictor for metastasis and ding to the primary cancer location. Med Glas (Ze- prognosis in lung adenocarcinoma; a large cohort nica) 2020; 17:66‒72. analysis. Lung Cancer 2018; 120:22‒6. 14. Travis WD, Brambilla E, Nicholson AG, Yatabe 21. Igai H, Matsuura N, Tarumi S, Chang SS, Misaki Y, Austin JHM, Beasley MB, Chirieac LR, Da- N, Go T Ishikawa S, Yakomise H. Prognostic factors cic S, Duhing E, Flieder DB, Geisinger K, Hirsc in patients after lobectomy for p-T1aN0M0 adeno- FR, Ishikawa Y, Kerr KM, Noguchi M, Pelosi G, carcinoma. Eur J Cardiothorac Surg 2012; 41:603–6. Powell CA, Tsao MS, Wistuba I. The 2015 World 22. Higgins KA, Chino JP, Ready N, D’Amico TA, Health Organization Classification of lung tumors Berry MF, Sporn T, Boyd J, Kelsey CR. Lympho- impact of genetic, clinical and radiologic advances vascular invasion in non-small-cell lung cancer: im- since the 2004 classification. J Thorac Oncol 2015; plications for staging and adjuvant therapy. J Thorac 10:1243‒60. Oncol 2012; 7:1141-47. 15. Funai K, Sugimura H, Morita T, Shundo Y, Shimizu 23. Hyun JK, Hai X, Chang-Min C, Joon SS, Hyeong K, Shiiya N. Lymphatic vessel invasion is a signifi- RK, Jung BL, Mi YK. Preoperative CT predicting cant prognostic indicator in stage IA lung adenocar- recurrence of surgically resected adenocarcinoma of cinoma. Ann Surg Oncol 2011; 18:2968‒72. the lung. Medicine (Baltimore) 2016; 95(2):e2513. 16. Yang F, Chen K ,Liao J, Li X, Sun K, Bao D, Wang J. 24. Moulla Y, Gradistinac T, Wittekind C, Eichfeld U, Risk factors of recurrence for resected T1aN0M0 inva- Gockel I, Dietrich A. Predictive risk factors for sive lung adenocarcinoma: a clinicopathologic study lymph node metastasis in patients with resected of 177 patients. World J Surg Oncol 2014; 12:285. non-small cell lung cancer: a case control study. J 17. Norifumi T, Takahiro M, Yoshihiro M, Shinsuke S, Cardiothorac Surg 2019; 14:11. Tomoharu Y, Kei K, Yukio T, Shuji M, Tomoyuki Y. Prognostic significance of vascular invasion in intermediate-grade subtype of lung adenocarcinoma. Jpn J Clin Oncol 2016; 46:1015–21.

368 ORIGINAL ARTICLE

ABO blood group genotypes and ventilatory dysfunction in patients with allergic and nonallergic asthma

Marina Lampalo1, Irena Jukić2,3, Jasna Bingulac-Popović4, Hana Safić Stanić4, Nikola Ferara5, Sanja Popović-Grle1,5

1Clinical Department for Lung Diseases Jordanovac, University Hospital Centre Zagreb; 2Medical Department, Croatian Institute of Transfusion Medicine (CITM); Zagreb, 3School of Medicine, Josip Juraj Strossmayer University, Osijek, 4Molecular Diagnostics Depart- ment, CITM, Zagreb, 5School of Medicine, University of Zagreb, Zagreb; Croatia

ABSTRACT

Aim ABO blood group genotypes are established as a genetic factor in pathophysiology of various diseases, such as cardiovascular disorders, cancers, infectious diseases and there is rising evidence of their involvement in other conditions. The aim of this study was to determine if ventilatory changes of lung function in asthma, measured by biomarkers/parameters, are connected to certain ABO blood group genotypes in Croatia.

Methods A case-control study included 149 patients with asthma and 153 healthy individuals (blood donors). ABO genotyping on five main alleles was performed using PCR-SSP method. All patients had spirometry performed and severity of asthma was es-

Corresponding author: timated. Clinical parameters of spirometry (FEV1, FEV/FVC, 2 Marina Lampalo PEF), biomarkers FeNO, IgE and pO2 were measured. The χ Clinical Department for Lung Diseases test, Fisher's test, Kruskal-Wallis test and Spearman's correlation coefficients with p˂0.05 were used as statistically significant. Jordanovac, University Hospital Centre Zagreb, Results There was no determined statistically significant differen- Jordanovac 104, 10000 Zagreb, Croatia ce in both ABO genotypes and phenotypes between patient and Phone: +385 91 1414 734; control groups. Comparison of the lung function in different ABO phenotypes in asthmatic patients also did not show any statistically Fax: +385 (0)1238 267 significant differences in FEV1 values, FEV/FVC ratio or PEF. E-mail: [email protected] Statistically significant differences in oxygenation between diffe- ORCID ID: 0000-0002-8529-9472 rent ABO blood types have not been noticed (p=0.326). Differen- ces in quantitative values of biomarkers (FeNO and IgE) between different ABO blood phenotypes in patients with asthma were not significant, except for IgE that had marginal values (p=0.074).

Original submission: Conclusion No correlation was found between certain ABO blo- 04 November 2019; od group genotypes and parameters/biomarkers of ventilatory Revised submission: dysfunction in patients with allergic and nonallergic asthma. 28 January 2020; Key words: asthma, blood group antigens, respiratory function Accepted: tests 07 April 2020 doi: 10.17392/1099-20

Med Glas (Zenica) 2020; 17(2): 369-374

369 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION during specific immunotherapy (23). Clinical biomarkers FeNO, IgE and pO are very important The ABO blood group system is the most impor- 2 for patients suffering from asthma. The FEV1 tant one, because it represents an immunological (forced expiratory volume in first second) is the barrier against transfusion of incompatible blood most used parameter of lung function, FEV1/ group or organ transplantation (1). Yamamoto FVC (forced vital capacity) ratio is a parameter cloned ABO genes in 1990 and that enabled more of airway obstruction and PEF (peak expiratory advanced ABO structural and functional analyses volume) represents important parameter in the (2). This was followed by studies about ABO ge- diagnosis and control of asthma (24). nes polymorphism and oligosaccharide antigens A and B, whose results are today successfully appli- There are contradictory results about ABO blood ed in forensics, transfusion, cell, tissue and organ group genotypes as a genetic risk factor in patho- transplantation and in cellular and evolutional bio- physiology of the asthma (13,14). logy (3). Years ago, the correlation between ABO The aim of this study was to determine if venti- blood group and incidence of certain diseases was latory changes of lung function in asthma, mea- established (4,5). That finding was followed by a sured by biomarkers/ parameters of disease, are series of studies that confirmed higher frequency connected with certain ABO blood group ge- of other tumours in blood type A phenotype (ne- notypes in Croatia. urological, salivary glands, colon, uterus, ovarian, PATIENTS AND METHODS pancreas, kidney, urinary bladder) (6,7). In O blood type carriers, higher incidence of skin tumours Patients and study design and melanomas was detected (8). In the last decade research discovered a higher frequency of This case-control study included 149 adult pa- thromboembolic incidents in non-O blood type tients with asthma using medical records. The carriers and marked ABO genotypes as one of the diagnosis of asthma was established through genetic risk factors for cardiovascular diseases patients’ history and the performance of clinical (9,10). There is a small number studies about the and laboratory tests. Blood samples were taken in correlation of the ABO blood group system and the Clinical Department for Lung Diseases Jor- incidence of various respiratory system diseases, danovac, the University Hospital Centre Zagreb such as rhinitis (11,12), asthma (13,14), tubercu- between January and May 2017. A sample of 8.5 losis (15), lung carcinoma (16) and the existing mL of blood was drawn in EDTA anticoagulant studies are based only on ABO phenotypes. tube (Vacutainer PPT, Beckton Dickinson, USA). Asthma is an inflammatory disease characterized The blood samples were stored at -20°C until ge- by bronchial hyper reactivity and periodical epi- nomic DNA extraction. sodes of airway obstructions. It is one of the most A control group included 153 blood donors from common chronic diseases, especially in children, Zagreb county, registered in the Croatian Insti- and it has a rising incidence in developed coun- tute of Transfusion Medicine (CITM) between tries (17,18). Prevalence of asthma on the global January and May 2017; all were in good health, scale today is 1-18%, depending on a geographi- without any respiratory disease and with normal cal area (19). Asthma is the result of complex spirometry results. interaction between genetic and environmental All participants signed an informed consent. The factors and it is hard to define all the genes and study was approved by two Medical Ethics Com- biomarkers that are connected to pathogenesis mittees: Clinical Department for Lung Diseases of atopy and asthma. Previous studies showed Jordanovac and CITM. some genes which have a great role in pathogenesis: IgE-receptor (FcεRI), cytokine genes and Methods ADAM33 gene (20-22). Clinical biomarkers/parameters measured in Defining asthmatic biomarkers and parameters, asthmatic patients. Spirometry was performed which is crucial for its diagnosis, and determining for all asthmatic patients. Ventilatory dysfunction target spots in immunological reaction, would be was diagnosed according to spirometry results, an option to block the inflammation in asthma and severity of asthma was estimated according

370 Lampalo et al. ABO blood group genotypes and asthma

to recommendations of the Global Initiative for and control group, 34 (22.8%) and 39 (25.5%), Asthma (GINA 2019) (19): intermittent asth- respectively. In blood type B phenotype group the ma (GINA I), mild persistent asthma (GINA most frequent genotype was O1B1 in both patients II), moderate persistent asthma (GINA III) and and control group, 25 (16.8 %) and 27 (17.6%), severe persistent asthma (GINA IV and V)(19). respectively. In AB blood type carriers, A1B ge- The following clinical parameters were also me- notype was more frequent than A2B genotype in asured: serum level of IgE, partial pressure of both groups, nine (6.0%) and three (2.0%) and 12 oxygen in arterial blood (pO2), FeNO in exhaled (7.8%) and three (2.0%), respectively. There was air, FEV1 value, FEV1/FVC ratio, and PEF. no statistically significant difference in the distri- ABO genotyping by PCR-SSP method (alle- bution of ABO genotypes and phenotypes betwe- le specific PCR). The samples of patients and en the patient and control groups (Table 1). blood donors were tested at the Department of Table 1. Distribution of ABO phenotypes and genotypes in Molecular Diagnostics of CITM using PCR-SSP patients with asthma and in control group No (%) of samples (partially modified) method according to Gassner ABO phenotype ABO genotype p et al. (25). Five main ABO alleles were determi- Patients Controls A ned (O1, O2, A1, A1 and B) through eight pa- A1A1 4 (2.7) 6(3.9) rallel PCR-SSP reactions with coamplification of A1A2 3 (2.0) 1(0.7) A2A2 1 (0.7) 0(0.0) human growth hormone (HGH) gene fragment as 0.167 positive internal control. 01A1 34 (22.8) 39(25.5) 01A2 17 (11.4) 6(3.9) 02A1 2 (1.3) 2(1.3) Statistical analysis B 01B 25 (16.8) 27(17.6) Descriptive statistics was used. Categorical and no- 02B 2 (1.3) 0(0.0) 0.135 minal variables were shown by absolute frequenci- BB 0 2(1.3) es and corresponding portions, while quantitative AB A1B 9 (6.0) 12(7.8) 0.756 values were shown by medians and interquartile A2B 3 (2.0) 3(2.0) ranges. Normality of data distribution was analysed O by Kolmogorov-Smirnov test and based on the re- 0101 41 (27.5) 46(30.1) sults of certain questionnaires; because of the small 0102 8 (5.4) 8(5.2) 0.624 0202 0 1(0.7) sample size, in further statistical analysis nonpara- Total 149 153 metric statistical tests were used. The differences in categorical variables between patients and control There were no statistically significant differen- group: ABO group was analysed using χ2 test, whi- ces in FEV1 values (p=0.375), FEV/FVC ratio le Fisher's test was used in the analysis of statistical (p=0.741) or PEF (p=0.843) according to ABO significance if there was at least one frequency in phenotypes (Table 2). contingency table smaller than 5. The differences Table 2. Comparison of lung function in different ABO pheno- in quantitative variables were analysed using Kru- types in 148 patients with asthma Ventilatory skal–Wallis test. Spearman's rank correlation co- ABO No of capacity Min. Max. Median p phenotype patients efficients between certain quantitative values were parameters calculated. The p˂0.05 was considered significant. FEV1 (%) O 49 34.80 125.30 68.00 RESULTS A 60 20.90 119.40 80.10 0.375 B 27 30.40 106.40 82.60 This case-control study included 149 patients AB 12 36.00 101.30 66.25 in the stable phase of asthma, 57 (38.3%) males FEV1/FVC and 92 (61.7%) females, with average age of 60 O 49 37.69 99.26 65.47 A 60 38.08 98.82 68.79 0.741 years. The control group included 153 healthy B 27 0.67 82.48 69.71 individuals, 71 (46.4%) males and 82 (53.6%) AB 12 40.26 83.24 68.40 females, with average age of 43 years. PEF O 49 37.20 690.00 400.00 A 60 17.30 650.00 116.10 The leading genotype was O1O1, in 41 (27.5%) of 0.843 the patients’ group and in 46 (30.1%) cases of the B 27 30.80 650.00 395.00 AB 12 41.10 640.00 410.00 control group. In blood type A phenotype group the Min., minimum; Max., maximum; FVC, forced vital capacity; FEV1, most frequent genotype was O1A1 in both patients forced expiratory volume in one second; PEF, peak expiratory flow

371 Medicinski Glasnik, Volume 17, Number 2, August 2020

In 81 (54.4%) patients spirometry results did not lung function parameters between different ABO show airway obstruction (FEV1 ≥80%), where- phenotypes, and quantitative values of FeNO. as in 68 (45.6%) patients obstructive ventilatory De la Vega et al. found a higher frequency of disorders were found (FEV<80%). Based on the blood type A in asthmatic children (26). Alo et spirometry, the severity of asthma was estimated al. came to a similar result, blood group A was as follows: 16 (10.7%) patients had intermittent significantly more frequent in asthmatic - pati asthma (GINA I), 27 (18.1%) patients had mild ents (37%) in comparison with healthy controls persistent asthma (GINA II), 45 (30.2%) had mo- (26%) (27). derate persistent asthma, and 61 (40.9%) had se- Brachtel et al. showed higher incidence of blood vere persistent asthma. group antigens A and B in 239 German patients Significant differences in oxygenation betwe- with atopic conditions (atopic dermatitis, hay fe- en different ABO blood types were not noticed ver, allergic rhinitis, bronchial asthma and acute (p=0.326) (Table 3). urticaria), in comparison with 151 controls (28). Table 3. Differences in quantitative values of oxygenation be- Topno et al. conducted case-control study in In- tween different ABO blood phenotype in 125 patients with asthma dian population that included 100 patients with ABO No of symptoms of allergic rhinitis and 100 healthy con- Oxygenation Min. Max. Median p phenotype patients trols, and found the most frequent blood group 0 O 41 55.00 97.00 79.00 among patients with allergic rhinitis (52%) (29). A 52 52.00 95.00 79.00 pO2 (kPa) 0.326 B 23 60.00 99.00 78.00 The hypothesis that ABO system could be the AB 9 68.00 103.00 80.00 one of genetic risk factors for the development of Min., minimum; Max., maximum; pO2, partial pressure of oxygen in arterial blood asthma is set by Ronchetti many years later (30), suggesting that ABO/secretor genes through their There were no statistically significant differen- oligosaccharide structures (glycosyltransferases) ces in aspect of both biomarkers, FeNO and control adhesion of infective agents. Thus, genetic IgE, although significance of IgE was marginal variations in ABO blood group system could result (p=0.074) (Table 4). in higher sensitivity to bacterial and viral infections Table 4. Differences in quantitative values of FeNO and and, because of that, could be an inductor of asth- IgE biomarkers between different ABO blood phenotypes in ma development (30). Recent evidence suggests patients with asthma* that asthma is associated with some kind of immu- ABO No of Biomarkers Min. Max. Median p phenotype patients nodeficiency responsible for an increased suscepti- O 44 1.90 144.20 20.75 bility to infection in asthmatic patients (31). A 55 4.30 134.00 19.00 FENO 0.915 Although there are not many studies about the B 26 3.50 112.50 18.55 AB 10 5.30 70.50 20.15 correlation between the ABO blood group ge- O 41 1.00 5000.00 165.00 notypes and the development of asthma, it can A 47 6.28 2847.00 187.00 IgE 0.074 be noticed that big variations in the study design, B 24 5.86 1418.00 108.50 great heterogeneity of the results and final sta- AB 8 12.00 452.00 37.35 *Correlation between biomarker FENO and ABO phenotypes were tistical analyses altogether make it difficult to examined on 135 asthmatic patients and IgE on 120, respectively; come to some final conclusion (32). Interpreta- Min., minimum; Max., maximum; FeNO, fractional exhaled nitric oxide; IgE, immunoglobulin E; tion of the results is difficult also due to genetic heterogeneity between various ethnic groups and DISCUSSION local environmental exposure to allergens (33). We conducted this study primarily to investi- The authors also emphasize inappropriate sam- gate the correlation between ABO blood group ple size, problems in the classification of asthma system genotypes and ventilatory dysfunction in phenotypes or inadequate coverage of suscepti- asthma, since only few studies investigated this ble genes (34). There are even more published relation (13,14), and this kind of research had not studies that investigate relationship of ABO blo- yet been conducted in Croatia. We found no sta- od group and asthma in asthmatic children than tistically significant difference in the distribution in adults (26, 30, 32). The limitation of our study of ABO phenotypes and genotypes between asth- is a small sample size of asthmatic patients and matic patients and healthy controls as well as of the fact that controls were not matched for age

372 Lampalo et al. ABO blood group genotypes and asthma

and sex, but our study design of case–control, and adults who suffered from bronchial asthma however, strengthens the evidence of the obtai- and 2000 controls in South India, showed no ned results. The time-span of conducting the stu- statistical correlation between the ABO system dy is, by authors’ opinion, not relevant, because and development of asthma, which is exactly the ABO genotype is a genetically dependent varia- same result as presented in our study (14). The ble, unchangeable through the passage of time. authors from Malaysia, who analysed data from Some results showed that carriers of blood type 14 studies about the ABO system and allergic di- O/ secretor (SE/SE) and O/ LE (a-B-) were signi- seases including asthma also emphasized a gap ficantly associated with the development of asth- in geographic data and a need for further studies ma in childhood in Taiwan (35). Also, the older focusing on different populations (37). study of Kauffman et al. on the 228 adult coal Considering the given results and previous stu- miners showed significantly lower lung function dies, further studies of correlation between the and higher prevalence of wheezing and asthma in ABO blood group system and risk for asthma de- Lewis-negative, non-secretor, blood type O (13). velopment are required, but with a much larger A recent study in Brazilian patients with allergic number of examined subjects and other genetic rhinitis found significant difference in the - inci factors included, in order to confirm/reject the dence of carriers of O blood groups in males, hypothesis that the ABO system could be one of but not females (11). It is an interesting result, the genetic risk factors for developing asthma. because asthma and allergic rhinitis have similar immunopathology mechanism, and due to the fact ACKNOWLEDGMENTS that females more frequently suffer from asthma The authors would like to thank all collaborators than males. Recently, Uwaezoke et al. critically from the participating institutions and other staff review current evidence about linking ABO hi- involved in this study. sto-blood group with the risk of respiratory atopy in children and adults published within the past FUNDING 45 years (36). There are only eight studies taken No specific funding was received for the present in consideration and conclusions are that seve- study. re asthma is associated with B phenotype, while mild and moderate asthma is associated with O TRANSPERENCY DECLARATION and A phenotypes (36). In contrast, the case-con- Competing interests: None to declare. trol study of Bijanzadeh et al. among 200 children

REFERENCES 1. Aymard JP. Karl Landsteiner (1868–1943) and the 8. Hakomori S. Antigen structure and a genetic basis of discovery of blood groups. Transfus Clin Biol 2012; histo blood groups A, B and O: their changes asso- 19:244-8. ciated with human cancer. Biochim Biophys Acta 2. Yamamoto F. Molecular genetics of the ABO histo- 1999; 1473:247-66. blood group system. Vox Sang 1995; 69:1. 9. Jukic I, Bingulac-Popovic J, Dogic V, Babic I, Culej 3. Seltsam A, Hallensleben, Kollomann A, Blasczyk J, Tomicic M, Vuk T, Sarlija D, Balija M. ABO blo- R. The nature of diversity and diversification at the od groups and genetic risk factors for thrombosis in ABO locus. Blood 2003; 102:3035-42. Croatian population. Croat Med J 2009; 50:550–8. 4. Garratty G. Relationship of blood groups to disease: 10. Franchini M, Marano G, Vaglio S, Catalano L, Pupe- do blood group antigens have a biological role? Re- lla S, Liumbruno GM. The Role of ABO Blood Type vista Medica del Instituto Mexicano Seguro Social in Thrombosis Scoring Systems. Semin Thromb He- 2005; 43:113-21. most 2017; 43:525-9. 5. Liumbruno GM, Franchini M. Beyond immunohae- 11. Falsarella N, Ferreira AI, Nakashima F, de Mattos matology: the role of the ABO blood group in human Cde C, de Mattos LC. Evidence of an association diseases. Blood Transfus 2013; 11:491-9. between the O blood group and allergic rhinitis. Rev 6. Franchini M, Liumbruno GM, Lippi G. The progno- Bras Hematol Hemoter 2011; 33:444-8. stic value of ABO blood group in cancer patients. 12. Carpeggiani C. Allergic rhinitis and association Blood Transfus 2016; 14:434-40. with the O blood group. Rev Bras Hematol Hemoter 7. Rummel SK, Ellsworth RE. The role of the histo- 2011; 33:406-7. blood ABO group in cancer. Future Sci OA 2016; 13. Kauffmann F, Frette C, Pham QT, Nafissi S, -Ber 2:FSO107. trand JP, Oriol R. Associations of blood group-related antigens to FEV1, wheezing, and asthma. Am J Respir Crit Care Med 1996; 153:76-82.

373 Medicinski Glasnik, Volume 17, Number 2, August 2020

14. Bijanzadeh M, Ramachandra NB, Mahesh PA, Sa- 25. Gassner C, Schmard A, Nussbaumer W, Schönitzer vitha MR, Manjunath BS, Jayaraj BS. Lack of asso- D. ABO glycosyltransferase genotyping by polyme- ciation between asthma and ABO blood group. Lung rase chain reaction using sequence-specific primers. 2009; 187:389-92. Blood 1996; 88:1852-6. 15. Ganguly S, Sarkar P, Chatterjee D, Bandyopadhyay 26. De la Vega AR, Gómez CJ, Bacallao GJ. Genetic AR. Association of ABO blood group polymorphi- polymorphism of ABO and Rh system in relation sm and tuberculosis: A study on Bengalee Hindu ca- to bronchial asthma: preliminary report. Allergol ste population, West Bengal, India. Indian J Tuberc Immunopathol (Madr) 1976; 4:305-10. 2016; 63:242-4. 27. Alo MN, Eze UA, Abdulhi Yaro S, Jubril B, Nwano- 16. Chrysanthakopoulos NA, Dareioti NS. ABO blood ke NN. Relationship between ABO and Rhesus blo- group and the risk of lung cancer in Greek adults: od groups and susceptibility to asthma within Soko- a case - control study. Exp Oncol 2018; 40:249-50. to metropolis, Nigeria. Int J Immunol 2015; 3:37-41. 17. Chang C. Treatment of asthma in children. In: 28. Brachtel R, Walter H, Beck W, Hilling M. Associa- Gershwin ME, Albertson TE, eds. Current Clinical tions between atopic diseases and the polymorphic Practice: Bronchial Asthma: a Guide for Practical systems ABO, Kidd, Inv and red cell acid phospha- Understanding and Treatment. New Jersey: Humana tase. Hum Genet 1979; 49:337-48. Press Inc., 2006: 65–111. 29. Topno N, Narvey VP, Jain AK. The correlation of 18. Pearce N, Aït-Khaled N, Beasley R, Mallol J, Keil allergic rhinitis with ABO phenotype. Indian J Oto- U, Mitchell E, Robertson C; ISAAC Phase Three laryngol Head Neck Surg 2019; 71(Suppl 3):1827- Study Group. Worldwide trends in the prevalence of 31. asthma symptoms: phase III of the International Stu- 30. Ronchetti F, Villa MP, Ronchetti R, Bonci E, Latini dy of Asthma and Allergies in Childhood (ISAAC). L, Pascone R, Bottini N, Gloria-Bottini F. ABO/Se- Thorax 2007; 62:758–66. cretor genetic complex and susceptibility to asthma 19. The global burden of asthma report. Global initiative in childhood. Eur Respir J 2001; 17:1236–8. for asthma (GINA). http://www.ginaasthma.org (09 31. Patella V, Bocchino M, Steinhilber G: Asthma is January 2020). associated with increased susceptibility to infection. 20. Van Eerdewegh P, Little RD, Dupuis J, Del Mastro Minerva Med 2015; 106:1-7. RG, Falls K, Simon J, Torrey D, Pandit S, McKenny 32. Bottini N, Ronchetti F, Gloria-Bottini F. Cooperative J, Braunschweiger K, Walsh A, Liu Z, Hayward B, effect of adenosine deaminase and ABO-secretor ge- Folz C, Manning SP, Bawa A, Saracino L, Thackston netic complex on susceptibility to childhood asthma. M, Benchekroun Y, Capparell N, Wang M, Adair R, Eur Respir J 2002; 20:1613-5. Feng Y, Dubois J, FitzGerald MG, Huang H, Gib- 33. Quinzii C, Belpinati F, Pignatti PF. Predictive Ge- son R, Allen KM, Pedan A, Danzig MR, Umland SP, netic Testing – New Possibilities in Determination Egan RW, Cuss FM, Rorke S, Clough JB, Holloway of Risk of Complex Diseases. Croat Med J 2001; JW, Holgate ST, Keith TP. Association of the 42:458-62. ADAM33 gene with asthma and bronchial hyperres- 34. Sanz-Lozano CS, García-Solaesa V, Davila I, Isi- ponsiveness. Nature 2002; 418:426–30. doro-García M. Applications of molecular genetics 21. Hakonarson H, Halapi E. Genetic analyses in asth- to the study of asthma. Methods Mol Biol 2016; ma: current concepts and future directions. Am J 1434:1-13. Pharmacogenomics 2002; 2:155-66. 35. Chen YL, Chen JC, Lin TM, Huang TJ, Wang ST, 22. Bijanzadeh M, Mahesh PA, Ramachandra NB. An Lee MF, Wang JY. ABO/secretor genetic complex is understanding of the genetic basis of asthma. Indian associated with the susceptibility of childhood asth- J Med Res 2011; 134:149–61. ma in Taiwan. Clin Exp Allergy 2005; 35:926-32. 23. Pevec B, Radulovic Pevec M, Stipic Markovic A, 36. Uwaezuoke SN, Eze JN, Ayuk AC, Ndu IK. ABO Batista I, Rijavec M, Silar M, Kosnik M, Korosec P. histo-blood group and risk of respiratory atopy in House dust mite-specific immunotherapy alters the children: a review of published evidence. Pediatric basal expression of T regulatory and FcERI pathway Health Med Ther 2018; 9:73–9. genes. Int Arch Allergy Immunol 2012; 159:287-96. 37. Dahalan NH, Tuan Din SA, Mohamad SMB. Asso- 24. Ciprandi G, Gallo F. The impact of gender on asth- ciation of ABO blood groups with allergic diseases: ma in the daily clinical practice. Postgrad Med 2018; a scoping review. BMJ Open 2020; 10:e029559. 130:271-3.

374 ORIGINAL ARTICLE

Epidemiology of neonatal sepsis caused by multidrug resistant pathogens in a neonatal intensive care unit level 3

Devleta Hadžić1, Fahrija Skokić1, Selmira Brkić2, Amina Saračević1, Delila Softić2, Dženana Softić1

1Paediatric Clinic, University Clinical Centre of Tuzla, 2School of Medicine University of Tuzla; Bosnia and Herzegovina

ABSTRACT

Aim Steady progress in intensive treatment worldwide has increased the survival of immature neonates, but with multiple invasive procedures, which have increased the risk of infection, thus the bacterial resistance to antibiotics. The aim of this study was to analyse the epidemiology of multidrug resistance pathogens as causative agents of neonatal sepsis in the neonatal intensive care unit.

Methods A retrospective cohort study conducted at the Intensive care unit of the Paediatric Clinic of Tuzla over a three-year period (2016-2018) analysed epidemiology of neonatal sepsis caused by multidrug resistance pathogens. Statistical analysis applied standard methods, and the research was approved by the Ethics Com- mittee of the institution. Corresponding author: Results Of the total of 921 treated neonates, multidrug resistance Devleta Hadžić (MDR) pathogens among causative agents of neonatal sepsis were Paediatric Clinic, found in 22 neonates (2.38%) with no gender difference. Prema- University Clinical Centre of Tuzla turity and low birth weight were confirmed as the most significant Prof.dr. Ibre Pašić bb, 75000 Tuzla, risk factors. From the maternal risk factors a significant difference Bosnia and Herzegovina was found in the first birth and in vitro fertilization. Clinically, Phone: +387 35 303 733; MDR sepsis manifested frequently as late onset sepsis, with longer hospital stay and higher mortality. The findings of leukope- Fax: +387 35 303 730; nia, thrombocytopenia and coagulation disorders were significant. E-mail: [email protected] Gram negative bacteria were frequently isolated, in particular Aci- ORCID ID: https://orcid.org/0000-0003- netobacter, which showed the greatest resistance to antibiotics. 4037-3736 Conclusion Neonatal MDR sepsis is a threat to life, it complicates the treatment, increases costs and mortality. Outcomes can be improved by preventive strategies, earlier and more accurate diagnosis and rational use of antibiotics. Original submission: 04 March 2020; Key words: antibiotic resistance, incidence, intensive care, neona- Revised submission: tal early onset sepsis, neonatal late onset sepsis 04 May 2020; Accepted: 23 May 2020 doi: 10.17392/1157-20

Med Glas (Zenica) 2020; 17(2): 375-382

375 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION World Health Organization (WHO) and the Euro- pean Centre for Diseases Prevention and Control Bacterial resistance to antibiotics is one of the le- (ECDC) (1), the control of the spread of multidrug ading problems of medicine in the new millenni- resistance pathogens must involve not only he- um (1). Non-rational use of antibiotics has led alth care professionals, but also the governments to the genetic mutation of bacteria, creating new and community (5). It is best to use our own re- resistant species that spread very quickly, prima- commendations for antibiotic administration, deri- rily in hospital settings, but also in the outpatient ved from analysis of local susceptibility and dyna- community (2). In recent decades, the creation of mics of the infection. Certainly, we are obliged to new antibiotics has stagnated significantly, which follow the pharmacological and antimicrobial gu- particularly affects the most critical categories of idelines for the choice of empirical and etiological patients (3). Some bacteria, such as Staphylo- therapy. Therefore, continuous epidemiological coccus, Enterococcus, Pseudomonas and Acine- monitoring of local susceptibility patterns to an- tobacter, are becoming resistant to most antibio- tibiotic agents is necessary to establish a rational tics that are in use (4). In addition, many invasive treatment strategy. There is insufficient data pu- diagnostic and therapeutic procedures in the blished on this topic from Bosnia and Herzegovi- modern medicine increase the risk of infection, na, even from the region, especially in the case of which requires permanent establishment of qua- critically ill neonates. lity prevention and efficient treatment (5). The aim of this study was to analyse the epidemi- The immune status of all critically ill patients ology of multidrug resistance pathogens as cau- is impaired, which facilitates the progression of sative agents of neonatal sepsis in the neonatal "super bacteria" (6). This is especially emphasi- intensive care unit level 3. zed in the most risky areas of medicine such as surgery, intensive medicine and neonatology (7). PATIENTS AND METHODS Neonates are at highest risk for morbidity and mortality, and despite the advancement of evi- Patients and study design dence-based medicine, the problem seems to be expanding (8). Possible sources of infection for A retrospective cohort study, which included all neonates are also spreading, increasing the risk of consecutive neonates with positive blood culture, health care-associated infections with multidrug from those treated at the Neonatal Intensive Care resistant (MDR) pathogens (9). Neonatal sep- Unit (NICU) of the Paediatric Clinic in Tuzla sis has continuously remained a top issue (10). (capacity of 20 beds, level III) over a three-year It is very important to have current information period (2016 to 2018). All neonates with proven on neonatal bacterial and other isolates and their multidrug resistant (MDR) pathogens isolated patterns of antimicrobial susceptibility, and thus from blood culture were designated as a test be guided in the choice of empirical antibiotic group, MDR sepsis group. They were compared therapy (1). The mortality rate of neonatal sep- with a control group consisting of the rest of nesis is different and ranges from 20-75% (10). The onates from the neonatal sepsis group, marked as outcome depends mainly on an early and timely non-MDR sepsis group. identification of the cause, as well as timely onset The study was approved by the Ethics Committee of an adequate causal therapy (10). of the University Clinical Centre of Tuzla. The rate of multidrug resistant pathogens is diffe- Methods rent at different centres (11). Commonly, MDRs include gentamicin-resistant Klebsiella species, Clinical and demographic data were obtained third-generation cephalosporin-resistant gram-ne- from medical records and electronic databases gative organisms, methicillin-resistant Staphylo- of patients treated in the NICU, including gen- coccus, and, more recently, carbapenem-resistant der, gestational age, birth weight, perinatal risk gram-negative species (3). Although the problem factors for neonatal sepsis, maternal and neonatal of antibiotic resistance is more pronounced in (presence of central venous catheter, length of some countries, the bacteria do not recognize state mechanical ventilation, parenteral nutrition and borders. According to the recommendations of the length of hospitalization), clinical presentation,

376 Hadžić et al. Multidrug resistant pathogens in neonatal sepsis

laboratory findings (C-reactive protein - CRP, Table 1. Characteristics of 921 patients admitted to the complete blood cell count - CBC, the highest and neonatal intensive care unit (NICU) lowest value for white blood cell count - WBC, Characteristics No (%) of patients Gestational age (weeks) absolute neutrophil count - ANC, immature -to- Term (≥37GW) 420 (45.6) total neutrophil ratio - I/T ratio, platelet count, Preterm (<37GW) 501(54.4) and coagulation status), applied therapy and Late preterm (34-36GW) 328 (35.6%) outcome. During admission, clinical status of ne- Very preterm (<32GW) 173 (18.7) Extremely preterm (<28GW) 58 (6.2) onates was scored by the SNAP-PE (Score for Birth weight (grams) Neonatal Acute Physiology - Perinatal Extensi- Low (<2500 g) 278 (30.1) on) and CRIB II (Clinical Risk Index for Babies) Very low (<1500g) 116 (12.5) Extremely low (<1000g) 38 (4.1) score (12). From laboratory findings we parti- Mechanical ventilation within NICU stay 330 (35.8) cularly analysed potential markers of infection. Surgery within NICU stay 46 (4.9) Early onset of neonatal sepsis was defined as an Sepsis Clinically confirmed 396 (42.9) infection that develops in the first 72 hours of Confirmed on blood cultures 187 (20.3) life, and late onset neonatal sepsis was defined MDR sepsis 22 (2.38) as an infection that develops after 72 hours of Total 921 (100) GW, gestational weeks; NICU, neonatal intensive care unit; MDR, life (10). Multidrug resistant bacteria were con- multidrug resistance pathogen sidered to be those showing resistance to three or Of the total of 921 treated neonates, multidrug more antimicrobial classes. Particular attention resistance (MDR) pathogens among causati- was paid to possible isolates of methicillin-resi- ve agents of neonatal sepsis were found in 22 stant Staphylococcus, vancomycin-resistant En- (2.38%), evenly in both genders. terococcus, gentamicin-resistant Klebsiellaspp, third-generation cephalosporin-resistant gram- Perinatal risk factors for the development of ne- negative organisms and carbapenem-resistant onatal sepsis were analysed in the MDR sepsis Enterobacteriaceae. There was no prophylactic group, and compared with the control non-MDR administration of fluconazole in the NICU at the sepsis group (Table 2). Immaturity and low birth time of the study, and empiric antibiotic therapy Table 2. Perinatal risk factors in the neonates with multidrug for neonatal sepsis included ampicillin with gen- resistant (MDR) and non MDR sepsis tamicin for suspected early onset neonatal sepsis, No (%) of patients and ceftazidime with amikacin for neonatal sep- in the group Variable MDR Non-MDR sis suspected of nosocomial pathogens. This ini- p tial antibiotic therapy was corrected according to (n=22) (n=165) confirmed isolates, including meropenem and/or Gestational age (weeks) Preterm 17 (77.3) 63 (38.2) 0.0005 vancomycin, if necessary, depending on further Late preterm 4 (18.2) 37 (22.4) 0.6546 clinical course and findings in particular patient. Very preterm 13 (59.1) 26 (15.8) 0.0001 Extremely preterm 2 (9.1) 3 (1.8) 0.0456 Statistical analysis Birth weight (grams) Low (<2500g) 15 (68.2) 49 (29.7) 0.0004 Standard methods of descriptive statistics (central Very low (<1500g) 2 (9.1) 9 (5.5) 0.5016 Gender tendency measures, dispersion measures) were Male 13 (59.1) 103 (62.4) 0.7645 used. Parametric and non- parametric significan- Female 9 (40.1) 62 (37.6) 0.7645 ce tests (χ2-test, Student's t- test) as well as linear Mode of delivery correlation method were used to test the signifi- Caesarean section 8 (36.4) 59 (35.8) 0.956 Vaginal 14 (63.6) 106 (64.2) 0.956 cance of differences between the samples. Statisti- Apgar score (AS) cal hypotheses were tested at a significance level AS in the first minute <5 9 (40.1) 47 (28.5) 0.2641 of α= 0.05, e.g. the difference between the samples AS in the fifth minute <5 6 (27.3) 25 (15.2) 0.1522 Maternal risk factors was considered significant at p< 0.05. No 12 (54.5) 92 (55.8) 0.9082 Firstborn 19 (86.4) 94 (57.0) 0.0081 RESULTS In vitro fertilization 4 (18.2) 4 (2.4) 0.0006 Infections in pregnancy 3 (13.6) 29 (17.6) 0.64 During the three-year period, 921 neonates were Amniotic infectious 2 (9.1) 24 (14.5) 0.4912 treated at the NICU (Table 1). PROM 2 (9.1) 38 (23.0) 0.1351 PROM, premature rupture of membranes

377 Medicinski Glasnik, Volume 17, Number 2, August 2020

weight were statistically significant among peri- es of neonatal disease severity scoring systems natal risk factors originating from neonates. Gen- (p<0.02 and p<0.008), comparing to the control der and mode of delivery did not show signifi- non-MDR sepsis group (Table 3). Neonates with cant difference between the groups. Maternal risk MDR sepsis often had multi-organ dysfunction factors showed low significance, especially for and required more supportive therapy. Almost the first birth and in vitro fertilization, that were all neonates with MDR sepsis received paren- significantly more frequent in the MDR group teral nutrition, required mechanical ventilation compared to the control. support, received intravenous immunoglobulins The MDR sepsis was observed in neonates with more often, comparing to the control non-MDR significantly lower gestational age (33.59±3.1 vs. sepsis group. They also required longer intensi- 36.27±3.3), lower birth weight (1852.08±746.3 ve treatment (20.7±10.8 vs.12.4±6.93 and had vs. 2841.78±805.9) and lower Apgar scores in a higher mortality rate comparing to the control the first and fifth minute (Table 3). non-MDR sepsis group (p<0.008) (Table 3). There were no significant differences in CRP Table 3.Clinical characteristics of the neonates with multidrug resistant (MDR) and non MDR sepsis levels between MDR and non-MDR sepsis MDR Non-MDR group. Also, there were no significant differ- Variable p (n=22) (n=165) ences in the absolute or immature neutrophil GA (weeks) (mean±SD) 33.59±3.1 36.27±3.3 0.0004 count to total neutrophil ratio between MDR BW (g) (mean±SD) 1852.08±746.3 2841.78±805.9 <0.0001 and non-MDR sepsis group (Table 4). Neonates AS 1st minute (mean±SD) 5.6±1.0 6.98±2.3 0.0061 AS 5th minute (mean±SD) 6.63±3.0 7.78±1.67 0.0073 with MDR sepsis had slightly lower leukocyte PRM (h) (mean±SD) 2.02±0.5 5.18±11.3 0.1924 values (10.56±3.1 vs. 14.12±8.4), and they had CRIB II (mean±SD) 6.3±3.8 3.94±4.08 0.0243 significantly more often recorded leukopenia, SNAPEPE II (mean±SD) 32.3±15.7 23.51±24.99 0.0873 Early onset of sepsis comparing to the control non-MDR sepsis group 7(31.8) 105 (63.6) 0.0043 (No, %) (p<0.0003)(Table 4). Significant differences in Late onset of sepsis 15(68.2) 60 (36.4) 0.0043 platelet counts were also recorded, neonates with (No, %) Gram-positive bacteria MDR sepsis had lower platelet count (160.31±3 4(18.2) 120(72.7) >0.0001 (No, %) vs. 223.27±116.6), and they had significantly Gram-negative bacteria 18(81.8) 45(27.3) >0.0001 (No, %) Table 4. Laboratory characteristics in the neonates with Meningitis (No, %) 6(27.3) 56 (34.0) 0.5308 multidrug resistant (MDR) and non MDR sepsis Pneumonia (No, %) 10(45.5) 35(21.2) 0.0123 Increase in PVR (No, %) 4(18.2) 6(3.6) 0.0041 MDR sepsis Non-MDR Variable p Severe IVH (No, %) 9(40.1) 5(3.0) >0.0001 (n=22) sepsis (n=165) Initial acute renal failure CRP (mg/L) (mean±SD) 38.06±0.5 21.69±40.1 0.0576 12(54.5) 34(20.6) 0.0005 (No, %) CRP <5 mg/ℓ (No, %) 12 (54.5) 88 (53.3) 0.9156 IVIG (No, %) 12(54.5) 27(16.4) <0.0001 CRP >5 mg/ℓ (No, %) 10 (45.5) 77 (46.7) 0.9156 Mechanical ventilation Htc (L/L) (mean±SD) 0.53±0.29 0.51±0.10 0.5169 14 (63.6) 25 (15.2) <0.0001 (No, %) Htc anaemia (No, %) 2 (9.1) 5 (3.0) 0.1554 Parenteral nutrition Htc normal (No, %) 14 (63.6) 124 (75.2) 0.2449 18 (81.8) 45 (27.3) <0.0001 (No, %) Htc polycythemia (No, %) 6 (27.3) 36 (21.8) 0.5614 Inotropes (No, %) 17(72.3) 40(24.2) 0.0001 Leukocytes (x109/L) 10.56±3.1 14.12±8.4 0.05 NICU stay (days) (mean±SD) 20.7± 10.8 12.40±6.93 <0.0001 (mean±SD) Leukocytes <6 (No, %) 9 (40.9) 19 (11.5) 0.0003 Outcome Leukocytes >13 (No, %) 6 (27.3) 69 (41.8) 0.1924 Survivors (No, %) 18 (81.8) 158(95.8) 0.0086 ANC (x109/L) (mean±SD) 4627.36±883.0 6056.70±4551.2 0.1444 Non-survivors (No, %) 4 (18.2) 7(4.2) 0.0086 ANC <1800 (No, %) 5 (22.7) 33 (20.0) 0.7675 GA, gestational age; SD, standard deviation; BW, birth weight; AS ANC >13000 (No, %) 1 (4.5) 23 (13.9) 0.2150 1st min, Apgar score in the first minute; AS 5th minute, Apgar score IT ratio 0.088±0.02 0.082±0.09 0,7562 in the fifth minute; PRM, premature rupture of membranes; CRIB Platelets (x109/L) 160.31±30.0 223.27±116.6 0.01 II, clinical risk index for babies scoring system; SNAPEPE II, score (mean±SD) for neonatal acute physiology-perinatal extension; PVR, pulmonary Platelets <150 (No, %) 11(50.0) 39 (23.6) 0.008 vascular resistance; IVH, intraventricular haemorrhage; IVIG, intra- Albumin (g/L) (mean±SD) 26.13±19.0 27.57±4.2 0.4 venous immunoglobulins NICU, Neonatal intensive care unit coagulation status normal 7 (31.8) 95 (57.6) 0.02 Clinically, neonatal sepsis caused by MDR pat- (No, %) coagulation status disrupted hogens manifested more frequently as late onset 15 (68.2) 70 (42.4) 0.02 (No, %) sepsis (p<0.004), more often with Gram-nega- CRP, C reactive protein; Htc, haematocrit; ANC, absolute neutrophil tive bacteria (p<0.0001), and with higher valu- count; IT ratio, immature/total neutrophil ratio;

378 Hadžić et al. Multidrug resistant pathogens in neonatal sepsis

more often recorded thrombocytopenia, com- even in the same place. This is attributed to diffe- paring to the control non-MDR sepsis group rences in quality of life, predisposing factors for (p<0.008). Also, coagulation disorders were infection, and usage of antibiotics (14 - 20). Neo- more frequently observed in the test compared to natal sepsis caused by multidrug resistant (MDR) the control group (p<0.02) (Table 4). pathogens is an important cause of morbidity and Gram-negative bacteria were more frequently mortality in critically ill neonates (7,14,20). In our isolated in the MDR sepsis group, in particular study multidrug resistance pathogens among cau- Acinetobacter, which showed the greatest resi- sative agents of neonatal sepsis were found in 22 stance to antibiotics (Table 5). Gram-negative neonates (2.38%). Generally, prevalence of MDRs bacteria were generally resistant to Gentamicin, is reported higher (21-24). According to Behma- while Enterobacter and Acinetobacter were re- di et al. (11), pathogens in late onset sepsis were sistant to several antibiotics; in two cases Aci- significantly resistant to antibiotics, ranging from netobacter was carbapenem-resistant, which has 13.6–47.8%. Analysing Gram-negative neonatal limited therapeutic antibacterial choices. We had sepsis in their 8-year cohort study, Tsai et al. (20) no recorded in vitro resistance to vancomycin, found that MDRs accounted for 18.6% of all ne- colistin and fluconazole. onatal Gram-negative bacteraemia in the NICU. The results of Yusef et al. (25) show that MDRs Table 5. Multidrug resistant (MDR) pathogens in neonatal sepsis are the most common cause of sepsis at their NICU No (%) of strains and are associated with higher mortality compa- Total EOS LOS red with non-MDR sepsis. Multiply, the neonates Bacterial isolate p (n=22) (n=7) (n=15) are at risk of infection, and as is known, sources Gram-positive bacteria 4 2 (28.6) 2 (13.3) 0.386 may originate from maternal disease, infections, Staphylococcus 2 1 (14.3) 1 (6.7) 0.5641 Enterococcus 2 1 (14.3) 1 (6.7) 0.5641 interventions during pregnancy and/or childbirth, Gram-negative bacteria 18 5 (71.4) 13 (86.7) 0.386 or postnatally, and come again out of the hospital, Escherichia coli 3 1 (14.3) 2 (13.3) 0.9492 or even from the community (10). Unfortunately, Klebsiella 4 1 (14.3) 3 (20.0) 0.7468 they can all be MDR sources (14). Perinatal risk Enterobacter 3 1 (14.3) 2 (13.3) 0.9492 Proteus 1 1 (14.3) - 0.1339 factors for neonatal infection have been investi- Pseudomonas 2 - 2 (13.3) 0.3116 gated in numerous studies (13-18). Gestational Acinetobacter 5 1 (14.3) 4 (26.7) 0.5182 immaturity and low birth weight in our study were EOS, early onset sepsis; LOS, late onset sepsis confirmed as the most significant risk factors for the onset of MDR sepsis. According to Afonso and DISCUSSION Blot (13) gradual decrease in susceptibility to rou- Despite the steady progress in the intensive tre- tine antibiotic is more highlighted in lower birth atment of neonates worldwide, neonatal sepsis weight and premature neonates. There were no remains a top issue, as one of the major causes gender differences between MDR and non-MDR of neonatal morbidity and mortality (10). Current sepsis group in our study, although, other studies reports indicate that sepsis causes about a quar- found male predominance (11). Townsel et al. (18) ter of all neonates deaths, and sepsis mortality in their contemporary review, exploring gender has increased by 10-15% every year in the last differences in critically ill preterm neonates, found 2 decades (10). Suspected neonatal sepsis is a that male predominance maintains a gender gap in common indication for admission to the NICU neonatal outcomes. Similar conclusion was made (13-16). Blood culture is the gold standard for the in a research by Roy et al. (19) reporting statisti- confirmation of sepsis. The prevalence of culture cally significant association between male gender proven neonatal sepsis is different in various stu- and mortality among culture-positive neonates. dies, from 10% to 50%, which depends on crite- In our study maternal risk factors showed low ria and sampling technique, as well as from qua- significance, especially for the first birth and in lity of health care and hospital services in various vitro fertilization, that were more frequent in the countries (17). The incidence of blood culture- MDR group compared to the control. This is qu- proven sepsis in our NICU currently is 20.3%. ite different from other studies that mostly repor- The diversity of etiology of sepsis varies from one ted infections, prolonged rupture of the membra- region to another region, and changes over time ne, and lack of prenatal care (16).

379 Medicinski Glasnik, Volume 17, Number 2, August 2020

Clinically, in our study neonatal sepsis caused by association of thrombocytopenia and coagulation MDR pathogens manifested more frequently as disorders in septic patients is known (27), but in late onset sepsis, with higher values of neonatal neonates, thrombocytopenia usually predomina- disease severity scoring systems, requiring more tes in reports, while reports of coagulation status supportive therapy and longer intensive treatment, of neonates are less frequent. Klingenberg et al. and finally higher mortality rate comparing to the (17) concluded that neonatal sepsis is an active control non-MDR sepsis group. The results of other condition, and that careful assessment of clinical studies are generally similar (20-22). Intensive tre- manifestations in combination with properly se- atment in critically ill neonates involves invasive lected biomarkers can be used to support or deny procedures, that are also risk factors of infection the diagnosis of sepsis, all aimed at a more rati- (23,24). These are the so-called specific points of onal use of antibiotics. Gram negative bacteria attention, given that most health-associated infec- were more frequently isolated in the MDR sepsis tions in intensive care units are associated with the group, in particular Klebsiella and Acinetobacter, use of therapeutic devices (24). Recommendations which showed the greatest resistance to antibio- for the proper use of all appliances and medical tics. Our results showed similarity to the studies materials (probes, catheters, suction of secretions, from Jordan (25), Greece (29), Taiwan (20), and maintenance of venous catheters, etc.), as well as Egypt (21), which report better sensitivity causes monitoring and preventing the spread of infection, of early onset sepsis, while late onset sepsis pat- are mandatory (1). hogens showed a greater level of resistance, par- The timing of exposure, neonatal immune status, ticularly in very preterm and extremely low birth and causative agent virulence influence the -cli weight neonates. The high prevalence of neonatal nical expression of neonatal sepsis (13). Immu- sepsis caused by MDR pathogens in some regi- nologically impaired response, especially in pre- ons, as reported in India (7,22), Taiwan (20) and mature neonates, whose prolonged stay in NICU, others, is partly explained by the use of antibiotics, with more invasive procedures, makes them sui- for a class above the World Health Organization table for healthcare associated infection and mul- recommendations. Differences between multidrug tidrug resistant pathogens (24). resistant pathogens and their corresponding drug sensitive complement were not analysed in our The length of intensive treatment was signi- study, because MDR infections are still rather rare. ficantly longer in the MDR group (20.7 ± 10.8 Because the treatment has not produced satisfac- days) compared to the control non-MDR sepsis tory results, modern medicine is increasingly pro- group (12.40 ± 6.93 days). The available eviden- moting the prevention, education, responsibility ce suggests a higher incidence and mortality rate and permanent control of hospital infections (1). of late-onset sepsis in premature and very low birth weight neonates, but pathogen distribution In conclusion, neonatal MDR sepsis has a risky and risk exposure for MDRs are similar for all clinical course and outcome. It is a threat to life, it neonates admitted to the NICU (15,16,21,22). complicates treatment, prolongs NICU stay, increases costs and mortality. Recovery of those neona- There are different reports on the utility of labo- tes depends on timely clinical suspicion, adequate ratory parameters in the assessment of neonatal treatment and supervision. Reasonable and rati- sepsis, and certainly, it is a significant tool in all onal use of antibiotics is the only right choice in sepsis scoring systems (17, 26-28). In our study, today's challenges. Outcomes may be improved neonates with MDR sepsis had slightly lower by preventative strategies, earlier and accurate di- leukocyte values, with more often recorded le- agnosis, which require monitoring of local epide- ukopenia, and lower platelet count with more miological data to improve the treatment. frequently recorded thrombocytopenia, and without significant difference in the values of CRP, FUNDING absolute neutrophil count or I/T ratio, comparing to the control non-MDR sepsis group. Also, coa- No funding was received for this study. gulation disorders were more frequently observed TRANSPARENCY DECLARATION in the test compared to the control group. The results of other studies are similar (27,28). The Competing interests: None to declare.

380 Hadžić et al. Multidrug resistant pathogens in neonatal sepsis

REFERENCES 1. RhodesA, Evans LE, Alhazzani W,Levy MM, Anto- 14. Cailes B, Kortsalioudaki C, Buttery J, Pattnayak S, nelli M, Ferrer R, Kumar A, Sevransky JE, Sprung Greenough A, Matthes J, Bedford Russell A, Kennea CL, Nunnally ME, Rochwerg B, Rubenfeld GD, N, Heath PT; neonIN network. Antimicrobial resi- Angus DC, Annane D, Beale RJ, Bellinghan GJ, stance in UK neonatal units: neonIN infection sur- Bernard GR, Chiche JD, Coopersmith C, De Backer veillance network. Arch Dis Child Fetal Neonatal Ed DP, French CJ, Fujishima S, Gerlach H, Hidalgo JL, 2018; 103:474-8. Hollenberg SM, Jones AE, Karnad DR, Kleinpell 15. Black CG, Tavares L, Stachel A, Ratner AJ, Randis RM, Koh Y, Lisboa TC, Machado FR, Marini JJ, TM. Distribution of late-onset neonatal sepsis patho- Marshall JC, Mazuski JE, McIntyre LA, McLean gens differs in inpatient and outpatient settings. Am AS, Mehta S, Moreno RP, Myburgh J, Navalesi P, J Perinatol 2018; 105:38-43. Nishida O, Osborn TM, Perner A, Plunkett CM, Ra- 16. Ozkan H, Cetinkaya M, Koksal N, Celebi S, Ha- nieri M, Schorr CA, Seckel MA, Seymour CW, Shi- cimustafaoglu M. Culture-proven neonatal sepsis in eh L, Shukri KA, Simpson SQ, Singer M, Thompson preterm infants in a neonatal intensive care unit over BT, Townsend SR, Van der Poll T, Vincent JL, Wier- a 7 year period: coagulase-negative Staphylococcus singa WJ, Zimmerman JL, Dellinger RP. Surviving as the predominant pathogen. Pediatrics Internatio- sepsis campaign: international guidelines for ma- nal 2014; 56:60–6. nagement of sepsis and septic shock: 2016. Intensive 17. Klingenberg C, Kornelisse RF, Buonocore G, Maier Care Med 2017; 43:304–77. RF, Stocker M. Culture-negative early-onset neona- 2. Blair JMA, Webber MA, Baylay AJ, Ogbolu DO, tal sepsis - at the crossroad between efficient sepsis Piddock LJV. Molecular mechanisms of antibiotic care and antimicrobial stewardship. Front Pediatr resistance. Nat Rev Microbiol 2015; 13:42-51. 2018; 6:285. 3. Laxminarayan R, Duse A, Wattal C, Zaidi AK, Wer- 18. Townsel CD, Emmer SF, Campbell WA, Hussain theim HF, Sumpradit N, Vlieghe E, Hara GL, Go- N. Gender differences in respiratory morbidity and uld IM, Goossens H, Greko C, So AD, Bigdeli M, mortality of preterm neonates. Front Pediatr 2017; Tomson G, Woodhouse W, Ombaka E, Peralta AQ, 5:6. Qamar FN, Mir F, Kariuki S, Bhutta ZA, Coates A, 19. Roy P, Kumar A, Kaur IR, Faridi MM. Gender diffe- Bergstrom R, Wright GD, Brown ED, Cars O. Anti- rences in outcomes of low birth weight and preterm biotic resistance-the need for global solutions. Lan- neonates: the male disadvantage. J Trop Pediatr cet. Infect Diss 2013; 13:1057-98. 2014; 60:480-5. 4. Graham CJ. The global threat of antibiotic resistan- 20. Tsai M, Chu S, Hsu J, Lien R, Huang H, Chiang M, ce: what can be done? JoGHR 2017; 1:e2017002. Fu R, Lee C, Huang Y. Risk factors and outcomes for 5. Presidental documents. Executive order 13676 of multidrug-resistant Gram-negative bacteremia in the September 18, 2014. Combating antibiotic- resistant NICU. Pediatrics 2014; 133:e322-9. bacteria. Federal Register 2014; 184:56931- 5. 21. Nour I, Eldegla HE, Nasef N, Shouman B, Abdel- 6. Laxminarayan R, Matsoso P, Pant S, Brower C, Hady H, Shabaan AE. Risk factors and clinical Rottingen JA, Klugman K, Davies S. Access to outcomes for carbapenem-resistant Gram-negative effective antimicrobials: a worldwide challenge. late-onset sepsis in a neonatal intensive care unit. J Lancet 2016; 387:168–75. Hosp Infect 2017; 97:52-8. 7. Bandyopadhyay T, Kumar A, Saili A, Randhawa VS. 22. Jajoo M, Manchanda V, Chaurasia S, Sankar MJ, Ga- Distribution, antimicrobial resistance and predictors utam H, Agarwal R, Yadav CP, Aggarwal KC, Che- of mortality in neonatal sepsis. J Neonatal Perinatal llani H, Ramji S, Deb M, Gaind R, Kumar S, Arya Med 2018; 11:145-53. S, Sreenivas V, Kapil A, Mathur P, Rasaily R, Deora- 8. Raymond SL, Stortz JA,Mira JC, Larson SD, Wynn ri AK, Paul VK, Investigators of the Delhi Neonatal JL, Moldawer LL. Immunological defects in neona- Infection Study (DeNIS) collaboration, New Delhi, tal sepsis and potential therapeutic approaches. Front India. Alarming rates of antimicrobial resistance and Pediatr 2017; 5:14. fungal sepsis in outborn neonates in North India. 9. McPherson C, Liviskie C, Zeller B, Nelson MP, PLoS One 2018; 13:e0180705. Newland JG. Antimicrobial stewardship in neonates: 23. Garcia H, Torres-Gutierrez J, Peregrino-Bejarano L, challenges and opportunities. Neonatal Netw 2018; Cruz-Castaneda MA. Risk factors for nosocomial in- 37:116-23. fection in a level III neonatal intensive care unit. Gac 10. Shane AL, Sanchez PJ, Stoll BJ. Neonatal sepsis. Med Mex 2015; 151:711-9. Lancet 2017; 390:1770-80. 24. Verstraete EH, De Coen K, Vogelaers D, Blot S. Risk 11. Behmadi H, Borji A, Taghavi-Rad A, Soghandi L, factors for health care-associated sepsis in critically Behmadi R. Prevalence and antibiotic resistance of Ill neonates stratified by birth weight. Pediatr Infect neonatal sepsis pathogen in Neyshabour, Iran. Arch Dis J 2015; 34:1180-6. Pediatr Dis 2016; 4:e33818. 25. Yusef D, Shalakhti T, Awad S, Algharaibeh H, Kha- 12. Dorling JS, Field DJ, Manktelow B. Neonatal dise- sawneh W. Clinical characteristics and epidemiology ase severity scoring systems. Arch Dis Child Fetal of sepsis in the neonatal intensive care unit in the era Neonatal Ed 2005; 90:11–6. of multi-drug resistant organisms: A retrospective 13. Afonso EDP, Blot S. Effect of gestational age on review. Pediatr Neonatol 2018; 59:35-41. the epidemiology of late-onset sepsis in neonatal in- 26. Hornik CP, Benjamin DK, Becker KC, Benjamin DK tensive care units - a review. Expert Rev Anti Infect Jr, Li J, Clark RH, Cohen-Wolkowiez M, Smith PB. Ther 2017; 15:917-24. Use of the complete blood cell count in late-onset neonatal sepsis. Pediatr Infect Dis J 2012; 31:803-7.

381 Medicinski Glasnik, Volume 17, Number 2, August 2020

27. Del Vecchio A. Evaluation and management of 29. Gkentzi D, Kortsalioudaki C, Cailes BC, Zaoutis thrombocytopenic neonates in the intensive care T, Kopsidas J, Tsolia M, Spyridis N, Siahanidou S, unit. Early Hum Dev 2014; 90:51-5. Sarafidis K, Heath PT, Dimitriou G; Neonatal infec- 28. Perrone S, Lotti F, Longini M, Rossetti A, Bindi I, tion surveillance network in Greece. Epidemiology Bazzini F, Belvisi E, Sarnacchiaro P, Scapellato C, of infections and antimicrobial use in Greek neona- Buonocore G. C reactive protein in healthy term tal units. Arch Dis Child Fetal Neonatal Ed 2019; newborns during the first 48 hours of life.Arch Dis 104:293-7. Child Fetal Neonatal Ed 2018; 103:163-6.

382 ORIGINAL ARTICLE

Interleukin-8 heterozygous polymorphism (-251 T/A and +781 C/T) increases the risk of Helicobacter pylori-infection gastritis in children: a case control study

Supriatmo1, Gontar Alamsyah Siregar2, Iqbal Pahlevi Adeputra Nasution3, Oke Rina Ramayani1

1Department of Child Health, 2Department of Internal Medicine, 3Department of Surgery; School of Medicine, Universitas Sumatera Utara, Medan, Indonesia

ABSTRACT

Aim To investigate the effects of interleukin-8 (IL-8) -251 T/A and +781 C/T polymorphism on the risk of Helicobacter pylori- infection gastritis in children, and the IL-8 level of children with or without gastritis H. pylori infection according to polymorphism.

Methods This prospective, case control clinical study included 64 children 2-18 years old. A disease group (32 gastritis patients with H. pylori-infection) was compared with a control group (32 gastritis patients without H. pylori infection). Demographic characteristics of patients were taken by a questionnaire; gastritis was confirmed by gastroscopy, H. pylori infection was confirmed with rapid urease test. Serum IL-8 level was measured by ELISA, and IL-8 -251 T/A and +781 C/T polymorphisms were analysed by Corresponding author: RT-PCR. Demographic characteristics, IL-8 level, polymorphism Supriatmo of patients, and IL-8 level according to polymorphisms were com- Department of Child Health, pared between the groups. School of Medicine, Results Children with tobacco exposure were associated with an Universitas Sumatera Utara increased risk of H. pylori-infection gastritis by 3.4-fold. There dr. Mansyur 5, Medan, Indonesia was a higher IL-8 level in the disease group compared to the con- Phone: +62 811 613 029; trol group. The disease group with IL-8 -251 AT polymorphism E-mail: [email protected]; had a higher risk compared to TT polymorphism by 8.7-fold, and [email protected] with IL-8 +781 CT polymorphism had a higher risk compared to ORCID ID: https://orcid.org/0000-0001- CC polymorphism by 10.7-fold. Children in the disease group with IL-8 -251 AT and TT, and +781 CT and CC polymorphisms 5681-9777 produced a higher IL-8 level than the control group in respective polymorphisms.

Original submission: Conclusion Children with H. pylori-infection gastritis have higher IL-8 production. There was an increased risk of developing H. 09 March 2020; pylori-infection in heterozygous -251 AT and +781 CT. Accepted: 07 April 2020 Key words: chemokine CXCL8, genetic heterogeneity, genetic doi: 10.17392/1160-20 predisposition, gastrointestinal disease

Med Glas (Zenica) 2020; 17(2): 383-388

383 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION PATIENTS AND METHODS Gastritis is a disease characterized by the inflam- Patients and study design mation process in the mucosa layer of the stomach (1). It happened as a response to damage, ei- This study was conducted as a prospective, case ther acute or chronic that causes mucosal atrophy control clinical study at H. Adam Malik Gene- and epithelial metaplasia (2). Helicobacter pylori ral Hospital, and Universitas Sumatera Utara (H. pylori) infection is the most common cause Hospital, Medan, Indonesia, during the period of gastritis globally, but it can also be caused by between September 2019 and December 2019. another microorganism infection, autoimmune, All patients were 2 to 18 years old with gastritis bile acid reflux, radiation exposure, and also the confirmed by gastroscopy, who did not consume exposure to tobacco, consumption of alcohol, antibiotics, bismuth containing drugs, histamine use of non-steroidal anti-inflammatory drugs H-2 receptor antagonists, proton pump inhibitor (NSAIDs) or steroids (3). (PPI), and immunomodulatory drugs for the past Infection of H. pylori comprises approximately 4 weeks before admission. Patients with mali- 50% of the world’s population (4). Oral inges- gnancy, immunosuppression, metabolic disor- tion of H. pylori gives the bacteria access to ders, upper gastrointestinal bleeding, or history gastric mucosa and induce immune responses, of gastrointestinal surgery were excluded. which cause tissue damage (5). Immune re- A written informed consent was taken from all sponses against H. pylori produce pro-inflam- subjects’ legal guardians after providing suffici- matory cytokines, especially interleukin 8 (IL-8) ent information about the study prior to the study (6) through several pathways, including toll-like enrolment. receptor 5 (TLR5), neutrophil activating protein The study was approved by the Health Research A (NAP), heat shock protein 60 (Hsp60), cyto- Ethical Committee, School of Medicine, Univer- toxin-associated gene A (CagA), and vacuolat- sitas Sumatera Utara in accordance with the prin- ing cytotoxin A (VacA) (4). A previous study (7) ciples of the Helsinki Declaration. showed that the production of IL-8 genetically A study group consisted of 34 gastritis children determined with homozygous for the AA geno- with the positive H. pylori infection (disease gro- type at -251 position demonstrated a higher lev- up), whereas a control group consisted of 34 ga- el of IL-8 production, whereas other studies (5, stritis children without H. pylori infection. 8-10) suggest IL-8 -251 A allele is an important risk factor for the development of H. pylori– Methods associated gastric disease, and it is associated The infection of H. pylori was confirmed by with increased inflammation and severity of the using rapid urease test. All patients were inter- disease (11). Another important IL-8 polymor- viewed using a structured questionnaire to note phism is IL-8 +781C allele, for which a previ- gender, age, socioeconomic status, and tobacco ous study showed the association between IL-8 exposure. Weight and height of all patients were +781C allele with increased risk of non-small also measured, then Z-score of weight-for-height cell lung carcinoma (12), ovarian cancer (13), were calculated to determine nutritional status and also gastric cardiac cancer (14). However, based on WHO paediatric growth indicators (19). some studies reported that IL-8 +781C was not associated with the incidence of gastric disease Six millilitres of blood sample were collected (15), glioma (16,17), and osteosarcoma (18). from all patients and were subjected to mea- The role of IL-8 +781 polymorphism for gastric suring serum IL-8 level by Quantikine ELISA diseases is still not clear. Human CXCK8/IL-8 Immunoassay kit (R&D System, Minneapolis, USA), using ELISA tech- The aim of this study was to investigate the nique. To extract DNA, high Pure PCR Tem- effects of IL-8 -251A and +781C polymorphisms plate Preparation Kit (Roche Applied Science, on the risk of H. pylori-infection gastritis in chil- Penzberg, Germany), using spin column met- dren, and also the IL-8 levels of gastritis in chil- hod was used. Polymorphism of IL-8 -251 T/A dren with H. pylori infection or without infection and IL-8 +781 C/T was examined by using the according to their genotype polymorphism.

384 Supriatmo et al. IL-8 polymorphism H. pylori gastritis

following primers: IL8 -251T/A forward primer Table 1. Demographic characteristics of 34 gastritis children 5’-ATTGGCTGGCTTATCTTCA-3’; reverse pri- with H. pylori (+) and without H. pylori (-) infection mer 5’-CAAATACGGAGTATGACGAAAG-3’ Characteristic H. pylori (+) H. pylori (-) p Gender (No, %) and gene IL8 +781C/T forward primer Male 13 (52) 12 (48) 0.801 5’-GTGGTATCACAGAGGATTATGC-3’; re- Female 21 (48.8) 22 (51.2) Age mean verse primer 5’-CAGTCATAACTGACAA- 11.83 (3.52) 11.96 (3.37) 0.880 (SD) (years), CATTGATC-3’. The PCR cycle conditions con- Weight mean 36.47 (10.91) 37.29 (9.91) 0.746 sisted of 95oC for 10 minutes, followed by 40 (SD) (kg) o o Height median cycles of 92 C for 15 seconds and 60 C for 60 141.5 (104 – 165) 141.5 (104 – 162) 0.825 (min. – max.) (cm) seconds. Amplification was done by using C1000 BMI-for-age 0.37 (1.58) 0.52 (1.42) 0.689 Thermal Cycler CFX96 Real Time System (Bio- (Z-score) mean (SD) Rad Laboratories, California, USA). BMI (No, %) Severely wasted 2 (100 ) 0 Statistical analysis Wasted 0 2 (100) Normal 29 (51.8) 27 (48.2) Demographic characteristics of the children were Overweight 1 (25) 3 (75) Obese 2 (50) 2 (50) analysed for differences between gastritis with H. Socioeconomic status (No, %) pylori infection and without H. pylori infection Moderate – high 28 (48.3) 30 (51.7) 0.493 by using χ2 test for categorical data, independent Low 6 (60) 4 (40) Tobacco exposure (No, %) t-test for normally distributed numerical data, Yes 27 (60 ) 18 (40) 0.021 and Mann-Whitney U test if the distribution was No 7 (30.4) 16 (69.6) not normal. Serum IL-8 was compared between BMI, body mass index; gastritis with H. pylori infection and without H. pylori infection by using independent t-test if 44 61 data were normally distributed, otherwise Mann- 125 45 Whitney U test was used. A comparison of ge- 34 notype polymorphism between the groups was 100 analysed by using χ2 test, or Freeman-Halton 75 extension of Fisher’s exact test if χ2 assumptions were violated. Serum IL-8 level was com- 50 pared between the groups according to genotype Interleukin 8, pg/ml polymorphism by using independent t-test for 25 normally distributed data, and Mann-Whitney U test for non-normal distribution. Differences were 0 considered statistically significant at p<0.05. H. pylori (+)H. pylori (-) Figure 1. Interleukin-8 level in gastritis children with (+) or RESULTS without (-) H. pylori infection Gender, age, weight, height, nutritional status (weight-for-height) and socioeconomic status AT polymorphism had a significantly higher were similar among the disease and control group. risk of H. pylori-infection gastritis compared to Children with tobacco exposure were associated TT polymorphism, by 8.7-fold (95% CI: 2.55, with increased risk of H. pylori-infection gastritis 30.01). There was also a significant increase in by 3.4-fold (95% CI: 1.18, 9.99) (Table 1). the frequency of IL-8 +781 CT polymorphism There was a significantly higher IL-8 level in between the disease and the control group, having the disease group (median = 20.64) compared CT polymorphism increase the risk of H. pylori- to the control group (median = 10.64), U=160.5, infection gastritis compared to CC polymorphi- r = 0.62 (p<0.001) regardless of the genotype sm by 10.7-fold (Table 2). polymorphism (Figure 1). When IL-8 level of each polymorphisms was The frequency of IL-8 -251 AT and AA polymor- compared between the groups, children in the di- phism was significantly higher in the disease sease group with -251 AT and TT, and +781 CT than in the control group, but only children with and CC polymorphisms were producing a signifi-

385 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 2. Comparison of genotype polymorphism between ue of serum IL-8 levels in children with H. pylori- gastritis children with H. pylori (+) or without H. pylori (-) infection gastritis compared to their counterparts. infection This finding is in concordance with a previous Polymorphism No (%) of patients Odds Ratio* p (No of patients)H. pylori (+) H. pylori (-) (95% CI) study that showed infection of H. pylori inducing IL-8 (-251 T/A) IL-8 secretion by gastric cell by some mechanism, AA (11) 6 (54.5) 5 (45.5) 4.56 (0.98-21.32) such as high levels of CagA expression (25,25) AT (33) 23 (69.7) 10 (30.3) 0.001 8.74 (2.55-30.01) and via TLR4-dependent mechanism (26). TT (24) 5 (20.8) 19 (79.2) --- IL-8 (+781 C/T) Production of IL-8 is also influenced by the poly- TT (4) 2 (50) 2 (50) 3.43 (0.41-28.94) morphism of IL-8 gene, such as -251 T/A gene CT (33) 25 (75.8) 8 (24.2) <0.00110.71 (3.36-34.14) CC (31) 7 (22.6) 24 (77.4) --- and +781 C/T gene (15). The result of this study *comparison of -251 TT and +781 CC polymorphisms showed an increased risk of having H. pylori- infection gastritis in patients with -251 AT poly- cantly higher IL-8 level than the control group in morphism than TT polymorphism and also +781 respective polymorphism (Table 3). CC polymorphism than CC polymorphism. A previous study by Chang et al. (9) showed that Table 3. Interleukin-8 levels of gastritis children with H. pylori (+) or without H. pylori (-) infection according to having IL-8 -251 AT polymorphism increased genotype polymorphism risk in developing severe gastritis A study by Ta- Mean (SD) or median Mean guchi et al. (27) however did not find any associ- Polymorphism (min.-max.) (pg/mL) difference p ation between IL-8 -251 AT polymorphism with (95% CI) or U H. pylori (+) H. pylori (-) atrophic gastritis and also gastric cancer. Previ- IL-8 (-251 T/A) 20.98 12.31 8.67 ous studies (14,28) showed no significant rela- AA 0.073 (8.20) (5.27) (-0.99, 18.33) tionship between IL-8 +781 CC polymorphisms 19.01 10,69 AT 41 0.004 and the incidence of gastric disease. In contrast, (10.05–128) (5.57 – 17.61) 37.85 10.14 this study showed an increased risk of develop- TT 0.00 0.001 (19.32–121.05) (5.88 – 17.21) ing H. pylori-infection gastritis in patients with IL-8 (+781 C/T) +781 CT polymorphism compared to CC poly- 22.91 12,87 TT 0.00 0,121 (20.84–24.98) (7.02 – 18.71) morphism by around 11-fold higher. 19.01 9,96 CT 24 0.001 This study finds significantly higher IL-8 pro- (8.81–128) (5.57 – 17.01) duction in H. pylori-infection gastritis patients 51.02 11.27 39.75 CC 0.049 (42.81) (3.50) (0.16, 79.34) with -251 AT, -251 TT, +781 CT, and +781 CC polymorphisms compared with gastritis patients DISCUSSION without H. pylori-infection. These findings are Infection of H. pylori may be acquired during consistent with a previous study (9) that shows childhood and lasts for lifetime if not properly significant differences in IL-8 levels between treated (20). It is known to have a strong associa- patients with H. pylori-infection compared with tion with socioeconomic status and poor personal patients without H. pylori-infection in IL-8 -251 and community hygiene, especially environmen- TT, AT, and AA polymorphisms. Our study did tal tobacco exposure (21), which is in concord- not find significant differences in IL-8 production ance with this study. Continuous exposure of in patients with -251 AA and +781 TT polymor- H. pylori caused chronic gastritis, peptic ulcer phisms, which might be due to the limited num- disease, gastric cancer (4), and even a huge va- ber of patients enrolled in this category. riety of extragastric diseases (22). High burden Some limitations of this study should be noted. of the disease caused by this infection entices the This study protocol did not perform tissue biop- mandatory measure in eradication and prevention sy for histologic and culture examinations due strategy (23), especially in more susceptible pop- to resources constraint. Also, it did not perform ulation for developing a more severe disease (5). any follow-up of esophagogastroduodenoscopy Interleukin-8 is known as a potent inflammation (EGD) after the establishment of the diagno- mediator with eminent role in angiogenesis, tu- sis and eradication therapeutic interventions. mour growth, invasion and metastases process in Patients with known gastric cancer were also malignancies (24). Our study showed a higher val- excluded from this study to focus on the risk of

386 Supriatmo et al. IL-8 polymorphism H. pylori gastritis

polymorphisms in gastritis patients, therefore it In conclusion, the presence of IL-8 -251 AT and showed a limited number of patients especially in TT, and +781 CT and CC polymorphism causes IL-8 -251 AA polymorphism and maybe in +781 higher IL-8 production in gastritis children with TT polymorphism. H. pylori infection. Heterozygous gene polymor- To the best of our knowledge, this study is the phisms (-251 AT and +781 CT) may be associa- first study to report the association of IL-8 -251 ted with the risk of developing gastritis in chil- T/A and IL-8 +781 C/T polymorphisms with se- dren with H. pylori infection. rum IL-8 levels in children with H. pylori-positi- FUNDING ve gastritis. This study is also the first study that shows a significantly higher risk in patients with No specific funding was received for this study. IL-8 -251 heterozygous AT and +781 heterozygous CT developing H. pylori-infection gastritis. CONFLICTS OF INTEREST Competing interests: None to declare

REFERENCES 1. Torbenson M. Common types of gastritis. In: Zhang 10. Ramis IB, Vianna JS, Gonçalves CV, von Groll A, L, Chandan VS, Wu T-T, editors. Surgical Pathology Dellagostin OA, da Silva PEA. Polymorphisms of the of Non-neoplastic Gastrointestinal Diseases. Cham/ IL-6, IL-8 and IL-10 genes and the risk of gastric pat- Switzerland: Springer International Publishing, 2019: hology in patients infected with Helicobacter pylori. 121–35. https://doi.org/10.1007/978-3-030-15573- J Microbiol Immunol Infect 2017; 50:153–9. 5_6 (28 Feb 2020) 11. Abbas IS. Interleukin-8 251- A/T polymorphism rela- 2. Park YH, Kim N. Review of atrophic gastritis and in- ted to peptic ulcer disease in H. pylori infected pati- testinal metaplasia as a premalignant lesion of gastric ent. Journal of Contemporary Medical Sciences 2016; cancer. J Cancer Prev 2015; 20:25–40. 2:138–40. 3. Azer SA, Akhondi H. Gastritis. StatPearls Treasure 12. Rafrafi A, Chahed B, Kaabachi S, Kaabachi W, Ma- Island (FL): StatPearls Publishing; 2020. http://www. almi H, Hamzaoui K, Sassi FH. Association of IL-8 ncbi.nlm.nih.gov/books/NBK544250/ (28 February gene polymorphisms with non small cell lung cancer 2020) in Tunisia: a case control study. Hum Immunol 2013; 4. Kao C-Y, Sheu B-S, Wu J-J. Helicobacter pylori in- 74:1368–74. fection: an overview of bacterial virulence factors and 13. Zarafshani M-K, Shahmohammadi A, Vaisi-Raygani pathogenesis. Biomed J 2016; 39:14–23. A, Bashiri H, Yari K. Association of interleukin-8 5. Saes M, de Labio RW, Rasmussen LT, Payão SLM. polymorphism (+781 C/T) with the risk of ovarian Interleukin 8 (-251 T>A) polymorphism in children cancer. Meta Gene 2018; 16:165–9. and teenagers infected with Helicobacter pylori. J Ve- 14. Savage SA, Abnet CC, Mark SD, Qiao Y-L, Dong nom Anim Toxins Incl Trop Dis 2017; 23:23. Z-W, Dawsey SM, Taylor PR, Chanock SJ. Variants 6. Kamali-Sarvestani E, Bazargani A, Masoudian M, of the IL8 and IL8RB Genes and risk for gastric car- Lankarani K, Taghavi A-R, Saberifiroozi M. Associa- dia adenocarcinoma and esophageal squamous cell tion of H pylori cagA and vacA genotypes and IL-8 carcinoma. Cancer Epidemiol Biomarkers Prev 2004; gene polymorphisms with clinical outcome of infec- 13:2251–7. tion in Iranian patients with gastrointestinal diseases. 15. Sugimoto M, Yamaoka Y, Furuta T. Influence of in- World J Gastroenterol 2006; 12:5205–10. terleukin polymorphisms on development of gastric 7. Hull J, Ackerman H, Isles K, Usen S, Pinder M, cancer and peptic ulcer. World J Gastroenterol 2010; Thomson A, Kwiatkowski D. Unusual haplotypic 16:1188–200. structure of IL8, a susceptibility locus for a common 16. Fu JW, Wang KW, Qi ST. Role of IL-8 gene polymor- respiratory virus. Am J Hum Genet 2001; 69:413–9. phisms in glioma development in a Chinese populati- 8. Ohyauchi M, Imatani A, Yonechi M, Asano N, Miura on. Genet Mol Res 2016; 15. A, Iijima K, Koike T, Sekine H, Ohara S, Shimose- 17. Liu H, Mao P, Xie C, Xie W, Wang M, Jiang H. Asso- gawa T. The polymorphism interleukin 8 −251 A/T ciation between interleukin 8-251 T/A and +781 C/T influences the susceptibility of Helicobacter pylori polymorphisms and glioma risk. Diagn Pathol 2015; related gastric diseases in the Japanese population. 10:138–138. Gut 2005; 54:330–5. 18. Chen Y, Yang Y, Liu S, Zhu S, Jiang H, Ding J. Asso- 9. Chang YW, Oh CH, Kim J-W, Lee JW, Park MJ, Shim ciation between interleukin 8 −251 A/T and +781 C/T J-J, Lee CK, Jang J-Y, Dong SH, Kim HJ, Kim SS, polymorphisms and osteosarcoma risk in Chinese Kim B-H. Combination of Helicobacter pylori infec- population: a case–control study. Tumor Biol 2016; tion and the interleukin 8 –251 T > A polymorphism, 37:6191–6. but not the mannose-binding lectin 2 codon 54 G > A 19. Dibley MJ, Staehling N, Nieburg P, Trowbridge FL. polymorphism, might be a risk factor of gastric can- Interpretation of Z-score anthropometric indicators cer. BMC Cancer 2017; 17:388. derived from the international growth reference. Am J Clin Nutr 1987; 46:749–62.

387 Medicinski Glasnik, Volume 17, Number 2, August 2020

20. Malaty HM, Graham DY. Importance of childhood 25. Ferreira RM, Pinto-Ribeiro I, Wen X, Marcos-Pinto socioeconomic status on the current prevalence of R, Dinis-Ribeiro M, Carneiro F, Figueiredo C. Heli- Helicobacter pylori infection. Gut 1994; 35:742–5. cobacter pylori cagA promoter region sequences in- 21. Kim YI, Kim HW, Kim JH, Lee DW, Jung JH, Cho fluence CagA expression and interleukin 8 secretion. SH, Kim SW, Jung SJ, Cho BS, Youn JH. Association J Infect Dis 2016; 213:669–73. between environmental tobacco smoke and Helico- 26. Lee H, Su Y-L, Huang B-S, Hsieh F-T, Chang Y-H, bacter pylori infection in children and adolescents: Tzeng S-R, Hsu C-H, Huang P-T, Lou K-L, Wang the United States national health and nutritional Y-T, Chow L-P. Importance of the C-terminal histi- examination survey 1999-2000. Korean J Fam Med dine residues of Helicobacter pylori GroES for Toll- 2018; 8:565–9. like receptor 4 binding and interleukin-8 cytokine 22. Korwin J-D de, Ianiro G, Gibiino G, Gasbarrini A. production. Sci Rep 2016; 6:37367. Helicobacter pylori infection and extragastric disea- 27. Taguchi A, Ohmiya N, Shirai K, Mabuchi N, Itoh A, ses in 2017. Helicobacter 2017; 22:e12411. Hirooka Y, Niwa Y, Goto H. Interleukin-8 promoter 23. Du Y, Zhu H, Liu J, Li J, Chang X, Zhou L, Chen M, polymorphism increases the risk of atrophic gastritis Lu N, Li Z. Consensus on eradication of Helicobacter and gastric cancer in Japan. Cancer Epidemiol Bio- pylori and prevention and control of gastric cancer markers Prev 2005; 14:2487–93. in China (2019, Shanghai). J Gastroenterol Hepatol 28. Kamangar F, Abnet CC, Hutchinson AA, Newschaffer [Internet] 2019 https://doi.org/10.1111/jgh.14947 (01 CJ, Helzlsouer K, Shugart YY, Pietinen P, Dawsey March 2020) SM, Albanes D, Virtamo J, Taylor PR. Polymorp- 24. Dakal TC, Kala D, Dhiman G, Yadav V, Krokhotin A, hisms in inflammation-related genes and risk of ga- Dokholyan NV. Predicting the functional consequen- stric cancer (Finland). Cancer Causes Control 2006; ces of non-synonymous single nucleotide polymorp- 17:117–25. hisms in IL8 gene. Sci Rep 2017; 7:1–18.

388 ORIGINAL ARTICLE

Endothelin-1 level as a predictor of hepatopulmonary syndrome in liver cirrhosis

Darmadi1, Riska Habriel Ruslie2

1Department of Internal Medicine, Faculty of Medicine, Universitas Sumatera Utara; 2Department of Child Health, Faculty of Medicine, Universitas Prima Indonesia

ABSTRACT

Aim To determine the role of endothelin (ET)-1 in predicting hepatopulmonary syndrome (HPS) in patients with liver cirrhosis.

Methods A cross sectional study involving 80 liver cirrhosis patients aged 18 years or older was conducted in Adam Malik Gene- ral Hospital Medan, Indonesia between July 2017 and June 2018. HPS diagnosis was confirmed from the presence of liver cirrhosis, abnormal oxygenation, and intrapulmonary vascular dilatations (IPVD). ET-1 level was measured from serum sample using ELI- SA method. Patients with coexisting primary pulmonary pathology and intrinsic heart disease were active smokers, and those who declined to participate were excluded. Statistical analysis was conducted at 95% confidence interval. p<0.05 was considered si- Corresponding author: gnificant. Darmadi Results Majority the patients were male (56.3%) and had higher Department of Internal Medicine, educational background (62.5%). Mean age of the patients was Universitas Sumatera Utara 51.3 (SD=12.6) years. The prevalence of HPS was 21.2%. The pa- dr. Mansyur 5, Medan, Indonesia tients with HPS had higher ET-1 level compared to those without Phone: +62 618 211 045; HPS (p<0.001). The patients with hepatic encephalopathy had 2.917 times higher risk for suffering from HPS, while the patients E-mail: [email protected] with Child Pugh score A had lower risk (0.738 times) for having ORCID ID: https://orcid.org/0000-0001- HPS compared to subjects with Child Pugh score B and C. ET-1 5281-168X level >187.5 mg/L had sensitivity and specificity for predicting HPS in subjects with liver cirrhosis of 82.35% and 81.25%, respectively.

Conclusion ET-1 could be used as a promising marker for HPS in patients with liver cirrhosis. ET-1 level of >187.5 mg/l had a good accuracy in predicting HPS in liver cirrhosis patient. Original submission: 30 March 2020; Key words: endothelin, hepatopulmonary syndrome, liver cirrhosis Accepted: 11 June 2020 doi: 10.17392/1177-20

Med Glas (Zenica) 2020; 17(2): 389-394

389 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION in this population and further to suggest a therapeutic target based on pathogenesis. Liver cirrhosis is the end stage of liver disease and carries significant morbidity and mortality. Liver The aim of this study was to determine the role cirrhosis is the 13th most common cause of death of ET-1 in predicting HPS in patients with liver globally (1,2). In the United Kingdom, mortality cirrhosis. rate from liver cirrhosis as end stage liver disease PATIENTS AND METHODS between 1968 and 2011 was 5.4 per 100 00 person-years (3). In Adam Malik General Hospital, Patients and study design the 4-year prevalence of liver cirrhosis was 4% (4). The most common etiologies for liver cirrhosis This cross sectional study was conducted at Haji are chronic hepatitis viral infection, alcoholic Adam Malik General Hospital Medan, North Su- liver disease, and nonalcoholic fatty liver dise- matera, Indonesia. A total of 80 patients were en- ase (1,5). Patients with liver cirrhosis are at risk rolled in this study from July 2017 to June 2018. of several complications such as hepatocellular An inclusion criterion was liver cirrhosis patients carcinoma, hepatic encephalopathy, hepatore- aged 18 years or older. The HPS was confirmed nal syndrome, and hepatopulmonary syndrome from the presence of liver cirrhosis, abnormal (HPS) (1,2,4). oxygenation, and intrapulmonary vascular dilatations (IPVD). Hepatopulmonary syndrome (HPS) is marked by end stage liver failure, arterial hypoxemia (PaO2 Abnormal oxygenation was defined by elevated al- <70 mmHg or alveolar-arterial oxygen gradient veolar-arterial oxygen gradient >15 mmHg or >20 >20 mmHg), and intrapulmonary vascular dila- mmHg in patients older than 64 years while breat- tation without underlying cardiopulmonary dise- hing room air at rest in the sitting position. Exclusi- ase (4). The prevalence of HPS ranges from 4 to on criteria were patients with coexisting pulmonary 47% among liver cirrhosis patients with mortality pathology, coexisting intrinsic heart disease, active rate of 16% (4,6). Liver failure causes imbalance smokers, and those who declined to participate in between vasodilators production and clearance, the study. The selection of the patients was made which is the underlying etiology of HPS (4). En- using consecutive sampling method. dothelin (ET)-1 is one of vasodilators produced A written consent was obtained from each patient in HPS. It is consisted of 21 amino acids with a before enrolling to the study. free amino terminus and C-terminal carboxyl acid. The Ethics Committee of the School of Medicine ET-1 is produced in endothelial cells and work in of Universitas Sumatera Utara was obtained prior autocrine fashion to induce vasodilatation (7). investigation. The HPS as a complication of liver cirrhosis im- Methods pairs survival rate of the patients, thus prompt diagnosis and management are mandatory. Classi- The IPVD was diagnosed by contrast-enhanced cally, HPS is diagnosed based on clinical manife- transthoracic echocardiography. stations supported by echocardiography findings, Every patients was interviewed using questionnai- and blood gas analysis (6,8). Appropriate marker re as a guideline to obtain demographic data. Rou- for HPS is needed to assist an accurate diagnosis. tine blood count, hepatitis viral marker, bilirubin Several studies (9,10) reported that ET-1 was level, liver enzymes, renal function test, CRP, and associated with severity of liver disease and HPS. ET-1 measurements were done for each patient. To our knowledge, there is no study regarding ET-1 was measured from serum sample using ET-1 as a predictor of HPS in patients with liver Quantikine ELISA kit DET100 (R&D Systems, cirrhosis. We hypothesized that ET-1 plays a si- Inc., Minneapolis, USA). gnificant role in predicting HPS in patients with The Child-Pugh scoring system was used to de- liver cirrhosis in Indonesian population, particu- termine the severity of liver cirrhosis. It consists larly in Medan. Due to high morbidity and mor- of several parameters. Each parameter has a ran- tality rates due to HPS in liver cirrhosis, ET-1 is ge of score from 1 to 3 (Table 1). Based on the expected to become a promising marker for HPS total score, a patient is categorized into grade A

390 Darmadi et al. Endothelin-1 and hepatopulmonary syndrome

Table 1. The Child-Pugh score system* Table 2. Demographic, social and clinical characteristics of Description/value according to the score the patients Variable 1 2 3 Characteristic No (%) of patients Encephalopathy None Minimal Advanced (coma) Gender Ascites Absent Controlled Refractory Males 45 (56.3) Bilirubin (µmol/L) <34 34-51 >51 Female 35 (43.8) Albumin (g/L) >35 28-35 <28 Mean age (SD) (years) 51.3 (12.60) Prothrombin (second) <4 4-6 >6 Education status *grade A (total score of 5-6), grade B (total score of 7-9), grade C Primary 8 (10.0) (total score of 10-15) Secondary (22 (27.5) Higher secondary and above 50 (62.5) (total score of 5-6), B (total score of 7-9), and C Ethnicity (total score of 10-15). The higher the score, the Bataknese 45 (56.3) Javanese 28 (35.0) more severe the disease becomes (11). Other 7 (8.8) Viral marker Statistical analysis HBsAg (+) 32 (40.0) Anti HCV (+) 10 (12.5) Normality of continuous data distribution was Hepatopulmonary syndrome (HPS) determined using Kolmogorov-Smirnov test. χ2 Absence 63 (78.8) Presence 17 (21.2) test was used to analyse the association between Child Pugh score gender, hepatic encephalopathy, and Child Pugh A 15 (18.8) score and HPS. To assess the difference of labora- B 17 (21.3) tory parameters in patients with and without HPS, C 48 (60.0) SD, standard deviation; HBsAG, hepatitis B surface antigen; HCV, we used independent T test with Mann-Whitney hepatitis C virus; test as an alternative. Receiver operating curve (ROC) analysis was conducted to determine the the most frequent. The prevalence of HPS was accuracy of ET-1 to predict HPS in patients with 21.2%. Sixty-percent patients had Child Pugh liver cirrhosis. Statistical analysis was conducted score C (Table 2). at 95% confidence interval, using p<0.05. From laboratory parameters, there was no si- RESULTS gnificant difference between patients with and A total of 80 patients were enrolled in this stu- without HPS except for ET-1 level. The patients dy, of whom majority were male, 45 (56.3%) and with HPS had higher ET-1 level compared to tho- had higher educational background, 50 (62.5%). se without HPS (p<0.001) (Table 3). Mean age of patients was 51.3 (SD=12.6) years. There were significant relationships between More than a half of patients, 42 (52.5 %) had po- hepatic encephalopathy, Child Pugh score, and sitive hepatitis viral marker with hepatitis B as HPS. Patients with hepatic encephalopathy had

Table 3. Difference in laboratory parameters between the patients with and without hepatopulmonary syndrome (HPS) Variables Reference values Absence of HPS (n = 63) HPS (n = 17) p Mean haemoglobin (SD) (g/dL) 12-17 10.2 (2.62) 9.7 (2.22) 0.517 Mean haematocrit (SD) (%) 36-51 30.6 (7.63) 28.9 (6.32) 0.414 Median WBC (min-max) (cells/µL) 4 500-11 000 7 170 (1 830-29 000) 7 470 (1 866-35 190) 0.711 Median platelet (min-max) (cells/µL) 150 000-350 000 142 000 (15 000-509 000) 164 000 (12 000 – 654 000) 0.353 Mean albumin (SD) (g/dL) 3.5-5.4 2.5 (0.6) 2.3 (0.4) 0.060 Median INR (min-max) 2.0-3.0 1.3 (0.8 – 3.7) 1.1 (0.9 – 1.6) 0.193 Median total bilirubin (min-max) (mg/dL) 0.3-1.2 2.1 (0.2-13.6) 3.3 (0.6 – 21.9) 0.104 Median direct bilirubin (min-max) (mg/dL) 0.0-0.3 0.9 (0.1 – 9.6) 1.5 (0.2 – 15.2) 0.189 Median AST (min-max) (U/L) 0-35 57 (15 – 216) 49 (15 – 377) 0.837 Median ALT (min-max) (U/L) 0-35 43 (8 – 246) 40 (11 – 146) 0.356 Median ALP (min-max) (U/L) 36-150 117 (39 – 331) 122 (43 – 178) 0.791 Median GGT (min-max) (U/L) 8-78 78 (13 – 346) 83 (13 -628) 0.742 Median RBG (min-max) (mg/dL) <140 105 (41 – 295) 100 (65 – 163) 0.298 Median ureum (min-max) (mg/dL) 8-20 30 (11 – 263) 40 (17 – 146) 0.270 Median creatinine (min-max) (mg/dL) 0.7-1.3 0.8 (0,4 – 7) 1.1 (0.5 – 3.6) 0.107 Median CRP (min-max) (mg/dL) <0.5 2.2 (0.1 – 8.4) 2.8 (0.8 – 14) 0.214 Median ET-1 level (min-max) (mg/L) N/A 86 (60 – 120) 320 (150 – 400) <0.001 ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, c-reactive protein; ET-1, endothelin-1; GGT, gamma-glutamyl transpeptidase (GGT); INR, international normalized ratio; RBG, random blood glucose; WBC, white blood cell;

391 Medicinski Glasnik, Volume 17, Number 2, August 2020

2.917 times higher risk for suffering from HPS, DISCUSSION while patients with Child Pugh score A had the The prevalence of liver cirrhosis increases signi- lower risk (0.738 times) for having HPS compa- ficantly probably due to alcohol consumption, red to patients with Child Pugh score B and C. obesity, and viral infection (3). Males were affec- There was no association between gender and ted more commonly compared to females with a HPS (Table 4). ratio of 2.1:1 (2,5). Most patients were older than Table 4. Associations between gender, hepatic encephalopathy, 50 years and had obesity and/or type 2 diabetes and Child Pugh score and hepatopulmonary syndrome (HPS) mellitus as comorbidities (1,5). In patients with No (%) of patients PR Variable p liver cirrhosis, the liver function deteriorates si- With HPS Without HPS Total (95%CI) gnificantly. The process involves changes in me- Gender 1.111 Male 10 (22.2) 35 (77.8) 42 (100.0) 0.810 tabolic enzymes and transport proteins (12). The (0.47-2.62) Female 7 (20.0) 28 (80.0) 28 (100.0) enzymes are downregulated and reduced functi- Hepatic encephalopathy 2.917 onally (12). Transport proteins suffer molecular Yes 5 (50.0) 5 (50.0) 10 (100.0) 0.031 (1.30-6.53) No 12 (17.1) 58 (82.9) 70 (100.0) changes, and as a result, energy expenditure and Child Pugh Score 0.738 wasting are disturbed (12). In our study involving Class A 0 (0.0) 15 (100.0) 15 (100.0) 0.032 (0.64-0.85) patients with liver cirrhosis, male gender was Class B+C 17 (26.2) 48 (73.8) 65 (100.0) CI, confidence interval; HPS, hepatopulmonary syndrome; PR, dominant (56.3%). Hepatitis viral infection was prevalence ratio the etiology of liver cirrhosis in 52.5% patients. Mean age of patients was 51.3 (SD 12.60) years. ROC analysis resulted in area under the curve of These findings are in accordance with the data 0.908 for ET-1 level >187.5 mg/L in predicting from the previous reports (1-3, 5). HPS (Figure 1). Further analysis showed that The association between chronic liver disease, ET-1 level >187.5 mg/L had sensitivity and sen- cyanosis, and clubbing finger was introduced sitivity for predicting HPS in patients with liver firstly by Fluckinger in 1884 (13). In 1966, Ber- cirrhosis of 82.35% and 81.25%, respectively thelot et al. reported pulmonary arterial dilatati- (Table 5). on in patients with liver cirrhosis (4). The HPS terminology was firstly introduced by Kennedy and Knudson in 1977 after reporting a patient with liver cirrhosis and hypoxemia (4,13). The prevalence of HPS ranges from 4 to 47% among liver cirrhosis patients (4,6). The pathogenesis of HPS remains uncertain. Liver damage causes an increase in synthesis and decrease in metabolism of vasodilator substances (13,14). On the other hand, vasoconstrictor substances production is decreased and lung’s sensitivity toward vasocon- strictors is decreased, which ends with vasodilatation in pulmonary capillaries and precapillaries (6,15,16). One of vasoactive substances is ET-1. ET-1 induces vasodilatation by increasing nitric oxide production (4,6,15,16). Vasodilatation leads to ventilation-perfusion mismatch and decreased oxygen saturation from longer distance for Figure 1. Receiver operating curve (ROC) for endothelin-1 oxygen molecules to reach red blood cells in the (ET)-1 in predicting hepatopulmonary syndrome (HPS) in liver centre of the pulmonary capillaries (4,6,15,16). cirrhosis In our study, the prevalence of HPS was 21.2%

Table 5. Accuracy of endothelin-1 level in predicting hepatopulmonary syndrome (HPS) in liver cirrhosis Cut off Sensitivity Specificity PPV NPV PLR NLR Accuracy ET-1 (>187.5 mg/l) 82.35% 81.25% 53.84% 94.44% 4.32 0.22 90.8% ET-1, endothelin-1; NLR, negative likelihood ratio; NPV, negative predictive value; PLR, positive likelihood ratio; PPV, positive predictive value

392 Darmadi et al. Endothelin-1 and hepatopulmonary syndrome

which is within the range of prevalence reported Abdel-Razik et al. conducted a study in Egypt in the literature (4,6). enrolling 42 liver cirrhosis patients with hepa- Reportedly, the mortality rate for HPS is 41% torenal syndrome. They found that ET-1/NO ra- (6). The presence of HPS significantly worsens tio can predict therapeutic response toward ter- the outcome of patients with liver cirrhosis and lipressin and albumin; early reduction of ET-1/ hampers their quality of life (6,13). Liver tran- NO ratio gave better response toward therapy in splantation is the only therapeutic option to over- their study (18). A prospective study by Koch et come HPS (6,13,15). Elevated ET-1 level also al. in the USA tried to determine the association induces proliferation, hypertrophy, and synthesis between ET-1 level and intrapulmonary vasodi- of extracellular matrix (13). Grebes et al. repor- latation in patients with liver cirrhosis, and found ted that ET-1 level is higher in patients with liver that ET-1 level in patients with intrapulmonary cirrhosis compared to normal one; their study in- vasodilatation is higher than those without vaso- volving 18 patients with liver cirrhosis and eight dilatation (9.1 pg/mL versus 2.1 pg/mL) (15). healthy patients shows a significant difference To date, there is no study determining the role of in arterial ET-1 level (17.8 pg/mL versus 9.2 pg/ ET-1 as a predictor of HPS in patients with liver mL) (14). Another study by Alam et al. (9) confir- cirrhosis. We tried to answer the question and fo- med this result: hepatic ET-1 level was measured und that ET-1 is a promising predictor for HPS and the result was higher in patients with liver in that population. Our results showed that ET-1 cirrhosis compared to controls. This condition level had sensitivity and specificity of 82.35% was still observed even if patients had undergone and 81.25%, respectively in predicting HPS with transjugular intrahepatic portosystemic shunt (9). a cut off value of >187.5 mg/L. The result of Kamath et al. study was in concert This study has a few limitations. The low pre- with this result; ET-1 levels in hepatic tissue, pe- valence of liver cirrhosis in our Centre makes it ripheral vein, hepatic vein, and portal vein were difficult to gather more patients for the study. In higher in patients with liver cirrhosis (17). addition, the proportion of patients with and wit- A study by Kaffarnik et al. reported that plasma hout HPS in this study is not balanced. A larger ET-1 level is positively correlated with the seve- study involving more patients and several centres rity of liver failure; the more severe the liver da- is mandatory to confirm the result of this study. mage, the higher the serum ET-1 level (10). This In conclusion, the patients with liver cirrhosis and is in line with the result from Alam et al., who HPS had a higher ET-1 level compared to those also found that ET-1 level has a positive corre- without HPS. ET-1 could be used as a promising lation with liver disease severity based on Child marker for HPS in patients with liver cirrhosis. Pugh score (9). Our study is in accordance with ET-1 level of >187.5 mg/l had a good accuracy in both studies, all patients were suffering from li- predicting HPS in liver cirrhosis patients. ver cirrhosis and we found that ET-1 level was significantly higher in patients with HPS compared FUNDING to those without HPS. But there are two studies No specific funding was received for this study that contradict this result. Koch et al. and Kamath et al. reported that ET-1 level does not correlate TRANSPARENCY DECLARATION with the severity of liver dysfunction (15,10). Conflicts of interest: None to declare.

REFERENCES 1. Li B, Zhang C, Zhan Y. Nonalcoholic fatty liver di- 2. Patasik YZ, Waleleng BJ, Wantania F. Profil pasien sease cirrhosis: a review of its epidemiology, risk sirosis hati yang dirawat inap di RSUP Prof. Dr. R. factors, clinical presentation, diagnosis, manage- D. Kandou Manado periode Agustus 2012-Agustus ment, and prognosis. Can J Gastroenterol Hepatol. 2014 (Profile of patients with liver cirrhosis admitted 2018; 2018:2784537. in Prof. Dr. R. D. Kandou general hospital Manado between August 2012-August 2014) [in Indonesian]. eCl 2015; 3:342-7.

393 Medicinski Glasnik, Volume 17, Number 2, August 2020

3. Ratib S, West J, Fleming KM. Liver cirrhosis in 11. Durand F, Valla D. Assessment of the prognosis England-an observational study: are we measuring of cirrhosis: Child-Pugh versus MELD. J Hepatol its burden occurrence correctly? BMJ Open 2017; 2005; 42:S100-7. 7:e013752. 12. Koch DG, Bogatkevich G, Ramshesh V, Lemasters 4. Elvira D. Diagnosis dan tatalaksana hepatopulmo- JJ, Uflacker R, Reuben A. Elevated levels of endo- nary syndrome (Diagnosis and management of hepa- thelin-1 in hepatic venous blood are associated with topulmonary syndrome [in Indonesian] MKA 2015; intrapulmonary vasodilatation in humans. Dig Dis 38:57-65. Sci 2012; 57:516-23. 5. Kalista KF, Lesmana CRA, Sulaiman AS, Gani RA, 13. He J, Yi B, Chen Y, Huang Q, Wang H, Lu K, Fu W. Hasan I. Clinical profile of cirrhotic patient with The ET-1-mediated carbonylation and degradation esophageal varices who undergone band ligation in of ANXA1 induce inflammatory phenotype and pro- Cipto Mangunkusumo Hospital. Jurnal Penyakit Da- liferation of pulmonary artery smooth muscle cells lam Indonesia 2019; 6:36-41. in HPS. PLoS One 2017; 12:e0175443. 6. Soulaidopoulos S, Cholongitas E, Giannakoulas 14. Grebes AL, Moller S, Gullberg V, Henriksen JH. En- G, Vlachou M, Goulis I. Review article: update on dothelin-1 and -3 plasma concentrations in patients current and emergent data on hepatopulmonary syn- with cirrhosis: role of splanchnic and renal passage drome. World J Gastroenterol 2018; 24:1285-98. and liver function. Hepatology 1995; 21:735-9. 7. Davenport AP, Hyndman KA, Dhaun N, Southan C, 15. Koch DG, Bogatkevich G, Ramshesh V, Lemasters Kohan DE, Pollock JS, Pollock DM, Webb DJ, Ma- JJ, Uflacker R, Reuben A. Elevated levels of endo- guire JJ . Endothelin. Pharmacol Rev 2016; 68:357- thelin-1 in hepatic venous blood are associated with 418. intrapulmonary vasodilatation in humans. Dig Dis 8. Voiosu AM, Daha IC, Voiosu TA, Mateescu BR, Dan Sci 2012; 57:516-23. GA, Baicus CR, Voiosu MR, Diculescu MM. Pre- 16. Nishiyama SK, Zhao J, Wray DW, Richardson RS. valence and impact on survival of hepatopulmonary Vascular function and endothelin-1: tipping the ba- syndrome and cirrhotic cardiomyopathy in a cohort lance between vasodilatation and vasoconstriction. J of cirrhotic patients. Liver Int 2015; 35:2547-55. Appl Physiol 2017; 122:354-60. 9. Alam I, Bass NM, Bacchetti P, Gee L, Rockey DC. 17. Kamath PS, Carpenter HA, Lloyd RV, McKusick Hepatic tissue endothelin-1 levels in chronic liver MA, Steers JL, Nagorney DM, Miller VM. Hepatic disease correlate with disease severity and ascites. localization of entothelin-1 in patients with idiopat- Am J Gastroenterol 2000; 95:199-203. hic portal hypertension and cirrhosis of the liver. Li- 10. Kaffarnik MF, Ahmadi N, Lock JF, Wuensch T, ver Transpl 2000; 6:596-602. Pratschke J, Stockmann M, Malinowski M. Corre- 18. Abdel-Razik A, Mousa N, Abdelsalam M, Abdelwa- lation between plasma endothelin-1 levels and se- hab A, Tawfik M, Tawfik AM, Hasan AS, Elhelaly verity of septic liver failure quantified by maximal R, El-Wakeel N, Eldars W. Endothelin-1/nitric oxide liver function capacity (LiMAx test). A prospective ratio as a predictive factor of response to therapy study. PLoS One 2017; 12:e0178237. with terlipressin and albumin in patients with type-1 hepatorenal syndrome. Front Pharmacol 2020; 11:9.

394 ORIGINAL ARTICLE

The effect of roselle flower petals extract (Hibiscus sabdariffa Linn.) on reducing inflammation in dextran sodium sulfate- induced colitis

Masrul Lubis1, Gontar Alamsyah Siregar1, Hakim Bangun2, Syafruddin Ilyas3

1Department of Internal Medicine, School of Medicine, 2Department of Pharmaceutical Technology, Faculty of Pharmacy, 3Department of Biology, Faculty of Mathematics and Natural Sciences; Universitas Sumatera Utara, Medan, Indonesia

ABSTRACT

Aim To determine the effect of roselle (Hibiscus sabdariffa) petals in dextran sodium sulfate (DSS)-induced colitis model by measuring pro-inflammatory cytokines expressions (IL-6, and TNF-α), anti-inflammatory cytokine expression (IL-10) and histological colitis inflammation score (HCIS).

Methods This study was conducted in two phases. For the first phase, five DSS-induced colitis mice were sacrificed (group 1) and compared with six healthy mice (group 2) after five-cycle induction (70 days). For the second phase, roselle-treated DSS-induced colitis mice were sacrificed on day 7, 14, 21, and 28 after induction and compared with mesalazine-treated DSS-induced colitis α Corresponding author: mice. Expressions of IL-6, TNF- , and IL-10 were determined by immunohistochemistry and HCIS were assessed by two experien- Masrul Lubis ced pathologists. Department of Internal Medicine, School of Medicine, Results The expressions of IL-6, TNF-α, and IL-10, and HCIS in Universitas Sumatera Utara DSS-induced colitis mice were increased compared with control. The expressions of IL-6, TNF-α were significantly higher in rose- dr. Mansyur 5, Medan, Indonesia lle-treated group on day 7 and 14, but not on day 21 and 28, whe- Phone: +62 618 211 045; reas, the expression of IL-10 was significantly lower only on day E-mail: [email protected] 7 compared with mesalazine-treated group. The inflammation was ORCID ID: https://orcid.org/0000-0002- higher in roselle-treated group assessed by using HCIS. Compared 8391-2341 to day 0, the reduction of HCIS was significant in roselle-treated and mesalazine-treated groups.

Conclusion Roselle flower petal can attenuate the inflammation in DSS-induced colitis in mice. The extract of roselle can be given Original submission: as an adjuvant therapy to the first-line therapy to enhance anti-in- 24 October 2019; flammatory effect by increasing expression of anti-inflammatory cytokines and decreasing pro-inflammatory cytokines. Revised submission: 06 January 2020; Key words: anti-inflammatory, cytokines, immunohistochemistry, Accepted: inflammatory bowel disease, plant extract 13 February 2019 doi: 10.17392/1095-20

Med Glas (Zenica) 2020; 17(2): 395-401

395 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION model of IBD with IL-10-deficient has shown a potential role of IL-10 in maintaining intesti- Inflammatory Bowel Disease (IBD) is a chronic nal immune regulation, which is very important or remitting/relapsing inflammation of the inte- in pathogenesis of IBD (7). Antigen-presenting stinal with two major groups: Crohn's Disease cell produces IL-10 that regulates homeostatic (CD), which can affect any region in digestive T-cell responses to commensal bacteria (8)me- tract from mouth to anus characterized by tran- diated by T-helper (Th. smural granulomatous inflammation, and ulcerative colitis (UC), which continuously affects Mesalazine [5-aminosalicylate (5-ASA)] is the colonic mucosa proximal from rectum and forms first-line therapy for IBD, and it also used for ulcers (1). The etiology of IBD is still unknown; remission maintenance in patients with UC; it is probably it is the result of excessive immune res- relatively safe and generally well tolerated altho- ponse to gastrointestinal microbiota that was sti- ugh its efficacy on CD is still limited (9). Other mulated by increased effector T cell activity and/ conventional treatments are corticosteroids, or decreased regulator T cell activity, changes in immunomodulatory drugs, biology agents, small the composition of gastrointestinal microorgani- molecular therapy as well as immunosuppressant sms and/or damage of the epithelial barrier (2) (10). Renewable treatments are still being studied including both innate and adaptive immunity. such as mesenchymal stem cell transplantation Dysregulation of the homeostasis between inte- and faecal microbiota transplantation (11). stinal immune systems and microbiota has been Roselle (Hibiscus sabdariffa Linn.) is a plant shown to be associated with the development of belonging to the family Malvaceae, which is inflammatory bowel diseases (IBD. ideal for developing countries; it can grow up to The process of immune response in IBD is not 2.4 meters with red or white flower petals, and it uniform. The inflammatory process in CD is is a good nutritious source of minerals and vita- mainly mediated by T helper (Th) type 1 or Th mins, especially in the flower petals, which also type 17 cells, whereas for UC it is generally me- have numerous health benefits (12,13). A study diated by Th type 2 cells. However, this model by Anokwuru et al. stated that roselle flower cannot be used generally because of the presen- petals contained several phenolic compounds, ce of phases of the disease (acute or chronic), with methanol extract giving the highest pheno- innate immunity pathways involvement, epithe- lic contents and antioxidant activity against free lial cells factor, and the use of anti-inflamma- radicals (14). Phenolic and flavonoid compo- tory drugs (3). The most important concept of unds are compounds that can be used as immu- immune response in the pathogenesis of IBD is nostimulator, as stated by Chiang et al. (15). the involvement of anti-inflammatory cytokines Research on the immunomodulatory activity of such as interleukin (IL)-10 and pro-inflamma- roselle flower petals using the hemagglutination tory cytokines, such as IL-6 and tumour ne- test method was carried out in vivo by Fakeye et crosis factor (TNF)-α. IL-10, which is known al.; researches also measured the concentration as cytokine synthesis inhibitory factor (CSIF), of interleukin-10 (IL-10) as an anti-inflamma- inhibits the release of pro-inflammatory cyto- tory cytokine (16). The results showed that rose- kines, hence reducing inflammatory response lle flower petals were able to increase the immu- in the mucosa (4). Pro-inflammatory cytokines, ne system. This is due to an increase in IL-10 such as IL-6 and TNF-α, were also known for production so that it can suppress the synthesis their roles in the pathogenesis of colitis to in- of other pro-inflammatory cytokines and affect duce Th17 to produce other pro-inflammatory B lymphocyte cells to produce antibodies (17). cytokines (5)evolution and, ultimately, the reso- The aim of this study was to determine the effect lution of these forms of inflammation. Studies of roselle (Hibiscus sabdariffa) flower petals in over the last two decades now provide a detailed dextran sodium sulfate (DSS)-induced colitis (but not yet complete. model by measuring pro-inflammatory cytoki- A study of Barman et al. reported that IL-10 nes expressions (IL-6, and TNF-α) and anti-in- had anti-proliferation effect of T helper cells flammatory cytokine expression (IL-10). on human intestinal lamina propria (6). Mice

396 Lubis et al. Hibiscus effect on inflammatory colitis

MATERIALS AND METHODS Methods Chemical induction of colitis. The experimental Animals and study design colitis was chemically induced with dextran sodi- This study was conducted in the pharmaceutical um sulfate (DSS) (MP Biomedicals LLC). Mice laboratory, biological laboratory, and anatomical were given 5% DSS in 5 cycles. Each cycle was pathology laboratory of the Universitas Suma- given for 7 days and then continued by distilled tera Utara, Medan, Indonesia during the period water for 7 days. After 5 cycles (70 days) the mice between June and September 2019. Fifty five were sacrificed or treated by roselle flower petals male healthy mice (Mus musculus) (Charles Ri- extract or mesalazine as gold standard therapy. ver, Inc., Kanagawa, Japan), 6-8 weeks old and Histopathological analysis. The colon speci- weighing 30 grams, were housed at 20 °C – 25 °C mens were paraffin-embedded, subsequently cut, with controlled 12 hour light/dark cycle. Labo- and they were stained with hematoxylin and eosin ratory-standardized cages were used to keep the (H&E) using standard technique. A histological animals with ad libitum access to food and water. score was assessed by two experienced pathologi- All animal procedures were based on the Helsin- sts from the Anatomy Pathology Department of the ki Declaration. School of Medicine, Universitas Sumatera Utara. The experiment was approved by the Institutio- Severity of inflammation was scored as follows: nal Ethics Committee of the Universitas Sumate- 0 - rare inflammatory cells in the lamina propria, ra Utara, Medan, Indonesia. 1 - increased number of granulocytes in the lamina propria, 2 - confluence of inflammatory cells exten- This study was conducted in two phases. ding into the submucosa, 3 - transmural extension In phase one, 11 mice were divided into 2 grou- of the inflammatory infiltrate. Crypt damage was ps. Group 1 consisted of six mice without DSS scored as follows: 0 - intact crypts, 1 - loss of one- induction and group 2 consisted of five DSS-in- third basal, 2 - loss of two-thirds basal, 3 - entire duced colitis mice. Both groups were sacrificed crypt loss, 4 - change of epithelial surface with ero- after group 2 completed five-cycle (70 days) sion, 5 - confluent erosion. Ulceration was scored induction. Colon of each mice was excised and as follows: 0 - absence of ulcer, 1 - 1 or 2 foci of ul- samples were fixated in phosphate buffered sa- cerations, 2 - 3 or 4 foci of ulcerations, 3 - confluent line (PBS) 10% formalin for the histopathologi- or extensive ulceration. Values were added to give cal analysis and for the examination of anti-in- a maximal histological score of 11 (19). flammatory (IL-10) and pro-inflammatory (IL-6, Immunohistochemical analysis of IL-6, IL- TNF-α) cytokines using immunohistochemistry. 10, and TNF-α. The chosen paraffin-embedded In phase 2, 44 mice were divided into two gro- slides were deparaffinized, rehydrated, and heat- ups (group 3 and group 4). Group 3 (24 mice) ed on microwave with 0.01 M citrate buffer (pH was DSS-induced colitis mice given roselle 6.0) for 30 minutes. Endogen peroxidase activity flower petals extract (300 mg/kg body weight/ was blocked with 3% hydrogen peroxide for 10 day) (16) as a therapy after the induction of co- minutes, and then washed with sulfate salt buffer. litis. One kg of roselle petals were macerated The specimens were incubated overnight in 4 °C, with 4.5 litres of 96% ethanol for 7 × 24 hours, then immune stained with primary antibody (rab- then concentrated with a rotary evaporator un- bit polyclonal IgG to bind the mice IL-6, IL-10, til concentrated extract was obtained. Group 4 and TNF-α cytokines) (Wuhan Fine Biotech Co., (20 mice) was DSS-induced colitis mice given Ltd., China) in concentration of 1 mg/mL diluted mesalazine (7.8 mg/day) (18) as gold standard by 1:600. Primary antibody was detected by av- therapy after the induction of colitis. Six mice idin-biotin peroxidase solution (ScyTek Labora- from group 3 and five mice from group 4 were tories, Inc., USA), and a signal was visualized by sacrificed on day 7, 14, 21, and 28. Histopat- using diaminobenzidine (ScyTek Laboratories, hological analysis and immunohistochemistry Inc., USA). The slides were then counterstained of IL-6, IL-10, and TNF-α were conducted for with hematoxylin and assessed by two blinded these mice as described above. experienced pathologists from the Anatomy Pa- thology Department of the School of Medicine,

397 Medicinski Glasnik, Volume 17, Number 2, August 2020

Universitas Sumatera Utara. The slides were cat- Table 1. Comparison of inflammatory cytokines expressions egorized as (0) for 0-15%, (1) for 15-25%, (2) on immunohistochemistry and histological colitis inflammatory score (HCIS) between healthy (group 1) and colitis mice for 26-50% and (3) for 51-100% stained cells for (group 2) groups TNF-α, IL-6 and IL-10. Scale 0 and 1 was cat- No (%) of mice Variable egorized as (-), 2 and 3 as (+) (20). Group 1 Group 2 p Immunohistochemical categorization of slides* Statistical analysis IL-6 0.002† - 5 (100) 0 (0) The results of IL-10, IL-6, and TNF-α on phase 1 + 0 (0) 6 (100) (group 1 and 2) were analysed by using Fisher’s TNF-α 0.002† - 5 (100) 0 (0) exact probability test and HCIS on phase 1 were + 0 (0) 6 (100) analysed by using Mann-Whitney U test. The IL-10 0.015† results of IL-10, IL-6, and TNF-α on phase 2 - 4 (80) 0 (0) + 1 (20) 6 (100) (group 3 and 4) were analysed by using Fisher’s HCIS 0 (0 – 1) 7.5 (6 – 9) 0.004‡ exact probability test, and HCIS on phase 2 were *(-): (0) for 0-15% and (1) for 15-25%; (+): (2) for 26-50% and (3) analysed by using independent t-test if the distri- for 51-100% stained cells for TNF-α, IL-6 and IL-10; †Fisher’s exact probability test; ‡Mann-Whitney U test bution was normal, or Mann-Whitney U test if the distribution was not normal to compare the groups inflammation was also significantly higher in in each week, and also using one-way ANOVA test Group 2 comparing to Group 1. or Kruskal-Wallis test. Bonferroni post-hoc test Phase 2: Effect of roselle compared to mesala- was conducted if there were significant differen- zine on inflammatory cytokines and histopat- ces using one-way ANOVA test. Mann-Whitney U hology of colon in DSS-induced mice colitis test comparing day 0 with each day 7, 14, 21, and The expression of pro-inflammatory (IL-6 and 28 was conducted if there were significant diffe- TNF-α) and anti-inflammatory (IL-10) cytoki- rences using Kruskal-Wallis test. Differences were nes by immunohistochemistry and the degree of considered statistically significant when p<0.05. inflammation by HCIS from six roselle-treated RESULTS DSS-induced colitis mice were compared with five mesalazine-treated DSS-induced colitis mice on Phase 1: Effect of DSS-induced colitis on in- day 7, 14, 21, and 28 (Table 2). There were signi- flammatory cytokines and histopathology of ficant differences in the number of mice in which mice colon pro-inflammatory cytokines (IL-6 and TNF-α) The expression of pro-inflammatory (IL-6 and were expressed in roselle-treated group comparing TNF-α) and anti-inflammatory (IL-10) cytokines to mesalazine-treated group on day 7 (p=0.015 by immunohistochemistry and the degree of in- and p=0.015, respectively) and 14 (p=0.015 and flammation by HCIS in five healthy mice were p=0.015, respectively), but not on day 21 (p=1.000 compared with six DSS-induced colitis mice (Ta- and p=0.182, respectively) and 28 (p=1.000 and ble 1). Pro-inflammatory cytokines were expre- p=1.000, respectively). Anti-inflammatory cytoki- ssed in all six mice of Group 2 (DSS-induced co- ne expression was significantly more frequent in litis group) compared to Group 1 (control) (none) the mesalazine-treated group comparing to roselle- (p=0.002). The expression of anti-inflammatory treated group on 7 (p=0.048), but not on day 14, 21 cytokines in Group 2 was also found in all mice and 28 (p=0.455, p=1.000, and p=1.000, respecti- compared to Group 1 (p=0.015). The degree of vely) (Table 2). Table 2. Comparison of inflammatory cytokine expression on immunohistochemistry between roselle group and mesalazine group* No (%) of mice IL-6 TNF-α IL-10 Day Roselle group Mesalazine group Roselle group Mesalazine group Roselle group Mesalazine group (n=6) (n=5) p (n=6) (n=5) p (n=6) (n=5) p + - + - + - + - + - + - 7 5 1 0 5 0.015 6 0 1 4 0.015 1 5 4 1 0.048 14 5 1 0 5 0.015 5 1 0 5 0.015 4 2 5 0 0.455 21 1 5 0 5 1.000 3 3 0 5 0.182 5 1 5 0 1.000 28 1 5 0 5 1.000 1 5 0 5 1.000 5 1 5 0 1.000 *(-): (0) for 0-15% and (1) for 15-25%; (+): (2) for 26-50% and (3) for 51-100% stained cells for TNF-α, IL-6 and IL-10;

398 Lubis et al. Hibiscus effect on inflammatory colitis

DISCUSSION Inflammatory bowel disease is caused by disturbance in the balance of pro-inflammatory and anti-inflammatory cytokines in colonic mucosa that leads to disease states (4). Some of pro-inflammatory cytokines that contribute to the pathogenesis of IBD are IL-6 and TNF-α (21). The anti-inflammatory cytokine that can help maintain the balance of inflammatory cytokines in colonic Figure 1. Histopathology of large intestine of mice. A) no IL-6 expression in control group or group without DSS induction; mucosa is IL-10 that can reduce the synthesis of B) IL-6 (+) expression in large intestinal of mice that had not pro-inflammatory cytokines by suppressing effec- been given intervention; C) IL-10 (–) in large intestine of mice tor Th1/Th17 cells, NK cells, and also macropha- that had been given roselle flower petals extract on day 7; D) expression of IL-10 (+) in large intestines of mice that had ges (7). Results from this study have shown that been given roselle flower petals extract on day 28 there is a significant increase in either pro-inflammatory or anti-inflammatory cytokines on Histopathology slides were examined and HCIS DSS-induced colitis. A conversion from acute to were assessed for each slide (Figure 1). The HCI- chronic colitis in this model has not been com- Ss in roselle-treated group were also significantly pleted yet. Chronic colitis is characterized by the higher than in mesalazine-treated group on day increased IL-6, TNF-α, but decreased expression 7, 14, 21, and 28 (p<0.001, p<0.001, p=0.005, of IL-10 (19). It is known that the mechanism and p=0.009, respectively). Compared to day 0, through which DSS induces inflammation of co- significant decrease of HCIS was found on either lonic mucosa is still not completely understood, roselle-treated and mesalazine-treated group some studies state that DSS disrupts intestinal (p<0.001 and p<0.001, respectively). From post- barrier function (20,21)mannan-binding lectin hoc analysis, HCIS from roselle-treated was si- (MBL, and other state that DSS is associated with gnificantly decreased on day 21 and 28 (p<0.001 medium-chain-length fatty acids (MCFAs) in the and p<0.001, respectively), and HCIS from me- lumen of colon prior induction of colitis (17). salazine-treated was significantly decreased on The results of this study have shown that rose- day 7, 14, 21, and 28 (all p<0.001) (Figure 2). lle flower petals are able to increase the immune system, by increasing anti-inflammatory cytokines production and also by reducing the synthesis of pro-inflammatory cytokines. Increased synthesis of IL-10 causes suppression of TNF-α production and induces B lymphocyte to produce antibodies (17). The result of our study correlates with study by Fakeye et al. which stated that the phenolic extract from roselle flower petals was not toxic at the doses of immunostimulatory activities, which increased production of anti-inflammatory cytokine (IL-10) (22). The roselle flower petals were also known to have anthocyanin (flavonoids) compound, which has antioxidant activity (14). Figure 2. Histological colitis inflammatory score (HCIS) after Roselle flower petal is still inferior compared intervention. In mice treated with roselle compared to mesala- with mesalazine. Mesalazine has increased levels zine HCIS was significantly higher (day 7, 14, 21 and 28 ; p< 0.05). Compared to day 0, HCIS decreased significantly of IL-10 and decreased levels of IL-6 and TNF-α in mice treated with roselle (p < 0.001) and mesalazine since the first week of the treatment, results that (p<0.001). In post-hoc analysis, compared to day 0, HCIS de- are far different from the roselle flower petal creased significantly by day 21 and day 28 in roselle group , and by day 7, 14, 21, and 28 in mesalazine group extract, which only resulted on day 21 and 28 for IL 10 and IL 6, on day 28 for TNF-α. This proves that mesalazine is indeed recognized as the first

399 Medicinski Glasnik, Volume 17, Number 2, August 2020

choice for colitis therapy because mesalazine has In conclusion, roselle flower petal can attenuate the excellent efficacy and ability to maintain colitis inflammation in DSS-induced colitis model. Ro- remission (23). selle flower petal extract can be given as an adju- Roselle flower petal is able to reduce the inflamma- vant to the first-line therapy to enhance the anti- tion by increasing expression of anti-inflammatory inflammatory by immunomodulation of cytokines. cytokines (IL-10) and therefore decreasing expre- ACKNOWLEDGEMENT ssions of pro-inflammatory cytokines (IL-6 and TNF-α). These could be attributed to the immuno- The authors acknowledge the assistance of the modulation properties of phenolic compound and Department of Pathology, School of Medicine, also antioxidative nature of anthocyanin within Universitas Sumatera Utara, Indonesia. roselle flower petals (12). Whether active phenolic compounds or anthocyanin that contribute the FUNDING most to the immunomodulation is within our stu- No specific funding was received for this study. dy limitations. In-depth investigation using those active compounds is needed in the future for the CONFLICTS OF INTEREST treatment of colitis. Conflicts of interest: None to declare.

REFERENCES 1. Matsuoka K, Kobayashi T, Ueno F, Matsui T, Hirai F, 11. Zhang H, Wang L, Li C, Yu Y, Yi Y, Wang J, Chen D. Inoue N, Kato J, Kobayashi K, Kobayashi K, Koganei Exosome-induced regulation in inflammatory bowel K, Kunisaki R, Motoya S, Nagahori M, Nakase H, disease. Front Immunol 2019; 10:1464. Omata F, Saruta M, Watanabe T, Tanaka T, Kanai T, 12. Da-Costa-Rocha I, Bonnlaender B, Sievers H, Pischel Noguchi Y, Takahashi K, Watanabe K, Hibi T, Suzuki I, Heinrich M. Hibiscus sabdariffa L. – A phytoche- Y, Watanabe M, Sugano K, Shimosegawa T. Eviden- mical and pharmacological review. Food Chem 2014; ce-based clinical practice guidelines for inflammatory 165:424–43. bowel disease. J Gastroenterol 2018; 53:305–53. 13. Ilodibia CV, Achebe UA, Arubalueze CU, Ibeh E. A 2. Sun M, He C, Cong Y, Liu Z. Regulatory immune comparative study on morphology, mineral and vita- cells in regulation of intestinal inflammatory response min compositions of two variants of Hibiscus sabda- to microbiota. Mucosal Immunol 2015; 8:969–78. riffa (L.). Arch Agri Environ Sci 2019; 4:50–6. 3. Bamias G, Kaltsa G, Ladas SD. Cytokines in the pat- 14. Anokwuru CP, Esiaba I, Ajbaye O, Adesuyi AO. hogenesis of ulcerative colitis. Discovery Medicine Polyphenolic content and antioxidant activity of hi- 2011; 11:459–67. biscus sabdariffa calyx. Res J Med Plant 2011; 5:557– 4. Műzes G, Molnár B, Tulassay Z, Sipos F. Changes of 66. the cytokine profile in inflammatory bowel diseases. 15. Chang Y-C, Huang H-P, Hsu J-D, Yang S-F, Wang World J Gastroenterol 2012; 18:5848–61. C-J. Hibiscus anthocyanins rich extract-induced 5. Strober W, Fuss IJ. Pro-inflammatory cytokines in apoptotic cell death in human promyelocytic leuke- the pathogenesis of IBD. Gastroenterology 2011; mia cells. Toxicol Appl Pharm 2005; 205:201–12. 140:1756–67. 16. Fakeye T, Pal A, Khanuja S. Anxiolytic and sedative 6. Barman S, Kayama H, Okuzaki D, Ogino T, Osawa effects of extracts of Hibiscus sabdariffa Linn (family H, Matsuno H, Mizushima T, Mori M, Nishimura J, Malvaceae). Afr J Med Med Sci 2008; 37:49–54. Takeda K. Identification of a human intestinal mye- 17. Xu Q, Katakura Y, Yamashita M, Fang S, Tamura T, loid cell subset that regulates gut homeostasis. Int Matsumoto S, Aiba Y, Teruya K, Osada K, Nishikawa Immunol 2016; 28:533–45. R, Shirahata S. IL-10 augments antibody production 7. Keubler LM, Buettner M, Häger C, Bleich A. A multi- in in vitro immunized lymphocytes by inducing a hit model: colitis lessons from the interleukin-10–de- Th2-type response and B cell maturation. Biosci Bio- ficient mouse. Inflamm Bowel Dis 2015; 21:1967–75. tech Bioch 2004; 68:2279–84. 8. Liu B, Tonkonogy SL, Sartor RB. Antigen-presenting 18. Laurence DR, Bacharach AL. Evaluation of drug ac- cell production of IL-10 inhibits T-helper 1 and 17 tivities: pharmacometrics, volume 1. London: Acade- cell responses and suppresses colitis in mice. Gastro- mic Press Inc. Ltd.; 1964 enterology 2011; 141:653–62. 19. Laroui H, Ingersoll SA, Liu HC, Baker MT, Ayya- 9. Williams C, Panaccione R, Ghosh S, Rioux K. Op- durai S, Charania MA, Laroui F, Yan Y, Sitaraman timizing clinical use of mesalazine (5-aminosalicylic SV, Merlin D. Dextran sodium sulfate (DSS) induces acid) in inflammatory bowel disease. Therap Adv Ga- colitis in mice by forming nano-lipocomplexes with stroenterol 2011; 4:237–48. medium-chain-length fatty acids in the colon. PLoS 10. Nielsen OH, Seidelin JB, Munck LK, Rogler G. Use one 2012; 7:e32084. of biological molecules in the treatment of inflammatory bowel disease. J Intern Med 2011; 270:15–28.

400 Lubis et al. Hibiscus effect on inflammatory colitis

20. Jammal MP, Araújo da Silva A, Martins Filho A, de binding lectin-associated immune response lead to Castro Côbo E, Adad SJ, Murta EFC, Nomelini RS. barrier dysfunction in dextran sodium sulfate-induced Immunohistochemical staining of tumor necrosis rat colitis. Gut Liver 2015; 9:734–40. factor-α and interleukin-10 in benign and malignant 24. Poritz LS, Garver KI, Green C, Fitzpatrick L, Ruggie- ovarian neoplasms. Oncol Lett 2015; 9:979–83. ro F, Koltun WA. Loss of the tight junction protein 21. Strober W, Fuss I, Mannon P. The fundamental basis ZO-1 in dextran sulfate sodium induced colitis. J Surg of inflammatory bowel disease. J Clin Invest 2007; Res 2007; 140:12–9. 117:514–21. 25. Fakeye T. Toxicity and immunomodulatory acti- 22. Eichele DD, Kharbanda KK. Dextran sodium sulfate vity of fractions of Hibiscus sabdariffa Linn (family colitis murine model: An indispensable tool for ad- Malvaceae) in animal models. Afr J Tradit Complem vancing our understanding of inflammatory bowel 2008; 5:394–8. diseases pathogenesis. World J Gastroenterol 2017; 26. Ham M, Moss AC. Mesalamine in the treatment and 23:6016–29. maintenance of remission of ulcerative colitis. Expert 23. Yuan B, Zhou S, Lu Y, Liu J, Jin X, Wan H, Wang Rev Clin Pharmacol 2012; 5:113–23. F. Changes in the expression and distribution of cla- udins, increased epithelial apoptosis, and a mannan-

401 ORIGINAL ARTICLE

Association between CD133 expression and clinicopathological profile in colorectal cancer

Imelda Rey1, Agung Putra2-4, Dharma Lindarto1,3, Fauzi Yusuf 3,5

1Department of Internal Medicine, School of Medicine, Universitas Sumatera Utara, Medan, 2Stem Cell And Cancer Research (SCCR), School of Medicine, Sultan Agung Islamic University (UNISSULA), Semarang, Indonesia, 3Doctoral Department of Medical Science, School of Medicine, Universitas Sumatera Utara, Medan, 4Pathology Anatomy Department, School of Medicine, Sultan Agung Islamic University (UNISSULA), Semarang, 5Internal Medicine Department, School of Medicine, Syiah Kuala Univeristy, Banda Aceh; Indonesia

ABSTRACT

Aim To investigate CD133 expression and its relationship to clinicopathological profile in colorectal cancer (CRC) patients.

Methods This cross-sectional study was performed at the Internal Medicine Department, School of Medicine, Adam Malik General Hospital. The colorectal cancer tissue was taken from surgical resection and colonoscopy biopsy from CRC patients. Clinical profile was obtained by a questionnaire. Histopathology examination was done using hematoxylin and eosin staining. Immunohistoche- mistry (distribution score and intensity score) combined with ROC analysis were conducted to determine CD133 expression. An asso- Corresponding author: ciation between CD133 expression and clinicopathological profile Imelda Rey was then analyzed. Department of Internal Medicine, Results Out of 118 patients, 690 (58.5%) were male. The high School of Medicine, and low level of CD133 expression were found in 44 (37.3%) and Universitas Sumatera Utara 74 (62.7%) patients, respectively. No difference between gender, Dr Mansyur Street 5, 20155 Medan, age, body mass index, hemoglobin, leucocytes, platelets, and hi- Indonesia stopathology with CD133 expression was found. There was a significant difference between CD133 and different CRC locations Phone: +628126500525 (p=0.002). CD133 expression was higher in the proximal colon E-mail: [email protected]; than the rectum (p=0.002), and it was higher in the distal colon [email protected] than the rectum (p=0.008), especially in terms of percentages of ORCID ID: https://orcid.org/0000-0002- stained cancer cells (distribution score). 8651-1084 Conclusion CD133 expression was associated with the tumour location, but not with other clinicopathological factors. Original submission: 27 November 2019; Key words: carcinoma, colon, immunohistochemistry, neoplastic stem cells, rectum Revised submission: 10 March 2020; Accepted: 24 March 2020 doi: 10.17392/1106-20

Med Glas (Zenica) 2020; 17(2): 402-407

402 Rey et al. CD133 expression in colorectal cancer

INTRODUCTION Sumatera, Indonesia. All patients with CRC in the period of September 2018 – July 2019 were Colorectal cancer (CRC) is still one of the most included in the study. Colorectal cancer or ade- common cancers worldwide. In global cancer nocarcionoma was proven by biopsy of colo- statistics, approximately 1,360,600 estimated rectal tissue obtained by either surgical resection new cases of CRC were clinically diagnosed (1). or colonoscopy biopsy. Patients whose tissue was In Pirngadi Hospital, Medan, Indonesia, in 25% inadequate or did not meet the requirements for (197 of 760) patients who underwent colonosco- conducting histopathological examinations were py CRC was diagnosed (2). In Adam Malik hos- excluded. Clinicopathological data included pital, Medan, Indonesia, most common CRC was gender, age, hemoglobin (Hb) level, leucocyte, rectal cancer (61.5%), followed by the left-sided thrombocyte, body mass index (BMI), and a colon cancer (23.1%), and the right-sided colon location of the tumour. The tumour locations cancer (15.4%) (3). were categorized into the proximal colon, distal Relapse, metastasis, resistance of chemotherapy colon, and rectum. and recurrence are still problems in CRC ma- Ethical approvals for this study were obtained nagement. They contribute to higher mortality from the Institutional Ethics Committee of Uni- and poor survival rate (4). Cancer stem cell (CSC) versitas Sumatera Utara, Medan, Indonesia. was considered to be responsible for tumour progression, relapse, metastases and therapeutic Methods resistance (5-7). Therefore, the identification of CSCs is crucial in the search for therapeutic tar- Histopathological examination was conducted gets and useful prognostic markers for CRC. to classify CRC based on its differentiation. Af- ter the tissue sample was processed (fixation, In CRC, putative CSCs can be identified by se- dehydration, clearing, impregnation) to make a veral markers such as CD44, CD133, CD24, paraffin block, they were stained using- hema EpCAM, LGR5, ALD, and many more. These toxylin-eosin. Colorectal cancer was classified markers are highly tumorigenic, chemoresistant into well-differentiated adenocarcinoma, mode- and could affect survival rate of colorectal can- rately-differentiated adenocarcinoma and poorly- cer (8,9). CD133 molecule (also known as pro- differentiated adenocarcinoma (12). minin-1) is five transmembrane glycoproteins with a molecular weight of 120 kDa. It is mainly Immunohistochemistry (IHC) was conducted to localized in membrane protrusions (10). Pre- determine the presence of CD133 in the colon or vious studies have shown that CD133+ tumour rectum. Primary CD133 antibody GTX100567 cells were more resistant to radiochemotherapy (C1C2) internal (1:100-1:1000) (GeneTex In- than CD133- cells in CRC (11). These findings ternational Corporation, California, USA) was showed that CD133 CSC burden in colorectal used. The assessment of the CD133 expression cancer was of relevance for patients’ treatment was performed in the central and peripheral por- outcome (11). tions of the tumour, and the strongest expression results were recorded. Two blinded experienced To date, there are no data about an expression of pathologists from Anatomic Pathology Depar- CD133 in CRC patient in Indonesia. Since these tment, School of Medicine, Universitas Sumate- data are important for t patient management and ra Utara, were assigned to assess the slides for prognostic predictors, we aimed to investigate CD133 expression. An inter-observer agreement the association of CD133 expression with several between both pathologists was calculated using clinicopathological data and tumour location in κ-statistics. The stained tumour cell was scanned CRC patients in centre at Medan, Indonesia. by using a high-power microscope. PATIENTS AND METHOD For CD133 expression analysis, the immunoreactivity score on the cell membrane and/or Patients and study design cytoplasm was calculated from the sum of both quantitative and qualitative parameters. For qu- This cross-sectional study was conducted at Haji antitative analysis (distribution score), the score Adam Malik General Hospital, Medan North was determined based on percentage of reactive

403 Medicinski Glasnik, Volume 17, Number 2, August 2020

cells (% of the total area). The score was categori- Statistical analysis zed as follows: 0 (0% of immunoreactive cells), 1 Receiver operating characteristic (ROC) analy- (<10% of immunoreactive cells), 2 (10%–50% of sis was conducted to determine a cut-off point of immunoreactivecells), and 3 (more than 50% of CD133 expression. Based on ROC analysis, total immunoreactive cells). For qualitative analysis, score of 0–3 indicated low level and a total score the immunoreactive staining intensity (intensity of 4-6 indicated high level. The associations be- score) was classified as follows: 0 (no immuno- tween clinicopathological data (gender, age, BMI, reactivity), 1 (weak immunoreactivity) (Figure Hb, leucocyte, thrombocyte, histopathology) and 1A), 2 (intermediate immunoreactivity), and 3 CD133 expression was analysed by using the χ2 (strong immunoreactivity) (Figure 2B). Strong test or Kolmogorov-Smirnov test. Kruskal-Wallis immunoreactivity staining was considered as the and Mann-Whitney test were used to assess as- most intense staining observed in the positive sociations between CD133 expression and CRC control of respective antibody. The total scores locations. For CD133 expression, distribution ranged from 0 to 6 for CD133 expression (13). score and intensity score were first analysed separately before the total of both scores was used. The p<0.05 was considered as statistically significant.

RESULTS One hundred and eighteen CRC patients were enrolled in this study. The patients’ mean age was 57.17 years, with more males enrolled than females (Table 1). Table 1. Characteristics of 180 patients with colorectal cancer Variable Gender (No, %) Male 69 (58.5) Female 49 (41.5) Age (mean) (±SD) (years) 57.30 (±12.99) Hb (mean) (±SD) (g/dL) 10.51(±2.1) Leucocyte (median) (min.-max.) (cells/mm3) 7,420 (1,609-34,180) A) Thrombocyte (median) (min.-max.) (cells/mm3) 286,350 (89,000-562,000) BMI (median) (min.-max.) (kg/m2) 22.22 (1.22-30.86) Tumour location (No, %) Proximal colon 32 (27.1) Distal colon 43 (36.4) Rectum 43 (36.4) Histopathology (No, %) Well-differentiated adenocarcinoma 55 (46.6) Moderately differentiated adenocarcinoma 48 (40.7) Poorly differentiated adenocarcinoma 15 (12.7) CD133 expression (No, %) High Level 44 (37.3) Low Level 74 (62.7) Distribution score (No, %) 0-1 43 (36.4) 2-3 75 (63.6) Intensity score (No, %) 0-1 86 (72.9) 2-3 32 (27.1) min., minimum; max., maximum B) There was no statistically significant difference Figure 1. A) Weak immunoreactivity of CD 133 in colorectal between gender (p=0.150), age (p=0.611), BMI cancer tissue. CD133 is marked by light brown colour on cy- (p=0.995), Hb (p=0.643), leucocyte (p=0.394), toplasm (white arrow); B) strong immunoreactivity of CD 133 and thrombocyte (p=0.227) with CD133 expre- in colorectal cancer tissue. CD133 is marked by dark brown colour on cytoplasm (white arrow) (200x magnification) (400x ssions both in distribution score, intensity score magnification) (Rey I, 2019) (data not shown) and total score (Table 2).

404 Rey et al. CD133 expression in colorectal cancer

Table 2. Association of clinical characteristics and CD133 DISCUSSION expression CD133 expression In this study, we did not find significant differences Variable High level Low level p in age, gender, and BMI with CD133 expression. 58.09 56.82 Age (±SD) (years) 0.611 This is in line with a previous study (14-16). The (±12.99) (±13.06) Gender previous study on Medan showed that the risk of 22/22 47/27 0.150 (Male/Female) CRC increased with age until 6th decade of life (2). BMI (mean) 22.05 22.05 0.995 However, it seems that the expression of CD133 (±SD) (kg/m2) (±3.58) (±4.14) depended on factors other than age and gender. Hb (mean) 10.63 10.44 0.643 (±SD) (gr/dL) (±2.19) (±2.06) Some factors are known to be involved in CRC, Leucocytes (median) 8,100 7,600 such as sex hormone, genetic profile (17), and also 3 0.394 (min.-max.) (cells/mm ) (1,609-34,180) (4,478-21,130) epigenetic factor (18) that is affected by the envi- Platelets (mean) 321,554 295,594 0.227 (±SD) (cells/mm3) (±112,588) (±112,158) ronment and lifestyle (19-20). These confounding BMI, body mass index; Hb, haemoglobin; min., minimum; max., factors were not assessed in this study. maximum The results of this study have shown that CD133 Intra-observer agreement coefficient κ was 0.923, expression was higher in the proximal colon than 1.000, and 0.927 for immunohistochemistry of in the rectum, and higher in the distal colon than the proximal colon, distal colon, and rectum, res- the rectum, especially in terms of percentages of pectively. stained cancer cells (distribution score). This is in Most tumours were located in the distal colon or line with the previous study where CD133 expre- rectum. Most tumours were either well-differen- ssion was very low in the rectum (21): CD133 tiated or moderately-differentiated adenocarcino- expression was detected in 28% cases in the right ma (p=0.933) (Table 3). colon, in 53.8% in the left colon, and in 18.2% cases in the rectum, considering CD133 positivity Table 3. CD133 expression based on histopathology as >5% of stained cancer cells (21). The result CD133 expression Histopathology Total p Low level High level of our study is different from the study in which Well-differentiated 21 34 55 CD133 expression was higher in the rectum than Moderately differentiated 17 31 48 0.933 in the colon (15). Several other studies reported Poorly differentiated 6 9 15 that there was no significant difference between Total 44 74 118 positive and negative CD133 and tumour loca- Locations of tumour were associated with different tion (16,22). However, in both studies (16,22) CD133 expression (p=0.002) (Table 4). The proxi- cut-off used was different from our study. For mal and distal colon had higher CD133 expression instance, Nosrati et al. stated that <50% staining than rectum (p=0.002 and p=0.008, respectively). was considered negative, and >50% staining was In further analysis, only the distribution score was considered positive for CD133 (16). found significantly higher in the proximal and dis- We suspected that the difference of our results tal colon than rectum (p=0.002), while intensity was due to several reasons. There was a variabi- score was not (data not shown). However, findings lity of cut-off used to define low and high level in histopathology did not relate to CD133 expre- of CD133 between studies. A different cut-off ssion, either when scores were analysed separately could produce a different result. In addition, sin- (p=0.620 for distribution score and p=0.66 for in- ce studies are conducted in several different co- tensity score, data not shown) or total score was untries, differences could be caused by different used (p=0.933) (Table 3). geographical area or race. Therefore, the result should be applied specifically for each country. Table 4. CD133 expression based on colorectal cancer (CRC) location Moreover, a study divided the CRC locations into CRC location 2 parts, the left and right colon (21), while our (median, min. – max.) study divided the CRC locations into 3 parts: the p Proximal Distal Rectum proximal colon, distal colon and rectum. colon colon CD133 expression We also analysed relationship between CD133 4 (2-6)* 3 (2-6)† 2 (2-6) 0.002 (immunoreactivity score) expression and tumour histopathology. Our study *p=0.002 (compared with rectum); †p=0.008 (compared with rectum) has shown that CD133 expression did not affect

405 Medicinski Glasnik, Volume 17, Number 2, August 2020

CRC histopathology. This result is in concordan- dy found that rectal cancer had a better survival ce with almost all previous studies (15-16, 21- rate than colon cancer (27). Studies assessing the 23). Kojima et al. study concluded that CD133 role of CSCs in CRC patients also found that pa- occurred mainly in well- to moderately-diffe- tients with high level CD133 expression had poor rentiated adenocarcinomas; however, well- and survival rates (14,28). In our study, since CD133 moderately-differentiated adenocarcinoma were expression was associated with the tumour loca- categorized into one group, thus the difference tion, it is interesting to further investigate about was significant compared to poorly-differentiated this effect also to the survival rate of colorectal adenocarcinoma (24). CD133 was known to be cancer in Medan. elevated in CRC than benign colorectal epithe- The results of the presented study have not shown lium (25). But it seems its contribution to the statistically significant differences between hae- patient’s prognosis does not go through histopat- moglobin level and CD133 expression. Anaemia hological changes of colon (15-16, 21-23). has an important role in response of first line The findings of this research could impact se- 5FU-based chemotherapy (29). In advance CRC veral things in the management of CRC in Me- patients, the group with higher haemoglobin dan, Indonesia. We found a high level CD133 showed a better response rate, slower progressi- expression in a relatively high number of patients vity, and higher survival rates than the group with (37.3%). This marker should be considered in the anaemia (29). The previous study indicated that management of CRC since CD133 could redu- anaemia could affect CD133 expression through ce the effectivity of chemotherapy (11). In this hypoxia, but the results were conflicting (30-32). study, we found CD133 expression was highest This area needs to be assessed in further research. in the proximal colon. Both proximal and distal In conclusion, CD133 expressions were associa- colon have a higher amount of CD133 expression ted with the tumour location. This marker is a po- than rectum. The tumour location had impact on tential to be used as a CRC prognostic predictor the survival of CRC patients in the previous stu- in Medan, Indonesia. dy (26, 27). Three- and 5-year survival rates were 87.6% and 81.6% for the right-sided CRC group FUNDING and 91.5% and 84.5% for the left-sided CRC, No specific funding was received for this study. respectively; univariate and multivariate analysis showed that the risk of death was increased in TRANSPARENCY DECLARATION the right-sided CRC location (26). Another stu- Conflicts of interest: none to declare.

REFERRENCES 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tie- 6. Visvader JE, Lindeman GJ. Cancer stem cells in so- ulent J, Jemal A. Global cancer statistics, 2012. CA lid tumours: accumulating evidence and unresolved Cancer J Clin 2015; 65:87-108. questions. Nat Rev Cancer 2008; 8:755-68. 2. Effendi R, Efendi D, Dairy LB, Sembiring J, Sihom- 7. Song L, Li Y, He B, Gong Y. Development of small bing M, Marpaung B, Soetadi SM, Siregar GA, Zain molecules targeting the Wnt signaling pathway in LH. Profile of colorectal cancer patients in Endos- cancer stem cells for the treatment of colorectal can- copic Unit at Dr. Pirngadi Hospital-Medan. Indones cer. Clin Colorectal Canc 2015; 14:133-45. J Gastroenterol Hepatol Dig Endosc 2008; 9:78-81. 8. Zhou Y, Xia L, Wang H, Oyang L, Su M, Liu Q, Lin 3. Rey I, Effendi R, Zain LH. Comparison of CEA le- J, Tan S, Tian Y, Liao Q, Cao D. Cancer stem cells in vel among tumor location and histopathological fin- progression of colorectal cancer. Oncotarget 2018; dings in colorectal cancer. J Gastroen Hepatol 2016; 9:33403-15. 31(Suppl. 3):200. 9. Munro MJ, Wickremesekera SK, Peng L, Tan ST, 4. Kobayashi H, Mochizuki H, Sugihara K, Morita T, Itinteang T. Cancer stem cells in colorectal cancer: a Kotake K, Teramoto T, Kameoka S, Saito Y, Taka- review. J of Clin Pathol 2018; 71:110-6. hashi K, Hase K, Oya M. Characteristics of recu- 10. Corbeil D, Röper K, Hellwig A, Tavian M, Mira- rrence and surveillance tools after curative resection glia S, Watt SM, Simmons PJ, Peault B, Buck DW, for colorectal cancer: a multicenter study. Surgery Huttner WB. The human AC133 hematopoietic stem 2007; 141:67-75. cell antigen is also expressed in epithelial cells and 5. Reya T, Morrison SJ, Clarke MF, Weissman IL. targeted to plasma membrane protrusions. J Biol Stem cells, cancer, and cancer stem cells. Nature Chem 2000; 275:5512-20. 2001; 414:105-11.

406 Rey et al. CD133 expression in colorectal cancer

11. Ong CW, Kim LG, Kong HH, Low LY, Iacopetta B, 23. Huang R, Mo D, Wu J, Ai H, Lu Y. CD133 expre- Soong R, Salto-Tellez M. CD133 expression pression correlates with clinicopathologic features and dicts for non-response to chemotherapy in colorectal poor prognosis of colorectal cancer patients: An cancer. Modern Pathol 2010; 23:4507. updated meta-analysis of 37 studies. Medicine 2018; 12. Fleming M, Ravula S, Tatishchev SF, Wang HL. Co- 97:e10446. lorectal carcinoma: pathologic aspects. J Gastroin- 24. Kojima M, Ishii G, Atsumi N, Fujii S, Saito N, Ochiai test Oncol 2012; 3:153-73. A. Immunohistochemical detection of CD133 expre- 13. Kato Y, Nishihara H, Mohri H, Kanno H, Kobayas- ssion in colorectal cancer: a clinicopathological stu- hi H, Kimura T, Tanino M, Terasaka S, Tanaka S. dy. Cancer Sci 2008; 99:1578-83. Clinicopathological evaluation of cyclooxygenase-2 25. Chew MF, Teoh KH, Cheah PL. CD133 marks for expression in meningioma: immunohistochemical colorectal adenocarcinoma. Malay J Pathol 2012; analysis of 76 cases of low and high-grade meningi- 34:25-8 oma. Brain Tumor Pathol 2014; 31:23-30. 26. Aoyama T, Kashiwabara K, Oba K, Honda M, Sa- 14. Horst D, Kriegl L, Engel J, Kirchner T, Jung A. Pro- dahiro S, Hamada C, Maeda H, Mayanagi S, Kan- gnostic significance of the cancer stem cell markers da M, Sakamoto J, Saji S. Clinical impact of tumor CD133, CD44, and CD166 in colorectal cancer. location on the colon cancer survival and recurren- Cancer Invest 2009; 27:844-50. ce: analyses of pooled data from three large phase 15. Hong I, Hong SW, Chang YG, Lee WY, Lee B, Kang III randomized clinical trials. Cancer Med 2017; YK, Kim YS, Paik IW, Lee H. Expression of the can- 6:2523-30. cer stem cell markers CD44 and CD133 in colorectal 27. Jafarabadi MA, Mohammadi SM, Hajizadeh E, Ka- cancer: an immunohistochemical staining analysis. zemnejad A, Fatemi SR. Does the prognosis of co- Ann Coloproctol 2015; 31:84-91. lorectal cancer vary with tumor site? Gastroenterol 16. Nosrati A, Naghshvar F, Maleki I, Salehi F. Cancer Hepatol Bed Bench 2011; 4:199-09. stem cells CD133 and CD24 in colorectal cancers 28. Chen S, Song X, Chen Z, Li X, Li M, Liu H, Li J. in Northern Iran. Gastroenterol Hepatol Bed Bench CD133 expression and the prognosis of colorectal 2016; 9:132-9. cancer: a systematic review and meta-analysis. PloS 17. Cook MB, Dawsey SM, Freedman ND, Inskip PD, One 2013; 8:e5638. Wichner SM, Quraishi SM, Devesa SS, McGlynn 29. Tampellini M, Saini A, Alabiso I, Bitossi R, Brizzi KA. Sex disparities in cancer incidence by period MP, Sculli CM, Berruti A, Gorzegno G, Magnino A, and age. Cancer Epidemiol Biomarkers Prev 2009; Sperti E, Miraglia S. The role of haemoglobin level 18:1174-82. in predicting the response to first-line chemotherapy 18. Gabory A, Attig L, Junien C. Sexual dimorphism in in advanced colorectal cancer patients. Brit J Cancer environmental epigenetic programming. Mol Cell 2006; 95:13-20. Endocrinol 2009; 304:8-18. 30. Soeda A, Park M, Lee D, Mintz A, Androutsellis- 19. Brait M, Ford JG, Papaiahgari S, Garza MA, Lee Theotokis A, McKay RD, Engh J, Iwama T, Kuni- JI, Loyo M, Maldonado L, Begum S, McCaffrey L, sada T, Kassam AB, Pollack IF. Hypoxia promotes Howerton M, Sidransky D. Association between li- expansion of the CD133-positive glioma stem cells festyle factors and CpG island methylation in a can- through activation of HIF-1α. Oncogene 2009; cer-free population. Cancer Epidemiol Biomarkers 28:3949-59. Prev 2009; 18:2984-91. 31. Seidel S, Garvalov BK, Wirta V, von Stechow L, 20. Bollati V, Baccarelli A. Environmental epigenetics. Schänzer A, Meletis K, Wolter M, Sommerlad D, Heredity 2010; 105:105-12. Henze AT, Nister M, Reifenberger G. A hypoxic nic- 21. Hongo K, Kazama S, Sunami E, Tsuno NH, Taka- he regulates glioblastoma stem cells through hypoxia hashi K, Nagawa H, Kitayama J. Immunohistoche- inducible factor 2α. Brain 2010; 133:983-95. mical detection of CD133 is associated with tumor 32. Matsumoto K, Arao T, Tanaka K, Kaneda H, Kudo regression grade after chemoradiotherapy in rectal K, Fujita Y, Tamura D, Aomatsu K, Tamura T, Yama- cancer. Med Oncol 2012; 29:2849-57. da Y, Saijo N. mTOR signal and hypoxia-inducible 22. Minoo P, Zlobec I, Peterson M, Terracciano L, Lugli factor-1α regulate CD133 expression in cancer cells. A. Characterization of rectal, proximal and distal co- Cancer Res 2009; 69:7160-4. lon cancers based on clinicopathological, molecular and protein profiles. Int J Oncol 2010; 37:707-18.

407 ORIGINAL ARTICLE

In vitro regulation of IL-6 and TGF-ß by mesenchymal stem cells in systemic lupus erythematosus patients

Dewi Masyithah Darlan1,2, Delfitri Munir2, Nelva Karmila Jusuf3, Agung Putra4-6, Riyadh Ikhsan3, Iffan Alif4

1Department of Parasitology, Faculty of Medicine, Universitas Sumatera Utara (USU), Universitas Sumatera Utara (USU), Medan; In- donesia; 2Pusat Unggulan Tissue Engineering; Faculty of Medicine, Universitas Sumatera Utara (USU), Medan; Indonesia; 3Department of Dermatology and Venereology, Faculty of Medicine, Universitas Sumatera Utara (USU), Medan; Indonesia; 4Stem Cell and Cancer Research (SCCR), Medical Faculty, Sultan Agung Islamic University (UNISSULA), Semarang; 5Department of Postgraduate Biomedi- cal Science, Medical Faculty, Sultan Agung Islamic University (UNISSULA), Semarang; 6Department of Pathological Anatomy; Medical Faculty, Sultan Agung Islamic University (UNISSULA), Semarang, Indonesia

ABSTRACT

Aim To analyse the ability of mesenchymal stem cells (MSCs) to regulate interleukin 6 (IL-6) and transforming growth factor (TGF-β) expression in vitro under co-culture conditions in human systemic lupus erythematosus (SLE).

Method This study used a post-test group design that used peripheral blood mononuclear cells (PBMCs) from SLE patients at Kariadi Hospital, Semarang, Indonesia, and MSCs from a human umbilical cord. The cells were divided into two groups. The control group of PBMCs was treated with a standard medium, and the Corresponding author: treatment group was co-cultured with the MSCs at a 1:40 ratio. Dr. H. Agung Putra, MD, M.Si. Med., Following 24 h incubation, the levels of IL-6 and TGF-β relea- Chairman of Stem Cell and Cancer sed in the culture medium were measured using a specific ELISA Research (SCCR) assay. Laboratory, Faculty of Medicine, Results This study showed a significant decrease in IL-6 level Sultan Agung Islamic University, (p< 0.05) and a significant increase in TGF-β level (p<0.001) Semarang, Jl. Raya Kaligawe following 24 h of co-culture incubation of human SLE PBMCs KM. 4 Semarang, Central Java 50112 cells and MSCs. Phone.+628164251646, Conclusion The PBMCs-to-MSCs ratio of 1:40 can regulate the Fax. 0246594366, IL-6 and TGF-β levels in human SLE PBMCs. Email: [email protected] Key words: autoimmune disease, cytokine dysregulation, immu- ORCID ID: http://orcid.org/0000-0003- noregulation, inflammatory disorder 4261-9437

Original submission: 24 April 2020; Revised submission: 28 May 2020; Accepted: 15 June 2020 doi: 10.17392/1186-20

Med Glas (Zenica) 2020; 17(2): 408-413

408 Darlan et al. IL-6 and TGF-β regulation by MSCs in SLE

INTRODUCTION immunoglobulins (9). The overactivation of Th2 in SLE may cause IL-6 overexpression leading to Systemic lupus erythematosus (SLE) is a pro- enhanced autoantibody secretion (10). Abnorma- totypical autoimmune disease characterized lities in these mechanisms, critical factors for se- by the excessive production of autoantibodies, lf-tolerance and immune homeostasis maintenan- immune complex formation, and systemic or ce, are also caused by downregulation of TGF-β, organ-specific inflammation triggering broad commonly produced by Tregs and known to inflammatory responses (1). Although SLE pati- regulate cellular functions such as proliferation, ent survival has improved in the past few years, differentiation, migration, and survival, particu- excess mortality has not improved substantially larly by inhibiting several Tregs, such as Th1 and (2). Mediation between the SLE cytokine tsuna- Th2 (11). mi and several T helper cell (Th) subsets is unba- lanced, abnormalities resulting from degradation A previous study revealed that under sufficient of production regulatory T cells (Tregs); this may levels of inflammatory niche, MSCs might im- affect the inflammatory niche (3). pede differentiation of Th1 and Th2 and simultaneously inhibit potent inflammatory cytokines Mesenchymal stem cells (MSCs) are a plastic- such as IL-6; this is a result of the production of adherent stromal cell with a multipotent differen- Tregs through the expression of anti-inflamma- tiation capacity (4); this may inhibit T helper cells tory cytokines such as TGF-β (12). These fin- and effector T cells, autoantibody production by dings demonstrate the possibility that MSCs im- B cells, and other homeostatic depletions leading pede the inflammatory niche in SLE. However, to immune system failure (5). The MSCs may the capacity of MSCs to regulate in-vitro IL-6 also fight excessive inflammatory niches by en- and TGF-β production in SLE patients remains hancing the production of Tregs. These qualities unclear. are supported by the expression of inducible nitric oxide synthase and indoleamine 2,3-dioxyge- Previously, we reported that MSCs induce the nase, in addition to several anti-inflammatory generation of Treg cell population in PBMCs of cytokines, including interleukin 10 and tran- SLE patients, indicating an enhancement of the sforming growth factor (TGF-β) (6)but the un- modulation of proinflammatory milieu (13). derlying mechanisms remain largely unknown. The aim of this study was to analyse the ability of The aim of this study was to investigate how MSCs to regulate IL-6 and TGF-β expression in allogeneic MSCs mediate immunosuppression vitro under co-culture conditions in human SLE. in lupus patients. Methods The effects of allogeneic umbilical cord-derived MSCs (UC-MSCs; PATIENTS AND METHODS these molecules may inhibit Th1, Th2, and Th17, Patients and study design which produce several excessive proinflammatory cytokines, including, specifically, interleu- This post-test control group study was conducted kin 6 (IL-6) (6,7)but the underlying mechanisms in the Stem Cell and Cancer Research Laboratory remain largely unknown. The aim of this study at the Sultan Agung Islamic University’s Faculty was to investigate how allogeneic MSCs mediate of Medicine in Semarang, Indonesia, between immunosuppression in lupus patients. Methods September and October 2018. The study used The effects of allogeneic umbilical cord-derived peripheral blood mononuclear cells (PBMCs) of MSCs (UC-MSCs. three SLE patients from Kariadi Hospital, Sema- Interleukin 6 usually mediates excessive in- rang, Indonesia, and MSCs from the last third of flammatory responses. IL-6 is a potent proin- a donor umbilical cord. The study samples were flammatory cytokine; its wide range of biologi- divided into two groups (three replications from cal activities play an essential role in immune three patients); the control group was treated regulation (8). The primary sources of IL-6 are with a standard medium, and the treatment group monocytes, fibroblasts, endothelial cells, B cells, received a treatment co-cultured with hUC-MSC. and T cells. Notably, IL-6 is produced by Th2 An informed consent was received from all pa- cells with a role in the differentiation of B lymp- tients’ legal guardians after providing beneficial hocytes into mature plasma cells that secrete information prior to the study.

409 Medicinski Glasnik, Volume 17, Number 2, August 2020

This study was approved by the Ethics Com- modified eagle medium (Sigma-Aldrich, Louis mittee of the institutional review board of Sul- St, MO) supplemented with 10% FBS (Gibco In- tan Agung Islamic University’s Faculty of Me- vitrogen, NY, USA) and 1% penicillin (100 U/ dicine, Semarang. mL)/streptomycin (100 µg/mL) (Gibco Invitro- gen, NY, USA). Upon reaching 95% confluence, Methods the standard medium was aspirated and replaced Mesenchymal stem cell isolation. The MSCs with an osteogenic differentiation medium con- were isolated from an umbilical cord obtained taining Human MesenCult Osteogenic Differen- from a donor. The umbilical cord was cut into tiation Basal Medium (Stem Cell Technologies, smaller pieces and transferred to a T25 culture Singapore), augmented with 20% Human Mes- flask (Corning, Tewksbury, MA, USA) contain- enCult Osteogenic Differentiation 5X Supple- ing Dulbecco’s modified eagle medium (Sigma- ment (Stem Cell Technologies, Singapore), and Aldrich, Louis St, MO), supplemented with 10% 1% L-Glutamine (Gibco, Invitrogen, NY, USA). fetal bovine serum (FBS) (Gibco, Invitrogen, Differentiation mediums were renewed every NY, USA) and 1% penicillin (100 U/mL)/strep- three days. Upon bone matrix formation, osteo- tomycin (100 µg/mL) (Gibco, Invitrogen, NY, genic differentiation was visualized by staining USA). Umbilical cord tissue was then incubated with 1 ml of 2% alizarin red solution. at 37 °C in a humid atmosphere consisting of 5% Isolation of PBMCs and MSCs co-culture. Hu-

CO2. The medium was renewed every three days, man PBMCs were separated using a Ficoll-Paque and the cells were passaged after reaching 80% (Sigma-Aldrich, Louis St, MO) density gradi- confluence (14 days). The following experiments ent centrifugation in a 15 mL conical tube from used hUC-MSCs at passage fourth. three SLE patient volunteers who gave specific, Characterization of MSCs. Surface antigens informed consent. The PBMCs were cultured and akin to MSCs were analysed by flow cytome- expanded in 2 mL of advanced RPMI 1640 culture try analysis at the fourth passage according to medium (Gibco, Invitrogen, NY, USA), supple- company protocols. The cells were subsequently mented with 10% FBS, 100 U/mL penicillin and incubated in the dark with allophycocyanin mo- streptomycin, 2 mM glutamine and incubated at

use anti-human CD73 antibodies, fluorescein 37 °C in a humidified atmosphere with 5% CO2. isothiocyanate mouse anti-human CD90 antibo- For the treatment group, following 24 h incubation, dies, perCP-Cy5.5.1 mouse anti-human CD105 PBMCs were co-cultured with MSCs on 24-well antibodies, and phycoerythrin mouse anti-human plates cultured in RPMI supplemented with 1% CD45/CD34/CD11b/CD19/HLA-DR-negative penicillin-streptomycin and 10% FBS at an MSCs- Lin antibodies. According to the International to-PBMCs ratio of 1:40 for 24 h, in accordance Society of Cellular Therapy (14), the specific with the aforementioned previous study (12). For marker profile expressed by MSCs comprises the control group, following 24 h incubation, the CD73, CD105, and CD90, and CD45/CD34/ isolated PBMCs were co-cultured with the stan- CD11b/CD19/HLA-DR negative Lin. dard medium in another 24-well plate for 24 h. The analysis was performed using BD Stemflow ELISA assay. A specific ELISA assay mea- (BD Biosciences, San Jose, CA, USA). The anti- sured the levels of IL-6 and TGF-β released in gens were stained with a specific antibody for 30 the culture mediums containing cells from both minutes at 4 °C, examined with a BD Accuri C6 the treatment and control groups. The levels of Plus flow cytometer (BD Biosciences, San Jose, IL-6 and TGF-β were measured according to the CA, USA) and analysed using BD Accuri C6 Plus manufacturer’s instructions (Fine Test, Wuhan, software (BD Biosciences, San Jose, CA, USA. China). Colorimetric absorbance was recorded at In-vitro osteogenic differentiation assay. To a wavelength of 450 nm. characterize the differentiation capacity of MSCs, Statistical analysis the cells were cultured at a density of 1.5 x 104 cells per well. The cells were grown at 37 °C, 5% Group comparisons were made using a paired -test followed by Fisher’s Least Significant Diffe- CO2, and ≥95% humidity in 24-well plates fea- turing a standard medium comprising Dulbecco’s rence test. A p<0.05 was considered significant. 푡

410 Darlan et al. IL-6 and TGF-β regulation by MSCs in SLE

RESULTS

Isolation and differentiation of MSCs Umbilical-cord-derived MSC-like cells were successfully isolated based on their plastic adherent capability under standard culture conditions. After passage four, MSCs had a stable fibroblast- like morphology with spindle shape (Figure 1A ). MSCs showed their multipotent characteristic through differentiating into osteoblast under in- ductive culture conditions (Figure 1 B).

Figure 2. Immunophenotyping analysis of mesenchymal stem cells, which was positive for CD73, CD90, and CD105 antibodies and negative for Lin antibodies

Figure 1. A) Umbilical cord mesenchymal stem cells characterized by being fibroblast-like with a spindle shape; B) calcium deposition shown in red by red alizarin staining (400x Figure 3. ELISA assays showed a significant decrease of inter- magnification) leukin 6 levels in the treatment group (639.35 ± 96.33 pg/mL) (left); a significant decrease was observed in transforming growth factor level (150.80 ±7.06 pg/mL) (right) MSCs were also positive for CD73, CD90 and CD105 and negative for the hematopoietic linea- DISCUSSION ge markers (CD45, CD34, CD11b, CD19, HLA- DR) (Figure 2). The ability of MSCs to regulate Treg prolifera- The results showed that IL-6 level was signifi- tion has promoted alternative clinical therapies cantly decreased (p<0.05) in the treatment group (13). However, though several studies have re- (639.35 ± 96.33 pg/mL) following 24 h incubati- vealed that MSCs can suppress the inflammatory on of the MSCs. However, a significant increase niche through different mechanisms (15,16), (p< 0.05) in TGF-β levels (150.80 ± 7.06 pg/mL) the underlying mechanisms of MSCs-mediated was observed (Figure 3). immunoregulation in SLE patients remained in- completely understood. This study has demon-

411 Medicinski Glasnik, Volume 17, Number 2, August 2020

strated that hUC-MSCs regulate the inflamma- Accordingly, our study demonstrated a significant tory niche in SLE patients by downregulating increase in TGF-β levels in human SLE PBMCs, IL-6 levels and upregulating TGF-β levels. following co-culturing at an MSCs-to-PBMCs ratio There are several immune tolerance breakdowns of 1:40 for 24 h. In inflammatory diseases, MSCs in SLE; these involve autoantigens and are present may express TGF-β, which may regulate multifa- in several cell types, making antigen-specific the- ceted cellular functions, including inhibiting seve- rapy design difficult (17). Current SLE therapies ral effectors and Th subsets. The previous study inhibiting both T and B lymphocytes carry the risk demonstrated that TGF-β could promote Tregs and of adverse events, such as infection (7,17)these inhibit effector T cell development (24). Another cells also display multiple potent immunomo- study demonstrated TGF-β’s critical role in regu- dulatory capabilities, including allosuppression, lating the signaling pathway that commences and making allogeneic cell therapy a possibility. The preserves the FoxP3 suppressor function (25)Th2, exact mechanisms involved in regulatory T cell in- or T regulatory (Treg. However, TGF-β expressed duction by allogeneic human MSC was examined, by Tregs occupy an autocrine position, excessively using purified CD4+ populations and well-charac- expressing anti-inflammatory cytokines and regu- terized bone marrow-derived adult human MSC. lating several inflammatory-mediated several Th Allogeneic MSC were shown to induce forkhead subsets, including, specifically, Th2. Inhibited Th2 box P3 (FoxP3. The unique immunoregulatory ca- development is caused by TGF-β expression; this pacities of MSCs include impeding T cell prolife- may inhibit the GATA-3 transcription factor (26) ration and inducing Treg generation, inhibiting B increased levels of cytokines and their receptors cell function and immunoglobulin production, and can be observed in target organs, and it is clear that regulating the excessive proinflammatory niche, they have important roles in disease pathogene- particularly in the context of human SLE (19). A sis. Recent therapeutic strategies have focused on previous study revealed that, under inflammatory proximal cytokines, such as interferon-α, interleu- conditions, MSCs might simultaneously inhibit kin (IL. Thus, degradation of Th2 generation also Th1 and Th2 differentiation and decrease seve- results in the downregulation of IL-6 molecules as ral potent inflammatory cytokines through the potent proinflammatory cytokines (27). These fin- production of Tregs activated by TGF-β (20)me- dings suggest that MSCs could regulate inflamma- senchymal stromal cells (MSCs. tory conditions in human SLE by downregulating IL-6 and upregulating TGF-β expression. Howe- This study demonstrated that MSCs might regu- ver, this study did not explore other potent cytoki- late the inflammatory niche by inhibiting the ove- nes, related transcription factors, or the generation rexpression of potent proinflammatory cytokines of different Th cell subsets. and simultaneously enhancing the anti-inflammatory mechanism in human SLE. The study showed In conclusion, we have shown here that MSCs that at an MSCs-to-PBMCs ratio of 1:40 in a 24 could hamper the IL-6 level and enhance the h co-culture MSCs might inhibit IL-6 expression TGF-β level in human SLE PBMCs at a 1:40 ra- in human SLE PBMCs. These data suggest that tio of MSCs to PBMCs. This finding is important MSCs may impede the inflammatory niche under for understanding the capability of MSCs as a SLE conditions. At a sufficient proinflammati- potential immunomodulator, particularly in SLE. on level MSCs may upregulate the expression of ACKNOWLEDGMENTS TLR-3, leading to the release of various anti-inflammatory molecules, including interleukin 10 The authors gratefully acknowledge that the (21). The NF-κβ and ERK pathways activated in present research is supported by the Ministry MSCs by several inflammatory cytokines binding of Research and Technology, the Higher Edu- may lead to cyclooxygenase-2 and TLR-4 upregu- cation Republic of Indonesia, and the Research and Community Service at Universitas Sumatera lation (22). Upregulation of PGE2 may lead those molecules to bind to EP2 and EP4 receptors, ac- Utara. For their assistance with this research, the tivating TRIF-TRAM mediated anti-inflammatory authors would like to thank the staff at Stem Cell signals and resulting in anti-inflammatory expre- and Cancer Research Laboratory at the Sultan ssions such as interleukin 10 and TGF-β (23). Agung Islamic University’s Faculty of Medicine in Semarang.

412 Darlan et al. IL-6 and TGF-β regulation by MSCs in SLE

FUNDING TRANSPARENCY DECLARATION Financial support was provided through the rese- Conflicts of interest: None to declare. arch grant TALENTA of the Year 2019 (Contract Number 499/UN5.2.3.1/PPM/KP-TALENTA Tanggal; 02 April 2019).

REFERENCES A, Prockop DJ, Horwitz EM. Minimal criteria for 1. Wang D, Zhang H, Liang J, Li X, Feng X, Wang H, defining multipotent mesenchymal stromal cells. The Hua B, Liu B, Lu L, Gary S. Gilkeson GS, Silver RM, International Society for Cellular Therapy position Chen W, Shi S, Sun L. Allogeneic mesenchymal stem statement. Cytotherapy ; 2006; 8:315–7 cell transplantation in severe and refractory syste- 15. Keating A. Mesenchymal stromal cells: New directi- mic lupus erythematosus: 4 years of experience. Cell ons. Cell Stem Cell 2012; 10:709–16. Transplant 2013; 22:2267–77. 16. Rodríguez-rodríguez N, Rosetti F, Crispín JC. T 2. Jorge AM, Lu N, Zhang Y, Rai SK, Choi HK. Cells. In: Tsokos GC, ed. Systemic Lupus Erythe- Unchanging premature mortality trends in systemic matosus: Basic, Applied and Clinical Aspects. New lupus erythematosus: A general population-based stu- York: Elsevier, 2016:113–9. dy (1999-2014). Rheumatol (United Kingdom) 2018; 17. Moulton VR, Suarez-Fueyo A, Meidan E, Li H, Mizui 57:337–44. M, Tsokos GC. Pathogenesis of human systemic lu- 3. Miyake K, Akahoshi M, Nakashima H. Th subset ba- pus erythematosus: a cellular perspective. Trends Mol lance in lupus nephritis. J Biomed Biotechnol 2011; Med 2017; 23:615–35. 2011: 980286. 18. Sanz I, Lee FE-H. Sanz 2010 Nat Rev Rheumatol - B 4. Sasaki M, Honmou O. Mesenchymal Stem Cell. In: cells as therapeutic targets in SLE. Nat Rev Rheuma- Houkin K, Abe K, Kuroda S, eds. Cell Therapy Aga- tol 2010; 6:326–37. inst Cerebral Stroke : Comprehensive Reviews for 19. Luz-Crawford P, Kurte M, Bravo-Alegría J, Contreras Translational Researcher and CLinical Trial. 1st ed. R, Nova-Lamperti E, Tejedor G, Noël D, Jorgensen Japan: Springer; 2017:147–55. C, Figueroa F, Djouad F. Mesenchymal stem cells 5. Fan L, Hu C, Chen J, Cen P, Wang J, Li L. Interaction generate a CD4+CD25+Foxp3 + regulatory T cell po- between mesenchymal stem cells and B-cells. Int J pulation during the differentiation process of Th1 and Mol Sci 2016; 17: 650. Th17 cells. Stem Cell Res Ther 2013; 4:65. 6. Wang D, Feng X, Lu L, Konkel JE, Zhang H, Chen 20. Bernardo ME, Fibbe WE. Mesenchymal stromal Z, Li X, Gao X, Lu L, Shi S, Wanjun Chen, Sun L. A cells: sensors and switchers of inflammation. Cell CD8 T cell/indoleamine 2,3-dioxygenase axis is requ- Stem Cell 2013; 13:392–402. ired for mesenchymal stem cell suppression of human 21. Aksoy E, Taboubi S, Torres D, Delbauve S, Hachani systemic lupus erythematosus. Arthritis Rheumatol A, Whitehead MA, Pearce WP, Berenjeno IM, Nock 2014; 66:2234–45. G, Filloux A, Beyaert R, Flamand V, Vanhaesebroeck 7. English K, Ryan JM, Tobin L, Murphy MJ, Barry FP, B. The p110δ isoform of the kinase PI(3)K controls Mahon BP. Cell contact, prostaglandin E2 and tran- the subcellular compartmentalization of TLR4 signa- sforming growth factor beta 1 play non-redundant ling and protects from endotoxic shock. Nat Immunol roles in human mesenchymal stem cell induction of 2012; 13:1045–54. CD4+CD25 High fork head box P3+ regulatory T 22. Putra A, Ridwan FB, Putridewi AI, Kustiyah AR, cells. Clin Exp Immunol 2009; 156:149–60. Wirastuti K, Sadyah NAC, Rosdiana I, Munir D. The 8. Tackey E, Lipsky P, Illei G. Rationale for interleu- role of tnf-α induced mscs on suppressive inflammati- kin-6 blockade in systemic lupus erythematosus. Lu- on by increasing tgf-β and il-10. Open Access Maced pus 2004; 13:333–43. J Med Sci 2018; 6:1779–83. 9. Hofmann K, Clauder AK, Manz RA. Targeting B 23. Wan YY, Flavell RA. “Yin-Yang” functions of tran- cells and plasma cells in autoimmune diseases. Front sforming growth factor-β and T regulatory cells in Immunol 2018; 9:835. immune regulation. Immunol Rev 2007; 220:199–213. 10. Su DL, Lu ZM, Shen MN, Li X, Sun LY. Roles of pro- 24. Chen W, Konkel JE. TGF- β and ''Adaptive'' Foxp 3+ and anti-inflammatory cytokines in the pathogenesis regulatory T cells. J Mol Cell Biol 2010; 2:30–6. of SLE. J Biomed Biotechnol 2012; 2012: 347141. 25. Mantel PY, Kuipers H, Boyman O, Rhyner C, Oua- 11. Talaat RM, Mohamed SF, Bassyouni IH, Raouf AA. ked N, Rückert B, Karagiannidis C, Lambrecht Th1/Th2/Th17/Treg cytokine imbalance in systemic BN, Hendriks RW, Crameri R, Akdis CA, Blaser lupus erythematosus (SLE) patients: correlation with K, Schmidt-Weber CB. GATA3-driven Th2 respon- disease activity. Cytokine 2015; 72:146-53. ses inhibit TGF-β1-induced FOXP3 expression and 12. Duff MM, Ritter T, Ceredig R, Griffi MD. Me- the formation of regulatory T cells. PLoS Biol 2007; senchymal stem cell eff ects on T-cell eff ector 5:2847–61. pathways. Stem Cell Res Ther 2011; 2:1–9. 26. Davis LS, Hutcheson J, Mohan C. The role of cyto- 13. Ikhsan R, Putra A, Munir D, Darlan DM, Suntoko B, kines in the pathogenesis and treatment of systemic Retno A. Mesenchymal stem cells induce regulatory lupus erythematosus. J Interf Cytokine Res 2011; T-cell population in human SLE. Bangladesh J Med 31:781–9. Sci 2020; 19:743–8. 27. Tanaka T, Narazaki M, Kishimoto T. IL-6 in In- 14. Dominici M, Blanc K Le, Mueller I, Slaper-Corte- flammation, immunity, and disease. Cold Spring Harb nbach I, Marini FC, Krause DS, Deans RJ, Keating Perspect Biol 2014; 6:a016295.

413 ORIGINAL ARTICLE

An ethanol extract of Senduduk fruit (Melastoma malabathricum L) inhibits the expression of vascular endothelial growth factor and tumour necrosis factor alpha in HeLa cells Deri Edianto1, Aznan Lelo2, Syafruddin Ilyas3, Marline Nainggolan4

1Department of Obstetrics and Gynaecology, 2Department of Pharmacology and Therapeutics; Faculty of Medicine Universitas Sumatera Utara, 3Faculty of Mathematics and Natural Science, 4Faculty of Pharmacy; Universitas Sumatera Utara, Indonesia

ABSTRACT

Aim Senduduk fruit (Melastoma malabathricum L) is native to Indomalaya and is believed to possess anticancer activity. This study investigated antiangiogenic and anti-inflammation effects of an ethanolic extract of Senduduk fruit (EESF) on HeLa cells.

Methods Cytotoxicity was assayed in HeLa cell cultures exposed

to a concentration series of 500–7.8 µg/mL of EESF. IC50 was determined with a methylthiazol-tetrazolium (MTT) cell viability assay. Antiangiogenic and anti-inflammation activity was evaluated by an immunocytochemical assay of vascular endothelial growth factor (VEGF) and tumour necrosis factor alpha (TNF-α) expression in HeLa cells cultured with 1× or 2× IC or as a control Corresponding author: 50 without EESF. Deri Edianto

Department of Obstetrics and Results IC50 of the EESF was 956 µg/mL. The intensity of VEGF Gynaecology, Faculty of Medicine, staining indicated moderate expression in HeLa cells in response to IC , weak expression in response to 2×IC , and strong expre- Universitas Sumatera Utara 50 50 ssion in the absence of the EESF. While the intensity of TNF-α Jalan Suka Menang no 18, 20146 Medan, staining indicated moderate expression in HeLa cells in response North Sumatera, Indonesia to IC50 and to the absence of EESF, and weak expression in res- Phone: +62 81 1639 232; ponse to 2× IC50. E-mail: [email protected] Conclusion Senduduk fruit extract inhibited VEGF and TNF-α ORCID ID: https://orcid.org/0000-0002- expression in HeLa cells in a concentration-dependent manner. 1459-9523 Key words: ethanol extract, HeLa cell, Melastoma malabathricum L fruit, TNF-α, VEGF

Original submission: 15 April 2020; Revised submission: 08 June 2020; Accepted: 22 June 2020 doi: 10.17392/1182-20

Med Glas (Zenica) 2020; 17(2): 414-419

414 Edianto et al. Senduduk Fruit Inhibits VEGF and TNF-α

INTRODUCTION of cancer (17). The chemopreventive anticancer Noncommunicable diseases are responsible for the activities of more than 1000 phytochemicals majority of deaths worldwide, and cancer is a le- have been investigated (18). Bioflavonoids are ading cause (1). The reasons are complex, but re- plant secondary metabolites with potential hu- flect both aging and growth of the population (2,3). man health benefits. They are relatively abundant According to Globocan 2018, cervical cancer is in plant foods and are nontoxic (19). Polyphenols the fourth most frequently diagnosed cancer and are a class of flavonoids that have low bioava- the fourth leading cause of cancer deaths in wo- ilability but are rapidly metabolized (20). The men, with an estimated 570,000 cases and 311,000 anticancer activity of polyphenols has been inve- deaths worldwide (4). Approximately 80% of new stigated, and a few trials in cervical cancer have cases occur in countries without effective scree- primarily involved green tea and curcumin (16). ning programs (5). Cervical cancer is generally tre- This study investigated the anticervical cancer ated by surgery or radiation. Both have side effects activity of a Melastoma (Senduduk) fruit extract. and recurrence may occur (6,7). Novel pharmaco- Ripe Senduduk fruits are purplish black, have a logical breakthroughs are needed, such as natural sweet and sharp taste. The Melastoma plants are extracts that are effective with minimal toxicity (8). native to Sumatera, Borneo, and the Malay Pe- The carcinogenesis of cervical cancer is a mul- ninsula (21). tistep process that causes a progressive transfor- Melastoma malabathricum L (Figure 1) is an or- mation of normal cells into cancer cells with namental plant with white or pink-purple flowers, disease-specific characteristics (9). Sustained included in the Melastomataceae family, which proliferative signaling, evading growth suppre- spread in Asia especially in Southeast Asia. Some ssors, resisting cell death, enabling replicative regions are approved under different names. Tra- immortality, inducing angiogenesis, and acti- ditionally, it has been known to have many me- vation of invasive and metastatic behaviour are dicinal benefits such as antibacterial, antiviral, key properties of cancer cells, named cancer anti-inflammatory, anti-parasite, antioxidant, an- hallmarks (10). Angiogenesis, the initiation and ticancer, anticoagulant, platelet activating factor growth of new blood vessels, ensuring the supply inhibitor, antiulcer, antidiarrhea, antivenom, fe- of oxygen and nutrients required to maintain tu- ver and wound healing (22). mour growth and support metastasis (11). Vascu- lar endothelial growth factor (VEGF) stimulates the formation of new blood vessels (12). In the early 19th century, it was perceived that cancer was related to inflammation. Recent efforts have shed new light on molecular and cellular pathways that connect inflammation and cancer. Inflammation is a target for pharmacologic intervention (13). One of the main chemical mediators involved in cancer-related inflammation is TNF-α (14). Treatment failure and the high systemic toxicity Figure 1. Senduduk Plant (Melastoma malabathricum L) (Edi- of conventional cancer therapies have prompted anto D, 2019) a search for a novel agent with anticancer activity and low toxicity (15). Polyphenols, alkaloids, ni- Melastoma genus in Southeast Asia consists of trogen compounds, and carotenoids are bioactive 22 species, 2 subspecies, and 3 varieties. These phytochemicals that can interfere with many si- varieties are based on the colour of flower petals, gnaling pathways that regulate cell proliferation which are pink, white, and purple. Pink flower and tumour initiation and growth (16). Vegeta- petal varieties are often found in Indonesia and bles, grains, fruits, and other plant products are Malaysia. The number of flower petals is five. rich in phytochemicals, and diets including a va- The fruit colour is deep purple, soft, and contains riety of fruits and vegetables may lower the risk many fine orange seeds (21).

415 Medicinski Glasnik, Volume 17, Number 2, August 2020

Senduduk fruit (Melastoma malabathricum L) is grow confluent at least 80%, which were evalu- known to contain anthocyanins such as cyanidin ated using inverted light microscope (Olympus di-hexoside, cyanidin hexoside, delphinidin hexo- CKX41, Japan). side, and pelargonidin, but the use of active ingre- Cytotoxicity assay. HeLa cells were transferred dients from this fruit has not been reported (22-25). to 96-well plates (104 cells/well) in 100 μL culture mediums containing 500, 250, 125, 62.5, MATERIAL AND METHODS 31.25, 15.625, or 7.8125 μg/mL of senduduk extract with six replicate wells for each concen- Materials and study design tration. The control cultures included culture me- This study was conducted at the Faculty of Phar- dium without cell and cell control with no added macy, Universitas Sumatera Utara, Medan and extract. The culture medium was discarded after Department Parasitology, Faculty of Medicine overnight incubation and replaced with 100 μL Universitas Gadjah Mada, Yogyakarta, Indonesia methylthiazol-tetrazolium (MTT – Sigma Al- in the period April - May 2019. drich, USA) solution and incubated for 2 hours. The approval for the investigation was obtained The MTT reaction was stopped with 10% sodi- from the Ethics Committee of the Health Resear- um dodecyl sulfate (SDS) in 0.01 N HCl (Sigma ch Ethical Committee, Faculty of Medicine Uni- Aldrich,USA) after formation of formazan, the versitas Sumatera Utara Indonesia. plates were kept overnight in the dark at room Whole ripe of Senduduk fruit (Melastoma mala- temperature, and the tetrazolium absorbance was bathricum L) collected from abandoned, vacant measured at 595 nm using a culture micro plate land in Pasar Rawa Village, Gebang District, reader (Bio-Rad Bechmark, (USA). Cell viability Langkat Regency, North Sumatera Province, was calculated as Indonesia was used in this study. The extract was made in the phytochemical laboratory of (C −B) % Cell viability = × 100%, the Faculty of Pharmacy, Universitas Sumatera (A −B) Utara, Medan, Indonesia. Prior to the extraction, the fruit was separated from it stalk, squeezed to obtain the pulp, and dried in a dehydrator at where A = cell control (cells + media), B = media 40°C until it was brittle. The ethanol extract of control, C = sample + media + cells. IC50 was cal- Senduduk fruit (EESF) was made by maceration culated with Microsoft Excel. of the pulp in 80% ethanol. The extracts were Vascular endothelial growth factor (VEGF) put in the dark red glass bottle and kept in free- and tumour necrosis factor alpha (TNF-α) ze on refrigerator before the use. The flavonoid expression. HeLa cells were transferred to 24- components of the extract were evaluated by well plates at 100,000 cells/well in RPMI with spectrophotometry (PT Saraswanti Indo Gene- coverslip inside and incubated overnight. The tech, Bogor, Indonesia). EESF was added at 1× or 2× IC50 to all but not Methods the control wells and reincubated overnight for each marker. VEGF and TNF-α expression was HeLa cells. HeLa cells were obtained from the assayed at the coverslip after overnight reincu- Laboratory of Parasitology at the Faculty of bation with a primary anti-VEGF C-1, sc-7269 Medicine, Universitas Gadjah Mada, Yogya- antibody (Santa Cruz Biotechnology Inc, USA) karta, Indonesia. The cells were cultured in Ro- and anti-TNF-α ab6671 antibody (Abcam, swell Park Memorial Institute medium (RPMI) USA), and a secondary antibody reagent univer- (Sigma Aldrich, USA) containing 10% fetal bo- sal detection kit (Paramount BGPD-0100, Bio- vine serum, 0.5% fungizone (Amphotericin B) Gear, USA). VEGF and TNF-α expression was and 2% penicillin-streptomycin and incubated assessed at the anatomical pathology laboratory in Heraeus Hera Cell incubator (Thermo Fischer using light microscope by two pathologists and Scientific, Waltham, MA, USA) at 37°C and 5% was graded as strong, moderate, or weak by the CO2. They were harvested with trypsin–EDTA intensity of brown staining. (Sigma Aldrich, USA) after single layer cells

416 Edianto et al. Senduduk Fruit Inhibits VEGF and TNF-α

RESULTS Spectrophotometry revealed that the EESF contained anthocyanin, quercetin, flavone, katheci- ne, and other components. EESF contains 115.70 mg/kg anthocyanin, and the derivatives are delphinidin and cyanidin 3-glucoside. In this research, HeLa cells were cultured for 24 hours in RPMI media that contain EESF in various concentrations. Cell viability was measured using the MTT assay. The highest concentration starts with 500 μg/mL, half of it was diluted at the next concentration to get IC50. And it turns out that the highest dose only inhibits the viability of HeLa cells by around 70%. The results of the MTT assay of EESF cytotoxicity at 500–7.8125 µg/mL were expressed as a percentage of viable HeLa cells

(Figure 2). IC50 of the EESF calculated by extrap- olation using Excel programme was 956 μg/mL.

Figure 2. Methylthiazol-tetrazolium (MTT) assay of citotoxicity as a percentage of viable HeLa cells; EESF, ethanol extract of senduduk fruit The presented research assessed anticancer benefits, especially the anti-angiogenesis and anti-inflammation effects. Because IC50 was high, VEGF and TNF-α expression were evaluated in HeLa cells treated with 1× and 2× IC level. Figure 3. Vascular endothelial growth factor (VEGF) expres- 50 sion in HeLa cells. A) in control cultures, B) cultured with 1x

By giving these plant extracts to HeLa cell cul- EESF IC50, C) cultured with 2x EESF IC50 tures with concentrations of 1X and 2X IC50, Without the administration of extracts, HeLa respectively, and as a control without extrac- cells only showed the same moderate expression ting, through immunocytochemical examination of TNF-α as if given extracts with a concentrati-

VEGF and TNF-α expression were assessed. on of 1X IC50. Increasing the dose of extract tur- Two pathologist gave the same examination ned down its expression. results, and there was no need for Kappa test. TNF-α expression in HeLa cell control, EESF 1× HeLa cells without administration of the extract IC50, and 2× IC50 cultures is shown in Figure 4. express strong VEGF, and their expression decre- The intensity was moderate in cells control cul- ased with increasing concentration of the extract. tures without EESF and also with 1× EESF IC50, VEGF expression in HeLa cell control, EESF 1× and weak in cultures with 2× EESF IC50. IC50, and 2× IC50 cultures is shown in Figure 3. VEGF expression was assessed by the intensity DISCUSSION of brown staining in HeLa cells. The intensity IC of 956 µg/mL was estimated in HeLa cell was strong in cells from control cultures without 50 cultures containing EESF at concentrations ran- EESF, moderate in cultures with 1× EESF IC , 50 ging from 500 µg/mL to 7.8125 µg/mL obtained and weak in cultures with 2× EESF IC50.

417 Medicinski Glasnik, Volume 17, Number 2, August 2020

it extract contains delphinidin, cyanidin 3-glucoside, quercetin, flavon, epicatechin and luteolin, and it is in line with previous reports that plant extracts containing phenol compounds like flavonoids and anthocyanin have anticancer activity (22). Amatori et al. reported that extracts of the alba variety of strawberries that contained phenol compounds inhibited the growth of A17 breast cancer cells compared with normal control cells (28). The extract used in this study contained constituents other than anthocyanin, which may contribute to its activity. Our study showed that the greater EESF had stronger effect on VEGF expression. Samad et al. reported that a methanol extract of Senduduk fruit had antiproliferative activity against HeLa cells but did not inhibit the proliferation of 3T3 cells, indicating its lack of toxicity to normal cells; the

high IC50 level observed will most likely to be harmless to normal cells (29). Angiogenesis is a hallmark cancer characteristic and is needed for continuing tumour development (10,30). Continuous blood vessels growth is required to supply the oxygen and nutrient support for increased growth of more than 1–2 mm3 in tumour size. VEGF stimulates angiogenesis and VEGF expression is triggered by hypoxia (31,32). Inflammation is one of the complex biological responses to the damages caused either by injury or microbial infection. The role of inflammation in tumorogenesis is now widely accepted. In many cases, chronic inflammation in the micro- Figure 4. Tumour necrosis Factor-α (TNF-α) expression in HeLa environment is essential for the initiation and cells. A) in control cultures, B) cultured with 1x ethanol extract progression the cancers (32). TNF-α is involved of senduduk fruit (EESF IC ) C) cultured with 2x EESF IC 50 , 50 in maintenance of the inflammation and host by serial dilution that reduced the dose by half at defence. However, there is a ‘dark side’ to this each step. It was much than 1.781 ± 1.2 µg/mL powerful cytokine in the pathological process reported by Alnajar et al. (26) for the inhibition of such as malignant disease (11). peripheral blood mononuclear-cell proliferation. A high dose of the EESF was required to inhibit VEGF and TNF-α expression in this in vitro stu- Roslen et al. (21) found that IC50 of a Senduduk fruit extract for Artemia salina brine shrimp lar- dy. Because the oral bioavailabilty of flavonoids is vae was 89.947 ppm. low, it will be difficult to achieve this high dose in Flavonoids, tanin, anthocyanin, and phenol com- an in-vivo studies. In conclusion, ethanol extract of pounds are among the bioactive components of Senduduk fruits reduced VEGF and TNF-α expre- Senduduk fruit with pharmacologic benefits (22). ssion in HeLa cells in a dose-dependent manner. Anthocyanins are water-soluble plant pigments FUNDING that are responsible for the red, blue, purple, and black colour of fruits. They are ubiquitous in plants No specific funding was received for this study and are the most abundant flavonoid constituent of TRANSPARENCY DECLARATION fruits (27). The study results through spectrofoto- metry examination revealed that the Senduduk fru- Conflicts of interest: None to declare.

418 Edianto et al. Senduduk Fruit Inhibits VEGF and TNF-α

REFERENCES 1. World Health Organization. Global health observa- 19. Hui C, Bin Y, Xiaoping Y, Long Y, Chunye C, Man- tory. Geneva: World Health Organization 2018. tian M. Anticancer activities of an anthocyanin-rich 2. Omran AR. The epidemiologic transition . A theory extract from black rice against breast cancer cells in of the epidemiology of population change. Milbank vitro and in vivo. Nutr Cancer 2010; 62:1128-36. Mem Fund Q 1971; 49:509-38. 20. Sak K. Cytotoxicity of dietary flavonoids on diffe- 3. Gersten O, Wilmoth JR. The cancer transition in Ja- rent human cancer types. Pharmacogn Rev 2014; pan since 1951. Demogr Res 2002; 7:271-306. 8:122-46. 4. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre 21. Roslen NA, Alewi NAM, Ah reamada H, Rasad LA, Jemal A. Global Cancer Statistic: GLOBOCAN MSBA. Cytotoxicity screening of Melastoma mala- estimates of incidence and mortality worldwide for bathricum extracts on human breast cancer cell line 36 cancers in 185 countries. CA Cancer J Clin 2018; in vitro. Asian Pac J Trop Biomed 2014; 4:545-48. 68:394-424. 22. Aslam MS, Ahmad MS, Ahmad MA. An up-to-date 5. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tiu- review on phytochemical constituents and pharma- lent, Jemal A. Global cancer statistic, 2012. CA Can- cological activities of Melastoma malabathricum. cer J Clin 2015; 65:65-87. Int J Pharm Sci Res 2017; 8:76-91. 6. Marnitz S, Kohler C. Advanced Cervical Cancer. 23. Anuar N, Adnan AFM, Saat N, Taha RM. Opti- In: Ayhan A, Reed N, Gultekin M, Dursun P, eds. mazitation of extraction by using response surface Textbook of Gynaecological Oncology. 2nd ed. An- methodology, purification and identification of ant- kara: Gunes Publishing, 2012:361-71. hocyanin pigments in Melastoma malabathricum 7. Carter J, Philp S, O’connell R. A 5-year review of fruit. Sci World J 2013; 1-10. gynecological oncology patients managed by fast 24. Nath D, Choudhury MD, Muzamber PB, Mitra A. tract surgery program. Int J Clin Med 2014 2014; Phytochemical profilling of Melastoma malabat- 5:36-41. hricum linnaeus (Melastomaceae): an ethnomedi- 8. Xu H, Zhou Z, Dong J, Lei M. Suppression of cer- cinally important plan of eastern Himalaya. Pleione vical cancer cell survival by ursulic acid extracted 2014; 8:478-85. from Catalpa bungei leaves. Pharmacogn Mag 2018; 25. Serna DMO, Martinez JHI. Phenolic and polyphe- 14:425-31. nolic from Melastoma malabathricum species. Mo- 9. Wang RA, Li QL, Li ZS, Zheng PJ, Zhang HZ, Hu- lecules; 2015:17818-47. ang XF, Chi SM, Yang AG, Cui R. Apoptosis drives 26. Alnajar ZAA, Abdullah MA, Ali HM, Alshawsh cancer cells proliferate and metastasize. J Cell Mol MA, Hadi AAH. Acute toxicity evaluation, antibac- Med 2013; 17:2015-11. terial, antioxidant and immunomodulatory effects 10. Hanahan D, Weinberg RA, Hallmarks of cancer: the of Melastoma malabathricum. Molecules 2012; next generation. Cell 2011; 144:646-74. 17:3547-59. 11. Zhong H, Bowen JP. Antiangiogenesis drug design: 27. Khoo HE, Azlan A, Tang ST, Lim SM. Anthocya- multiple pathways targeting tumor vasculature. Curr nidins and anthocyanins: colored pigment as food Med Chem 2006; 13:849-62. pharmaceutical ingredients, and the potential health 12. Kukreja I, Kapoor P, Deshmukh R, Kulkarni V. benefits. Food Nutr Res 2017; 61:1-21. VEGF and CD34: A correlation between tumor an- 28. Amatori S, Mazzoni L, Alvarez-Suarez JM, Giam- giogenesis and microvessel density – an immunohi- perri F, Gasparrini M, Forbes-Hernandez TY, Afrin stochemical study. J Oral Maxillofac Pathol 2013; S, Prorenzano AE, Persico G, Mezzeti B, Augusto 17:367-73. A, Faneli M, Battino M. Polyphenol-rich strawberry 13. Mantovani A. Tumor-associated macrophage in extract (PRSE) shows in vitro and in vivo biological cancer-related inflammation. Immunotherapy 2011; activity against invasive breast cancer cells. Sci Rep 3:21-2. 2016; 6:1-14. 14. Balkwill F. TNF-α in promotion and progression of 29. Samad NA, Nik NNS, Kamal M, Yahya N, Aziz MY, cancer. Cancer Metastasis Rev 2006; 25:409-16. Zain NNM, Yusoff NAM, Lim V. Ethnobotanical, 15. Fimognari C, Lenzi M, Hrelia P. Chemopreventi- phytochemical and pharmacological aspects of Me- on of cancer by isothiocyanates and anthocyanins: lastoma sp. Mal J Med Health Sci 2018; 14:153-63. mechanisms of action and structure activity relati- 30. Mirossay L, Varinska L, Mojzis J. Aantiangiogenic onship. Curr Med Chem 2008; 15:440-7. effect of flavonoids and 16. Moga MA, Dimienescu OG, Arvatescu CA, Mi- 31. Zetter BR. Angiogenesis and tumor metastasis. ronescu A, Dracea L, Ples L. The role of natural Annu Rev Med 1998; 49:407-24. polyphenols in the prevention and treatment of cer- 32. Chung SS, Okobi Q, Adekoya D, Atefi M, Clarke vical cancer – an overview. Molecules 2016; 21:1- O, Dutta P, Vadgama JV. Proinflammatory cytokines 32. IL-6 and TNF-α increased telomerase activity thro- 17. Batra P, Sharma AK. Anti-cancer potential of flavo- ugh NF-κB/STAT1/STAT3 activation and witheferin noids: recent trends and future perspectives. 3 Bio- a inhibited the signaling in colorectal cancer cells. tech 2013; 3:439-59. Mediat Inflamm 2017; 2017-5958429. 18. Karikas GA. Chemoprevention molecular and biochemical mechanisms involved in cancer control and management. Health Sci J 2011; 5:149-56.

419 ORIGINAL ARTICLE

Snakehead fish extract as an enhancer of vascular endothelial growth factor and nitric oxide levels in cerebral angiogenesis: an insight of stroke therapy

Iskandar Nasution1, Hasan Sjahrir1, Syafruddin Ilyas2, Muhammad Ichwan3

1Department of Neurology, School of Medicine, 2Department of Biology, Faculty of Mathematics and Natural Sciences, 3Department of Pharmacology, School of Medicine; Universitas Sumatera Utara, Indonesia

ABSTRACT

Aim To assess the effect of snakehead fish extract administration in angiogenesis focusing on the level of vascular endothelial growth factor (VEGF), nitric oxide (NO) and VEGF receptor 2 (R2) expression is ischemic stroke models.

Methods An experimental study was conducted on 5 groups of ischemic stroke rats models: Group K- without carotid artery ligation, Group K+ with artery ligation, Group P1 with artery ligation and administration of 200 mg/day extract, Group P2 with artery ligation and 400 mg/day extract, and Group P3 with artery ligation and 800 mg/day extract. The VEGF expression and NO levels Corresponding author: were assessed on day 3. Iskandar Nasution Results Snakehead fish extract significantly increased VEGF Medical Doctor, Neurologist levels along with increasing doses, in which the highest VEGF Department of Neurology, School of level was observed in P3 group (361.7±40.2; p<0.001). The NO Medicine, Universitas Sumatera Utara level also increased along with an increasing dose of snakehead fish extract, in which the highest NO level was found in P3 group Jalan Persatuan no 29 A, Sei Agul, (59.43±0.88 μmol/gr; p<0.001). The VEGFR2 expression also in- Medan, Sumatera Utara, creased significantly after snakehead extract administration along Indonesia, 20117 with increasing doses (p<0.001) in which administration of 800 Phone: +62 81 2650 3305; mg extract yielded the highest VEGFR2 expression compared E-mail: [email protected] with lower doses (17.7 vs. 15.6; p<0.001) ORCID ID: https://orcid.org/ 0000-0003- Conclusion Snakehead fish extract administration increased angi- 1342-0613 ogenesis process marked by an increased level of VEGF, NO and VEGFR2 expression in ischemic stroke rat models. Original submission: Key words: arginine, nitric oxide, cerebral ischemia 27 November 2019; Revised submission: 23 January 2020; Accepted: 06 January 2020 doi: 10.17392/1107-20

Med Glas (Zenica) 2020; 17(2): 420-424

420 Nasution et al. Snakehead fish extract for angiogenesis

INTRODUCTION Methods Various studies have raised the possibility of Stroke induction was done by the unilateral ca- snakehead fish extract as a potential treatment in rotid artery occlusion method at the internal and ischemic stroke (1,2). Snakehead fish extract is external carotid arteries of rats on the left side rich in arginine. Arginine acts as a coenzyme in (8,9). Distribution of microglia cells in occlusion the formation of nitric oxide (NO), an important area of the internal and external carotid arteries mediator for vasodilation, from endothelial nitric was assessed in all groups. Neurological deficits oxide synthase (eNOS) (3-5). Several studies have in rats were assessed within 24 hours after an is- also found that snakehead fish extract may increa- chemic event (10,11). se vascular endothelial growth factor (VEGF), in The dosage of snakehead fish extract (VipAlbu- which binding of VEGF to its receptor, VEGF-Re- min, Royal Medicalink Pharmatab, Makassar, ceptor-2 (VEGFR2) is involved in an increase of Indonesia) given was calculated based on animal proliferation of endothelial cells, and are expected equivalent dose (AED) formula (11). Snakehead to enhance angiogenesis in patients with ischemic fish extract 100% were given intraorally by inser- stroke (6,7). However, studies focusing on the de- ting feeding tube daily. The focus of this research finite role of snakehead fish extract in cerebral anis the formation of vascular (histological) structu- giogenesis are still very limited. res of brain tissue in rats ischemic stroke models This study aimed to assess the effect of snakehead after the treatment with Snakehead fish extract. fish extraction administration in angiogenesis focu- The NO, VEGF and VEGFR2 levels were asse- sing on the level of VEGF, NO and VEGF receptor ssed on the third day after stroke, according to 2 (R2) expression is ischemic stroke models. previous studies (9). The VEGF and NO level measurement was done METHODS from rat plasma on the third day after stroke, e.g. after occlusion of the internal and external carotid Study design arteries. Histological examination of brain tissue This experimental study was conducted in Biomo- was done by routine staining technique of hema- lecular Laboratory, Brawijaya University Malang- toxylin – eosin. The brain tissue of rat’s ische- Indonesia from January – April 2019. The study mic stroke models in each group was processed was conducted on 5 groups of ischemic stroke for immunohistochemistry examination. This rats models to assess the effect of snakehead fish showed that there was an activation of the immu- extract administration in angiogenesis focusing nological process. Measurement of VEGFR2 on the level of vascular endothelial growth factor expression was carried out on immune labelled (VEGF), nitric oxide (NO) and VEGF receptor 2 tissue (Ab182981 anti-VEGFR2 antibody Flk-1 (R2) expression is ischemic stroke models. A-3: sc-6251, and ELISA kit LS-F542, LifeS- The experimental animals used were 25 Sprague pan-Bio Life Bioscience Inc, Seattle, USA), then - Dawley strain rats (Rattus norvegicus sp.) with viewed using a fluorescence microscope (Nikon body weight of 305 - 425 grams. Twenty five rats E-100, with attachment SonyA7 Camera, Tokyo, were divided randomly into 5 treatment groups, Japan). Cells with increased area of vascular for- each group consisted of 5 rats: group K (-) was mation for each antibody were counted in six ran- control group without carotid artery ligation, and domly selected fields from one optical field on without extract administration, group K (+) was microscope located along the border of the ische- control group with artery ligation, but without mic lesion. Two pieces of brain tissue specimens extract administration; groups P1, P2 and P3 all of each animal were counted in random order, underwent artery ligation and were then treated and the average value was determined (7). with 200 mg, 400 mg and 800 mg/day, respecti- The VEGFR2 examination and calculation was vely, snakehead extract for three days. observed by looking at the brown colour in the All animal procedures were based on the Helsinki cytoplasm of cells, then calculated according to Declaration and approved by the Animal Research a formula by Soini et al. (13) and Pizem et al. Ethics Commission, Universitas Sumatera Utara. (14). To guarantee representation and reduce the

421 Medicinski Glasnik, Volume 17, Number 2, August 2020

errors, it was needed to observe approximately dose increase (Table 1). The level of NO in group 20 fields with 1000x magnification, each contai- P3 was the highest (59.4 µM) compared with P2 ning approximately 1500 cells (12-14). (53.6 µM; p˂0.001), P1 (51.9 µM; p=0.001) and K (+) group (control group without extract admi- Statistical analysis nistration) group (40.2 µMq; p˂0.001).

All data were analysed by investigators for which Table 1. Level of vascular endothelial growth factor (VEGF), the experimental group was blinded. An analysis nitric oxide (NO) and VEGF receptor 2 (R2) expression in of increased levels of VEGF, NO and VEGFR2 each treatment group VEGFR-2 expression was done by ANOVA test and post VEGF (pg/mL) NO (µM) Group* (No of cells/field) hoc comparison, in which p<0.05 was stated as Mean ± SD p Mean ± SD p Mean ± SD p significant (15). K (-) 202.1 ± 54.7 25.4 ± 2.54 Ref 7.0 ± 2.44 Ref K (+) 128.0 ± 34.0 0.07 40.2 ± 3.01 <0.001 5.9 ± 2.3 0.96 RESULTS P1 182.0 ± 28.7 0.28 51.9 ± 5.3 0.001 12.0 ± 2.47 0.01 P2 267.4 ± 47.3 <0.001 53.6 ± 4.75 <0.001 15.6 ± 2.79 <0.001 The results showed that in the occlusion area of P3 361.7 ± 40.2 <0.001 59.4 ± 0.87 <0.001 17.7 ± 2.95 <0.001 the internal and external carotid arteries, there *K (-),control group without carotid artery ligation, and without extract administration, K (+), control group with artery ligation, but without was an increase of the number of microglia cells extract administration; P1, P2 and P3, treated with 200 mg, 400 mg and in the K (+) group. This showed that there was 800 mg/day, respectively, snakehead extract for three days activation of the immunological process in the K (+) group (Figure 1). A significant increase of VEGFR2 expression (no of cells) in the groups P1 (12.0; p=0.01), P2 (15.6; p˂0.001) and P3 (17.7; p˂0.001) was noticed comparing to K+ group (control group without extract administration) (Table 1). The increase of VEGFR2 expression occurred along with the dose extract indicating that the vascularization process began after the administration of snakehead fish extract for 3 days of exposure (Figure 2).

Figure 1. Histological picture of brain tissue due to occlusion of the internal and external carotid arteries (hematoxilen- eosin staining). A) Normal tissue - control, A1) 400x magnification normal tissue, B) Occlusion of the artery, B1) 400x magnification in the tissue with occlusion artery. White arrows indicated microglia cells and black arrows indicated neuron cells (Nikon E-100 microscope with magnification of 100x and 400x; photomicrographs were carried out with a Sony ICLE-A7 camera)

A significant increase was found in plasma VEGF level in rat groups treated with snakehead fish extract (P1, P2 and P3) compared to K (+) group (control group without extract administration). The increase of VEGF occurred along Figure 2. Histological representation of vascularization in the with an increasing dose (Table 1). The level of groups. K (-) control group without carotid artery ligation, and VEGF in group P1 (200 mg extract administrati- without extract administration, K (+), control group with ar- on), group P2 (400 mg) and group P3 (800 mg) tery ligation, but without extract administration, P1, treated with 200 mg of snakehead fish extract, P2, treated with 400 was 182.0 (p=0.28), 267.4 (p˂0.001) and 361.7 mg of snakehead fish extract, P3 treated with 800 mg of pg/mL (p<0.001). snakehead fish extract. Black arrows indicated vascular for- The administration of snakehead fish extract pro- mation by VEGFR2 expression (Nikon E-100 microscope with magnification of 100x and 400x; photomicrographs were carried duced also an increase in NO level, along with the out with a Sony ICLE-A7 camera)

422 Nasution et al. Snakehead fish extract for angiogenesis

DISCUSSION In hypoxic state, increased VEGF and VEGFR2 expression stimulates angiogenesis (30), proved Angiogenesis is a process of vascular compensa- by the presence of vascular formation area in bra- tion or protection, which is also a target of stro- in tissue. This study found that after the admini- ke therapy (16,17). An immunohistochemistry stration of 200 mg dose of snakehead fish extract, analysis indicates angiogenesis begins to be acti- there was a significant increase in VEGFR2 ve on day 2 to 7 after stroke attack (9). expression compared to control group, and the Snakehead fish extract contained many -impor increase of VEGFR2 expression occurred along tant amino acids, particularly arginine. Arginine, with dose extract increase. which will be oxidized to citrulline, is a cofactor After all, it can be concluded that administration in the conversion of eNOS to NO. The NO has of snakehead fish extract for 3 days after expo- a protective effect by reducing platelet aggrega- sure to ischemic stroke accelerated the ongoing tion, and increases vasodilation, blood flow and vascularization process. regulates vascular tone (18-22). It in turn promotes blood flow and tissue oxygenation (23). A This is by far the first study with histologically study by Zhang et al. showed that administered proven effect of administered snakehead fish exogenous NO/donor NO significantly increased extract to increase cerebral angiogenesis. Howe- vascularization and proliferation of cerebral en- ver, it is important to understand that VEGF, NO dothelial cells (24). and VEGFR2 expression are highly determined by the hypoxic condition. This study did not asse- This study found VEGF level increased signifi- ss the levels of inducible-NOS (iNOS) that are cantly with increasing doses of snakehead fish known to play a role in stimulating NO expendi- extract. Interestingly, VEGF levels in K (-) group ture. Knowing the iNOS and NO ratio will help (without carotid artery ligation) were higher com- to make clear that NO rise comes from arginine pared with K (+) group (control group with artery and not from iNOS. ligation, but without extract administration). The decrease of VEGF level in the K (+) group proba- The administration of snakehead fish extract can bly occurred because it was still in the acute phase increase the level of VEGF, NO and VEGRF2 (3 days). Longer observation period is needed to expression in cerebral angiogenesis process of obtain increased plasma VEGF levels. ischemic stroke rat models. The VEGF is very important for endothelial cell ACKNOWLEDGMENTS function associated with angiogenesis. The signal transduction initiated by VEGF/VEGFR2, lea- The authors thank staff of Molecular Biology ding to proliferation and formation of new blood Laboratory, School of Medicine, University vessels has multiple protective effects in regula- Brawijaya Malang, Indonesia for their support ting angiogenesis and neurogenesis (25-27). to this research. The VEGF-VEGFR system is an important target FUNDING for pro-angiogenic therapy (28). It is involved in the growth of new blood vessels, endothelial cell No specific funding was received for this study. mitogenesis, vasodilatation (via the NO depen- TRANSPARENCY DECLARATION dent pathway) and vascular permeability (29,30). Conflicts of interest: None to declare.

REFERENCES

1. Hoeben A, Landuyt B, Highley MS, Wildiers H, Van 3. Rahayu P, Marcelline F, Sulisttyaningrum E, Suhar- Oosterom AT, De Bruijn EA. Vascular endotheli- tono MT. Potential effect of striatin (DLBS0333), al growth factor and angiogenesis. Pharmacol Rev a bioactive protein fraction isolated from Channa 2004; 56:549–80. striata for wound treatment. Asian Pac J Trop Bio- 2. Oentaryo G, Istiati I, Soesilawati P. Acceleration med 2016; 6:1001-7. of fibroblast number and FGF-2 expression using 4. Pudjonarko D, Abidin Z. Clinical outcome and argi- Channa striata extract induction during wound he- nine serum of acute ischemic stroke patients supple- aling process: in vivo studies in Wister rats. Dental mented by snakehead fish extract. In IOP Conferen- Journal 2016; 49:125-32. ce Series: Earth and Environmental Science 2018; 116:12-28.

423 Medicinski Glasnik, Volume 17, Number 2, August 2020

5. Mustafa A, Sujuti H, Permatasari N, Widodo MA. 18. Majewska I, Gendaszewska-Darmach E. Proangio- Determination of nutrient contents and amino acid genic activity of plant extracts in accelerating wound composition of Pasuruan Channa striata extract. healing - a new face of old phytomedicines. Acta Bi- IEESE International Journal of Science and Techno- ochim Pol 2011; 58:449–60. logy 2013; 2:1-11. 19. Gurtner GC, Werner S, Barrandon Y, Longaker MT. 6. Greenberg D, Jin K. Vascular endothelial growth Wound repair and regeneration. Nature 2008; 453: factors (VEGF) and stroke. Cell Mol Life Sci 2013; 314–21. 70:1753–61. 20. Chen J, Zacharek A, Zhang C, Jiang H, Li Y, Rober- 7. Sköld MK, Gertten CV, Sandbergnordqvist AC, ts C, Lu M, Kake A, Chopp M.. Endothelial nitric Mathiesen T, Holmin S. VEGF and VEGF receptor oxide synthase regulates brain-derived neurotrophic expression after experimental brain contusion in rat. factor expression and neurogenesis after stroke in J Neurotrauma 2005; 22:353-67. mice. J Neurosci 2005; 25: 2366-75. 8. Indra MR, Gasmara CP. Metode UCAO (Unila- 21. Chen K, Pittman R, Popel A. Nitric oxide in the vas- teral cerebral artery occlusion) meningkatkan ka- culature: where does it come from and where does dar MMP-9 jaringan otak pada model tikus stroke it go? A quantitative perspective. Antioxid Redox iskemik (The UCAO methods succesfully increased Signal 2008; 10:1185-95. MMP-9 level in brain tissue of ischemic stroke rat 22. Sahid N, Hayati FR. Snakehead consumption enhan- models) [Indonesian]. Malang Neurology Journal ces wound healing? From tradition to modern cli- 2016; 2:1-5. nical practice: a prospective randomized controlled 9. Gandin C, Widmann C, Lazdunski M, Heurteaux C. trial. Evid Based Complement Altternat Med 2018; Mlc901 favors angiogenesis and associated recovery 2018:3032790. after ischemic stroke in mice. Cerebrovasc Dis 2016; 23. Carmeliet P, Collen D. Molecular basis of angioge- 42:139–54. nesis. Role of VEGF and VE-Cadherin. Ann. N. Y. 10. Bederson, JB, Pitts LH, Tsuji M, Nishimura MC, Acad Sci 2000; 902:249–64. Davis RL, Bartkowski H. Rat middle cerebral artery 24. Zhang R, Wang L, Zhang L, Chen J, Zhu Z, Zhang ZC. occlusion: evaluation of the model and development Nitric oxide enhances angiogenesis via the synthesis of a neurologic examination. Stroke 1986; 17:472-6. of Vascular Endothelial Growth Factor and cGMP af- 11. Siswanto A, Dewi N, Hayatie L. Effect of haruan ter stroke in the rat. Circ Res 2003; 92:308-13. (Channa striata) extract on fibroblast cells count in 25. Xiao B, Wang S, Yang G, Sun X, Lin L, Cheng J. wound healing. J Dentomaxillofac Sci 2016; 1:89-94. HIF 1α contributes to hypoxia adaptation of the na- 12. Norhalifah N, Rahmawanty D, Nurlely N. Uji efek- ked mole rat. Oncotarget 2017; 8:109941-51. tivitas ekstrak air ikan haruan (Channa striata) asal 26. Sen CK, Khanna S, Venojarvi M, Trikha P. Cooper- Kalimantan selatan terhadap bleeding time dan induced vascular endothelial growth factor expre- clotting time secara in vivo (Efficacy of snakehead ssion and wound healing. Am J Physiol Heart Circ fish extract (Channa striata) from South Borneo on 2002; 282:1821-7. bleeding time and clotting time in vivo) [Indonesi- 27. Talwar T, Vashanta M, Srivastava P. Role of vascular an]. Media Farmasi 2016; 13:237-49. endothel growth factor and other growth factors in 13. Soini Y, Paakko P, LehtoVP. Histopathological eva- post stroke recovery. Ann Indian Acad Neurol 2014; luation of apoptosis in cancer. Am J Pathol 1997; 17:1-6. 153:1041-8. 28. Yao RQ, Zhang L, Wang W, Li L. Cornel iridoid 14. Pizem J, Cor A. Detection of apoptosis cells in tumour glycoside promotes neurogenesis and angiogenesis paraffin section. Radiol Oncol 2003; 37: 225-32. and improves neurological function after focal cere- 15. Notoadmojo, S. Metodologi penelitian kesehatan bral ischemia in rats. Brain Res Bull 2009; 79:69–76. (Research methodology on health sciences) [Indone- 29. Martin A. Imaging of perfusion, angiogenesis and sian] 2012 Jakarta: PT. Rineka Cipta, 2012. tissue elasticity after stroke. J Cerebr Blood F Met 16. Liu J, Wang Y, Akamatsu Y, Lee C, Stetler RA, 2012; 1296-507. Lawton YG. Vascular remodeling after ischemic 30. Simon M, Gordon E, Welsh C. Mechanism and regu- stroke: mechanism and therapeutic potentials. Prog lation of endothelial VEGF receptor signalling. Nat Neuobiol 2014; 115:138-56. Rev Mol Cell Biol 2016; 17:611-25. 17. Chopp M, Li Y, Chen J, Zhang R, Zhang Z. Brain 31. Das A, Kim Y, Youn S, Varadarajan S. Cysteine oxi- repair and recovery from stroke. Eur Neurol 2008; dation of Cu importer CTR1 promotes VEGF signa- 3:2-5. ling and angiogenesis. Circulation 2016; 134:1.

424 ORIGINAL ARTICLE

Combination of vitamin A and D supplementation for ischemic stroke: effects on interleukin-1ß and clinical outcome

Alfansuri Kadri1, Hasan Sjahrir1, Rosita Juwita Sembiring2, Muhammad Ichwan3

1Department of Neurology, 2Department of Clinical Pathology, 3Department of Pharmacology; Faculty of Medicine, Universitas Sumatera Utara, Indonesia

ABSTRACT

Aim Accumulated evidence suggests that vitamin A and D agonists can alleviate the development of atherosclerosis. Therefore, the aim of this study was to determine the effect of vitamin A and D combination supplement on interleukin-1β (IL-1β) and clinical outcome in ischemic stroke.

Methods A single-blind, randomized controlled trial was conducted on ischemic stroke patients at Adam Malik Hospital between March 2018 to February 2019. The patients were randomized into 4 groups of the treatment consisting of supplementation using vitamin A or D only, combination of vitamin A and D, and placebo group, all given for 12 weeks. Clinical outcome was determined using the National Institute of Health Stroke Scale (NIHSS). At the time of admission and after the treatment was completed, all Corresponding author: patients were measured for vitamin A, vitamin D, and IL-1β serum level, and NIHSS score. Alfansuri Kadri Department of Neurology, Results From the total of 120 patients, in the combination group Faculty of Medicine, there were significant increments on both vitamin A (p=0.04) and Universitas Sumatera Utara vitamin D (p=0.01) serum level after 12 weeks of the treatment, compared to the other groups. In conjunction, IL-1β serum le- Kompleks Taman Setia Budi Indah Blok G, vel showed a significant decrement in the combination group No. 17, 20132 Medan, Sumatera Utara, (p<0.001). Lastly, the biggest improvement of NIHSS could be Indonesia seen in the combination group, which was marked by the highest Phone: +62 81 260 100 64; decrement of NIHSS score (p<0.001). E-mail : [email protected] Conclusion Administration of combination of vitamin A and D ORCID ID: http://orcid.org/ 0000-0002- supplementation can significantly increase vitamin A and D serum 7420-0588 level, decrease IL-1β serum level, and ultimately improve clinical outcome in ischemic stroke patients.

Original submission: Key words: cerebrovascular disorders, cytokines, nutrients, stroke/drug therapy 06 February 2020; Revised submission: 08 May 2020; Accepted: 28 May 2020 doi: 10.17392/1137-20

Med Glas (Zenica) 2020; 17(2): 425-432

425 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION ar receptor superfamily, mediates the action of the biological form of vitamin D (1.25[OH] D ). This Stroke is one of the vascular diseases of the brain 2 3 superfamily comprises more than 60 nuclear recep- which is categorized as the third leading cause of tors for lipophilic ligands such as steroid/thyroid death after heart disease and malignancy, and is hormones, vitamin A and vitamin D (13). The VDR the number one cause of long-term disability in activation by 1.25(OH) D or analogues has been the world (1,2). Mortality and morbidity of ce- 2 3 reported to show a protective effect of atherosclero- rebrovascular disease are a large socio-economic sis in several animal experiments. After activation, burden, and also constitute a very large burden for VDR binds specifically to the Vitamin D Respon- global health services (3). Ischemic or hypoxic se Element (VDRE) as a heterodimer with the X injury often causes irreversible brain damage and Retinoid Receptor (RXR) and then modulates the is a leading cause of disability and death throu- expression of the target gene (3,14). The retinoid X ghout the world (4). Much effort has been made receptor (RXR) is a type of nuclear receptor that is to find pharmacological therapies to reduce brain activated by 9-cis-retinoic acid, a form of vitamin A injury due to cerebral ischemia (5). In recent ye- (retinol) and is a first-generation retinoid (15-18). ars, vitamins, minerals and other essential nutrients show a great potential as primary therapies Recent evidence suggests that vitamin A (retinol) and attract much attention from researchers. This is also a risk factor for cardiovascular disease and therapy, known as Nutraceutical therapy, can be mortality (19). Plasma retinol levels are inversely an excellent choice for the treatment of patients proportional to the risk of cardiovascular disease with brain injuries (6,7). mortality (19). Research has shown that patients The process of atherosclerosis is characterized with coronary heart disease or atherosclerosis by arterial remodelling that causes the accumu- show significantly lower levels of retinol than con- lation of progressive subendothelic plaques thro- trols (19,20). There is increasing evidence that the ugh a series of complex cellular processes that retinoic acid (RA) signalling pathway provides an occur in the walls of the arteries with inflamma- important mechanism for regulation of the blood tion playing an important role in its various pha- brain barrier (BBB) in the neurovascular system. ses (8). In the past decade, a number of studies The protective effect of RA on the BBB further have examined the relationship between serum reduces cerebral damage in ischemic stroke (21). vitamin D concentrations and the risk of cere- The development of future clinical experiments brovascular event. The results show that hypo- can identify RA as a potential target for the tre- vitaminosis D can cause atheroma and vitamin D atment of neurovascular disease (21,22). supplementation can prevent the risk of vascular Retinol and its derivatives carry out their biolo- disease (3,9,10). gical actions via specific nuclear receptors that Laboratory data also show a potential association regulate gene transcription. The RA receptors of vitamin D deficiency as a risk factor for stroke, can also interact with other nuclear receptors that through systemic and vascular inflammatory sti- have neuroprotective effects such as vitamin D3, mulation that causes atherogenesis both directly which is neuroprotective against stroke (5,23,24). and indirectly (3,11). Studies in experimental ani- Overall, accumulated evidence suggests that VDR mals given a low vitamin D diet for 8 weeks, com- and RXR agonists can alleviate the development pared with the control group showed that infarct of atherosclerosis through inhibition of oxidative volume and neurological disorders in the low vi- and inflammatory stress (24,25). Nevertheless, the tamin D diet group were significantly greater than role of combined vitamin A and D as beneficial those in the control group. This occurs due to the nutrients in health as well as disease such as ische- up-regulation of ischemia-induced interleukin-6 mic stroke is still a controversial topic. In addition, (IL-6) on a low vitamin D diet (3,11,12). there have been no studies that have examined the Vitamin D has roles in a variety of biological acti- effect of combined vitamin A and D supplementa- ons and most of these actions of vitamin D are con- tion on ischemic stroke patients. sidered to be exerted through the nuclear vitamin D The aim of this study was to determine the effect receptor (VDR)-mediated control of target genes. of the combination of vitamin A and D supplemen- Vitamin D Receptor (VDR), a member of the nucle- tation on changes in IL-1β and also clinical outco-

426 Kadri et al. Vitamin A and D and ischemic stroke

me using the National Institutes of Health Stroke group patients were given vitamin D3 50,000 IU

Scale (NIHSS) in ischemic stroke patients. per week, using vitamin D3 soft capsule (50,000

IU vitamin D3 per capsule). The combination MATERIALS AND METHODS group was given 50,000 IU vitamin A combined with 50,000 IU vitamin D, also given once a Patients and study design week, and the placebo group was given capsules This was a prospective cohort, single-blind, place- containing saccharose lactis as the placebo. Both bo-controlled, pre and post-test study. The patients the supplement and placebo were given in capsu- were selected from the Stroke Unit of the Neuro- les of the same size, colour, and weight to ensure logy Department, Adam Malik General Hospital that patients were blinded toward the treatments. in Medan - Indonesia, in the period March 2018 to All of the treatments (vitamin D supplementation February 2019. The study was conducted in accor- and placebo) were given upon admission until 12 dance with the Declaration of Helsinki (1964), and weeks. The standard medical treatment for ische- the protocol was approved by the Health Research mic stroke and physiotherapy were administered Ethical Committee, Medical Faculty, Universitas in both groups. No lost to follow-up was found Sumatera Utara. Informed consents were taken until the end of this study on 3 months. from the patients prior to participation to ensure At the time of admission, all patients were me- complete satisfaction. asured for NIHSS score, vitamin D and vitamin The inclusion criteria were the patients with acu- A serum level, and IL-1β serum level. After 12 te ischemic stroke diagnosed clinically and evi- weeks, all of the measurements were repeated denced by neuroimaging examination, admitted once again to evaluate the differences. to the Hospital on day ≤3 after onset and were Initial stroke severity and clinical outcome were >18 years of age. The exclusion criteria were the determined using the National Institute of Health patients with impaired liver function (26), impa- Stroke Scale (NIHSS). The scale consists of five ired renal function (serum urea concentration > score sections: score 0 (no stroke symptoms), 2 mg/dL, nephrolithiasis) (27), patients taking score 1–4 (minor stroke), score 5–15 (moderate supplements containing vitamin A and D at the stroke), score 16–20 (moderate to severe stroke), time of the study (27,28), hypercalcemia, car- and score 21–42 (severe stroke) (29). dioembolic stroke, and patients with sepsis and Serum vitamin D was evaluated using enzyme- diabetes mellitus. linked immunosorbent assay (ELISA). Serum As the ischemic stroke patients got admitted in was extracted from venous blood, and quantita- the Neurology Department, every consecutive tive measurement of total 25-OH vitamin D3 in patient was assessed for inclusion and exclusion serum was performed using a monoclonal anti- criteria and was randomized by a simple rando- body that binds to 25-OH vitamin D3 (30). Se- mization method, where random numbers were rum vitamin A was measured by high-pressure computer generated using Microsoft Excel Spre- liquid chromatography (HPLC). Serum was ex- adsheet Software. tracted from venous blood and protected from One hundred and twenty three ischemic stroke light and chilled until centrifugation (31-32). patients were assessed for the study, 120 were Serum IL-1β was measured using the Quanti- found to be eligible and were randomized, and kine® HS ELISA for Human IL-1β (R&D Sys- assigned to four groups: supplementation using tem, a Biotechne brand) according to the manu- vitamin A only, using vitamin D only, combinati- facturer’s protocol (33). on of vitamin A and D, and also placebo group of Statistical analysis 30 patients in each group. To analyse baseline characteristics across the Methods treatment groups, one-way ANOVA and χ2 The patients in the vitamin A group were given tests were used. To determine the differences vitamin A 50,000 IU per week (5,000 IU vitamin of serum vitamin A and D levels, Serum IL-1β, A per tablet, grinded and inserted into a capsule and NIHSS between before and after treatment to make 50,000 IU dose), while the vitamin D paired t-test was used.

427 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 1. Baseline characteristics of 120 patients with ischemic stroke Treatment groups Parameter Reference values p Vitamin A + D Vitamin A Vitamin D Placebo Age (Year; Mean ± SD) 65.85 ± 4.23 66.85 ± 4.62 62.65 ± 5.66 66.35 ± 5.30 0.066 Gender (No) 0.122 Male 13 15 8 17 Female 17 15 22 13 Smoking (No) 0.100 Yes 11 14 8 13 No 19 16 23 17 Alcohol drinking (No) Yes 9 11 8 14 No 21 19 22 16 0.150 NIHSS on admission (Mean ± SD) 13.25 ± 1.61 12.10 ± 1.92 13.20 ± 1.24 13.15 ± 1.39 0.067 Systolic blood pressure (Mean ± SD) (mmHg) 120 mmHg 152.05 ± 9.85 151.15 ± 9.72 157.85 ± 8.78 155.65 ± 9.48 0.097 Blood sugar level (Mean ± SD) (mg/dL) 70 – 130 mg/dL 121.10 ± 13.21 127.75 ± 10.76 123.10 ± 12.22 127.70 ± 19.27 0.352 Dyslipidaemia (No) Yes 16 15 13 14 0.848 No 14 15 17 16 Vitamin A serum level (Mean ± SD) (μg/L) 200 – 800 μg/L 462.90 ± 246.78 476.85 ± 201.79 487.90 ± 107.18 473.80 ± 273.49 0.720 Vitamin D serum level (Mean ± SD) (ng/mL) > 30 ng/mL 19.71 ± 7.15 20.42 ± 7.58 18.19 ± 6.90 20.69 ± 7.41 0.038 IL-1β serum level (Mean ± SD) (pg/mL) < 0.70 pg / mL 0.49 ± 0.17 0.47 ± 0.42 0.30 ± 0.19 0.43 ± 0.39 0.170 NIHSS, National Institute of Health Stroke Scale; RESULTS Table 3. Changes in serum vitamin D levels before and after treatment Treatment Vitamin D serum level (Mean ± SD) (μg/L) From the total of 120 patients, the baseline data p on admission showed no significant difference group Before After across the four treatment groups. The overall Vitamin A 20.42 ± 7.57 19.67 ± 8.02 0.08 Vitamin D 18.19 ± 6.91 20.37 ± 8.66 0.02 mean of vitamin D level of the patients on admis- Vitamin A+D 19.71 ± 7.15 25.70 ± 10.39 0.01 sion showed insufficiency level (20.75±7.53 ng/ Placebo 20.69 ± 7.41 20.62 ± 7.10 0.07 mL) (Table 1). The overall mean vitamin A level There was a significant difference between serum was normal (422.90 ± 214.45 μg/L). Initial mean IL-1β level before and after the treatment, in the of IL-1β serum level was above the normal value vitamin A+D group (p<0.001), and also in the (0.75 ± 1.72 pg/mL). placebo group (p<0.001) (Table 4). In the vitamin A group and combination group,

serum vitamin A level after the treatment was Table 4. Changes in serum IL-1ß levels before and after increased comparing to initial levels (p=0.21 and treatment p=0.04, respectively), whereas in the vitamin D Treatment IL-1β serum level (Mean ± SD) (pg/mL) p group and placebo groups a slight (no significant) group Before After decrease was found (p=0.92 and p=0.18, respec- Vitamin A 0.47 ± 0.42 0.49 ± 0.29 0.14 Vitamin D 0.30 ± 0.19 0.27 ± 0.17 0.19 tively) (Table 2). Vitamin A+D 0.49 ± 0.17 0.21 ± 0.14 <0.001 Placebo 0.43 ± 0.39 0.79 ± 0.26 <0.001 Table 2. Changes in serum vitamin A level before and after treatment Treatment Vitamin A serum level (Mean ± SD) (μg/L) Although the result of this study indicated a si- p group Before After gnificant decrease in the NIHSS value after the Vitamin A 476.85 ± 201.78 498.55 ± 204.59 0.21 treatment in all treatment groups (p<0.001), the Vitamin D 487.90 ± 107.18 486.70 ± 126.55 0.92 Vitamin A+D 462.90 ± 246.78 511.95 ± 282.45 0.04 biggest decrement was found in the combination Placebo 473.80 ± 463.95 463.95 ± 266.34 0.18 group (Table 5).

In the vitamin D group and vitamin A+D supple- Table 5. Changes in NIHSS (National Institute of Health mentation group, serum vitamin D level was Stroke Scale) before and after treatment Treatment NIHSS (Mean ± SD) increased significantly (p=0.02 and p=0.01, res- p pectively) after the treatment, whereas in the pla- group Before After cebo group and vitamin A group, serum vitamin Vitamin A 12.10 ± 1.92 10.30 ± 1.59 < 0.001 Vitamin D 13.20 ± 1.24 10.40 ± 1.23 < 0.001 D decreased slightly (no significant) (p=0.08 and Vitamin A+D 13.25 ± 1.61 6.00 ± 1.52 < 0.001 p=0.07, respectively) (Table 3). Placebo 13.15 ± 1.39 11.75 ± 1.29 < 0.001

428 Kadri et al. Vitamin A and D and ischemic stroke

DISCUSSIONS taking place, so that pro-inflammatory cytokine levels would also increase (42). This is consistent The results of the study found that the baseline with the theory of increasing serum IL-1β, the characteristics of the study patients did not differ pro-inflammatory cytokines; in a healthy brain, significantly between the treatment groups. This IL-1 is not detected, but will be expressed after indicates that the initial conditions of the patients an injury or during an illness (43). For example, were quite homogeneous when the study begins. in animal stroke models (induced by occlusion of In all groups, the mean of serum vitamin D level cerebral artery media), a rapid increase in IL-1α, at the beginning of the study were below normal and IL-1β expression is observed in microglia values (insufficiency). This is consistent with within a few hours of damage and in the brain research on vitamin D level (insufficiency) also area, which experiences cell death (42,43). conducted in Indonesia by Sari et al. on 2017 (34) and with the study of Sari et al. (2014), which Vitamin D is one of the nutraceuticals that has was also carried out in the North Sumatra region a major role in modulating the inflammatory (17.71 ng/mL). There is an influence of the type process in ischemic stroke, through its anti-in- of work and also the source of food on the adequ- flammatory and antioxidant properties, especi- acy of vitamin D (35). This finding is consistent ally in modulation of interleukin. In this research, with research conducted in Indonesia (36). groups that received vitamin D supplementation (vitamin D only and vitamin A+D) showed a de- In the vitamin A and combination A+D supple- crement in serum IL-1β levels, whereas the de- mentation groups, serum vitamin A levels after crement was more significant in the combination supplementation showed an increase compared group. This could be attributed to the fact that to other groups. In the combination A+D group, a giving vitamin A alongside vitamin D will pro- significant increase in serum vitamin A level was duce a synergism effect which can improve the seen. In the vitamin D and placebo groups, the- efficacy of both vitamins( 3,12,25, 43-45). re was a slight decrease in serum vitamin A level, but it was not significant. This shows that, without All of the findings of this study support the theory adequate intake from outside, serum vitamin A le- that vitamin A can increase the effectiveness of vel can become even lower. Likewise, supplemen- vitamin D. Optimal vitamin D activation requires tation containing vitamin A can increase serum vitamin A, so that the administration of vitamin A vitamin A level. It turns out that a combination of together with vitamin D as supplementation can vitamin A+D provides the highest increase among maximize the expected anti-inflammatory and the other 3 groups. This supports the theory that neuroprotective effects. The study results show the action of vitamins A and D work together, in- that co-administration of VDR and RXR ago- volving nuclear receptors (37-41). nists synergistically can reduce atherosclerosis and correlate with endothelial protection through After the treatment, in the vitamin D and vitamin suppression of oxidative stress and inflammation. A+D supplementation groups, serum vitamin D le- The combination of treatment with VDR and RXR vels were found to increase significantly. Whereas agonists is a promising strategy for the prevention in the vitamin A and placebo groups, serum vitamin and treatment of atherosclerosis (25,46,47). When D levels showed a slight decrease. This is consi- examined more deeply, among the 4 treatment stent with research conducted in the North Sumatra groups, the highest decrease in NIHSS was seen region in 2017 (34). The similarity of these findings in the combination of vitamin A and D supple- is possible because, in general, the population of the mentation; it seems to have the most inflammatory two studies was the same, and also the areas whe- effect, characterized by an increase in serum vita- re they were carried out. Thus, it can be concluded min D and A levels, as well as the most significant that different types of supplementation will affect decrease in serum IL-1β level in this supplementa- changes in serum vitamin D levels (34). tion group compared with the other 3 groups. The increased initial IL-1β serum levels were pro- This study showed that type’s supplementation bably due to the initial examinations carried out had a significant relationship with clinical outco- in the acute phase of stroke (<3 days after onset), mes (NIHSS) where the combination of vitamins where the inflammatory process in stroke was A and D provided the most significant improve-

429 Medicinski Glasnik, Volume 17, Number 2, August 2020

ment in clinical outcomes. Vitamins A and D can changes in serum vitamin levels and changes in affect inflammation in a number of ways, such cytokines, but there are still other factors that also as antioxidant/anti-inflammatory mechanisms, play a role in clinical outcomes of stroke, such as inhibition of nitric oxide synthase, regulation of age and location of lesions (56,57). neuronal calcium, and detoxification pathways Some limitations in this study were that, other (48). Over time, it appears that vitamin D defi- factors that affect the metabolism of vitamins A ciency often occurs throughout the world and is and D such as daily food intake, were not inve- associated with several chronic health problems, stigated in more details. To ensure that the me- including cardiovascular, musculoskeletal, infec- tabolism of vitamins A and D in the body goes tion, autoimmune, and malignancy (4,11,39). In well, an examination should be made of several vitro and animal studies, vitamin D deficiency things such as the content of food intake, intesti- affects the activity/expression of macropha- nal absorption and genetic factors. However, the ges and lymphocytes in atherosclerotic plaque, main purpose of this study was not to determine which causes chronic inflammation of the artery the nutritional status of vitamins in the subjects, walls. Also, vitamin D modulates the immune but rather to see the effect of vitamin supplemen- system by regulating the production of inflamma- tation on clinical outcomes, through cytokine tory cytokines and inhibits pro-inflammatory cell modulation. Furthermore, serum vitamin A and proliferation, both of which are important in the D levels after supplementation, can be used as an pathogenesis of vascular inflammation that cau- overall representation of the success of supple- ses atherogenesis (3,4,49,50). mentation to improve vitamin levels in the body, Vitamin A is a nutrient that influences the immu- and thus also affect changes in cytokine levels, ne response through multiple mechanisms, such and ultimately, clinical outcomes. as helping the production of B cells from immu- Ultimately, the findings of this study suggested noglobulin A (IgA) by up-regulating IL-6 (51). that administration of combination of vitamin A Retinoic Acid (RA), a bioactive derivative of vi- and D supplementation can significantly improve tamin A and an important differentiation factor clinical outcome in ischemic stroke when compa- during vertebrate development, has been found to red to patients without supplementation. participate in neuron and vascular development. In conclusion, administration of the combination In animal models, it has been proven that RXR of vitamin A and D supplementation significantly agonists inhibit the development of atherosclero- increased vitamin A and D serum level, decrea- sis (25,46,47,52). Retinoic Acid receptors can also sed IL-1β serum level, and ultimately improved interact with other nuclear receptors that have ne- clinical outcome in ischemic stroke patients. uroprotective effects such as vitamin D3, which is neuroprotective against stroke (5, 22–24). ACKNOWLEDGMENTS Vitamins A and D reduce the production of T cells, Authors would like to thank Prodia Laboratory increase the production of anti-inflammatory T cell Medan for their support to this research. mediators and promote regulation of T cell differentiation. Vitamin A also reduces IL-6 and IFN-γ FUNDING production and limits the toxic effects of Reactive No specific funding was received for this study. Oxygen Species (ROS) released during inflammation (53–55). Nevertheless, factors that influence CONFLICTS OF INTEREST stroke outcomes certainly do not only depend on Competing interests: None to declare. REFERENCES 1. Gofir A. Pengantar Manajemen Stroke Komprehen- 3. Muscogiuri G, Annweiler C, Duval G, Karras S, Ti- sif (Comprehensive Stroke Management) [Indonesi- rabassi G, Salvio G. Vitamin D and cardiovascular an]. Jogjakarta: Pustaka Cendikia Press, 2007. disease: From atherosclerosis to myocardial infarcti- 2. Lakhan SE, Kirchgessner A, Hofer M. Inflamma- on and stroke. Int J Cardiol 2017; 230:577-84. tory mechanisms in ischemic stroke: therapeutic 4. Makariou SE, Michel P, Tzoupi MS, Challa A. Mili- approaches. J Transl Med 2009; 7:97. on HJ. Vitamin D and stroke: promise for prevention and better outcome. Curr Vasc Pharmacol 2014; 12:117-24.

430 Kadri et al. Vitamin A and D and ischemic stroke

5. Sato Y, Meller R, Yang T, Taki W, Simon R. Stereo- 21. Kong L, Wang Y, Wang XJ, Wang XT, Zhao Y, Wang selective neuroprotection against stroke with vita- LM, Chen ZY. Retinoic acid ameliorates blood–bra- min A derivatives. Brain Res 2008; 1241:188-92. in barrier disruption following ischemic stroke in 6. Vonder Haar C, Peterson T, Martens K, Hoane M. rats. Pharmacol Res 2015; 99:125–36. Vitamins and nutrients as primary treatments in expe- 22. Rist PM, Jiménez MC, Tworoger SS, Hu FB, Man- rimental brain injury: Clinical implications for nutra- son JE, Sun Q, Rexrode KM. Plasma retinol-bin- ceutical therapies. Brain Res 2016; 1640:114–29. ding protein 4 levels and the risk of ischemic stro- 7. Badimon L, Vilahur G, Padro T. Nutraceuticals and ke among women. J Stroke Cerebrovasc Dis 2018; atherosclerosis: human trials. Cardiovasc Ther 2010; 27:68–75. 28:202–15. 23. Yu S, Airavaara M, Wu K, Harvey BK, Liu HS, Yang 8. Aggarwal A, Aggarwal P, Khatak M, Khatak S. Y, Zacharek A, Chen J, Wang Y. 9-cis retinoic acid Cerebral ischemic stroke: sequels of cascade. Int J induces neurorepair in stroke brain. Sci Rep 2017; Pharma Bio Sci 2010; 1:1-24. 7:1–12. 9. Sun Q, Pan A, Hu FB, Manson JE, Rexrode KM. 24. Panickar KS, Dawson HD. Acute effects of All- 25-Hydroxyvitamin D levels and the risk of stroke. trans-retinoic Acid in ischemic injury. Transl Neu- A prospective study and meta-analysis. Stroke 2012; rosci 2012; 3:143-52. 43:1470-77. 25. Lin LM, Peng F, Liu YP, Chai DJ, Ning RB, Xu CS, 10. Judd SE, Morgan CJ, Panwar B, Virginia J, Wadley Lin JX. Co-administration of VDR and RXR agoni- VG, Jenny NS, Kissela BM, Gutiérrez OM. Vitamin sts synergistically alleviates atherosclerosis through D deficiency and incident stroke risk in community inhibition of oxidative stress: an in vivo and in vitro living black and white adults. Int J Stroke 2017; study. J Atheroscler 2016; 251:273-81. 11:93–102. 26. Witham MD, Crighton LJ, Gillespie ND, Struthers 11. Nargesi AA, Heidari B, Esteghamati S, Hafezi-Ne- AD, McMurdo MET. The effects of Vitamin D jad N, Sheikhbahaei S, Pajouhi A, Nakhjavani M, Supplementation on physical function and quality of Esteghamati A. Contribution of vitamin D defici- life in older patients with heart failure. J Am Heart ency to the risk of coronary heart disease in subjects Assoc 2010; 3:195-201. with essential hypertension. Atherosclerosis 2016; 27. Schleithoff SS, Zitterman A, Tenderich G, Berthold 244:165-71. HK, Stehle P, Koerfer R. Vitamin D supplementation 12. Agbalalah T, Hughes SF, Freeborn EJ, Mushtaq S. Im- improves cytokine profiles in patients with congesti- pact of vitamin D supplementation on endothelial and ve heart failure: a double-blind, randomized, place- inflammatory markers in adults: A systematic review. J bo-controlled trial. Am J Clin Nutr 2006; 83:754 –9. Steroid Biochem Mol Biol 2017; 173:292-300. 28. Zittermann A, Frisch S, Berthold HK, Götting C, 13. Kato S. The function of vitamin D receptor in vita- Kuhn J, Kleesiek K, Stehle P, Koertke H, Koerfer R. min D action. J Biochem 2000; 127:717-22. Vitamin D supplementation enhances the beneficial 14. Pike JW, Meyer MB, Martowicz ML. New tech- effects of weight loss on cardiovascular disease risk niques in transcription research extend our under- markers. Am J Clin Nutr 2009; 89:1321-7. standing of the molecular actions of the vitamin D 29. Kasner SE. Clinical interpretation and use of stroke hormone. IBMS Bone Key 2009; 6:169-80. scales. Lancet Neurol 2006; 5:603–12. 15. Germain P, Chambon P, Eichele G, Evans RM, Lazar 30. Kennel KA, Drake MT, Hurley DL. Vitamin D defi- MA, Leid M, De Lera AR, Lotan R, Mangelsdorf ciency in adults: when to test and how to treat. Mayo DJ, Gronemeyer H. International Union of Pharma- Clin Proc 2010; 85:752–57. cology. LXIII. Retinoid X receptors. Pharmacol Rev 31. Gibson R, ed. Principles of nutritional assessment. 2006; 58:760–72. Oxford, UK: Oxford University Press, 2005. 16. De Lera AR, Krezel W, Rühl R. An Endogenous 32. Sommer A. Vitamin A deficiency and its consequ- mammalian retinoid X receptor ligand, at last! Chem ences: a field guide to detection and control. 3rd ed. Med Chem 2016; 11:1–12. Geneva: World Health Organization, 1995. 17. Allenby G, Bocquel MT, Saunders M, Kazmer S, 33. Quantikine® HS ELISA Human IL-1β/IL-1F2 Speck J, Rosenberger M, Lovey A, Kastner P, Grip- Immunoassay. https://resources.rndsystems.com/ po JF, Chambon P. Retinoic acid receptors and reti- pdfs/datasheets/hslb00c.pdf. (20 April 2018) noid X receptors: interactions with endogenous reti- 34. Sari DK, Zaimah ZT, Lestari S, Hutagalung SV, Ga- noic acid. Proc Natl Acad Sci USA 1993; 90:30–4. nie RA. Vitamin D supplementation in women with 18. Rühl R, Krzyżosiak A, Niewiadomska-Cimicka A, vitamin d receptor gene polymorphisms: a randomi- Rochel N, Szeles L, Vaz B, Wietrzych-Schindler M, zed controlled trial. Asian J Clin Nutr 2017; 9:89-96. Álvarez S, Szklenar M, Nagy L, de Lera AR, Krężel 35. Sari DK, Alrasyid H, Nurindrawaty L, Zulkifli L. W. 9-cis-13,14-Dihydroretinoic acid is an endoge- Occurrence of vitamin D deficiency among women nous retinoid acting as RXR ligand in mice. PLOS in North Sumatera, Indonesia. Mal J Nutr 2014; Genetics 2015; 11:e1005213. 20:63-70. 19. Min K, Min J. Relation of serum vitamin A levels to 36. Tanumihardjo SA, Muherdiyantiningsih, Permaesih all-cause and cause-specific mortality among older D, Komala, Muhilal, Karyadi, D. Daily supplements adults in the NHANES III population. Nutr Metab of vitamin A (8.4 μmol, 8000 IU) improve the vita- Cardiovasc Dis 2014; 24:1197-203. min A status of lactating Indonesian women. Am J 20. Wei Z, Kuang J. Vitamin D deficiency in relation to Clin Nutr 1996; 63:32–5. the poor functional outcomes in nondiabetic patients 37. Golzarand M, Shab-Bidar S, Koochakpoor G, Spea-

with ischemic stroke. Biosci Rep 2018; 38:1–8. kman JR, Djafarian K. Effect of vitamin D3 supplementation on blood pressure in adults: An updated meta-analysis. Nutr Metab Cardiovasc Dis 2016; 26:663-73.

431 Medicinski Glasnik, Volume 17, Number 2, August 2020

38. Hsia J, Heiss G, Ren H, Allison M, Dolan NC, Gre- 48. Ren W, Gu Y, Zhu L, Wang L, Chang Y, Yan M, Han enland P, Heckbert SR, Johnson KC, Manson JE, B, He J. The effect of cigarette smoking on vitamin Sidney S, Trevisan M. Calcium/vitamin D supple- D level and depression in male patients with acute mentation and cardiovascular events. Circulation ischemic stroke. Compr Psychiatry 2015; 65:9–14. 2007; 115:846-54. 49. Chehaibi K, Trabelsi I, Mahdouani K, Slimane M. 39. Daraghmeh A, Bertoia M, Al-Qadi M, Abdulbaki Correlation of oxidative stress parameters and in- A, Roberts M, Eaton C. Evidence for the vitamin flammatory markers in ischemic stroke patients. J D hypothesis: The NHANES III extended mortality Stroke Cerebrovasc Dis 2016; 25:2585-93. follow-up. Atherosclerosis 2016; 255:96-101. 50. Poole KE, Loveridge N, Barker PJ, Halsall DJ, Rose 40. Christakos S, Ajibade D, Dhawan P, Fechner A, C, Reeve J, Warburton EA. Reduced Vitamin D in Mady L. Vitamin D: metabolism. Endocrin Metab Acute Stroke. Stroke 2006; 37:243-5. Clin 2010; 39:243-53. 51. Penkert R, Jones B, Häcker H, Partridge J, Hurwitz 41. Challoumas D. Vitamin D supplementation and lipid J. Vitamin A differentially regulates cytokine expre- profile: What does the best available evidence show? ssion in respiratory epithelial and macrophage cell Atherosclerosis 2014; 235:130-39. lines. Cytokine 2017; 91:1–5. 42. Brough D, Galea J. The role of inflammation and 52. Jimenez-Lara AM, Aranda JA. Interaction of vita- interleukin-1 in acute cerebrovascular disease. J In- min D and retinoid receptors on regulation of gene flamm Res 2013; 6:121-8. expression. Horm Res 2001; 54:301–5. 43. Denes A, Pinteaux E, Rothwell N, Allan S. Interle- 53. Røsjø E, Myhr KM, Løken-Amsrud KI, Bakke SJ, ukin-1 and Stroke: Biomarker, harbinger of dama- Beiske AG, Bjerve KS, Hovdal H, Lilleås F, Midga- ge, and therapeutic target. Cerebrovasc Dis 2011; rd R, Pedersen T, Benth JS, Torkildsen Ø, Wergeland 32:517-27. S, Michelsen AE, Aukrust P, Ueland T, Holmøy T. 44. Jain S, Buttar HS, Chintameneni M, Kaur G. Pre- Increasing serum levels of vitamin A, D and E are vention of cardiovascular diseases with anti-in- associated with alterations of different inflammation flammatory and antioxidant nutraceuticals and her- markers in patients with multiple sclerosis. J Neuro- bal products: an overview of pre-clinical and clinical immunol 2014; 271:60–5. studies. Recent Pat Inflamm Allergy Drug Discov 54. Barker T, Martins TB, Hill HR, Kjeldsberg CR, 2018; 12:145–57. Dixon BM, Schneider ED, Henriksen VT, Weaver 45. Kamal K, Majeed F, Naqvi I. Vitamin D supplemen- LK. Vitamin D sufficiency associates with an increa- tation for preventing recurrent stroke and vascular se in anti-inflammatory cytokines after intense exer- events in patients with stroke or transient ischemic cise in humans. Cytokine 2014; 65:134–7. attack (Protocol). Cochrane Database Syst Rev 55. Anand PK, Kaul D, Sharma M. Synergistic action of 2012; 41-8. vitamin D and retinoic acid restricts invasion of ma- 46. Lips P, Eekhoff M, van Schoor N, Oosterwerff M, crophages by pathogenic mycobacteria. J Microbiol de Jongh R, Krul-Poel Y, Simsek S. Vitamin D and Immunol Infect 2008; 41:17-25. type 2 diabetes. Steroid Biochem Mol Biol 2017; 56. Park SY, Lee IH. The influence of age, lesion side 173:280-5. and location on rehabilitation outcome after stroke. J 47. Michos ED, Carson KA, Schneider AL, Lutsey PL, Phys Ther Sci 2011; 23:817-9. Xing L, Sharrett AR, Alonso A, Coker LH, Gross M, 57. Belanger HG. Recovery from stroke: factors affec- Post W, Mosley TH, Gottesman RF. Vitamin D and ting prognosis. Clin Neuropsychol 2019; 33:813-6. subclinical cerebrovascular disease. JAMA Neurol 2014; 71:863.

432 ORIGINAL ARTICLE

Association between working memory impairment and activities of daily living in post-stroke patients

Fasihah Irfani Fitri, Aida Fithrie, Aldy S Rambe

Department of Neurology, School of Medicine, Universitas Sumatera Utara, Indonesia

ABSTRACT

Aim Stroke is one of the leading causes of adult disability and functional impairment worldwide. Cognitive impairments including memory dysfunction are common after stroke and may have a negative impact to the functional status and activities of daily living (ADL). The study aimed to determine the impact of working memory impairment after stroke on ADL.

Methods This cross-sectional study involved post-stroke patients who underwent neurologic examination and mini mental state examination (MMSE), forward digit span (FDS) and backward digit span (BDS) to assess cognitive function, and assessment of da- Corresponding author: ily life activities (ADL) and instrumental activities of daily living Fasihah Irfani Fitri (IADL) scales. Department of Neurology, Results This study included 38 patients, 23 (60.5%) males and 15 School of Medicine, Universitas Sumatera (39.5%) females. The mean MMSE, FDS and BDS scores were Utara, Adam Malik General Hospital 24.60±4.49, 4.87±1.166 and 3.47±1.158, respectively. There were Medan 17 (44.7%) patients with cognitive impairment (MMSE Score was Jalan Bunga Lau No 17, Kemenangan lower than 24). The mean ADL and IADL scores were significantly higher in patients with cognitive impairment, showing gre- Tani, Medan Tuntungan, Kota Medan, ater dependency in this group of patients. There was a significant Sumatera Utara-20136, Indonesia negative correlation between global cognitive function, working Phone: +08 116 550 721 memory and ADL and IADL scores in post-stroke patients. E-mail: [email protected] Conclusion Cognitive impairment has a negative impact on daily ORCID ID: https://orcid.org/0000-0002- life activities in post-stroke patients. 0883-3029 Key words: cognition, function, quality of life, stroke Original submission: 24 January 2020; Revised submission: 12 February 2020; Accepted: 04 May 2020 doi: 10.17392/1135-20

Med Glas (Zenica) 2020; 17(2): 433-438

433 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION and psychological. Other than global cognitive function, working memory has also been found Stroke remains a major healthcare problem and is to be related to functional outcome and might the third leading cause of death after heart disease affect physical rehabilitation and the quality of and cancer. Stroke is also a leading cause of functi- life subsequently (9). Post-stroke cognitive impa- onal impairments, with 20% of survivors requiring irment can be assessed through neuropsychologi- institutional care after 3 months, and 15-30% be- cal assessments that ideally are able to cover all ing permanently disabled (1). Stroke is considered cognitive domains, including working memory a major cause of long-term physical disabilities in (10). However, in clinical practice setting, the adults; it is the second most common cause of co- brevity of this neuropsychological assessment gnitive impairment and dementia (2). Physical im- should also be taken in consideration (11). pairments tend to improve, to a greater or lesser degree, following stroke; however, for reasons which The study aimed to determine the impact of glo- remain unknown, cognitive impairments progre- bal cognitive and working memory impairment ssively worsen (3). Several studies have shown that after stroke on daily life activities. cognitive impairment might have a negative effect PATIENTS METHODS on functional outcome and activities of daily living (ADL) (2,4,5). The occurrence of cognitive impa- Patients and study design irment among post-stroke patients was high. Mul- tiple cognitive domains are affected and this may This cross-sectional study involved 38 post-stroke hamper recovery and negatively impact indepen- patients who were recruited from the Memory dence and quality of life after stroke (5). Cognitive Clinic, Neurology Department of the Adam Malik impairment may also lead to decreased functional General Hospital Medan, North Sumatera, Indo- capacity and thus affect rehabilitation outcome in nesia, between April and June 2018. Inclusion cri- stroke. Paker et al. reported that cognitive dysfunc- teria were: age equal or more than 18 years, history tion interfered with community ambulation in pa- of stroke for more than three months to two years, tients with stroke, but did not have a significant modified Rankin scale score ˂3, compos mentis effect on ADL and global recovery (6). and fully cooperative, speak Bahasa Indonesia Stroke mainly affects attention and executive func- fluently, able to read and write, and gave written tion, but may also impair memory function, ma- consent to be included in the study. Exclusion cri- inly working memory and episodic memory (2,7). teria were: patients who were medically unstable Post-stroke memory dysfunction (PMD) is a pre- (delirium) or had other psychiatric disorders, had requisite for the diagnosis of post-stroke dementia. an aphasia and history of dementia before stroke. The prevalence of PMD varied from 23-55% in the All patients underwent physical and neurologic three-month post-stroke period. Larger stroke vo- examination. lume, pre-stroke medial temporal lobe atrophy, and The Health Research Ethical Committee of white matter lesions were related with decreased the Medical Faculty of Universitas Sumatera post-stroke memory function (7). A deficit in wor- Utara/H. Adam Malik General Hospital approved king memory also negatively impacts post-stroke this study. physical rehabilitation. Malouin et al. studied the relationship between working memory and motor Methods improvement in post-stroke patients and found that Cognitive function was evaluated using mini working memory correlated significantly with the mental state examination (MMSE), forward digit level of improvement. Their result suggested that span (FDS) and backward digit span (BDS). The the functional outcome after practice depended on MMSE, which consisted of 6 subscales (orien- the ability to maintain and manipulate information tation, registration, attention, calculation, recall, in working memory (8). language) and praxis test, was used to assess glo- The magnitude of stroke makes it important for bal cognitive function, to estimate the severity of patients and their families to be aware of their cognitive impairment and to classify patients as future possibilities and factors associated with re- having a clinical level of cognitive impairment habilitation not only physical but also cognitive (12): patients with cognitive impairment - the

434 Fitri et al. Working memory in post-stroke patients

MMSE score <24, and patients with normal co- RESULTS gnitive function – the MMSE score ≥24. This study included 38 patients, 23 (60.5%) Measures of forward (FDS) and backward di- males and 15 (39.5%) females. The mean age git span (BDS) are among the oldest and most was 58.8±10.38 years. widely used neuropsychological tests of short- The mean MMSE score was 24.60±4.49. There term verbal memory (13). were 17 (44.7%) patients with cognitive impair- Digit span was measured for forward and rever- ment (the MMSE score <24), and 21 (55.3%) pa- se-order (backward) recall of digit sequences. tients with normal cognitive function (the MMSE In the forward condition, sequences of digits of score ≥24). There were no significant differences on increasing length have to be repeated in the same demographic characteristics and stroke risk factors order as presented. In the backward condition, di- between the patients with cognitive impairment and git sequences have to be repeated in reverse order those who were cognitively intact (Table 1). (14). Digit sequences are presented beginning with Table 1. Clinical characteristics of the patients a length of two digits, and two trials are presented No (%) of patients in the group at each increasing list length. Testing ceases when Cognitively Cognitive Total the patient fails to accurately report either trial at Variable normal group impairment (n=38) p one sequence length or when the maximal list len- (n=21) group (n=17) gth is reached (9 digits forward, 8 backward) (13). Gender 0.847 Male 13 (61.9) 10 (58.8) 23 (60.5) The ability to perform activities of daily living Female 8 (39.1) 7 (41.2) 15 (39.5) Age (mean±SD) was assessed using two scales: Activities of Da- 59.82±9.14 57.50±12.06 58.84±10.38 0.524 (years) ily Living (ADL) and Instrumental Activities of Age group (years) 0.775 Daily Living (IADL). <45 1 (4.8) 1 (5.9) 2 (5.3) 46-50 3 (14.3) 3 (17.6) 6 (15.8) The ADL scale is a career-rated instrument consi- 51-55 3 (14.3) 2 (11.8) 5 (13.2) sting of six daily-living abilities, including basic 56-60 4 (19.0) 5 (29.4) 9 (23.7) 61-65 3 (14.3) 4 (23.5) 7 (18.4) tasks of personal care in everyday life, e.g. bathing, 66-70 4 (19.0) 0 (0.00) 4 (10.5) continence, transfers, feeding, dressing and tran- 71-75 2 (9.5) 1 (5.9) 3 (7.9) sferring. For this study, questionnaire responses >75 1 (4.8) 1 (5.9) 2 (5.3) Educational level 0.403 were made on to a Likert scale (15) ranging from Primary 1 (4.8) 3 (17.6) 4 (10.5) 0 (independent) to 2 (completely dependent) (16). Junior high school 3 (14.3) 1 (5.9) 4 (10.5) The Instrumental Activities of Daily Living High School 12 (57.1) 4 (23.5) 16 (42.1) Diploma 1 (4.8) 1 (5.9) 2 (5.3) (IADL) scale measures the activities related to University 4 (19.0) 8 (47.1) 12 (31.5) independent living. It is a career-rated instrument Occupation 0.747 consisting of seven daily-living abilities inclu- Employee 7 (33.3) 7 (41.2) 14 (36.8) Housewife 4 (19.0) 2 (11.8) 6 (15.8) ding items related to using the telephone, prepa- Entrepreneur 6 (28.6) 1 (5.9) 7 (18.4) ring meals, taking medicine, traveling, shopping Farmer 0 1 (5.9) 1 (2.6) for groceries or personal items, performing light Unemployed 4 (19) 6 (35.3) 10 (26.3) Stroke aetiology 0.878 or heavy housework and managing money. In a Ischemia 20 (95.2) 16 (94.1) 36 (94.7) similar way to the ADL, every activity is rated on Haemorrhage 1 (4.8) 1 (5.9) 2 (5.3) a scale that includes three choices: person is in- Hypertension 0.426 Yes 20 (95.2) 15 (88.2) 35 (92.1) dependent, person requires assistance and person No 1 (4.8) 2 (11.8) 3 (7.9) is completely dependent on others. For this study, Diabetes mellitus 0.578 the questionnaire responses were made using a Yes 8 (38.1) 8 (47.1) 16 (42.1) Likert scale ranging from 0 (independent) to 2 No 13 (61.9) 9 (52.9) 22 (57.9) Atrial fibrillation 0.828 (completely dependent) (17). Yes 1 (4.8) 1 (5.9) 2 (5.3) No 20 (95.2) 16 (94.1) 36 (94.7) Statistical analysis Coronary heart disease 0.239 Yes 2 (9.5) 4 (23.5) 6 (15.8) The Spearman correlation to measure the correlati- No 19 (90.5) 13 (76.5) 32 (84.2) on between MMSE scores and ADL and IADL was used. Statistical significance was assumed at <0.05. The FDS (p=0.0124) and BDS (p=0.026) scores were significantly higher in patients with normal

435 Medicinski Glasnik, Volume 17, Number 2, August 2020

cognitive function, 5.29±1.007 and 4.00±1.304, ce of neuropsychological impairment could affect respectively, comparing to the patients with the functional abilities in post-stroke patients, par- impaired cognitive function, 4.35±1.169 and ticularly the IADL (5). Orso et al. studied 75 post- 2.82±1.704, respectively. stroke patients and found that cognitive dysfunc- Mean scores of ADL (p=0.001) and IADL tion had a strong impact on ADL dysfunction (6). (p=0.001) were significantly lower in the - pati Park et al. found that post-stroke cognitive impair- ents without cognitive impairment, 0.19±0.512 ment without dementia also interfered with health- and 0.38±1.071, respectively, showing that pati- related quality of life (19). Cognitive impairment ents with normal cognitive function were more has also been reported to be associated with post- independent than those who were cognitively stroke rehabilitation. Deficits in working memory impaired, 3.29±3.19 and 5.41±5.06, respectively (ability to hold information in mind temporarily) (Table 2). caused by stroke can present a significant barrier to independence (9). Malouin et al. reported that Table 2. Activities of daily living (ADL) and instrumental activities of daily living (IADL) scores of post-stroke patients working memory negatively correlated with the with and without cognitive impairment level of improvement, suggesting that functional Mean±SD outcome after practice depended on the ability to Cognitively Cognitive maintain and manipulate information in working Total Variable normal group impairment p (n=38) memory (8). Although a study by Paker et al. did (n=21) group (n=17) not find any statistically significant difference in FDS 5.29±1.007 4.35±1.169 4.87±1.166 0.0124 BDS 4.00±1.304 2.82±1.704 3.47±1.158 0.026 the functional status improvement between stroke ADL 0.19±0.512 3.29±3.19 1.58±2.647 0.001 patients with and without cognitive impairment, IADL 0.38±1.071 5.41±5.06 2.63±4.258 0.001 the community ambulation rate was higher in co- FDS, forward digit span; BDS, backward digit span; gnitively normal group at the sixth month visit There was a significant positive correlation (6). Our study did not investigate this relationship between global cognitive function (MMSE further, but the Modified Rankin Scale score score) and working memory function, FDS (r= (MRS) score in our study was lower than 3, mea- 0.620) and BDS (0.536) (p<0.001). There was a ning the patients had only minimal symptoms and significant negative correlation between each co- mild to moderate physical disability (able to walk gnitive score (MMSE, FDS and BDS) with ADL without assistance). and IADL scores (p=0.005 and p=0.010, respec- After stroke, the most prominent impairment can tively). Higher scores on cognitive assessment be recognized in the patient’s processing speed, showed better cognitive function related to lower attention, and executive function. (10) Post-stroke scores on ADL and IADL scales, e.g. the patients dementia, particularly vascular dementia, causes were more independent (Table 3) slowing in cognitive flexibility, perceptual disorder, Table 3. Correlation between working memory and activities and impairment information retrieval at the time of of daily living (ADL) and instrumental activities of daily living stroke diagnosis. Up to 20-50% of stroke patients (IADL) suffer from the memory problem that manifests du- Cognitive ADL IADL scores r p r p ring a period following the stroke diagnosis (2). MMSE -0.772 <0.001 -0.757 <0.001 There were 17 patients (44.7%) patients with co- FDS -0.430 0.007 -0.443 0.005 gnitive impairment in our study. This proportion BDS -0.362 0.026 -0.413 0.010 MMSE, mini mental state examination; FDS, forward digit span; is similar to the result from a previous study by BDS, backward digit span; Ferreira et al. that reported a proportion of 37.8% patients with cognitive impairment in post-stroke DISCUSSION patients (5). Our data have shown a negative association betwe- Post-stroke cognitive dysfunction spectrum is de- en cognitive function and activities of daily living termined by the size and location of the infarcti- in post-stroke patients: the patients with better co- on, but there is no clear indication on the presen- gnitive function had better performance on daily ce of memory dysfunction due to stroke mainly life activities. This is in line with several previous episodic since it is believed that almost never studies. Ferreira et al. have shown that the presen- an infarction in the medial temporal lobe, brain

436 Fitri et al. Working memory in post-stroke patients

structure is predominantly involved in memory terplay of a number of interacting components. encoding and retrieval (7,18). Working memory, also labelled short-term me- This study showed that FDS and BDS scores were mory, is involved in the on-line maintenance and higher in cognitively normal group compared to active manipulation of information; it is generally the group with cognitive impairment. Our results conceived as a multicomponent system, which re- showed that a higher MMSE score is also associa- lies on a complex network of brain areas including ted with a better performance on working memory temporo parietal and frontal areas. tasks. This is consistent with previous studies that It is postulated to include an attentional control reported higher scores of all cognitive domains in component, the central executive, as well as sto- patients with normal cognitive function compared res involved in the short-term maintenance of to cognitively impaired post-stroke patients (4,5). material of different natures (8). The emergence The highest impact of stroke at the time of diagno- of post-stroke memory dysfunction fits in current sis is on the attention and executive function rather thinking of memory as a function of an intact ce- than on memory, which may be impaired at vario- rebral network, connecting several parts of the us post-stroke intervals, but memory dysfunction brain, including medial temporal lobes, anterior are also common after stroke (2). thalamic nucleus, mammillary body, fornix, and Cognitive impairment, particularly memory pro- prefrontal cortex with each other (the so-called blems following a stroke, can be evaluated and circuit of Papez). Any stroke in either of these assessed through neuropsychological assessments. structures or in the connections in between could Clinically, different neuropsychological asse- result in PMD (7). ssments are used to assess cognitive dysfunction Assessing the level of working memory impair- in terms of cognitive domain: memory evaluation ment in post-stroke patients may improve a reha- is proposed to be associated with memory types, bilitation plan. Approaches to enhancing memory short-term memory and working memory refer to may aim to reduce distraction, or increase effici- the perceptual and learning areas of the cognitive ency of encoding and retrieval (9). domain, which are processed by the frontal lobe This study had several limitations. The sample (2). Working memory is the ability to maintain size was relatively small, so the result cannot and manipulate information for a brief period of be generalized to all post-stroke patients. Also, time. The model consists of two slave systems: the we did not analyse the stroke type, location and phonological loop for verbal information and the severity that could affect the cognitive function, visuospatial sketchpad for visual and spatial in- although we had excluded patients with aphasia formation. The capacity of the phonological loop and history of prior dementia. is typically assessed using digit span tasks, while In conclusion, post-stroke cognitive impairment, the visuospatial sketchpad can be assessed using particularly deficit in working memory may ne- spatial span tests. Both working memory compo- gatively affect daily life activities. This result is nents function under the control of the central exe- important because cognitive dysfunction after cutive, which is recruited under higher memory stroke has implications for rehabilitation and tre- loads (14). Our study used MMSE and digit span atment strategies. to assess working memory. Subtest of MMSE can assess attention and concentration (by serial FUNDING subtraction), verbal memory (repetition of senten- This research is funded by Research Institute ces) and visuospatial (2 pentagons drawing). Digit Universitas Sumatera Utara according to Rese- span can assess attention, concentration and men- arch Contract TALENTA Universitas Sumatera tal control (2,13,14). Utara (No: 2590/UN5.1.R/PPM/2018; March Based on stroke location and severity, memory 16th 2018). disorder may occur for one or more memory types, eventually ending in memory decline and loss (2). TRANSPARENCY DECLARATION Working and episodic memory were found to be Conflicts of interest: None to declare. more regularly diminished in stroke victims (18). Memory is now considered as relying on the in-

437 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES 1. Goldstein LB, Adams R, Alberts MJ, Appel LJ, Bra- 9. Shigaki CL, Frey SH, Barrett AM. Rehabilitation of ss LM, Bushnell CD, Culebras A, DeGraba TJ, Go- poststroke cognition. Semin Neurol 2014; 34:496- relick PB, Guyton JR, Hart RG, Howard G, Kelly- 503. Hayes M, Nixon JV, Sacco RL. Primary prevention 10. Levine DA, Galecki AT, Langa KM, Unverzagt FW, of ischemic stroke: a guideline from the American Kabeto MU, Giordani B, Wadley V. Trajectory of Heart Association/ American Stroke Association cognitive decline after incident stroke. JAMA 2015; Stroke Council. Stroke 2006; 37:1583-633. 314:41-51. 2. Al-Qazzaz NK, Ali SH, Ahmad SA, Islam S, Moha- 11. Nakling AE, Aarsland D, NAess H, Wollschlaeger mad K. Cognitive impairment and memory dysfunc- D, Fladby T, Hofstad H, Wehling E. Cognitive de- tion after a stroke diagnosis: a post-stroke me- ficits in chronic stroke patients: neuropscyhological mory assessment. Neuropsychiatr Dis Treat 2014; assessment, depression, and self-reports. Dement 10:1677-91. Geriatr Cogn Disord Extra 2017; 7:283-96. 3. Mijajlovic MD, Pavlovic A, Brainin M, Heiss WD, 12. Folstein MF, Folstein SE, McHugh PR. "Mini-men- Quinn TJ, Hansen HB, Hermann DM, Assayag EB, tal state." A practical method for grading the cogniti- Richard E, Thiel A, Kliper E, Shin YI, Kim YH, Choi ve state of patients for the clinician. J Psychiatr Res S, Jung , Lee YB, Sinanovic O, Levine DA, Schle- 1975; 12:189–98. singer I, Mead G, Milosevic V, Leys D, Hagberg G, 13. Woods DL, Kishiyama MM, Yund EW, Herron TJ, Ursin MH, Teuschl Y, Prokopenko S, Mozheyko E, Edwards B, Poliva O, Hink RF, Reed B. Improving Bezdenezhnykh, Matz K, Aleksic V, Muresanu DF, digit span assessment of short-term verbal memory. Korczyn AD, Bornstein NM. .Post-stroke dementia J Clin Exp Neuropsychol 2011; 33:101–11. a comprehensive review. BMC Med 2017; 15:11. 14. Kessels RPC, Overbeek A, Bouman Z. Assessment 4. Oros, RI, Popescu CA, Iova CA, Mihancea P, Iova of verbal and visuospatial working memory in mild SO. The impact of cognitive impairment after stro- cognitive impairment and Alzheimer’s dementia. ke on activities of daily living. HVM Bioflux 2016; Dement Neuropsychol 2015; 9:301-5. 8:41-4. 15. Joshi A, Kale S, Chandel S, Pal DK. Likert scale: 5. Ferreira MGR, Moro CH, Franco SC. Cognitive explored and explained. British journal of applied performance after ischemic stroke. Dement Neurop- science and technology 2015; 7:396-403. sychol 2015; 9:165-75. 16. Katz S, Ford AB, Moskowitz RW, Jackson BA, Ja- 6. Paker N, Bugdayci B, Tekdos D, Kaya B, Dere C. ffe MW. Studies of illness in the aged. The index Impact of cognitive impairment on functional outco- of ADL: a standardized measure of biological and me in stroke. Stroke Res Treat 2010; pii652612. psychological function. JAMA 1963; 185:914–9. 7. Snaphaan L, Leeuw FE. Poststroke memory func- 17. Lawton MP, Brody EM. Assessment of older people: tion in nondemented patients: a systematic review self-maintaining and instrumental activities of daily on frequency and neuroimaging correlates. Stroke living. Gerontologist 1969; 9:179–86. 2007; 38:198-203. 18. Chellappan K, Mohsin NK, Ali SH, Islam S. Post- 8. Malouin J, Belleville S, Richards CL, Desrosiers stroke brain memory assessment framework. 2012 J, Doyon J. Working memory and mental practice IEEE-EMBS Conference on Biomedical Enginee- outcomes after stroke. Arch Phys Med Rehab 2004; ring and Sciences, Langkawi, Malaysia,17-19 De- 85:177-83. cember 2012. IEE, 2012: pp. 189-94.

438 ORIGINAL ARTICLE

The impact of pain intensity on quality of life of postherpetic neuralgia patients

Wizar Putri Mellaratna, Nelva K. Jusuf, Ariyati Yosi

Department of Dermatology and Venereology, School of Medicine, Universitas Sumatera Utara, Medan, Indonesia

ABSTRACT

Aim To investigate the impact of pain intensity of postherpetic neuralgia (PHN) patients.

Methods This cross sectional study included 30 PHN patients attended to the Dermatology and Venereology Department of the School of Medicine of Universitas Sumatera Utara during the period between April 2019 and October 2019. Zoster Brief Pain In- ventory questionnaire including the worst pain in 24 hours and the level of interference with seven domains of quality of life (QoL) (general activity, mood, walking ability, working, relationship with other people, sleep and enjoyment of life) were used. Pearson’s correlation (r) was categorized as very weak (0.0-< 0.2), weak (0.2-< 0.4), moderate (0.4-< 0.6), strong (0.6-<0.8), and very Corresponding author: strong (0.8-1.0). Wizar Putri Mellaratna Results There was a very strong correlation between worst pain Department of Dermatology and intensity and mood disturbance (r=0.846) and working status Venereology, School of Medicine, (r=0.818). Worst pain intensity had a strong correlation with ge- Universitas Sumatera Utara neral activity (r=0.673), relationship with other people (r=0.653), Jl. Dr. Mansyur No.5 20222 Medan, sleep (r=0.774) and disturbance of enjoyment of life (r=0.783). Therefore, the correlation between worst pain intensity and wal- Indonesia king ability was moderate (r=0.475). The worst pain intensity had Phone: +62 852 6014 1026; a significant correlation with all seven domains of QoL (p< 0.05). E-mail: [email protected] ORCID ID: https://orcid.org/0000-0002- Conclusion Pain intensity influences the QoL with directly proportional correlation between pain intensity and disturbance of 5945-8898 QoL.

Key words: cross-sectional herpes zoster, pain, questionnaire, Original submission: survey 13 December 2019; Revised submission: 03 January 2020; Accepted: 11 March 2020 doi: 10.17392/1111-20

Med Glas (Zenica) 2020; 17(2): 439-444

439 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION than 90 days of herpes zoster rash. The shortest duration of PHN was 3 months, while the longest Postherpetic neuralgia (PHN) is a long term neu- was 3 years after the HZ onset. Inclusion criteria ropathic pain experienced commonly within more were patients who had been diagnosed with PHN than three months after onset of herpes zoster (HZ) in the last 5 years (2015-2019), aged over 40 years (1). Based on epidemiological data taken from 13 (40-44, 45-49, 50-54, 55-59, and ˃60 age groups), hospitals in Indonesia from 2011 to 2013, it was who signed informed consent for investigation. found that about 26.5% of the HZ total cases (593) Patient demographic data included age, sex, and developed into PHN. The risk factors of PHN employment history from medical records. include an age over 50 years, females, duration of pain during HZ, and psychological stress (2). The research protocol was approved by Ethics Commission in the School of Medicine of Univer- Chronic pain of PHN causes sleeping disorders, sitas Sumatera Utara and Adam Malik Hospital. depression which leads to unsocialised behaviour, anorexia, weight loss, fatigue and disturbances of Methods dressing, shower, general activity, traveling, shopping, cooking and household activities (3). Chen Demographic data included age, sex, employment et al. have reported that economic burden caused record, phone number and patient’s address. Af- by PHN was higher than in HZ, and this included ter being reconfirmed for their willingness to medicines costs, consultation, and hospital tre- participate in the research, data were collected in atment (4). Deterioration of quality of life (QoL) the patients’ houses. The researcher requested the of PHN patients is assessed with a questionnaire patients to fill out the ZBPI questionnaire (with or interview (5), while Zoster Brief Pain Inventory permission) (8). The questionnaire evaluated the (ZBPI) is one of the questionnaires that assess pain level in 24 hours with Likert scale 0-10 (8): pain intensity and reflect QoL in the same patients. score 0 indicated no pain, up to score 10 which Thus, an improvement to evaluate the impact of indicated the worst pain felt by the patients. The pain intensity on QoL of PHN patients remains a patients filled out the level of interference with matter of urgency, considering the real impact in seven domains of quality of life (QoL) questi- disruption of activities of daily living (ADL) (6). onnaire (8) including general activity, mood, Coplan stated that intensity of pain from ZBPI walking ability, working, relationship with other questionnaire strongly correlated with disturbance people, sleep and enjoyment of life. The patients of quality of life of PHN patients (7). There was no responded with a score of 0 to 10 according to the research about the correlation of pain intensity and level of disturbance they had experienced. Score quality of life of PHN patients in Indonesia. 0 indicated no disturbance, up to score 10 indicated worst disturbance in the daily activity. The aim of this study was to investigate the impact of pain intensity of post herpetic neuralgia Statistical analysis (PHN) patients. Normality test was performed by using one sam- PATIENTS AND METHODS ple Kolmogorov Smirnov test. Pearson’s correlation was used as hypothesis testing. The p<0.05 Patients and study design was considered as statistically significant. This analysis was classified into correlation categori- This a cross sectional study conducted at the de- es: very weak (r 0.0-< 0.2), weak (r 0.2-< 0.4), partment of Dermatology and Venereology School moderate (r 0.4-< 0.6), strong (r 0.6-< 0.8), and of Medicine of Universitas Sumatera Utara, Me- very strong (r 0.8-1.0) (9). The increase of pain dan, and 30 PHN patients were recruited during intensity is accompanied by the increase of QoL the period between April 2019 and October 2019. interference of PHN patients. Data on patients with postherpetic neuralgia diagnosis were collected from medical records of RESULTS Rumah Sakit Pendidikan Universitas Sumatera Utara, and two local public health centres located This research involved 30 PHN patients, most in Teladan and Padang Bulan. Postherpetic neu- of them were over the age of 60, 16 (53.3%) ralgia was defined as persisting pain after more and one (3.3%) was in the 50-54 years age

440 Mellaratna et al. Postherpetic neuralgia and quality of life

group. Based on gender, the patients were pain intensity had positive correlation with all mostly female, 17 (56.7%), compared with seven domains of QoL (Table 3). male, 13 (43.3%). Majority of patients worked as private employees and were retired, seven Table 3. Correlation between pain intensity and QoL The worst pain intensity in Correlation coefficient (23.3%) of each, followed by housewives and p category (No of patients) farmers, six (20.0%) and two (6.7%), respecti- General activity r = 0.673 (30) < 0.001 vely (Table 1). Mood r = 0.846 (30) < 0.001 Walking ability r = 0.475 (30) 0.008 Table 1. Demographic characteristics of 30 patients with Working ability r = 0.818 (30) < 0.001 postherpetic neuralgia Relationship with other people r = 0.653 (30) < 0.001 Variable No (%) of patients Sleep r = 0.774 (30) < 0.001 Enjoyment of life r = 0.783 (30) < 0.001 Age (years) 40-44 3 (10.0) 45-49 4 (13.3) DISCUSSION 50-54 1 (3.3) 55-59 6 (20.0) The worst pain intensity had a very strong corre- ≥ 60 16 (53.3) Gender lation with increasing interference of QoL and it Female 17 (56.7) affected all seven domains of ADL in ZBPI with Male 13 (43.3) significant impact on mood, sleep and general Working status activity (10). In this research, PHN was mostly Salesman 4 (13.3) Farmer 2 (6.7) found in the age group over 60 years. Munoz- Government employees 4 (13.3) Quiles et al. reported the increasing risk of PHN Retired 7 (23.3) was two times higher in age group 60-69 years, Private employee 7 (23.3) Housewife 6 (20.0) three times higher in the age group 70-79 years, and 3.67 higher in the age group over 80 years Majority of patients had worst pain intensity, (11).Weaver reported that incidence of PHN was with score 9 accounted for nine (30.0%) pati- the highest at the age of 70 (73%), and it was ents, and minority of patients had score, 4 and 5, followed by the age 55 years (27%) (12). two (6.7%) patients each. Mean (SD) was 7.17 Incidence of PHN mostly involved females com- (2.321) (Table 2). pared with the males because females have a low pain threshold. Whilst curiosity about pain Table 2. Pain intensity in postherpetic neuralgia patients for females is greater than that in males, and wo- Pain intensity No (%) of patients Men (SD) men have lower sensitization to sensory stimulus 0 0 7.17 (2.321) (13). In this research, it is shown that the female 1 0 2 0 proportion was greater than male. Schmidt-Ot et 3 4 (13.3) al. and Nahm et al. reported that about 63% and 4 2 (6.7) 62.65% of PHN patients were female (14,15). 5 2 (6.7) 6 0 (0) Majority of patients had already retired becau- 7 6 (20.0) se generally PHN involved older patients. Po- 8 4 (13.3) stherpetic neuralgia was caused by the economic 9 9 (30.0) 10 3 (10.0) burden due to the high absenteeism rate at work (16). In this research, majority of patients wor- There was a very strong correlation between ked as private employee and already retired. This the worst pain intensity and mood disturbance corresponds to Nakamura et al. that proposed (r=0.846), as well as working status (r=0.818) the highest occurrence of PHN patients in reti- (p<0.001). The worst pain intensity had a strong red conditions (39.6%), housewives (33.7%) and correlation with general activity (r=0.673), re- those still actively working (23%) (17). Johnson lationship with other people (r=0.653), sleep et al. also stated the same that 51.9% of PHN pa- (r=0.774), and disturbance on enjoyment of life tients were already retired, those with full-time (r=0.783) (p< 0.001). Therefore, the correlation employment status (16%), housewives (13.7%), between worst pain intensity and walking abi- part-time employment status (11.4%), and the lity was moderate (r=0.475; p<0.05). The worst employed (6.9%) (18).

441 Medicinski Glasnik, Volume 17, Number 2, August 2020

From the dimension of pain intensity items, pain very strong correlation between worst pain in- at its worst has the highest internal validity com- tensity and general activity (7). Nevertheless, the pared with others (Cronbach’s alpha 0.77-0.85), results of our research correspond with Oster et and also has the strongest correlation with distur- al. reporting moderate to strong impact of pain bance of ADL (Cronbanch’s alpha 0.76-0.85) intensity value (with mean SD) on general acti- (19). Worst pain intensity was also recommen- vity (40% ±3.7-3.1) (27). ded by Food and Drugs Administration (FDA) The present research indicated worst pain inten- to assess patient reported outcomes (PRO) (20). sity strongly correlated with relationship with Our research indicated that most of PHN patients others. Coplan et al. research reported that worst had a worst pain intensity score of 9. This is in pain intensity moderately correlated with relati- accordance with Wyrwich et al. who reported on with others (7). Deshpande et al. also stated mean (SD) worst pain intensity score from the that there was a decline on QoL from relationship ZBPI questionnaire was 6.8 (2) (21). Van Seven- with others (using SF-36 questionnaire) with sco- ter et al. reported that mean (SD) in placebo gro- re 70.3 (24.2) in PHN patients with VAS score 5 up (who have not yet received pregabalin) was (2.1), while in the control group normative valu- 6.86 (1.49) with median (minimum-maximum) 7 es were obtained for relationship with other peo- (1.71-10) (10). ple at around 88.3 (18.6) (28). This research indicated that worst pain inten- Our results indicated that worst pain intensity sity was very strongly correlated with mood strongly correlated with sleep disturbance, which disturbance. This is different from Coplan et al. corresponds with Coplan et al. study (7). Vinik who stated that worst pain intensity was stron- et al. found that in PHN patients who receive no gly correlated with mood disturbance (7). This treatment with 6.64 of mean pain score inten- research is in accordance with Schelereth et al. sity, a disturbance of sleep at the level of seve- who mentioned Pain Rating Index (PRI) had a re (43.3%), moderate (38.1%) and mild (18.6%) positive correlation with allodynia and Hospital was experienced; relationship between pain in- Anxiety and Depression Scale (HADS) for anxi- tensity and sleep interference can occur in two ety and depression (22). Saxena et al. reported directions, improvement and derivation of one a negative correlation between Brain Derived component which can affect others. Vinik et al. Neurotropic Factor (BDNF) and pain intensity also mentioned that by using pregabalin in PHN in PHN patients, where an increase in pain inpatients, pain intensity decreased, which can im- tensity is followed by a decrease in BDNF level, prove the sleep quality. It is because pregabalin and vice versa (23). Chen et al. also has reported works in the pathways responsible for pain res- that PHN/HZ was related to depression through a ponse and sleep (29). Cao et al. reported that mechanism mediated by cellular immunity (24). chronic pain in PHN caused sleep disturbance, This research showed that there was a very strong based on an analysis of brain regional activities correlation between worst pain intensity and discovering a limbic system disturbance, which working interference. Aunhachoke et al. menti- regulates sleep activity, otherwise it caused persi- oned that pain intensity was moderately corre- stent chronic pain in PHN patients (30). lated with work (25). Coplan et al. also menti- This research indicated that worst pain intensity oned strongly correlated occurrence between strongly correlated with enjoyment of life which worst pain intensity and work disturbance (7). is consistent with Coplan et al. research (7). Oster Meanwhile, Serpel et al. reported a decrease in reported that in PHN patients mean (SD) score effectiveness in their work in PHN patients who of worst pain intensity was 6 (2.4), about 48% were still actively working (6). experienced impaired ability to enjoy life with This research indicated there was a strong corre- moderate to severe level (27). lation between worst pain intensity and general We reported that worst pain intensity moderately activity. Laurent et al. stated the disturbance of correlated with walking ability. Aunhachoke et general activity including ability to dress (54%), al. reported that correlation between pain inten- move (47%), and wash (26%) (26). This is diffe- sity and walking ability was low/moderate (25). rent from Coplan et al. research that reported a This is different from Coplan et al. who menti-

442 Mellaratna et al. Postherpetic neuralgia and quality of life

oned the presence of strong correlation between Worst pain intensity had a very strong correlati- worst pain intensity and walking ability (7). on to mood and work, and a strong correlation to Laurent et al. reported that impact of pain inten- general activity, relationship with other people, sity in PHN was significant to general activity, sleep and enjoyment of life, but it has moderate mood and enjoyment of life (26). This is also re- correlation with walking ability. levant to Freeman’s research that reported signi- Intensity of pain level in PHN patients has a si- ficant impact in pain intensity score with mean gnificant impact of the QoL, so the authors re- value of 7.3 on disturbance of sleep, enjoyment of commend to treat the PHN patients adequately life and mood (31). Correlation test analysis re- based on patient’s characteristic, routine monito- sults in this study indicated the influence of worst ring of the treatment outcomes and prevention of pain intensity in PHN patients to their QoL. An PHN complication after onset of HZ by admini- increase in pain intensity followed by an increase stering vaccination. in QoL interference, with significant impact on mood, work, general activity, relationship with ACKNOWLEDGEMENT others, sleep and enjoyment of life were found. We would like to express our gratitude to Rumah This study has some limitations including no Sakit Pendidikan Universitas Sumatera Utara, exclusion of patients who were currently using and two local public health centres located in Te- analgesics and the wide range of persisted pain ladan and Padang Bulan, Medan for permitting after HZ rash (3 months-3 years), which might the use of patients’ data from medical records. influence the study results. Pain in PHN patients was normally intense in 3 months after HZ onset FUNDING and reduced gradually in several years after (32). No specific funding was received for this study. In conclusion, the pain intensity influences the TRANSPARENCY DECLARATION QoL, with directly proportional correlation between pain intensity and disturbance on QoL. Conflicts of interest: None to declare.

REFERENCES

1. Schmader KE, Dworkin RH. The epidemiology and 6. Serpell M, Gater A, Carrol S, Abetz-Webb L, natural history of herpes zoster and post herpetic ne- Mannan A, Johnson R. Burden of post-herpetic neuralgia. In: Watson CPN, Gershon AA, Oxman MN. uralgia in a sample of UK residents aged 50 years Herpes Zoster: Post Herpetic Neuralgia and other or older: findings from the Zoster Quality of Life Complications. Switzerland: Springer, 2017. https:// (ZQOL) study. Health Qual Life Outcomes 2014; www.springer.com/gp/book/9783319443461// (10 12:92. February 2018). 7. Coplan PM, Schmader K, Nikas A, Chan IS, Choo 2. Pusponegoro E, Nilasari H, Lumintang H, Niode NJ, P, Levin MJ, Johnson G, Bauer M, Williams HM, Daili SF, Djauzi S. Buku Panduan Herpes Zoster di Kaplan KM, Guess HA, Oxman MN. Development Indonesia (Indonesia Shingles Book Guide) [in In- of a measure of the burden of pain due to herpes zo- donesian]. Jakarta: BP FKUI, 2014. ster and postherpetic neuralgia for prevention trials: 3. Schmader KE, Oxman MN. Varicella and Herpes adaptation of the brief pain inventory. J Pain 2004; Zoster. In: Goldsmith LA, Katz SI, Gilchrest, Paller 5: 344-56. AS, Leffell DJ, Wolff K. Fitzpatrick‘s Dermatology 8. Cleeland CS. The Brief Pain Inventory User Guide. in General Medicine. 8th Ed. New York: The Mc Texas: The University of Texas MD Anderson Can- Graw Hill Company Inc., 2012:2393- 8. cer Center; 2009 (05 May 2018) https://www.mdan- 4. Chen Q, Hsu TY, Chan R, Kawai K, Acosta CJ, derson.org/research/departments-labs-institutes/ Walia A, Pan JY. Clinical and economic burden of departments-divisions/symptom-research/symptom- herpes zoster and postherpetic neuralgia in patients assessment-tools/brief-pain-inventory.html from the national skin center, Singapore. Dermatol 9. Dahlan MS. Statistik untuk kedokteran dan keseha- Sin 2015; 33:201-5. tan (Statistic for Medicine and Health). [in Indonesi- 5. Lukas K, Edte A, Bertrand I. The impact of herpes an] 6th Ed. Jakarta: Salemba Medika, 2014. zoster and post herpetic neuralgia on quality of life: 10. Van Seventer R, Sadosky A, Lucero M, Dukes E. patient-reported outcomes in six European countri- A cross-sectional survey of health state impairment es. Z Gesundh Wiss 2012; 20:441-51. and treatment patterns in patients with postherpetic neuralgia. Age Ageing 2006; 35:132-7.

443 Medicinski Glasnik, Volume 17, Number 2, August 2020

11. Muñoz-Quiles C, López-Lacort M, Orrico-Sánchez 22. Schlereth T, Heiland A, Breimhorst M, Féchir M, A, Díez-Domingo J. Impact of postherpetic neural- Kern U, Magerl W, Birklein F. Association between gia: A six year population-based analysis on people pain, central sensitization and anxiety in postherpe- aged 50 years or older. J Infect 2018; 77:131-6. tic neuralgia. Eur J Pain 2014; 19:193-201. 12. Weaver BA. The burden of herpes zoster and po- 23. Saxena AK, Lakshman K, Sharma T, Gupta N, Ba- stherpetic neuralgia in the United States. J Am Oste- nerjee BD, Singal A. Modulation of serum BDNF opath Assoc 2007; 107:S2-7. levels in postherpetic neuralgia following pulsed 13. Belfer I. Pain in women. Agri Pain 2017; 29:51-4. radiofrequency of intercostal nerve and pregabalin. 14. Schmidt-Ott R, Schutter U, Simon J, Nautrup BP, Pain Manag 2016; 6:217-27. Von Krempelhuber AV, Gopala K, Anastassopoulou 24. Chen MH, Wei HT, Su TP, Li CT, Lin WC, Chang A, Guignard A, Curran D, Matthews S, Espie E. In- WH, Chen TJ, Bai YM. Risk of depressive disorder cidence and costs of herpes zoster and postherpetic among patients with herpes zoster: a nationwide po- neuralgia in German adults aged ≥ 50 years: a pros- pulation-based prospective study. Psychosom Med pective study. J Infect 2018; 76:475-82. 2014; 76:285-91. 15. Nahm FS, Kim SH, Kim HS, Shin JW, Yoo SH, 25. Aunhachoke K, Bussaratid V, Chirachana- Yoon MH, Lee DI, Lee YW, Lee JH, Jeon YH, Jo kul P, Chua-Intra B, Dhitavat J, Jaisathaporn K, DH. Survey on the treatment of postherpetic neu- Kaewkungwai J, Kampirapap K, Khuhaprema T, ralgia in Korea; multicenter study of 1.414 Patients. Pairayayutakui K, Pitisuttihum P, Sindhvananda J, Korean J Pain 2013; 26:21-6. Thaipisuttikul Y. Measuring herpes zoster, zoster- 16. Vinik A, Emir B, Cheung R, Whalen E. Relationship associated pain, post-herpetic neuralgia-associated between pain relief and improvements in patient loss of quality of life, and healthcare utilization and function/quality of life in patients with painful dia- costs in Thailand. Int. J. Dermatol 2011; 50:428-35. betic peripheral neuropathy or postherpetic neuralgia 26. Laurent B, Vicaut E, Leplège A, Bloch K, Leute- treated with pregabalin. Clin ther 2013: 35:612-22. negger E. Prevalence and impact on quality of life 17. Nakamura H, Mizukami A, Adachi K, Matthews S, of post-herpetic neuralgia in French medical cen- Holl K, Asano K, Watanabe A, Adachi R, Kiuchi M, ters specialized in chronic pain management: The Kobayashi K, Sato K, Matsuki T, Kaise T, Curran D. ZOCAD study. Med Mal Infect 2014; 44:515-24. Economic burden of herpes zoster and postherpetic 27. Oster G, Harding G, Dukes E, Edelsberg J, Cleary neuralgia in adults 60 years of age or older: results PD. Pain, Medication use, and health-related quality from a prospective, physician practice-based cohort of life in older persons with postherpetic neuralgia: study in Kushiro, Japan. Drugs Real World Outco- results from a population-based survey. J Pain 2005; mes 2017; 4:187-98. 6:356-63. 18. Johnson RW, Wasner G, Saddier P, Baron R. Po- 28. Deshpande MA, Holden RR, Gilron I. The impact of stherpetic neuralgia: epidemiology, pathophysio- therapy on quality of life and mood in neuropathic logy and management. Expert Rev Neurother 2014; pain: what is the effect of pain reduction?. Anesth 7:1581-95. Analg 2006; 102:1473-9. 19. Atkinson TM, Mendoza TR, Sit L, Passik S, Scher 29. Vinik A, Emir B, Parsons B, Cheung R. Prediction HI, Cleeland C, Basch E. The brief pain inventory of pregabalin-mediated pain response by severity of and its “pain at its worst in the last 24 hours” item: sleep disturbance in patients with painful diabetic clinical trial endpoint considerations. Pain Med neuropathy and postherpetic neuralgia. Pain Med 2010; 11:337-46. 2014; 15:661-70. 20. U.S. Food and Drug Administration. FDA 30. Cao S, Song G, Zhang Y, Xie P, Tu Y, Li Y, Yu T, Draft Guidance for Industry. Analgesic Yu B. Abnormal local brain activity beyond the pain Indications: Developing Drugs and Biological matrix in postherpetic neuralgia patients: a resting- Products (05 May 2018). https://www.policymed. state functional MRI Study. Pain Physician 2017; com/2014/02/fda-draft-guidance-analgesic-indicati- 20:E303-14. ons-developing-drug-and-biological-products.html 31. Freeman R, Wallace MS, Sweeney M, Backonja 21. Wyrwich KW, Kawata AK, Thompson C, Holm- MM. Relationships among pain quality, pain impact, strom S, Stoker M, Wiklund I. Validation of the and overall improvement in patients with postherpe- self-assessment of treatment questionnaire among tic neuralgia treated with gastroretentive gabapentin. patients with postherpetic neuralgia. Pain Res Treat Pain Med 2015; 16:2000-11. 2012; 621619. 32. Mick G, Hans G. Postherpetic neuralgia in Europe: the scale of the problem and outlook for the future. J. Clin. Gerontol Geriatr 2013; 4:102-8.

444 ORIGINAL ARTICLE

The level of tumour necrosis factor-alpha and its relationship to the cognitive function of Malayan-Mongoloid patients with schizophrenia

Mustafa M. Amin1, Abdul Rasyid2, Elmeida Effendy1, Nurmiati Amir3, Dwi Anita Suryandari4

1Department of Psychiatry, 2Department of Radiology; Faculty of Medicine, Universitas Sumatera Utara, Medan, 3Department of Psy- chiatry, 4 Department of Biology; Faculty of Medicine, University of Indonesia, Jakarta; Indonesia

ABSTRACT

Aim Schizophrenia is a mental disorder and one of the suspected causes is cytokines. One of them is tumour necrosis factor-alpha (TNF-α). Cytokines have the potential to affect cognitive function. The study aimed to find a correlation of TNF-α level with the Mini-Mental State Examination (MMSE) score in patients with schizophrenia (PwS), and comparing the level of TNF-α levels between PwS and healthy controls.

Methods We conducted a cross-sectional analytic study and the study designs were correlation and comparative analysis, i.e. using a Mann-Whitney U test. A total number of 100 subjects were collected, and they were divided into two groups of PwS and con- Corresponding author: trol group, respectively. Mustafa M. Amin Results The results found that most of the PwS subjects were 39 Department of Psychiatry, men (78.0%), while the control group were 28 men (56.%). The Faculty of Medicine, differences in TNF-α levels between PwS and control groups were Universitas Sumatera Utara found to be significant p <0.001, there was no significant correlation between TNF-α level and the score of MMSE of the PwS with Jl. dr. T. Mansur No. 5, Medan, Indonesia p = 0.938, with a very weak correlation that was r = -0.011, and a Phone: +62 81 2600 1772; negative correlation direction. E-mail: [email protected]; [email protected] Conclusion There was a significant difference between TNF-α level of PwS and control group, i.e. PwS group had lower TNF-α ORCID ID: http://orcid.org/0000-0003- level compared to the control group. The TNF-α level of PwS gro- 0912-9372 up had a very weak effect on the cause of cognitive dysfunction in PwS group, yet the higher level of it could reduce MMSE score in PwS group. Original submission: Key words: cytokines, cognitive dysfunction, psychotic disorders 28 November 2019; Revised submission: 21 January 2020; Accepted: 28 February 2020 doi: 10.17392/1108-20

Med Glas (Zenica) 2020; 17(2): 445-450

445 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION Malayan. Our study focused on the Proto Ma- layan type only. Schizophrenia is a mental disorder that usually appears in late adolescence or early adulthood, The aim of this study was to investigate the effect which affects about 21 million people worldwide of TNF-α on the cognitive function of patients of (1). Thus far, the aetiology of schizophrenia is un- the Malayan-Mongoloid race with schizophrenia known, but several hypotheses related to this have in North Sumatera, and it was the first study in been stated: schizophrenia is a neurodevelopmen- North Sumatera which assessed this. tal disease; it is a neurodegenerative disease; and PATIENTS AND METHODS it is a progressive neurodevelopmental disease (2). Another important point is the implication of cy- Patients and study design tokines in the aetiology or pathology of schizophrenia (3). One of the cytokines that play an es- This cross-sectional analytical study was carried sential role for schizophrenia is tumour necrosis out at the Outpatient and Inpatient Installation of factor-alpha (TNF-α). TNF-α level also increased the North Sumatera Psychiatric Hospital (R. S. significantly in patients with schizophrenia (PwS) J. Provsu) Prof. M. Ildrem during the period of 3 compared to healthy controls (4-7). Studies with months in 2019. This psychiatric hospital is a re- the same results were reported by O'Brien, Scully, ferral psychiatric hospital in North Sumatra Pro- and Dinan (8), i.e. there were significant differ- vince, and had an inpatient capacity of 400 beds. ences in TNF-α level in the schizophrenia group Patients who came to this psychiatric hospital were of 13.49 pg/ ml±0.42 and 6.79 pg/ml±0.42 in the almost entirely Malayan-Mongoloid Race. The tar- control group. This result is supported by Czerski get population was PwS that were the patients of the et al. (9) who found a significant difference in the Outpatient and Inpatient Installation of North Su- frequency of TNF-α between PwS and control matera Psychiatric Hospital. The patients included group, 87.9% and 82.9%; this frequency increased the PwS group, and the healthy control group that to 90.7% of those who had a history of schizophre- fulfilled the inclusion and exclusion criteria. Inclu- nia in the first and second-degree relatives. Similar sion criteria of PwS were: schizophrenic patients th results have also been reported by Rosario García- diagnosed based on the 10 edition of the Interna- Miss et al. (10) who found that there was a sig- tional Classification of Disease and Related Health nificant difference between PwS and control group Problems criteria (14), aged 15-40 years, Malayan- regarding the TNF-α values of 10.63 ± 4.65 pg/mL Mongoloid Race, cooperative and willing to be in- and 6.74±4,00 pg/mL, TNF-α levels in PwS were terviewed. Exclusion criteria were: having a history higher than in the group. of previous psychiatric disorders, having a general medical condition that affected brain structure, and Inflammation is thought to have a role in the pat- obesity (defined by body mass index - BMI of ≥ hophysiological process of schizophrenia, and cu- 30). The inclusion criteria for the control group rrently, various studies are being conducted to find were: age 15-40 years, Malayan-Mongoloid Race, out the anti-inflammatory effect in PwS (11). A cooperative and willing to be interviewed, and no literature review from Baune et al. (12) concluded family history of having a mental disorder. Exclu- that TNF-α has a neuroprotective or neurodege- sion criteria for the control group were: having a nerative function. Therefore, it remains debatable history of previous psychiatric disorders, having whether TNF-α can maintain or decrease cogniti- a general medical condition that affected brain ve function. A study conducted by Xiu et al. (13) structure, and obesity. We took the control group found that TNF-α could mediate cognitive severity from the people who lived near the hospital. towards PwS. Thus far, different results have been found in previous studies. Therefore, more resear- The sample size calculation used the following ch is needed to support the statement of cytokines, formula: such as conducting a study of the TNF-α, which has a relation with schizophrenia. Malayan-Mongoloid people are one of the ra- The sample size was based on the sample size ces living in Medan, Indonesia. It consists of table with alpha 5% two-sided beta hypothesis two sub-races, i.e. Proto Malayan and Deutro of 10%, in which the assumption of the standard

446 Amin et al. TNF-α and cognitive function

deviation ratio was 1, it was found that n1 = n2 = higher average age of 35.42±2.78 years in PwS 21 subjects. In this study, 50 subjects were taken group. Higher BMI, of 23.79±2.98, in the control for each group. group was found. The average MMSE score was Next, the subjects and their relatives were asked to 21.00±4.56 in PwS group (Table 1). read the letter of statement of the research and sign Table 1. Baseline characteristics of patients with schizophre- the consent form after an explanation about the nia (PwS) and controls participation in the study. The study was approved Group Variable by the Health Research Ethical Committee of Fa- PwS (n=50) Control (n=50) culty of Medicine, Universitas Sumatera Utara. Gender (No, %) Male 38 (76.0) 28 (56.0) Female 12 (24.0) 22 (44.0) Methods Age (average±SD) (years) 35.42±2.78 29.38±5.75 BMI (average±SD) 21.57±2.02 23.79±2.98 Blood plasma sampling was carried out as MMSE* 21.00±4.56 - followed: blood was drawn with a sterile syringe *Data was only for the patients with schizophrenia BMI, body mass (aseptic) from a vein in the area where the upper index; MMSE, Mini-Mental State Examination arm meets the forearm (median cubital vein) for 6 mL. The blood was then kept in a vacuta- The TNF-α level for PwS was 3.24 (0.65-43.80) iner containing ethylenediaminetetraacetic acid pg/dL, while for the control group it was 16.25 (EDTA) and stored at 4-8 oC until the serum and (4.80-56.10) pg/dL (p<0.001) (Table 2). The box- plasma were separated. plot comparison of these two groups was shown in Figure 1. We found that the PwS group had the The enzyme-linked immunosorbent assays (ELI- lowest TNF-α level i.e. 0.65 pg/dL, much less var- SA) examination was then performed using the ied in the TNF-α level and lower median compared human Quantikine TNF-alpha kit (R&D systems, to the control group. The highest of the TNF-α lev- Minneapolis, MA, USA) and read the results el was found in the control group, i.e. 56.10 pg/dL. using the Thermo-Fisher machine. The cognitive function of the PwS was measured using Mini- Table 2. Comparative analysis of TNF-α level between pa- Mental State Examination (MMSE) rating scale tients with schizophrenia (PwS) and the control group (15) that had been validated in the Indonesian Group language by Geriatric Psychiatry Section of In- Variable PwS (n=50) Control (n=50) p donesia Psychiatry Association (16); the total TNF-α level (mean) 3.24 16.25 <0.001 score was divided into normal (24-30), probable (pg/dL) (0.65-43.80) (4.80-56.10) cognitive disorder (17-23), and definite cognitive disorder (0-16). The rating scale itself only took less than 10 minutes to complete, and was relatively easy to use.

Statistical analysis Statistical analysis began by normalizing the data using Saphiro-Wilk normality test. We found that the result was abnormal data, and continued the analysis using Mann-Whitney U test to compare the difference between the TNF-α level of PwS and control, and the Spearman correlation test to find the relationship of TNF-α level with MMSE Figure 1. Boxplot graph results of comparison analysis of score in the PwS group. TNF-α level between patients with schizophrenia (PwS) and the control group RESULTS There was no significant correlation between Among 100 patients analysed (50 patients in each TNF-α level and MMSE score (p=0.938); the cor- PwS and control group) the males predominated relation coefficient was very weak, and the direc- over female patients in both groups, 38 (76.0%) tion was negative (r= -0.011). The higher TNF-α in PwS and 28 (56.05) in control group, with level resulted in lower MMSE score (Figure 2).

447 Medicinski Glasnik, Volume 17, Number 2, August 2020

drenaline, and glutamate (33). A low level of inflammatory cytokines in the brain could still affect complex brain functions such as memory, mood, anxiety, cognition, and nerve activity (34,35). Cytokines, in this case, TNF-α had an essential role in regulating complexity including immunity and inflammation. Thus, the low TNF-α level in the PwS group indicated that there had been a defect during the induction of inflammatory pathways or active inhibition of these cytokines (17). Our results show that there was no significant Figure 2. Scatter graph correlation analysis results between correlation between TNF-α level and MMSE the TNF-α level and Mini-Mental State Examination (MMSE) score of patients with schizophrenia score, and it had a very weak correlation coefficient. In schizophrenia, cognitive dysfunction DISCUSSION happened before the appearance of positive and negative symptoms (12). The existence of TNF-α The results in this study found predomination of had an important role related to immunity and in- male PwS patients, and it was the same as repor- flammation in the brain (36), and changes in the ted by Lv et al. (17), Garcia-Miss et al. (10), Kunz level of TNF-α start when the PwS is still in the et al. (18), Tian et al. (19), Naudin et al. (4), and mothers’ womb (37). One mechanism that allows Kubistova et al. (20). Different results found that TNF-α to influence schizophrenia is through neu- there were more females with schizophrenia than regulin-1. The neuregulin-1 gene along with the males as reported by Kowalski et al. (6), and Hope erb84 receptor acted for the occurrence of plasti- et al. (21). It was explained by the fact that males city, myelination, and the formation of long-term with schizophrenia were more frequently hospita- potentiation, thus it had an important role in the lised compared to females, because males are less cognitive function (12). The results of our study adherent to the treatment, and often commit sui- were supported by a study conducted by Hennessy cide (22), or because females with schizophrenia et al. (38), which found that high TNF-α could in- were more responsive to the treatment and less duce a decrease of working memory. In addition, than 50% experienced inpatient care (23). inhibition of TNF-α could also improve memory In this study the mean of TNF-α level in the PwS loss and spatial learning (39). The existence of group was lower compared to the control group. TNF-α was also important in learning activities Similar results were reported by Chiang et al. and memory because the presence of TNF-α could (24) who found lower TNF-α in males. Other stu- interfere with both processes. As stated before, in- dies found the same results such as the study by hibition of TNF-α could restore cognitive function Tian et al. (19), and Zhu et al. (25). We assumed (40). Some of the mechanisms offered related to that the similarities that we found were due to all the above statement are TNF-α having a contribu- of these studies conducted in Mongoloid race. tion in astrogliosis, apoptosis, neurogenesis and Many studies showed different results, e. g. the permeability in the endothelial cell layer; thus, TNF-α level in the PwS group was higher com- it can influence the cognitive function (41). This pared to the control group (4,26-30), and most process started with prenatal inflammation resul- of them were done in Caucasian race, except one ting in abnormalities of cytokine levels including study that was done in Indonesia. TNF-α resulting in neurodevelopment disorders. Some conditions that potentially caused lower Specifically, concerning cognition, these effects TNF-α level included antipsychotic drugs con- included its relationship with the quality of life and sumed by PwS (14,31), the chronicity of the overall function of PwS, which later influenced the schizophrenia (17), and variations in vitamin D outcome of the disease (42). This is supported by levels in the body (32). Cytokines were thought a literature review written by Misiak et al. (43), to be involved in the regulation of several neuro- they also confirm that TNF-α has contributed to transmitters, such as dopamine, serotonin, nora- cognitive impairment in PwS.

448 Amin et al. TNF-α and cognitive function

In conclusion, no relationship between TNF-α and nurses at North Sumatera Psihiatric Hospital and cognitive function of PwS was found. We who had been cooperative with us, especially to thought that the age of the patients affected our Universitas Sumatera Utara for granting a fund results. Based on our findings of the coefficient for this study. correlation, the level of TNF-α may have a very weak effect on the cognitive function of PwS. It FUNDING contradicts with previous studies that suggested The study was funded by TALENTA funding of that cytokines had an important role in influen- Universitas Sumatera Utara. cing the cognitive function. TRANSPARENCY DECLARATIONS ACKNOWLEDGEMENTS Competing interests: None to declare. We want to thank all the patients and people who were willing to join our study, and the physicians

REFERENCES

1. World Health Organization. Schizophrenia. www. tumor necrosis factor alfa levels, but not methyle- who.int/mental_health/management/schizophrenia/ netetrahydrofolate reductase C677T polymorphism, en (20 April 2018) are risk factors for schizophrenia. J Psychiatric Res 2. Gupta S, Kulhara P. What is schizophrenia: a neu- 2010; 4:441-6. rodevelopmental or neurodegenerative disorder or 11. Fan X, Goff DC, Henderson DC. Inflammation and a combination of both? A critical analysis. Indian J schizophrenia. Expert Rev Neurother 2007; 7:789- Psychiatry 2010; 52:21-7. 96. 3. Nawa H, Takahashi M, Paterson PH. Cytokine 12. Baune BT, Camara, M-L, Jawahar C, Anscomb H, and growth factor involvement in schizophrenia – Körner H. Tumour necrosis factor-alpha mediated support for the developmental model. Mol Psychia- mechanisms of cognitive dysfunctions. Transl Neu- try 2000; 5:594-603. rosci 2012; 3:263-77. 4. Naudin J, Capo C, Giusan B, Mège JL, Azorin JM. 13. Xiu MH, Man L-J, Wang D, Du X, Yin G, Zhang Y, A differential role for interleukin-6 and tumor necro- Tan YL, Cehn N, Cehn S, Teixeira AL, Cassidy RM, sis factor-α in schizophrenia? Schizophr Res 1997; Soares JC, Zhang XY. Tumor necrosis factor-alpha 26:227-33. -1031T/C polymorphism is associated with cogni- 5. Erbagci AB, Herken H, Köyloüglu O, Yilmaz N, tive deficits in chronic schizophrenia patients with Tarakçioglu M. Serum IL-1β, sIL-2R, IL-6, IL-8 healthy controls. Am J Med Genet B Neuropsychiatr and TNF-α in schizophrenic patients, relation with Genet 2018:177:379-87. symptomatology and responsiveness to risperidone 14. World Health Organization. The ICD-10 classifica- treatment. Mediators Inflamm 2001; 10:109-15. tion of mental and behavioural disorders. Geneva: 6. Kowalski J, Blada P, Kucia K, Madej A, Herman World Health Organization, 1992. ZS. Neuroleptics normalize increased release of 15. Folstein M, Folsten S, McHugh P. Mini-mental sta- interleukin-1β and tumor necrosis factor-α from te: a practical method for grading the cognitive state monocytes in schizophrenia. Schizophr Res 2001; of patient for the clinician. J Psychiatry Res 1975; 50:169-75. 12:189-98. 7. Theodoropoulou S, Spanakos G, Baxevanis CN, Eco- 16. Kolegium Psikiatri Indonesia. Modul psikiatri ge- nomou M, Gritzapis AD, Papamichail MP, Stefanis riatri (Geriatric psychiatry module) [in Indonesian]. CN. Cytokine serum levels, autologous mixed lymp- Jakarta: Kolegium Psikiatri Indonesia, 2008 hocyte reaction and surface marker analysis in never 17. Lv MH, Tan Y.L, Yan SX, Tian L, Chen DC, Tan medicated and chronically medicated schizophrenic SP, Wang ZR, Yang FD, Yoon JH, Zunta-Soares patients. Schizophr Res 2001; 47:13-25. GB, Soares JC, Zhang XY. Decreased serum TNF- 8. O’Brien SM, Scully P Dinan TG. Increased tumor ne- alpha levels in chronic schizophrenia patients on crosis factor-alpha concentrations in excacerbations long-term antipsychotics: correlation with psycho- of schizophrenia. Psychiatry Res 2008; 160:256-62. pathology and cognition. Psychopharmacol 2015; 9. Czerski PM, Rybakowski F, Kapelski P, Rybakowski 232:165-72. JK, Dmitrzak-Węglarz M, Leszczyńska-Rodziewicz 18. Kunz M, Ceresér KM, Goi PD, Fries GR, Teixera A, Slopien A, Skibinska M, Kaczmarkiewicz-Fass AL, Fernandes BS, Belmonte-de-Abreu PS, Kauer- M, Hauser J. Association of tumor necrosis factor – Sant’Anna M, Kapczinski F, Gama CS. Serum le- 308G/A promoter polymorphism with schizophrenia vels of IL-6, IL-10, TNF-α in patients with bipolar and bipolar affective disorder in a Polish population. disorder and schizophrenia: differences in pro- and Neuropsychobiology 2008; 57:88-94. anti-inflammatory balance. Braz J Psychiatry 2011; 10. García-Miss M.d.R, Pérez-Mutul J, López-Canul 33:268-74. B, Solis-Rodriguez F, Puga-Machado P, Oxté-Ca- brera A, Gurubel-Maldonado J, Arankowsky-San- doval G. Folate, homocysteine, interleukin-6, and

449 Medicinski Glasnik, Volume 17, Number 2, August 2020

19. Tian L, Tan Y, Chen D, Lv M, Tan S, Soares DA, Jansen AFM, Lemmers H, Toenhake-Dijkstra JC, Zhang XY. Reduced serum TNF alpha level in H, van Herwaarden AE, Janssen M, van der Molen chronic schizophrenia patients with or without tar- RG, Joosten I, Sweep FCGJ, Smit JW, Netea-Maier dive dyskinesia. Prog Neuro-Psychopharmacol Biol RT, Koenders MMJF, Xavier RJ, van der Meer Psychiatry 2014; 54:269-4. JWM, Dinarello CA, Pavelka N, Wijmenga C, No- 20. Kubistova A, Horacek J, Novak T. Increased inter- tebaart RA, Joosten LAB, and Netea MG. Host and leukin-6 and tumor necrosis factor alpha in first epi- environmental factors influencing individual human sode schizophrenia patients versus healthy controls. cytokine responses. Cell 2016; 167:1111-24. Psychiatr Danub 2012; 24:153-6. 33. Mansur RB, Zugman A, Asevedo EM, da Cunha GR, 21. Hope S, Melle I, Aukrust P, Steen NE, Birkenaes Bressan RA, Brietzke E. Cytokines in schizophre- AB, Lorentzen S, Agartz I, Ueland T, Andreassen nia: possible role of anti-inflammatory medications OA. Similar immune profile in bipolar disorder and in clinical and preclinical stages. Psychiatry Clin schizophrenia: selective increase in soluble tumor Neurosci 2012; 66:247-60. necrosis factor receptor I and von Willebrand factor. 34. Pollmächer T, Haack M, Schuld A, Reichenberg A, Bipolar Disord 2009; 11:726-34. Yirmiya R. Low levels of circulating inflammatory 22. Ochoa S, Usall J, Cobo J, Labad X, Kulkarni J. cytokines-do they affect human brain functions? Gender differences in schizophrenia and first-epi- Brain Behav Immun 2002; 16:525-32. sode psychosis: a comprehensive literature review. 35. Hoseth EZ, Ueland T, Dieset I, Birnbaum R, Shin Schizophr Res Treatment 2012; 2012:916198. JH, Kleinman JE, Hyde TM, Mørch RH, Hope 23. Li R, Ma X, Wang G, Yang J, Wang C. Why sex S, Lekva T, Abraityte AJ, Michelsen AE, Melle differences in schizophrenia? J Transl Neurosci I, Westlye LT, Ueland T, Djurovic S, Aukrust P, We- 2016; 1:37-42. inberger DR, Andreassen OA. A study of TNF 24. Chiang SSW, Riedel M, Gruber R, Müller N, pathway activation in schizophrenia and bipolar Schwarz M. In vivo type II T-helper cells shift in disorder in plasma and brain tissue. Schizophr Bull schizophrenia compared to sex- and age-matched 2017; 43:881-90. healthy controls. Eur J Psychiatry 2011; 25:192-204. 36. Strieter RM, Kunkel SL, Bone RC. Role of tumor 25. Zhu F, Zhang L, Liu F, Wu R, Guo W, Ou J, Zhang necrosis factor-alpha in disease state and inflamma- X, and Zhao J. Altered serum tumor necrosis factor tion. Crit Care Med 1993; 21:S447-63. and interleukin-1β in first-episode drug-naïve and 37. Kirkpatrick B, Miller BJ. Inflammation and chronic schizophrenia. Front Neurosci 2018; 12:296. schizophrenia. Schizophr Bull 2013; 39:1174-9. 26. Boerrigter D, Weickert TW, Lenroot R, O’Donnell 38. Hennessy E, Gormley S, Lopez-Rodriguez AB, M, Galletly C, Liu D, Burgess M, Cadiz R, Jacomb Murray C, Murray C, Cunningham C. Systemic I, Catss VS, Fillman SG, and Weickert CS. Using TNF-α produce acute cognitive dysfunction and blood cytokine measures to define high inflamma- exaggerated sickness behaviour when superimposed tory biotype of schizophrenia and schizoaffective upon progressive neurodegeneration. Brain Behav disorder. J Neuroinflamm 2017; 14:1-15. Immun 2017; 59:233-44. 27. Lee EE, Hong S, Martin AS, Eyler LT, Jeste DV. 39. Şahin TD, Karson A, Balci F, Yazir Y, Bayramgürler Inflammation in schizophrenia: cytokine levels and D, Utkan T. TNF-alpha inhibition prevents cognitive their relationships to demographic and clinical varia- decline and maintains hippocampal BDNF levels in bles. Am J Geriatr Psychiatry 2017; 25:50-61. the unpredictable chronic mild stress rat model of 28. Ergün S, Yanartas Ö, Kandemir G, Yaman A, Yildiz depression. Behav Brain Res 2015; 292:233-40. M, Haklar G, and Sayar K. The relationship between 40. Belarbi K, Jopson T, Tweedie D, Arellano C, Luo W, psychopathology and cognitive functions with cyto- Greig NH, and Rosi S. TNF-α protein synthesis inhi- kines in clinically stable patients with schizophrenia. bitor restores neuronal function and reverses cogni- Psychiatry Clin Psychopharmacol 2018; 28:66-72. tive deficits induced by chronic neuroinflammation. 29. Simamora RH, Loebis B, Husada MS. Compari- J Neuroinflamm 2012; 9:23. son of serum levels of tumor necrosis factor alpha 41. Thomasy H. Tumor necrosis factor α as a potential (TNF-α) in Batak male schizophrenic patients ver- mediator of the effects of phosphodiesterase 4B in- sus healthy controls. Int J Life Sci Scienti Res 2018; hibition on cognition after traumatic brain injury. J 4:1652-56. Neurosci 2016; 36:11587-9. 30. Trovão N, Prata J, VonDoellinger O, Santos S, Bar- 42. Miller BJ, Culpepper N, Rapaport MH, Buckley bosa M, Coelho R. Peripheral biomarkers for first- P. Prenatal inflammation and neurodevelopment in episode psychosis-opportunities from the neuroin- schizophrenia: a review of human studies. Prog Ne- flammatory hypothesis of schizophrenia. Psychiatry uropsychopharmacol Biol Psychiatry 2013; 42:92- Investig 2018; 16:177-84. 100. 31. Xu, H. Neuroinflammation in schizophrenia focused 43. Misiak B, Stańczykiewicz B, Kotowicz K, Ryba- on the pharmacological and therapeutic evidence. kowski JK, Samochowiec J, Frydecka D. Cytokines Pharmacol 2015; 6:438-53. and C-reactive protein alterations with respect to 32. Ter Horst R, Jaeger M, Smeekens SP, Oosting cognitive impairment in schizophrenia and bipolar M, Swertz MA, Li Y, Kumar V, Diavatopoulos disorder: A systematic review. Schizophr Res 2018; 192:16-29.

450 ORIGINAL ARTICLE

Poguntano effect on cell viability in cholesteatoma keratinocyte of chronic suppurative otitis media

Harry A. Asroel1, Aznan Lelo2, Delfitri Munir1, Jenny Bashiruddin3, Restu Syamsul Hadi4

1Ear, Nose and Throat - Head and Neck Surgery Department, School of Medicine, Universitas Sumatera Utara, Medan, 2Pharmacology Department, School of Medicine, Universitas Sumatera Utara, Medan, 3 Ear, Nose and Throat - Head and Neck Surgery Department, School of Medicine, Universitas Indonesia (UI), Jakarta, 4Stem Cell Research Centre, Universitas Yarsi, Jakarta; Indonesia

ABSTRACT

Aim Chronic Suppurative Otitis Media (CSOM) with or without cholesteatoma is a global major problem and it is becoming a burden especially in developing countries. Studies have found that ethanol extracts of Puguntano leaves obtained by the percolation and socletation methods have phytochemical contents that provide anti-inflammatory effect. The aim of this study was to investigate the effects of ethanol extract and Puguntano on the viability of cholesteatoma keratinocytes.

Methods This in vitro experimental study included 8 groups of the cholesteatoma keratinocyte culture: not given puguntano leaf Corresponding author: ethanol extract, three positive control groups given different con- Asroel, Harry Agustaf centrations of dexamethasone, and four groups that were given ENT-Head and Neck Surgery Department, different concentrations of ethanol extract of leaf Puguntano. School of Medicine, Results A statistically significant decrease in procalcitonin level Universitas Sumatera Utara (p<0.001) and an amount of bacterial colonies (p<0.001) in four 20155 Medan, Indonesia groups were found. Phone: +62 812 6517 710 Conclusion This study showed that Puguntano leaf extract has the Fax: +62 61 8216264 same effect as dexamethasone in terms of suppressing cell viabi- E-mail: [email protected] lity and has lower side effects compared to dexamethasone as other ORCID ID: https://orcid.org/0000-0003- herbal medicines. Thus, the ethanol extract of Puguntano leaves 3096-2166 can be used as an alternative for the prevention of cholesteatoma. Key words: inflammation, otitis, treatment

Original submission: 04 February 2020; Revised submission: 29 April 2020; Accepted: 19 June 2020 doi: 10.17392/1155-20

Med Glas (Zenica) 2020; 17(2): 451-456

451 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION to be widely cultivated by the local community as a medicinal plant (7). In some areas, this plant Chronic suppurative otitis media (CSOM) with or is known as “kukurang, tamah raheut” and “em- without cholesteatoma is a global major problem peduh tanah” (6). Juwita et al. (8) and Harfina and becomes a burden especially in developing et al. (9) reported an anti-inflammatory potential countries (1)Cipro HC. The World Health Orga- of Puguntano leaf in the form of ethanol extract. nization (WHO) reported that there are 65-330 million people with CSOM globally and 94% of Studies have found that ethanol extracts of them are in developing countries (1). About 60% Puguntano leaves obtained by the percolation (39 – 200 million people) died because of CSOM and socletation methods have phytochemical (1,2)Cipro HC. Commonly, CSOM prevalence in contents, such as alkaloids, flavonoids, glyco- developing countries, such as India, is higher with sides, saponins, tannins and steroids/terpenoids total 5.2% (2). In Indonesia, CSOM prevalence is (6). One of the classes of chemical compounds reported as many as 3.6% (2). There are two types responsible for anti-inflammatory effects is ste- of CSOM: tubotympanic (without cholesteatoma) roids/terpenoids (8). and atticoantral (with cholesteatoma) (3). A study in Jakarta which conducted research on Atticoantral type of CSOM can cause a serious the effect of dexamethasone on proinflammatory complication such as intratemporal complicati- cytokines, namely TNF-α and IL-1α, got an ave- on (facial nerve paralysis) and intracranial com- rage cell viability of 92.95% when planted, and plication (meningitis, cerebral abscess) which 96.60% after 48h cultivation (9)tissue culture increases morbidity and mortality rates signifi- methods can be applied also to cholesteatoma of cantly (2,4). The CSOM resulted from acute oti- chronic suppurative otitis media patients that can tis media, which is characterized with continuous be used to study the nature of cholesteatomas to otorrhea from the middle ear through perforated the cellular level. Purpose: The purpose of this stu- tympanic membrane (1). This process is the main dy was to get tissue culture methods for cholestea- cause of hearing disorder (1)Cipro HC. tomas, so that the steps, techniques and specific tools for culturing cholesteatoma can be determined, Cholesteatoma characterized by hyperprolifera- also the number of cells that were planted, good tion is associated with chronic inflammation and environment in order to obtain cell viability and bone destruction (2). Ossicle resorption occurs in good culture results. Methods: This study consi- local area surrounding cholesteatoma perimatrix sts of three phases: 1; with 0.05 significance level, or granulation tissue (5). Cholesteatoma perima- from the matrix it was seen that cell viability was trix contains lymphocytes, monocytes, fibroblasts significantly different between the control group and endothelial cells, which are sources of proin- and 40 mg and 100 mg dose groups (9). flammation cytokines, such as TNF-α and IL-6 (5). TNF-α and IL-6 expression in cholesteatoma The aim of this study was to investigate the tissue is higher compared to granulation tissue (5). effects of ethanol extract and Puguntano on the TNF-α plays a direct role in bone destruction thro- viability of cholesteatoma keratinocytes. ugh osteoclast differentiation and through opened MATERIALS AND METHODS bone matrix, indirectly (4). This process occurs simultaneously with IL-1 expression (4). Materials and study design Puguntano (Curanga fel-terrae [Lour] Merr.) is a herb from Scrophulariaceae family that grows in This study was conducted at the Universitas Su- Asian regions, such as Indonesia, Malaysia, the matera Utara (USU), Medan, from January 2019 Philippines, China and India (6). In Indonesia, in to September 2019. the Tiga Lingga village community of Dairi Dis- This in vitro experimental study included 8 grou- trict in North Sumatra province Puguntano leaf ps of the cholesteatoma keratinocyte culture: one is empirically used as a medicinal herb (7). This group which was not given puguntano leaf etha- plant is believed to be efficacious as a painkiller, nol extract, three positive control groups given it increases host immunity, and is even used in different concentrations of dexamethasone, and an anti-aging process (7). This plant has begun four groups that were given different concentrations of ethanol extract of leaf Puguntano.

452 Asroel et al. Poguntano effect on CSOM

Cholesteatoma tissue was taken during the sur- method relies on introducing a live-cell membra- gery of the patients with atticoantral type of ne-permeant DNA fluorophore, such as Hoechst CSOM in the Central Surgery Installation at the 33342, into the culture medium of cells at the end Tangerang General Hospital. Cholesteatoma ke- of any live-cell imaging experiment and measuring ratinocyte cultivation was carried out at the In- each cell's integrated nuclear fluorescence to quan- tegrated Laboratory Research Institute of YARSI tify DNA content. Importantly, our method over- Universitas, Jakarta. Cell viability tests are carri- comes the toxicity and induction of DNA damage ed out in the same laboratory. typically caused by live-cell dyes through strategic The approval for the investigation was obtained timing of adding the dye to the cultures; allowing from the Ethics Committee of the Faculty of Me- unperturbed cells to be imaged for any interval of dicine, Universitas Sumatera Utara. time before quantifying their DNA content. We assess the performance of our method empirically Methods and discuss adaptations that can be implemented using this technique. Results: Presented in con- Samples. Patients with atticoantral type of junction with cells expressing a histone 2B-GFP CSOM, who had agreed to be included into the fusion protein (H2B-GFP. Puguntano extract was study and had undergone surgery in Tangerang diluted within 5 mL ethanol. We then examined the General Hospital, were included in the study. Pa- cells viability under fluorescent microscope. tients were not allowed to take steroids at least 1 week before the surgery. Cholesteatoma tissues Cholesteatoma keratinocyte culture that was su- were immediately stored in a 50 mL conical itable for the analysis, was divided into 8 groups bottle containing sterile phosphate buffer saline using microplate based on the treatment obta- plus antibiotics-antimycotics in a cold state. Then ined (11,12): Group 1- control group that was the samples were cultured using a keratinocyte not given puguntano ethanol extract, Group 2 growing medium, Epilife plus Human Kerati- - positive control group by giving dexamethaso- nocyte Growth Supplement (HKGS) (10–12) ne solution at a dose of 1 μM, Group 3 - positive control group by giving dexamethasone at a Ethanol extract preparation. The preparation dose of 10 μM, Group 4: positive control group of ethanol extract of leaf Puguntano was done at with dexamethasone at a dose of 100 μM, Gro- Faculty of Pharmacy of USU. Puguntano extracts up 5 - cholesteatoma keratinocytes given ethanol were made by the maceration method (13). Pugun- extract of puguntano at a dose of 1%, Group 6 tano leaves that had been washed and dried, were - cholesteatoma keratinocytes were given etha- then ground into powder. Puguntano extracts nol extract of puguntano at a dose of 2%, Group were made by the maceration method. Ten parts 7 - cholesteatoma keratinocytes given ethanol of Simplisia powder were incorporated into a clo- extract of puguntano at a dose of 4%, and Gro- sed container and left for 5 days protected from up 8 - cholesteatoma keratinocytes treated with lights while stirred occasionally. The mixture was ethanol extract of puguntano at a dose of 8%. then squeezed until the pulp was left. The pulp was washed so that 100 parts of a solution were Cell viability test. Cell viability is the number obtained. The solution was transferred into a clean of living cells and it is usually expressed as a container and then left for two days to be filtered. percentage of the number in the control group. A cell viability assay is performed based on the Keratinocytes cells were cultured in Epilife plus ratio of live and dead cells. This assay is based HKGS medium and then stained using Hoecsht on an analysis of cell viability in cell culture for protocol (14,15)the capacity to simultaneously me- evaluating in vitro drug effects in cell-mediated asure DNA content in cells being tracked over time cytotoxicity assays for monitoring cell prolifera- remains challenged by dye-associated toxicities. tion (16). In this study, we assayed cell viability The ability to measure DNA content in single cells with EVOS fluorescence microscopy. by means of LCFM would allow cellular stage and ploidy to be coupled with a variety of imaging di- Statistical analysis rected analyses. Here we describe a widely applicable nontoxic approach for measuring DNA con- The data obtained were processed and analysed tent in live cells by fluorescence microscopy. This with univariate analysis (normal or abnormal dis-

453 Medicinski Glasnik, Volume 17, Number 2, August 2020

tribution) using the Kolmogorov Smirnov test, by variate analysis (comparing before and after the treatment) using t-test for normal data and Wilcoxon test for non-data normal, as well as multivariate analysis (comparing many groups) using the ANOVA test for normal distribution and the A) B) Kruskal-Wallis test for abnormal distribution. Bon- ferroni test was performed to see whether there are differences in each of the two experimental groups. A p <0.05 was used as statistically significant.

RESULTS C) D) Figure 2. Puguntano effect on cells viability of cholesteatoma Cell viability in the control group showed the keratinocytes with Hoecst staining. A) there was a decreasing highest mean far above the group given dexa- number of cells' viability in PG1 group (when added ethanol extract of Puguntano at a dose of 1%) compared to group K methasone and puguntano with a mean of 45.6. (it was not given Puguntano ethanol extract – Figure 1); B) The mean of viability between groups given there was a decreasing number of cells' viability in PG2 group dexamethasone and puguntano did not show si- B (when added ethanol extract of puguntano at a dose of 2%) gnificant differences in mean viability (p>0.05). compared to PG1 group ; C) there was a decreasing number of cells' viability in PG4 group (when added ethanol extract Compared to the control group, we found that the of Puguntano at a dose of 4%) compared to PG2 group; D) more dexamethasone was given, the lower num- there was a decreasing number of cells' viability in PG8 group ber of cells’ viability was noticed (Figure 1). compared to PG2 group

A) B)

C) D)

Figure 1. Dexamethasone effect on cells viability of choles- Figure 3. Number of cell’s viability in the control group and teatoma keratinocytes with Hoecst staining. A) Control group samples, which were given dexamethasone and Puguntano (K) (it was not given Puguntano ethanol extract) where cells’ extract in various doses. K, total percentage of cells’ viability; viability was 100%; B) there was a decreasing number of Dex1, cell’s viability after given 1 μM dexamethasone; Dex10, cells' viability in group DEX1 (positive control group adding cell’s viability after given 10 μM dexamethasone; Dex100, dexamethasone at a dose of 1 μM) compared to group DEX10; cell’s viability after given 1000 μM dexamethasone; PG1, C) there was a decreasing number of cells' viability in group cells’ viability after given Puguntano extract 1%; PG2, cells’ DEX10 (positive control group adding dexamethasone at a viability after given Puguntano extract 2%; PG4, cells’ viability dose of 10 μM) compared to group DEX1; C) there is a de- after given Puguntano extract 4%; PG8, cells’ viability after creasing number of cells' viability in group DEX100 (positive given Puguntano extract 8% control group adding dexamethasone at a dose of 100 μM) compared to group DEX10 DISCUSSION The same effect was also seen in samples that Studies that assess cell viability, especially in cho- were given Puguntano ethanol extract (Figure 2). lesteatoma are still very rare. This research may be There was an increasing number of cell’s viabi- a novel study that can be used as a pilot study to lity on samples given 2% of Puguntano extract. see the effect of Puguntano ethanol extract on the However, when the samples were given 4% and viability of cholesteatoma culture cells. We found 8% of Puguntano extract, the number of cell’s that the culture group with the highest cell viabi- viability was decreasing (Figure 3). lity rate was the control group (group 1 that was

454 Asroel et al. Poguntano effect on CSOM

not given Puguntano ethanol extract) with a mean 1 (COX-1) and cyclooxygenase 2 (COX-2) in- of 45.6 %. This value is significantly greater than hibition. Active substance of flavonoid is also either the culture group that received dexametha- considered as an effective scavenger in reactive sone or those who received ethanol extract from oxygen species due to its phenolic hydroxyl gro- Puguntano leaves. Kara et al at Turkish Kocaeli up, and thus, is considered to be a strong anti-oxi- University (2019) in their 3D cholesteatoma cell dant (13,20). Furthermore, effect of the drug on culture model study using sodium diclofenac 1% COX-2 decrease in cell viability and the presence and 2%, 5-fluorouracil 1% and negative controls of necrotic processes shown by Fourier transform received lower cell viability and higher apoptosis infrared analysis, suggesting a direct correlation in the sodium diclofenac group compared to cho- of prostanoids in cellular homeostasis and survi- lesteatoma cell cultures group without additional val (22). Currently, there is evidence that prosta- drugs (18)and there is no medical treatment for this glandins produced by COX-2 intervene in tumour disease. Exploring new medical treatment options cell proliferation as NSAIDs and selective COX-2 May help to create treatment alternatives instead of inhibitors inhibit proliferation of different can- surgery. Materials and Methods: Required chole- cer cell types expressing COX-2(23). NS-398, a steatoma tissues for cell culture were excised from COX-2 specific inhibitor, was described to reduce 4 different participants who underwent surgery in cell proliferation of MC-26 cell line, a highly inva- our clinic and agreed to give tissue for the study. sive mouse CRC cell model expressing constituti- Cholesteatoma-derived keratinocytes and fibrobla- vely COX-2 (24),(25). sts were cocultured in temperature-sensitive culture Appropriate host immune response could offer dishes to make a three-dimensional (3D. In our stu- protection against infectious threats; however, dy, on the other hand, cells’ viability, when samples excessive inflammatory immune response can were given dexamethasone, was not significantly lead to the uncontrolled growth and proliferation different when compared to the culture groups that of cholesteatoma (26). received Puguntano ethanol extract. In conclusion, this study showed that Puguntano Puguntano ethanol extract contain chemical com- leaf extract has the same effect as dexamethasone pounds of glycoside, flavonoid, saponin, tannin, in terms of suppressing cell viability compared to and steroid/triterpenoid (19). Yassine et al. (20) in dexamethasone as other herbal medicines. Thus, their research have demonstrated flavonoid role in the ethanol extract of Puguntano leaves can be anti-inflammation process. The early step where used as an alternative for the prevention of cho- histamine and serotonin release is followed with lesteatoma. oedema caused by bradykinin and prostaglandin release (this phase is known to be sensitive with ACKNOWLEDGEMENT anti-inflammatory agent, both steroid and non- We thank Ashadi Budi, MD, Otologic Surgeon steroid), flavonoid is shown to inhibit important (Tangerang General Hospital) for dedication enzyme that has a role in biosynthesis of tissue and support to the cholesteatoma matrix for this activator, especially prostaglandin and nitric oxide research. (20). Flavonoid shows several biological effects, such as anti-inflammation, anti-hepatotoxic, anti- FUNDING allergy, anti-virus, and anti-cancer (21). Flavonoid No specific funding was received for this study. is also effective in inhibiting arachidonic acid metabolism that mediates prostaglandin biosynthe- TRANSPARENCY DECLARATION sis through gene expression of cyclooxygenase Conflicts of interest: None to declare. REFFERENCES 1. World Health Organization. Chronic suppurative 2. Yarisman L, Asroel HA, Aboet A, Zaluchu F. Hubun- otitis media - Burden of Illness and Management gan ekspresi RANKL dengan derajat destruksi tulang Options. Geneve: World Health Organization, 2004 akibat kolesteatoma pada otitis media supuratif kro- https://apps.who.int/iris/handle/10665/42941 (14 Ja- nik (The relationship of RANKL expression with the nuary 2020) degree of bone destruction due to cholesteatoma in chronic suppurative otitis media) [In indonesian] Oto Rhino Laryngol Indones 2017; 47:1–9.

455 Medicinski Glasnik, Volume 17, Number 2, August 2020

3. Park M, Lee B, Lee J. Expression of tumor necrosis 14. Gomes CJ, Harman MW, Centuori SM, Wolgemuth factor-like weak inducer of apoptosis in human midd- CW, Martinez JD. Measuring DNA content in live le ear cholesteatoma. J Otorhinolaryngol Relat Spec cells by fluorescence microscopy. Cell Div 2018; 2013;75:221–7. 13:1–10. 4. Vitale RF, Ribeiro FDAQ. The role of tumor necrosis 15. Chan LLY, McCulley KJ, Kessel SL. Assessment of factor-alpha (TNF-alpha) in bone resorption present cell viability with single-, dual-, and multi-staining in middle ear cholesteatoma. Rev Bras Otorrinolarin- methods using image cytometry. Methods Mol Biol gol (English ed) 2007; 73:123–7. 2017; 1601:27–41. 5. Kuczkowski J, Sakowicz-Burkiewicz M, Iycka-Świ- 16. Mukherjee PK. Bioassay-Guided Isolation and Eva- eszewska E, Mikaszewski B, Pawełczyk T. Expressi- luation of Herbal Drugs. In: Quality Control and on of tumor necrosis factor-α, interleukin-1α, interle- Evaluation of Herbal Drugs. Neederlands: Elsevier ukin-6 and interleukin-10 in chronic otitis media with Mosby, 2019: 515–37. bone osteolysis. ORL J Otorhinolaryngol Relat Spec 17. Kara A, Duman BO, Yazır Y, SinanYilmaz M, 2011; 73:93–9. Halbutogulları ZSU, Demir D, Kara RO, Bayraktar 6. Hijriani A, Chandra A, Hardiansyah N, Riki TA. H, Guven M. Evaluation of the effect of diclofenac Which Potential as Antiasma) Profil Fitokimia Ek- sodium and 5-fluourasil in a 3D cholesteatoma cell strak Etanol Daun Puguntano [Curangafel-terrae culture model. Otol Neurotol 2019; 40:1018-25. (Merr.) Lour.] (Phytochemical Profile of Ethanol 18. Sibangriang. HE. Pengaruh pemberian ekstrak daun Extract of Puguntano Leaves [Curangafel-terrae Poguntano (Picria fel-terrae (Lour.) Merr.) terhadap (Merr.) Lour.] [In indonesian] Semin Nas Sains Te- Kadar Superoxide Dismutase (SOD) pada penderi- knol 2013;422–6. ta diabetes mellitus tipe 2 (Effect of administration 7. Depkes RI. Inventaris Tanaman Obat Indonesia In- of Poguntano (Picria fel-terrae (Lour.) Merr.) Leaf ventaris Tanaman Obat Indonesia. 1st ed. (Inventory extract to levels of Superoxide Dismutase (SOD) in of Indonesian Medicinal Plants Inventaris Tanaman patients with type 2 diabetes mellitus) [In indonesian] Obat Indonesia. 1st ed.) Jakarta: Departemen Keseha- Medan: Universitas Sumatera Utara, 2017. tan Indonesia, 2000. 19. Yassine EZ, Dalila B, Latifa EM, Smahan B, Lebtar 8. Barnes PJ. How corticosteroids control S, Sanae A.. Phytochemical screening, anti-inflamma- inflammation:Quintiles Prize Lecture 2005. Br J tory activity and acute toxicity of hydro-ethanolic, Pharmacol 2006;148:245–54. flavonoid, tannin and mucilage extracts of Lavan- 9. Restuti RD. Metode biakan jaringan kolesteatoma dula stoechas L. from Morocco. Int J Pharmacogn pasien otitis media supuratif kronik tipe bahaya (Cho- Phytochem Res 2016; 8:31–7. lesteatoma tissue culture method for chronic suppura- 20. Satria D, Silalahi J, Haro G, Ilyas S, Hsb PAZ. An- tive otitis media patients with hazard type) [In indo- tioxidant and antiproliferative activities of an ethyla- nesian] Oto Rhino Laryngol Indones 2013; 43:13–8. cetate fraction of picria fel-terrae Lour. herbs. Asian 10. Raynov AM, Choung Y-H, Park HY, Choi SJ, Park Pacific J Cancer Prev 2017; 18:399–403. K. Establishment and characterization of an in vitro 21. Sakane KK, Monteiro CJ, Silva W, Silva AR, Santos model for cholesteatoma. Clin Exp Otorhinolaryngol PM, Lima KF, Moraes KCM. Cellular andmolecular 2008; 1:86–91. studies of the effects of a selective COX-2 inhibitor 11. Iwamoto Y, Nishikawa K, Imai R, Furuya M, Uena- celecoxib in the cardiac cell line H9c2 and their corre- ka M, Ohta Y, Morihana T, Itoi-Ochi S, Penninger lation with death mechanisms. Brazilian J Med Biol JM, Katayama I, Inohara H, Ishi M. Intercellular Res 2014; 47:50–9. communication between keratinocytes and fibroblasts 22. Cao Y, Prescott SM. Many actions of cyclooxygena- induces local osteoclast differentiation: a mechanism se-2 in cellular dynamics and in cancer. J Cell Physiol underlying cholesteatoma-induced bone destruction. 2002; 190:279-86. Mol Cell Biol 2016; 36:1610–20. 23. Yao M, Lam EC, Kelly CR, Zhou W, Wolfe MM. 12. Raffa S, Leone L, Scrofani C, Monini S, Torrisi MR, Cyclooxygenase-2 selective inhibition with NS-398 Barbara M. Cholesteatoma-associated fibroblasts mo- suppresses proliferation and invasiveness and de- dulate epithelial growth and differentiation through lays liver metastasis in colorectal cancer. Br J Cancer KGF/FGF7 secretion. Histochem Cell Biol 2012; 2004; 90:712–9. 138:251–69. 24. Sobolewski C, Cerella C, Dicato M, Ghibelli L, Die- 13. Rambe AYM, Munir D, Sembiring RJ, Ilyas S. derich M. The role of cyclooxygenase-2 in cell proli- Effect of poguntano leaves extract (Picria fel-terrae feration and cell death in human malignancies. Int J merr.) to procalcitonin level in acute bacterial rhino- Cell Biol 2010; 2010:215158. sinusitis model of wistar mice. Med Glas (Zenica) 25. Kuo CL, Shiao AS, Yung M, Sakagami M, Sudhoff H, 2020;17:110–6. Wang CH, C-H, Lien C-F. Updates and knowledge gaps in cholesteatoma research. Biomed Res Int 2015.

456 ORIGINAL ARTICLE

Pregnancy outcome in women who survived genital or extragenital cancer

Anis Cerovac¹,², Dženita Ljuca³, Gordana Bogdanović³, Gordana Grgić³, Haris Zukić³

¹Department of Gynaecology and Obstetrics, General Hospital Tešanj, Tešanj, ²Department of Anatomy, School of Medicine, University of Tuzla, Tuzla, ³Clinic for Gynaecology and Obstetrics, University Clinical Centre Tuzla, Tuzla; Bosnia and Herzegovina

ABSTRACT

Aim To investigate clinical and obstetrical characteristics, an outcome and a prognosis for pregnant women with diagnosed and treated genital or extragenital cancer and their newborns.

Methods This retrospective cohort study included pregnant and childbearing women with a history of cancer diagnosed before pregnancy during the period between 1 January 2014 and 31 December 2018. Data related to the course of pregnancy and childbirth were collected from medical records (mothers‘ disease history and partogram). The analysis covered clinical and histopathological characteristics of cancers, type of the treatment (surgery, chemotherapy, radiotherapy), demographic data, obstetric characteristics, comorbidities of women, and outcome of the newborns. Corresponding author: Results The study recorded 18 414 deliveries, of which 30 (0.16%) Anis Cerovac were pregnancies in women who had been diagnosed and treated General Hospital Tešanj earlier for genital or extragenital cancer. The average age of the Braće Pobrić 17, 74260 Tešanj, women at the time of delivery was 29.43±5.97 years. There were Bosnia and Herzegovina six (20%) women with genital and 24 (80%) with extragenital can- Phone: +387 32 650 662; cer. The most frequent extra genital cancer was Hodgkin lympho- Fax: +387 32 650 605; ma, in eight (26.6%) cases; ovarian cancer was the most frequent genital cancer, in four (13.3%) cases. The average time span from E-mail: [email protected] the cancer diagnosis and start of the treatment to the delivery was ORCID ID: https://orcid.org/0000-0002- 59.2±44.4 months (5 years) (range 12 months - 15 years). Two 7209-382X (6.6%) women died.

Conclusion Our data demonstrate a favourable obstetric and neonatal outcome for women who have survived cancer.

Original submission: Key words: fertility, malignancy, obstetric outcome, pregnancy 06 December 2019; Revised submission: 21 January 2020; Accepted: 17 February 2019 doi: 10.17392/1109-20

Med Glas (Zenica) 2020; 17(2): 457-464

457 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION health are an important issue (8). Melanoma is the sixth most commonly diagnosed cancer in women; A delay of childbearing to the later reproductive age the diagnosis of melanoma is established prior to increases the number of women who had a cancer their first pregnancy in many women. The available cured or survived a cancer (1-3). Recent advances evidence does not show an adverse effect of pre- in oncologic diagnostic methods and treatments, gnancy on disease-free and overall survival, pro- and better access to fertility sparing and infertility gression and/or mortality in melanoma patients (3). treatment imply that both pregnancy and childbirth are now real options for women with history The data about complications related to pregnancy of cancer (1,2). The most common types of cancer and neonatal outcome of women with history of in women of reproductive age are breast, genital, cancer are scarce. Some studies have reported hematologic, thyroid, and skin (melanoma) (3,4). increased risks of diverse obstetrical, perinatal, Fertility and pregnancy rates have decreased in and neonatal complications concerning these pre- women with history of cancer (cancer survivors, e. gnancies (1,2,9,10). However, larger studies are g. five years after the cancer treatment; completed required with longer follow-up periods in order initial cancer treatment and no apparent evidence of to further verify reproductive outcome following active disease), especially in those with leukaemia fertility-sparing techniques for women diagnosed and breast cancer (3% and 8%, respectively) (3,5). with genital and extragenital cancers. However, these rates do not significantly differ in Despite of the increasing importance of knowled- malignant melanoma, thyroid cancer, and Hodgkin ge about the course and outcome of pregnancy lymphoma survivors (29 %, 33 % and 32%, res- in cancer survivors in Bosnia and Herzegovina, pectively) (6). Genital cancers (including breast, there have been no such studies, except for one ovaries, uterine tubes and uterus) are the fourth case report (11). most commonly diagnosed cancers in women of The aim of this study was to investigate clinical childbearing age (15–45 years), accounting for and obstetric characteristics, perinatal and neo- 16% of all cancers (6). Cancer therapy comprises natal outcome and prognosis in pregnant women chemotherapy with agents of high gonadotoxicity; with the history of cancer. radiotherapy treatment of the ovaries, vagina, or uterus can have detrimental effects on the patient’s PATIENTS AND METHODS ability to achieve reproductive function and carry pregnancy to term (2,6). Fertility-sparing surgeries Patients and study design are aimed at preserving woman’s fertility, provi- This retrospective cohort study included pregnant ding options for girls and women of reproductive and childbearing women with the history of can- age with genital cancer. The women with low stage/ cer diagnosed before pregnancy during the period grade/borderline genital carcinoma are eligible for between 1 January 2014 and 31 December 2018 at fertility-sparing surgical techniques (6); the fertility the Clinic for Gynaecology and Obstetrics, Univer- sparing treatment in these women has not negati- sity Clinical Centre Tuzla. Inclusion criteria were: vely affected overall women survival rate or qua- pregnant women who gave birth with the establis- lity of life, with good reproductive and pregnancy hed diagnosis (histopathologically confirmed) and outcomes (6). Pregnancies of women after breast treatment (surgery, chemotherapy or radiotherapy) cancer do not carry a worsened prognosis, but some for cancer before pregnancy. Excluding criteria studies suggest an increased risk of miscarriage, were: women without histopathologically confir- preterm birth, low birth weight, increased Caesa- med diagnosis of cancer, pregnant women with hi- rean delivery rate and antepartum and postpartum stopathologically confirmed diagnosis but whose haemorrhage (3). Haematological cancers acco- pregnancy ended with miscarriage (spontaneous unt for 17% of all cancers diagnosed in girls and or induced abortion), and women with histopatho- women of reproductive age (7). Thyroid cancer is logically confirmed benign borderline tumours be- the most common endocrine cancer (8). Because fore and during pregnancy, and carcinoma in situ of half of the patients are in the reproductive age, of uterine cervix (four patients). effects of different treatments, such as radioactive The survey was approved by the Ethics Com- iodine (RAI), on future gonadal and reproductive mittee of the University Clinical Centre of Tuzla.

458 Cerovac et al. Pregnancy and cancer

Methods RESULTS Data on the course of pregnancy and childbirth During the observed period 18,414 deliveries were collected on the basis of available medi- were recorded, of which 30 (0.16%) were cases cal records (mother’s disease history and parto- of pregnancy in women who had been earlier diagrams). Oncologic data analysis covered clinical gnosed and treated for genital or extragenital can- and histopathological characteristics of cancers, cer. There were six (20%) women with genital date of diagnosis, type of treatment (surgery, che- and 24 (80%) with extragenital cancer. motherapy, radiotherapy), and maternal survival. The average women’s age at the time of the dia- Obstetrical data included the age of women at gnosis and treatment was 23.6±67.85 (range 3-36 delivery, time from diagnosis and treatment to years); the youngest woman had brain cancer delivery, parity, week of gestation at delivery, (17.5 years old), and the oldest one had thyroid mode of delivery, obstetric comorbidities and cancer (28.1 years old). The average women’s complications. Neonatal data included gender, age at the time of delivery was 29.43±5.97 (ran- birth weight and length, intrauterine growth re- ge 16-38 years). striction (IUGR), Apgar score (AS) at the first Women with thyroid (32.5 years), haematologi- and fifth minute, congenital malformation, admi- cal (31 years) and other (malignant melanoma, ssion to the neonatal intensive care unit (NICU), epipharyngeal cancer, hepatic angiosarcoma, stillbirths, and early neonatal death. osteosarcoma of the humerus, abdominal gangli- Statistical analysis oneuroblastoma) types of cancers (28.66 years) were older at the time of delivery than women Descriptive statistics, mean value, standard de- with genital (27 years) and brain (23.5 years) viation (SD), and percentage were used in stati- cancers (Table 1). stical data processing.

Table 1. Obstetric characteristics of 30 pregnant women according to type of cancer Extra genital Characteristics Genital Others* Total N(%) Haematological Thyroid Brain Number of women (No, %) 6 (20) 10 (33.3) 6 (20) 2 (6.66) 6 (20) 30 Age at time of diagnosis and treatment (mean SD) (years) 21±8.31 25±6.1 28.1±5.03 17.5±6.3 21.83±11.4 23.66±7.85 Age at time of delivery (mean SD) (years) 27±6.72 30.7±4.2 32.5±4.5 23.5±3.5 28.6±8.3 29.43±5.97 Time span from treatment to delivery (mean SD) (years) 4.58±3.92 5.8±3.6 4.3±2.06 6.5±2.12 8.5±5.8 5 Parity (No, %) Primiparous 5 (83.3) 6 (60) 3 (50) 2 (100) 4 (66.6) 19 (63.33) Secundiparous 1 (16.6) 3 (30) 3 (50) - 2 (33.3) 10 (33.33) Third and multiparous - 1 (10) - - - 1 (3.33) Weeks of gestation (mean SD) 38.13±2.02 Delivery less than 37 weeks of gestation (No, %) 1 (16.66) 1 (10) - - 1 (16.66) 3 (10) Mode of delivery (No, %) Vaginal birth - 5 (50) 5 (83.3) - 2 (33.3) 12 (40) Caesarean section 6 (100) 5 (50) 1 (16.6) 2 (100) 4 (66.6) 18 (60) Preterm rupture of membranes (No, %) Yes 1 (16.6) 6 (60) 4 (66.6) 1 (50) 2 (33.3) 14 (46.6) No 5 (83.3) 4 (40) 2 (33.3) 1 (50) 4 (66.6) 16 (53.3) Obstetrics comorbidities (No, %) Miscarriages in reproductive anamnesis 1 (16.6) 2 (20) - - 1 (16.6) 4 (13.3) Infertility/Assisted reproductive technology 1 (16.6) - - - 1 (16.6) 2 (6.6) Genital infection (colpitis and pelveoperitonitis) - 2 (20) 3 (50) - 1 (16.6) 6 (17.64) Previous Caesarean section, uterine septum resection, myomectomy 1 (16.6) 2 (20) 1 (16.6) - 1 (16.6) 4 ( 13.3) Gestational hypertension /preeclampsia - - 1 (16.6) 1 (33.3) 1 (16.6) 3 (8.82) Hypothyroidism - 1 (10) 6 (100) - 1 (16.6) 8 (23.52) Other maternal comorbidities† - 3 (30) 1 (16.6) - - 4 (13.33) Maternal mortality 1 (16.6) 1 (10) - - - 2 (6.6) Type of treatment of cancers (No, %) Surgery 6 (100) - 6 (100) 2 (100) 5 (83.3) 19 (63.3) Chemotherapy 5 (83.3) 10 (100) - 2 (100) 5 (83.3) 22 (73.3) Radiotherapy - 8 (80) 6 (100) - 2 (33.3) 16 (53.3) *malignant melanoma (2), epipharyngeal cancer (1), hepatic angiosarcoma (1), osteosarcoma of the humerus (1) and abdominal ganglioneuroblastoma (1). † varicose veins, earlier surgery of persistent ductus arteriosus, avascular necrosis of the femoral head, bilateral hydronephrosis

459 Medicinski Glasnik, Volume 17, Number 2, August 2020

The most frequent were primiparous women, 8.01±1.78 (range 2- 9) in the first minute, and in all type of cancers; one woman had twins. 8.63±0.76 (range 6-9) in the fifth minute. Immi- The average weeks of gestation was 38.13±2.02 nent fetal asphyxia (Apgar score ˃7) was found weeks (range 29- 40). The overall prevalence in 13.3% newborns, and incipient (Apgar score of deliveries with less than 37 weeks gestation ˂7) in 16.6% newborns; 16.6% newborns were was 10%. Caesarean section was a more frequ- admitted to neonatal intensive care unit. No ent mode of delivery (60%); in genital and brain stillbirths and congenital malformations were cancers it was 100%, and the lowest in women recorded (Table 2). with thyroid cancer, 16.6%. The most frequently encountered cancer types Preterm rupture of membranes was found in 14 were haematological, in ten (33.3%), genital, (46.6%) pregnancies; most frequently in women in six (20%), and thyroid cancer, in six (20 %) with haematological and thyroid cancers, 60% women. The most frequent was Hodgkin lymp- and 66.6%, respectively. Miscarriages with less homa, in eight (26.6%) women. Among geni- than 10 weeks of gestation in reproductive ana- tal cancers ovarian cancer was represented in mnesis were found in 13.3% women, most frequ- four (13.3%) and breast cancer in two (6.6%) ently in haematological cancers, 20% (Table 1). women. Among extra genital cancers, haema- Infertility and assisted reproductive technology tological (Hodgkin lymphoma, acute myeloid was found in two (6.6%) women. Genital infec- leukaemia, and multiple myeloma), thyroid tion (colpitis and pelveoperitonitis) was found cancer, brain cancers (cerebellar and lateral in six (17.64%) women, most frequently in wo- ventricle cancer), and others (malignant mela- men with thyroid and haematological cancers noma, epipharyngeal cancer, hepatic angiosar- (50% and 20%). Gestational hypertension/preec- coma, osteosarcoma of the humerus, abdominal lampsia was found in three (8.8%) women with ganglioneuroblastoma) were found. thyroid (16.6%), brain (33.3%) and other cancers One women accidentally found out that she was (16.6%). Hypothyreoidism was found in eight pregnant (20 weeks) during the treatment for re- (23.52%) women, mostly in women with thyroid lapses of acute myeloid leukaemia M4, which cancers (100%). had previously been discovered and treated. The most frequent were female newborns, in all The average time span from the diagnosis and type of cancers. The average birth weight was treatment to delivery was 59.2±44.4 months (5 3202±557.36 (range 1250- 4070 grams). The years), from 12 months to 15 years, the highest in highest birth weight was found in women with other cancers (8.5 years), and the lowest in thyro- thyroid cancer (3471). Average birth length was id cancer (4.3 years). Nineteen (63.3%) women 52.76±3.4 (range 40-57 centimetres) (Table 2). underwent surgery, 22 (73.3%) chemotherapy The prevalence of intrauterine growth restric- and 16 (53.3%) radiotherapy. During the follow tion was 13.3%. The average Apgar score was up period (from 1 January 2014 to 31 Decem-

Table 2. Obstetric characteristics of newborns according to type of mother cancer Extra genital Characteristics Genital Others* Total (No, %) Haematological Thyroid Brain Gender (No, %) Male 2 (33.3) 7 (70) 2 (33.3) 1 (50) 2 (33.3) 14 (46.6) Female 4 (66.6) 3 (30) 4 (66.6) 1 (50) 4 (66.6) 16 (53.3) Birth weight (mean SD) (g) 3170±409.14 3144±472.02 3471.6±463.48 3345±360.62 2815.71±989.3 3202±557.36 Birth length (mean SD) (cm) 52.5±1.97 52.5±2.5 54.6±2.65 55±1.41 49.42±6.57 52.76±3.45 Intrauterine growth restriction (No, %) 1 (16.6) 1 (10) - - 2 (33.3) 4 (13.3) Apgar score (AS) (mean SD) (min) First min. 8.5±0.83 7.4±2.01 9±0 9±0 8±2.44 8.1±1.78 Fifth min. 8.83±0.40 8.3±0.82 9±0 9±0 8.5±1.22 8.63±0.76 Condition of fetus after delivery (No, %) Imminent fetal asphyxia (AS>7) 1 (16.6) 2 (20) - - 1 (16.6) 4 (13.3) Incipient fetal asphyxia (AS<7) 1 (16.6) 3 (30) - 1 (16.6) 5 (16.6) Neonatal Intensive Care Unit 1 (16.6) 3 (30) - - 1 (16.6) 5 (16.6) *the same as in Table 1

460 Cerovac et al. Pregnancy and cancer

Table 3. Types of cancer/stage, treatment and time span from treatment to delivery Time Radiotherapy Total Type of surgery Chemotherapy (YES/NO) span from Type/stage of cancers (YES/NO) (No, %) (No of women) (No of women) treatment to (No of women) delivery Genital 6 (20) Ovarian cancer FIGO Ia Unilateral 1 (3.3) NO NO 24 months (Cystadenocarcinoma serosum) adnexectomy (1) Unilateral Ovarian cancer FIGO IIIc 18 and 24 2 (6.6) adnexectomy, 6 cycles (taxol/carboplatina) (2) NO (Cystadenocarcinoma serosum) months Omentectomy (2) Unilateral Embryonic ovarian cancer FIGO IIIc 6 cycles (bleomycin, etoposid, 1 (3.3) adnexectomy, NO 8 years (Ca embrionale ovarii) cysplatina) Omentectomy Breast cancer 6 cycles+hormonotherapy (letrosol, - pT1bN0M0 (1) 24 months and 2 (6.6) Mastectomy (2) tamoxifen) +imunotherapy (trastuzu- 1 - pT2bN1M0 (1) 11 years mab) (2) (Carcinoma ductale invasivum) Extragenital 24 (80) Cerebelar Craniotomy and 1 (3.3) 6 cycles chemotherapy NO 8 years (Astrocytoma anaplasticum) tumour ablation Tumorectomy, Lateral ventricle Ventriculocysterno- 1 (3.3) 6 cycles chemotherapy NO 5 years (High grade ependimoma) stomy, Ventriculo- peritoneostomy Hepatic angiosarcoma 1 (3.3) Tumour extirpation 6 cycles chemotherapy NO 13 years Amputation of neoadjuvant and adjuvant Osteosarcoma of the humerus 1 (3.3) NO 15 years extremity chemotherapy Abdominal ganglioneuroblastoma 1 (3.3) Tumour extirpation 6 cycles NO 13 years Excision of the Malignant melanoma pT1N0M0 1 (3.3) NO NO 12 months tumour Excision of the Malignant melanoma pT2N1M0 1 (3.3) tumour, lymph 4 cycle (dacarbazine) YES 4 years node extirpation Epipharingeal cancer pT1N0M0 6 cycles (docetaxel, cisplatin, florou- 1 (3.3) NO YES 5 years (Carcinoma planocellulare) racil) Thyroid gland (Carcinoma papillare Thyreoidectomy Radioactive iodine 1 (3.3) NO 24 months invasivum glandulae thyroideae) (1) treatment (1) Thyroid gland (Carcinoma papillare Thyreoidectomy Radioactive iodine invasivum glandulae thyroideae) 5 (16.6) NO 3 – 8 years (5) treatment (5) pT1bN0M0 (5) Lymphoma Hodgkin IIA/B stage (Nodu- ABVD regimen (adriamycin+bleomyci Radiotherapy of the lar sclerosis) (6) 8 (26.6) NO n+Vinblastine+dacarbazine) (7) neck and mediasti- 3 – 10 years IIIB stage (2) ABVD+BEA COPP regimen 4 cycle (1) num (8) Multiple myeloma 1 (3.3) NO 6 cycles NO 3 years Before pregnancy: AD (adriamycin+dexamethason) 3+7 Acute myeloid leukaemia M4 (FAB) 1 (3.3) NO NO 24 months protocol for induction and reinduction. In pregnancy: cytosar ber 2019) two (6.66%) women died. Their deaths higher than 21 and 27, respectively found in our were caused by acute myeloid leukaemia M4 study. Approximately 12% of ovarian cancers (two months after delivery) and ovarian cancer occurred under the age of 44 (6). Gerstl et al. re- FIGO IIIC, respectively (17 months after deli- ported that most women with ovarian cancer had very) (Table 3). grade 1- stage IA, and underwent fertility-sparing treatment including unilateral oophorectomy with DISCUSSION adjuvant chemotherapy, similar to our study (6). In a systematic review by Gerstl et al. (6) the In our study we have found four women with women’s mean age at the time they were diagno- ovarian cancer (three were with epithelial ovarian sed with genital cancer, as well as the time of de- cancers). According to the stage of ovarian cancer livery was 30.5 and 30.3, respectively, which is one women was in stage FIGO IA, two in stage

461 Medicinski Glasnik, Volume 17, Number 2, August 2020

FIGO IIIC, and one with embryonic ovarian can- Two studies (14,15) reported that the mean age cer in stage FIGO IIIC, and all had been treated during the pregnancy and at delivery for thyroid with unilateral adnexectomy and chemotherapy, cancer women was 29.9 and 29.7 years, respec- and gave birth to healthy newborns. Kashima et al. tively, which means younger women compared (12) reported among 18 women who underwent to our data (32.5). The same studies (14,15) re- fertility sparing surgery with epithelial ovarian ported that the duration between the diagnosis of cancer FIGO stage IC, seven singleton pregnan- thyroid cancer and the beginning of pregnancy cies for five women, compared to our results, four was 40.5 and 60.9 months, respectively, in com- singleton pregnancies in four women; they suggest parison with 51.6 months in our study. In the two that fertility sparing surgery for ovarian cancer is a studies the median age of thyroid cancer diagno- valid treatment option for women of reproductive sis was 36 and 24.7 years, comparing to 28.1 in age who strongly desire to conceive (12). Gerstl our study (14,16). Hirsch et al. reported 47.6% et al. (6) reported 15% of miscarriages and 10% and 17.5% of women in T1 and T2 stage of the of preterm birth rates, which is lower than in our disease, respectively, whereas in our study 83.3% study, 16.6% and 16.6%, respectively. and 16.6% of women were in T1 and T2 stage of In a systematic review and meta-analysis con- the disease, respectively (14). In our study all pa- ducted by Gerstl et al. (7) the most commonly re- tients underwent thyroidectomy and subsequent ported haematological diagnoses were Hodgkin radioactive iodine treatment (RAI) compared to or non-Hodgkin lymphoma (40%), chronic mye- 93.6% and 92%, respectively, in an Israeli study loid leukaemia (21%), acute myeloid leukaemia (14). No women in our study had thyroid cancer (15%), and acute lymphoblastic leukaemia (7%). progression/recurrence during pregnancy, com- In our study we have found 33.3% of haemato- pared to 9.5% in the Israeli study (14). In a syste- logical diagnoses out of all cancers: Hodgkin matic review by Sawka et al. (8) it was found that lymphoma 26.6%, multiple myeloma 3.3%, and RAI treatment for thyroid carcinoma was not ge- acute myeloid leukaemia 3.3%, which had been nerally associated with an increase of long-term treated with chemotherapy and radiotherapy. risk of infertility, miscarriage, stillbirths, or congenital defects, which was the same for women De Sanctis et al. reported that the median time with thyroid carcinoma not treated with RAI; it from diagnosis to delivery was nine years, com- correlates with our results. Blackborn et al. (16) pared to 5.7 years in our study (13). The female/ reported complications connected with increased male ratio in the Italian study was 51% / 49% risks for haemorrhage and diabetes/or abnormal (10), which is different from our finding, 30%/ glucose tolerance during pregnancy, childbirth, 70%. The median birth weight found in the Itali- and the puerperium in thyroid cancer survivors. an study was similar to our data (3220 vs. 3144) We did not record these complications. Our stu- (13). De Sanctis et al. reported 12% of miscarri- dy confirms that pregnancy does not have an images, 7% of premature births, 2% of low birth pact on the recurrence of the disease in women weight infants in women treated for Hodgkin who survived thyroid cancer without evidence of lymphoma, which is lower in comparison to our disease’s persistence before the conception. results (13). In the study by De Sanctis et al. two cases of congenital malformations were recorded A Turkish study (1) reported 68 pregnant women and no cases of stillbirths, while in our study no who survived cancer, where the most frequently cases of congenital malformations or stillbirths encountered cancer types were thyroid (26.4%), were found (13). In our study 80% of women haematological (22.1%), genital (19.1%), and bre- with haematological cancer underwent radiothe- ast cancer (13.2%), which slightly differs from our rapy (neck and mediastinum) and all women had results (20%, 33.3%, 13.3% and 6.6%, respec- chemotherapy before pregnancy, which is higher tively). The mean maternal age (31.7 vs. 29.4), compared to the study by De Sanctis et al. (91% birth weight (3030 vs. 3202) and gestational age and 66%, respectively) (13). A Serbian study (4) (37.5 vs. 38.1) in Davutoğlu et al. study were very reported good maternal and neonatal outcome similar to our findings (1); the prevalence of nulli- in women with acute myeloid leukaemia M4, in parity (63.3%), miscarriages (13.3% vs. 2.9%), contrast to our study. preterm birth (10% vs. 8.8%), intrauterine growth

462 Cerovac et al. Pregnancy and cancer

restriction (IUGR) (8.8% vs. 5.8%) and gestational comparing to our finding, e.g. two maternal de- hypertension / preeclampsia (8.8% vs. 2.9%) was aths (out of 30); both deaths in our study were higher in our study, respectively. The most proba- due to late diagnosis and advanced-stage cancers, ble reason for this difference is that Davutoğlu et and both babies survived. The main objective in al. excluded women with other coexisting medical such advanced cancers with poor maternal pro- conditions and any sequelae associated with che- gnosis is to prolong the delivery to avoid extreme motherapy and radiotherapy from the analysis. A prematurity (1,19). Georgian study reported higher incidence of pre- Oncologists should offer oncofertility counse- term delivery in breast cancer and leukaemia sur- lling and fertility preservation in women before vivors in contrast to survivors of Hodgkin lymp- the start of cancer treatment (6,11,19,20). Wo- homa, melanoma, and thyroid cancer (17). We men with cancer are generally advised to delay reported a high incidence of Caesarean section (60 pregnancy for up to 2 years following the cancer %), which correlates with Davutoglu et al. (66.6%) treatment in order to identify possible relapse, and Hartnett et al. studies (1,17). The high number and because of the time needed for the oocyte to of women with miscarriages, IUGR and Caesarean recover from the damage caused by chemothe- section in our study is probably a result of numero- rapy and radiotherapy (5,11,19,20). Women us comorbidities in pregnancies, earlier surgeries, who successfully conceive subsequently to the chemotherapy /radiotherapy /radioactive iodine treatment should be monitored throughout their treatment, and older age of pregnant women. The pregnancies by the treating gynaecological on- preterm rupture of membranes was found in 46.6% cologist, foetal medicine and obstetrics and re- of term gestations, mostly in haematological and productive specialist in order to reduce potential thyroid cancers (60% and 66.6%, respectively). pregnancy and birth complications (6,11,19,20). This is high frequency when compared to healthy There are several limitations that should be addre- adolescent and adult pregnancies (39.44% and ssed. Firstly, we describe women presented with 21.33%), probably because of comorbidities and a combination of different cancers and treatments consequences of cancer treatment (18). without specifying which treatment resulted in a The prevalence of admission to neonatal intensi- specific reproductive outcome. Furthermore, the- ve care unit (NICU) in our study was higher than re were inconsistencies and underreporting in in the Turkish study (16.6% vs. 8.8%), mostly some patients in a dosage of chemo- and radi- due to prematurity (1). There were no early ne- otherapy, and study population was small. Des- onatal deaths or congenital anomalies detected pite the limitations associated with this study, we in the newborns in both studies, but the Turkish reported several interesting findings. study reported one stillbirth (1). Miscarriage, In conclusion, although in small sample, our data preterm birth and IUGR are common findings demonstrate a favourable obstetric and neonatal among the offspring of female childhood cancer outcome for women who have survived cancer. survivors who received chemotherapy and abdo- The women considering pregnancy after cancer minal, pelvic or total body irradiation (9,10,15). treatment should not be necessarily afraid, but Women intending to become pregnant after sur- should be counselled carefully about perinatal ri- viving cancer should be strongly supported, the- sks and risks of recurrence. Prenatal care should ir pregnancy and delivery should be monitored be tailored to the specific cancer and risks, and closely and in a multidisciplinary manner. The appropriate support should be offered by the ob- influence of pregnancy on the course of cancer stetrician and oncologist guiding that pregnancy. and the risk of relapse has to be discussed indivi- FUNNDING dually with the patient, depending on the tumour type and its stage (1). No specific funding was received for this study. Davutoglu et al. (1) reported four (out of 31) ma- TRANSPARENCY DECLARATION ternal deaths due to the advanced stage of breast and gastrointestinal cancers, which is higher Conflict of interest: None to declare.

463 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES 1. Alıcı Davutoğlu E, Madazlı R, Yılmaz N, Ozel A, 11. Cerovac A, Ljuca D, Nevacinovic E, Tulumovic A, Uludag S, Sozen I. Pregnancy in cancer patients and Iljazovic E. Giving birth after fertility sparing tre- survivors; experience of a university hospital in Tur- atment of embrional carcinoma Figo III C: case re- key. J Obstet Gynaecol 2017; 37:1015-9. port and literature review. Med Arch 2018; 72:371-3. 2. Clark H, Kurinczuk JJ, Lee AJ, Bhattacharya S. Ob- 12. Kashima K, Yahata T, Fujita K, Tanaka K. Outco- stetric outcomes in cancer survivors. Obstet Gyne- mes of fertility sparing Surgery for women of repro- col. 2007; 110:849-54. ductive age with FIGO stage IC epithelial ovarian 3. Landa A, Kuller J, Rhee E. Perinatal considerations cancer. Int J Gynaecol Obstet 2013; 121:53-5. in women with previous diagnosis of cancer. Obstet 13. De Sanctis V, Filippone FR, Alfò M, Muni R, Ca- Gynecol Surv 2015; 70:765-72. valieri E, Pulsoni A, Annechini G, Valeriani M, Osti 4. Jeremic K, Stefanovic A, Dotlic J, Kadija S, Kontic MF, Minniti G, Enrici RM. Impact of different tre- O, Gojnic M, Jeremic J, Kesic V. Cancer during pre- atment approaches on pregnancy outcomes in 99 gnancy – clinical characteristics, treatment outco- women treated for Hodgkin lymphoma. Int J Radiat mes and prognosis for mothers and infants. J Perinat Oncol Biol Phys 2012; 84:755-61. Med 2018; 46:35-45. 14. Hirsch D, Levy S, Tsvetov G, Weinstein R, Lifshitz 5. Stensheim H, Cvancarova M, Møller B, Fosså SD. A, Singer J, Shraga-Slutzky I, Grozinski-Glasberg Pregnancy after adolescent and adult cancer: a po- S, Shimon I, Benbassat C. Impact of pregnancy on pulation-based matched cohort study. Int J Cancer outcome and prognosis of survivors of papillary 2011; 129:1225-36. thyroid cancer. Thyroid 2010; 20:1179-85. 6. Gerstl B, Sullivan E, Vallejo M, Koch J, Johnson M, 15. Budak A, Gulhan I, Aldemir OS, Ileri A, Tekin E, Wand H, Webber K, Ives A, Anazodo A. Reproducti- Ozeren M. Lack of influence of pregnancy on the ve outcomes following treatment for a gynecological prognosis of survivors of thyroid cancer. Asian Pac J cancer diagnosis: a systematic review. J Cancer Sur- Cancer Prev 2013; 14:6941-3. viv 2019; 13:269-81. 16. Blackburn BE, Ganz PA, Rowe K, Snyder J, Wan Y, 7. Gerstl B, Sullivan E, Koch J, Wand H, Ives A, Deshmukh V, Newman M, Fraser A, Smith K, Her- Mitchell R, Hamad N, Anazodo A. Reproductive get K, Kim J, Kirchhoff AC, Porucznik C, Hanson H, outcomes following a stem cell transplant for a hae- Abraham D, Monroe M, Hashibe M. Reproductive matological malignancy in female cancer survivors: and gynecological complication risks among thyroid a systematic review and meta-analysis. Support Care cancer survivors. J Cancer Surviv 2018; 12:702-11. Cancer 2019; 27:4451-60. 17. Hartnett KP, Ward KC, Kramer MR, Lash TL, Mer- 8. Sawka AM, Lakra DC, Lea J, Alshehri B, Tsang RW, tens AC, Spencer JB, Fothergill A, Howards PP .The Brierley JD, Straus S, Thabane L, Gafni A, Ezzat S, risk of preterm birth and growth restriction in pre- George SR, Goldstein DP. A systematic review exa- gnancy after cancer. Int J Cancer 2017; 141:2187-96. mining the effects of therapeutic radioactive iodine 18. Marković S, Bogdanović G, Cerovac A. Prematu- on ovarian function and future pregnancy in fema- re and preterm premature rupture of membranes in le thyroid cancer survivors. Clin Endocrinol (Oxf) adolescent compared to adult pregnancy. Med Glas 2008; 69:479-90. (Zenica) 2020; 17:136-40. 9. Signorello LB, Cohen SS, Bosetti C, Stovall M, Kas- 19. Peccatori FA , Azim HA Jr, Orecchia R, Hoekstra per CE, Weathers RE, Whitton JA, Green DM, Do- HJ, Pavlidis N, Kesic V, Pentheroudakis G; ESMO naldson SS, Mertens AC, Robison LL, Boice JD Jr. Guidelines Working Group. Cancer, pregnancy and Female survivors of childhood cancer: preterm birth fertility: ESMO Clinical Practice Guidelines for di- and low birth weight among their children. J Natl agnosis, treatment and follow-up. Ann Oncol 2013; Cancer Inst. 2006; 98:1453-61. 24:vi160-70. 10. van der Kooi ALF, Kelsey TW , van den Heuvel- 20. Barton SE , Najita JS, Ginsburg ES, Leisenring WM, Eibrink MM , Laven JSE , Wallace WHB, Anderson Stovall M, Weathers RE, Sklar CA, Robison LL, RA. Perinatal complications in female survivors of Diller L. Infertility, infertility treatment, and achie- cancer: a systematic review and meta-analysis. Eur J vement of pregnancy in female survivors of childho- Cancer. 2019; 111:126-37. od cancer: a report from the Childhood Cancer Sur- vivor Study cohort. Lancet Oncol 2013; 14:873-81.

464 ORIGINAL ARTICLE

Preterm and term birth in twin pregnancies during a seven-year period: a call for obstetricians to declare about amnionicity and chorionicity

Anis Cerovac¹,², Gordana Grgić³, Dženana Softić³, Dženita Ljuca³, Sergije Marković4, Amer Mandžić³

ABSTRACT

Aim To compare maternal, foetal and neonatal characteristics, and perinatal outcome of preterm and term deliveries in twins pregnancies in order to improve perinatal care in Bosnia and Herzegovina.

Methods This retrospective cohort study included pregnant women with twin pregnancy who delivered during the period between 1 January 2012 and 31 December 2018 at the Clinic for Gynae- cology and Obstetrics, University Clinical Centre Tuzla.

Results During the seven-year period 26 734 deliveries were recorded, out of which 362 (1.35 %) were twin pregnancies, 226 (62.4%) preterm and 136 (37.5%) term ones. In the preterm group Corresponding author: 38 (16.8%) pregnancies were assisted medical reproduction, and 16 (11.7%) of those were in the term group. The average birth Anis Cerovac weight was significantly higher for the first twin in both grou- General Hospital Tešanj ps (p<0.00001). Incipient intrauterine foetal asphyxia was more Braće Pobrić 17, 74260 Tešanj, frequent in the preterm group (p<0.05). The most common indica- Bosnia and Herzegovina tion for Caesarean section was abnormalities of foetal presentation Phone: +387 32 650 662; and lie, 176 (68.2%) for the overall sample. Fax: +387 32 650 605; Conclusion Cornerstone of twin pregnancy antenatal care is to get E-mail: [email protected] correct data about amnionicity and chorionicity. Since majority of ORCID ID: https://orcid.org/0000-0002- prenatal data did not have this information we call all obstetricians 7209-382X to declare about amnionicity and chorionicity in twin pregnancies during the first trimester ultrasound examination.

Key words: foetal membranes, multiple pregnancy, premature labour, prenatal care Original submission: 29 February 2020; Revised submission: 13 March 2020; Accepted: 04 May 2020 doi: 10.17392/1148-20

Med Glas (Zenica) 2020; 17(2): 465-471

465 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION obstetrician, anaesthesiologist and neonatologist is crucial for successful delivery (1). Twin pregnancy is associated with a higher risk of perinatal mortality and morbidity, especially in Many studies about the outcomes of twin pre- monochorionic twin pregnancies (1). The inciden- gnancies have shown higher rates of miscarriages ce of multiple pregnancies is rising and accounts (missed abortion one of the twin or vanishing for about 3% of pregnancies. The leading cause is twin), chromosomal and structural abnormaliti- an induction of multiple ovulations by medications es (5), anaemia, pre-eclampsia, gestational dia- in order to stimulate ovulation and medically assi- betes, higher rates of prematurity (8), preterm sted reproduction (MAR) (2-4). The frequency of premature rupture of membranes, lower birth twin pregnancies also depends on the age, parity weight, stillbirth, higher rates of CS, postpartum and heredity (2,3). Twin pregnancies should be ca- haemorrhage (1), and more frequent admissions red for by an experienced multidisciplinary team to neonatal intensive care unit (NICU) (6,9,10). (1,5). Both chorionicity and amnionicity are cru- Studies that compared preterm and term twin cial for antenatal management of twin pregnancies pregnancies are rare. and should be determined by ultrasound in the first Despite of the increasing maternal age and the trimester by identification of the lambda or ‘’T“ prevalence of twin pregnancies in Bosnia and sign (1,5). There are specific recommendations for Herzegovina there were no studies about the ultrasound examinations in twin pregnancy when course and outcome these pregnancies. it comes to timing, frequency and content of ultra- The aim of this study is to compare maternal, fo- sound assessment, as well as screening for and pre- etal, neonatal characteristics, caesarean section natal diagnosis of aneuploidy and structural abnor- indications and perinatal outcome between pre- malities (1). Monozygotic twins have increased the term and term deliveries in twin pregnancies in risk of foetal/perinatal loss, various foetal anomali- order to improve the perinatal care in Bosnia and es and intrauterine growth restriction (IUGR) (1,5). Herzegovina. Preterm births are more common among twins PATIENTS AND METHODS compared with singleton pregnancies. It is estimated that 40-60% of twins will be delivered Patients and study design spontaneously before 37 weeks of gestation, This retrospective cohort study included pregnant most probably because they reach maturity ear- women with a twin pregnancy who delivered du- lier (1). The timing of delivery for monochori- ring the period between 1 January 2012 and 31 onic/monoamniotic twin pregnancy is usually December 2018 at the Clinic for Gynaecology between 32–34 weeks, monochorionic/diamnio- and Obstetrics, University Clinical Centre Tuzla. tic at 36 weeks, dichorionic at 37 weeks (1,5). Inclusion criteria were pregnant women who de- Vaginal delivery is usually recommended for un- livered twins. Exclusion criteria were twin pre- complicated dichorionic/diamniotic twins if the gnancies less than 24 weeks of gestation, twins leading twin is cephalic and if there are no other birth weight less than 500 grams, trigeminy, pre- obstetric indications for Caesarean section (CS). gnancies with one missed or vanishing twin and If twin one is breech, CS might be a more fa- pregnancies with stillbirths. vourable option (1,5,6). Monochorionic/diamniotic twins will commonly be delivered by CS, Patients were divided into two groups: preterm the option of vaginal birth might be offered to birth (below 37 weeks of gestation) and term the parents if the leading twin is cephalic (1,2). birth (above 37 weeks of gestation). Monochorionic/monoamniotic twins, according The survey was approved by the Ethics Com- to recommendations, will be delivered by CS (7). mittee of the University Clinical Centre Tuzla. Active management for delivery of the second twin is usually recommended to avoid a prolon- Methods ged intertwine delivery (1). If needed, stabilizati- Data on the course of pregnancy and childbirth on of lie, internal podalic version, breech extrac- were collected from medical records (patient his- tion or immediate CS can be employed. In such tory and partograms). Obstetrical data included circumstances, experience and competence of the maternal age at delivery, parity, gestational age

466 Cerovac et al. Twin pregnancies

at delivery, mode of delivery (vaginal/CS, indi- stical data processing. Statistical significance of cations for CS and obstetric comorbidities (pre- the difference between mean values of variables term delivery <37 and early preterm delivery <32 in the groups was tested by ANOVA test and weeks of gestation, extragenital/genital disease, Tukey's post hoc test. The Student's t-test (sta- gestational hypertension/preeclampsia, premature tistical comparison test) was used to determine rupture of membrane, placental, amniotic fluid and statistical significance of the difference between umbilical cord abnormalities, MAR). Foetal data a set of two data. Statistically significant diffe- included intrauterine asphyxia, abnormalities of rence was set at p<0.05. foetal presentation and lie, anomalies and IUGR. Neonatal data included: gender, birth weight (ex- RESULTS tremely low birth weight – ELBW below 1000 g, During the seven-year period recorded 26 734 very low birth weight - VLBW 1000-1500 g, low deliveries were recorded, out of which 362 (1.35 birth weight - LBW 1500-2500 g) and birth length. %) were twin pregnancy, 226 (62.4%) were pre- Perinatal outcome was assessed according to the term, and 136 (37.5%) were term births. Out Apgar score at first and fifth minute, as imminent of the total number of deliveries, 1860 (6.9%) (Apgar score ˂9 and ˃7, respectively) and incipi- were preterm, of which 226 (12.1%) were twin ent (Apgar score ˂8) intrauterine asphyxia, and pregnancies. Early preterm birth was recorded admission to a NICU (11). in 42 (out of 362; 11.6%). Spontaneous preterm delivery was recorded in 131 (57.9%) twin pre- Statistical analysis gnancies. During the observed period significant Descriptive statistics, mean value, standard de- changes were not noticed in the prevalence of viation (SD) and percentage were used in stati- twin pregnancies (p=0.397).

Table 1. Maternal characteristics of twin pregnancies No (%) of women in the group Maternal characteristics p Preterm Term Total Twin pregnancies 226 (62.4) 136 (37.5) 362 (1.3) 0.437 Maternal age at the time of delivery (years) Mean (±SD) 29.25 (±5.65) 29.92 (±5.18) 29.5 (±5.48) 0.130 <19 7 (3) 0 7 (1.9) 19-35 176 (77.89 106 (77.9) 282 (77.9) >35 42 (18.5) 30 (22.1) 72 (19.8) 0.324 Parity Primiparous 137 (60.6) 58 (42.6) 195 (53.8) Secundiparous 60 (26.5) 52 (38.2) 112 (30.9) Thirdparous and multiparous 28 (12.3) 26 (19.1) 54 (14.9) 0.215 Extragenital diseases* 19 (8.4) 16 (11.7) 35 (9.6) 0.5 Genital diseases Gynaecological surgeries 6 (2.6) 5 (3.6) 11 (3) Cerclage 8 (3.5) 2 (1.4) 10 (2.7) Previous caesarean section 14 (6.1) 12 (8.8) 26 (7.1) Corrected/uncorrected genital anomalies 7 (3) 0 7 (1.9) 0.127 Genital infections 12 (5.3) 12 (8.8) 24 (6.6) Colpitis 8 (3.5) 8 (5.8) 16 (4.4) Chorioamnionitis 2 (0.8) 0 2 (0.5) 0.5 Pelveoperitonitis 2 (0.8) 4 (2.9) 6 (1.6) Abnormalities of placenta, amniotic fluid and umbilical cord 27 (11.9) 25 (18.3) 52 (14.3) Placental abruption 9 (3.9) 2 (1.4) 11 (3) Placenta previa 1 (0.4) 1 (0.7) 2 (0.5) Placenta suspecta 2 (0.8) 7 (5.1) 9 (2.5) Polihydramnios 4 (1.7) 5 (3.6) 9 (2.5) Complications with umbilical cord (wrapped around neck and body, true node and prolapse) 11 (4.8) 10 (7.3) 21 (5.8) 0.451 Gestational hypertension/preeclampsia 24 (10.6) 24 (17.6) 48 (13.2) 0.5 Medical assisted reproduction 38 (16.8) 16 (11.7) 54 (14.9) 0.5 Premature and preterm rupture of membranes 95 (42) 18 (13.2) 113 (31.2) 0.5 Spontaneous preterm delivery 131 (57.9) - 131 (36.1) Earlier miscarriages 5 (2.2) 8 (5.8) 13 (3.6) 0.5 Mode of delivery Vaginal delivery 56 (24.7) 48 (35.2) 104 (28.7) Caesarean section 170 (75.2) 88 (64.7) 258 (71.27) 0.266 *cardiovascular, pneumological, gastrointestinal, endocrinological, ophthalmological and neurological

467 Medicinski Glasnik, Volume 17, Number 2, August 2020

The prevalence of twin pregnancies was ranged (16.8% and 20.5%), compared with the second from 49 (out of 3688; 1.3%) deliveries in 2015 to one (22.1% and 22.7%), in the preterm and the 64 (out of 3720; 1.7%) in 2017. The overall aver- term group, respectively (p=0.07) (Table 2). age maternal age at the time of the delivery was The average Apgar score in the first (p<0.025) and 29.5±5.48, in the preterm birth group 29.25±5.65 the fifth (p<0.027) minute was significantly higher (ranged 16-43), and in the term birth group for the first than for the second twin in the pre- 29.92±5.18 (ranged 20-43) years. term group. The average Apgar score in the first Adolescent (below 19 years) twin pregnancies were and fifth minute was significantly lower in the pre- represented only in the preterm group, in seven (3%) term than in the term group (p<0.05). Imminent women, resulting in the overall prevalence of 1.9%. and incipient intrauterine asphyxia of the first and The overall prevalence of pregnant women above the second twin were more frequent in the preterm 35 years of age was 19.9% (out of 362). Abnorma- than the term group (p<0.05) (Table 3). lities of placenta, amniotic fluid and umbilical cord were present in the overall sample in 52 (14.3%), Table 3. Neonatal characteristics from twin pregnancies Neonatal No (%) of neonates in the group and in preterm and term group, in 27 (11.9%) and p characteristics Preterm Term Total 25 (18.3%) cases, respectively. Gestational hyper- Average Apgar score (±SD) tension/preeclampsia was presented with prevalence 7.50 8.53 7.89 1. minute I twin <0.00001 of 10.6% in the preterm group and in the term gro- (±1.82) (±0.99) (±1.64) 7.15 8.36 7.61 1. minute II twin <0.00001 up with 17.6 %, in the overall sample the prevalen- (±1.95) (±1.01) (±1.76) ce was 13.2%. Methods of MAR were implied in 7.99 8.72 8.26 5. minute I twin <0.00001 16.8% women in the preterm group and in 11.7% (±1.33) (±0.74) (±1.2) 7.73 8.65 8.08 in the term group. Preterm rupture of membranes, 5. minute II twin <0.00001 (±1.53) (±0.61) (±1.34) as expected, was higher in the preterm than the term Birth weight (gr) Average I twin 2189.23 2788.45 2414.35 group, 42% and 13.2%, respectively. Spontaneous <0.00001 preterm delivery was represented with 131 (57.9%) (±SD) (±543.36) (±400.50) (±573.1) Average II twin 2164.51 2757.79 2387.4 <0.00001 of all preterm twin deliveries. The most frequent (±SD) (±547.75) (±450.88) (±588.06) mode of delivery was CS in the overall sample ELBW in I twin 5 5 0 (<1000 g) (2.2) (1.4) (71.2%), as well as in the preterm and the term gro- 0.0246 ELBW in II twin 8 8 0 up, 75.2 % and 64.7%, respectively (Table 1). (<1000 g) (3.5) (2.2) VLBW in I twin 20 20 The most common foetal presentation and lie 0 (<1500 >1000 g) (8.85) (5.5) in the overall sample were cephalic/cephalic 0.0285 VLBW in II twin 12 12 0 (51.6%), transverse/oblique lie (19.8%), cepha- (<1500 >1000 g) (5.3) (3.3) lic/breech (12.9%), breech/cephalic (10.5%) and LBW in I twin 122 28 150 (<2500 >1500 g) (54) (20.5) (41.43) breech/breech (5.2%). Foetal anomalies were re- <0.002 LBW in II twin 135 31 166 corded in three (0.8%) newborns. The IUGR was (<2500 >1500 g) (59.7) (22.7) (45.85) recorded with lower prevalence in the first twin Average birth length (±SD) (cm) 46.83 50.71 48.29 I twin <0.00001 Table 2. Foetal characteristics in twin pregnancies (±4.59) (±2.57) (±4.38) No (%) of foetuses in the 47.07 50.52 48.37 II twin <0.00001 Foetal characteristics group P (±4.69) (±2.86) (±4.42) Preterm Term Total 452 272 724 Twin gender Foetal anomalies* 2 (0.8) 1 (0.7) 3 (0.82) 0.259 (226) (136) (362) Foetal presentation and lie 79 47 126 Male Cephalic/breech 34 (15) 13 (9.5) 47 (12.9) (17.4) (17.2) (34.8) Breech/cephalic 21 ( 9.2) 17 (12.5) 38 (10.5) 79 47 126 Female <0.00004 Breech/breech 10 (4.4) 9 (6.6) 19 (5.2) (17.4) (17.2) (34.8) Transverse/oblique lie I/II twin 33 (14.6) 39 (28.6) 72 (19.9) 75 38 113 Both male Cephalic/cephalic 128 (56.6) 59 ( 43.3) 187 (51.6) 0.232 (33.1) (27.9) (31.2) Intrauterine asphyxia 72 51 123 Both female Imminent I twin 101 (44.6) 40 (29.4) 141 (38.9) <0.05 (31.8) (37.5) (33.9) Imminent II twin 116 (51.3) 49 (36) 165 (45.6) 0.009 Admission to NICU Incipient I twin (AS<7) 44 (19.4) 3 (2.2) 47 (13) <0.05 44 3 47 I twin Incipient II twin (AS<7) 55 ( 24.3) 10 (7.3) 65 (18) 0.01 (19.4) (2.2) (13) 0.01 Intrauterine growth restriction 55 10 65 II twin I twin 38 (16.8) 28 (20.5) 66 (18.2) ( 24.3) (7.3) (18) II twin 50 (22.1) 31 (22.7) 81 (22.3) 0.071 ELBW Extremely low birth weight; VLBW, very low birth weight; *two hearth and one brain anomalies LBW Low birth weight; NICU, neonatal intensive care unit

468 Cerovac et al. Twin pregnancies

The average birth weight was significantly higher increase medically indicated preterm birth in mul- for the first twin in the preterm (p<0.00001) and tiple pregnancies as well (12). Israelian and Greek the term group (p<0.00001). The average birth (3,9) study reported 56% and 63% deliveries, res- length was lower in the first twin in the preterm pectively, prior to 37 weeks of gestation among group (p<0.289), and higher in the first twin in twin pregnancies, similarly to our results, 62.4%. the term group (p<0.288). Our study recorded 11.6% of early preterm birth, The average birth weight and length were signi- which is lower comparing to the groups below and ficantly lower in the preterm group (p<0.00001). above 35 years in the Greek study (23% and 28%, Extremely low birth weight was recorded in 13 respectively) (3). An Alabama study (13) reported twins, five (2.2%) in the first, and eight (3.5%) 1.3% twins of the total population contributing to in the second twin pregnancies. Very low birth 12.2% of the total preterm birth rate, which is simi- weight was recorded in 20 (8.8%) first twins, and lar to our report (1.35% and 12.1%, respectively). in 12 (5.3%) second twins. Low birth weight was The prevalence of spontaneous preterm deliveries more frequent in the preterm group (p<0.002). in our study is similar to other studies (13). These data from developed countries correspond to our The more frequent twin gender was male in the data, indicating good perinatal care in Bosnia and preterm group, and female in the term group Herzegovina (3,9,10,13). (p<0.00004). The male-female ratio was 352:372. Admission to NICU was found in 30.9% twins Aisien et al. reported the age range of twin for the overall sample (Table 3). mothers was 16-47 (mean of 28), which correlates with our study (10). Abnormalities of foetal presentation and lie were a predominant indication for CS, in 98 (57.6%) ca- Obstetrics comorbidities found in our study, even ses in the preterm, 78 (88.6%) in the term group though more common in the preterm group, were and 176 (68.2%) for the overall sample (Table 4). not significantly higher than in the term group. This was also shown by Prapas et al. (3). Israeli Table 4. Indications for Caesarean section study reported 29.3% preeclampsia/eclampsia and No (%) of caesarean 41.3% hypertensive disease, which is higher in Indications section in the group p comparison to our 13.2% (for both hypertension/ Preterm Term Total preeclamsia), probably because of the age diffe- 170 88 258 Total number of caesarean section 0.5 (75.2) (64.7) (71.2) rences (46.9 vs. 29.5 years of age) (9). A Tennesse- Uterine scar (previous caesarean 15 12 27 an study reported very similar frequency gestatio- 0.5 section, myomectomy) (8.8) (13.6) (10.4) nal hypertension/preeclampsia (12.9% vs 12.7%) 12 10 22 Placental abnormalities <0.05 with our study; however, a Turkish study reported (7) (11.3) (8.5) low frequency of preeclampsia (6%) (14). Prema- 11 10 21 Umbilical cord complications 0.477 ture rupture of membranes was found by Avnon et (6.4) (11.3) (8.1) al. in 28%, comparing to 31.2% in our study (9). Incipient intrauterine foetal 99 13 112 0.01 asphyxia (58.2) (14.7) (43.4) A Greek and Turkish study reported higher frequ- Abnormalities of foetal presenta- 98 78 176 ency of MAR than in our study, especially in older 0.287 tion and lie (57.6) (88.6) (68.2) women, probably because of greater availability DISCUSSION of MAR (3,15). Childbirth in twin pregnancies is often accompanied by abnormal presentations, The results of presented study have shown umbilical cord prolapse, placental abruption, pla- 13.5/1000 pregnancies average rate of twin pre- centa previa and low placental insertion, bleeding gnancies fitting within the framework of our po- due to atony and irregular peeling of the placenta pulation, as well as with some reports for Cauca- (16,17). All these complications were noted in our sians, 10-16.8/1000 pregnancies (3,10). study. We have found higher frequency of placen- Preterm birth in twin pregnancies is the leading tal abruption and umbilical cord complications cause of perinatal mortality and morbidity (2,11). It compared to the Turkish and Greek study (3,15). is believed that the high incidence of preterm birth This higher frequency of placental abruption in is most often due to overgrowth and overstretching our study is probably because of the higher frequ- of the uterus (2). Schaaf et al. reported significant ency of preeclampsia, as a risk factor for placental

469 Medicinski Glasnik, Volume 17, Number 2, August 2020

abruption (3,15). Experiences have been divided Admission to NICU in our study, as well as the considering the application of prophylactic cerc- frequency of Apgar score below 7 in the first lage in twin pregnancies (2). Murray et al. conclu- and the fifth minute correlated with other studies ded that there was no evidence of the usefulness (3,9,13). of cerclage in reducing the risk of preterm births The mode of delivery in premature twins, as in twins (17). The placement of cerclage with well as in all other combinations when the the usual indication is not contraindicated, on the first twin does not lead cephalic, is a matter contrary, it significantly reduces the proportion of of debate and controversy (7,16). Several stu- birth in the gestational group up to 31 weeks of dies reported different frequency of CS from gestation (2). A Turkish study reported a higher 26% (low), 52% (moderate) to 70% and 100% frequency of urgent and elective cerclage than we (high), matching our results in the high frequ- found in our study, probably because of a high ency group, 71.27% (3,9,10). In comparison frequency of MAR (15). with the results of our study, all known indica- Reported twins presentation during labour is ro- tions for CS in one Nigerian study were found ughly similar to the foetal presentation and lie in lower prevalence (10). These differences, in our study (10). Shorter gestation is associated respectively, the higher frequency of CS and with newborns with lower birth weight, more the CS indication can be explained by populati- amniotic fluid, and a greater chance of change in on differences: poor obstetric practice, as well the position and presentation (16). as a decline in obstetric skills, causes more and The actual incidence of the IUGR is unknown, more pregnancies, especially, twin pregnanci- due to the use of growth curves for singleton es to be delivered by CS (19). There were no pregnancies. There are growth curves for twin reports of maternal death in a Nigerian study pregnancies as well, but most researchers belie- among twin bearing mothers, and the results ve they are result of small population data wit- correlate with ours (10). hout considering the impact of the chorionicity The strength of this study is a large number of and outcome of the pregnancy (2,8). The IUGR analysed twin pregnancies. Limitations are the found in our study is similar to Alrahmani et al. retrospective nature of the study and the absence study for adolescent and adult group (18). In our of data about amnionicity and chorionicity (be- study, foetal anomalies occurred in 0.82%, which cause primary care obstetricians in our country correlates with other studies (2). do not record such data during the first trimester Newborns from twin pregnancies have lower ultrasound examination). However, lack of those birth length and weight compared to singleton data led to fail in stratification of twin pregnan- pregnancies of the same gestational age, which cies for the differences in amnionicity and hori- has a significant influence on perinatal outcome onicity. Furthermore, the data concerning clear (2,10). The neonatal outcome of the second twin information about population characteristics like is usually more complicated due to abnormaliti- ethnicity, educational status, socioeconomic con- es of presentation and lie and asphyxia associa- dition, were restricted. ted with operative manipulation during delivery, In conclusion, medically assisted reproduction which is demonstrated in our study, as well (10). does not significantly increase the incidence of Israeli study reported lower average birth weight twin pregnancies in our country. Maternal co- in comparison to our results in both, the first and morbidities and obstetric complications, thou- second twin (9). Aisien et al. reported a higher gh more frequent, do not have a significantly incidence of low birth weight newborns than in higher prevalence in preterm births in twin pre- our study probably because of differences among gnancies. As expected, preterm delivery is the study populations (10). Israeli study reported major cause of neonatal morbidity and admissi- higher frequency of VLBWI in comparison to on to NICU in twin pregnancies. Cornerstone of our study, probably because of the difference in twin pregnancy antenatal care is to get correct age (advanced maternal age in Israeli group) (9). data about amnionicity and chorionicity. Since Older women have a significantly higher frequ- majority of prenatal data did not have this in- ency of delivery with VLBWI (3). formation we call upon all obstetricians to dec-

470 Cerovac et al. Twin pregnancies

lare about amnionicity and chorionicity in twin FUNDING pregnancies during the first trimester ultraso- No specific funding was received for this study. und examination. These data are crucial for the future antenatal planning. TRANSPARENCY DECLARATION Conflict of interest: None to declare.

REFERENCES 1. Di Renzo GC, Fonseca E, Gratacos E, Hassan S, 9. Avnon T, Ovental A, Many A. Twin versus singleton Kurtser M, Malone F, Nambiar S, Nicolaides K, pregnancy in women ≥ 45 years of age: comparison Sierra N, Yang H, Fuchtner C, Berghella V, Caste- of maternal and neonatal outcomes. J Matern Fetal lazo Morales E, Hanson M, Hod M, Ville Y, Vis- Neonatal Med 2019; 18:1-6. ser G, Simpson JL, Adra A, Bataeva R, Roura LC, 10. Aisien AO, Olarewaju RS, Imade GE. Twins in Jos Chmait RH, Cheng Y, Hyett J, Khalil A, Malhotra Nigeria: a seven-year retrospective study. Med Sci N, Giardina I, Morrison J, Nazareth A, Poon CYL, Monit 2000; 6:945-50. Purandare CN, Quintero R, Sepulveda W, Tosto V. 11. Cerovac A, Grgic G, Ljuca Dž. Mode of delivery in FIGO Working Group on Good Clinical Practice in preterm births - Bosnian and Herzegovinian experi- Maternal-Fetal Medicine. Good clinical practice ad- ence. Mater Sociomed 2018; 30:290-3. vice: management of twin pregnancy. Int J Gynaecol 12. Schaaf JM, Mol BW, Abu-Hanna A, Ravelli AC. Obstet 2019; 144:330-7. Trends in preterm birth: singleton and multiple 2. Škrablin S. Višeplodne trudnoće (Multiple pregnan- pregnancies in the Netherlands, 2000-2007. BJOG cies). In: Šimunić et al. Reprodukcijska endokrino- 2011; 118:1196-204. logija i neplodnost (Reproductive endocrinology and 13. Gardner MO, Goldenberg RL, Cliver SP, Tucker JM, infertility) [in Croatian]. Zagreb: Školska knjiga, Nelson KG, Copper RL. The origin and outcome of 2012: 637-56. preterm twin pregnancies. Obstet Gynecol 1995; 3. Prapas N, Kalogiannidis I, Prapas I, Xiromeritis P, 85:553-7. Karagiannidis A, Makedos G. Twin gestation in ol- 14. Sibai BM, Hauth J, Caritis S, Lindheimer MD, der women: antepartum, intrapartum complications, MacPherson C, Klebanoff M, VanDorsten JP, Lan- and perinatal outcomes. Arch Gynecol Obstet 2006; don M, Miodovnik M, Paul R, Meis P, Thurnau G, 273:293-7. Dombrowski M, Roberts J, McNellis D. Hypertensi- 4. Malamitsi-Puchner A, Voulgaris K , Sdona E, Chri- ve disorders in twin versus singleton gestations. Am stou C, Briana DD. Twins and socioeconomic fac- J Obstet Gynecol 2000; 182:938-42. tors: changes in the last 20 years. J Matern Fetal 15. Kazandi M, Turan V. Multiple pregnancies and the- Neonatal Med 2019; 32:455-60. ir complications. J Turk Soc Obstet Gynecol 2011; 5. Taylor MJ. The management of multiple pregnancy. 8:21-4. Early Hum Dev 2006; 82:365-70. 16. Škrablin S. Višeplodne trudnoće (Multiple pregnan- 6. Ayres A , Johnson TR. Management of multiple pre- cies). In: Đelmiš J., Orešković S. et al. Fetalna me- gnancy: labor and delivery. Obstet Gynecol Surv dicina i opstetrcija (Fetal medicine and obstetrics) 2005; 60:550-4. [in Croatian]. Zagreb: Medicinska naklada, 2014: 7. Dagenais C, Lewis-Mikhael AM, Grabovac M, 371-84. Mukerji A, McDonald SD. What is the safest mode 17. Murray SR, Stock SJ, Cowan S, Cooper ES, Norman of delivery for extremely preterm cephalic/non- JE. Spontaneous preterm birth prevention in multi- cephalic twin pairs? A systematic review and meta- ple pregnancy. Obstet Gynaecol 2018; 20:57–63. analyses. BMC Pregnancy Childbirth 2017; 17:397- 18. Alrahmani L, Abdelsattar ZM, Adekola H, Gonik B, 409. Awonuga A. Adolescence and risk of preterm birth 8. Norwitz ER , Edusa V, Park JS. Maternal physiology in multifetal gestations. J Matern Fetal Neonatal and complications of multiple pregnancy. Semin Pe- Med 2019; 32:1321-4. rinatol 2005; 29:338-48. 19. Cerovac A, Ljuca Dž, Bogdanović G, Grgić G, Zu- kić H. Pregnancy outcome in women who survived genital or extragenital cancer. Med Glas (Zenica) 2020; 17:270-7.

471 ORIGINAL ARTICLE

A therapeutic effect of Nigella sativa extract on female Wistar rats vulvovaginal candidiasis model

Muhammad Rusda1, Muhammad Fidel Ganis Siregar1, Aznan Lelo2, Syafruddin Ilyas3, Ratna Akbari Ganie4, Yusuf Effendi5, Poppy Anjelisa Zaitun Hasibuan6, Raden Roro Lia Iswara7

1Obstetrics and Gynaecology Department, 2Pharmacology Department; School of Medicine, 3Biology Department, Faculty of Mathemat- ics and Natural Sciences, 4Clinical Pathology Department, School of Medicine; Universitas Sumatera Utara, 5Obstetrics and Gynaecol- ogy Department, School of Medicine, Universitas Sriwijaya, 6Faculty of Pharmacy, 7Microbiology Department, School of Medicine, Universitas Sumatera Utara; Indonesia

ABSTRACT

Aim Vulvovaginal candidiasis (VVC) is a disease mostly caused by Candida albicans and affects the quality of life of women especially in the form of chronic recurrent vulvovaginal candidiasis (RVVC). Nigella sativa is known to have several effects such as antimicrobial, anti-inflammatory, immune stimulation and anticancer properties. The aim of this study was to evaluate the effect of Nigella sativa on vulvovaginal candidiasis.

Methods This study is a true experimental design, we used 28 Wistar strain rats divided into 4 groups, all groups were conditio- ned in a pseudoestrus state. Candida albicans was inoculated into Corresponding author: the rats’ vagina for 3 consecutive days. All groups were observed Muhammad Rusda every 24 hours, 48 hours and 72 hours to evaluate the number of Obstetrics and Gynaecology Department, Candida albicans colonies, IgG and IgM anti Candida. School of Medicine, Universitas Sumatera Utara Results After administration of intervention, there was a significant difference in the amount of fungal colonization after the Jl. Dr. Mansyur No.5 Medan, treatment in each group (p<0.001). There was no significant diffe- North Sumatera, Indonesia, 20155 rence in the effectiveness of Nigella sativa and fluconazole in su- Phone: +62 811 616 694; pressing Candida albicans colonies after 72 hours (p=0.101). The Fax: +62 61 8216 264; administration of Nigella sativa showed a significant difference in Email: [email protected] the increase in IgM levels compared to the others group (p<0.001), ORCID ID: https://orcid.org/0000-0002- while the IgG level did not show a significant difference (p=0.423). 2268-6838 Conclusion Nigella sativa provides a therapeutic effect by decreasing the number of fungal colonies and increasing IgM levels. Original submission: Key words: Candida albicans, fungal inoculation, pseudoestrus 16 December 2019; Revised submission: 13 January 2020; Accepted: 27 January 2020 doi: 10.17392/1112-20

Med Glas (Zenica) 2020; 17(2): 472-476

472 Rusda et al. Therapeutic effect of Nigella sativa

INTRODUCTION of this study was to compare Candida albicans colony count, anti-Candida IgG and IgM, before Among human diseases caused by Candida al- and after the treatment of vulvovaginal candidia- bicans, vulvovaginal candidiasis (VVC), espe- sis by Nigella sativa. cially in chronic and recurrent form (RVVC), is arguably the most frequent candidiasis bacterial MATERIALS AND METHODS infection of humans. Epidemiological investigations provide a global estimate of the incidence Materials and study design of RVVC approaching 2% (1). It is important to deal with this disease immediately because of This was experimental study to assess the effect complications that can be caused. Complications of Nigella sativa extract on vulvovaginal candi- that often occur in this disease, if it is not handled diasis. This study was held between February and properly, are infertility in sexually active women October 2019 in four institutions: Multidicipline due to pelvic inflammatory disease and chorio- Laboratory of Medical Faculty Universitas Su- amnionitis that can cause abortion or premature matera Utara (USU) to assess anti-Candida IgG birth in pregnant women (2). and IgM; Microbiology Laboratory of Universi- tas Sumatera Utara General Hospital; Faculty of Nigella sativa Linn is a member of the Ranun- Mathematics and Natural Sciences’ Animal Ho- culaceae family, which grows in Asia and the use of Universitas Sumatera Utara for acclimati- Middle East and has been used as food and zation, intervention and blood sample collection natural medicine. The ingredients contained of the animal; and Pharmacology Laboratory of by Nigella sativa are thymoquinone, thymo- Faculty of Pharmacy Universitas Sumatera Utara hydroquinone, dithymoquinone, and thymol. to produce the extract of Nigella sativa. The inflammatory process balances with lipo-oxygenizes and thymoquinone inhibits the The sample included 28 female Wistar rats oxygen cycle. Several pharmacological effects (Rattus norvegicus sp.) aged 2-3 months with a of Nigella sativa including antimicrobial, anti- range of body weight between 101-241 grams, inflammatory, immune stimulation, and anti- which met the inclusion and exclusion criteria. cancer properties have been reported (3). Before inoculation of Candida albicans, estradiol valerate 2 mg was given subcutaneously 3 Basically, this plant has been used in inflammati- days before inoculation and 4 days after inocu- on treatment since ancient times. Appearance of lation to make the rat in the condition of pseu- synthetic anti-fungal agents in the middle of the doestrus to maintain the Candida albicans and last century led to a reduced interest in usage of prevent self-healing of the rat. plants as a natural source of anti-fungal drugs. In recent years the situation has changed, and the Methods field of ethnobotanical research has evolved (4). The Nigella sativa extract was produced by Thymoquinone on Nigella sativa showed a selec- Pharmacology Laboratory using ethanol 96% tive anti-bacterial effect on seven bacteria, espe- solvent and sodium carboxymethyl cellulose cially Gram-positive strains with a low minimal (CMC Na) in order to obtain the extract of Ni- inhibitory concentration (MIC). Thymoquino- gella sativa 5 mg/mL. ne showed a potent growth inhibitory effect on The samples were divided into 4 groups: K1-7 Gram-positive bacteria, with MICs ranging from rats as the control group, K2-7 rats were given a 8 to 64 μg/mL (5). fluconazole 10 mg/kg of body weight, K3-7 rats Zataria multiflora extract on Candida albicans were given Nigella sativa extract with the dose has shown the MIC and minimum fungicidal con- of 6.6 mL/kg of body weight, K4-7 rats were gi- centration (MFC) value (50% and 90%) of 0.13, ven a combination of Nigella sativa extract 6.6 0.38, 0.74, and 1.03 mg/mL, respectively (6). mL/kg of body weight and fluconazole 10 mg/kg Research to assess the MIC and MFC of Nigella of body weight. All treatments were given orally sativa in Candida albicans has been conducted with an oral-gastric tube. Twenty-four hours after in vitro (6,7), but no in vivo studies have been fo- inoculation, we gave the therapy for three days. und investigating vulvovaginitis in rats. The aim The propagation of Candida albicans was held in

473 Medicinski Glasnik, Volume 17, Number 2, August 2020

the Microbiology Laboratory. Candida albicans RESULTS (ATCC 14053, USA) inoculated on Sabouraud There were differences in the average of colony Chloramphenicol 2 Agar Media/ SAB CHL 2-D forming units based on the time of observation (Biomeruex, L’etoile, France) with 4 quadrants (p<0.001). The number of colonies sharply incre- method and after 48 hours incubation in O incu- 2 ased on all groups after inoculation with the Can- bator at 25 ºC, Candida albicans suspension (3 dida albicans. The day after the treatment, the K2 McFarland) was prepared from grown colonies. group and K4 group showed the greatest decrea- The colonies was identified macroscopically se in the CFU and even showed a mean below the by observing the shape by Gram-staining, and mean before inoculation. On the third day after confirmed with MALDI-TOF mass spectro- the treatment, it appeared that the K2 group, K3 metry and with Vitex 2 Compact (Biomeruex, group, and K4 group showed the mean number L’etoile, France) in Microbiology Laboratory of colonies below the number of colonies before of USU Hospital. inoculation (Table 1). Identification began by preparing a fungal Table 1. The differences of Candida albicans colony forming suspension of 3 Mc-Farland turbidity using units (CFU) based on the observation of each group a nephelometer (DensiCHECK, Biomeruex, Candida albicans mean (SD) of CFU Group Pre inocu- Post inocu- p L’etoile, France). Then the tubes containing fun- Day-1 Day-2 Day-3 gal suspension were arranged in a cassette and lation lation 0.14 237.86 229.29 110.29 130.43 K1 <0.001 inserted into YST ID identification (Biomeruex, (0.38) (106.46) (120.89) (131.48) (117.76) L’etoile, France) and AST-YS08 antifungal sen- 10 136.86 4.29 0.14 0.43 K2 <0.001 sitivity test (Biomereux, L’etoile, France) cards. (18.68) (115.77) (1.11) (0.38) (1.13) 2.14 164.57 88.29 6.43 1.14 K3 <0.001 Barcode was given on each card to connect with (2.48) (124.69) (99.63) (5.29) (0.9) the identity of the isolates that had been entered, 0.57 210 2.86 0.57 0.14 K4 <0.001 then inserted into filler room and proceeded to (0.79) (95.67) (1.86) (1.13) (0.38) K1, control group; K2, fluconazole group; K3,Nigella sativa group; the next room. The Vitek 2 Compact machine K4, fluconazole + Nigella sativa group automatically identifies and gives antifungal sensitivity result. When observing each group, almost all groups showed an increase of IgM, but in the K1 gro- Inoculation of the rat’s vagina was done by swab- up (control) the mean of IgM level decreased on bing the vagina with a cotton swab dipped in the the 7th day from 769.88 pg/mL to 547.72 pg/mL, fungal suspension. The colonies (colony forming then on the 14th day a sharp increase from 547.72 unites, CFU) of Candida albicans on the rat’s pg/mL to 1094.14 pg/mL. There were significant vagina were measured before inoculation, 24, mean differences in each group for IgM level 48 and 72 hours after. The rats underwent taking (p<0.05) (Table 2). vaginal smear to take samples, which were incu- Table 2. IgM level changes according to inoculation time bated for 48 hours at 25 ºC. period The IgM and IgG level were measured by ELISA Mean (SD) IgM level (pg/mL) Group p (Enzym Linked Immunosorbent Assay) method Pre-Inoculation Day-7 Day-14 before the inoculation, 7 days and 14 days after K1 769.88 (191.91) 547.72 (130.46) 1094.14 (136.24) 0.006 K2 682.03 (178.09) 866.79 (418.23) 1795.39 (318.01) 0.018 the intervention. The blood sample was taken K3 982.57 (291.49) 1000.81 (660.05) 2295.23 (1165.42) 0.018 with 1mL disposable syringe from rat’s tail as K4 448.97 (113.1) 1026.95 (843.94) 1106.49 (338.18) 0.028 much 0.25 mL. K1, control group; K2, fluconazole group; K3,Nigella sativa group; K4, fluconazole + Nigella sativa group Statistical analysis There were significant difference in the mean The data were analysed using Kruskal-Wallis and value of IgG between the K1 (3553.67 pg/mL), Mann-Whitney tests to assess differences of the K2 (3879.48 pg/mL) and K4 (4215.15 pg/mL) colonies forming units (CFU) of Candida albi- group based on the time of observation (p<0.05); cans, and Friedman test to assess the differences the K3 (4021.07 pg/mL) group did not show si- of IgM and IgG level. gnificant difference in IgG level in relation to the time of observation (p=0.058) (Table 3).

474 Rusda et al. Therapeutic effect of Nigella sativa

Table 3. IgG level changes according to inoculation time period against vaginitis. In fact, some women have ele- Mean (SD) IgG level (pg/mL) Group p vated levels of IgA and IgG specific candida in Pre inoculation Day-7 Day-14 serum or vaginal secretions (9). In rats, candida- 3588.05 3553.67 2209.72 K1 <0.001 (1037.50) (1139.07) (971.53) specific IgM and IgG antibodies have shown to 4005.81 3879.48 3013.53 be protective against systemic and vaginal candi- K2 0.001 (974.02) (637.57) (918.45) da infections (10). In addition, in a mouse model 4031.89 4021.07 3765.14 K3 0.058 (570.05) (598.33) (614.55) of vaginal candidiasis, specific infection-induced 4373.40 4215.15 2795.78 aspartyl proteinase IgA antibodies contribute to K4 0.027 (492.20) (426.89) (835.41) protection against infection. Rats with primary K1, control group; K2, fluconazole group; K3,Nigella sativa group; vaginal infection showed resistance to vaginal K4, fluconazole + Nigella sativa group rinse containing IgA antibodies directed against the secretion of aspartyl proteinase Candida al- DISCUSSION bicans (11). These antibodies are able to provide Creating, supporting and maintaining pathogens protection to recipients who have not been infec- for vaginal candidiasis rat model need a special ted (12). treat. The vaginal candidiasis condition in rats Prince et al. stated that detection of candida will quickly recover, so that the infection will IgG, IgM, and IgA by ELISA method showed heal quickly without creating a pseudoestrus situ- excellent performance, as indicated by accep- ation. The estradiol valerate 2-5 mg was admini- table linearity, low intra-assay and inter-assay stered subcutaneously 3 days before inoculation coefficient of variation (CV) values, and low and 4 days after to decrease rat immunity or crea- specimen stability. Strong inter-laboratory repro- ting a pseudoestrus situation. If needed, it can be ducibility, as demonstrated by large panel sera repeated weekly (7). testing at two different locations, was also very The results of this study showed an increase of good, which confirms the strong nature of this Candida albicans CFU in all rat groups after ELISA. Seroprevalence level in a positive sera inoculation. IgM level showed an increase in all panel for antibodies to other fungal pathogens the study groups with an exception of group K1 are similar to those observed in healthy individu- (control) in which a decrease in IgM mean level als, suggesting that cross-reactivity due to other on the 7th day after the treatment was found, and fungal infections continues to show consistent re- on the 14th day a sharp increase was noticed. sults (10,13). The number of fungal pathogens in secondary infection rats was less than in primary There were significant mean differences in grou- infections (a secondary infection is an infection ps K2, K4 and K1 for IgG level based on inocula- that occurs during or after treatment for another tion time. This is consistent with a trout fish study infection, it may be caused by the first treatment from Awad, et al., in which an increase of IgM le- or by changes in the immune system), and this is vel was found after Nigella sativa administration related to a higher level of anti-candida antibodi- compared to the control group; they also found es in secondary infections compared to primary an increase in lysozyme, the amount of serum infections (11). protein, antiprotease and antibacterial activity in the group that received Nigella sativa (8). In conclusion, Nigella sativa has therapeutic effect that reduces the number of Candida al- The role of humoral immunity against vaginitis is bicans colonies and increases the level of IgM. not very clear. In patients with recurrent vulvova- There is a need for further studies to asses the ginal candidiasis (RVVC), elevated levels of IgA effect of Nigella sativa in the treatment of human and specific IgG in serum or vaginal secretions vulvovaginal candidiasis. show little or no protective role for antibodies

475 Medicinski Glasnik, Volume 17, Number 2, August 2020

REFERENCES

1. Cassone A, Sobel JD. Experimental models of vagi- 8. Awad E, Austin D, Lyndon AR. Effect of black nal candidiasis and their relevance to human candi- cumin seed oil (Nigella sativa) and nettle extract diasis Infect Immun 2016; 84:1255–61. (Quercetin) on enhancement of immunity in rainbow 2. Sahoo S, Mohanty I, Parida B, Patnaik S. Prevalence trout, Oncorhynchus mykiss (Walbaum). Aquacultu- of vulvovaginal candidiasis in sexually active fema- re 2013; 388:193-7. les with antifungal susceptibility patterns of the iso- 9. Cleff MB, Wendisch I, Cabana, Angela L, Rodrigues lates. Int J Biomed Res 2018; 4:38. MR, Mello JRB, Meireles MCA, Hernandez-Escare- 3. Fard FA, Zahrani ST, Bagheban AA, Mojab F. no JJ. Experimental vaginal candidiasis: assessment Therapeutic effects of Nigella sativa Linn (Black of Origanum vulgare for its treatment. Afr J Micro- Cumin) on Candida albicans vaginitis. Arc Clin In- biol 2011; 5:420711. fect 2015; 10:1–5. 10. Prince HE. Evaluation of enzyme-linked immuno- 4. Aljabre SHM. Antidermatophyte activity of ether sorbent assays for detecting circulating antibodies to extract of Nigella sativa and its active principle, Candida albicans. J Clin Lab Anal 2008; 22:234–8. thymoquinone. J. Ethnopharmacol 2006; 101:116–9. 11. Karen L, Floyd L, Wormley Jr, Paul LF Jr. Candida- 5. Chaieb K, Kouidhi B, Jrah H, Mahdouani K, Bak- specific antibodies during experimental vaginal can- hrouf A. Antibacterial activity of thymoquinone, an didiasis in Mice. Infect Immun 2002; 5790–9. active principle of Nigella sativa and its potency 12. De Bernardis F, G Santoni, M Boccanera, E Spreghi- to prevent bacterial biofilm formation. BMC 2011; ni, D Adriana, L Morelli, Cassone A. Local antican- 29:1–6. didal immune responses in a rat model of vaginal in- 6. Moghim H, Taghipoor S, Shahinfard N, Kheiri S, fection by and protection against Candida albicans. Panahi R. Antifungal effects of Zataria multiflora Infect Immun 2000; 68:3297–304. and Nigella sativa extracts against Candida albi- 13. Fidel PL, Lynch ME, Tait L, Sobel JD. Mice immu- cans. J Herb Med Pharmacol 2015; 4:138–41. nized by primary vaginal Candida albicans infection 7. Calderon L, Williams L, Martunez M, Clemons KV, develop acquired vaginal mucosal immunity. Infect Stevans DA. Genetic susceptibility to vaginal candi- Immun 1995; 63:547–53. diasis. Medical Mycology 2013; 41:143-7.

476 ORIGINAL ARTICLE

Endovascular or open surgical treatment of high-risk patients with infrainguinal peripheral arterial disease and critical limb ischemia Dragan Totić1, Vesna Ðurović Sarajlić1, Haris Vranić1, Amel Hadžimehmedagić1, Nedžad Rustempašić1, Muhamed Djedović1, Haris Vukas2, Alen Ahmetašević1

1Clinic for Cardiovascular Surgery, University Clinical Centre, Sarajevo, 2Department of Surgery, Cantonal Hospital, Zenica; Bosnia and Herzegovina

ABSTRACT

Aim To determine preferable type of treatment in our clinical circumstances by following two groups of patients with critical limb ischemia (CLI), who were treated endovascularly and surgically.

Methods Research was carried out in the form of a prospective study of 80 patients with CLI and Trans-Atlantic Inter-Society Consensus (TASC) C or D type of arterial disease, with Ame- rican Society of Anesthesiology (ASA) class III risk, who were Corresponding author: randomly divided in two groups as per the treatment they received, surgical and endovascular. Patients were followed during 28 Dragan Totić months using clinical examination and Duplex Ultrasound (DUS) Clinic for Cardiovascular surgery, in accordance with prescheduled control visits. University Clinical Centre Bolnička 25, 71000 Sarajevo, Results There was a statistical difference between surgical and endovascular group in two years patency (82.5% vs. 55%; p=0.022) Bosnia and Herzegovina but it did not result in the difference in amputation free survival Phone/fax: +387 33 29 77 05: (AFS) (95% vs. 85%; p=0.171) or two-year freedom from major E-mai: [email protected] adverse limb events (MALE) (87.5 vs. 77.5; p=0.254). Also, the- ORCID ID: https://orcid.org/0000–0002- re was no difference in the overall survival of patients (100% vs. 8929-5428 97.5%; p=0.317). Conclusion Initial endovascular treatment is a preferred form of the treatment for selected patient population.

Key words: amputation-free survival, endovascular procedure, Original submission: infrainguinal bypass, patient survival critical limb ischemia, endovascular, peripheral arterial disease, surgical 13 February 2020; Revised submission: 21 April 2020; Accepted: 18 June 2020 doi: 10.17392/1143-20

Med Glas (Zenica) 2020; 17(2): 477-484

477 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION The aim of this study was to investigate for a preferred treatment, endovascular or surgical, Currently, the treatment of critical limb ischemia of high-risk patients with CLI and infrainguinal is complicated by the development of endovas- peripheral arterial disease (PAD) and limited life cular procedures (1-4). While in the recent past expectancy, as well as to validate our endovascu- the main problem was presence of conditions su- lar/surgical skills. itable for open surgery or ability of the patient to withstand that treatment, today the main problem PATIENTS AND METHODS is how to treat the patient – by endovascular procedure or by open surgery (5-7). Endovascular Patients and study design procedures involve lesser invasiveness for decreased patency (6-8). On the other side, bypass This prospective clinical study was conducted procedures involve greater invasiveness and risk from 2012 until 2016 on 80 consecutive pati- to the patient for prolonged patency (7-9). ents admitted to the Clinic for Vascular Surgery of Clinical Centre of the University of Sarajevo The dilemma is particularly pronounced in sub- (CCUS) that suffered from CLI. population of patients with critical limb ischemia (CLI) and serious comorbidities that make these Inclusion criteria were: patients with critical limb patients poor candidates for open surgical proce- ischemia, Trans-Atlantic Inter-Society Consensus dures (1, 6-9). Endovascular treatment that losses (TASC) II B or C disease, and American Society its effect after a couple of months, puts a patient of Anesthesiology Class III surgical risk (4,16). in even more risk without the resolution of ische- During the study period, precisely in 2014, there mia (10,11). Repeat endovascular procedures often was a revision of TASC II classification but we do not produce wanted results (with the assumpti- remained adhered to 2007 TASC II classification on that health care system could afford it) and, if as per the original design of the study (17). we exclude amputation as a form of preferred tre- Exclusion criteria were: lost to follow-up, patho- atment, the patient is left with surgical intervention logy of the extremity not related to the interven- as the only option after a lot of wasted time (11,12). tion or occlusion of the bypass (trauma, infection The patients with CLI are not good candidates for or deep vein thrombosis), disease of the patient both treatments at the same time, as shown in the that could cause disturbance of the coagulation BASIL study (1). However, there were attempts system (malignancy). to standardize the treatment such as the one in The patients were divided into two groups of 40: a study by Goodney et al. (2) on a validation of patients with infrainguinal revascularization by the society for vascular surgery’s objective per- endovascular intervention (ET group), and pati- formance goals (OPGs) for critical limb ische- ents with infrainguinal revascularization by vein mia, or in a study by Conte et al. (3) on suggested bypass (BX group). objective performance goals in catheter-based The sample was calculated based on knowledge treatment of critical limb ischemia (2,3). of the population related to the Clinic for Vascular Anyway, at the time we designed and started this Surgery Sarajevo (600,000 inhabitants), prevalen- study, literature review did not show without do- ce of CLI in the local population (6%), confidence ubt a preferred treatment for high risk patients interval (95%) and 5% of margin of error. We tried with CLI, having ischemic ulceration that needed to randomize patients to different groups by the more than 6 months for healing (1-15). Additio- time of their admission to hospital. Consecutive nally, there is another very important dilemma. and eligible patients were intermittently assigned Is it correct for centres with low annual number to endovascular intervention or to surgical bypass of operations and with limited resources to base and that schedule was followed as long as techni- their decision making on guidelines developed in cal and personal resources allowed us doing so. In centres with high annual number of operations (e. cases of insufficient equipment for endovascular g. referral centres)? Also, gaining sufficient en- intervention, or lack of staff, we performed sur- dovascular skills is complicated by progressive gical revascularization and vice versa – number development of technology – the problem enco- of patients treated in this manner was reimbursed untered by most developed countries (5,7,12,13). once the equipment and staff were sufficient.

478 Totić et al. Endovascular or surgical treatment of limb ischemia

All participants signed a written consent upon from the hospital. A few days before the dischar- admission to the study. The study was approved ge, patents were given aspirin (100 mg/day) as by the Ethics Committee of the CCUS based on well, and they remained on aspirin for lifetime its compliance with the Helsinki Declaration. (150 mg/day). Patients were controlled on prescheduled visits Methods which were: thirty days post discharge, three Endovascular intervention considered angioplasty months post discharge, six months post dischar- with or without stenting. Stenting was reserved for ge for three consecutive visits, and then yearly. cases with complications such as dissection or rup- We examined patency of bypass or intervention ture of blood vessel or when international guidelines using clinical examination and DUS. The study recommended usage of stents (superficial femoral was designed to follow up patients for at least artery). Usage of popliteal stents was limited by lack two years. Patients that did not show up on con- of specialized stents for that delicate region. Acce- trol visits were contacted by phone and those that ss artery was most frequently common femoral ar- could not be reached in that way were considered tery. Immediately upon access we administered 100 lost to follow up. IU/kg of heparin with the aim of Activated Partial The recommendation of the Society for Vascular Thromboplastin Time (APTT) of 250 seconds. A Surgery (SVS) is to incorporate all adverse limb desirable result was absence of residual stenosis or events into survival analysis. That includes recu- persistent stenosis of less than 30% on control angi- rrence of ischemic suffering due to the graft occlu- ography and continuity of patency to the foot. Befo- sion or intervention failure and the need for addi- re the intervention, patients received low-molecular- tional bypass or endovascular reinterventions, so weight heparin (LMWH) administered according called MALE (Major Adverse Limb Event) (2,3). to the body weight. From the moment of the intervention, patients were kept on conventional heparin Statistical analysis infusion for 24 h with ttt aim of maintaining APTT Demographic data and other characteristic of the two to three times of the normal. After 24 h, during patients were expressed as nominal or categori- the first postintervention month, patients were kept cal variables. Student T test was used for com- on dual antiplatelet therapy, acetyl salicylic acid parison of groups where applicable or adequate (ASA) 100 mg 1x1 and Clopidogrel 75 mg 1x1. nonparametric test. In case of skewed distribu- Starting with the second postintervention month all tion, Mann-Whitney test was used. Pearson χ2 patients received lifelong ASA 150 mg 1x1. test with Yates correction for small samples was Great saphenous vein (GSV) was checked and used for categorical variables. mapped by duplex ultrasound (DUS) to all sur- Patency, amputation free survival and overall surgical patients the day before surgery. Bypass was vival were calculated using Kaplan-Meier analysis performed using the technique of reversed gre- and were graphically expressed by Kaplan-Meier at saphenous month (GSV). Only veins greater curve. Log rank, Breslow (generalized Wilcoxon) than 3 mm in diameter, perfused without resi- or Tarone-Ware analysis were used for measuring stance and without fibrotic segments were used. statistical significance. Binary logistic regression We preferred tunelization within vascular bed. was used to show influence of different input va- Only in reoperations we performed subcutaneous riables on likelihood of negative outcome (occlu- tunnels. Surgery was performed with the aim to sion or amputation). In all tests p<0.05 was used as bypass all occluded segments to achieve conti- statistically significant. nuous patency until the foot. The most frequent inflow site was the common fe- RESULTS moral artery (CFA). Only in cases dictated by spe- A total of 80 patients were enrolled in the study cific anatomic distribution of the disease or insuffi- from during the period 2012 to 2014 (plus two cient GSV length, superficial femoral artery (SFA) years of follow up). The average age was 66 ye- or popliteal artery (PA) was used as the inflow site. ars; 58 (73%) patients were males; 43 (53%) pa- Patients were given LMWH preoperatively and tients were diabetics and 30 (38%) were treated that regime was held by the time of discharge for tissue loss (Table 1). The ET and BX groups

479 Medicinski Glasnik, Volume 17, Number 2, August 2020

were a good match. Significant differences were found in gender distribution e.g. more females in the ET group, 16 (40%) (p=0.012), and in TASC II classification, e.g. TASC II B was more prevalent in the ET group, 24 (60%) and TASC II C in the surgical group, 28 (70%) (p=0.007). Fibular artery was a more prevalent target artery in the ET group, 4 (10%) (p=0.002) (Table 1).

Table 1. Characteristics of the patients with critical limb ischemia (CLI) treated endovascularly and surgically No (%) of patients in the group Characteristic p Endovascular Surgical Age mean (SD) (years) 67 (8.8) 65 (7.1) 0.267 Male 24 (60) 34 (85) 0.012 Diabetes 23 (57.5) 20 (50) 0.501 Elevated cardiac risk 38 (95) 32 (80) 0.043 HTA 34 (85) 29 (72.5) 0.076 Figure 1. Patency of surgical grafts and edovascular inter- Atrial fibrillation 8 (20) 5 (12.5) 0.152 vention; Fem-pop, all infraingvinal bypasses; PTA, percutanous IM in history (CAD) 6 (15) 5 (12.5) 0.632 transluminal angioplasty; Episodes of heart failure 3 (7.5) 2 (5) 0.468 Elevated pulmo risk (COPD) 16 (40) 20 (50) 0.369 Two patients in the BX group had major amputa- Low (FEV ≥80%) 11 (27.5) 17 (42.5) 0.073 tion (above knee) in early stages of the study. In Mild (FEV1 50-79%) 3 (7.5) 2 (5) 0.468 the ET group six patients were amputated; four High (FEV1 3 -49%) 2 (5) 1 (2.5) 0.312 had below knee amputation and two above knee Elevated renal risk (kidney 19 (47.5) 18 (45) 0823 failure) amputation (p=0.150) (Figure 2). Low (creatinine 110 – 160 16 (40) 16 (40%) 1 mmol/L) Mild (creatinine 161 – 240 3 (7.5) 2 (10%) 0.468 mmol/ L) High (creatinine > 241 0 0 mmol/ L) Smoking 34 (85) 36 (90%) 0.499 Diagnosis 26 (65) / 24 (60) / Pain / gangrene 0.501 14 (35) 16 (40) TASC 24 (60/ 12 (30%) / B/C 0.007 16 (40) 28 (70) Target artery of revascularization SFA 18(45) 35 (87.5) 0.001 PA 25 (62.5) 28 (70) 0.306 PTA 9 (22.5) 12 (30) 0.320 ATA 1 (2.5) 3 (7.5) 0.223 FA 4 (10) 1 (2.5) 0.002 FA, fibular artery; superficial femoral artery; PA, popliteal artery; ATA, anterior tibial artery; PTA, posterior ttibial artery; CAD, coronary artery disease; FEV1, forced expiratory volume at 1st sec.; HTA, arterial hypertension; SD, standard deviation; TASC, transatlantic Figure 2. Amputation free survival of the patients treated by sur- inter-society consensus; gical bypass and by endovascular intervention; Fem-pop, all infraingvinal bypasses; PTA, percutanous transluminal angioplasty; Technical success was achieved in 100% of patients. Patency of bypass was 82.5% and patency In relation to the patency analysis, not all occlu- of endovascular intervention was 55% (p=0.013); sions resulted in the recurrence of CLI, especially the difference was not statistically significant in ET. Recurred CLI after failed bypass or en- until the 240th day of the study (or in the first 8 dovascular intervention was always considered months). Seven patients in BX group experien- for revascularization. In the BX group occlusion ced bypass failure and intervention failure affec- of grafts in two patients resulted in recurrence ted 18 patients in the ET group. Most of the failu- of CLI and both patients got additional bypass res in theET group happened in he second half of with contralateral great saphenous vein. Rema- the study period (Figure 1). ining two patients with graft occlusion did not

480 Totić et al. Endovascular or surgical treatment of limb ischemia

develop CLI and did not need reintervention. Binary logistic regression model showed signifi- Additionally, one patient had stenosis of the graft cance for amputation free survival [χ2 (3, N=80) = close to proximal anastomosis that was resolved 14.510] (p=0.002). Age had significant influence by endovascular intervention (crossover patient). on amputation free survival (AFS) (QP=1.175; There was also one crossover patient in ET and p=0.008). Gender had a trend toward significance two patients with recurrence of ischemia resol- (QP=4.8; p=0.068). ved by endovascular reintervention. Nine pati- The most frequent systemic complication in the ents did not experience recurrent CLI upon the BX group was respiratory failure, in two (5%) intervention failure. In six patients reintervention patients, followed by myocardial infarction (1%) was not possible and they eventually ended up and heart failure (1%). The most frequent local with amputation. complication in the BX group was lymphedema, Survival without MALE in BX and ET group in 8 (20%) patients, followed by superficial wo- was 87.5% and 77.5% respectfully (p=0.243) und infection (10%), deep wound infection (5%) (Figure 3). and bleeding (2.5%). The most frequent systemic complication in the ET group was renal failure, in two (5%) patients, followed by myocardial infarction (1%). The most frequent local complication in the ET group was access site hematoma, in three (7.5%) patients, followed by pseudoaneurysm, dissection and embolization – each occurred in 2.5% of patients. Median of postoperative hospitalization in the BX group was 10 days and 2 days in the ET group (p=0.0001).

DISCUSSION After 28 months of follow up, patency of bypass was 82.5% and patency of endovascular intervention was 55%. Results of our study are in Figure 3. Survival analysis of the patients without major ad- accordance with recommended range of OPG verse limb events (MALE) treated by surgical bypass and by (Objective Performance Guidelines) (2,3). Also, endovascular intervention; Fem-pop, all infraingvinal bypasses; when compared with patency of endovascular PTA, percutanous transluminal angioplasty; intervention in similar studies, significant diffe- One patient from the ET group died due to massi- rence was not detected (1,2,3, 6-13). ve myocardial infarction at the start of the second Venous bypasses patency in our study was better half of the study period. However, overall survi- or equal to that reported in the literature. In some val of patients between groups was not different. cases, results were significantly better, but detailed Binary logistic regression model showed sta- analysis showed that these studies with inferior tistical significance as a whole2 [χ (5, N=80) = results dealt predominantly with especially vulne- 48.242] (p=0.0001). Three independent variables rable patient subpopulations, such as diabetics, or had unique, significant contribution to model. The patients treated exclusively for tissue loss (3,5,14). strongest predictor of the failure was renal insuf- Again, patency of venous bypasses in this study ficiency (QP=0.044; p=0.0001), followed by dia- was within the recommended range of OPG (2,3). betes (QP=0.09; p=0.0009), and pulmonary insuf- The results of our study found three risk factors ficiency (QP=0.177; p=0.020). When similar was with significant influence on patency. The best applied for other independent variables such as predictor of occlusion was renal insufficiency diagnosis, age and gender, we also got a statisti- (95% greater chances for occlusion than those cally significant model 2 [χ (3, N=80) = 10.569] without renal pathology), followed by diabetes (p=0.014). The strongest predictor of failure was mellitus (91% greater chances for occlusion than gender (QP=2.896) followed by age (QP=1.079). nondiabetics), and pulmonary insufficiency (82%

481 Medicinski Glasnik, Volume 17, Number 2, August 2020

greater chances for occlusion). Negative effects score (4,17). Accordingly, we could assume that of renal failure and diabetes were well recogni- affection of superficial femoral artery in CLI is zed in numerous studies (3,15,20,21). Some of just a marker but not a cause of bad outcome. Our them also found renal insufficiency to be a better study is not alone in this finding. Similar influence predictor of occlusion than diabetes (3,15). of superficial femoral artery on outcome is repor- Regarding demography, best predictor of occlu- ted by Belvins and Beard in their review of forms sion found in our study was female gender – wo- of revascularization of lower extremity (6,14). men had nearly three times greater chances to Two-year survival of patients in this study was experience occlusion than men, followed by age 97.5% and 100% in the ET and the BX group, res- – each year above the average increased chances pectfully; one patient in the ET group died from for occlusion by factor 1.1. There are numerous massive myocardial infarction at the start of the studies showing gender related disparity in the second half of the study period after suffering treatment and outcome of CLI (9,15,18,19). Alt- from occlusion of treated segment and consequent hough other factors could confound inference, we irreparable recurrent ischemia resulting in major can consider female gender as a risk factor per se, amputation one month later. Of course, there is a probably because women were older at presenta- cause and effect of relationship between failure of tion and with more advanced PAD (18,19). treatment and patients’ demise. It looks counte- Amputation free survival (AFS) and limb salvage rintuitive that it happened in the group exposed to (LS) are the most frequent outcome measures in lesser invasive treatment. However, a lot of studies literature (1,2,3,8,10). The AFS is a more realistic with similar design did not find mortality increase indicator of outcome than LS since the latter cen- in the surgical group. For instance, BASIL study sors from analysis patients who died with intact did not find significant difference in survival of extremity (1), and it is the reason why we used AFS patients in surgical and endovascular group (68% and MALE free survival as measures of outcome. vs. 71%, respectively) (1). The results of two-year survival of other studies show similar findings Two-year MALE free survival in this study was (7,9, 13-17). None of the studies reported better better or equal compared to the results of other survival in endovascular group. Even more, Ma- similar studies (2,3,5-9), which also means that saki et al. reported significantly better survival in our result was within the range recommended by the surgical group with constant hazard ratio thro- OPG (2,3). ughout five years of follow up (9). We could relate Four variables were statistically significant pre- this to the fact that higher risk patients were treated dictors of AFS as found in our study: age, gen- in the endovascular group. der (with tendency toward significance), TASC Systemic complications in infrainguinal bypa- “C” class of PAD and superficial femoral artery ss surgery on 7% were frequently reported, of as a target artery of revascularization. Also, it is which, 4.5% was myocardial infarction and heart understandable that more advanced PAD predicts failure and 1.5% cerebrovascular insult (20,21); worse outcome and numerous studies justified the remaining 1% were other systemic complica- that (2,3,15,20). An effect of advanced PAD is tion, such as renal and respiratory failure (20,21). well recognized with special recommendation by In our study 5% of patients suffered from heart OPG for patients with high risk anatomy of the complications (2.5% MI and 2.5% CHF), and disease in the form of TASC C or D lesions (2,3). 5% from respiratory complications, latter being Also, there are special OPG recommendations significantly different from reports in theworld li- for patients older than 80, recognizing age as an terature (21). The most frequently reported syste- influential factor on outcome (2,3). mic complications related to the endovascular Influence of superficial femoral artery on outcome treatment were myocardial infarction (1.2%), he- is interesting and not easy to explain. Superficial art failure (1.5%), cerebrovascular insult (1.4%) femoral artery is most frequently affected by PAD and renal failure (2%), of which 1% needed hae- and its isolated occlusion rarely produces a limb modialysis treatment) (3, 22-25). Seven percent threatening condition (6,7,8,14). However, its is also frequently reported percentage of systemic affection in CLI is usually combined with exten- complications in endovascular interventions (3, sive infrapopliteal lesions, resulting in high TASC 22-25). In our study we found myocardial infarc-

482 Totić et al. Endovascular or surgical treatment of limb ischemia

tion in one patient (2.5%), and renal impairment to wider population (7,10,13). That could be the in two (5%), latter being significantly different reason for serious insufficiency of valid evidence from reference literature (22-25). related to this issue; probably, the most important The most frequent local postoperative compli- limiting factor is human. In that context we could cation in the BX group was superficial wound consider this study as an attempt to offer guide- infection, found in 10% of patients, what was lines for preferred treatment with local technical exactly the percentage reported in the literature and human resources. (20), followed by deep wound infection (5%), This study has several limitations. Most impor- also in accordance with reports in the literature – tantly, the process of randomization was impro- (4.8%) (16). Bleeding requiring intervention was vised due to technical issues. It is very hard to found in one patient (2.5%), significantly diffe- accomplish true randomization in surgical studi- rent than literature reports (0.4%), probably due es performed by several surgeons. Also, it is very to small sample size in our study. Lymphoedema difficult to adhere to strict randomization proto- was present in 20% of patients as reported in the col in the situation of unpredictable equipment literature (2,16,20,21). provision and lack of staff. The most frequent local complication in postin- Additionally, if we were in a position to recruit tervention period in the ET group was access more patients we could expect significant diffe- site hematoma (7.5%), which was significantly rence in the outcome between treatment groups different from the data in the literature (3.5%) (insufficient randomization affects this as well). (2,16,20,21). Postintervention dissection happe- The sample size was actually small, possibly in ned in 2.5% of patients also different from the accordance with sample size calculation. We were literature (0.5%) (2,16,20,21). Access site pseu- forced to this because of slow accrual of patients doaneurysm was found in 2.5% of patients,,,,,, into the study, which might be related to the small while average from the literature was 0.4% target population and narrow inclusion criteria. (2,16,20,21). Prevalence of patients who suffered In conclusion, there was statistically significant from distal embolization was not different from difference in two-year patency of bypass and endo- the literature (2.2%) (2,16,20,21). vascular intervention. However, difference in pa- Ambiguity related to best therapeutic approach tency did not affect amputation free survival. The to patients with CLI has lasted for years. Despite difference remained insignificant even with MALE marked advancement in pharmacologic therapy, it included. Also, overall survival of patients, crosso- remains insufficient in the treatment of CLI (26- ver percentage and morbidity were the same in the 28). However, pharmacologic modification of risk groups. Therefore, endovascular treatment could factors is indisputable and is highly recommended be considered as preferred initial treatment in se- (28). With rapid development of minimally inva- lected patients that match our study population. sive percutaneous technologies, discussion on best FUNDING treatment of patients with CLI, surgical or endovascular, became more complicated (1,6-10,13). Publishing of this work was financially supported Powerful technological progress and improve- by Bosnalijek d.d. Sarajevo, Bosnia and Herze- ments in the equipment on a daily basis complica- govina. te the creation of a study with reliable results that TRANSPARENCY DECLARATION could serve as durable guidelines on the treatment Conflict of interest: None to declare

REFERENCES 1. Bradbury AW, Adam DJ, Bell J, Forbes JF, Fowkes 2. Goodney PP, SchanzerA, DeMartino RD, Nolan, FG, Gillespie I, Ruckley CV, Raab GM; BASIL Tri- DW, Nathanael D, Hevelone ND. Validation of al Participants. Bypass Versus Angioplasty in Severe the Society for vascular surgery’s Objective Per- Ischaemia of the Leg (BASIL) trial. An intention-to- formance Goals (OPGs) for critical limb ischemia treat analysis of amputation-free and overall survival in everyday vascular surgery practice. J Vasc Surg in patients randomized to a bypass surgery-first or a 2011; 54:100–8. balloon angioplasty-first revascularization strategy. J VascSurg 2010; 51(5 Suppl):5S-17S.

483 Medicinski Glasnik, Volume 17, Number 2, August 2020

3. Conte MS, Geraghty PJ, Bradbury AW. Suggested 17. Jaff MR, White CJ, Hiatt WR, Fowkes GR, Dor- objective performance goals and clinical trial desi- mandy J, Razavi M, Reekers J, Norgren L. An upda- gn for evaluating catheter-based treatment of critical te on methods for revascularization and expansion limb ischemia. J Vasc Surg 2009; 50:1462–73. of the TASC lesion classification to include below- 4. 4. Norgren L, Hiatt WR, Dormandy JA. TASC II the-knee arteries: A supplement to the inter-society Working Group: Inter-Society Consensus for the consensus for the management of peripheral arterial management of peripheral arterial disease (TASC disease (TASC II): The TASC Steering Committee. II). J Vasc Surg 2007; 45:S29-30. Ann Vasc Dis 2015; 8:343-57. 5. Losa S, Ruscazio A, Faglia E, Mantero M, Gabrielli 18. Lejey A, Schaeffer M, Georg Y, Lucereau B, Ro- L. Endovascular and surgical treatment of chronic ussin M, Girsowicz E. Gender related long-term limb ischaemia in diabetics. Ital J VascEndovasc differences after open infrainguinal surgery for cri- Surg 2014; 11:351-9. tical limb ischemia. Eu J Vasc Endovasc Surg 2015; 6. Belvins Jr WA, Schneider PA. Endovascular ma- 50:506-12. nagement of critical limb ischemia. Eur J VascEndo- 19. Jackson EA, Munir K, Schreiber T, Rubin JR, Cuff vasc Surg 2010; 39:756-61. R, Gallagher KA, Henke PK, Gurm HS, Grossman 7. Masaki H, Tabuchi A, YunokiY, Watanabe Y, Mimu- PM. Impact of sex on morbidity and mortality rates ra D, Furukawa H. Bypass vs. endovascular therapy after lower extremity interventions for peripheral of infrapopliteal lesions for critical limb ischemia. arterial disease: observations from the Blue Cross Ann Vasc Dis 2014; 7:227-31. Blue Shield of Michigan cardiovascular consortium. 8. Dosluoglu HH, Lall P, Harris LM, Dryjski ML. J Am Coll Cardiol 2014; 63:2525-30. Long-term limb salvage and survival after endovas- 20. Meltzer AJ, Graham A, Connolly PH, Meltzer EC, cular and open revascularization for critical limb is- Karwowski JK, Bush HL, Schneider DB. The com- chemia after adoption of endovascular-first approach prehensive risk assessment for bypass (CRAB) faci- by vascular surgeons. J Vasc Surg 2012; 56:361-71. litates efficient perioperative risk assessment for pa- 9. Söderström MI, Arvela EM, Korhonen M, Halme- tients with critical limb ischemia. J Vasc Surg 2013; smäki KH, Albäck AN, Biancari F, Lepäntalo MJ, 57:1186-95 Venermo MA. Infrapopliteal percutaneous translu- 21. Patel MR, Conte MS, Cutlip DE, Dib N, Geraghty minal angioplasty versus bypass surgery as first-line P, Gray W. Evaluation and treatment of patients with strategies in critical leg ischemia: a propensity score lower extremity peripheral artery disease: consen- analysis. Ann Surg 2010; 252:765-73. sus definitions from Peripheral Academic Resear- 10. Dosluoglu H, Lall P, Blochle R, Harris LM, Dryjski ch Consortium (PARC). J Am Coll Cardiol 2015; MLJ. Clinical presentation and outcome after failed 65:931-41. infrainguinal endovascular and open revascularizati- 22. Simson EL, Kearns B, Stevenson MD, Cantrell AJ, on in patients with chronic limb ischemia. Vasc Surg Littelwood C and Michaels JA. Enhancement to an- 2013; 58:98-104 gioplasty for peripheral arterial disease: systematic 11. Sandford RM, Bown MJ, Sazers RD, London JN, review, cost-effectiveness assessment and expected Nazlor AR and McCarthy MJ. Is infrainguinal bypa- value of information analysis. Health Technol Asse- ss grafting successful following failed angioplasty? ss 2014; 18:1-252. Eur J Vasc Endovasc Surg 2007; 34:29-34. 23. Zimmermann A, Ludwig U, Eckstein HH. Indicati- 12. Santo VJ, Dargon P, Azarbal AF, Liem TK, Mitchell ons and results of endovascular therapy of critical EL, Landry GJ, Moneta GL. Lower extremity auto- limb ischemia. Radiologe 2016; 21:543-9 logous vein bypass for critical limb ischemia is not 24. Nordanstig J, Smidfelt K, Langenskiöld M, Krag- adversely affected by prior endovascular procedure. sterman B. Nationwide experience of cardio and J Vasc Surg 2014; 60:129-35. cerebrovascular complications during infrainguinal 13. Abu Dabrh AM, Steffen MW, Asi N, Undavalli C, endovascular intervention for peripheral arterial di- Wang Z, Elamin MB, Conte MS, Murad MH. Bypa- sease and acute limb ischaemia. Eur J Vasc Endo- ss surgery versus endovascular interventions in se- vasc Surg 2013; 45:270-4. vere or critical limb ischemia. J Vasc Surg 2016; 25. Banerjee S, Pershwitz G, Sarode K, Mohammad A, 63:244-53. Abu-Fadel MS, BaigMS. Stent and non-stent based 14. Beard JD. Which is the best revascularization for cri- outcomes of infrainguinal peripheral artery interven- tical limb ischemia: Endovascular or open surgery? J tions from the multicenter XLPAD Registry. J Inva- Vasc Surg 2008; Suplement 48:11S-16S. sive Cardiol 2015; 27:14-8. 15. Korhonen M, Biancari F, Soderstrom M, 26. Forster R, Liew A, Bhattacharya V, Shaw J, Stansby ArvelaE,HalmesmakiK, AlbackA. Femoropopliteal G. Gene therapy for peripheral arterial disease. balloon angioplasty vs. bypass surgery for CLI: a Cochrane Database Syst Rev 2018 31;10. propensity score analysis. Eur J Vasc Endovasc Surg 27. Tsigkou V, Siasos G, Rovos K, Tripyla N, Tousou- 2011; 41:378-84. lis D. Peripheral artery disease and antiplatelet tre- 16. Sidi A, Lobato EB, Cohen JA. The American Soci- atment. Curr Opin Pharmacol 2018; 39:43-52. ety of Anesthesiologists' Physical Status: category V 28. Lawall H, Huppert P, Espinola-Klein C, Silke- revisited. J Cli Anesth 2000; 12:328-34 Zemmrich C, Ruemenapf G. German guideline on the diagnosis and treatment of peripheral artery disease - a comprehensive update 2016. Vasa 2017; 46:79-86.

484 ORIGINAL ARTICLE

The effect of Aloe vera ethanol extract on the growth inhibition of Candida albicans Via Karina Nabila1, Imam Budi Putra2

1Faculty of Medicine, 2Department of Dermatology and Venereology, Faculty of Medicine; Universitas Sumatera Utara, Indonesia

ABSTRACT

Aim Candida albicans can cause two major types of infections: superficial infections (such as oral or vaginal candidiasis) as well as life-threatening systemic infections, and Aloe vera extract is one of the potentially useful therapeutic options. The aim of this study was to determine antifungal properties of Aloe vera ethanol extract on vulvovaginal candidiasis caused by Candida albicans.

Methods Aloe vera ethanol extract was prepared by the maceration method with 70% ethanol and dissolved in DMSO into multiple concentrations (6.25%, 12.5%, 25%, and 50%). Candida albicans was cultured in Sabouraud dextrose agar for 72 hours and disc diffusion method was used to evaluate the inhibitory effect of each Corresponding author: concentration in comparison with fluconazole. Zones of inhibition at 72 hours were measured and documented, then analysed to get Imam Budi Putra the mean inhibition zone (MIZ). Department of Dermatology and Venereology, Faculty of Medicine, Results After 72 hours, all concentrations of Aloe vera etha- Universitas Sumatera Utara nol extract showed inhibition effect against C. albicans with mean inhibition zones of each concentration, 12.450±0.208 mm Jl. Dr. Mansur No. 66, (6.25%), 13.975±0.457 mm (12.5%), 15.650±0.420 mm (25%), Universitas Sumatera Utara Hospital, and 17.225±0.512 mm (50%), respectively. Fluconazole revealed Medan 20136, North Sumatera, Indonesia comparable antifungal effect with MIZ of 11.025±0.478 mm. One- Phone: +62 812 6011 965; way ANOVA test showed a significant effect of Aloe vera ethanol E-mail: [email protected] extract on inhibition zone of Candida albicans growth (p<0.005). ORCID ID: http://orcid.org/0000-0003- Conclusion Aloe vera ethanol extract possesses concentration de- 1805-7278 pendent activity against Candida albicans that is comparable with standard antifungal agents.

Original submission: Key words: herbal, therapy, yeast infections 01 November 2019; Revised submission: 20 January 2020; Accepted: 22 June 2020 doi: 10.17392/1098-20

Med Glas (Zenica) 2020; 17(2): 485-489

485 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION been used in folk medicine for over 2000 years and has remained an important component in Candida species is one of the commensal micro- the traditional medicine of many countries. Aloe organisms in humans that can be found on the vera belongs to the Liliaceae family. It is repor- skin, gastrointestinal, genitourinary tracts and in ted that Aloe vera contains over 75 nutrients and the oral and conjunctival space (1). There are se- 200 active compounds including sugar, anthraqu- veral Candida species, such as Candida (C.) al- inones, saponins, vitamins, enzymes, minerals, bicans, C. tropicalis, C. krusei, C. parapsilosis, lignin, salicylic acid and amino acids, and other C. pseudotropicalis, and C. glabrata (1,2). different potentially active compounds including C. albicans is a polymorphic fungus that can simple/complex polysaccharides, phenolic com- grow either as ovoid-shaped budding yeast, as pounds, and organic acids (9). Based on in vitro elongated ellipsoid cells with constrictions at and animal studies, which used total leaf extract, the septa (pseudohyphae) or as parallel-walled Aloe vera exhibits anti-inflammatory, anti-arthri- true hyphae (3). C. albicans can cause two major tic, antibacterial, and hypoglycemic properties types of infections in humans: superficial infecti- (10). Several studies have proven the antifungal ons (such as oral or vaginal candidiasis) and life- properties of Aloe vera extract on the inhibition threatening systemic infections (4). It is estima- of Candida albicans growth (11). However, we ted that approximately 75% of all women suffer have not found any study in the usage of Aloe at least once in their lifetime from vulvovaginal vera on human vulvovaginal candidiasis. candidiasis (VVC), and 40–50% experience at The aim of this study was to determine the anti- least one additional episode of infection (5). A fungal properties of Aloe vera ethanol extract on small percentage of women (5–8%) suffer from vulvovaginal candidiasis species including Can- at least four recurrent VVC per year. dida albicans. Predisposing factors for VVC include diabetes mellitus, use of antibiotics, oral contraception, MATERIAL AND METHODS pregnancy and hormone therapy (6). The clinical signs of VVC are oedema and erythema of the Study design and sample selection vulva and the vagina accompanied by an abnor- This laboratory-controlled prospective study was mal vaginal discharge that may be watery, chee- conducted at the Microbiology Faculty of Uni- se-like, or minimal. The vaginal discharge typi- versitas Sumatera Utara in the period June to Oc- cally resembles cottage cheese (7). tober 2018. Aloe vera extract was obtained from Recently, there has been an increasing challenge Pharmacology Laboratory, Pharmacology Fa- for patients with refractory vulvovaginal candi- culty of Universitas Sumatera Utara. The inclu- diasis caused by azole-resistant Candida species. sion criteria were the number of 0.5 McFarland Fluconazole resistant C. albicans is a growing standard, which consists of 1.5 x 108 colony/mL and perplexing problem following years of indis- of Candida albicans on Sabouraud dextrose Agar criminate drug prescription and unnecessary drug (SDA). The exclusion criteria was the culture exposure and for which there are few therapeutic contaminated with bacteria. alternatives. Regrettably, although the azole class The protocol of this study was approved by the of drugs has expanded, new classes of antifungal Health Research Ethical Committee, Universitas drugs have not been newly registered, so limited Sumatra Utara/H. Adam Malik General Hospital, treatment options exist for patients with azole re- Medan, Indonesia. sistant Candida vaginitis (8). Potential alternative therapies include the use of new active princi- Methods ples, as natural products, that have been active in Preparation of extract. Dried Aloe vera in the vitro. Among the natural products, plants contain powder form as much as 7 g, extracted by using diverse components that are important sources of maceration method with 70% ethanol. The ma- biologically active molecules (1). ceration process was carried out for 3 days with Many plants have been utilized for their medici- occasional stirring. Maceration solution was then nal properties worldwide. Aloe vera species has filtered to obtain filtrate. Maceration filtrate was

486 Nabila et al. Aloe vera as anti-Candida

concentrated to obtain as much as 1.5 gr extract. RESULTS The result of the marinade extract was 21.42% After 72 hours, all concentrations of Aloe vera Aloe vera extract. This thick extract was left alo- ethanol extract showed inhibition activity on ne at room temperature until all the ethanol eva- Candida albicans. Minimum concentration of the porated. Then the extract was stored in a sealed extract to have antifungal activity was 6.25% with bottle before the examination. the mean inhibition zone of 12.450±0.208 mm, In vitro test. The tested fungal isolates were C. followed by 12.5% with the mean inhibition zone albicans obtained from vulvovaginal candidiasis of 13.975±0.457 mm, 25% with the mean inhibi- patient at Microbiology Laboratory of Sumatera tion zone of 15.650±0.420 mm, and 50% with the Utara Universitas. The selected microorganism mean inhibition zone of 17.225±0.512 mm. These was identified and confirmed by conventional results are comparable with or even more superior and biochemical test. Candida albicans strea- than fluconazole (positive control), which showed ked on SDA plates using sterile swabs, previo- the inhibition zone of 11.025±0.478 mm. usly suspended in sterile distilled water (for 15 A significant effect of Aloe vera ethanol extract minutes) and adjusted to 1×106 colony forming on the inhibition zone of the growth of Candida units (CFU/mL) (0.5 McFarland standard) using albicans was found (p<0.005). Each concentra- a nephelometer. tion showed distinct significant results in the Antifungal activity of Aloe vera extracts was inhibition of the growth of Candida albicans evaluated using disc diffusion method. The (Table 1, Figure 1). inoculum size of each clinical isolate was stan- Table 1. Distribution of the inhibition zone of Candida dardized matching a turbidity equivalent to a 0.5 albicans by Aloe vera ethanol extract McFarland standard. A total of 24 cultures were Diameter of inhibition zone (mm) Groups divided into 4 Candida species in 6 test groups, Mean±SD Minimum Maximum as followed: group I (positive control by using Positive control 11.025±0.478 10.50 11.60 Negative control 0.0000 0.00 0.00 fluconazole, 50 mg/mL ), group II (negative 50% Aloe vera ethanol extract 17.225±0.512 16.70 17.80 control by using 1% DMSO solution), group 25% Aloe vera ethanol extract 15.650±0.420 15.20 16.10 III (using 50% Aloe vera extract in 70% ethyl 12.5% Aloe vera ethanol extract 13.975±0.457 13.50 14.50 ethanol, diluted in 1% DMSO), Group IV (using 6.25% Aloe vera ethanol extract 12.450±0.208 12.20 12.70 25% Aloe vera extract in 70% ethyl ethanol, A) B) C) diluted in 1% DMSO), group V (using 12.50% Aloe vera extract in 70% ethyl ethanol, diluted in 1% DMSO), and group VI (using 6.25% Aloe vera extract in 70% ethyl ethanol, diluted in 1% DMSO). The swab was drawn over the enti- D) E) F) re surface of already prepared plates of SDA to get a uniform distribution of bacteria. The SDA plates were then kept lid side up in an incubator at 25 ºC for 72 hours. The plates were checked daily for spillage and growth of other organisms. Figure 1. Inhibition zone of Candida albicans on Sabouraud dex- Measurement of the inhibition zone was done on trose agar for four different Aloe vera ethanol extract concen- the third day when the margin of inhibition was trations (disc in the center): A) positive control showed a clear inhibition zone around disc diffusion of fluconazole, B) negative clearly visible. The zones of complete inhibition control showed no inhibition zone around disc diffusion, C) 50% were measured using a Vernier caliper in milli- concentration showed a mean inhibition zone of 17.225 ± 0.512 metres by gross visual inspection. mm, D) 25% concentration showed the mean inhibition zone of 15.650 ± 0.420 mm, E) 12.5% concentration showed the mean Statistical Analysis inhibition zone of 13.975 ± 0.457 mm, F) 6.25% concentration showed the mean inhibition zone of 12.450 ± 0.208 mm Statistical analysis was performed using one-way ANOVA test. The p<0.05 was considered statisti- DISCUSSION cally significant. In the present investigation, in vitro anti-fungal activity of the ethanol extracts of Aloe vera on

487 Medicinski Glasnik, Volume 17, Number 2, August 2020

pathogenic VVC species, Candida albicans, was of C. albicans is formed. Aloe vera leaf extract with quantitatively evaluated on the basis of the zone concentrations of 100%, 50% and 25% have the of inhibition. All four concentrations of Aloe vera inhibitory capabilities that are equivalent to nysta- ethanol extracts exhibited a varying degree of inhi- tin 0.50 mg, 0.24 mg and 0.20 mg, respectively. bitory effect against selected fungal pathogens. The A study by Renisheya et al. reported anti-bacteri- highest concentration of Aloe vera ethanol extracts al and antifungal activity of Aloe vera gel extract performed the maximum zone of inhibition. on pure five bacterial (Bacillus subtilis, Salmone- In Mpila et al. study, it was stated that antimicro- lla typhi, Escherichia coli, Staphylococcus aureus bial activity of Mayana leaves (Coleus atropur- and Proteus vulgaris) and three fungal (Aspergillus pureus [L] Benth) ethanol extract to the growth fumigatus, Candida albicans and Penicillium sp.) of Staphylococcus aureus, Escherichia coli and cultures. They found that the maximum inhibition Pseudomonas aeruginosa by active ingredients zone of 11 mm for C. albicians and 9 mm for Peni- could be grouped into four categories as weak, cillium sp. (17). A different study by Saniasiaya et moderate, strong and very strong (inhibition al. (10) reported antifungal effect of the Malaysian zone <6 mm, 6-10 mm, 11-20 mm and 20-30 Aloe vera leaf extract on selected fungal species of mm, respectively) (12). The results of our study pathogenic otomycosis species in a culture medi- have shown Aloe vera ethanol extracts with the um. For Aspergillus niger, a zone of inhibition for concentration of 6.25%, 12.5%, 25%, and 50% alcohol and aqueous extract was seen for all con- strong growth inhibition on Candida albicans. centrations except 3.125 g/mL; there was no zone of inhibition for both alcohol and aqueous extracts The study by Kurniawan et al. reported that etha- of Aloe vera leaf for C. albicans (10). The peparati- nol extract of Aloe vera contained flavonoid, al- on of the Aloe vera extract at high temperature may kaloids, tannins, saponins and steroids (10,13). have affected the active ingredient leading to the These compounds were considered to play a role ineffectiveness towards C. albicans (18). in inhibiting the growth of Candida albicans (10). Flavonoids can cause coagulation or clumping of In conclusion, the extract of Aloe vera has shown proteins. Clotting proteins undergo denaturation concentration-dependent antifungal effect on so they cannot function anymore (14). Tannins Candida albicans. A possible mechanism of the can interfere with the function of the cytoplasmic obtained antifungal properties is related with fla- membrane. At low concentrations it can damage vonoids, alkaloids, tannins, saponins and steroids/ the cytoplasmic membrane causing leakage of terpenoids, contained in Aloe vera extracts. Com- important metabolites that activate the enzyme parable inhibitory effect with fluconazole might system, whereas at high concentrations it can da- support the application of these extracts like effi- mage the cytoplasmic membrane and precipitate cient antifungal agents. Therefore, further clinical cell proteins. Saponins can reduce surface tension trials are needed to confirm efficacy and safety of resulting in increased permeability or cell leaka- Aloe vera extracts as potential antifungal agents. ge and cause intracellular compounds to come ACKNOWLEDGMENT out. Steroids can inhibit protein synthesis because they accumulate and cause changes in cell consti- The authors are grateful to the Head of Depar- tuent components (14,15). Terpenoids exhibited tment of Dermatology and Venereology, Medical excellent activity against C. albicans yeast and Faculty of Universitas Sumatera Utara / Univer- hyphal form growth. Thus, terpenoids may be use- sitas Sumatera Utara Hospital, Medan, Indone- ful in the near future not only as an antifungal che- sia, for their support during this study. motherapeutic agent but also to synergize effects of conventional drugs like fluconazole (15). FUNDING A study by Huslina (16) reported that the admini- No specific funding was received for this study. stration of Aloe vera leaf extract affects the length TRANSPARENCY DECLARATION of the inhibition zone in the growth of C. albicans: the greater concentration of Aloe vera leaf extract Conflict of interest: None to declare. is given, the greater the zone of growth inhibition

488 Nabila et al. Aloe vera as anti-Candida

REFERENCES 1. Spampinato C, Leonardi D. Candida infections, cau- 13. Kurniawan J. Uji efek hepatoprotektor ekstrak eta- ses, targets, and resistance mechanisms: traditional nol daun lidah buaya (Aloe vera l.) terhadap kadar and alternative antifungal agents. BioMed Research malondialdehid plasma tikus jantan galur wistar International 2013; 2013: 204237. yang diinduksi parasetamol (Test of hepatoprotector 2. MacCallum DM. Hosting infection: experimental effect of ethanol extract of Aloe vera [Aloe vera l.] models to assay Candida virulence. Int J Microbiol leaves against malondialdehyde levels in paraceta- 2012; 2012:363764. mol-induced male Wistar strain mouse) [in Indone- 3. Berman J, Sudbery PE. Candida albicans: a mole- sian]. Jurnal Mahasiswa Fakultas Kedokteran Untan cular revolution built on lessons from budding yeast. 2015; 3:1-18. Nat Rev Genet 2002; 3:918-30. 14. Dwijoseputro D. Dasar-dasar mikrobiologi (Basics 4. Pfaller MA, Diekema DJ. Epidemiology of invasi- of Microbiology) [in Indonesian]. Jakarta: Djamba- ve mycoses in North America. Crit Rev Microbiol tan, 2005. 2010; 36:1-53. 15. Zore GB, Thakre AD, Jadhav S, Karuppayil SM. 5. Sobel JD. Vulvovaginal candidosis. Lancet 2007; Terpenoids inhibit Candida albicans growth by 369:1961-71. affecting membrane integrity and arrest of cell cycle. 6. Mayer FL, Wilson D, Hube B. Candida albicans pat- Phytomedicine 2011; 18:118-90. hogenicity mechanisms. Virulence 2013; 4:119–28. 16. Huslina F. Pengaruh ekstrak daun lidah buaya (Aloe 7. Dovnik A, Golle A, Novak D, Arko D, Takač I. vera l.) terhadap pertumbuhan jamur Candida albi- Treatment of vulvovaginal candidiasis: a review of cans secara in vitro (The effect of Aloe vera [Aloe the literature. Acta Dermatovenerol Alp Pannonica vera l.] leaf extract on the growth of Candida albi- Adriat 2015; 24:5-7. cans in in vitro) [in Indonesian]. Jurnal Ilmiah Bio- 8. Sobel JD, Sobel R. Current treatment options for logi Teknologi dan Kependidikan 2017; 5:72-7. vulvovaginal candidiasis caused by azole-resistant 17. Kaur H, Goyal RR, Bhattacharya A, Gupta R, Lal Candida species. Expert Opin Pharmacother 2018; NK, Arora B, Barua A, Yadav R, Balha M, Rana- 19:971-7. nmay B, Kuanr N, Dhawan D, Sharma K, Chahal 9. Radha MH, Laxmipriya NP. Evaluation of biological N, Prasad T. Antifungal activity of phyto-extracts of properties and clinical effectiveness of Aloe vera: a Piper longum, Aloe vera, and Withania somnifera systematic review. J Tradit Complement Med 2014; against human fungal opportunistic pathogen Candi- 5:21-6. da albicans. DU Journal of Undergraduate Research 10. Saniasiaya J, Salim R, Mohamad I, Harun A. Anti- and Innovation. 2015; 1: 107-15. fungal effect of Malaysian Aloe vera leaf extract on 18. Renisheya JJMT, Johnson M, Nancy BS, Laju RS, selected fungal species of pathogenic otomycosis Anupriya G, Renola JJET. Anti-bacterial and anti- species in in vitro culture medium. Oman Medical fungal activity of Aloe vera gel extract. International Journal 2017; 32:41–6. Journal of Biomedical and Advance Research 2012; 11. Kumar S, Yadav M, Yadav M, Yadav JP. Compara- 3:184-7. tive analysis of antimicrobial activity of methanolic extracts of Aloe vera and quantification ofAloe emo- din collected from different climatic zones of India. Arch Clin Microb 2015; 6:1. 12. Mpila D, Fatimawali F, Weny W. Uji aktivitas anti- bakteri terhadap ekstrak etanol daun Mayana (Co- leus antropurpereus (l.) benth) terhadap Staphylo- coccus aureus, Escherichia coli, dan Pseudomonas aeruginosa secara in-vitro (Antibacterial activity test against ethanol extract of Mayana (Coleus antropurpereus (l.) benth) leaves against Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa in-vitro) [in Indonesian]. Jurnal of Fakultas MIPA, Universitas Sam Ratulangi, Manado, 2012.

489 ORIGINAL ARTICLE

Strut graft vs. traditional plating in the management of periprosthetic humeral fractures: a multicentric cohort study Giuseppe Rollo1, Michele Biserni2, Gazi Huri3, Christian Carulli4, Mario Ronga5, Michele Bisaccia6, David Gomez-Garrido7, Nezih Ziroglu8, Enrico Maria Bonura9, Andrea Alberto Ruberti10, Andrea Schiavone11, Luigi Meccariello1

1Department of Orthopaedics and Traumatology, Vito Fazzi Hospital, Lecce, Italy, 2Department of Orthopaedics and Traumatology, Santa Maria della Stella, Orvieto, Terni, Italy, 3Orthopaedics and Traumatology Department, Hacettepe University School of Medicine, Ankara, Turkey, 4Orthopaedics Clinic, University of Firenze, Firenze, Italy, 5Department of Medicine and Health Sciences "Vincenzo Ti- berio" University of Molise, Campobasso, Italy, 6Orthopaedics and Traumatology Unit, Department of Surgical and Biomedical Science, S.M. Misericordia Hospital, University of Perugia, Sant'Andrea delle Fratte, Perugia, Italy, 7Division of Orthopaedics and Trauma Surgery, Hospital Quirónsalud y Hospital Laboral Solimat, Toledo, Spain, 8Bakirkoy Dr. Sadi Konuk Training & Research Hospital Istanbul, Tur- key, 9 Department of Orthopaedics and Traumatology, Poliambulanza Foundation Hospital, Brescia, Italy, 10 Department of Orthopaedics and Traumatology, Inselspital Bern, Berne, Switzerland, 11Department of Orthopaedics and Traumatology, Maggiore Hospital, Lodi, Italy

ABSTRACT

Aim The gradual increase in shoulder implants in active elderly patients has appeared in a parallel increase in periprosthetic humeral fractures. The aim of this study was to investigate the advantages of using strut grafting with plate fixation during periprosthetic humerus fractures.

Methods Thirty patients diagnosed with periprosthetic humeral fracture were divided into two groups. The first group of 15 pa- Corresponding author: tients (PS) underwent plate, ring, screws and strut allografts. The Luigi Meccariello second group with resting 15 patients (PWS) was treated with only Department of Orthopaedics and plate and screws. The criteria to evaluate the groups during follow- Traumatology, Vito Fazzi Hospital up were the Constant Shoulder Score (CSS) and Oxford Shoulder Piazzetta Muratore, Block: Score (OSS); the bone healing was measured by X-rays, controls A Floor:V, 73100, Lecce, Italy measured by radiographic union score (RUS), and complications. The follow up was terminated at 12nd month in both groups. Phone: +39 329 9419 574; Fax: +39 082 3713 864; Results The difference between the two groups in all parameters E-mail: [email protected] was not significant. However, all patients gained adequate sho- ORCID: https://orcid.org/0000-0002- ulder motor skills for normal daily living activities. All fractures were healed. Only two complications were registered, and blood 3669-189X loss was minimal.

Conclusion We believe a revision to reverse shoulder prosthesis Original submission: with a long-stem implant with or without cortical strut allograft augmentation to be safe and appropriate in the management of the- 12 February 2019; se complex injuries, though technically challenging, and having Revised submission: good results for normal activities daily life. 10 March 2020; Accepted: Key words: bone strut, outcomes, bone healing, humeral, ORIF, periprosthetic 23 March 2020 doi: 10.17392/1141-20

Med Glas (Zenica) 2020; 17(2): 490-497

490 Rollo et al. ORIF+ in periprosthetic humeral

INTRODUCTION The types of fractures according to the Worland classification (10) were: type A fractures occur The quality-of-life improvement achieved about the tuberosities, type B fractures occur following total shoulder replacement is satisfac- about the stem and they are sub-classified by tory, which has resulted in a steady increase in fracture anatomy and implant stability: B1 spiral the number of shoulder arthroplasties performed fractures with a stable implant, B2 transverse or each year (1). The gradual increase in shoulder short oblique fractures about the tip of the stem implants in active elderly patients has resulted in a with a stable implant, and B3, those about the parallel increase of complications like periprosthe- stem with an unstable implant; type C fractures tic humeral fractures (2), whose incidence varies occur well distal to the tip of the stem. between 0.6% and 3% of all shoulder prostheses (3). Complications include glenohumeral instabi- The patients were divided into two groups accor- lity, aseptic loosening, rotator cuff pathology, sep- ding to their treatment preferences. The first tic complication, neural injury, deltoid dysfunction group, plate and strut (PS), was represented by and periprosthetic fracture (4). Periprosthetic hu- 15 patients treated with plate, strut allografts, meral fractures account for approximately 11% of screws and cerclage for the humeral periprosthe- all complications in shoulder prosthesis (5). tic fracture. The second group, plate without strut (PWS), was represented by 15 patients suffering Majority of registered complications were found from periprosthetic humeral fractures treated so- in patients over the age of 80 years with peripro- lely with plates and screws. sthetic fracture due to osteoporosis, while in the patients under the age of 75 due to medium or Non-Union Scoring System (NUSS) in retrospec- high energy trauma (6). Loss of stem fixation to tive mode was used to examine bone healing (11). the humeral bone is rare; however, bone resorp- All patients were informed clearly and thoro- tion around the humeral stem is more common ughly about all possible operative and conser- and can be a risk factor for periprosthetic fracture vative treatment options. Patients were treated and failure in the revision surgery (7). Only a few according to the ethical standards of the Helsinki reports have investigated bone resorption after Declaration and read, conceded, approved and si- shoulder prosthesis (8,9). gned the informed consent form. Significant risk factors for periprosthetic fracture All patients of both groups underwent the same include female gender, humeral head replacement postoperative rehabilitation protocol (see rehabi- accompanying rotator cuff repair, on-growth litation protocol). stem coating, and high size of the implant (7). Azienda Sanitaria Locale (ASL) Lecce/Italy Et- Using cortical bone transplants allows the ortho- hical Committee approved this research. paedic surgeon to combine biological materials Methods with traditional metallic fixation implants. This study aims to investigate the superiority of Plate and Strut (PS) surgical technique. Sur- bio metallic fixation (plate with allograft bone gery was executed in a beach chair position strut) compared with plate fixation in the surgery in 100% of the cases, using the deltopectoral of humeral periprosthetic fracture. approach with a lateral extension approach. Once the fracture areas were exposed, the previous im- PATIENTS AND METHODS plants were removed and the fracture line was reduced by surgical means. If the stem was unsta- Patients and study design ble it was revised with a long one. Strut allograft From January 2010 to December 2018, at the se- was prepared on a separate table and the humeral ven Trauma Level I Centres (five in Italy, one of shaft plate was temporarily fixed with a K-wire. each in Turkey and Spain) 30 diaphyseal peripro- To stabilize the strut bone allograft and hume- sthetic humeral fractures were treated. Exclusion ral shaft in the bottom, screws were performed. criteria included: fractures caused by oncologi- (We used metal cerclage cable in the long stem). cal and/or haematological disease, patients under Metal cerclage cable was implanted to stabilize the age of 65 years. Patients who underwent a medium and top of strut allograft; this was then follow-up of less than 12 months were excluded. placed to support the medial humeral wall. Bone

491 Medicinski Glasnik, Volume 17, Number 2, August 2020

grafts were placed to support the humeral shaft and to fill bone defects (Figure 1). At the end of the operation, the final reduction was evaluated by fluoroscopy in three plans (anterior-posterior - AP, external rotation, and axial).

Figure 2. A 71-year-old woman with low-energy trauma to the right upper limb, treated three years before with reverse shoulder prosthesis for total tear rator cuff. A-C) fracture according to Worland type B2; D) treatment with osteosynthesis by locking plate utilizing eccentrically placed screw holes and screws (Dyaphesal Humeral Plate, Intrauma, Rivoli, Torino, Italy); E, F, G) perfect anatomical reduction (Rollo G, 2016)

bilitation course, and to standardize and direct the whole patient population to follow a single physio-kinesiotherapy program to reduce the bias. Stage 1 (first month): Immobilization collar for 4 weeks, Codman exercises, forbidden external rotation movement for the first 6 weeks.

Figure 1. A 78-year-old man, who had had a reverse shoulder Stage 2 (about weeks 5-9) only if there were prosthesis for 10 years, implanted due to fracture of the proxi- clinical data of healing and fragments move as mal humerus. A) After a fall from the top, X-ray shows a Type a part, and no displacement was evident on the BIII fracture according to Worland; B) the level of fracture and the mobilization of the ascend stem; C1) the re-reimplantation x-ray, then: lightly passive movement, actively with long stem and higher size liner (arrow), C2) the osteosyn- supported forward elevation and abduction, so- thesis with metal cerclage cables and opposed allograft bone othing functional use week 3-6 (avoid resistant strut, C3) bone chips; D) postsurgical X-ray; E) X-ray inspection after 12 months from surgery (Rollo G, Biserni M, 2017) abduction), progressively decrease help during movement from week 6 on. Plate without Strut (PWS) surgical technique. Stage 3 (after two months): add isotonic, concen- In every case, surgery was done in the beach cha- tric, and eccentric strengthening activities. ir position using the deltopectoral approach with the lateral extension approach. After having pre- If there was bone healing but joint stiffness, then sented the fracture areas, implants were removed supplement passive stretching by a physiotherapist. and the fracture was reduced. If the stem was un- The chosen principles to evaluate the two groups stable, it was revised with a long one. The hume- during the follow-up were: the proximal humeral ral shaft plate was provisionally implanted with a complication after the two types of surgery, the k-wire or a clamp. To secure the strut bone allo- time of surgery, Constant Shoulder Score (CSS) graft and humeral shaft in the bottom, metal ca- (12), Oxford Shoulder Score (OSS) (13), radio- ble cerclage and screws were applied. (We used graphic Union Score (RUS) (14), and postopera- free metal cerclage cable in the long stem). For tive complications. the stabilization of the medium area, an anatomic The Constant-Shoulder Score (CSS) (12) is a pre-shaped low-profile semi-tubular plate was 100-points scale composed of a number of indivi- used with multi-planar screw holes (Dyaphesal dual parameters. These parameters define the level Humeral Plate, Intrauma Rivoli, Torino, Italy), of pain and the ability to carry out the normal da- which provide primary stability to fix the peri- ily activities of the patient. The Constant-Murley prosthetic fracture (Figure 2). At the end of the score was introduced to determine the functionality procedure, the final fracture was evaluated by after the treatment of a shoulder injury. The test is fluoroscopy in all plans (anteroposterior, external divided into four subscales: pain (15 points), activi- and internal rotation, and axial view). ties of daily living (20 points), strength (25 points) Recovery protocol. The aim of our protocol was and range of motion (forward elevation, external to provide an orientation of postoperative reha- rotation, abduction and internal rotation of the sho-

492 Rollo et al. ORIF+ in periprosthetic humeral

ulder) (40 points). The higher the score represent ent (r) was approximated at the second decimal. the higher the quality of the function. Subjective Cohen’s kappa (k) was used to determine the relia- findings (severity of pain, activities of daily living bility and validity of the correlation between func- and working in different positions) of the partici- tional osteosynthesis and bone healing. pants are responsible for 35 points and objective measurements (AROM without pain, measure- RESULTS ments exo -and endorotation via reference points The patients of the first group at the time of the and measuring muscle strength) are responsible for injury had a mean age of 76.27 (±12.68; range 68- the remaining 65 points. The Constant-Murley sco- 88), the association between the genders (M: F) re is used in almost every language without official was 0.25 (3: 12). Eight (53.33%) patients were due translations. In French, a validated translation has to trauma and the remaining seven (46.67%) due been published. The time needed to complete the to degenerative joint disease; reverse shoulder im- Constant-Murley test is between 5 to 7 minutes. plantation was applied. In all cases, cementless im- The Oxford Shoulder Score (OSS) (13) is a 12- plants were applied. The mean in years of survival item patient-reported specifically designed and of the implant before the fracture was 3.2 (±2.34; developed for assessing outcomes of shoulder 1-6). The type of accidents included a fall in six surgery, e.g. for assessing the impact on patients’ (40.00%), normal daily actives in eight (53.33%), quality of life of degenerative conditions such as sport in one (6.67%) patient. The types of frac- arthritis and rotator cuff problems. tures according the Worland classification (10) The bone healing was measured by Radiographic were: type A in 0 patients, B1 in five (33.33%), Union Score (RUS) (14), the RUS score provided B2 in five (33.33%), B3 in one (6.67%), C in four by Litrenta et al. (12) and derived from the RUST (26.67%) patients. The most affected dominant scoring system. The RUS provides four compo- side was the left side, in four (26.67%) patients. nent scores of cortical bridging, cortical disappea- The average point of the NUSS in PS was 54.89 rance, trabecular consolidation, and trabecular di- (±13.74; range 25-70) (Table 1). sappearance. Each component can be scored from Table 1. Characteristics of the patients treated with (PS) and 1 to 3. Similarly, the two trabecular indices were without plate and strut (PWS) scored from1-3, each based on consolidation for Variable PS PWS p Number of patients 15 15 p>0.05 one of the indices, and fracture line disappearance Average age (SD) 76.27 (±12.68) 76.87(±11.9) p>0.05 for the other. The overall RUS score therefore ran- (years) ged from a minimum of 10 to a maximum of 30. Range of age (years) 68-88 66-89 p>0.05 Gender ratio (M:F) 0.25 (3:12) 0.25 (3:12) p>0.05 The follow-up period was set at 12 months in the Reason for reverse Trauma: 8 (53.33) Trauma: 8 (53.33) p>0.05 shoulder implant Degenerative joint Degenerative joint two groups. p>0.05 (No, %) disease: 7 (46.67) disease: 7 (46.67) Statistical analysis Uncemented: 15 Uncemented: 15 p>0.05 Type of implant (No, (100) (100) %) To summarize the characteristics of this study gro- Cemented: 0 Cemented: 0 p>0.05 up and subgroups we used descriptive statistics. Average implant time before periprosthetic 3.2 (±2.34; 1-6) 3.1 (±2.38 ; 1-7) p>0.05 This included both means and standard deviations fracture (SD; range) of all continuous variables. The t-test was used to (years) compare continuous outcomes. To compare cate- Fall: 6 (40.00) Fall: 5 (33.33) Type of accident Daily actives: 8 Daily actives: 8 p>0.05 gorical variables the Fisher exact test was applied (No, %) (53.33) (55.33) (these groups were smaller than 10 patients). The Sport: 1 (6.67) Sport: 2 (13.34) A: 0 A: 0 statistical significance was defined as p<0.05. To B1: 5 (33.33) B1: 4 (26.67) Worland Classification B2: 5 (33.33) B2: 6 (40.00) p>0.05 make a comparison between the predictive score (No, %) B3: 1 (6.67) B3: 1 (6.67) of quality of life and outcomes we used the Pe- C: 4 (26.67) C: 4 (26.67) arson correlation coefficient (r). For simplicity of Injured humeral side Right: 5 (33.33) Right: 5 (33.33) p>0.05 data, mean age was rounded to the closest year, (No, %) Left: 10 (66.67) Left: 10 (66.67) Dominant injured Right: 2 (13.34) Right: 3 (20.00) p>0.05 including their standard deviations. Predictive upper limb (No, %) Left: 4 (26.67) Left: 3 (20.00) score of outcomes and quality of life and their Average Nonunion 54.89 (±13.74) 54.9 (±13.79) p>0.05 standard deviations were approximated at the first Scoring System (SD) Nonunion Scoring decimal, while the Pearson correlation coeffici- 25-70 25-70 p>0.05 system range

493 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 2. The comparison of the patients treated with (PS) and without plate and strut (PWS) Characteristics PS PWS p Average surgical time in minutes 113.2 (±13.5; range 81 -176) 102.6(±13.7; range 79 -175) p>0.05 The mean follow up (SD; range) (years) (years) 1.4 (±1.8; 1-7) 1.4 (±1.8; 1-7) p>0.05 Average time of X-rays bone healing (SD; range) (days) 122.4 (±23.8; 94 -156) 122.7 (±23.7; 95 -158) p>0.05 Blood loss: 465±224.08 mL Blood Loss: 474±229.01 mL p>0.05 Intraoperative fracture: 0 Intraoperative fracture: 1(6.67) p<0.05 Type of complications (No, %) Postoperative fracture: 1 (6.67) Postoperative fracture: 0 p<0.05 Died after 1 year of follow up: 1 (6.67) Died After 1 year of follow up: 1 (6.67) p>0.05 Arc of flexion (SD; range) 129.57° (±34.36; 67°–168°) 129.47° (± 34.28 (66°–169°) p>0.05 Arc of extension (SD; range) 32.1° (±12.8; 30°–54°) 31.9° (±12.2; 30°–54°) p>0.05 Arc of abduction (SD; range) 117.54° (± 36.55; 69°–173°) 116.99° (± 36.79; 68°–172°) p>0.05 Arc of intrarotation with abduction at 90° (SD; range) 61.42° (± 24.54; 38-86) 61.32° (± 24.27; 38-86) p>0.05 Arc of extrarotation in adduction (SD; range) 42.20° (± 22.82; 25°–65°) 43.20° (± 21.90; 25°–65°) p>0.05 Average correlation between osteosynthesis and bone k=0.8123±0.137795 k=0.81236±0.119355 p>0.05 healing at the moment of X-rays bone callus

The main purpose of the reverse shoulder im- The objective quality and shoulder function of plantation was the trauma in eight (53.33%) pa- PS's life before the trauma, measured by CSS, tients and degenerative shoulder disease in seven was about 74.3 points (range 46-96) while the (46.67%) patients. In all cases, cementless im- quite quality of life before the trauma, measured plants were applied. The mean in years of sur- by CSS, was about 74.2 points (range 48-94) in vival of the implant before the fracture was 3.1 PWS, p>0.05(Graph.1). At the moment of trau- (±2.38; 1-7). ma, in the PS group, the CSS was 14.1 (range At the endpoint, the complications in PS were: 0-30) at the same moment PWS, the CSS was blood loss of 465±224.08 mL, post-operative 13.9 (range 0-30) (p>0.05). After 1 month from fracture in one (6.67%), death after one year of the revision surgery the CSS score was in 23.5 follow up in one (6.67%) patient; complications (range 10-45) PS and 23.6 (range 10-45) in PWS in PWS were: blood loss of 474±229.01 mL, intra (p>0.05). Also on the third month after the sur- operative fracture in one (6.67%), death after one gery (p>0.05), the difference at three CSS scores year of follow up in one (6.67%) patient. There was 48.4 in PS (range 25-70 ) and 48.3 in PWS( was a significant statistical difference between range 25-70), as well as in the sixth month of PS and PWS in intra and post-operative fracture follow-up. At 6 months from the revision sur- (p<0.05) (Table 2). gery, in PS the CSS was 54.6 (range 30-85), while in PWS was 54.4 (range 30-85) (p>0.05). The average duration of surgical time in the PS At twelve months after the surgery, we had CSS group was 113.2 (±13.5; range 81 -176) minutes; in score in PS 66.7(range 30-90), while in PWS was the PWS group it was 102.6 (±13.7; range 79 -175) 66.7(range 30-90) (p>0.05) (Figure 3A). (p> 0.05). The PS's X-rays bone healing scaled by radiographic union score occurred in our group The subjective quality and shoulder function of on average of 122.4 (±23.8; range 94 -156) days PS's life before the trauma, measured by OSS, after the surgery, while PWS's X-rays bone heal- was about 71.2 points (range 40-100), while the ing scaled by radiographic union score occurred quite quality of life before the trauma, measured on average 122.7 (±23.7; range 95 -158) days after by OSS, was about 70.9 points (range 40-100) in the surgery (p>0.05). Also, the arch of abduction PWS (p>0.05). At the moment of trauma, in the averaged 117.54° (± 36.55; range 69°–173°) in PS group, the OSS was 11.2 (range 0-24), and in group PS and 129,47° (±116.99° (± 36.79; range the same moment in PWS, the OSS was 11.1 (ran- 68°–172°) in group PWS (p>0.05). The arc of in- ge 0-24) (p>0.05). After one month from the revi- trarotation with abduction at 90° was in average sion surgery, the OSS score was in PS 20.4 (range 61.42°(± 24.54; range 38-86) in PS while 61.32°(± 5-40) and 20.8 (range 5-40) in PWS (p>0.05). At 24.27; range 38-86) in PWS (p>0.05). The arc of twelve months after the surgery, OSS score in PS extrarotation in adduction was in average 42.20° (± was 63.4 (range 40-92), while in PWS it was 63.5 22.82; range 25°–65°) in PS while 43.20° (± 21.90; (range 40-92) (p>0.05) (Figure3B). range 25°–65°) in PWS (p>0.05) (Table 2).

494 Rollo et al. ORIF+ in periprosthetic humeral

sthesis in the humeral, these areas have no traction force from muscles, which may explain the more distal extensity of bone resorption and a greater tendency towards critical bone resorption (7,8). Another factor is the Rotator cuff gives to osteoporosis at the proximal humerus (17). The lack of weight-bearing can lead to medial cortical reabsorption more easily in humerus than femur (7). Periprosthetic humeral fractures are estimated to account for around 11% of all complications associated with total shoulder prosthesis (16). The latest radiographic studies show that bone resorption was noted: in about 85.7% of patients with humeral stem arthroplasty; full-thickness cortical bone resorption occurred on average 17.7%; mostly happening within the initial year after surgery, and the stress-shielding was exclusively seen with uncemented humeral stems (7,8, 16-18). The biometallic solution associated with stem revision in instability cases is the one most practiced in the treatment of this fracture challenge pattern Figure 3. A) Trend of Constant Shoulder Score (CSS) pre and (2,3,10,11, 19-21). However, even if the stem is at 1 year after the traumatic event. At twelfth month, there was no statistical differences between PS and PWS group not revised and the correct anatomy is not resto- (p>0.05); B) Trend of Subjective Oxford Elbow Score (OES) red, good functional results can be obtained (9). pre and at 1 year after the traumatic event. At twelfth month, The ability of the humerus to adapt to deformities there was no statistical differences between PS and PWS (< 20° anterior angulation; < 30° varus/valgus an- group (p>0.05) gulation; < 3 cm shortening) (22), the orthopaedic The average correlation between osteosynthe- surgeon does not seek the anatomy and biomecha- sis and bone healing at the moment of X-rays nics needed in the treatment of periprosthetic sho- was (k=0.8123±0.137795) in PS as in PWS ulder fracture (15) or of osteosynthesis failures (k=0.81236±0.119355) (p>0.05) (Table 2). (23). The time of humeral bone healing or nonunion are also multiple predisposing factors for the DISCUSSION origin of these complications: general factors lin- Bone resorption in the humeral stem following ked to the subject and the situation of fracture (24). shoulder prosthesis has been reported (15), lack of Kurowicki et al. (25) describe osteosynthesis met- stem fixation to the humeral bone did not happen; hod as employing a locking plate using eccentri- nevertheless, the bone nearby the humeral stem cally disposed screw holes to set "skive screws" in can be a risk agent for failure at the time of revisi- the proximal end of the plate to perform fixation on operation and periprosthetic fracture (15). The around the stem of the implant. A clinical series pathology of humeral bone resorption after total of five consecutive cases managed with this pro- shoulder arthroplasty (TSA) has been described as cedure was performed with an average follow-up stress shielding (15). The principal mechanism of of 29 months (range, 12–48). Two further patients ossein absorption following TSA is also conside- had less than 1-year follow-up. All patients showed red to be stress shielding (7,8,16). The situation of fracture healing. Functional outcomes were limi- bone resorption is confined to the proximal zone ted with only patient achieving forward elevation in TSA, whereas it prolongs more distally in sho- above 90 degrees, and the average American Sho- ulder arthroplasty. Besides, full-thickness cortical ulder and Elbow Surgeons Function score were 28 bone resorption is rare after TSA (7). Bone resor- (25,26). Pain relief was nearly uniform with an ave- ption was generally seen at the greater tuberosity rage visual analog scale pain score of 0.5 (25). The (Zone 1), lateral diaphysis (Zone 2), and calcar prosthesis revision should always be performed region (Zones 7) (7). Compared to shoulder pro- except in patients with poor prognosis (26). Indeed,

495 Medicinski Glasnik, Volume 17, Number 2, August 2020

we have reconstituted both in reading and in our In conclusion, the osteosynthesis of a fracture in study that the revision of periprosthetic fracture (fa- general terms has the purpose of obtaining a mor- ilure) can restore the functionality of the shoulder phological reconstruction and sufficient mecha- and its quality of life in most cases to sufficient le- nical stability to allow early mobilization to cure vels for normal daily life activities and the survival healing. In the case of periprosthetic fractures, of the system (9, 19-21, 27-30). We have two com- the problems are greater because of the presen- plications and blood loss normally described in the ce of the prosthesis that conditions the use of scientific literature (2,3,9,10, 25-28) traditional synthesis means. The use of multiple The limitations of the current study were the limi- plaques would have the disadvantage of creating ted number of patients, non-probability sample of an overly "metallic" engagement, while resorting convenience, due to low incidence in the populati- to the use of bank cortical bone strands has the on. Another limit is that it is a retrospective study. advantage of supporting a biologically valid biological material as well as osteointegration even Disadvantages of retrospective studies are inferi- if only partially. Finally, we believe a revision to or level of evidence compared with prospective reverse prosthesis with long-stem with or without studies, the patients to confounding (other risk cortical strut allograft augmentation to be safe factors may be present that were not measured), and proper in the management of these complex cannot determine causation, only association, injuries, though technically challenging and ha- some key statistics cannot be measured. ving good results for normal daily activities Another limitation was that the measurements and intervention were made without randomiza- FUNNDING tion of the researcher to the experimental groups, No specific funding was received for this study. which have the potential for bias. Finally, other limiting factors of the study acknowledged by the TRANSPARENCY DECLARATION authors can be the potential for regression to the Conflict of interest: None to declare. mean, the presence of temporal confounders and the mention of the subjective score.

REFERENCES 1. Boorman RS, Kopjar B, Fehringer E, Churchill RS, 7. Inoue K, Suenaga N, Oizumi N, Yamaguchi H, Smith K, Matsen FA 3rd. The effect of total shoulder Miyoshi N, Taniguchi N, Munemoto M, Egawa T, arthroplasty on self-assessed health status is compa- Tanaka Y. Humeral bone resorption after anatomic rable to that of total hip arthroplasty and coronary shoulder arthroplasty using an uncemented stem. J artery bypass grafting. J Shoulder Elbow Surg 2003; Shoulder Elbow Surg 2017; 26:1984-9. 12:158–63. 8. Spormann C, Durchholz H, Audigé L, Flury M, 2. Trompeter AJ, Gupta RR. The management of com- Schwyzer HK, Simmen BR, Kolling C. Patterns of plex periprosthetic humeral fractures: a case series proximal humeral bone resorption after total sho- of strut allograft augmentation, and a review of the ulder arthroplasty with an uncemented rectangular literature. Strat Traum Limb Recon 2013; 8:43–51. stem. J Shoulder Elbow Surg 2014; 23:1028-35. 3. Athwal GS, Sperling JW, Rispoli DM, Cofield RH. 9. Carta S, Fortina M, Riva A, Meccariello L, Cataldi Periprosthetic humeral fractures during shoulder ar- C, Colasanti GB, Ferrata P. Is there a rationale in the throplasty. J Bone Joint Surg Am 2009; 91:594–603. use of the bone strut allograft for management and 4. Sewell MD, Kang SN, Al-Hadithy N, Higgs DS, internal fixation in periprosthetic humeral fractures? Bayley I, Falworth M, Lambert SM. Management of A case report. EMBJ 2017; 12:115–8. peri-prosthetic fracture of the humerus with severe 10. Kirchhoff C, Kirchhoff S, Biberthaler P. Periprosthe- bone loss and loosening of the humeral component tic humeral fractures: strategies and techniques for after total shoulder replacement. J Bone Joint Surg osteosynthesis. Unfallchirurg 2016; 119:273-80. Br 2012; 94:1382-9. 11. Calori GM, Colombo M, Mazza EL, Mazzola S, 5. Bohsali KI, Wirth MA, Rockwood CA Jr. Complica- Malagoli E, Marelli N, Corradi A. Validation of the tions of total shoulder arthroplasty. J Bone Joint Surg Non-Union Scoring System in 300 long bone non- Am 2006; 88:2279–92. unions. Injury 2014; 45: (Suppl 6):S93-7. 6. Bergdahl C, Ekholm C, Wennergren D, Nilsson F, 12. Ban I, Troelsen A, Christiansen DH, Svendsen SW, Möller M. Epidemiology and patho-anatomical Kristensen MT. Standardised test protocol (Constant pattern of 2,011 humeral fractures: data from the Score) for evaluation of functionality in patients with Swedish Fracture Register. BMC Musculoskelet shoulder disorders. Dan Med J 2013; 60:A4608. Disord 2016; 17:159. 13. Dawson J, Rogers K, Fitzpatrick R, Carr A. The Oxford shoulder score revisited. Arch Orthop Trau- ma Surg 2009; 129:19-23.

496 Rollo et al. ORIF+ in periprosthetic humeral

14. Litrenta J, Tornetta P 3rd, Mehta S, Jones C, OʼToole 23. Rollo G, Tartaglia N, Falzarano G, Pichierri P, Stasi A, RV, Bhandari M, Kottmeier S, Ostrum R, Egol K, Medici A, Meccariello L. The challenge of non-union Ricci W, Schemitsch E, Horwitz D. Determination of in subtrochanteric fractures with breakage of intrame- radiographic healing: an assessment of consistency dullary nail: evaluation of outcomes in surgery revi- using RUST and Modified RUST in metadiaphyseal sion with angled blade plate and allograft bone strut. fractures. J Orthop Trauma 2015; 29:516-20. Eur J Trauma Emerg Surg 2017; 43:853-61. 15. Carta S, Fortina M, Riva A, Meccariello L, Manzi 24. Calori GM, Colombo M, Bucci MS, Fadigati P, Co- E, Di Giovanni A, Ferrata P. The biological metallic lombo AI, Mazzola S, Cefalo V, Mazza E. Compli- versus metallic solution in treating periprosthetic cations in proximal humeral fractures. Injury. 2016; femoral fractures: outcome assessment. Adv Med 47(Suppl 4):S54-8. 2016; 2016:2918735. 25. Kurowicki J, Momoh E, Levy JC. Treatment of 16. Raiss P, Edwards TB, Deutsch A, Shah A, Bruckner periprosthetic humerus fractures with open reduc- T, Loew M, Boileau P, Walch G. Radiographic chan- tion and internal fixation. J Orthop Trauma 2016; ges around humeral components in shoulder arthro- 30:e369-74. plasty. J Bone Joint Surg Am 2014; 96:e54. 26. Casier S, Middernacht B, Van Tongel A, De Wilde 17. Oh JH, Song BW, Kim SH, Choi JA, Lee JW, Chung L. Revision of reversed shoulder arthroplasty: is a SW, Rhie TY. The measurement of bone mineral reoperation possible? Obere Extrem 2017; 12:16-24. density of bilateral proximal humeri using DXA in 27. Hernandez NM, Chalmers BP, Wagner ER, Sperling patients with unilateral rotator cuff tear. Osteoporos JW, Cofield RH, Sanchez-Sotelo J. Revision to re- Int 2014; 25:2639-48. verse total shoulder arthroplasty restores stability for 18. Razfar N, Reeves JM, Langohr DG, Willing R, patients with unstable shoulder prostheses. Clin Ort- Athwal GS, Johnson JA. Comparison of proximal hop Relat Res 2017; 475:2716-22. humeral bone stresses between stemless, short stem, 28. Terragnoli F, Zattoni G, Damiani L, Cabrioli A, Li and standard stem length: a finite element analysis. J Bassi G. Treatment of proximal humeral fractures Shoulder Elbow Surg 2016; 25:1076-83. with reverse prostheses in elderly patients. J Orthop 19. Martinez AA, Calvo A, Cuenca J, Herrera A. Inter- Traumatol 2007; 8:71-6. nal fixation and strut allograft augmentation for pe- 29. Rollo G, Rotini R, Pichierri P, Giaracuni M, Stasi A, riprosthetic humeral fractures. J Orthop Surg (Hong Macchiarola L, Bisaccia M, Meccariello L. Grafting Kong) 2011; 19:191-3. and fixation of proximal humeral aseptic nonunion: 20. Kumar S, Sperling JW, Haidukewych GH, Cofield a prospective case series. Clin Cases Miner Bone RH. Periprosthetic humeral fractures after shoulder Metab 2017; 14:298-304. arthroplasty. J Bone Joint Surg Am 2004; 86-A:680–9. 30. Rollo G, Pichierri P, Marsilio A, Filipponi M, Bi- 21. Wutzler S, Laurer HL, Huhnstock S, Geiger EV, saccia M, Meccariello L. The challenge of nonunion Buehren V, Marzi I. Periprosthetic humeral fractures after osteosynthesis of the clavicle: is it a biomecha- after shoulder arthroplasty: operative management nical or infection problem? Clin Cases Miner Bone and functional outcome. Arch Orthop Trauma Surg Metab 2017; 14:372-8. 2009; 129:237-43. 22. Zhang BS, Li WY, Liu XH, Wei J, He L, Wang MY. Comparative results of non-operative and operative treatment of humeral shaft fractures. Beijing Da Xue Xue Bao 2017; 49:851-4.

497 ORIGINAL ARTICLE

Surgical treatment of multifragmentary segmental femur shaft fractures with ORIF and bone graft versus MIPO: a prospective control-group study Giuseppe Rollo1, Mario Ronga2, Enrico Maria Bonura3, Rocco Erasmo4, Michele Bisaccia5, Paolo Pichierri1, Antonio Marsilio1, Andrea Pasquino1, David Gomez Garrido6, Raffaele Franzese7, Andrea Schi- avone8, Luigi Meccariello1 1Department of Orthopaedics and Traumatology, Vito Fazzi Hospital, Lecce, Italy, 2Department of Medicine and Health Sciences 'Vincenzo Tiberio' University of Molise, Campobasso, Italy, 3Department of Orthopaedics and Traumatology, Poliambulanza Foundation Hospital, Brescia, Italy, 4Department of Orthopaedics and Traumatology, Santo Spirito Hospital, Pescara, Italy, 5Department of Ortho- paedics and Traumatology, Azienda Ospedaliera "Santa Maria della Misericordia", Perugia, Italy, 6Orthopaedics and Traumatology Unit, Hospital General de Villalba, Madrid, Spain, 7Orthopaedics and Traumatology Unit, Villa del Sole Hospital , Caserta, Italy, 8Department of Orthopaedics and Traumatology, Maggiore Hospital, Lodi, Italy

ABSTRACT Aim Multifragmentary segmental femoral shaft fracture is a high energy injury frequently associated with life-threatening conditions. The aim of this study was to compare the use of bio metallic open reduction internal fixation (ORIF) (plate with allograft bone strut) with minimally invasive plate osteosynthesis (MIPO) fixation for the treatment of multi-segmental femoral shaft fracture in terms of outcomes, bone healing and complications. Methods Forty patients with segmental femoral shaft fractures were included and divided into two groups: 20 patients treated with ORIF+, 20 with MIPO. All fractures were classified accor- Corresponding author: ding to AO (Arbeitsgemeinschaft für Osteosynthesefragen) and Luigi Meccariello Winquist and Hansen Classification. Evaluation criteria were: Department of Orthopaedics and duration of follow up and surgery, Non-Union Scoring System, Traumatology, Vito Fazzi Hospital Pain Visual Analogic Scale (VAS), objective quality of life and Piazzetta Filippo Muratore, hip function, subjective quality of life and knee function, quality of life the Short Form-12 Survey Questionnaires (SF-12), bone Block: A- Floor: V, Lecce, Italy healing and femoral alignment (radiographs), Radiographic Union Phone: +39 32 9941 9574; Score for Hip (RUSH). Fax: +39 08 2371 3864; Results Better results of ORIF in terms of complication rate, E-mail: [email protected] RUSH, VAS, regression between RUSH and VAS, average corre- Giuseppe Rollo ORCID: https://orcid. lation clinical-radiographic results and patients’ outcomes (Cohen org/0000-0003-1920-1286 k) were obtained, and similar results for the length of follow up, surgery duration, perioperative blood transfusion, wound healing. No statistical difference for Harris Hip Score (HHS), Knee Society Original submission: Score (KSS), quality of life (SF-12). 25 February 2020; Conclusions The ORIF and bone strut allograft technique had Revised submission: better results compared to the MIPO technique with regards to 24 March 2020; complication rate, RUSH, VAS, regression between RUSH and VAS, and average correlation clinical-radiographic results and pa- Accepted: tients’ outcomes (Cohen k) in the surgical treatment of multifra- 31 March 2020 gmentary segmental femoral shaft fractures. doi: 10.17392/1150-20 Key words: allograft, bone strut, limb reconstruction, trauma, outcomes Med Glas (Zenica) 2020; 17(2): 498-508

498 Rollo et al. Better results of ORIF+bone strut than MIPO

INTRODUCTION centre for surgical treatment of segmental femoral shaft fractures. Patients had to be fit for surgery, Multifragmentary segmental femoral shaft aged between 16 and 65 for males and 16 and 50 fractures are high energy injuries frequently for females. Exclusion criteria were: haemato- associated with life-threatening conditions (1). logical or oncological patients, acute or chronic The common femoral shaft fracture incidence is infections, previous lower limb trauma, nerve 21/100,000 person/years (2). Non-operative tre- injuries, segmental contralateral fracture, vessels atment of femoral shaft fractures in adults is an injuries, non 3.2 type of fracture according to AO exception (3). The concept of damage control in (Arbeitsgemeinschaft für Osteosynthesefragen) orthopaedics especially treating femoral fractu- AO Classification (7), all Winquist and Hansen res in polytraumatized patients is well establis- Classification’s types, e. g. 0, I and II femoral shaft hed, but clear indicators of which patients benefit (8), age under 16 or over 65 for males and 50 years from this approach are missing (4). for females, amputee or subamputee lower limb, Today intramedullary nailing of femoral shaft ipsilateral neck femoral injuries, intramedullary fractures is the gold standard of treatment. In a nailing or definitive external fixation treatment, recent analysis comparing different treatment op- bone metabolism diseases, skeletal immaturity, tions in femoral shaft fractures, it could be clearly mental or neurologic disorder. stated that intramedullary fixation was associated All patients were informed in a clear and com- with the lowest complication rates and loss of prehensive way of the two type of treatments and reduction rates compared to external fixation or other possible surgical and conservative alterna- plating strategies (5). Plate osteosynthesis is par- tives. Patients were treated according to the Et- ticularly advantageous in certain situations whe- hical Standards of the Helsinki Declaration, and re an intramedullary nail may be contraindicated were invited to read, understand, and sign the in- or technically not feasible. These may include formed consent form. the polytrauma patient, ipsilateral femoral neck and shaft fractures, fracture in the proximal or Azienda Sanitaria Locale (ASL) Lecce/Italy Et- distal shaft, paediatric femoral shaft fracture, or hical Committee approved this research. an excessively narrow intramedullary canal (6). Methods Due to high complication rates with infection, refracture, delayed healing, non-union and soft- All fractures were classified according to AO tissue problems, the concept of biological bridge Classification (7) and Winquist and Hansen Cla- plating (6) was developed with a minimally-inva- ssification (8). Forty patients were divided in two sive fixation technique (MIPO) in order to impro- groups: 20 patients treated with ORIF plus tibial ve results and outcomes. bone strut allograft (ORIF+) and 20 patients trea- The aim of this study is to compare the use of bio- ted with MIPO technique (Table 1, Table 2). metallic fixation (plate with allograft bone strut) All patients underwent the same rehabilitation for the Open Reduction Internal Fixation (ORIF) protocol (see rehabilitation protocol). To study with plate fixation with minimally invasive plate the bone healing on radiographs, the Non-Union osteosynthesis (MIPO) for the treatment of mul- Scoring System (NUSS) in retrospective mode ti-segmental femoral shaft fracture in terms of was used (9) (Table 1). outcomes, bone healing and complications. The criteria to evaluate the patient groups during the follow-up were: the mean follow up, durati- PATIENTS AND METHODS on of surgery, Pain Visual Analogic Score (VAS) collected the same day that the X-rays were taken Patients and study design (10), objective quality of life and hip function From a total of 120 femoral shaft fracture patients measured by the Harris Hip Score (HHS) (11), admitted and treated at five trauma centres (five subjective quality of life and knee function mea- Italian and one Spanish) from January 2016 to sured by Knee Society Score (KSS) (12), the Ra- December 2019 we finally included in our study diographic Union Score for Hip (RUSH), quality 40 patients with segmental femoral shaft fractures. of life measured by the Short Form Survey (SF- Inclusion criteria were: patients admitted to our 12) questionnaires (11). The bone healing and

499 Medicinski Glasnik, Volume 17, Number 2, August 2020

Table 1. Description of the patients with open reduction inter- Table 2. Description of associated injures in the patients nal fixation plus bone graft (ORIF+) and patients with plate with open reduction internal fixation plus bone graft (ORIF+) fixation with minimally invasive plate osteosynthesis (MIPO) and patients with plate fixation with minimally invasive plate Characteristic ORIF+ (n=20) MIPO (n=20) osteosynthesis (MIPO) Average age, years (standard Characteristic ORIF+ MIPO p 39.67 (±12.34) 39.56 (±11.87) deviation, SD) Injuries associated with the segmental shaft fracture (No, %) Age range (years) 16-65 16-65 Brain Injury/ Cerebral concussion 9 (45) 10 (50) >0.05 Gender ratio (No) (male:female) 9 (18:2) 9 (18:2) Fat Embolism 1 (5) 1 (5) 1.00 Previous type of accident (No, %) Hemopneumothorax 4 (20) 3 (15) >0.05 Fall from height 5 (25) 4 (20) Liver injuries 7 (35) 6 (30) >0.05 Car accident 4 (20) 5 (25) Spleen injuries 2 (10) 2 (10) 1.00 Motorcycle accident 4 (20) 5 (25) Bowel injuries 5 (25) 6 (30) P>0.05 Work accident 6 (30) 5 (25) Ipsilateral tibial injuries 4 (20) 5 (25) P>0.05 Agricultural accident 1 (5) 1 (5) Contralateral femoral injuries 4 (20) 3 (15) P>0.05 Previous type of femoral shaft fractures according AO (7) (No, %) Contralateral tibial injuries 3 (15) 4 (20) P>0.05 1(5) 1(5) Proximal 1/3 intact segment Rib fractures 15 (75) 16 (80) P>0.05 Clavicle fractures 4 (20) 5 (25) P>0.05 Proximal 1/3 fragmentary se- 1 (5) 1(5) gmental Humerus fractures 3 (15) 4 (20) P>0.05 6 (30) 6 (30) Forearm fractures 5 (25) 4 (20) P>0.05 Middle 1/3 intact segment Metatarsal fractures 6 (30) 5 (25) P>0.05 6 (30) 5 (25) Middle 1/3 fragmentary segmental Patella fractures 2 (10) 2 (10) 1.00 3 (15) 4 (20) Acetabulum fractures 2 (10) 2 (1) P>0.05 Distal 1/3 intact segment 3 (15) 3 (25) Pelvic injury 1 (5) 1 (1) P>0.05 Distal 1/3 fragmentary segmental Spine fractures 6 (30) 5 (25) P>0.05 Winquist and Hansen Classification’s femoral shaft type (No, %) Total 83 (100) 84 (100) P>0.05 III 11(55) 12 (60) Associate knee’s injuries type II: (No, %) IV 9 (45) 8 (40) Lateral meniscus 6 (30) 5 (25) P>0.05 Type of Fracture (No, %) Medial meniscus 12 (60) 13 (65) P>0.05 14 (70) Closed 15 (75) Posterior cruciate 3 (15) 2 (10) P>0.05 Anterior cruciate 11 (55) 10 (50) P>0.05 Open 6 (30) 5(25) Medial collateral ligament 5 (25) 4 (20) P>0.05 Open fracture according Gustilo Anderson Classification (14) (No, %) Lateral collateral ligament 8 (40) 9 (45) P>0.05 Type II: 10 (50) 10 (50) 26.1 25.7 Type IIIA 5 (25) 6 (30) ISS average (SD; range) P>0.05 (±4.77;12-52) (±4.86;12-53) TypeIIIB 5 (25) 4 (20) 11.8 11.9 Orthopaedic device used in the surgery for the damage control (No, %) GCS average (SD; range) P>0.05 4 (20) (±1.64;8-15) (±1.66;8-15) Skeletal traction 3 (15) Orthopaedic device – plate used in the definitive surgery (No, %) 16 (80) External fixation 17 (85) Straight 4 (20) 5 (25) P>0.05 Work occupation (No, %) Curved 5 (25) 5 (25) P>0.05 Agricultural industry 6 (30) 7 (35) Condylar 11 (55) 10 (25) P>0.05 Industrial sector 10 (50) 9 (45) ISS, injury severity Score; GCS, Glasgow Coma Score; Tertiary industry 4 (20) 4 (20) Injured lower limb side (No, %) 8 (40) multifragmentary fractures or long spiral fractures, Right 7 (35) the opposite uninjured limb was also prepared to Left 12 (60) 13 (65) Average Non Union Scoring allow intraoperative comparison with the fracture 50.89 (±18.33) 50.74 (±18.47) System (SD) side. The image intensifier was positioned on the Range Non Union Scoring system 21-75 21-75 opposite site of the operating table. The 4.5-mm broad locking compression plate system (LCP, De- femoral alignment were measured using plain puySynthes, Oberdorf, Switzerland) was used for radiographs films. The evaluation endpoint was all cases. The plates chosen were short or long, de- set at 12 months after surgery. Bone union was pending on the fracture location and configuration. measured using the radiographic union score as As a general rule, the plate should be long enough described by Litrenta et al. (13). to allow the insertion of at least three screws each ORIF + bone strut allograft technique. The pati- into the proximal and distal main fragments. The ent was placed in a supine position on the radiolu- modification of the surgical access consisted of cent operating table. A supporting pad was placed saving the vastus lateralis and perforating arteries. under the knee in 20 degrees of flexion with the After exposing the fracture site, the fracture site patella pointed upward. The limb was draped free was bloodied and the tibial shaft allogeneic bone from the iliac crest to the foot to allow intraoperati- strut was prepared on a separate table after perfor- ve assessment of length and rotation. In segmental ming tampon dye tests for the risk of infection. The

500 Rollo et al. Better results of ORIF+bone strut than MIPO

modelling of the tibial bone strut always measured to the plane between the periosteum and the vastus two and a half times the extent of the fracture site. lateralis muscle. The lateral cortex of the femur The distal portion of the strut was modelled to "fla- was exposed using two Hohmann retractors— me" to be as congruent as possible to the anatomi- one ventral and one dorsal on both incisions. A cal shape of the metaphyseal passage to the medial tunnelling instrument was then tunnelled from the femoral condyle, when we needed to use the bone proximal incision toward the distal incision betwe- strut to fix the distal segmental shaft fracture. The en both Hohmann retractors to create a submus- margins of the fracture site were modelled in such cular, extraperiosteal tunnel. One end of the plate a way to create a wide surface for a compression was tied to the hole at the tip of the tunnelling in- osteosynthesis. The hardware for osteosynthesis strument by means of a suture. The tunnelling in- used in all cases was an anatomically pre-con- strument was then withdrawn, pulling the attached toured low profile plate LCP or the condylar pla- plate along the prepared tunnel. te LISS (DepuySynthes, Oberdorf, Switzerland). Once the plate was fully advanced into the tunnel, Longitudinal traction or with external fixation or the image intensifier was used to check the correct great distractor was applied to restore the length position of the plate. Longitudinal traction or with and rotation alignment of the femur. The alignment external fixation or great distractor was applied was checked with the image intensifier in both an- to restore the length and rotation alignment of the teroposterior (AP) and lateral views. The compre- femur. The alignment was checked with the ima- ssion cortical screws were also applied to stabilize ge intensifier in both anteroposterior (AP) and the strut bone, and placed to reinforce the medi- lateral views. The length and angulation were al wall of the diaphysis or distal metaphysic too. re-checked. If reduction was satisfactory, we did Furthermore, the free space between the splint and the fixation with two compression screws near the bone was filled in with morcelized bone and bone fracture side (one distal and one proximal). After paste. At this step, the hip rotation test was done we completed the insertion of screws, the hip ro- by flexion of the hip and knee to 90 degrees, and tation test was done by flexion of the hip and knee internal and external rotation of the hip was perfor- to 90 degrees, and internal and external rotation med. Finally, the soft tissues and skin were sutured. of the hip was performed. If alignment was achie- MIPO technique. The patient was placed in a su- ved, the fixation was completed using at least three pine position on the radiolucent operating table. A bicortical screws on each main fragment; these supporting pad was placed under the knee in 20 de- screws could require longer or separate incisions grees of flexion with the patella pointed upward. for other screws insertion or percutaneous screws. The limb was draped free from the iliac crest to the Screw placement was done by different techniques foot to allow intraoperative assessment of length depending on surgeon’s preference. and rotation. In segmental multifragmentary fractu- Rehabilitation protocol. The aim of our proto- res or long spiral fractures, the opposite uninjured col was to provide the clinician an orientation of limb was also prepared to allow intraoperative com- postoperative rehabilitation course, and to stan- parison with the fracture side. The image intensifier dardize and direct the whole patient population was positioned on the opposite site of the operating to follow a single physio-kinesiotherapy program table. The 4.5-mm broad locking compression plate to reduce the bias. From the post-operative up to (LCP) system or the condylar plate LISS (Depu- the third week all patients wore a long leg splint ySynthes, Oberdorf, Switzerland) was used for all plaster with a flexed knee at 20°. cases. The plates chosen were short or long, depending on the fracture location and configuration. As Phase I (from 3 to 6 weeks). Target: protect the a general rule, the plate should be long enough to fixation and early bone healing avoiding weight allow the insertion of at least three screws each into bearing if instructed to do so, ensure wound hea- the proximal and distal main fragments. ling, attain and maintain full hip and knee extension, gain hip and knee flexion to 90 degrees, Small (4–5 cm) proximal and distal incisions were decrease hip and knee and leg swelling, promote made over the lateral aspect of the femur with quad muscle strength and control. Precautions: deep dissection down through the ilio-tibial tract weight bearing as ordered by surgeon, no resi- and vastus lateralis muscle in line with their fibres sted closed chain exercises x 6 weeks, no resi-

501 Medicinski Glasnik, Volume 17, Number 2, August 2020

sted open chain exercises x 6 weeks, limit knee tomical and biomechanical axis of the humerus flexion with strengthening to 45 degrees, avoid from the radiological point of view. loading knee at deep flexion angles. RESULTS Phase II – intermediate phase (from 6 to 12 weeks). Target: no effusion, full hip and knee There were no statistically significant differences extension, single leg stand control, normalize between the two populations according to age, gait, regain full motion, regain full muscle stren- gender, type of fracture, NUSS, etc. (Table 1). gth, good control and no pain with functional mo- All patients were polytraumatized patients, in the vements (including step up/down, walking in flat ORIF Group there were 83 associated lesions and floor, partial lunge) (staying less than 60° of knee eight in the MIPO group (p>0.05) (Table.2). flexion). Precautions: limit knee flexion with The most present associated knee injury was the strengthening to 45 degrees, avoid loading knee medial meniscus injury in both the ORIF + and at deep flexion angles, post-activity swelling, sta- MIPO group (p>0.05) (Table.2). ir stepper, deep knee bends and squats. There was no statistically significant difference Phase III (week 12+). Target: regain full muscle (p>0.05) between the two groups for Injury Se- strength, no thigh atrophy, gradual return to full verity Score, the Glasgow Coma Scale and the activity, walk without pain for 2 kilometres. Pre- plates used for osteosynthesis (Table 2). cautions: avoid pain along skin wounded and all The average time from trauma to definitive care lower limb, build up resistance and repetitions was 16.16 (±13.19; range 0-42) days in ORIF+, gradually, perform exercises slowly avoiding qu- while in MIPO it was 13.8 (±11.86; range 0-36) ick direction changes and impact loading, exer- days (p>0.05). cise frequency should be 4-6 times per week to build strength, be consistent and regular with The mean of follow-up was 18.36 (±1.12; range the exercise schedule, do not do knee extension 12–48) months for ORIF+ and 18.48 (±1.18; ran- weights with machine (running, jumping, pivo- ge 12–48) months for MIPO (p>0.05). ting or cutting, lunges, stairmaster, step exercise The surgery lasted for an average of 92.9 (±21.6; with impact). range 73-123) minutes in ORIF+, while 88.7 (±20.8; range 67-118) minutes for MIPO (p>0.05). Statistical analysis The RBC (red blood cell) of perioperative Descriptive statistics were used to summarize the transfusions was on average 4.2 (±1.42; range characteristics of the study group and subgroups, 1-9) in ORIF+, while 4.1 (±1.25; range 1-8) for including mean and standard deviation of all con- MIPO (p>0.05). tinuous variables. The t-test was used to compare In both groups, patients demonstrated appropria- continuous outcomes. The χ2 test or Fisher’s ex- te wound healing within 25 days. act test (in subgroups smaller than 10 patients) During the follow up no complications were noti- were used to compare categorical variables. The ced in ORIF+ group; instead there were two bre- statistical significance was defined as p<0.05. akage of plate cases and one plating bending case Pearson correlation coefficient (r) was used to (p<0.05) for MIPO. The time of plate breakage compare the predictive score of outcomes and with respect to surgery was: 110 days in one case quality of life. Mean age (and the range) of the and 87 days in the second case. The bending plate patients was rounded at the closest year. The pre- case was recorded 67 days after the surgery. dictive score of outcome and quality of life and the range were approximated at the first decimal, All three MIPO cases were re-operated using while the Pearson correlation coefficient (r) was lateral compression locking plate screws and approximated at the second decimal. medial anterior bone strut allograft. All these surgeries were successfully performed and were Cohen's kappa coefficient (κ) used to measure uneventful. inter-rater agreement for qualitative (categorical) items; through this parameter we calculated the The average time of bone healing was 145.4 concordance between different qualitative values (±24.8; 108 -173) days after the surgery in of the outcomes and the bone healing, the ana- ORIF+, while it was 154.7 (±25.3; 106 -187)

502 Rollo et al. Better results of ORIF+bone strut than MIPO

Figure 1. A 47-year-old male patient, closed trauma of right femur after car accident. A) 1/3 middle multi fragmentary frac- Figure 2. A 46-year-old female patient, closed trauma of right ture according to the new AO’s Classification while 32.C3 ac- femur after car accident; A-C) 1/3 middle multi fragmentary cording to the old AO’s Classification; B-D) after trans-skeletal fracture according to the new AO Classification while 32.C3 traction, the fractures were treated for 7 days after the trauma according to the old AO Classification (7). D-F) After trans- with ORIF technique; E-G) at 3 months from the surgery forma- skeletal traction, the fractures were treated for 5 days after tion of exuberant callus bone already present (arrows); H-J) at the trauma according to MIPO technique; G-I) After 20 months one year from the surgery X-rays show complete fusion of the from the surgery plate and screws were removed (G,H,I) and medial bone splint (Rollo G, 2016) the patient experienced pain during weight bearing; J-L) Af- ter 20 days from onset of symptoms no pain during weight days for MIPO (p>0.05). The average time of bearing was experienced and there was radiological proof of bone healing in reoperated MIPO cases was definitive bone healing (Bonura EM, 2017) 138.2 (±21.4; 102 -168) days. At average day of the bone healing the RUSH was of 29.4 (±0.2; range 27.9-30) points in ORIF+ while 27.8 (±0.8; range 26.4-30) in MIPO (p<0.05). At the last X-rays control before the breakage of the plates the RUSH was of 14.6 points in one case and 13.9 in the other case, while in bending plate case it was 16.8 points. At average day of the bone healing in the MIPO reoperated group, the RUSH was of 27.73 (±0.75; range 27.4-28.2). In the average day of the bone healing the VAS was 1.7 (±0.7; range 0-3) point in ORIF+ while Figure 3. A-B) A 41-year-old, fell from a height, polytrauma, it was 2.8 (±0.9; range 0-4) in MIPO (p<0.05). open fracture of the right femur; 1/3 middle multi segmental fracture according to the new AO Classification while 32.C3 At the last follow up review before the breakage according to tttt old AO Classification; B) Associated injury: of the plates, the VAS was 4.7 points in one case brain injury. The patient was provided with damage control and 3.9 in the other case, while in bending plate procedure with external fixation; D-E) At the 25th post-operative day, definitive fracture reduction and fixation with MIPO case it was 3.7 points. At average day of the bone technique; F) At 30 days from the surgery with MIPO, the ra- healing in the MIPO reoperated group, the VAS diographs showed a fracture’s breakdown despite a formation was 2.3 (±0.57; range 2-3). of bone callus; G) On the 67th postoperative day the patient presented limb shortening; the radiographs show the bending We found that on average day of bone healing the effect (top arrow and bottom arrow pointing to the inter-frag- regression between RUSH and VAS scores showed mentary screw indicated by the letter; F) of the plate; H) The a p= 0.042 in ORIF+, while p=0.072 in MIPO X-rays show the surgical revision with ORIF + technique; I-J) The X-rays show the X-ray result at 12 months from the surgi- (p<0.05). We found that at average day in MIPO cal revision. The line shows the reconstruction of the medial reoperated group of bone healing the regression wall femoral (Rollo G, 2016) between RUSH and VAS scores showed a p=0.051. sured by HHS, was about 87.8 points (±6.53; The ORIF+’s objective functionality of the range 82.2-100). The quite same functionality hip and quality of life before the trauma, mea- of the hip and quality of life before the trauma,

503 Medicinski Glasnik, Volume 17, Number 2, August 2020

measured by HHS, was about 88.1 points (±6.51; cond case, in the bending plate case it was 86.8. range 68.8-100) in MIPO (p>0.05). At the mo- The quality of ORIF+’s life before the trauma, ment of trauma, the HHS was 22.4 points (±3.27; measured by SF-12, was about 94.6 points (range range 11.8-42.7) as at the moment of subtrochan- 86.4-100), while the quality of life before the teric fracture in MIPO, the HHS was 22.6 (±3.22; trauma, measured by S-F12, was about 94.2 range 11.8-42.8) (p>0.05). There was no statisti- points (range 88.2-100) in MIPO (p>0.05). At cal significance (p>0.05) between the two groups the moment of trauma, in ORIF+ group the SF- at 1 month of follow up after the surgery. Also 12 was 34.8 (±6.47; range 16.4-46.6) as at the in the scoring of HHS there was a statistically moment of subtrochanteric fracture in MIPO, significant difference after the third month after HHS was 35.2 (±6.39; range 16.2-45.8) (p>0.05). the surgery in favour of the ORIF+ group as the There was no statistical significance (p>0.05) be- breaking of the fixation hardware prevented the tween the two group ORIF+ at 1 month of fol- injured lower limb function, and therefore, the low up after the surgery. After the third month patient's life quality, as well as the sixth month of after the surgery, there was a statistically signifi- follow-up. In ORIF+ group, at the twelve month cant difference at three months after the surgery follow up, the HHS score was 84.2 (±8.95; range in favour of the ORIF+ group as the breaking 77.6-100) points while in MIPO’s 17 cases the of the fixation hardware prevented the patient's HHS was 83.9 (±8.78; range 77.2-100) (p>0.05). quality life measured by SF-12, as well as the Twelve months after the revision surgery we had sixth month of follow-up. In ORIF+ group, at the same results in MIPO’s reoperated group, twelve months after the surgery, the SF-12 score according to HHS. The results were: 88.2 points was 90.4 (±12.77; range 76.8-100) points, while in the first case of breakage while 87.6 in the se- in MIPO’s 17 cases the SF-12 score was 89.6.4 cond case, in the bending plate case it was 90.4. (range 76.6-100) (p>0.05). The ORIF+’s objective functionality of the knee At twelve months after the revision surgery we and quality of life before the trauma, measured by had the same results in MIPO’s reoperated group, KSS, was about 92.4 points (±4.42; range 86.6- according to SF-12. He results were: 94.2 points in 100). Quite the same functionality of the hip and the first case of breakage while 96.6 in the second quality of life before the trauma, measured by case, in the bending plate case was 96.4. The higher KSS, was about 92.2 points (range 86.4.8-100) SF-12 score in this subgroup is due to the ORIF+ in MIPO, (p>0.05). At the moment of trauma, the psychological aspect because all of 3 patients had a KSS was 19.4 points (±6.53; range 11.8-36.7) sense of more security and lesser anxiety. as at the moment of subtrochanteric fracture in The average correlation of clinical-radiographic MIPO, the KSS was 19.6 (±6.47; range 11.8- results and patients outcomes was high accor- 36.7) (p>0.05). There was no statistical signifi- ding Cohen κ: 0.853±0.100 for ORIF+, while κ: cance (p>0.05) between the two groups, ORIF+ 0.727±0.095 for MIPO (p<0.05). at 1 month of follow up after the surgery. Also in the scoring of the KSS there was a statistically DISCUSSION significant difference after the third month after Multifragmentary segmental femoral shaft fractu- the surgery in favour of the ORIF+ group as the re is not a very common pattern of injury. These breaking of the fixation hardware prevented the types of fractures are usually associated with high injured lower limb function and therefore the energy injuries frequently associated with life- patient's life quality, as well as the sixth month threatening conditions (1). Damage control is the of follow-up. In ORIF+ group, at twelve month strategy temporarily used in polytrauma patients follow up, the KSS score was 81.6 (±5.89; range (which is followed by definitive treatment when 74.2-100) points while in MIPO’s 17 cases the patients’ conditions allow so). Non-operative KSS was 81.2 (±5.72; range 73.8-100) (p>0.05). options are an exception and used only in very Twelve months after the revision surgery we had selected cases (very short life expectancy, other the same results in MIPO’s reoperated group, life-threatening injuries, etc.) (3,4) according to HHS. The results were: 82.6 points Notoriously, the treatment of choice for diaphyse- in the first case of breakage while 80.8 in the se- al femoral fractures is internal fixation with intra-

504 Rollo et al. Better results of ORIF+bone strut than MIPO

medullary nail (IMN) stabilization; it achieves bone grafting as well as the rates of postoperati- correct alignment and high rate of bone healing ve complications, and preserves vascular supply with low complications rate and early limb mo- to the fracture site. This technique is considered bility (15). In fact recent studies have proved worth further studies, but also considered to have that IMN stabilization allows the lowest possi- proven to be a promising technique (20,21). ble complication rate and loss of reduction rate We present a control-group study including pati- (5,16). Doubts rose on the possibility to perform ents with sustained multifragmentary segmental reamed or unreamed nailing procedures, but se- femoral shaft fracture, treated either with ORIF+ veral studies have shown that reamed intrame- and bone strut allograft or with MIPO. We wan- dullary nailing is correlated with shorter time to ted to compare results of the two techniques and union and lower rates of delayed-union, nonuni- compare our results with those present in the li- on, and reoperation, and it should therefore be terature. We must say that our results are in con- considered as the favourite option compared to tradiction with the authors supporting the MIPO undreamed IMN stabilization (15,17) option. Better results in terms of the complica- However, IMN stabilization is not always the tre- tion rate, average day of proven bone healing atment of choice, and plating techniques play an (RUSH), VAS at the average day of bone healing, important role in treating such cases. Plate oste- regression between RUSH and VAS at average osynthesis is particularly advantageous in certain day of bone healing and the average correlation situations where an intramedullary nail may be clinical-radiographic results and patients outco- contraindicated or technically not feasible. The- mes (Cohen k) were obtained in the ORIF Gro- se may include the polytrauma patient, ipsilate- up. Differently similar results were recorded for ral femoral neck and shaft fractures, fracture in other clinical aspects, such as length of follow the proximal or distal shaft, paediatric femoral up, surgery duration, perioperative blood tran- shaft fracture, multifragmentary segmental shaft sfusion and wound healing. fractures or an excessively narrow intramedu- We did not record any significant complication in llary canal. Rigid locked plates and dynamic the ORIF+ Group, while we recorded 3 compli- fixation with active locking plates are both stu- cations for the MIPO Group: 2 metalwork brea- died options in this context with controversial kages and 1 cases of metalwork bending. These results (6,18). Metal plates and the use of strut 3 patients were all reoperated using lateral com- bone graft have become a very common stra- pression locking plate screws and medial ante- tegy to treat femur shaft fractures, with the aim rior bone strut allograft; results were good with to correct as much as possible the lack of bone no further significant complications. We suspect stock and to provide the highest grade of stability that these complications have a biomechanical and allow good fracture union in the shorter peri- background causing a higher rate of failure of od of time (19). These methods are however very MIPO construct, but we do not have enough data invasive procedures and carry a relatively high to support this theory. But as a matter of fact, the risk rate of malunion, nonunion, infection, poor complication rate of the MIPO group was signifi- blood supply, refracture (20). cantly higher than that in the ORIF group. Therefore, minimally-invasive fixation technique No statistical difference was again recorded throu- (MIPO) have been studied and introduced in or- ghout the entire follow up with regards to objecti- der to improve results and outcomes in specific ve functionality of the hip and quality of life before selected cases, when a plate fixation is indica- the trauma (measured by HHS), objective functio- ted. As a result of technical advancement, mini- nality of the knee and quality of life before the tra- mally invasive plate osteosynthesis (MIPO) has uma (measured by KSS), quality of ORIF+’s life gained popularity in recent years and has achie- before the trauma (measured by SF-12). ved satisfactory clinical outcomes (6). Several Basing the discussion on our results, we would authors claimed that MIPO, compared with the like to highlight that the ORIF+ and bone strut traditional approaches, decreases the union time technique to treat multifragmentary segmental fe- by minimizing the disruption of the soft tissues moral shaft fractures could deserve prioritization (including periosteum), lowers the incidence of against the MIPO technique. In fact the ORIF+

505 Medicinski Glasnik, Volume 17, Number 2, August 2020

technique could allow a very good anatomical time of bone healing (Radiographic Union Score), reduction of the fracture, together with an appro- complication rates and radiological imaging (plain priate reconstruction of the medial wall. Our stu- radiographs and CT scans) (24,25). dy suggests that ORIF+ can also provide better Another allegedly positive aspect of the use of results in terms of fracture healing and VAS com- MIPO proposed in the literature is the fact the pared to MIPO. Despite the fact that several stu- MIPO reduces the rate of cases needing bone dies had highlighted that MIPO could reduce the graft if compared to conventional plating in com- risk of infection, complications and union time plex femur fractures. The MIPO technique is (mainly due to a better preservation of the vascu- thought to allow biological fracture healing by lar supply to the fracture site and minimization of preserving the vascularity of all bone fragments, soft tissue disruption, with consequent undistur- thus serving as a living bone graft. Therefore, pri- bed rapid callus bone healing) (20,21), we did mary bone graft could not be necessary. However, not record matching results in our study, which bone grafting is recommended in cases with no was composed of two good size control-groups signs of callus on the radiographs at three months if considering the incidence of multisegmentary or cases with extreme destruction of vascularity segmental femoral shaft fractures. by trauma, open fracture, or bone loss where hea- A target of treatment of femur fractures should ling takes more time (6,26,27). According to our mainly be early functional recovery, avoidance of experience we cannot confirm this aspect as we complications (bedsores, pulmonary complicati- reported favourable results in the group where ons, osteoporosis from disuse), restoration of axial bone graft was used. alignment, good stabilization, early mobilization The analysis and discussion of our results seem to (to prevent stiffness and muscle atrophy), return to take the reader to favour the ORIF+ option aga- good quality of life (22,23). Compared to the data inst the MIPO option. Further studies with bigger in the literature, the ORIF technique allows results cohorts and more powered studies are needed in in keeping with this accepted target. In fact we order to finally validate or reject our hypothesis. recorded a low complication rate, good fracture As far as we know, we recommend the use of stabilization and healing, good clinical and func- ORIF and bone graft for the treatment of mul- tional final results which allowed very satisfactory tifragmentary segmental femur shaft fractures ROMs (range of motion) and quality of life scores when IMN stabilization is not indicated. Speci- throughout the entire follow up. fic indications that could favour MIPO have not Another advantage of ORIF+ against MIPO that been studied by the authors, but we encourage we would like to stress is the possibility to use the development of research towards this aspect bone strut allograft. This is potentially able to re- in order to have a more clear scenario on ORIF duce stress shielding, increase the percentage of vs. MIPO indications and advantages/disadvan- probability of fracture consolidation, makes the tages. Most of the current data in literature are system more stable, reduces complications, and based on case series and other studies with a low improves patients’ quality of life due to a shorter level of evidence. functional recovery. However, the customization The use of the ORIF option should however be ta- of the transplant must be considered against the ken with caution, trying to provide an appropriate potential disadvantages of the lengthening of sur- surgical technique and tissue handling, good hae- gical time and the complexity of the surgery, the mostasis throughout the surgery, reduced surgery risk of infections, nonunion, mortality, and tran- times, less possible invasive approach and exposu- smission of infectious diseases (22). Furthermore, re, satisfactory reduction and alignment, early mo- the advantages of the use of bone graft have been bilization. Good results are possible only if these reported in several studies for the treatment of fe- aspects are taken into account as widely known. mur fractures. This aspect was proved by reporting The results of our study show that the ORIF and better results of patients treated with metal fixa- bone strut allograft technique has better results tion and bone graft against patients treated with compared to the MIPO technique with regar- metal fixation alone. Advantages were recorded in ds to complication rate, average day of proven terms of quality of life scores, functional scores, bone healing (RUSH), VAS, regression between

506 Rollo et al. Better results of ORIF+bone strut than MIPO

RUSH and VAS and average correlation of cli- reconstruction of the medial wall. We advocate nical-radiographic results and patients outcomes for the need of a more powered study and bigger (Cohen k) in the surgical treatment of multifra- size cohorts in order to validate or reject our re- gmentary segmental femoral shaft fractures. sults and achieve consensus on ORIF+ vs MIPO. Perhaps similar results are obtained in terms of FUNNDING length of follow up, surgery duration, perioperative blood transfusion and wound healing. Both No specific funding was received for this study. techniques are indicated when IMN stabilization is not indicated to manage this pattern of femur TRANSPARENCY DECLARATION fractures. However, ORIF+ offers advantages re- Conflict of interest: None to declare. lated to a better anatomical fracture reduction and

REFERENCES 1. Testa G, Aloj D, Ghirri A, Petruccelli E, Pavone V, 12. Acharya KN, Rao MR. Retrograde nailing for distal Massé A. Treatment of femoral shaft fractures with third femoral shaft fractures: a prospective study. J monoaxial external fixation in polytrauma patients. Orthop Surg (Hong Kong) 2006; 14:253-8. F1000Res 2017; 6:1333. 13. Litrenta J, Tornetta P 3rd, Mehta S, Jones C, OʼToole 2. Enninghorst N, McDougall D, Evans JA, Sisak K, RV, Bhandari M, Kottmeier S, Ostrum R, Egol K, Balogh ZJ. Population-based epidemiology of fe- Ricci W, Schemitsch E, Horwitz D. Determination of mur shaft fractures. J Trauma Acute Care Surg 2013; radiographic healing: an assessment of consistency 74:1516-20. using RUST and modified RUST in metadiaphyseal 3. Gänsslen A, Gösling T, Hildebrand F, Pape HC, Oe- fractures. J Orthop Trauma 2015; 29:516–20. stern HJ. Femoral shaft fractures in adults: treatment 14. Saleeb H, Tosounidis T, Papakostidis C, Giannoudis options and controversies. Acta Chir Orthop Trau- PV. Incidence of deep infection, union and malunion matol Cech 2014; 81:108-17. for open diaphyseal femoral shaft fractures treated 4. D'Alleyrand JC, O'Toole RV. The evolution of dama- with IM nailing: a systematic review. Surgeon 2019; ge control orthopedics: current evidence and prac- 17:257-69. tical applications of early appropriate care. Orthop 15. El-Menyar A, Muneer M, Samson D, Al-Thani H, Clin North Am 2013; 44:499-507. Alobaidi A, Mussleman P, Latifi R. Early versus late 5. Jiang M, Li C, Yi C, Tang S. Early intramedullary intramedullary nailing for traumatic femur fractu- nailing of femoral shaft fracture on outcomes in pare management: meta-analysis J Orthop Surg Res tients with severe chest injury: a meta-analysis. Sci 2018; 24;13:182. Rep 2016; 6:30566. 16. Bucholz RW, Jones A. Fractures of the shaft of the 6. Apivatthakakul T, Chiewcharntanakit S. Mini- femur. J Bone Joint Surg Am 1991;73-A:1561–6. mally invasive plate osteosynthesis (MIPO) in the 17. Li AB, Zhang WJ, Guo WJ, Wang XH, Jin HM, treatment of the femoral shaft fracture where intra- Zhao YM. Reamed versus undreamed intramedu- medullary nailing is not indicated. Int Orthop 2009; llary nailing for the treatment of femoral fractures: A 33:1119-26. meta-analysis of prospective randomized controlled 7. Garnavos C, Kanakaris NK, Lasanianos NG, Tzortzi trials. Medicine (Baltimore) 2016; 95:e4248. P, West RM. New classification system for long-bo- 18. Madey SM, Tsai S, Fitzpatrick DC, Earley K, Lutsch ne fractures supplementing the AO/OTA classificati- M, Bottlang M. Dynamic fixation of humeral shaft on. Orthopedics 2012; 35:e709-19. fractures using active locking plates: a prospective 8. Carsen S, Park SS, Simon DA, Feibel RJ. Treatment observational study. Iowa Orthop J 2017; 37:1–10. With the SIGN nail in closed diaphyseal femur 19. Coquim J, Clemenzi J, Salahi M, Sherif A, Tavakko- fractures results in acceptable radiographic ali- li Avval P, Shah S, Schemitsch EH, Bagheri ZS, Bo- gnment. Clin Orthop Relat Res 2015; 473:2394-401. ugherara H, Zdero R. Biomechanical analysis using 9. Calori GM, Colombo M, Mazza EL, Mazzola S, FEA and experiments of metal plate and bone strut Malagoli E, Marelli N, Corradi A. Validation of the repair of a femur midshaft segmental defect. Biomed Non-Union Scoring System in 300 long bone non- Res Int. 2018; 2018:4650308. unions. Injury. 2014; 45(Suppl 6):S93-7. 20. Andalib A, Sheikhbahaei E, Andalib Z, Tahririan 10. Maiettini D, Bisaccia M, Caraffa A, Rinonapoli G, MA. Effectiveness of minimally invasive plate oste- Piscitelli L, Bisaccia O, Rollo G, Meccariello L, osynthesis (MIPO) on comminuted tibial or femoral Ceccarini P, Rebonato A. Feasibility and value of fractures. Arch Bone Jt Surg 2017; 5:290–5. Radiographic Union Score Hip fracture after tre- 21. Link BC, Apivatthakakul T, Hill BW, Cole PA, Babst atment with intramedullary nail of stable hip fractu- R. Minimally invasive plate osteosynthesis (MIPO) res. Acta Inform Med 2016; 24:394-6. of periprosthetic femoral fractures with percutaneo- 11. Dettoni F, Pellegrino P, La Russa MR, Bonasia DE, us cerclage wiring for fracture reduction: tips and Blonna D, Bruzzone M, Castoldi F, Rossi R. Valida- technique. JBJS Essent Surg Tech 2014; 4:e13. tion and cross cultural adaptation of the Italian version of the Harris Hip Score. Hip Int 2015; 25:91-7.

507 Medicinski Glasnik, Volume 17, Number 2, August 2020

22. Carta S, Fortina M, Riva A, Meccariello L, Manzi 25. Alam QS1, Alam MT, Reza MS, Roy MK, Kamru- E, Di Giovanni A, Ferrata P. The biological metallic zzaman M, Sayed KA, Alamgir MK, Mohiuddin versus metallic solution in treating periprosthetic AM. Evaluation of outcome of exchange nailing femoral fractures: outcome assessment. Adv Med with autogenous bone graft for treating aseptic no- 2016; 2016:2918735. nunion of femoral shaft fracture. Mymensingh Med 23. Ehlinger M, Bonnomet F, Adam P. Periprosthetic J 2019; 28:378-81. femoral fractures: the minimally invasive fixation 26. Wenda K, Runkel M, Degreif J, Rudig L. Minimally option. Orthop Traumatol Surg Res 2010; 96:304–9. invasive plate fixation in femoral shaft fractures. 24. Rollo G, Tartaglia N, Falzarano G, Pichierri P, Stasi A, Injury 1997; 28(Suppl 1):A13–9. Medici A, Meccariello L. The challenge of non-union 27. Rollo G, Pichierri P, Marsilio A, Filipponi M, Bi- in subtrochanteric fractures with breakage of intrame- saccia M, Meccariello L. The challenge of nonunion dullary nail: evaluation of outcomes in surgery revi- after osteosynthesis of the clavicle: is it a biomecha- sion with angled blade plate and allograft bone strut. nical or infection problem? Clin Cases Miner Bone Eur J Trauma Emerg Surg 2017; 43:853-61. Metab 2017; 14:372-8.

508 ORIGINAL ARTICLE

Tibial diaphyseal fractures in children: indications and limitations of the treatment with monolateral and hybrid external fixator Michele Bisaccia1, Giuseppe Rinonapoli1, Salvatore Di Giacinto2, Andrea Schiavone3, Simone Lazzeri2, David Gomez-Garrido4, Juan Antonio Herrera-Molpeceres5, Umberto Ripani6, Cristina Ibáñez-Vicente7, Giuseppe Rollo8, Enrico Maria Bonura9, Raffaele Franzese10, Luigi Meccariello8, Auro Caraffa1

1Orthopaedics and Traumatology Unit, Department of Surgical and Biomedical Science, S.M. Misericordia Hospital, University of Perugia, Sant’Andrea delle Fratte, Perugia, Italy, 2Department of Paediatrics, Orthopaedics and Traumatology, Meyer University Children’s Hospital, Firenze, Italy, 3Department of Orthopaedics and Traumatology, Maggiore Hospital, Lodi, Italy, 4Department of Orthopaedics and Traumatology, Orthopaedic and Traumatology Unit, Hospital Quirón Salud Toledo and Hospital Solimat Toledo, Toledo, Spain, 5Department of Orthopaedics and Traumatology, Hospital Universitario “Virgen de la Salud”, Toledo, Spain, 6Division of Anaesthesia, Analgesia and Intensive Care and Pain Therapy, Department of Emergency and Major Trauma, “Ospedali Riuniti di Ancona”, Ancona, Italy, 7Department of General Medicine, "Hospital de Getafe", Madrid, Spain, 8Department of Orthopaedics and Traumatology, Vito Fazzi Hospital, Lecce, Italy, 9Department of Orthopaedics and Traumatology, Poliambulanza Foundation Hospital, Brescia, Italy, 10Orthopaedics and Traumatology Unit, Villa del Sole Caserta, Caserta, Italy

ABSTRACT

Aim To report our indications and limitations about the use of external fixation in children.

Methods It was retrospectively reviewed all tibial fractures treated with monolateral and hybrid external fixator, at our three Cen- tres. It was included 32 fractures which did not show an acceptable reduction after an attempt under anaesthesia. The exclusion crite- Corresponding author: ria were: open fractures, children with previous fractures of the Luigi Meccariello lower limbs, with skeletal congenital diseases, fractures involving Department of Orthopaedics and the physis and with neurovascular involvement. All fractures were Traumatology, Vito Fazzi Hospital classified according to the AO (Arbeitsgemeinschaft für Osteo- Piazzetta Filippo Muratore, synthesefragen) classification. An outcome was evaluated according to the time needed to obtain radiographic bone healing, the Block: A- Floor: V, Lecce, Italy range of motion (ROM) of the ankle, the asymmetry of the lower Phone: +39 329 941 9574; limbs, the malunion, and complications. Fax: +39 082 371 3864; E-mail: [email protected] Results The average time of consolidation was 10.66 weeks (6-17 weeks). There were no cases of deep infection, but only seven cas- Michele Bisaccia ORCID: https://orcid. es of superficial pin infections. No patients reported loss of ROM org/0000-0001-6608-9245 of the knee or ankle. We had zero cases of residual angle greater than 5°, and in all cases the difference in length between the limbs was ˂1 cm. Original submission: Conclusion The external fixation is a viable technique in the tre- 06 March 2020; atment of tibial fractures in children. Therefore, the external fixa- Revised submission: tion, both monolateral and hybrid, should be considered a viable 10 March 2020; treatment for this type of fracture. Accepted: Key words: bone fixation, paediatric, tibia fractures, trauma 27 March 2020 doi: 10.17392/1159-20

Med Glas (Zenica) 2020; 17(2): 509-516

509 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION PRIENTS AND METHODS Tibial fractures account for about 15% of all fractu- Patients and study design res in children taking the third place in frequency (1). In 70% of cases, the fracture affects only the From January 20014 to December 2019 we en- tibia, while in the remaining 30% there is also an rolled all tibial shaft fractures treated at five Trau- involvement of the fibula. In 10% of cases they are ma Level I Centres (four in Italy, and one in Spa- open (1,2). About 60% of the fractures are loca- in) with external monolateral fixator (Hoffmann ted at the distal tibial shaft (1,2). An oblique line II, Stryker, Kalamazoo, Michigan, United States) fracture shows 35%, 32% are comminuted, 20% and hybrid fixator (Tenxor, Stryker, Kalamazoo, transversal line and 13% are spiral (2). Michigan, United States). Data were collected re- The most frequent causes are road accidents, trospectively, including patient demographics and and above all, sport injuries, like mountain bike, fracture classification according to the AO(Ar- rollerblade, ski, snowboard, etc.) (3). In spi- beitsgemeinschaft für Osteosynthesefragen) pae- ral fractures the traumatic mechanism is almost diatric comprehensive classification of long-bone always represented by a torsional force, with the fractures (2): diaphyseal fractures 42-D: fractures foot blocked to the ground, having as the initi- of both bones, simple - complete transverse (< 30°), al site of injury the distal portion at the level of 42-D/4.1; fractures of both bones, simple - com- the anteromedial cortical and then developing in plete oblique or spiral (>30°), 42-D/5.1; fractures posterolateral direction (4). Transverse or multi- of both bones, simple - complete oblique or spi- fragmented fractures are instead caused mostly ral (> 30°), 42t-D/5.1; fractures of both bones, by direct trauma (5). The complications such as multifragmentary - complete transverse (< 30°), vascular or nerve injuries or compartment syn- 42-D/4.2; fractures of both bones, multifragmen- drome are rare (4,5). Secondary complications tary - complete oblique or spiral (>30°), 42-D/5.2; include malunion and premature physeal closure fractures of both bones, multifragmentary - com- in fractures extending to physis (4,5). plete transverse (< 30°), 42t-D/4.2; fractures of both bones, multifragmentary - complete oblique The most commonly used treatment for the tibi- or spiral (> 30°), 42t-D/5.2; isolated fractures of al child fracture is a reduction under anaesthesia the fibula, simple - complete transverse (< 30°), and cast-immobilization (6). In those cases where 42f-D/4.1; isolated fractures of the fibula, simple adequate reduction cannot be obtained, surgical - complete oblique or spiral (> 30°), 42f-D/5.1; treatment is indicated (6). Other indications for isolated fractures of the fibula, multifragmentary surgical treatment are: open fractures, compar- - complete transverse (< 30°), 42f-D/4.2; isolated tment syndrome, fractures in children with spasti- fractures of the fibula, multifragmentary - comple- city, floating knee and highly unstable fractures, te oblique or spiral (> 30°), 42f-D/5.2. for which it is not possible to obtain or maintain an adequate stability (7,8). Although conservative The inclusion criteria were fractures of the tibial treatment has been practiced for many years with shaft which did not obtain an acceptable reduction satisfactory results, several aspects have led to after an attempt under anaesthesia. The exclusion an increase in the number of surgical procedures criteria were: open fractures, children with previo- including changes of living, sport habits, econo- us fractures of the lower limbs, with systemic and mics, and patient’s request to treatment (7,8). metabolic disorders, children with skeletal congenital diseases, fractures involving the physis and The most widely used means of synthesis are: fractures with neurovascular involvement. titanium intramedullary nails, metal or absorbable pins, plates and screws; external fixators Azienda Ospedaliera Universitaria Sanitaria Peru- are instead used almost exclusively in open gia/Italy Ethical Committee approved this research. fractures (9-11). Methods The aim of our study is to evaluate clinical and radiographic results of external fixation, both All patients underwent anterior posterior (AP) hybrid and monolateral in the non-open tibial and lateral radiographs (L) of the tibia to assess shaft fractures. the location and type (transverse, oblique, spiral or comminuted) of the fracture. The type of inter-

510 Bisaccia et al. Tibial diaphyseal fractures in children

vention was explained to the patients' parents and delayed union, nonunion: union after 9 months or a written consent was received. In all cases the after the second procedure; malalignment greater intervention was performed within 48 hours after than 10° and/or heterometry of the limbs greater injury under general anaesthesia with the patient than 20 mm; skin infection, osteomyelitis. in the supine position on a radio transparent ta- To describe the possible infection we used the ble. All patients received a prophylactic dose of CkChan's scale (16). Checketts–Otterburns gra- cefazolin 30 minutes before the incision at a dose ding system, describes the possible infection in of 25mg / kg body weight (12). 5 grades: grade 1 - slight erythema, little disc- In all cases it was possible to obtain closed reduc- harge, treat with improved local pin care; grade tion. All patients were discharged within two days 2 - erythema, discharge, pain, warmth, treat with after surgery, with instructions for the daily dre- improved local pin care and oral antibiotics; gra- ssing of the pin with Sodium hypochlorite 1.1%. de 3 - as per grade 2, but no improvement with At discharge, the patients were explained about oral antibiotics, pins/ex fix can be continued; the home rehabilitation program, aimed at the grade 4 - severe soft tissue infection involving range of motion (ROM) recovery of the ankle several pins ± pin loosening, ex fix must be dis- and knee, the recovery of the quadriceps femoral continued; grade 5 - as per grade 4, but with bone muscle, ischio-crural and gluteal muscles. From involvement visible on radiographs, ex fix must the second day after the surgery a partial load was be discontinued; grade 6 - major infection occu- granted and gradually increased thereafter. rring after ex fix removal, treatment requires cu- rettage of pin track. On average, the removal of the fixators was performed at 10 weeks (8-12 weeks). Surgical technique. In 21 cases we used a monolateral fixator and in the remaining 11 cases a Malunion was defined as a residual deformity hybrid fixator.The choice between the two types exceeding 10° of angulation on the coronal pla- of fixators was imposed by the location of the ne and/or 10° of angulation on the sagittal plane fracture. Fractures located at the third medial and/or obvious clinical malrotation (14,15). of the tibial shaft were treated with monolateral The return to recreational activities was gradual fixation, while fractures localized at the distal and secondary to muscle tone recovery. On ave- third of the shaft (were treated) with hybrid fixa- rage all patients resumed their activities between tion. In the case of the monolateral fixator after th th the 12 and 16 week. having realigned the fracture under fluoroscopy Radiographs were made at 2, 5, 10, 16 weeks, 6 as much as possible, first we placed the distal months and one year. The fracture was conside- screws (3) and then the proximal (3) and connec- red united when callus was visible on the radio- ted them with bars and clamps. Where necessary, graphs in at least three cortices with no tenderne- in order to reduce the lever arm, we placed an ss at the fracture site. intermediate pin, therefore reducing the fracture We took into consideration the angles on the frontal under fluoroscopic control. and sagittal planes both in the immediate postope- Statistical analysis rative and at longitudinal controls and at the final visit. At the last radiographic control at one year To summarize the characteristics of this study gro- we performed x-rays in the erect standing position up and subgroups we used descriptive statistics. (orthostatism) to obtain measurements for a possi- This included both mean and standard deviation ble heterometry of the limbs. An inter-rater reliabi- (SD) of all continuous variables. The t-test was lity analysis using the Kappa statistic was perfor- used to compare continuous outcomes. To com- med to determinate consistency among the ratings. pare categorical variables the Fisher exact test The clinical evaluation at follow-up included the was applied (these groups are smaller than 10 pa- ROM measurement of the ankle and knee, rotati- tients). The statistical significance was defined as on and limb alignment, possible skin infections. p<0.05. To make a comparison between the pre- Clinical outcomes were evaluated using modified dictive score of quality of life and outcomes we criteria described by Flynn (13). The complicati- used the Pearson correlation coefficient (r). For ons taken into consideration were the following: simplicity of data, mean ages was rounded to the

511 Medicinski Glasnik, Volume 17, Number 2, August 2020

closest year, including their standard deviations. In 11 cases in which the hybrid fixator was used Predictive score of outcomes and quality of life we first positioned three distal wires (1.5 or2 and their standard deviations were approximated mm) and then connected the semicircle to the at the first decimal, while at the second decimal it proximal screws, in a number of 3 (Table 1). The was approximated Pearson correlation coefficient average surgical time was of 38 min (±12.23; (r). Cohen’s kappa (k) was used to determine the range 25-61). Under no circumstances open re- reliability and validity of the correlation between duction of the fracture was necessary. functional osteosynthesis and bone healing. All 32 cases resulted in fracture healing and there were no cases of delayed union and nonunion. RESULTS The average healing time was of 10.66 weeks A total of 32 patients were treated, 21 were males (±3.09; range 6-17). The average coronal defor- and 11 females. mity was 3.47° (±1.74; range 0-7). The average According to the AO classification three patients sagittal deformity was 3.62° (±1.88; range 0-6). were classified as A1, seven as A2, nine as B1 The average residual deformity heterometry was and five patients as C type (Table 1). 3.47mm (±2.51; range 0-8) (Table 1). The mean age was 9.88 (±2.50; range 5-15y/o) We had no case when we had an angle >5°, and years. The right side was affected in 15 cases and in all cases the difference in length between the the left side in 17 cases (Table 1). A road accident limbs was ˂1 cm (p>0.05). was the cause of fracture in 17 cases, a fall from No child developed a deep infection, and only height in nine and a sport injury in six cases. in seven cases grade 1 superficial infections of

Table 1. AO classification, results, complications, outcome, modified Flynn’s criteria of 32 children with tibial diaphyseal fractures (42-D) Residual Residual Time of Residual deformity Patient’s Age AO-OTA deformity deformity Localization union Sagittal plane Complications Outcome number (years) CLASS Coronal plane heterometry (weeks) (degrees) (degrees) (mm) 1 5 A1 Distal 6 0 0 0 None Excellent 2 8 A1 Medial 6 0 0 0 None Excellent 3 10 A1 Distal 7 0 0 0 Skin infection Excellent 4 15 A2 Distal 7 2 3 3 None Satisfactory 5 9 A2 Medial 6 3 5 2 Soft tissue irritation Excellent 6 10 A2 Medial 6 1 3 0 Skin infection Excellent 7 8 A2 Medial 7 2 6 4 None Satisfactory 8 14 A2 Distal 8 5 5 3 None Excellent 9 11 A2 Distal 9 4 0 2 Soft tissue irritation Excellent 10 9 A2 Medial 11 5 5 2 None Excellent 11 6 B1 Distal 15 4 4 5 None Excellent 12 7 B1 Distal 11 4 4 2 Skin infection Satisfactory 13 14 B1 Distal 16 3 5 7 None Excellent 14 12 B1 Distal 9 5 5 1 None Excellent 15 13 B1 Distal 13 3 3 7 None Excellent 16 7 B1 Distal 14 4 5 7 Soft tissue irritation Excellent 17 6 B1 Medial 15 2 0 0 None Satisfactory 18 11 B1 Medial 11 5 4 4 Skin infection Excellent 19 10 B1 Distal 11 5 5 4 None Excellent 20 9 B2 Distal 11 2 2 8 None Excellent 21 9 B2 Distal 14 5 4 4 Soft tissue irritation Excellent 22 8 B2 Medial 13 4 5 4 Skin infection Excellent 23 7 B2 Distal 11 4 5 1 None Excellent 24 11 B2 Medial 10 1 5 4 Soft tissue irritation Excellent 25 12 B2 Medial 9 4 5 1 None Excellent 26 12 B2 Distal 9 5 5 1 None Excellent 27 12 B2 Distal 17 5 4 8 Skin infection Satisfactory 28 11 C Medial 12 4 4 5 Skin infection Excellent 29 8 C Medial 11 5 4 6 None Excellent 30 10 C Distal 12 4 5 5 Soft tissue irritation Excellent 31 12 C Distal 11 7 1 6 None Satisfactory 32 10 C Distal 13 4 5 5 None Excellent AO-OTA CLASS, Arbeitsgemeinschaft für Osteosynthesefragen (AO) / Orthopaedic Trauma Association (OTA) classification

512 Bisaccia et al. Tibial diaphyseal fractures in children

Figure 1. A 10-year-old patient. Fracture type 4.2 A3 (AO classification). A-C) Fist treatment: reduction under anaesthesia and cast immobilization; B, C) breakdown of the fracture and rotation defect; D, E) good reduction after application of external fixator; F, G) good healing without bone consolidation defects after 16 weeks (Meccariello L, 2017)

Figure 2. A 15-year-old patient, fracture type 4.3.C3 AO classification. A, B) breakdown of the fracture and rotation defect with cast immobilization; C,D) good reduction after application of external hybrid fixator; E, F) good healing of bone and skin without bone consolidation defects after 16 weeks; G, H) after 6 months the bone healing after the remontion of external fixator (Bisaccia M, 2015) the pins according to CkChan's scale, all resol- fractures and/or with impossible reduction to ma- ved with oral antibiotics. In all cases, infections intain conservatively (21-23). were resolved with oral antibiotic therapy with Currently the intramedullary stabilization tech- amoxicillin+clavulanic acid 45 mg/6.4 mg/kg/ niques are the most used in the majority of fractu- day divided in two doses, for 7 days. res of long bones in the paediatric population In twenty-six cases the outcome was excellent (24). Other techniques involve the use of pins, (Table.1) screws of percutaneous and plaques (24,25). External fixation was well tolerated by all chil- The elastic intramedullary nails considering their dren. No patients reported loss of ROM of the mini-invasiveness are very useful in this type of ankle or knee. fracture. This type of synthesis requires an entry The average correlation of clinical-radiographic site that does not involve the physis, allows early results and patients outcomes was high (Cohen κ: mobilization, rapid recovery of the ROM, lack 0.80). A total of 26 excellent results and 6 satis- of any stiff joint, short-term hospitalization, low factory were noticed (Table 1, Figures 1, 2). costs and minimal surgical scar (26,27) Elastic stable intramedullary nailing may be com- DISCUSSION plicated by the loss of reduction following push Tibial shaft fractures in the paediatric population out of the nails at the entry site in unstable tibial are the ones that most often require hospitaliza- fractures and following technical failures (28). tion (17). Usually they do not present particular This system in fact showed some limitations, complications and can be treated with reduction especially in paediatric patients over 12 years of and casting (18); however, this method requires a age and with high body weight, particularly in prolonged immobilization and a careful follow- unstable fractures and in case of inadequate sur- up. Several studies, in fact, report good results gical technique (29,30). In these cases we may with cast treatment, compared with the risk of have shortening and angulation of the fragments secondary fracture displacement and the need in 5-12% of the cases (31). of new surgery under anaesthesia to replace the In a study of 35 adolescents Deakin (32) reports cast and correct the displacement (19,20). Surgi- 38% of malunion in tibial fractures treated with cal treatment is recommended in cases of open flexible intramedullary nails and a mean time for fractures, in polytrauma patients, in the case of union of 17 weeks higher than as shown in our compartment syndrome, in fractures with neu- case histories. ro-vascular involvement and in highly unstable

513 Medicinski Glasnik, Volume 17, Number 2, August 2020

Moreover, even the current indications are still heterometry, osteomyelitis or secondary systemic greatly debated. infection and loss of range of motion. A recent review (25) claims that the current gold The goal of our work was to analyse the results standard treatment of paediatric tibial fracture: obtained with external fixation in the treatment ESIN (elastic stable intramedullary nailing) has of this type of fracture. External fixation allows been repeatedly modified using end caps, pre- a good reduction of the fracture and an excellent bent nails, and optimized surgical techniques. In healing; it also has the advantage of being mini- addition, new methods such as rigid locking nails mally invasive; it does not require a second sur- and plates have been included in the treatment gical intervention for the removal of synthetic approaches for femur and tibia shaft fractures. means, allows early mobilization of the joints, All these methods of paediatric fractures care and in the event of fracture of the secondary de- carry inherent advantages that require considera- composition it is possible to change the exter- tion for each clinical situation (33). nal fixator without the need for re-intervention Instead, for what pertains the rigid intramedullary (37). In addition, the minimally invasiveness at nailing, unlike the adult population, it may not be the soft tissue level, the absence of periosteal possible in paediatric tibial fractures due to limi- damage, less bloody supply destruction in the tations including the proximal tibial physis plate fracture site make this a very attractive method and the small canal diameter in these patients (34) (38,39). Furthermore, this method shows a very low risk of infection. In fact in our case histories Other problems to keep into consideration are the we only had 7 cases of superficial infections of postoperative knee pain, destruction of the intra- the pin resolved with oral antibiotics, but no ca- medullary blood supply with more blood loss, ia- ses of deep infection. Skin infections are a pro- trogenic propagation of the fracture, inadequate blem commonly encountered by other authors distal fixation and hardware failure, leading to (27). The disadvantage of the external fixation malunion (35). could be in the poor tolerability of the procedu- External fixation, all things consistent with re, especially in young patients, and the need for the other techniques described in the literatu- better compliance by the patients towards other re, allows a good reduction of the fracture and means of synthesis. In our survey the fixator excellent healing. It also has the advantage of was well tolerated by all patients. being minimally invasive, does not require a se- In conclusion, more research is needed to deter- cond surgery for removal of synthetic means and minate the optimum treatment strategy for this allows quick mobilization of joints and in case common paediatric injury. The existing literature of secondary displacement of the fracture it is is of poor quality; consisting mainly of retrospec- possible to modify the external fixator to correct tive reviews of patients' medical records, charts, the displacement without the need for a second and radiographs. Carefully designed, high-qu- surgical intervention. ality prospective cohort studies utilizing a nati- The disadvantage of external fixation implant onalized multi-hospital approach are needed to could be the risk of a poor tolerance of this pro- improve understanding before protocols and gu- cedure, especially in young patients, and a requ- idelines can be developed and implemented. irement for a greater rate of compliance by the From our results we can state that external fixa- patient as compared to other synthetic means. In tion represents a valid alternative to intramedu- our patients, there were no reports of complaints llary nails for this type of fracture. There are a during daily activities or medication of pins. lot of benefits, such as a quick mobilization of In the literature, the external fixation in paediatric the joint, low invasiveness, only one surgery tibial fractures is reported for open fractures (36). and the possibility to correct any secondary Aslani et al. (34) compared the external fixation displacement. It is important to point out that and elastic intramedullary nails without finding the application of the external fixator should be statistically significant differences between the done by a surgeon experienced in this method two techniques of bone healing time, the end re- and that the patient and family must be coopera- sult, the percentage of malalignment or residual tive until its removal.

514 Bisaccia et al. Tibial diaphyseal fractures in children

FUNNDING TRANSPARENCY DECLARATION No specific funding was received for this study. Conflict of interest: None to declare.

REFERENCES 1. Wegmann H, Orendi I, Singer G, Eberl R, Castellani 16. Bisaccia M, Manni M, Colleluori G, Falzarano G, C, Schalamon J, Till H. The epidemiology of fractures Medici A, Meccariello L, Rinonapoli G, Schiavone in infants -Which accidents are preventable? Injury A, Ibáñez Vicente C, Ferraro A, Caraffa A. The ma- 2016; 47:188-91. nagement of pin-care in external fixation technique: 2. Joeris A, Lutz N, Wicki B, Slongo T, Audigé L. An povidone-iodine versus sodium hypochlorite 0.05% epidemiological evaluation of pediatric long bone (amukina-med) medications. Euromediterranean Bi- fractures - a retrospective cohort study of 2716 pati- omedical Journal 2016;11:81-7. ents from two Swiss tertiary pediatric hospitals. BMC 17. Galano GJ, Vitale MA, Kessler MW, Hyman JE, Vita- Pediatr 2014; 14:314. le MG. The most frequent traumatic orthopaedic inju- 3. Mashru RP, Herman MJ, Pizzutillo PD. Tibial shaft ries from a national pediatric inpatient population. J fractures in children and adolescents. J Am Acad Ort- Pediatr Orthop 2005; 25:39-44. hop Surg 2005; 13:345-52. 18. Stenroos A, Laaksonen T, Nietosvaara N, Jalkanen 4. Sauli A Palmu, SampoAuro, Martina Lohman, Reijo J, Nietosvaara Y. One in three of pediatric tibia shaft T Paukku, Jari I Peltonen, and YrjänäNietosvaara- fractures is currently treated operatively: a 6-year epi- Tibial fractures in children. A retrospective 27-year demiological study in two university hospitals in Fin- follow-up study. Acta Orthop 2014; 85:513–7. land treatment of pediatric tibia shaft fractures. Scand 5. Yang J, Letts M. Isolated fractures of the tibia with J Surg 2018; 107:269-74. intact fibula in children: A review of 95 patients. J Pe- 19. Rose S, Marzi I. Tibial shaft fractures in children. Tra- diatrOrthop 1997; 17:347–51. uma und Berufskrankheit. 2013; 15(Suppl. 2):212-6. 6. Heinrich SD, Mooney JF. Fractures of the shaft of the 20. Malhotra, Karan ; Pai, Sunil ; Radcliffe, Graham. Do tibia and fibula. Chapt. 25 In: Beaty JH, Kasser JR, minimally displaced, closed tibial fractures in chil- eds. Rockwood and Wilkins’ Fractures in Children. 7th dren need monitoring for compartment syndrome? ed. Philadelphia PA: Lippincott Williams &Wilkins, Injury 2015; 46:254-9. 2010:931–66. 21. Griffet J, Leroux J, Boudjouraf N, Abou-Daher A, 7. Bauer J, Hirzinger C, Metzger R. Quadruple ESIN El Hayek T. Elastic stable intramedullary nailing of (Elastic Stable Intramedullary Nailing): modified tibial shaft fractures in children. J Child Orthop 2011; treatment in pediatric distal tibial fractures. J Pediatr 5:297-304. Orthop 2017; 37:e100-3. 22. Zarad AL. Flexible intramedullary nails for unstable 8. Bartlett GS III, Weiner LS, Yang EC. Treatment of fractures of the tibia in children: a retrospective eva- type II and type III open tibia fractures in children. J luation effectiveness. Egypt Orthop J 2014; 49:281-5. Orthop Trauma 1997; 11:357–62. 23. Slongo T, Audigé L, Hunter JB, Berger SM. Clinical 9. Humphrey JA, Gillani S, Barry MJ. The role of exter- evaluation of end caps in elastic stable intramedullary nal fixators in paediatric trauma. Acta Orthop Belg nailing of femoral and tibial shaft fractures in chil- 2015; 81:363-7. dren. Eur J Trauma Emerg Surg 2011; 37:305. 10. Goodwin RC, Gaynor T, Mahar A, et al. Intramedu- 24. Vallamshetla VR, De Silva U, Bache CE, Gibbons PJ. llary flexible nail fixation of unstable pediatric tibi- Flexible intramedullary nails for unstable fractures of alshaft fractures. J Pediatr Orthop 2005; 25:570–6. the tibia in children. An eight-year experience. J Bone 11. Kubiak EM, Egal KA, Scher D, et al. Operative tre- Joint Surg Br 2006; 88:536-40. atment of tibial fractures in children: Are elastic sta- 25. Lieber J, Schmittenbecher P. Developments in the tre- ble intramedullary nails an improvement over exter- atment of pediatric long bone shaft fractures. Eur J nal fixation? J Bone Joint Surg Am 2005; 87:1761–8. Pediatr Surg 2013; 23:427-33. 12. Sviestina I, Mozgis J, Mozgis D. Analysis of anti- 26. Heo J, Oh CW, Park KH, Kim JW, Kim HJ, Lee JC, biotic surgical prophylaxis in hospitalized children Park IH. Elastic nailing of tibia shaft fractures in suffering upper and lower extremity injuries. Int J young children up to 10 years of age. Injury 2016; Clin Pharm 2016; 38:233-7. 47:832-6. 13. Flynn JM, Hresko T, Reynolds RA, Blasier RD, Da- 27. Tu KK, Zhou XT, Tao ZS, Chen WK, Huang ZL, Sun vidson R, Kasser J. Titanium elastic nails for pediatric T, Zhou Q, Yang L. Minimally invasive surgical tech- femur fractures: a multicenter study of early results nique: Percutaneous external fixation combined with with analysis of complications. J Pediatr Orthop titanium elastic nails for selective treatment of tibial 2001; 21:4-8. fractures. Injury 2015; 46:2428-32. 14. Setter KJ, Palomino KE. Pediatric tibia fractures: cu- 28. Sink EL at all Complications of pediatric femur rrent concepts. Curr Opin Pediatr 2006; 18:30-5. fractures treated with titanium elastic nails: a com- 15. Lee SH, Hong JY, Bae JH, Park JW, Park JH. Factors parison of fracture types. J Pediatr Orthop 2005; related to leg length discrepancy after flexible intra- 25:577-80. medullary nail fixation in pediatric lower-extremity 29. Narayanan Uget al. Complications of elastic stable in- fractures. J Pediatr Orthop B 2015; 24:246-50. tramedullary nail fixation of pediatric femoral fractures, and how to avoid them. J Pediatr Orthop 2004; 24:363-9.

515 Medicinski Glasnik, Volume 17, Number 2, August 2020

30. Marengo L, Paonessa M, Andreacchio A, Dimeglio A, 36. Ramasubbu RA, Ramasubbu BM. Surgical stabili- Potenza A, Canavese F. Displaced tibia shaft fractures zation for open tibial fractures in children: External in children treated by elastic stable intramedullary na- fixation or elastic stable intramedullary nail - which iling: results and complications in children weighing method is optimal? Indian J Orthop 2016; 50:455-63. 50 kg (110 lb) or more. Eur J Orthop Surg Traumatol 37. Bisaccia M, Meccariello L, Manni M, Falzarano G, 2016; 26:311-7. Medici A, Rinonapoli G, Di Giacinto S, Colleluori G, 31. Gordon JE, Gregush RV, Schoenecker PL, Dobbs Vicente CI, Ceccarini P, Bisaccia O, Caraffa A. Tre- MB, Luhmann SJ. Complications after titanium ela- atment of acute proximal humeral fractures in chil- stic nailing of pediatric tibial fractures. J Pediatr Ort- dren with modular external fixator. Journal of Acute hop 2007; 27:442-6. Disease 2016; 5:497-501. 32. Deakin DE, Winter H, Jain P, Bache CE. Malunion 38. Della Rocca GJ, Crist BD. External fixation versus following flexible intramedullary nails for tibial and conversion to intramedullary nailing for definitive femoral fractures in adolescents. J Child Orthop management of closed fractures of the femoral and ti- 2010; 4:571-7 bial shaft. J Am Acad Orthop Surg 2006; 14(10 Spec 33. Mashru RP, Herman MJ, Pizzutillo PD Tibial shaft No.):S131-5. fractures in children and adolescents. J Am AcadOrt- 39. Franzese R, Conte M, Gagliardo N, Pieretti, G. Chil- hop Surg 2005;13:345-52. dren vs elderly in orthopedic surgery site of infecti- 34. Aslani H, Tabrizi A, Sadighi A, Mirblok AR. Tre- on. Are there difference? Experience of a high volu- atment of open pediatric tibial fractures by external me plastic surgeon consultant. Acta Med Sal 2019; fixation versus flexible intramedullary nailing: a com- 49:24–9. parative study. Arch Trauma res 2013; 2:108-12 35. Dendrinos GK, Kontos S, Katsenis D, Dalas A. Tre- atment of high-energy tibial plateau fractures by the Ilizarov circular fixator. J Bone Joint Surg Br 1996;78:710-7.

516 ORIGINAL ARTICLE

Ten-year risk assessment for type 2 diabetes mellitus using the Finnish Diabetes Risk Score in family medicine

Suzana Savić1, Saša Stanivuković², Biljana Lakić1

1Department of Family Medicine, School of Medicine, 2.School of Medicine; University of Banja Luka

ABSTRACT

Aim To assess ten-year risk of diabetes mellitus type 2 (T2DM) using the Finnish Diabetes Risk Score (FINDRISC) in respondents over 18, in Primary Health Centre in Banja Luka.

Methods A prospective study was conducted using data from a population with undiagnosed T2DM in Primary Health Centre in Banja Luka. Eligible respondents were those aged 18 to 70 years. Sociodemographic, behavioural and anthropometric variables were those related to the risk models evaluated by FINDRISC.

Results Data were collected from 520 individuals, 58.8% female and 41.2% male (p=0.005). A very high risk of developing T2DM in the next ten years was found in 5.6% females and 3.7% males. A high risk was found in 12.4% females and 15.9% males, 34.2% respondents ≥ 65 years, 28.8% with body mass index >30 kg/m2, Corresponding author: 26.6% who were not practicing physical activity (p=0.000), 24.0% Suzana Savić who took antihypertensive drugs, 42.3% who were diagnosed with Department of Family Medicine, impaired glycaemia, 30.4% and 22.9% respondents whose parents School of Medicine, and distant relatives, respectively, had T2DM. A moderate risk occurred in 31.4% females with waist circumference >88 cm. University of Banja Luka Kordunaška 10 78000 Banja Luka, Half (50%) males with waist circumference >102 cm and 33.2% Republic of Srpska, respondents who were not eating fruits and vegetables every day Bosnia and Herzegovina had a slightly increased risk of developing T2DM (p<0.05). Phone: +387 65 544 472; Conclusion The FINDRISC may be used as a tool which would Fax: +387 51 216 813; help general practitioners in everyday work, to detect patients E-mail: [email protected] with T2DM risk factors and to encourage them to change life style ORCID ID: https://orcid.org/000-0002- towards healthy habits. 0677-6323 Key words: Diabetes Registry, life style changes, primary health centre Banja Luka, prospective study Original submission: 30 April 2020; Revised submission: 19 May 2020; Accepted: 23 June 2020 doi: 10.17392/1189-20

Med Glas (Zenica) 2020; 17(2): 517-522

517 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION Questions included age, body mass index, waist circumference, physical activity, daily consump- Diabetes mellitus type 2 (T2DM) is a growing tion of fruits, berries or vegetables, history of pandemic worldwide (1,2). According to the treatment with antihypertensive drugs, and hi- 2017 International Diabetes Federation, the pre- story of high blood glucose (2,17). Although it is valence of diabetes among people aged 20 to 64 widely used to assess the risk of the development years was 327 million and among people aged 65 of T2DM in the next ten years, FINDRISC was to 79 was 98 million. It is estimated that by 2045 also evaluated as a tool for identifying undiagno- this number will increase to 438 million among sed T2DM, abnormal glucose tolerance, and me- people aged 20 to 64 years and among people 65 tabolic syndrome (18-20). to 79 years it will be 191 million, unless preventive measures are implemented (2,3). Globally, We had decided to carry out this study becau- T2DM is responsible for about 2 million deaths se of the population with T2DM risk factors is per year (4,5), while it is estimated that about 825 growing. One of duties of family medicine prac- billion USD cost to the Health Care system will titioners is to recognize these factors and to edu- be linked to this disease (6,7). cate the patients how to diminish them. According to the Republic of Srpska Population The aim of this study was to assess ten-year risk Health and Diabetes Registry, the prevalence of of diabetes mellitus type 2 using FINDRISC in T2DM in Republic of Srpska has increased, ran- subjects over the age of 18 in Primary Health ging from 3.1% (2014), 3.5% (2015), 4% (2016) Centre in Banja Luka. to 5.2% (2017) in relation to the environment PATIENTS AND METHODS and the trend, and the prevalence of this disease is estimated to be between 4-6%. The inciden- Patients and study design ce rate in the observed period has been increa- This was a prospective study conducted in four sing and ranged from 79.4 / 100,000 inhabitants teams of family medicine in Primary Health Cen- in 2014 to 90/100.000 in 2017 (8,9). According tre in Banja Luka using data from patients with to the World Health Organization in Bosnia and undiagnosed T2DM. The study was conducted as Herzegovina, every second person is overweight an anonymous from 15 October 2019 to 15 Janu- (54.6%), while one in five is obese (19.2%) and ary 2020 with 520 respondents. The researchers physically inactive (20.6%) (1,10,11). conducted the survey on Mondays, Wednesdays Identifying individuals with undiagnosed T2DM and Fridays, from 9 a.m. to 5 p.m. on respondents represents an important approach for preventing who came to the family doctor's office for any or delaying complications of this disease (12). The reason. The nurses measured body mass index American Diabetes Association recommends te- (BMI) and waist circumference. sting for individuals at high risk of T2DM aged 40 The protocol informed consent and questionna- and over, with a family history of T2DM, obesity, ires were approved by the Ethical Institutional physical inactivity or dyslipidaemia (2,13). In the Committees at the Primary Health Centre Banja first step of T2DM case identification, an objecti- Luka. This work had been carried out in accor- ve assessment of the likelihood of the presence or dance with the Declaration of Helsinki. Eligible future development of an adverse health condition respondents were those aged 18 to 70 years who (14) is performed, while in the second step, an oral provided an informed consent. Females who re- glucose tolerance test (OGTT) or measurement of ported being pregnant or individuals having any glycosylated haemoglobin (HbA1c) may be per- physical disability preventing anthropometric formed, but only among those who, in the previo- measurements (weight, height, blood pressure us step, were categorized as at high risk (15). or waist circumference) or those bedridden were During the Finnish Diabetes Prevention Study, excluded from the study the Finnish Diabetes Risk Score (FINDRISC) was formed to allow early detection of individu- Methods als with increased risk for T2DM in the next ten Sociodemographic, behavioural and anthropome- years. The questionnaire consists of eight simple tric variables were those related to the risk models questions, while each answer is scored (2,16). evaluated by FINDRISC (18). These variables

518 Savić et al. Decade risk assessment for type 2 diabetes by FINDRISC

were: age (< 45 years 0 points; 45-54 years 2 po- eight (3.7%) males (FINDRISC >20), while 38 ints; 55-64 years 3 points; ≥65 years 4 points), (12.4%) females and 24 (11.2%) males (FIN- BMI (<25 kg/m2 0 points; 25-30kg/m2 1 point; ≥ DRISC 15-20) had a high risk. (p>0.05). 2 30 kg/m 3 points), waist circumference (male: < According to the distribution of respondents by 94 cm and female: <80 cm 0 points; male: 94–102 BMI and waist circumference, there was a sta- cm and female: 80–88 cm 3 points; male: >102 cm tistically significant difference (p<0.05) between and female: >88 cm 4 points), physical activity, at genders of the respondents. There was no signifi- least 30 min/day (yes 0 points; and no 2 points), cant difference between genders in relation to the daily consumption of fruits and vegetables, at least exercise of physical activity for at least 30 mi- one portion per day (every day 0 points; not every nutes (p=0.856). Female respondents consumed day 1 point), history of antihypertensive drug tre- fruits and vegetables every day more frequently atment NO 0 points and YES 2 points), history of than males (p <0.05). There was no statistically high blood glucose (whether the respondent had significant difference (p>0.05) between genders ever been found to have high blood glucose in in relation to the usage of antihypertensive dru- a health examination during an illness or during gs, the frequency of hyperglycemia and diabetes pregnancy; NO 0 points and YES 5 points) and occurrence in distant or close relatives (Table 1). family history of T2DM, score according to rela- Table 1. Prevalence of components of the Finnish Diabetes tives with T2DM diagnosis (NO 0 points; YES: Risk Score according to gender in 520 respondents grandparent, relative, uncle, aunt 3 points; YES: N (%) of respondents parents, siblings, son, daughter 5 points). Variable Total Male Female p 520 (100.0) 214 (41.2) 306 (58.8) ''Low“ ten-year risk assessment for T2DM (one in Age (years) 100 people will develop diabetes, 1%) had respon- < 45 221 (42.5) 79 (36.9) 142 (46.4) dents with scoring <7. „Slightly high risk“ for ten- 45-54 111 (21.3) 53 (24.8) 58 (19.0) 0.162 year risk assessment for T2DM (one in 25 people 55-64 115 (22.1) 50 (23.4) 65 (21.2) >64 73 (14.0) 32 (15.1) 41 (13.4) will develop diabetes, 4%) had respondents with BMI (kg/m²) score 7-11. ''Moderate risk“ for ten-year risk asse- <25 169 (32.5) 42 (19.6) 127 (41.5) ssment for T2DM (one in 6 people will develop 25-30 271 (52.1) 129 (60.3) 142 (46.4) 0.000 >30 80 (15.4%) 43 (20.1) 37 (12.1) diabetes, 17%) was found in respondents with sco- Waist circumference (cm) re 12-14. ''High risk“ for ten-year risk assessment M: <94; W:<80 203 (39.0) 69 (32.2) 134 (43.8) for T2DM (one in 3 people will develop diabetes, M:94-102; W: 80-88 244 (46.9) 107 (50.0) 137 (44.8) 0.013 33%) was found in respondents with score 15-20. M:>102; W: >88 73 (14.0) 38 (17.8) 35 (11.4) Physical activity (30 min/d) ''Very high risk“ for ten-year risk assessment for Yes 328 (63.1) 134 (62.6) 194 (63.4) 0.856 T2DM (one in 2 people will develop diabetes, No 192 (36.9) 80 (37.4) 112 (36.6) 50%) occurred in respondents with score >20. Vegetables-fruits Daily 336 (64.6) 128 (58.4) 211 (69.0) 0.013 No daily 184 (35.4) 89 (41.6) 95 (31.0) Statistical analysis Hypertension Without medication 312 (60.0) 136 (63.6) 176 (57.5) 0.167 Statistical methods included descriptive statistics With medication 208 (40.0) 78 (36.4) 130 (42.5) using measures of central tendency and standard Hyperglycaemia antecedent Yes 111 (21.3) 47 (22.0) 64 (20.9) deviation (Pearson's χ2 test, Fisher´s Extract, 0.774 Mann-Whitney U), and the differences between No 409 (78.7) 167 (78.0) 242 (79.1) Familiar antecedents DM2 individual groups of respondents were tested by No 310 (59.6) 127 (59.3) 183 (59.8) the χ2 test. There was a statistical significance Grandparent 118 (22.7) 46 (21.5) 72 (23.5) 0.713 between compared data if the probability was Parents 92 (17.7) 41 (19.2) 51 (16.7) BMI, Body mass index; M, Male; F, Female; DM2, Type 2 diabetes less than 5% (p<0.05). mellitus The lowest average risk of developing T2DM in RESULTS the next ten years was found in 148 (67%) res- A total of 520 individuals were invited to par- pondents younger than 45 years, and 25 (34.2%) ticipate in the study, out of which 306 (58.8%) older than 65 years had a high risk (p <0.05). were female and 214 (41.2%) male (p=0.005). The lowest risk was found in 109 (64.5%) res- Very high risk of developing T2DM in the next pondents with BMI < 25kg/m2, 106 (39.1%) with ten years was found in 17 (5.6%) females and BMI 25-30 kg/m2 had a slightly increased risk,

519 Medicinski Glasnik, Volume 17, Number 2, August 2020

while 23 (28.8%) respondents with BMI>30kg/ and 47 (42.3%) who had hyperglycemia measured m2 had a high risk (p<0.05). Slightly increased during routine screening (p <0.05). A high risk of risk of T2DM was found in 19 (50%) males with developing T2DM in the next 10 years was found waist circumference >102 cm, while 11 (31.4%) in 28 (30.4%) respondents whose close relatives females with waist circumference >88 cm had a had diabetes and in 27 (22.9%) respondents whose moderate risk (p <0.05) (Table 2). distant relatives had diabetes (p<0.05) (Table 3).

Table 2. The risk of developing diabetes mellitus type 2 Table 3. The risk of developing diabetes mellitus type 2 (T2DM) according to age, body mass index (BMI), waist (T2DM) according to physical activity, intake of fruits and circumference and gender vegetables, taking antihypertensive drugs, impaired blood No (%) of respondents at the risk* sugar and family burden diabetes Variable Slightly Very p No (%) of patients at the risk* Low Moderate High Total elevated high Question Slightly Very p Low Moderate High Total Age (years) 0.000 elevated high 148 61 5 6 1 221 ≤ 45 Do you normally have at least 30 minutes of physical (67.0) (27.6) (2.3) (2.7) (0.5) (100.0) activity at work and / or in your free time during the day 25 47 14 23 2 111 including normal daily activity? 45 do 54 (22.5) (42.3) (12.6) (20.7) (1.8) (100.0) 168 110 26 21 3 328 Yes 18 43 28 18 8 115 (51.2) (33.5) (7.9) (6.4) (0.9) (100.0) 55 do 64 (15.7) (37.4) (24.3) (15.7) (7.0) (100.0) 25 56 38 51 22 192 No 0.000 2 15 17 25 14 73 (13.0) (29.2) (19.8) (26.6) (11.5) (100.0) > 65 (2.7) (20.5) (23.3) (34.2) (19.2) (100.0) 193 166 64 72 25 520 Total 193 166 64 72 25 520 (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) Total (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) How often do you eat fruits and vegetables? 2 BMI (kg/m ) 0.000 Every 152 105 38 34 7 336 109 40 10 10 0 169 day (45.2) (31.3) (11.3) (10.1) (2.1) (100.0) ≤ 25 (64.5) (23.7) (5.9) (5.9) (0.0) (100.0) Not every 41 61 26 38 18 184 0.000 82 106 36 39 8 271 day (22.3) (33.2) (14.1) (20.7) (9.8) (100.0) 25.0–30.0 (30.3) (39.1) (13.3) (14.4) (3.0) (100.0) 193 166 64 72 25 520 Total 2 20 18 23 17 80 (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) ≥ 30.0 (2.5) (25.0) (22.5) (28.8) (21.3) (100.0) Have you ever taken antihypertensive drugs? 193 166 64 72 25 520 165 105 18 22 2 312 Total No (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) (52.9) (33.7) (5.8) (7.1) (0.6) (100.0) Waist circumference (cm) 28 61 46 50 23 208 Yes 0.000 Female 0.000 (13.5) (29.3) (22.1) (24.0) (11.1) (100.0) 97 26 8 3 0 134 193 166 64 72 25 520 < 80 Total (72.4) (19.4) (6.0) (2.2) (0.0) (100.0) (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) 33 50 16 26 12 137 80–88 Have you ever been measured for an impaired blood sugar level (24.1) (36.5) (11.7) (19.0) (8.8) (100.0) during a routine check-up during illness or pregnancy? 1 9 11 9 5 35 190 146 48 25 0 409 > 88 cm No (42.8) (25.7) (31.4) (25.7) (14.3) (100.0) (46.5) (35.7) (11.7) (6.1) (0.0) (100.0) 131 85 35 38 17 306 3 20 16 47 25 111 Total Yes 0.000 (2.9) (27.8) (11.4) (12.4) (5.6) (100.0) (2.7) (18.0) (14.4) (42.3) (22.5) (100.0) Male 0.000 193 166 64 72 25 520 Total 37 21 8 3 0 69 (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) < 94 (53.6) (30.4) (11.6) (4.3) (0.0) (100.0) Has anyone in your family had or now has diabetes? 23 41 15 23 5 107 162 100 31 17 0 310 94–102 No (21.5) (38.3) (14.0) (21.5) (4.7) (100.0) (52.3) (32.3) (10.0) (5.5) (0.0) (100.0) 2 19 6 8 3 38 24 42 19 27 6 118 > 102 cm Yes† (5.3) (50.0) (15.8) (21.1) (7.9) (100.0) (20.3) (35.6) (16.1) (22.9) (5.1) (100.0) 0.000 62 81 29 34 8 214 7 24 14 28 19 92 Total Yes‡ (29.0) (37.9) (13.6) (15.9) (3.7) (100.0) (7.6) (26.1) (15.2) (30.4) (20.7) (100.0) *Finnish Diabetes Risk Score (FINDRISC); 193 166 64 72 25 520 Total (37.1) (31.9) (12.3) (13.8) (4.8) (100.0) Over a quarter, 51 (26.6%) patients who were not *Finnish Diabetes Risk Score (FINDRISC); †grandparent, aunt, practicing daily physical activity had a high risk of uncle or first relative, but not parents, brothers and sisters; ‡parents, developing T2DM (p=0.000); 61 (33.2%) who had brothers, sisters or child not consumed fruits and vegetables every day had a DISCUSSION slightly increased risk (p <0.05), 38 (20.7%) had a The FINDRISC is a well-known risk score cre- high risk, and 18 (9.8%) had a very high risk. A high ated initially for incidental T2DM cases, but can risk of developing T2DM was found in 50 (24.0%) currently be used for T2DM screening (21). Using respondents who had used antihypertensive drug, this questionnaire of 520 respondents in our study,

520 Savić et al. Decade risk assessment for type 2 diabetes by FINDRISC

in 12.4% of women and 11.2% of men a high risk physical activity promotes health and reduces the (FINDRISC 15-20), and in 5.6% of women and risk of impaired glucose tolerance and others (27). 3.7% of men a very high risk (FINDRISC over 20) In our study, almost half of respondents who used to develop T2DM in the next ten years was deter- fruits and vegetables in their daily diet had a low mined. The study by Atayoglu et al. conducted in risk, while a quarter who did not use antihyperten- Turkey to 1500 adults aged ≥18 years using the sive drugs had a high risk of developing T2DM. FINDRISC as Diabetes Risk Questionnaire found A Colombian study (28) using FINDRISC found 13.5% respondents were in the high-risk group 46.95% respondents had a daily physical activity, (15.2% of females vs. 12.4% of male) (22). A stu- which is lower than in our study (63%). An intake dy conducted on Peruvian population (using The of fruits in the diet was more frequent than in the FINDRISC questionnaire) indicated that 37.1% Columbian study (28), while taking antihyperten- of respondents had a high risk of T2DM (23). A sive therapy was similarly represented. cross-sectional analysis of Vandersmissen at al. in- A prospective cohort study conducted on the dicated a prevalence of unknown dysglycemia of Spanish population in Primary Health Care for 1.8% among 275 healthy employees: 12% had a people who have had FINDRISC 15 and higher moderate risk and 5.5% had a high - very high risk showed that intensive lifestyle changes signifi- (24). The study by Bergmann et al. conducted on cantly reduced the risk of developing diabetes over 500 German subjects at increased risk of de- (29). Kolb et al. have also shown that reducing veloping T2DM indicated FINDRISC as a simple, the prevalence of diabetes is directly related to high-performance tool to predict the risk of deve- lifestyle changes (30). In our study, in a group of loping this disease. It has also been pointed out that respondents who had an elevated glycaemia, most subjects with lower FINDRISC will soon benefit of them had a positive family history for T2DM, more from implementing preventive interventions and at same time, according to FINDRISC, they to reduce the risk factor of developing T2DM (25). had a high risk of developing T2DM. In our study, a significant prevalence in respon- In conclusion, this study confirmed a high risk of dents aged 45-54 was at high and very high risk developing T2DM in a larger number of female for a development of T2DM, while in respon- respondents, respondents over 65, respondents with dents >65 years old, this prevalence was higher. BMI over 30 kg/m2, respondents who did not exer- Our results have shown higher risk of diabetes cise and did not eat fruits and vegetables daily, as was in correlation with higher BMI and with lar- well as those whose parents had T2DM. The FIN- ger waist circumference. DRISC may be used as a tool which would help The study Meijnikman et al. showed that FIN- general practitioners, in everyday work, to detect DRISC can serve as a good questionnaire for patients with T2DM risk factors and to encourage predicting visceral obesity (50.4% had pre-dia- them to change life style towards healthy habits. betes and 11.1% were diagnosed T2DM) (26). In our study, over half of respondents who practiced FUNDING physical activity had a low risk, while among No specific funding was received for this study. those who were not practicing physical activity, 26.6% and 11.5% had a high and very high risk, TRANSPARENCY DECLARATION respectively. A study conducted in Banja Luka in Conflicts of interest: None to declare. the population 25-75 age of years indicated that REFERENCES dicine, University of Banja Luka), 2019:165-83. 3. IDF Diabetes Atlas. 8th ed. International Diabetes 1. Kuzmanović R, Mirjanić D, Danelišen D, Vulić D, Federation. 2017 https://www.idf.org/e-library/epi- Šošić M, Vuković M, eds. Dijabetes mellitus: Sa- demiology-research/diabetes-atlas/134-idf-diabetes- vremena dostignuća i izazovi (Diabetes mellitus: atlas-8th-edition.html (15 November 2019) Contemporary Achievements and Challenges) [In 4. GBD 2016 Causes of Death Collaborators. Global Serbian]. Banja Luka: Akademija nauka i umjetnosti regional and national age-sex specific mortality Republike Srpske (Republic of Srpska Academy of for 264 causes of death, 1980–2016: a systematic Sciences and Arts); 2017. analysis for the Global Burden of Disease Study 2. Savić S. Dijabetes mellitus. In: Petrović V, Savić S, 2016. Lancet 2017; 390:1151–210. eds. Njega u primarnoj zdravstvenoj zaštiti, porodi- 5. Global Burden of Metabolic Risk Factors for Chro- ci i zajednici (Nursing in Primary Care, Family and nic Diseases Collaboration. Cardiovascular disea- Community) [In Serbian]. Banja Luka: Medicinski se. Chronic kidney disease. and diabetes mortality fakultet Univerziteta u Banjoj Luci (Faculty of Me- burden of cardiometabolic risk factors from 1980 to

521 Medicinski Glasnik, Volume 17, Number 2, August 2020

2010: a comparative risk assessment. Lancet Diabe- DiabVasc Dis Res 2005; 2:67–72. tes Endocrinol 2014; 2:634–47. 19. Schwarz PE, Li J, Reimann M, Schutte AE, 6. NCD Risk Factor Collaboration (NCD-RisC). Bergmann A, Hanefeld M, Bornstein SR, Schulze Worldwide trends in diabetes since 1980: a pooled J, Tuomilehto J, Lindstrom J. The Finnish Diabetes analysis of 751 population-based studies with 4.4 Risk Score is associated with insulin resistance and million participants. Lancet 2016; 387:1513–30. progression towards type 2 diabetes. J Clin Endocri- 7. International Diabetes Federation. IDF Diabetes nol Metab 2009; 94:920–6. Atlas. 8th edition. Brussels, Belgium: IDF. 2017 20. Zhang L, Zhang Z, Zhang Y, Hu G, Chen L. Eva- http://fmdiabetes.org/wp-content/uploads/2018/03/ luation of Finnish Diabetes Risk Score in screening IDF-2017.pdf (15 November 2019) undiagnosed diabetes and prediabetes among US 8. Šiljak S, Štrikić D, Jandrić Lj, Bratić R, Danojević adults by gender and race: NHANES 1999–2010. D, Grujić Rudić V, Petković V, Marinković J, Dimi- PloS One 2014; 9:e97865. trijević S, Kvaternik M, Janjić B, Šipovac V, Arse- 21. Jølle A, Midthjell K, Holmen J, Magnus Carlsen S, nović S. Analysis of Population Health in Republic Tuomilehto J, Håkon Bjørngaard J, Olav Åsvold B. of Srpska, 2014. Banja Luka: Public Health Institute, Validity of the FINDRISC as a prediction tool for Republic of Srpska; 2015 http://www.phi.rs.ba/pdf/ diabetes in a contemporary Norwegian population: publikacije/publikacija_zdr_stanje_2014.pdf (05 a 10-year follow-up of the HUNT study. BMJ Open October 2019) Diabetes Research and Care 2019; 7:e000769. 9. Šiljak S, Štrikić D, Jandrić Lj, Bratić R, Danojević 22. Atayoglu AT, Inanc N, Başmisirli E, Çapar AG. Eval- D, Grujić Rudić V, Petković V, Marinković J, Dimi- uation of the Finnish Diabetes Risk Score (FIND- trijević S, Kvaternik M, Janjić B, Šipovac V, Arse- RISC) for diabetes screening in Kayseri, Turkey. Prim nović S. Analysis of Population Health in Republic Care Diabetes 2020; pii:S1751-9918(19)30514-5. of Srpska, 2016. Public Health Institute, Republic of 23. Perelb P, Miranda J, Smeeth L. Diagnostic accuracy Srpska, 2017 http://www.phi.rs.ba/pdf/publikacije/ of the Finnish Diabetes Risk Score (FINDRISC) for Zdravstveno_stanje_stanovnistva_RS_2016_web. undiagnosed T2DM in Peruvian population. Primary pdf (5 October 2019) Care Diabetes 2018; 12: 517–25. 10. WHO. Diabetes country profiles. 2016. World He- 24. Vandersmissen GJM, Godderis L. Evaluation of alth Organization 2016 http://www.who.int/diabe- the Finnish Diabetes Risk Score (FINDRISC) for tes/publications/grd-2016/en/ (13 December 2019) diabetes screening in occupational health care. Int J 11. Savić S. Prevencija dijabetes melitusa tip 2. U: Occup Med Environ Health 2015; 28:587–91. Stanetić K, Račić M, Petrović V, Jatić Z, Savić S, 25. Bergmann A Li J, Wang L, Schulze J, Bornstein SR, Kusmuk S. Prevencija najčešćih hroničnih bolesti Schwarz PE. A simplified Finnish diabetes risk score (Prevention of the most common chronic diseases) to predict type 2 diabetes risk and disease evolution [In Serbian] Banja Luka: Narodna i univerzitetska in a German population. Horm Metab Res 2007; 39: biblioteka Republike Srpske. Udruženje doktora po- 677-82. rodične medicine Republike Srpske (National and 26. Meijnikman AS, De Block CE, Verrijken A, Mertens University Library of Republika Srpska. Association I, Corthouts B, Van Gaal LF. Screening for type 2 of Doctors of Family Medicine of Republic of Srp- diabetes mellitus in overweight and obese subjects ska); 2017:77–98. made easy by the FINDRISC Score. J Diabetes 12. Selph S, Dana T, Blazina I, Bougatsos C, Patel H, Complications 2016; S1056-8727(16): 30131-3. Chou R. Screening for type 2 diabetes mellitus: A 27. Rudić Grujić V, Grabež M, Petković V, Novako- systematic review for the US. Preventive Services vić B, Prtina A. Smanjen nivo fizičke aktivnosti Task Force. Ann Intern Med 2015; 162:765–76. kao činilac rizika za poremećaj tolerancije glukoze 13. American Diabetes Association. Stan- (Reduced level of physical activity as a risk factor dards of Medical Care in Diabetes-2018. for impaired glucose tolerance) In: Abstract book Lifestyle Management. Diabetes Care 2018; Treći međunarodni kongres Ekologija zdravlje rad 41(Suppl 1): S38–51. sport (Third International Congress Ecology Health 14. Brown N, Critchley J, Bogowicz P, Mayige M, Work Sport) [ In Serbian] Banja Luka. Udruženje Unwin N. Risk scores based on self-reported or ava- ''Zdravlje za sve“ (Associations ''Health for all“): ilable clinical data to detect undiagnosed type 2 dia- 2009,170-3. betes: a systematic review. Diabetes Res Clin Pract 28. Diego GA, Laura AJ, Miguel AC, Leonardo FN, 2012; 98:369–85. Paul AC, Patricio LJ. Evaluation of the Finnish Dia- 15. Khunti K, Mani H, Achana F, Cooper N, Gray LJ, betes Risk Score to predict type 2 diabetes mellitus Davies MJ. Systematic review and meta-analysis of in a Colombian population: A longitudinal observa- response rates and diagnostic yield of screening for tional study. Journal ListWorld J Diabetes 2015; 6: type 2 diabetes and those at high risk of diabetes; 1337-44. PloS One 2015; 10:e0135702. 29. Costa BL, Barrio F, Cabré JJ, Piñol JL, Cos X, Solé 16. Lindstrom J,Tuomilehto J. The diabetes risk score: a C, Bolíbar B, Basora J, Castell C, Solà-Morales O, practical tool to predict type 2 diabetes risk. Diabe- Salas-Salvadó J, Lindström J, Tuomilehto J. DE- tes Care 2003; 26:725–31. PLAN-CAT Research Group. Delaying progression 17. Schwarz PE, Li J, Lindstrom J, Tuomilehto J. To- to type 2 diabetes among high-risk Spanish individu- ols for predicting the risk of type 2 diabetes in daily als is feasible in real-life primary healthcare settings practice. Horm Metab Res. 2009; 41:86–97. using intensive lifestyle intervention. Diabetologia 18. Saaristo T, Peltonen M, Lindstrom J, Saarikoski L, 2012; 55:1319-28. Sundvall J, Eriksson JG. Tuomilehto J. Cross-secti- 30. Kolb H, Martin S. Environmental/lifestyle factors in onal evaluation of the Finnish Diabetes Risk Score: the pathogenesis and prevention of type 2 diabetes. a tool to identify undetected type 2 diabetes. Ab- BMC Med 2017;15:131. normal glucose tolerance and metabolic syndrome.

522 ORIGINAL ARTICLE

Glycaemia, renal function and body mass in men and women with type 2 diabetes

Marko Pirić1,2, Dunja Šojat1,2, Valentina Mišković1, Matej Šapina1,3,6, Zvonimir Popović1,4, Tatjana Bačun1,5

1Faculty of Medicine Osijek, Josip Juraj Strossmayer University of Osijek, 2Health Centre Osijek, 3Department of Pediatrics, University Hospital Centre Osijek, 4Department of Neurology, University Hospital Centre Osijek, 5Division of Endocrinology, Department of Internal Medicine, University Hospital Centre Osijek; Osijek, Croatia, 6Faculty of Dental medicine and health Osijek, Department of informatics and statistics,

ABSTRACT

Aim To determine parameters of glycaemic control, renal function and anthropometric measurements in patients with type 2 diabetes in family medicine offices and to examine whether there is a difference in these parameters between genders.

Methods This cross-sectional study included 136 patients of both genders diagnosed with type 2 diabetes, with an average age of 69.33±10.87. General and demographic data were collected, anthropometric measurements were taken, as well as data on fasting

plasma glucose, HbA1c and creatinine level from laboratory findings. Estimated glomerular filtration rate (eGFR) was calculated.

Corresponding author: Results The average results of fasting plasma glucose test were Tatjana Bačun 8.43 mmol/L, of HbA1c 7.15%, and of creatinine 79.00 µmol/L. In Faculty of Medicine Osijek, 19.12% of patients eGFR was <60 mL/min/1.73m2. 80.15% were Josip Juraj Strossmayer overweight and 38.97% had a body mass index (BMI) of 30 kg/ 2 University of Osijek, m or higher. Morbid obesity was recorded in 2.94% of patients. Females had a statistically significantly higher hip circumference Josipa Huttlera 4, 31 000 Osijek, Croatia (p=0.002) and BMI (p=0.019), while males had a statistically sig- Phone: +38531512800; nificantly higher waist-to-hip ratio (p=0.006) and BMI (p=0.007). Fax: +38531512833; E-mail: [email protected] Conclusion The patients did not reach the target value of fasting plasma glucose (<7mmol/L) and HbA1c was above recommended Marko Pirić ORCID ID: https://orcid. (<7%). Given that the patients were elderly, glucoregulation can org/0000-0003-0287-7861 be considered as adequate. The average eGFR classified the patients into G2 group (mildly decreased glomerular filtration). The mean BMI was not within the recommended values. It is important to educate patients on a healthy diet and physical activity, to con- Original submission: trol their weight, but also to choose medications that reduce weight 28 May 2020; in addition to glycaemic control. Accepted: Key words: blood glucose, body weight, glomerular filtration rate 19 June 2020 doi: 10.17392/1210-20

Med Glas (Zenica) 2020; 17(2): 523-529

523 Medicinski Glasnik, Volume 17, Number 2, August 2020

INTRODUCTION blem of overweight people, which was primarily related only to the developed countries, has now Diabetes is a chronic metabolic disease and one of become a major public health problem in the deve- the most important public health problems in the loping countries, but also in the least developed co- Republic of Croatia and worldwide (1). The Inter- untries (7). The main cause is considered to be the national Diabetes Federation (IDF) estimates that modern way of life in which physical activity level more than 420 million people worldwide suffer is unsatisfactory and the diet is based on industri- from diabetes, and the number of patients incre- ally processed food with a high percentage of car- ases every year (2). Currently, as many as 40% of bohydrates and saturated fatty acids (7). Adipose patients have not yet been diagnosed with diabe- tissue, which has long been considered exclusively tes, making the problem even more complex (3). a storehouse of energy, is an active endocrine organ Furthermore, among the diagnosed patients there that greatly affects insulin resistance, endothelial are many of those who do not get adequate therapy dysfunction and the development of chronic in- or do not take the recommended therapy, which flammation, and all of the above often leads to the often results in the development of microvascu- development of the metabolic syndrome – a com- lar and macrovascular complications of diabetes, bination of visceral obesity, arterial hypertension, requiring further efforts in managing complex cli- dyslipidaemia, and impaired glucose metabolism nical conditions (3). According to the data from (8). Many studies have confirmed the correlation the National Diabetes Registry (CroDiab), there between higher body mass index and a higher risk are 254,296 persons with diabetes registered in of developing diabetes, and special attention is paid the Republic of Croatia, and if patients who have to visceral obesity, i.e. the increased accumulation not yet been diagnosed are taken into account, the of visceral adipose tissue since it is metabolically total number reaches as many as 400,000 (3). The more active and more significantly involved in the complex pathophysiology of diabetes includes the creation of higher levels of proinflammatory cyto- interaction of various factors: increased lipolysis, kines (9,10). Consequently, it is evident that weight increased gluconeogenesis in the liver, increased loss and body mass index reduction, especially secretion of glucagon in pancreatic α-cells, de- waist circumference, prevents the progression of creased insulin secretion from pancreatic β-cells, prediabetes to diabetes, but also slows down the decreased incretin effect, increased renal glucose natural course of type 2 diabetes, having a benefi- reabsorption and disturbance of neurotransmitters cial effect on preventing or alleviating the compli- (4). That is why the treatment of diabetes requires cations of diabetes (10). individual determination of treatment objectives, but it also relies on universal recommendations The objectives of this research were to determi- primarily related to encouraging physical activity ne the parameters of glycaemic control (fasting and changing eating habits. plasma glucose, HbA1c), renal function and anthropometric measurements (body height, body The adequate glycaemic control in diabetes impli- mass, waist circumference, hip circumference, es the lowest possible glycaemic values without body mass index and waist-to-hip ratio) in pati- the occurrence of side effects, and treatment is ai- ents with type 2 diabetes in family medicine offi- med at achieving HbA1c values < 7% (5). Howe- ces and to examine whether there is a difference ver, target values have recently been determined in these parameters between males and females. individually, taking into account the patients’ age, Also we wanted to examine a quality of achie- the frequency of hypoglycaemia and comorbidities ving recommended goals for patients with type (5). Target HbA1c value of < 8% is recommended 2 diabetes in family medicine offices in Croatia. for elderly patients with advanced microvascular and macrovascular complications, numerous co- PATIENTS AND METHODS morbidities, and frequent hypoglycaemia (5,14). Patients and study design The correlation between being overweight and the risk of developing type 2 diabetes has been exami- This cross-sectional research was conducted in ned in detail, and evidence of this correlation is the two family medicine offices in the Health Centre occurrence of excessive weight gain in as many as Osijek during the period of three months (March 80% of patients with type 2 diabetes (6). The pro- to June 2019).

524 Pirić et al. Laboratory parameters in Type 2 diabetes

Patients were over 18 years of age, of both sexes, it is defined as a waist circumference ≥102 cm in diagnosed with type 2 diabetes. Before the rese- men and ≥88 cm in women (12). arch patients were provided with detailed information on the planned research and they read and Statistical analysis signed the informed consent document. Categorical data were presented as absolute and A total of 136 patients were included in the rese- relative frequencies. Numerical data were descri- arch, of which 55 were men and 81 were women, bed as the arithmetic mean and standard deviation with an average age of 69.33±10.87. Selection of in case of normal distribution and as the median the patients and the size of the sample ensured and interquartile range in other cases. Differen- the representativeness of the sample and objec- ces between categorical variables were tested by tive results. The patients were anonymised, with χ2 test and Fisher's exact test if necessary. Nume- each patient being assigned a unique code. rical variable differences between two indepen- Before the research, the patients were provided dent groups were tested by Student's t-test and with detailed information on the planned resear- Mann–Whitney U test in case of deviation from ch and they read and signed the informed consent normal distribution. Correlation between the va- document. riables was expressed as Pearson's correlation coefficient in cases where variables follow a normal The research was approved by the Ethics Com- distribution or Spearman's correlation coefficient mittee of the Health Centre Osijek and the Ethics in cases where variables do not follow a normal Committee of the Faculty of Medicine Osijek of distribution. All P values are two-tailed. The le- Josip Juraj Strossmayer University of Osijek. vel of significance was set at p=0.05. Methods RESULTS The following data were collected: demographic The research included 136 patients, of which 55 data (sex, age), data on duration of diabetes, data were males and 81 were females, with an avera- on body mass and height, data on waist and hip ge age of 69.33±10.87. The average duration of circumference. Body mass index (BMI) was cal- diabetes was 9 years, with an interquartile range culated by dividing the body weight in kilogram by of 4 to 14; no statistically significant difference in body height in meter square and waist-to-hip ratio duration of diabetes between the sexes was found was calculated by dividing waist circumference in (p=0.984). The recommended target values of fa- centimetres by hip circumference in centimetres. sting plasma glucose were not achieved in patients, Laboratory findings were used to collect data on but the recommended HbA1c values were achie- fasting plasma glucose, HbA c and creatinine le- 1 ved. No statistically significant difference was fo- vels (reference values 4.4 – 6.4 mmol/L, < 6 % und in values of fasting plasma glucose and HbA1c and 79 – 125 µmol/L, respectively), and estimated between males and females (p=0,402) (Table 1). glomerular filtration rate (eGFR) was calculated. Creatinine level was determined photometrically Table 1. Average values of fasting plasma glucose and average values of HbA1c –according to the genders using the Jaffe reaction by picric acid. Samples Parameter No of patients Mean (SD) p of serum, plasma (EDTA or heparin) and urine Plasma glucose (mmol/L) (24-hour or single sample) were used. The level Male 55 8.20 (2.45) 0.458 of glycated haemoglobin was determined by tur- Female 81 8.59 (3.31) Total 136 8.43 (2.99) bidimetric inhibition immunoassay, where the HbA1c (%) sample for analysis was whole blood with EDTA Male 55 7.20 (1.63) 0.402 as an anticoagulant. Visceral obesity was deter- Female 81 7.35 (1.75) Total 136 7.15 (1.60) mined according to the NCEP (National Cho- 1 SD, standard deviation; HbA1c, glycated haemoglobin A C lesterol Education Program) measures and IDF (International Diabetes Federation) measures. The average values of creatinine were within the According to the IDF visceral obesity is defined reference range. The average value of creatinine as a waist circumference ≥ 94 cm in male and ≥ in plasma in males was 89.00 (72.25 – 98.50) 80 cm in female (11) and according to the NCEP µmol/L, in females it was 71.00 (58.00 – 81.25) µmol/L (p<0.001).

525 Medicinski Glasnik, Volume 17, Number 2, August 2020

The estimated glomerular filtration rate in pati- in females than in males (p=0.019 and p=0.002, ents was 80.56±20.28 mL/min/1.73 m2, and the respectively). No statistically significant diffe- difference in values between the genders was not rence in waist circumference was found between statistically significant (p=0.536). A total of 110 males and females (p=0.693) (Table 4). (80.88%) patients had normal and mildly decre- Table 4. Body height and weight, body mass index (BMI), waist ased glomerular filtration, while 26 (19.2%) had circumference, hip circumference and waist-to-hip ratio in pa- eGFR < 60 mL/min/1.73 m2 (Table 2). tients with type 2 diabetes – comparison between the genders Parameter No of patients Mean (SD) p Table 2. Distribution of patients into glomerular filtration rate Body height (cm) (GFR) categories according to estimated (e)GFR values Male 55 173.29 (7.65) < 0.001 Female 81 159.06 (6.29) eGFR (mL/ No (%) of GFR categories Total 136 164.68 (9.81) min/1.73m2) patients Body mass (kg) G1 normal or high GF ≥ 90 47 (34.56) Male 55 84.20 (13.54) 0.007 G2 mildly decreased GF 60-89 63 (46.32) Female 81 77.09 (15.68) G3a mildly to moderately decreased GF 45-59 20 (14.71) Total 136 79.96 (15.21) G3b moderately to severely decreased GF 30-44 3 (2.20) BMI (kg/m2) G4 severely decreased GF 15-29 1 (0.74) Male 55 28.11 (4.27) 0.019 G5 kidney failure < 15 2 (1.47) Female 81 30.43 (6.27) Total 136 29.48 (5.64) The majority of patients, 109 (80.15%) were Waist circumference (cm) overweight; 53 (38.97%) patients met the cri- Male 55 94.36 (14.23) 0.693 teria for diagnosing obesity, i.e., had a BMI of Female 81 93.30 (16.18) Total 136 93.73 (15.38) 30 or higher. Morbid obesity was recorded in 4 Hip circumference (cm) (2.94%) patients. Male 55 103.33 (8.52) 0.002 Female 81 109.30 (12.45) Visceral obesity was determined in 66 (48.53%) Total 136 106.88 (11.38) patients if NCEP (National Cholesterol Education Waist-to-hip-ratio Program) measures were used as limit values and in Male 55 0.88 (0.13) 0.006 88 (64.71%) if IDF (International Diabetes Federa- Female 81 0.84 (0.13) Total 136 0.94 (0.17) tion) measures were used as limit values (Table 3). SD, standard deviation

Table 3. Body mass index (BMI) and waist circumference in DISCUSSION patients with type 2 diabetes Reference values No (%) of Parameter Patients with type 2 diabetes in our research do male/female (cm) patients not have fully satisfactory glycaemic control ta- 2 BMI (kg/m ) king only values of fasting plasma glucose and < 25 27 (19.85) 25 – 29.99 56 (41.18) HbA1c into account. Taking into account the 30 – 34.99 30 (22.06) mean age of patients, which was 69.33 ± 10.87 35 – 39.99 19 (13.97) and by applying an individualized approach in >40 4 (2.94) Waist circumference according to NCEP classification setting target values of fasting glucose values and Normal ≤ 102/≤ 88 70 (51.47) HbA1c, depending on the patients’ age, glycae- Increased >102/>88 66 (48.53) mic control can be considered appropriate. The Total 136 (100.00) Waist circumference according to IDF classification problem of achieving appropriate glycaemic re- Normal ≤94/≤80 48 (35.29) gulation exists not only in the Republic of Croa- Increased >94/>80 88 (64.71) tia but also in many other countries across Europe Total 136 (100.00) NCEP, National Cholesterol Education Programme; IDF, Internatio- and the World, and the values of parameters are nal Diabetes Federation variable in some countries. In England, 28-66.5% patients achieve target values of HbA1c ≤7.5%, The average body mass of patients in this research and in France, 24 to 52% (13,14). Patients who was 79.96 kg, and the average of BMI was 29.48 took part in our research achieved the mean va- kg/m2. Body height, body weight, and waist-to- lue of fasting plasma glucose of 8.43 mmol/L and hip ratio were statistically significantly higher value of HbA c of 7.15%, which is similar to the in males than in females (p<0.001, p=0.007, and 1 results of the study conducted in the Republic of p=0.006, respectively), while body mass index Croatia in 2015 including 10,275 patients treated and hip circumference were significantly higher in family medicine offices, where the mean va-

526 Pirić et al. Laboratory parameters in Type 2 diabetes

lue of fasting plasma glucose was 8.60 mmol/L had BMI of 30 or higher, which does not deviate and the mean value of HbA1c was 7.60% (15). from the available results on the BMI at the level of After a long period in which universal and strict the entire Republic of Croatia, according to which guidelines for glycaemic control in patients with 25.3% of men and 34.1% of women are conside- type 2 diabetes were recommended and applied, red obese. Obesity is nowadays one of the biggest the ADA/EASD 2018 guidelines recommend an public health issues in the Republic of Croatia and individualized approach to patients when deter- worldwide, and it is pathophysiologically related to mining target HbA1c values. HbA1c value of the development of cardiovascular diseases, diabe- <7% is recommended for most patients with dia- tes, but also many other chronic diseases (22). The betes, but special emphasis is placed on indivi- average value of BMI in our research was statisti- dual patient characteristics, where target HbA1c cally significantly higher in women than in men, value <8% is recommended for elderly patients but this research does not fully explain the cause with advanced microvascular and macrovascu- of this difference between the genders. It would be lar complications, numerous comorbidities and necessary to examine more thoroughly the possible frequent hypoglycaemia (16). differences in eating habits and the level of physical In addition to glucoregulation, it is extremely im- activity between men and women, and the possible portant to monitor renal function in patients with correlation between changes in diet and physical type 2 diabetes, and the best indicator of the pre- activity with regard to age. Higher BMI can be reserved renal function is the estimated glomerular lated to the patients’ sedentary lifestyle and poor filtration rate (eGFR) (17). Serum creatinine level eating habits, suggesting a healthy diet and weight is often used as an indirect indicator of glomeru- loss are generally recommended for the treatment lar filtration, but it largely depends on body mass of diabetes and are valid regardless of the achieved and a daily diet and is therefore not recommended glucoregulation parameters and therapy used for as a substitute for estimated glomerular filtration the treatment of diabetes (23). values (18). Estimated glomerular filtration rate There is a cause-and-effect relationship betwe- in our patients was 80.56 ± 20.28 mL/min/1.73 en the amount of visceral adipose tissue and the m2 measured by the CKD-EPI (Chronic Kidney occurrence of type 2 diabetes and cardiovascular Disease Epidemiology Collaboration) formula, diseases, and data on the existence of visceral and the difference in values between the sexes obesity is considered a better predictor of mor- was not statistically significant. According to the bidity and mortality from cardiovascular disea- KDIGO 2012 guidelines, the average measured ses and diabetes than BMI calculation (24,25). value of estimated glomerular filtration classifies According to the NCEP, 48.53% of patients in our patients into G2 group, i.e. the group of per- our research met the criteria for visceral obesity, sons with mildly decreased glomerular filtration while 64.71% of patients met the criteria accor- (19). The values of estimated glomerular filtra- ding to the classification of the IDF. A large-scale tion rate in our patients can be compared to the research conducted in the Republic of Croatia, values obtained. In the research conducted on also on the population of patients with type 2 dia- 146 patients in family medicine offices in Brazil, betes, had similar results. According to the NCEP 34.2% of patients had eGFR < 60 mL/min/1.73 visceral obesity was determined in 68.60% and m2, while in our country only 19.12% of patients according to the IDF in as many as 87.40% pati- had eGFR < 60 mL/min/1.73 m2 (20). ents (15). Taking into account the average values The average creatinine values were slightly of waist circumference it can be concluded that higher in men than in women and were within the they are not within the recommended values in reference interval for both genders confirming either men or women. the previously known physiological difference, Waist-to-hip ratio is an indicator of the abdominal which can be explained, among other things, by obesity, and the recommended values in our pati- lower muscle mass in women than men (21). ents were achieved in both genders. Several large- The average body mass index in our patients was scale European studies have classified higher hip 29.48 ± 5.64 kg/m2, and as many as 38.97% of pati- circumference as a protective factor for the deve- ents met the criteria for diagnosing obesity, i.e. they lopment of type 2 diabetes, but there are also studies

527 Medicinski Glasnik, Volume 17, Number 2, August 2020

whose results show exactly the opposite (26). The the importance of body mass on type 2 diabetes correlation of the higher hip circumference with a mellitus. In addition to a quality medical care, lower risk for developing type 2 diabetes is attribu- success of the treatment always depends on acti- ted to the presence of a larger amount of adipose ve participation of the patients. tissue in the gluteofemoral region (27,28). Adipose In conclusion, monitoring of patients with type tissue in the gluteofemoral region was shown to be 2 diabetes is complex and each patient should be more sensitive to insulin than visceral abdominal approached individually, while adhering to the adipose tissue, and therefore lower insulin levels, universal recommendations for reference values, lower plasma glucose levels, and lower HbA1c but also taking into account sex differences. In pa- levels were associated with the higher hip circum- tients with type 2 diabetes, good glucoregulation ference (27-30). A correlation between hip circum- is important, as well as monitoring of renal para- ference and risk for developing type 2 diabetes in meters, primarily the estimated glomerular filtrati- men was not found, which is explained by the fact on rate, both for the observation or monitoring of that, due to men’s body constitution, higher hip renal complications of diabetes and the possibility circumference usually does not indicate to a larger of correction of therapy. Regular measurements of amount of adipose tissue in that area, but to a larger body mass, body mass index, waist and hip cir- amount of muscle tissue that does not have the afo- cumference and waist-to-hip ratio are indispensa- rementioned adipose tissue characteristics (31,32). ble in any monitoring, which makes the basis for Our research also showed significant difference in proper education on nutrition and physical acti- the mean values of hip circumference comparing vity, as well as the use of medications that reduce to the women. weight in addition to glycaemic control. This study showed that our patients did not achieve recommended values for type 2 diabetes FUNDING mellitus but taking into account that the patients No specific funding was received for this study. were elderly, glucoregulation can be considered as adequate. However, the mean BMI was not TRANSPARENCY DECLARATION within the recommended values and it is very im- Conflicts of interest: None to declare. portant to educate patients in order to show them

REFERENCES 1. Topić E, Primorac D, Janković S, Štefanović M. 8. Proença ARG, Sertié RAL, Oliveira AC, Campana Medicinska biokemija i laboratorijska medicina. 2nd AB, Caminhotto RO, Chimin P, Lima FB. New con- ed. Zagreb: Medicinska naklada, 2018. cepts in white adipose tissue physiology. Braz J Med 2. International Diabetes Federation. Diabetes Atlas – Biol Res 2014; 47:192 – 205. 9th ed. https://www.diabetesatlas.org/en/ (05 April 9. Vinciguerra F, Baratta R, Farina MG, Tita P, Padova 2020) G, Vigneri R, Frittitta L. Very severely obese patients 3. Croatian Institute of Public Health. CroDiab registar. have a high prevalence of type 2 diabetes mellitus and https://www.hzjz.hr/sluzba-epidemiologija-prevencija- cardiovascular disease. Acta Diabetol 2013; 50:443–9. nezraraznih-bolesti/crodiab-registar/ (05 April 2020). 10. Freemantle N, Holmes J, Hockey A, Kumar S. How 4. De Fronzo RA. From the triumvirate to the omnius strong is the association between abdominal obesity octet: a new paradigm for the treatment of type 2 dia- and the incidence of type 2 diabetes? Int J Clin Prac betes mellitus. Diabetes 2009; 58:773-95. 2008; 62:1391–6. 5. Qaseem A, Vijan S, Snow V, Cross T, Weiss TB, 11. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome Owens DK. Glycemic control and type 2 diabetes – a new worldwide definition. A consensus statement mellitus: the optimal hemoglobin A1c targets. A gui- from the International Diabetes Federation. Diabet dance statement from the American College of Physi- Med 2006; 23:469–80. cians. Ann Intern Med 2007; 147:417-22. 12. National Institutes of Health. Clinical guidelines 6. International Diabetes Federation. IDF Diabetes on the identification, evaluation, and treatment of Atlas. 8th ed. https://www.idf.org/e-library/epidemio- overweight and obesity in adults – the evidence re- logy-research/diabetes-atlas.html (15 April 2020) port. Obesity Res 1998; 6:51–209. 7. Nolan CJ, Damm P, Prentki M. Type 2 diabetes across 13. Kanavos P, van den Aardweg S, Schurer W. Diabetes generations: from pathophysiology to prevention and expenditure, Burden of Disease and Management in 5 management. Lancet 2011; 378:169–81. EU Countries. LSE Health, London School of Econo- mics, 2012 http://www.lse.ac.uk/LSEHealthAndSoci- alCare/research/LSEHealth/MTRG/LSEDiabetesRe- port26Jan2012.pdf (30 March 2020)

528 Pirić et al. Laboratory parameters in Type 2 diabetes

14. Organisation for Economic Co-operation and De- 24. Matsuzawa Y. The role of fat topology in the risk of velopment. The diabetes epidemic and its im- disease. Int J Obes 2008; 32:83–92. pact on Europe http://www.oecd.org/els/health- 25. Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, systems/50080632.pdf (30 March 2020) Lanas F, McQueen M, Budaj A, Pais P, Varigos J, Lis- 15. Bralić Lang V. Klinička inercija liječnika obiteljske heng L. Effect of potentially modifiable risk factors medicine u regulaciji glikemije oboljelih od šećerne associated with myocardial infarction in 52 countries bolesti tip 2. Zagreb: School of Medicine, University (the INTERHEART study): case-control study. Lan- of Zagreb, 2015; Ph. D. thesis. cet 2004; 364:937–52. 16. American Diabetes Association. Standards of medical 26. Conway B, Xiang YB, Villegas R, Zhang X, Li H, Wu care in diabetes—2018 abridged for primary care pro- X, Yang G, Gao YT, Zheng W, Shu XO. Hip circum- viders. Clin Diabetes 2018; 36:14–37. ference and the risk of type 2 diabetes in middle-aged 17. Creatinin Clearence. El Camino Health. https://www. and elderly men and women: the Shanghai women elcaminohealth.org/library/creatinine-clearence (31 and Shanghai men’s health studies. Ann Epidemiol March 2020) 2011; 21:358–66. 18. Levey AS, Perrone RD, Madias NE. Serum creatinine 27. Dowling H, Fried S, Pi-Sunyer F. Insulin resistance in and renal function. Annu Rev Med 1988; 39:465-90. adipocytes of obese women: effects of body fat distri- 19. Levin A, Stevens PE, Bilous RW, Coresh J, De Fran- bution and race. Metabolism 1995; 44:987–95. cisco ALM, De Jong PE. Griffith KE, Hemmelgarn 28. Wahrenberg H, Lonnqvist F, Arner P. Mechanisms BR, Iseki K, Lamb EJ, Levey AS, Riella MC, Shlipak underlying regional differences in lipolysis in human MG, Wang H, White CT, Winearls CG. Kidney disea- adipose tissue. J Clin Invest 1989; 84:458–67. se: Improving global outcomes (KDIGO) CKD work 29. Seidell J, Perusse L, Despres J, Bouchard C. Waist and group. KDIGO 2012 clinical practice guideline for the hip circumferences have independent and opposite evaluation and management of chronic kidney disease. effects on cardiovascular disease risk factors: the Que- Kidney International Supplements 2013; 3:1-150. bec family study. Am J Clin Nutr 2001; 74:315–21. 20. Fontela PC, Winkelmann ER, Ott JN, Uggeri DP. 30. Snijder M, Dekker JM, Visser M, Yudkin JS, Ste- Estimated glomerular filtration rate in patients with houwer CDA, Bouter LM, Heine RJ, Nijpels G, Sei- type 2 diabetes mellitus. Rev Assoc Med Bras 2014; dell JC. Larger thigh and hip circumferences are asso- 60:531-7. ciated with better glucose tolerance: the Hoorn study. 21. National Kidney Foundation. K/DOQI clinical prac- Obes Res 2003; 11:104–11. tice guidelines for chronic kidney disease: evaluati- 31. Chowdhury B, Lantz H, Sjostrom L. Computed to- on, classification, and stratification. Am J Kidney Dis mography-determined body composition in relation 2002; 39(2 Suppl 1):S1-266. to cardiovascular risk factors in Indian and matched 22. Medanić D, Pucarin-Cvetković J. Obesity-a public Swedish males. Metabolism 1996; 45:634–44. health problem and challenge Acta Med Croatica 32. Parker E, Pereira M, Stevens J, Folsom A. Association 2012; 66: 347-55. of hip circumference with incident diabetes and coro- 23. Church TS, Cheng YJ, Earnest CP, Barlow CE, nary heart disease: the atherosclerosis risk in commu- Gibbons LW, Priest EL. Exercise capacity and body nities study. Am J Epidemiol 2009; 169:837–47. composition as predictors of mortality among men with diabetes. Diab Care 2004; 27:83-8.

529 ORIGINAL ARTICLE

Lifestyle risk factors and comorbidities of cancer patients in a country with limited resources

Nejra Mlačo1, Armin Šljivo1, Ahmed Mulać1, Amina Kurtović-Kozarić2, Anes Pašić3, Semir Bešlija3, Šejla Cerić4, Timur Cerić3

1School of Medicine, University of Sarajevo; Sarajevo, Bosnia and Herzegovina 2 Department of Clinical Pathology, Cytology and Human Genetics, Clinical Centre of the University of Sarajevo; Sarajevo, Bosnia and Herzegovina, 3Clinic of Oncology, Clinical Centre of the Uni- versity of Sarajevo; Sarajevo, Bosnia and Herzegovina, 4Clinic of Nuclear Medicine and Endocrinology, Clinical Centre of the University of Sarajevo; Sarajevo, Bosnia and Herzegovina

ABSTRACT

Aim To investigate quality of life and exposure to lifestyle risk factors of cancer patients in Bosnia and Herzegovina and a correlation of cancer type with lifestyle risk factors.

Methods This was a cross-sectional study conducted on 200 cancer patients from the Clinical Centre of the University of Sara- jevo. The respondents completed an anonymous questionnaire consisting of seven sections: basic patient information, physical activity, dietary habits including alternative medicine, tobacco use, alcohol consumption, anxiety, and comorbidities.

Corresponding author: Results A total of 150 (75%) patients were overweight with 113 Timur Cerić (56%) of them being less physically active after the confirmed di- Clinic of Oncology, Clinical Centre of the agnosis. After the diagnosis, 79 (40%) patients ate less food, and 154 (77%) healthier; 130 (65%) reported consumption of alter- University of Sarajevo native medicine and food supplements, 39 (30%) spent >1/4 of Bolnička 25, 71000 Sarajevo, average monthly salary on these products. Majority never consu- Bosnia and Herzegovina med alcohol, 135 (68%) and 101 (51%) patients reported history Phone: +387 33 297 289; of tobacco use. Being obese was an independent predictor for Fax: +387 33 298 526; colorectal carcinoma; being less obese was linked to a decreased Email: [email protected] risk of breast cancer diagnosis. Physical activity was linked to a decreased risk of lung cancer diagnosis. Many patients (122; 61%) Nejra Mlačo ORCID ID: https://orcid. reported having chronic comorbidities, mostly hypertension, while org/0000-0002-3299-6899 44 (22%) patients were proven to be clinically anxious.

Conclusion Our data suggest lack of public awareness of the consequences of unhealthy lifestyles. Risk factors such as alco- Original submission: hol consumption and tobacco use differed from other European countries. Significance of lifestyle changes after the diagnosis for 09 May 2020; reducing mortality and cancer recurrence requires further research. Revised submission: Prevention programs and more data are needed. 01 June 2020; Accepted: Key words: body weight, Bosnia and Herzegovina, diet, medical 03 June 2020 oncology, tobacco use doi: 10.17392/1199-20

Med Glas (Zenica) 2020; 17(2): 530-537

530 Mlačo et al. Lifestyle factors of cancer patients

INTRODUCTION 30% to 40% of patients with various types of cancer have some combination of mood disorders (13). Incidence of cancers has increased in the last years due to aging population, as well as greater Screening is of great importance for early can- exposure to cancer risk factors (1). According to cer detection. Screening programs significantly Globocan (2), estimated cancer incidence in 2018 improve public and individual health, but only if was 18 million cases worldwide, accounting for they are well organized and if they include ove- 9.5 million deaths. Data from Bosnia and Her- rall target population (14). These programs for zegovina (B&H) in 2018 estimated 14.385 new secondary cancer prevention are not a legal obli- cancer cases and 9.012 deaths, with lung and bre- gation in B&H and they are available for health ast cancer as the most common among males and insured people. Another problem are long wai- females, respectively (2). ting lists for diagnostic procedures in public health facilities. This is one of the few countries in Up to 95% of cancers are linked to environmen- Europe which does not have a national program tal and lifestyle factors (1), therefore preventi- for early detection of specific cancers. on is the best strategy in the fight against cancer. Overweight individuals are at a greater risk There is a significant lack of information about of multiple malignancies, including colorectal, the quality of life of cancer patients in B&H, pancreatic, liver, kidney, breast and prostate since this is a developing country with limited cancer (3,4). After the diagnosis and during tre- resources. There is a lack of similar data in the atment, patients often have reduced physical acti- region as well. Most of the studies related to qu- vity and well-being, as well as significant weight ality of life had small sample size for individual loss and cachexia, leading to malnutrition (5). cancers (15-17). Deficiency of statistical data di- Malnutrition can influence the effectiveness and sables health costs planning and rationalization. success of chemotherapy, radiotherapy, and can- The aim of this study was to investigate quality cer-related surgery due to changes in metaboli- of life of cancer patients in B&H and a correlati- sm, pharmacokinetics and healing dynamics (6). on of cancer type with exposure to lifestyle risk Various lifestyle behaviours, such as diet, smo- factors, with a specific focus on habits after di- king and alcohol consumption, have also been agnosis, considering parameters of physical and linked to the development of common cancers mental state. Our hypothesis is that some risk and may contribute to the overall survival rate factors and habits are unique in B&H and diffe- of patients (7,8). Tobacco use, primarily ciga- rent from other Balkan countries. rette smoking, is the largest exogenous carcino- PATIENTS AND METHODS gen linked to lung, oesophagus, larynx, mouth, kidney, bladder, liver, pancreas and other can- Patients and study design cers (9). Most studies evaluating the effects of smoking cessation after a cancer diagnosis on This was a cross-sectional study on patients trea- overall mortality, demonstrate a significant be- ted at the Clinic of Oncology, Clinical Centre of nefit of quitting smoking (10). Chronic alcohol the University of Sarajevo, B&H. The study was consumption is associated with increased risk conducted from 02 April to 30 October 2019. The for respiratory tract, upper digestive tract, liver, research was accepted and approved from the In- colon, rectum and breast cancer (11). Studies su- stitute for Scientific Research and Development ggest that adherence to a healthy dietary pattern of the Clinical Centre of the University of Sara- is inversely associated with overall mortality, jevo and completed in compliance with the Hel- whereas an unhealthy Western dietary pattern is sinki Declaration. positively associated with the risk of overall mor- The respondents completed an anonymous que- tality among cancer survivors (12). stionnaire, which took approximately 10 minutes Cancer patients often develop a chronic, clinically to complete. The study followed a stratified mul- significant syndrome of psychosocial distress ha- tistage random sampling design to represent the ving depressive disorders, anxiety, and reduced qu- socio-epidemiological characteristics of Bosnian ality of life. A previous meta-analysis reported that population. A total of 200 patients completed the questionnaire.

531 Medicinski Glasnik, Volume 17, Number 2, August 2020

Methods The vast majority reported living in an urban area, 166 (83.0%) The highest prevalence of malignan- The questionnaire consisted of 7 sections: the cies in our sample was for breast cancer patients basic information about patient (i.e. age, gender, 81 (41.8%), followed by colorectal carcinoma 43 occupation, date of oncological diagnosis etc.), (22.2%) and lung cancer 15 (7.7%) (Table 1). physical activity practices, food and alternative medicine consumption habit, smoking habit, Table 1. Distribution of cancer diagnoses in patient cohort alcohol intake anxiety level and comorbidities according to place of residence No (%) of patients among the patients. Type of cancer Urban Rural Total Body Mass Index (BMI) was calculated from pa- Breast cancer 70 (86.5) 11 (13.5 81 (41.8) tients height and weight. Having BMI below 18.5 Colorectal carcinoma 37 (86.1) 6 (13.9) 43 (22.2) kg/m2 was considered as underweight, from 18.5 Lung cancer 12 (80.0) 3 (20.0) 15 (7.7) to 24.9 kg/m2 as normal weight, from 25 to 29.9 Cervical cancer 11 (78.6) 3 (21.4) 14 (7.2) Ovarian cancer 7 (70.0) 3 (30.0) 10 (5.2) 2 2 kg/m as overweight and above 30 kg/m as obese. Prostate cancer 6 (85.8) 1 (14.2) 7 (3.4) Monthly expenditure on alternative medicine Pancreatic cancer 3 (60.0) 2 (40.0) 5 (2.6) Testicular cancer 5 (100) 0 (0.0) 5 (2.6) supplements compared to average salary was cal- Stomach cancer 1 (33.3) 2 (66.66) 3 (1.5) culated by dividing the reported expenditure on Melanoma 2 (100) 0 (0.0) 2 (1.0) these products with average salary in B&H. Endometrial cancer 2 (100) 0 (0.0) 2 (1.0) Larynx cancer 2 (100) 0 (0.0) 2 (1.0) To assess anxiety levels a Self-rating Anxiety Sca- Nasopharyngeal cancer 1 (100) 0 (0.0) 1 (0.5) le (SAS) was used. The 20-question SAS is based Cholangiocarcinoma 1 (100) 0 (0.0) 1 (0.5) on the Diagnostic and Statistical Manual of Men- Myoepithelioma 0 (0.0) 1 (100) 1 (0.5) Mesothelioma 1 (100) 0 (0.0) 1 (0.5) tal Disorders (DSM) definitions of anxiety (18). Thyroid cancer 1 (100) 0 (0.0) 1 (0.5) Respondents were asked to indicate how they felt according to “None or a little of time” (scored as The BMI ranged from 16.5 to 47.3 kg/m2 with 1), “Some of the time” (2), “Most of the time” (3) a median value of 25.9 kg/m2. Most of the pa- or “All of the time” (4). Total score from this scale tients were overweight, 150 (75%), followed by ranged from 20 to 80, and higher score indicated 39 (19.5%) obese and 5 (2.5%) underweight pa- the person with greater anxiety. Zung (18) stated tients. Only 6 (3%) patients had BMI within the that raw scores above 36 indicated that patients normal range. had “clinically significant” anxiety. Majority of patients reported doing low intensity Statistical analysis type of physical activities, which included long walks, low intensity cycling or yoga, 142 (71%). Descriptive statistics summarized the data. The Being diagnosed made a change in physical acti- results were shown in frequencies and percen- vity: 113 (56.5%) patients being less physically tages. Differences between age, gender, BMI active, 21 (10.5%) reported being involved in and diagnosis in respect of frequency were also more physical activity, and 65 (32.5%) reported assessed by conducting binary logistic regressi- no change in physical activity habits. Our study on and the Mann Whitney U test. Correlation showed that higher BMI significantly increased between parameters was assessed by Pearson the probability of developing colorectal carcino- Correlation test. ma (OR=1.063; 95% CI 1.001-1.130; p=0.047) and breast cancer (OR=1.463; 95% CI 1.263- RESULTS 1.694; p<0.001), and decreased the risk for de- In total, 206 patients treated at the Clinic of Onco- veloping lung cancer (OR=0.603; 95% CI 0.447 logy at Clinical Centre of the University of Sara- -0.814; p=0.001). Being physically active also jevo responded to the survey. After excluding six decreased the risk for developing lung cancer patients because of lack of data regarding their di- (OR=0.138; 95% CI 0.032-0.604; p=0.009). agnosis, the final sample consisted of 200 patients. Most of the patients consumed 3 meals/day, 153 Female patients were predominant, 139 (69.5%). (76.5%), followed by 36 (18%) patients who The mean age was 57.89±11.62 years (range 24- consumed 2 meals/day and 11 (5.5%) patients 83 years). who consumed only 1 meal/day (Table 2).

532 Mlačo et al. Lifestyle factors of cancer patients

Table 2. Food consumption habits of patients with cancer No (%) of patients Consumption Fruits and Refined sugars Milk and da- habit Red meat vegetables and sweets iry products Every day 168 (84.0) 27 (13.5) 3-4 days a week 18 (9.0) 12 (6.0) 144 (72.0) 171 (85.5) 1-2 days a week 12 (6.0) 47 (23.5) Never 2 (1.0) 114 (57.0) 56 (28.0) 29 (14.5)

After the diagnosis, 89 (44.5%) patients did not change the amount of food they ate per day, while 79 (39.5%) ate less, and 32 (16%) patients started Figure 2. Average daily cigarettes consumption of active and former smokers among the patients with cancer eating more. On the other hand, 154 (77%) patients reported that their diet was healthier after the diagnosis, and almost two thirds, 130 (65%) had (12.3%) consumed more than 7 glasses of wine. been using some form of supplements and alter- In total, 18 (27.1%) patients reported that they native medicine products. Breast cancer patients stopped drinking alcohol after the diagnosis and were more prone to consume alternative medica- 4 (6.2%) were drinking less. tions than other oncological patients (p=0.009). Regarding the association between cancer type Monthly expenditures on these supplements and alcohol consumption, breast cancer was si- was for the vast majority, 87 (43.4%) patients, gnificantly higher among former and current between 25 and 100 €, followed by 44 (22.1%) alcohol consumers (p<0.001). Among current patients who spent between 100 and 250 € alcohol consumers higher amounts of alcohol monthly in average (Figure 1). consumption (more than three glasses of wine) was significantly associated with breast cancer (p=0.001). A total of 122 (61%) patients reported having one chronic comorbidity and 63 (31.5%) two or more chronic comorbidities. Hypertension, diabetes and hyperlipidaemia were the most common, with 80 (43%), 28 (15%) and 18 (10%), respectively (Figure 3).

Figure 1. Monthly expenditure in euro (€) on alternative medicine in the patients with cancer Current tobacco use was reported by 57 (28.5%) patients, but more than half, 101 (50.5 %) had smoking history. The majority of smokers were tobacco users for more than 20 years, 116 (73.3%). The diagnosis changed the pattern of tobacco consumption, 34 (21.3%) reported that they quit smoking, 32 (20%) smoking less, and 2 (1.3%) reported smoking more (Figure 2). The majority patients never consumed alcohol, Figure 3. Comorbidities in the patients with cancer 135 (67.5%), 36 (18%) patients consumed it in the past and 29 (14.5%) patients were current alcohol consumers. In the group of current and past alco- Our results showed that 44 (22%) patients were hol consumers, the usual dosage of drinks per >36 on SAS scale (clinically anxious). Correlati- session was less than one glass of wine for 13 on between anxiety measured on SAS scale and (20%) patients, between one and two glasses of physical activity showed negative correlation (r= wine for 30 (46.2%) patients, between three to -0.173, n= 191; p<.0005), indicating that patients six glasses of wine for 14 (21.5%) patients, and 8 with higher anxiety were less physically active.

533 Medicinski Glasnik, Volume 17, Number 2, August 2020

DISCUSSION Consistent with our study, 56.1% patients in an Italian study (27) reported to have changed their This study revealed that patients with malignan- food habits since the diagnosis, either reducing cies in B&H usually make changes in daily habits or increasing one or more food items. A Finnish after cancer diagnosis. Most of them are poorly study showed that 31.9% breast cancer patients physically active, but strive to eat healthy food. had changed their diet after the diagnosis (28). The majority use alternative medicine substances Globally, patients tend to reduce food items more and have a history of smoking. It is a worrying frequently than increasing them. Dietary factors fact that even 61% of patients have comorbiditi- are related to one-third of cancers, especially hor- es, which indicates a high prevalence of chronic mone-related malignancies (29). Reduction of diseases in B&H. milk and dairy products has been reported in 61% Compared with national data, BMIs of our pa- of patients in Italian study, significantly less than tients are reasonably well. Data from 2016 (19) in our results (27). Reduction of dairy products is shows 56.4% overweight and 19.4% obese peo- not in line with the World Cancer Research Fund ple in adult population in B&H. Lack of physical (WCRF) recommendations (30), and probably activity and increased caloric intake are the main has the origin in media misinformation. explanations for these largely preventable mo- A study in the United States (31) showed that dern diseases (20). Cultural factors such as high around half of cancer survivors had smoked re- caloric and low nutritional meals are also contri- gularly prior to their cancer diagnosis, which buting to obesity (20). Obesity is also associated is similar as to our findings, and only 36% of with a higher risk of breast cancer, particularly in them quit smoking after their cancer diagnosis. postmenopausal women, and with worse disease Our data show smoking cessation among much outcome for women of all ages (20). Our study less patients, which could be linked to tobacco confirms this finding. smoking as cultural heritage and post-war peri- Interestingly, in neighbouring countries, Croatia od. Croatia and Serbia have slightly less current and Serbia, there are 25.6% and 21.1% obese peo- tobacco smokers, 33% and 32.4%, respectively, ple, respectively, which is more than in B&H (21). while there is 35.1% of current tobacco smokers These data support lower risk for various chronic in B&H, according to WHO (24). The standard diseases, including cancer, but in combination of care for all patients with cancer who use to- with other risk factors like smoking and anxiety, bacco should include either direct tobacco cessa- total risk cannot be considered less in B&H. tion support by the clinician or referral to structu- Being less involved in physical activity after the red smoking cessation program (32). diagnosis is usually associated with debilitating On the contrary, many of our patients never con- illness and exhausting treatment (22). Mental sumed alcohol-which could be linked to cultural state has a significant role in the quality of life believes of this area, which is shown also in low of cancer patients as well, considering there is prevalence of alcohol related cancer compared to an estimated prevalence of major depression and tobacco related cancers. Total alcohol per capi- minor depressive disorders in 16% and 22% of ta consumption, in litters of pure alcohol is 12.6 patients with cancer, respectively. (23). Our stu- in Serbia and 12.2 in Croatia; it is much less in dy proved that high BMI and obesity, as well as B&H, 7.1 (24). Our results are comparable with physical activity, are protective factors for lung the US study where 58% patients did not chan- cancer. Numerous studies confirm that physi- ge their drinking habits after the diagnosis (33). cally active persons have 20 to 50% lower risk Alcohol is also associated with 7–10% increased for lung cancer (24). Previous studies also proved risk for breast cancer for each 10 g (~1 drink) unexpected and still unexplained inverse relati- alcohol consumed daily by adult women (34). onship between obesity and lung cancer morta- Breast cancer survivors have the highest alcohol lity, called the obesity paradox (25). This fact is consumption among cancer survivors (33). Our confounded by smoking since individuals who study confirms these findings. smoke are generally less active and have lower A study which included more than 10.000 cancer body weight than nonsmokers (26). patients confirmed that 19% of patients showed

534 Mlačo et al. Lifestyle factors of cancer patients

clinical anxiety, which is close to our results, but Disease prevention programs are limited in they also proved clinical anxiety almost twice as B&H. Reducing specific lifestyle risk factors often in female patients (35). We did not find a demands joint efforts of health workers, media, statistically significant correlation between anxi- civil society organizations, the private sector and ety and gender. There may be a gender difference municipalities. in willingness to report distress (36) which was In conclusion, unhealthy lifestyle of cancer patients not noticed in our patients. Our study, as well as in B&H is particularly related to smoking history other studies (35), showed that younger patients and physical inactivity. Patients often tend to im- were more likely to be anxious. There is more prove life habits after the diagnosis. Nevertheless, disruption of everyday living in younger cancer they sometimes develop anxiety and depression, patients, whereas older patients may already have which, along with the consequences of the disease impairments in physical function and are cogni- and its treatment, reduces their quality of life. tively and emotionally better prepared to accept Data were collected from a small number of parti- the illness (37). The post-war period probably cipants, because patients were interviewed using has had a very significant contribution to anxiety a long survey which was time-consuming. These of our cancer population. findings require further research in larger studies, Data from Medicare for patients aged ≥65 years as their overall impact might be large. It is nece- in the United States indicate that 40% of patients ssary to investigate the relationship between risk with cancer have at least one other chronic con- factors exposure after the diagnosis and the risk dition, and 15% have two or more (37). The va- of subsequent disease recurrence, progression, or lues in our study are about twice as high and we death. Our data suggest lack of public awarene- included patients of all ages. High prevalence of ss on the consequences of unhealthy lifestyles. chronic diseases in B&H could be connected to Primary and secondary prevention programs in a complex constellation of social, economic and B&H must be improved. Awareness of healthy behavioural factors leading to increased exposu- lifestyles and the importance of regular screening re to lifestyle risk factors. Cancer patients with programs should be raised. This includes impro- comorbidity have poorer survival rate, quality of vement of legislation in the field of tobacco con- life, and a higher health care cost (37). trol, implementation of programs for smoking Awareness of early symptoms and signs of ma- and obesity prevention and healthy diet among lignancies is generally low in B&H. More than children and youth in schools, implementation of half of patients are diagnosed with metastatic campaigns on healthy lifestyle, better coopera- or locally advanced disease (38). Programs for tion with mass media for health promotion pur- early diagnosis of malignancies should be com- poses, as well as the development of mandatory prehensive and continuous, but in B&H they take screening programs accessible to the entire target the form of projects and are not part of the regu- population and implemented throughout the year. lar health care system. These programs include screening for cervical cancer, breast cancer and FUNDING colorectal cancer, as they represent three most si- No specific funding was received for this study. gnificant screenings according to EU recommendations as well (38). TRANSPARENCY DECLARATION Conflicts of interest: None to declare. REFERENCES 1. Anand P, Kunnumakkara AB, Kunnumakara AB, 3. Berger NA. Obesity and cancer pathogenesis. Ann N Sundaram C, Harikumar KB, Tharakan ST, Lai OS, Y Acad Sci 2014; 1311:57-76. Sung B, Aggarwal BB. Cancer is a preventable dise- 4. Allott EH, Masko EM, Freedland SJ. Obesity and ase that requires major lifestyle changes. Pharm Res prostate cancer: weighing the evidence. Eur Urol 2008; 25:2097-116. 2013; 63:800-9. 2. International Agency for Research on Cancer (IARC). 5. Von Haehling S, Anker SD. Cachexia as a major un- Global Cancer Observatory 2018. http://gco.iarc.fr derestimated and unmet medical need: facts and num- (26 February 2020) bers. J Cachexia Sarcopenia Muscle 2010; 1:1-5.

535 Medicinski Glasnik, Volume 17, Number 2, August 2020

6. Schulz KH, Patra S, Spielmann H, Klapdor S, Schlüter 19. World Health Organization (WHO). WHO Global K, Van Eckert S. Physical condition, nutritional sta- Health Observatory. Data Repository. Geneva: WHO, tus, fatigue, and quality of life in oncological out-pa- 2013 http://apps.who.int/gho/data/view.main (26 Fe- tients. SAGE Open Med 2017; 5:2050312117743674. bruary 2020) 7. Ligibel J. Lifestyle factors in cancer survivorship. J 20. Picon-ruiz M, Morata-tarifa C, Valle-goffin JJ, Fried- Clin Oncol. 2012; 30:3697-704. man ER, Slingerland JM. Obesity and adverse breast 8. Peters R, Ee N, Peters J, Beckett N, Booth A, Ro- cancer risk and outcome: mechanistic insights and ckwood K, Anstey KJ. Common risk factors for ma- strategies for intervention. CA Cancer J Clin 2017; jor noncommunicable disease, a systematic overview 67:378-97. of reviews and commentary: the implied potential for 21. World Health Organization (WHO). Cancer country targeted risk reduction. Ther Adv Chronic Dis 2019; profiles 2014 https://www.who.int/cancer/country- 10:2040622319880392. profiles/en/ (26 February 2020) 9. National Center for Chronic Disease Prevention 22. Holmes MD, Chen WY, Feskanich D, Kroenke CH, and Health Promotion (US) Office on Smoking and Colditz GA. Physical activity and survival after breast Health. The Health Consequences of Smoking—50 cancer diagnosis. JAMA 2005; 293:2479-86. Years of Progress: A Report of the Surgeon General. 23. Mitchell AJ, Chan M, Bhatti H, Halton M, Grassi Atlanta (GA): Centers for Disease Control and Pre- L, Johansen C, Meader N. Prevalence of depressi- vention (US), 2014 https://www.ncbi.nlm.nih.gov/ on, anxiety, and adjustment disorder in oncological, books/NBK179276/ (26 February 2020) haematological, and palliative-care settings: a meta- 10. Warren GW, Alberg A. Smoking cessation after a can- analysis of 94 interview-based studies. Lancet Oncol cer diagnosis and survival in cancer patients. J Clin 2011; 12:160-74. Oncol 2018; 36:1561. 24. Patel AV, Carter BD, Stevens VL, Gaudet MM, 11. Seitz HK, Becker P. Alcohol metabolism and cancer Campbell PT, Gapstur SM. The relationship betwe- risk. Alcohol Res Health 2007; 30: 38-41, 44-7. en physical activity, obesity, and lung cancer risk by 12. Schwedhelm C, Boeing H, Hoffmann G, Aleksandro- smoking status in a large prospective cohort of US va K, Schwingshackl L. Effect of diet on mortality adults. Cancer Causes Control 2017; 28:1357-68. and cancer recurrence among cancer survivors: a 25. Zhang X, Liu Y, Shao H, Zheng X. Obesity Paradox systematic review and meta-analysis of cohort studi- in Lung Cancer Prognosis: Evolving Biological Insi- es. Nutr Rev 2016; 74:737-48. ghts and Clinical Implications. J Thor Oncol 2017; 13. Holland JC, Andersen B, Breitbart WS, Buchmann 12:1478-88. LO, Compas B, Deshields TL, Dudley MM, Fleish- 26. Audrain-McGovern J, Benowitz NL. Cigarette smo- man S, Fulcher CD, Greenberg DB, Greiner CB, Han- king, nicotine, and body weight. Clin Pharmacol Ther dzo GF, Hoofring L, Hoover C, Jacobsen PB, Kva- 2011; 90:164-8. le E, Levy MH, Loscalzo MJ, McAllister-Black R, 27. Gavazzi C, Sieri S, Traclò F, Sproviero A, Vandoni Mechanic KY, Palesh O, Pazar JP, Riba MB, Roper G, Ricci R, Stragliotto S, Zagonel V; Changes in food K, Valentine AD, Wagner LI, Zevon MA, McMillian habits in cancer patients in Italy: a survey. AIOM- NR, Freedman-Cass DA. Distress Management: Cli- SINPE-FAVO. Nutrition 2018; 55-56:140-5. nical practice guidelines in oncology. J Natl Compr 28. Salminen EK, Lagström HK, Heikkilä S, Salminen Canc Netw 2013; 11:190-209. S. Does breast cancer change patients' dietary habits? 14. Institute of Medicine (US) Committee on Perinatal Eur J Clin Nutr 2000; 54:844-8. Transmission of HIV; National Research Council 29. Levi F. Cancer prevention: epidemiology and per- (US) and Institute of Medicine (US) Board on Chil- spectives. Eur J Cancer 1999; 35:1046-58. dren, Youth, and Families; Stoto MA, Almario DA, 30. World Cancer Research Fund International. Meat, fish McCormick MC, editors. Reducing the Odds: Pre- and dairy products and the risk of cancer. World Can- venting Perinatal Transmission of HIV In The United cer Research Fund, 2018 https://www.wcrf.org/sites/ States. Washington (DC): National Academies Press default/files/Meat-Fish-and-Dairy-products.pdf (26 (US); 1999. 2, Public Health Screening Programs February 2020) https://www.ncbi.nlm.nih.gov/books/NBK230552/ 31. Tseng TS, Lin HY, Moody-Thomas S, Martin M, (28 May 2020) Chen T. Who tended to continue smoking after cancer 15. Salibasic M, Delibegovic S. The Quality of Life and diagnosis: the national health and nutrition examina- Degree of Depression of Patients Suffering from Bre- tion survey 1999-2008. BMC Public Health 2012; ast Cancer. Med Arch. 2018; 72:202-5. 12:784. 16. Trninić Z, Vidacak A, Vrhovac J, Petrov B, Setka V. 32. Warren GW, Marshall JR, Cummings KM, Toll Quality of life after colorectal cancer surgery in pa- BA, Gritz ER, Hutson A, Dibaj S, Herbst R, Mulshine tients from University Clinical Hospital Mostar, Bo- JL, Hanna N, Dresler CA. Addressing tobacco use in snia and Herzegovina. Coll Antropol. 2009; 33 Suppl patients with cancer: a survey of American Society 2:1-5. of Clinical Oncology members. J Oncol Pract 2013; 17. Glavić Z, Galić S, Krip M. Quality of life and per- 9:258-62. sonality traits in patients with colorectal cancer. 33. Corrao G, Bagnardi V, Zambon A, Arico S. Exploring Psychiatr Danub. 2014; 26:172-80. the dose-response relationship between alcohol con- 18. Dunstan DA, Scott N. Assigning Clinical Significan- sumption and the risk of several alcohol-related con- ce and Symptom Severity Using the Zung Scales: ditions: a meta-analysis. Addiction 1999; 94:1551-73. Levels of Misclassification Arising from Confusion 34. Liu Y, Nguyen N, Colditz GA. Links between alcohol between Index and Raw Scores. Depress Res Treat consumption and breast cancer: a look at the eviden- 2018; 2018: 9250972. ce. Womens Health (Lond) 2015; 11:65-77.

536 Mlačo et al. Lifestyle factors of cancer patients

35. Saquib J, Madlensky L, Kealey S, Saquib N, Natara- 37. Sarfati D, Koczwara B, Jackson C. The impact of co- jan L, Newman VA, Patterson RE, Pierce JP. Classifi- morbidity on cancer and its treatment. CA Cancer J cation of CAM use and its correlates in patients with Clin 2016; 66:337-50. early-stage breast cancer. Integr CancerTher 2011; 38. Federal Institute for Development Programming, 10:138-47. Federation of Bosnia and Herzegovina. Strategija za 36. Linden W, Vodermaier A, Mackenzie R, Greig D. prevenciju, tretman i kontrolu malignih neoplazmi u Anxiety and depression after cancer diagnosis: preva- Federaciji Bosne i Hercegovine 2012-2020 (Strategi- lence rates by cancer type, gender, and age. J Affect es for prevention, treatment and control of malignant Disord 2012; 141:343-51. neoplasms in Federation of Bosnia and Herzegovina) [in Bosnian] http://fzzpr.gov.ba/bs/docs/24/3/strate- ski-dokumenti-fbih (28 May 2020)

537 ORIGINAL ARTICLE

Factors associated with overweight and obesity in preschool children Amira Kurspahić-Mujčić1, Amra Mujčić2

1Department of Social Medicine, School of Medicine, University of Sarajevo, 2Secondary Medical School; Sarajevo, Bosnia and Herzegovina

ABSTRACT

Aim To determine factors associated with overweight/obesity in preschool children.

Methods This cross-sectional study was carried out in paediatric outpatient departments of the Public Institution Primary Health Care Centre of Canton Sarajevo, Bosnia and Herzegovina. The study included 300 preschool children who were divided into two groups: normal weight (n=150) and overweight/obese (n=150). Children were weighed and measured and body mass index was calculated. Data concerning potential childhood overweight/obesity risk factors (sociodemographic characteristics of child and mother, physical activity and sedentary behaviour of the child) were collected using a designed questionnaire filled in by the mothers. Logistic regression analysis was used to estimate the association between potential risk factors and overweight/obesity Corresponding author: in preschool children. Amira Kurspahić-Mujčić School of Medicine, Results Male gender of the children (p=0.043) and maternal University of Sarajevo overweight (p=0.000) were positively associated with child overweight/obesity. Male children were 1.6 times more likely to Čekaluša 90, 71000 Sarajevo, be overweight/obese than female (95% CI: 1.01-2.53). Children of Bosnia and Herzegovina overweight mothers were 3.34 times more likely to be overweight/ Phone: +387 33 202 051; obese than children with mothers of normal weight (95% CI: 1.77- Fax: +387 33 202 051; 6.28). Physical activity <60 minutes/day ) (p=0.014), screen time Email: [email protected] >180 minutes/day (p=0.020), regular snacking while watching ORCID ID: http://www.orcid.org/0000- television, using computer/tablet/ mobile phones (p=0.000) were associated with overweight/obesity status of the preschool chil- 0002-0741-5490 dren.

Conclusion Public health programs that aim to reduce overweight/ Original submission: obesity in preschool children should mainly help mothers understand the serious risk of childhood obesity and the importance of 01 April 2020; them creating a healthy lifestyle in childhood. Accepted: 05 May 2020 Key words: paediatric obesity, screen time, exercise, mothers doi: 10.17392/1175-20

Med Glas (Zenica) 2020; 17(2): 538-543

538 Kurspahić-Mujčić et al. Overweight /obesity in preschoolers

INTRODUCTION The aim of this study was to determine factors associated with overweight/obesity in preschool The obesity epidemic in preschool children is children in Sarajevo Canton. on the rise and represents a serious public health challenge (1). Recent evidence suggests that in EXAMINEES AND METHODS obese children most of their excess weight is gained before the age of five years (2). Studies have Examinees and study design indicated that childhood obesity is associated This cross-sectional study was carried out at the with the development of noninsulin-dependent Paediatric Outpatient Department of the Public diabetes at an early age (3). Obese children are Institution Primary Health Care Centre of Canton more likely to suffer mental health issues and Sarajevo, Bosnia and Herzegovina (B&H) in the psychosocial problems and even discrimination period March – July 2019. (4). Childhood obesity predicts obesity in adult- The study included preschool children. According hood (5) and later cardiovascular disease (6). to Kail (19), preschoolers are defined as children Studies have shown that many factors can po- between 4 and 6 years of age. A total of 300 chil- tentially be associated with overweight/obesity dren (150 of normal weight and 150 overweight/ in preschool children (7,8). From this, factors obese) were selected from patients who used which are on the maternal side were marital sta- health care services at the Primary Health Care tus, level of education, employment and weight Centre during the study period. It was combined status (9). Some previous studies found that overweight and obese children, based on pre- maternal characteristics had a stronger effect on vious research suggesting that children who are children’s weight status than paternal characte- overweight are at risk of becoming obese (20). ristics (10). This effect is mainly accounted for The inclusion criteria were children aged 4-6 by a combination of genetic and behavioural years who had a medical record in the Primary factors, as mother has a strong influence on their Health Care Centre of the Sarajevo Canton, wit- children’s eating and drinking habits, as well as hout contraindications for physical activity. The on their activity levels (11). exclusion criteria were children younger than 4 One of the factors related to overweight and obe- or older than 6 years, children who did not have sity in preschool children is insufficient physical medical records at the Primary Health Care Cen- activity (12). Today, young children are sedentary tre of the Sarajevo Canton, known contraindica- for a significant portion of the day (13). Most tions for physical activity. studies focused on screen time as an indicator of The Ethics Committee of the School of Medici- sedentary behaviour (14). There is solid eviden- ne, University of Sarajevo, approved the study. ce that increased screen time has been associated For this research, a written consent of the Gene- with overweight, obesity in preschool children ral Director of the Primary Health Care Centre of within multiple cohorts and studies (15,16). the Sarajevo Canton was obtained. An informed Marsh et al. reported that there was strong evi- consent for participation in the study was taken dence that screen time in the absence of food ad- from all mothers. vertising was associated with increased dietary intake compared with non-screen behaviour (17). Methods Considering the rapid rise in prevalence of early- A questionnaire was administered to mothers onset obesity in preschool children, and its links accompanying their children to the Primary He- to later childhood and adult obesity, particular alth Care Centre. The questionnaires were given attention should be paid to identifying the early to mothers while they were waiting for the appo- life risk factors for overweight /obesity. intment. The sociodemography section requested No studies examined the factors associated with information on the child and mother. There were overweight and obesity (sociodemographic cha- two items about the child (age, gender) and three racteristics of child and mother, physical activity items concerning the mother: marital status (li- and sedentary behaviour of child) in preschool ving with a spouse/co-habiting subject), single, children in Bosnia and Herzegovina, but only divorced and widowed), level of education (in- those which examined obesity in elementary and complete elementary school, completed elemen- secondary school students (18). tary school, completed secondary school, com-

539 Medicinski Glasnik, Volume 17, Number 2, August 2020

pleted high school/university) and employment between normal weight and overweight/obese status based on mothers’ stated employment du- group was performed by χ2 test. ring an interview (employed or not employed). The individual effects of potential childhood The mothers reported average minutes that the- overweight/obesity risk factors on the presence of ir child spent in moderate to vigorous intensity child overweigh/obesity were obtained by logistic physical activity (that makes him/her out of bre- regression analysis through the calculation of the ath or warmer than usual) daily. According to the odds ratio (OR). The level of significance was set World Health Organization (WHO) physical acti- at p<0.05, and the confidence level of 95%. vity guidelines, preschool children who are physically active for at least 60 minutes of moderate-to RESULTS vigorous intensity physical activity (MVPA) da- The study evaluated 300 children in two groups ily were considered as compliant with the WHO of 150 each (i.e. normal weight and overweight/ physical activity recommendations (21). obese). The mothers reported average minutes per day their More male children were in the overweight/obe- child sat and watched television, and also the time se than in the normal weight group, 82 (55.0%) the child used a computer/ tablet/ mobile phones. and 65 (43.3%). More female children were in The specific questions asked were, ''Over the past the normal weight group than in the overweight/ 30 days, for how many minutes on average per day obese, 85 (56.7%) and 67 (45.0%). did the child sit and watch TV?“ and ''Over the past Most of the children in the normal weight and 30 days, for how many minutes on average per day overweight/obese group, 124 and 123 (82.7% and did the child use a computer/tablet/mobile phone?" 82.0%, respectively) resided in 2-parent househol- We summed up average minutes of television and ds, with both their biological mother and father. computers/ tablet/ mobile phones to estimate "scre- Maternal education level in the normal weight en time." We defined high screen time as greater and overweight/obese group was not significantly than 120 minutes (2 hours) per day (22). different (p=0.448). Maternal education level Variables “snacks in front of TV” and “snacks in included mainly high school/ university educati- front of computers/tablet/mobile phones ” were on, 86 (57.3%) in the normal weight group and grouped in “snacks in front of screens”, with 92 (61.3%) in the overweight/ obese group. three options of answers: never, sometimes, and Maternal employment in the normal weight and always eats snacks in front of screens. overweight/obese group was not significantly The mothers were asked to indicate their own height different (p=0.803). Slightly more children of and weight. Maternal body mass index (BMI) was employed mothers were in the overweight/obese calculated as underweight (BMI <18.5), normal- group than in the normal weight, 105 (70.0) and weight (18.5< BMI <24.9), overweight (BMI ≥25- 102 (68.0), respectively. 29.9), and obese (BMI ≥30). Children were weighed Among normal weight children, 16 (10.7%) li- and measured, and BMI was calculated. Weight was ved with an overweight mother, and five (3.3%) measured with the child in light clothing, to the ne- lived with a mother who was obese. Among arest 0.1 kg. Standing height was recorded without overweight/obese children, 43 (28.7%) lived shoes to the nearest 0.1 cm. The BMI was calculated with an overweight mother, and six (4.0%) lived as weight (kg) divided by height in square meters with a mother who was obese. (m2). The BMI categories were defined according to the WHO reference curves for different age and A total of 87 (58.0%) children in the normal weight gender groups (23): normal weight (BMI between group and 65 (43.3%) in the overweight/obese gro- 5th-85thpercentiles), overweight (BMI between up spent at least 60 minutes each day in MVPA. 85th-95th percentiles), and obese (BMI ≥85th-95th A total of 14 (9.3%) children in the normal weight percentiles). The mothers were also asked to report group and eight (5.3%) in the overweight/obese their weight and height. group achieved screen time recommendations (<2 hr per day). Statistical analysis Four times more children with regular snacking Testing of the difference in the distribution of po- while watching television, using computer/tablet/ tential childhood overweight/obesity risk factors mobile phones was in the overweight/obese gro-

540 Kurspahić-Mujčić et al. Overweight /obesity in preschoolers

Table 1. Distribution and association of risk factors and overweight/obesity in preschool children

No (%) of children Logistic regression analysis Potential risk factor p Normal weight Overweight / obese OR (95% CI) p Child gender Female 85 (56.7) 67 (45.0) 0.049 1 Male 65 (43.3) 82 (55.0) 1.6 (1.01-2.53) 0.043 Family status Two-parent family (married) 124 (82.7) 123 (82.0) 0.582 1 Single parent family (divorced) 25 (16.7) 27 (18.0) 1.09 (0.59-1.98) 0.780 Single parent family (single) 1 (0.7) 0 (0.0) 0.34 (0.01-8.33) 0.505 Maternal education level Completed high school /university 86 (57.3) 92 (61.3) 0.448 1 Completed secondary school 64 (42.7) 57 (38.0) 1.2 (0.46-1.91) 0.437 Incomplete/completed elementary school 0 (0.0) 1 (0.7) 0.36 (0.01-8.87) 0.529 Maternal employment Yes 102 (68.0) 105 (70.0) 0.803 1 No 48 (32.0) 45 (30.0) 1.09 (0.68-1.79) 0.7081 Maternal body mass index status Normal weight 123 (82.0) 99 (66.0) 0.001 1 Underweight 6 (4.0) 2 (1.3) 0.41 (0.08-2.09) 0.286 Overweight 16 (10.7) 43 (28.7) 3.34 (1.77-6.28) 0.000 Obese 5 (3.3) 6 (4.0) 1.49 (0.44-5.03) 0.519 Physical activity (minutes/day) ≥60 87 (58.0) 65 (43.3) 0.039 1 30-59 59 (39.3) 79 (52.7) 1.79 (1.12-2.86) 0.014 < 30 4 (2.7) 6 (4.0) 2.01 (0.54-7.41) 0.295 Sedentary behaviour Screen time (minutes/day) < 120 14 (9.3) 8 (5.3) 0.022 1 120-180 118 (78.7) 107 (71.3) 1.59 (0.64-3.93) 0.318 > 180 18 (12.0) 35 (23.3) 3.40 (1.20-9.61) 0.020 Snacks in front of screens Never 49 (32.7) 14 (9.3) 0.000 1 Always 14 (9.3) 56 (37.3) 3.22 (1.65-6.27) 0.000 Sometimes 87 (58.0) 80 (53.3) 1.14 (0.49-2.63) 0.753 OR, odds ratio; CI, confidence interval up, 56 (37.3%), than in the normal weight group, DISCUSSION 14 (9.3) (Table 1). This study investigated the factors associated with Male gender of the children (p=0.043) and mater- overweight/obesity in preschool children in Sa- nal overweight (p=0.000) were positively asso- rajevo Canton. The results of this study showed ciated with child overweight/obesity. Male chil- that maternal overweight had a significant asso- dren were 1.6 times more likely to be overweight/ ciation with child overweight/obesity. Similar to obese as female (95% CI: 1.01-2.53). Children our finding, Janjua et al. found that children of of overweight mothers were 3.34 times more li- overweight and obese mothers were more likely kely to be overweight/obese than children with to be obese and overweight at the age of 5 (24). mothers of normal weight (95% CI: 1.77-6.28). Those results match previous studies and suggest Family status, maternal education level and ma- that maternal excess weight has an important role ternal employment were not significantly asso- on child’s BMI (25,26). This relationship can be ciated with overweight/obesity status of the pres- explained through both genetic and lifestyle fac- chool children. tors passed on from mothers to their children (27). Physical activity <60 minutes/day ) (p=0.014), In this study, male children are found more likely screen time >180 minutes/day (p=0.020), regular than female to be overweight/obese. Overweight snacking while watching television, using com- and obesity rates in preschool children in the Eu- puter/tablet/mobile phones (p=0.000) were asso- ropean Union are generally higher in boys than in ciated with overweight/obesity status of the pres- girls (28). This may be attributable to different body chool children. Regular snacking in front of the standards for boys and girls. For example, mothers screens is a practice that triples children’s risk of may think that boys should eat more to grow faster, becoming obese (95% CI: 1.65-6.27). (Table 1). which encourages overeating and can lead to obesity. In addition, the prevalence of overweight and

541 Medicinski Glasnik, Volume 17, Number 2, August 2020

obesity in Turkish children aged less than 7 years related to consumption of unhealthy food thro- was without gender difference (29). ughout the day (40). Studies have revealed that Research on relationship between maternal edu- children consume a large proportion of their daily cation level and overweight /obesity in prescho- calories and meals while watching screen media ol children has not provided a definite answer. (41). In this study, regular snacking in front of Some studies, as well as this one, have not found screens is a practice that triples children’s risk for a statistically significant relationship (30). Pre- becoming obese. vious studies across 11 European countries have This study aimed to focus on the role of sociode- indicated that low maternal education could yield mographic characteristics of child and mother, a substantial risk of early childhood obesity (31). physical activity and sedentary behaviour of the Interestingly, the association between maternal child, but many other factors of the family envi- education and children’s weight status in China ronment might influence children’ obesity. The- is different from that in western countries. In refore, future research is needed to explore other China, the obesity rate among children with high potential factors associated with overweight/obe- maternal education is higher than that of lower sity in preschool children that were not included maternal education (32). in the present study. A growing research literature has explored the In conclusion, this study demonstrated a signi- relationship between maternal employment and ficant association between maternal overweight children’s body mass index (BMI) (33,34). A syste- and children’s weight status. This suggests that matic review using six studies from the United Sta- early interventions for childhood obesity sho- tes of America, the United Kingdom, Germany and uld focus on children of overweight or obese Japan, concludes that maternal employment is asso- mothers. In addition, factors including screen ciated with an increased risk of overweight /obe- time viewing and physical activity are also asso- sity for children (35). The employed mothers spent ciated with children’s weight status. Educating less time on meal preparation and healthy weight parents, specially mothers on the screen-use re- management than unemployed mothers as docu- commendations and the negative health risks of mented by Savage et al. (36). In this study maternal excessive screen use, as well recommendations employment was not significantly associated with for physical activity may improve parental awa- overweight/obesity status of the preschool children. reness and monitoring of their children’s seden- Regarding other risk factors for overweight / tary behaviour and physical activity. obesity, our results fit into known patterns; as in FUNDING other studies, screen time viewing was positively (37,38), whereas physical activity was inversely No funding was received for this study. associated with childhood overweight or obesity in preschool age children (39). TRANSPARENCY DECLARATION The number of hours in front of TV is directly Competing interests: None to declare.

REFERENCES 6. Nadeau KJ, Maahs DM, Daniels SR, Eckel RH. Childhood obesity and cardiovascular disease: links 1. Güngör NK. Overweight and obesity in children and and prevention strategies. Nat Rev Cardiol 2011; adolescents. J Clin Res Pediatr Endocrinol 2014; 8:513–25. 6:129–43. 7. Toselli S, Zaccagni L, Celenza F, Albertini A, Gu- 2. Lanigan J, Barber S, Singhal A. Prevention of obe- aldi-Russo E. Risk factors of overweight and obesity in preschool children. Proc Nutr Soc 2010; sity among preschool children with different ethnic 69:204-10. background. Endocrine 2015; 49:717–25. 3. Pulgaron ER, Delamater AM. Obesity and type 2 8. Rossen LM, Talih M. Social determinants of dispa- diabetes in children: epidemiology and treatment. rities in weight among US children and adolescents. Curr Diab Rep 2014; 14:508. Ann Epidemiol 2014; 24:705–13. 4. Tomiyama JA, Carr D, Granberg EM, Major B, Ro- 9. Barroso CS, Roncancio A, Hinojosa MB, Reifsni- binson E, Sutin AR, Brewis A. How and why weight der E. The association between early childhood stigma drives the obesity ‘epidemic’ and harms he- overweight and maternal factors. Child Obes 2012; alth. BMC Medicine 2018; 16:123. 8:449–54. 5. Simmonds M, Llewellyn A, Owen CG, Woolacott 10. Keane E, Layte R, Harrington J, Kearney PM, Perry N. Predicting adult obesity from childhood obesity: IJ. Measured parental weight status and familial a systematic review and meta-analysis. Obes Rev socio-economic status correlates with childhood 2016; 17:95-107. overweight and obesity at age 9. PLoS ONE 2012; 7:e43503.

542 Kurspahić-Mujčić et al. Overweight /obesity in preschoolers

11. van Ansem WJ, Schrijvers CT, Rodenburg G, van de 27. Kurspahić Mujčić A, Zećo E. Socioeconomic and Mheen D. Maternal educational level and children’s demographic factors associated with abdominal obe- healthy eating behaviour: role of the home food envi- sity in women of childbearing age. Med Glas (Zeni- ronment (cross-sectional results from the INPACT ca) 2017; 14:218-23. study). Int J Behav Nutr Phys Act 2014; 11:113. 28. Cattaneo A1, Monasta L, Stamatakis E, Lioret S, Ca- 12. Sijtsma A, Koller M, Sauer PJJ, Corpeleijn E. Te- stetbon K, Frenken F, Manios Y, Moschonis G, Savva levision, sleep, outdoor play and BMI in young S, Zaborskis A, Rito AI, Nanu M, Vignerová J, Caroli children: the GECKO Drenthe cohort. Eur J Pediatr M, Ludvigsson J, Koch FS, Serra-Majem L, Szponar 2015; 174:631-9. L, van Lenthe F, Brug J. Overweight and obesity in in- 13. Barbosa SC, Coledam DHC, Stabelini Neto A, Elias fants and pre-school children in the European Union: RGM, Oliveira AR. School environment, sedentary a review of existing data. Obes Rev 2010;11:389–98. behavior and physical activity in preschool children. 29. Kondolot M, Balci E, Ozturk A, Mazicioglu MM, Rev Paul Ped 2016; 34:301–8. Hatipoglu N, Kurtoglu S, Ustunbas HB. Body mass 14. Uijtdewilligen L, Nauta J, Singh AS, van Mechelen index percentiles for Turkish children aged 0-84 W, Twisk JWR, van der Horst K, Chinapaw MJM. months. Ann Hum Biol 2011; 38:676-80. Determinants of physical activity and sedentary 30. Savva1 SC, Tornaritis M, Chadjigeorgiou C, Kou- behaviour in young people: a review and quality rides YA, Savva ME, Panagi A, Chrictodoulou E, synthesis of prospective studies. Br J Sports Med Kafatos A. Prevalence and socio-demographic asso- 2011; 45:896–905. ciations of undernutrition and obesity among pres- 15. Jiang J, Rosenqvist U, Wang H, Greiner T, Ma Y, chool children in Cyprus. Eur J Clin Nutr 2005; 59: Toschke AM. Risk factors for overweight in 2- to 1259–65. 6-year-old children in Beijing, China. Int J Pediatr 31. Ruiz M, Goldblatt P, Morrison J, Porta D, Forastie- Obes 2006; 1:103–8. re F, Hryhorczuk D, Antipkin Y, Saurel Cubizolles 16. Mendoza JA, Zimmerman FJ, Christakis DA. Tele- MJ, Lioret S, Vrijheid M, Torrent M, Iñiguez C, vision viewing, computer use, obesity, and adiposity Larrañaga I, Bakoula C, Veltsista A, van Eijsden in US preschool children. Int J Behav Nutr Phys Act M, Vrijkotte TG, Andrýsková L, Dušek L, Barros 2007; 4:44. H, Correia S, Järvelin MR, Taanila A, Ludvigsson 17. Marsh S, Ni Mhurchu C, Maddison R. The non-ad- J, Faresjö T, Marmot M, Pikhart H. Impact of low vertising effects of screen-based sedentary activities maternal education on early childhood overweight on acute eating behaviours in children, adolescents, and obesity in Europe. Paediatr Perinat Epidemiol and young adults. A systematic review. Appetite 2016; 30:274-84. 2013; 71:259. 32. Liu W. Lin R, Li B. Pallan M, Cheng KK, Adab 18. Hasanbegovic S, Mesihovic-Dinarevic S, Cuplov P. Socioeconomic determinants of childhood obe- M, Hadzimuratovic A, Boskailo H, Ilic N, Njuhović sity among primary school children in Guangzhou, A, Čengić N, Bajramović E, Brković Š. Epidemio- China. BMC Public Health 2016; 16:473. logy and etiology of obesity in children and youth 33. Morrissey TW, Dunifon RE, Kalil A. Maternal em- of Sarajevo Canton. Bosn J Basic Med Sci 2010; ployment, work schedules, and children’s body mass 10:140–6. index. Child Dev 2011; 82: 66–81. 19. Kail RV. Children and Their Development 6th ed. En- 34. Chia Y. Maternal labour supply and childhood obe- glewood Cliffs, NJ: Prentice Hall, 2011. sity in Canada: evidence from the NLSCY. CJE 20. Hawkins SS, Law C. A review of risk factors for 2008; 41:217–44. overweight in preschool children: a policy perspecti- 35. Mindlin M, Jenkins R, Law C. Maternal employment ve. Int J Pediatr Obes 2006; 1:195–209. and indicators of child health: a systematic review in 21. WHO. Global recommendations on physical activity pre-school children in OECD countries. J Epidemiol for health. Geneva, Switzerland: WHO, 2010. Community Health 2009; 63:340-50. 22. Briefel RR, Deming DM, Reidy KC. Parents’ per- 36. Savage JS, Fisher JO, Birch LL. Parental influence ceptions and adherence to children’s diet and action eating behaviour: conception to adolescence. J vity recommendations: the 2008 feeding infants and Law Med Ethics 2007; 35:22–34. toddlers study. Prev Chronic Dis 2015;12:150110. 37. Jiang J, Rosenqvist U, Wang H, Greiner T, Ma Y, 23. WHO Multicentre Growth Reference Study Group. Toschke AM. Risk factors for overweight in 2-to WHO child growth standards: length/height-for-age, 6-year-old children in Beijing, China. Int J Pediatr weight-for-age, weight-for-length, weight-for-height Obes 2006; 1:103–8. and body mass index-for-age: methods and deve- 38. Ariza AJ, Chen EH, Binns HJ, Christoffel KK. Risk lopment. Geneva: WHO, 2006. factors for overweight in five-to six-year-old Hispa- 24. Janjua NZ, Mahmood B, Islam MA, Goldenberg nic American children: a pilot study. J Urban Health RL. Maternal and early childhood risk factors for 2004; 81;150–61. overweight and obesity among low-income predo- 39. Brophy S, Cooksey R, Gravenor MB, Mistry R, minantly black children at age five years: a prospec- Thomas N, Lyons RA, Williams R. Risk factors for tive cohort study. J Obes 2012; 2012:1–9. childhood obesity at age 5: analysis of the millenni- 25. Olson CM, Demment MM, Carling SJ, Strawder- um cohort study. BMC Public Health 2009; 9:467. man MS. Associations between mothers’ and their 40. Hare-Bruun H, Nielsen BM, Kristensen PL, Møller children’s weights at 4 years of age. Child Obes NC, Togo P, Heitmann BL. Television viewing, food 2010; 6:201–7. preferences, and food habits among children: a pros- 26. Heslehurst N, Vieira R, Akhter Z, Bailey H, Slack pective epidemiological study. BMC Public Health E, Ngongalah L, Pemu A, Rankin J. The association 2011; 11:311. between maternal body mass index and child obe- 41. Matheson DM, Wang Y, Klesges LM, Beech BM, sity: a systematic review and meta-analysis. PLoS Kraemer HC, Robinson TN. African-American girls’ Med 2019;16:e1002817. dietary intake while watching television. Obes Res 2004; 12:32–7. 543 Medicinski Glasnik, Volume 17, Number 2, August 2020

544 545 Medicinski Glasnik, Volume 17, Number 2, August 2020

546