Vol. 78 2019 No. 2

ISSN 0015–5659 Impact Factor: 0.497

Folia Morphol. Vol. 78, No. 2, May 2019 2018 Vol. 77 No. 2, pp. 171–408 2019, Vol. 78, No. 2, pp. 221–454 CONTENTS

ORIGINAL ARTICLES Evaluation of PECAM-1 and p38 MAPK expressions in cerebellum tissue of rats treated with caffeic acid phenethyl ester: a biochemical and immunohistochemical study...... 221 A. Çetin, E. Deveci The effect of prolonged formalin fixation on the staining characteristics of archival human brain tissue...... 230 A. Alrafiah, R. Alshali Aging changes in the retina of male albino rat: a histological, ultrastructural and immunohistochemical study...... 237 M.E.I. Mohamed, E.A.A. El-Shaarawy, M.F. Youakim, D.M.A. Shuaib, M.M. Ahmed Anatomical variations and dimensions of arteries in the anterior part of the circle of Willis...... 259 J. Shatri, S. Cerkezi, V. Ademi, V. Reci, S. Bexheti Morphological study of myelinated and unmyelinated fibres in the sacrococcygeal dorsal roots of the rat...... 267 J.-C. Lee, C.-H. Cheng, C.-T. Yen Combination of vitamin E and L-carnitine is superior in protection against isoproterenol-induced cardiac injury: histopathological evidence...... 274 E.A. Huwait Observations of foetal heart veins draining directly into the left and right atria...... 283 J.H. Kim, O.H. Chai, C.H. Song, Z.W. Jin, G. Murakami, H. Abe Renal vascularisation anomalies in the Polish population. Coexistence of arterial and venous anomalies in the vascular pedicle of the kidney...... 290 H. Sośnik, K. Sośnik Can constant light exposure affect the thyroid gland in prepubertal male albino rats? Histological and ultrastructural study...... 297 F. A . Abdel Gawad, E.A.A. El-Shaarawy, S.F. Arsanyos, T.I. Abd El-Galil, G.N. Awes The effect of diclofenac sodium on neural tube development in the early stage of chick embryos...... 307 T. Ertekin, A. Bilir, E. Aslan, B. Koca, O. Turamanlar, A. Ertekin, S. Albay The effects of leptin on F-actin remodelling in type 1 diabetes...... 314 C. Guven, E. Taskin, H. Akcakaya, R. Nurten Evaluation of the factors associated with sublingual varices: a descriptive clinical study...... 325 N. Akkaya, D. Ölmez, G. Özkan Morphometric evaluation and surgical implications of the infraorbital groove, canal and foramen on cone-beam computed tomography and a review of literature...... 331 İ. Bahşi, M. Orhan, P. Kervancioğlu, E.D. Yalçin Assessment of paranasal sinus parameters according to ancient skulls’ gender and age by using cone-beam computed tomography...... 344 K.O. Demiralp, S. Kursun Cakmak, S. Aksoy, S. Bayrak, K. Orhan, P. Demir Vol. 78 2019 No. 2 Prevalence of second canal in the mesiobuccal root of permanent maxillary molars from a Turkish subpopulation: a cone-beam computed tomography study...... 351 G. Magat, S. Hakbilen Incidence, number and topography of Wormian bones in Greek adult dry skulls ...... 359 K. Natsis, M. Piagkou, N. Lazaridis, N. Anastasopoulos, G. Nousios, G. Piagkos, M. Loukas Greater palatine foramen: assessment with palatal index, shape, number and gender...... 371 A. Ortug, M. Uzel ← INDEXED in: BIOSIS Previews, CAS, CINAHL, CrossRef, Dental Abstracts, EBSCO, Elsevier BIOBASE, EMBIOLOGY, FMJ, Google Scholar, Index Copernicus (143.41), Index Medicus/MEDLINE, Index Scholar, Ministry of Science and Higher Education (15), NCBI/National Center for Biotechnology Information, Polish Medical Bibliography, Scopus, SJR, Thomson Reuters, Thomson Scientific Products — Biological Abstracts, Ulrich’s Periodicals Directory, Veterinary Bulletin, WorldCat and Zoological Record. May 2019

Cover picture: Ventral view of the first cat. The right circumcaval ureter (RU) can be seen together with the right renal veins (RRV); RK — right kidney; LK — left kidney; CVC — caudal vena cava; LRV — left renal vein. For details see: M. Abidu-Figueiredo et al., Folia Morphol 2019; 78, 2: 433–436. http://journals.viamedica.pl/folia_morphologica An international multidisciplinary journal devoted to fundamental research in the morphological sciences Official Journal of the Polish Anatomical Society Investigation of the styloid process length in a Greek population...... 378 N. Zokaris, I. Siska, K. Natsis, M. Piagkou, N. Lazaridis, A. Skolka, E. Piehslinger (a Constituent Member of European Federation for Experimental Morphology — EFEM) Studies on clinical anatomy of the maxillofacial and mandibular regions of the Madras EDITOR-IN-CHIEF Red sheep (Ovis aries) in India...... 389 Janusz Moryś V. Sundaram, P. Dharani, R. Gnanadevi, R. Kavya Department of Anatomy and Neurobiology Accessory head of the flexor pollicis longus muscle: anatomical study and clinical significance.....394 Medical University of Gdańsk D.R. Ballesteros, P.L. Forero, L.E. Ballesteros http://journals.viamedica.pl/folia_morphologica Age-related structural-mechanical property changes in human peroneus longus muscle...... 401 A. Niyomchan, B. Panichareon, A. Siriphorn, T. Wongtawatchai See our website for information on manuscript status, Differences in foetal topographical anatomy between insertion sites of the iliopsoas aims and scope, instructions for authors and gluteus medius muscles into the proximal femur: a consideration of femoral torsion...... 408 as well as editorial board. P. Zhao, Z.W. Jin, J.H. Kim, H. Abe, G. Murakami, J.F. Rodríguez-Vázquez Blind mole rat (Spalax leucodon) masseter muscle: structure, homology, diversification Folia Morphologica and nomenclature ...... 419 A. Yoldas, M. Demir, R. İlgun, M.O. Dayan Publishing, Subscription and Advertising Office: VM Media sp. z o.o. VM Group sp.k., Grupa Via Medica Dermatoglyphics as a possible morphological biomarker in myopia: ul. Świętokrzyska 73, 80–180 Gdańsk, Poland analysis of finger ridge counts and fluctuating asymmetry...... 425 L. Sretić, N. Labus, T. Filipović, M. Filipović tel. (+48 58) 320 94 94, fax (+48 58) 320 94 60 Managing editor SHORT COMMUNICATION Joanna Niezgoda e-mail: [email protected] Pancreatic theory of relativity… ...... 431 Cover designer S. Hać Sylwia Scisłowska CASE REPORTS The journal is published at: www.fm.viamedica.pl in one volume per year consisting of four numbers. Right circumcaval ureter and double right renal vein in the Brazilian shorthair cat (Felis catus): The annual price for access to electronic version of the journal is 120 PLN for individual subscribers and 240 PLN two case reports...... 433 for institutions from Poland and 40 EURO for foreign individual subscribers and 80 EURO for foreign institu- M. Abidu-Figueiredo, A.V. Stocco, C.A. Santos-Sousa, P. Souza Junior, L.A.S. Pires, M.A. Babinski tions. Payment should be made to: VM Media sp. z o.o. VM Group sp.k., Grupa Via Medica, Bank BGŻ Paribas SA account number: 15 1600 1303 0004 1007 1035 9021; SWIFT: PPABPLPK. Electronic orders option available at: Circumaortic left renal vein (circumaortic renal collar) associated with the presence www.fm.viamedica.pl of vascular anomalies: a case series and review of literature...... 437 The publisher must be notified of a cancellation of access to electronic version not later then two months before R. Haładaj, M. Polguj, G. Wysiadecki, A. Żytkowski, M. Topol the end of a calendar year. After that date electronic access will be automatically prolonged for another year. An atypical biceps brachii and coracobrachialis muscles associated with Advertising. For details on media opportunities within this electronic version of journal please contact multiple neurovascular aberrations: a case report with clinical significance...... 444 the advertising sales department, ul. Świętokrzyska 73, 80–180 Gdańsk, Poland, tel: (+48 58) 320 94 57, M. Piagkou, T. Totlis, N. Anastasopoulos, N. Lazaridis, K. Natsis e-mail: [email protected] Unusual arterial supply of the segment IV with interlobar bridge and right replaced The editors accept no responsibility for advertisement contents. hepatic artery: a case report...... 450 Folia Morphologica is the official journal of the Polish Anatomical Society. For information about the Society, M. Tiryakioğlu, T. Koç, N.S. İlgi please contact: Prof. Marek Grzybiak, Department of Clinical Anatomy, Medical University of Gdansk, ul. Dębinki 1, 80–211 Gdańsk, Poland, tel: +48 58 349 14 22, e-mail: [email protected] All rights reserved, including translation into foreign languages. No part of this periodical, either text or illustration, may be used in any form whatsoever. It is particularly forbidden for any part of this material to be copied or translated into a mechanical or electronic language and also to be recorded in whatever form, stored in any kind of retrieval system or transmitted, whether in an electronic or mechanical form or with the aid of photocopying, microfilm, recording, scanning or in any other form, without the prior written permission of the publisher. The rights of the publisher are protected by national copyright laws and by international conventions, and their violation will be punishable by penal sanctions. Editorial policies and author guidelines are published on journal website: http://journals.viamedica.pl/folia_morphologica Legal note: http://journals.viamedica.pl/FM/about/legalNote Folia Morphologica is indexed by: BIOSIS Previews, CAS, CINAHL, CrossRef, Dental Abstracts, EBSCO, Elsevier BIOBASE, EMBIOLOGY, FMJ, Google Scholar, Index Copernicus (143.41), Index Medicus/MEDLINE, Index Scholar, Ministry of Science and Higher Education (15), NCBI/National Center for Biotechnology Information, Polish Medical Bibliography, Scopus, SJR, Thomson Reuters, Thomson Scientific Products — Biological Abstracts, Ulrich’s Periodicals Directory, Veterinary Bulletin, WorldCat and Zoological Record. Position in Index Copernicus ranking systems is available at: www.indexcopernicus.com. Current Impact Factor of Folia Morphologica (2017) is 0.497.

Za prenumeratę czasopisma Folia Morphologica przysługuje 5 pkt edukacyjnych* *na podstawie rozporządzenia Ministerstwa Zdrowia z dnia 6 października 2004 r. w sprawie sposobów dopełnienia obowiązku doskonalenia zawodowego lekarzy i lekarzy dentystów (Dz.U. 04.231.2326 z dnia 22 października 2004 r.)

ISSN 0015–5659 Editor-in-Chief JANUSZ MORYŚ Department of Anatomy and Neurobiology, Medical University of Gdańsk ul. Dębinki 1, 80–211 Gdańsk, Poland tel. (+48 58) 349 1401, fax (+48 58) 349 1421, e-mail: [email protected]

Editorial Advisory Board

Ryszard Aleksandrowicz, Department of Anatomy, Alexander J. McDONALD, Department of Cell Biology Medical University of Warsaw, Poland and Neuroscience, USC School of Medicine, Franciszek BURDAN, Experimental Teratology Unit Columbia, United States of the Human Anatomy Department, Medical University Stanisław Moskalewski, Department of Histology of Lublin, Poland and Embryology, Medical University of Warsaw, Poland Małgorzata Bruska, Department of Anatomy, Orlando PACIELLO, Dipartimento di Patologia e Sanita animale, University Medical School, Poznań, Poland Univesita degli Studi di Napoli Federico II, Napoli, Italy Mafalda CACCIOTTOLO, USC Leonard Davis School Asla Pitkänen, Department of Neurobiology, of Gerontology, University of Southern California, A.I. Virtanen Institute, University of Kuopio, Finland Los Angeles, United States Michał POLGUJ, Department of Angiology, Stephen W. Carmichael, Department of Anatomy, Medical University of Łódź, Poland Mayo Clinic, Rochester, United States Michał K. Stachowiak, Department of Molecular and Bogdan CISZEK, Department of Human Anatomy, Structural Neurobiology and Gene Therapy, State University Medical University of Warsaw, Poland of New York, Buffalo, United States Carla D’AGOSTINO, Neuromuscular Center, University of Paweł SYSA, Department of Histology and Embryology, Southern Califonia, Los Angeles, CA, United States Warsaw University of Life Sciences, Poland Zygmund Antoni DOMAGAŁA, Department of Anatomy, Medi- Michał SZPINDA, Department of Anatomy, Nicolaus cal University of Wrocław, Poland Copernicus University in Toruń, Collegium Medicum Rastislav DRUGA, Department of Functional Anatomy, in Bydgoszcz, Poland nd 2 Medical Faculty Charles University, Prague, Czech Republik Stanisław Szteyn, Department of Comparative Anatomy, Jochen FanghÄnel, Department of Anatomy, Warmia and Masuria University, Olsztyn, Poland Ernst-Moritz-Arndt University, Greifswald, Germany Edyta SZUROWSKA, 2nd Department of Radiology, Marek GRZYBIAK, Department of Cinical Anatomy, Medical University, Gdańsk, Poland Medical University of Gdańsk, Poland Jean-Pierre Timmermans, Laboratory of Cell Biology and Hans Jorgen GUNDERSEN, Stereological Research Laboratory, Histology/Central Core Facility for Microscopic Imaging, University of Aarhus, Denmark Department of Veterinary Sciences, Kazimierz Jędrzejewski, Department of Anatomy, University of Antwerp, Belgium Medical University of Łódź, Poland Mirosław TOPOL, Department of Angiology, Leszek Kaczmarek, Department of Molecular Cell Medical University of Łódź, Poland Neurobiology, Nencki Institute, Warsaw, Poland Krzysztof TURLEJSKI, Department of Neurophysiology, Zbigniew KMIEĆ, Department of Histology, Nencki Institute of Experimental Biology, Warsaw, Poland Medical University of Gdańsk, Poland Jiro USUKURA, Ecolopia Science Institute, Henryk KOBRYŃ, Department of Morphological Sciences, Nagoya University, Japan Warsaw, Agricultural University, Poland Jerzy WALOCHA, Department of Anatomy, Przemysław Kowiański, Department of Anatomy and Jagiellonian University, Collegium Medicum, Neurobiology, Medical University of Gdańsk, Poland Kraków, Poland Dariusz KOZŁOWSKI, 2nd Department of Cardiology, Mark J. West, Department of Neurobiology, Medical University of Gdańsk, Poland Institute of Anatomy, Åarhus University, Denmark Marios LOUKAS, Department of Anatomical Sciences, School Shonen YOSHIDA, Laboratory of Cancer Cell Biology, of Medicine, St. George’s University, Grenada, West Indies Nagoya University School of Medicine, Nagoya, Japan Mirosław Łakomy, Department of Animal Anatomy, Maciej ZABEL, Department of Histology and Embryology, Warmia and Masuria University, Olsztyn, Poland Medical University of Wrocław, Poland Andrzej Łukaszyk, Department of Histology and Embryology, Marco ZEDDA, Department of Veterinary Medicine, University Medical School, Poznań, Poland University of Sassari, Italy

Evaluation of PECAM-1 and p38 MAPK expressions in cerebellum tissue of rats treated with caffeic acid phenethyl ester: a biochemical and immunohistochemical study A. Çetin1, E. Deveci2

1Department of Neurosurgery, University of Health Sciences, Gazi Yaşargil Education and Research Hospital, Diyarbakır, Turkey 2Department of Histology and Embryology, Faculty of Medicine, Dicle University, Diyarbakır, Turkey

[Received: 7 June 2018; Accepted: 16 July 2018]

Background: This study aimed to investigate the antioxidative and anti-inflammatory effects of caffeic acid phenethyl ester (CAPE) on damage caused to cerebellum tissue by diffuse traumatic head trauma via biochemical, histopathologic, and immuno-histochemical methods. Materials and methods: Male Sprague-Dawley (300–350 g) rats were subjected to traumatic brain injury with a weight-drop device (300 g/1 m weight-height impact). Twenty-four adult rats were randomly divided into three equal groups of 8, including a control group, traumatic brain injury (TBI) group, and TBI + CAPE treatment group (10 μmoL/kg/i.p.). Cerebellum tissue samples taken from anterior lobe from all rats were taken 7 days after traumatic injury and were subjected to biochemical and histopathological analysis, as well as immunohistochemical analysis for platelet endothelial cell adhesion molecule-1 (PECAM-1) and phosphate 38-mitogen-activated protein kinase (p38 MAPK). Results: In the TBI group, the granular layer had dilatation and haemorrhage in the capillary vessels and inflammatory cell infiltration around the periphery of the blood vessels. In the TBI + CAPE group, the small capillaries in the white matter were slightly dilated, there were no inflammatory cells, and dense chromatin/ granular cells were observed in the granular layer. Also in the TBI + CAPE group, the Purkinje cells of the ganglion cell layer had ovoid nuclei, were chromatin- -rich, and their extensions protruded to the molecular layer. CAPE is thought to regulate inflammation,cell damage, and angiogenetic development by affecting the PECAM-1 and p38 MAPK proteins. Conclusions: These proteins are key modulators of endothelial integrity and neuroinflammation in vessels in response to endothelial damage as well as of the proinflammatory response in the cerebellum in response to traumatic damage. (Folia Morphol 2019; 78, 2: 221–229)

Key words: traumatic brain injury, cerebellum, PECAM-1, p38 MAPK, rat

Address for correspondence: E. Deveci, Prof., PhD, Dicle University, Faculty of Medicine, Department of Histology and Embryology, 21280, Diyarbakır, Turkey, tel: +90 412 2488001 ext. 4443 (faculty room), fax: +90 412 2488440, e-mail: devecie32@hotmail

221 Folia Morphol., 2019, Vol. 78, No. 2

INTRODUCTION Experimental studies suggest that the phosphate The cerebellum is vulnerable to damage from 38 mitogen-activated protein kinase signalling path- a variety of sources, including developmental defects, way plays a major role in ischaemia-reperfusion in- degenerative diseases, infectious processes, chronic jury. An increase in p38 MAPK expression in kidney, alcoholism, trauma, and tumours [35]. Caffeic acid liver, vascular and myocardial cells and pulmonary phenethyl ester (CAPE) is a flavonoid-like compound ischaemia-reperfusion models has been reported that is the active component of propolis, which is [2, 40, 50, 62]. Additionally, it has been reported that obtained from honey bee hives and has been long- MAPK participates in the control of complex cellular used in traditional medicine. It has many protective processes, such as gene expression, synaptic plastic- properties, including antioxidant, anti-inflammatory, ity in the nervous system, cell differentiation, cell anti-carcinogenic, antiviral, neuroprotective, and im- proliferation, cell death, and inflammation [20, 31]. munomodulation [3]. Research in a number of trau- The MAPK pathway is an evolutionarily conserved matic brain injury (TBI) animal models has shown that mechanism of transducing external stress and injury pathophysiological changes in the cortex, thalamus, to internal cellular responses; these responses play and hippocampus contribute to motor and cognitive a major role in the balance between cell survival and disorders. Impairments in motor function, coordina- cell death [55, 58]. Platelet endothelial cell adhesion tion, and cerebellar function are common outcomes molecule-1 (PECAM-1) is a restricted receptor in blood of TBI [36]. Some of the classically described conse- and vascular cells [39]. quences of direct traumatic injury to the cerebellum There are five members of the immunoglobu- are hypotonia, ataxia, dysmetria, tremor, and vertigo lin superfamily. It has been shown that intracellu- [38]. Cells from various brain regions respond dif- lar adhesion molecule-1 and -2 (ICAM-1, ICAM-2), ferently to mechanical injury and it is known that vascular adhesion molecule-1 (VCAM-1), PECAM-1, the Purkinje neurons are the cells most affected by mucosal vascular addressing cell adhesion molecule-1 cerebellar trauma [4]. (MAdCAM-1). PECAM-1 has been reported to play an Among the many signalling pathways that re- important role in endothelial cells and in leukocyte spond to mitogens and stress factors, mitogen-ac- adhesion to transduction. MAdCAM-1, a ligand of tivated protein kinase (MAPK) family members are L-selectin and a4P7 integrin, has been reported to crucial for cell repair by regulating the activities of play a key role in the selective targeting of lympho- nuclear transcription factors [43, 53, 61]. There are cytes [17, 47]. Furthermore, it has also been shown three known types of MAPKs: ERKs, SAPK/JNKs and that ICAM-1, VCAM-1 and PECAM-1 serve as signal p38 MAPKs. Initially, ERKs were found to be important transducers not only for the transendothelial migra- for cell survival, SAPK/JNKs and p38 MAPKs were more tion of leukocytes but also for inducing endothelial likely to be stressed and thus play a role in apopto- signalling and influencing the progression of neuro- sis [27, 28, 52, 57]. The magnitude of the balance inflammation [18]. Among the molecules known to between ERK and SAPK/JNK/p38 MAPK activation (secondarily) activate the integrins are the chemokines is assumed to be a key to whether cells continue to and PECAM-1, CD31 or endoCAM. Once a leukocyte survive or whether apoptosis has entered. However, binds to endothelium via an integrin-immunoglobulin many researchers have recently shown that the regu- (Ig) superfamily (IgSF) interaction, it “searches” for lation of apoptosis by MAPKs is more complex than junctions between endothelial cells, first squeesing originally thought [11, 27, 33, 52]. Mitogen-activated between these potential discontinuities, and then protein kinases are a highly conserved family of ser- penetrating the underlying basement membrane to ine/threonine protein kinases involved in a variety of reach the tissues [42, 64]. Mouse CD31 has also been fundamental cellular processes such as proliferation, isolated and found to be 63% identical at the aa differentiation, motility, stress response, apoptosis, level to human CD31 [59]. While monoclonal anti- and survival. Conventional MAPKs include the extra- bodies to CD31 are reportedly species specific [59], cellular signal-regulated kinase 1 and 2 (Erk1/2 or polyclonal antibodies to CD31 seem to show species p44/42) [22, 48], the c-Jun N-terminal kinases 1-3 cross-reactivity [34]. PECAM-1 plays a critical role in (JNK1-3)/stress activated protein kinases (SAPK1A, 1B, damage recovery after pathological conditions. The 1C) [63, 65], the p38 isoforms (p38a, b, g, and d), and function of PECAM-1 in endothelial cells in the context Erk5 [21, 44]. The lesser-studied, atypical MAPKs in- of inflammation and apoptosis has been investigated clude Nemo-like kinase (NLK), Erk3/4, and Erk7/8 [20]. [29, 56]. It is not known that PECAM-1 is exercised on 222 A. Çetin, E. Deveci, Evaluation of PECAM-1 and p38 MAPK expressions in cerebellar tissue

neurons or glia [13]. In the brain, PECAM-1 has been from the inferior vena cava for serum biochemical reported to be exclusively in cerebral endothelial cells markers determination in all groups. In this way, forming the blood-brain barrier [30]. MDA, SOD, CAT, and GSH-x values were measured The current study aimed to investigate the anti- biochemically. And, cerebellum tissue taken from oxidative and anti-inflammatory effects of CAPE on anterior lobe was extracted, fixed in a 10% formalin damage caused to the cerebellum tissue by diffuse solution, and embedded in paraffin blocks for histo- traumatic head trauma via biochemical, histopatho- pathologic examination in all groups. Sections (5 µm logic, and immunohistochemical methods. thick) were obtained from paraffin blocks and stained with haematoxylin and eosin (H&E). MATERIALS AND METHODS Animals Immunohistochemistry method Every single surgical methodology and the con- Antigen retrieval process was performed in citrate sequent care and healing of the animals utilised as buffer solution (pH 6) two times first 5 min, later a part of this investigation were in strict understand- 4 min boiled in microwave oven at 700 W (Bosch®). ing with the National Institutes of Health (NIH Pub- They were allowed to cool to room temperature for lications No. 8023, revised 1978) rules for animal 20 min and washed in distilled water for 5 min two times. care. All techniques performed in this examination Endogenous peroxidase activity was blocked in 0.1% were approved by the Ethics Committee for Animal hydrogen peroxide [K-40677109,64271 hydrogen

Experimentation of the Faculty of Medicine at Di- peroxide (H2O2), Dortmudt + Germany, MERCK] (3 mL cle University, Turkey (Protocol number: 2017-32). 30% hydrogen peroxide (H2O2) + 27 mL methanol) Twenty-four male Sprague-Dawley rats (300–350 g) for 10 min. Ultra V block (Histostain-Plus Kit, Invitro- were maintained under 23 ± 2°C and 12 h light/dark gen, Carlsbad, CA) was applied for 8 min prior to the cycles with ad libitum access to standard pelleted application of primary antibodies p38 MAPK mouse food and water. monoclonal, 1/100 and PECAM-1 mouse monoclonal, All rats toward the finish of the analysis were 1/100 for overnight. Secondary antibody (Histostain- healthy and no distinction in nourishment/water con- Plus Kit, Invitrogen, Carlsbad, CA) was applied for sumption and body weight pick up amongst experi- 20 min. Slides then were exposed to streptavidin-perox- mental and control rats were noticed. The rats were idase for 25 min. Diaminobenzidine (DAB, Invitrogen, divided into three equal groups as control, trauma Carlsbad, lot: HD36221, Thermo Fischer, Fremont, CA, (TBI), and trauma (TBI) + CAPE. USA) was used as a chromogen. Control slides were prepared as mentioned above but omitting the prima- Traumatic brain injury model ry antibodies. After counterstaining with haematoxy- Sedation procedure. The animals were anes- lin (product number: HHS32 SIGMA, Haematoxylin thetised via an intraperitoneal injection of 5 mg/kg Solution, Harris Modified, Sigma-Aldrich, 3050 Spruce xylazine HCl (Rompun, Bayer HealthCare AG, Ger- Street, Saint Louis, MO 63103, USA), washing in tap many) and 40 mg/kg ketamine HCl (Ketalar, Pfizer Inc., water for 3 min and in distilled water for 2 × 3 min, USA) [45], after which they were allowed to breathe the slides were mounted with Entellan® (lot: 107961, spontaneously. Sigma-Aldrich, St. Louis, MO, United States). Control group. Isotonic saline solution (an equal volume of CAPE) was administered i.p. for 7 days. Malondialdehyde and glutathione peroxidase assays TBI group. The diffuse brain injury model de- Malondialdehyde (MDA) levels and glutathione scribed by Marmarou et al. [32] was used Briefly, peroxidase (GSH-Px) activities were determined in a trauma device dropped a constant weight (300 g) the cerebellum of each rat, and the average values from a specific height (1 m) through a tube, inducing of each group were calculated. Each cerebellum mild trauma. Following trauma, the rats’ cerebellum sample was prepared as a 10% homogenate (ac- tissues were removed. cording to weight) in 0.9% saline using a homog- TBI + CAPE group. 15 min following injury, the enizer on ice. Then, the homogenate was centri- rats were given CAPE (10 μmoL/kg/i.p) intraperito- fuged at 2000 rpm for 10 min, and the superna- neally every day for 7 days [26]. At the end of the tant was collected. MDA levels were determined study, animals were sacrificed and cerebellum tis- using the double heating method of Draper and sues were removed. Blood samples were collected Hadley [14]. MDA is an end product of fatty acid 223 Folia Morphol., 2019, Vol. 78, No. 2

peroxidation that reacts with thiobarbituric acid cm for H2O2. CAT activity was expressed as mmol H2O2 (TBA) to form a coloured complex. Briefly, 2.5 mL consumed/min per mg tissue protein. of TBA solution (100 g/L) was added to 0.5 mL of homogenate in a centrifuge tube, and the tubes were Statistical analysis placed in boiling water for 15 min. After cooling All of the statistical analyses for this study were with flowing water, the tubes were centrifuged at carried out using GraphPad Prism 4.0 software 1000 rpm for 10 min, and 2 mL of the supernatant (GraphPad Software, 2003, San Diego, CA, USA). All was added to 1 mL of TBA solution (6.7 g/L); these data are presented as mean ± standard deviation tubes were placed in boiling water for another 15 min. (SD). Groups of data were compared with an analysis After cooling, the amount of TBA-reactive species was of variance (ANOVA) followed by Tukey’s multiple measured at 532 nm, and the MDA concentration comparison test. Values of p < 0.05 were considered was calculated using the absorbance coefficient of significant. the MDA–TBA complex. MDA values were expressed as nanomoles per gram (nmol/g) of wet tissue. RESULTS The GSH-Px activity was measured by the method Biochemical findings of Paglia and Valentine [37]. An enzymatic reaction was The following biochemical parameters were compared between groups (Fig. 1). initiated by the addition of hydrogen peroxide (H2O2) to a tube that contained reduced nicotinamide adenine dinucleotide phosphate, reduced glutathione, sodium Tissue malondialdehyde levels azide, and glutathione reductase. The change in absorb- Malondialdehyde values in the trauma (TBI) group ance at 340 nm was monitored by spectrophotometry. were significantly higher than those of the control Data were expressed as U/g protein. group (p < 0.001), while the TBI + CAPE group had significantly lower levels than those of the trauma Measurement of superoxide dismutase activity (TBI) group (p < 0.001). Total superoxide dismutase (SOD) activity was determined with a SOD detection kit (RANSOD kit, Tissue glutathione peroxidase levels Randox Co., UK) according to the manufacturer’s Tissue glutathione peroxidase levels of the control instructions. SOD accelerates the conversion of group were significantly higher than those of the trau- the toxic superoxide (produced during oxidative ma group (p < 0.001), while those of the TBI + CAPE energy processes) to hydrogen peroxide and mo- group were also significantly higher than those of lecular oxygen. This method employs xanthine and the trauma group (p < 0.001). xanthine oxidase to generate superoxide radicals that react with 2-(4-iodophenyl)-3-(4- nitrophenol)- Superoxide dismutase levels 5-phenyltetrazolium chloride (INT) to form a red Tissue SOD levels of the control group were signifi- formazan dye. The SOD activity is measured by the cantly lower than those in the trauma group, while degree of inhibition of this reaction. One unit of those of the TBI + CAPE group were significantly SOD causes 50% inhibition of the rate of reduction higher than those of the trauma group (p < 0.001). of INT under the assay’s conditions. Absorbance measurements were taken at 505 nm, and SOD Catalase levels levels were determined through a standard curve Tissue CAT levels in the control group were signifi- and expressed as U/mg protein [49]. cantly lower than those of the trauma group (p < 0.001), while those of the TBI + CAPE group were significantly Measurement of catalase activity higher than those of the trauma group (p < 0.001). Tissue catalase (CAT) activity was assayed spectrophotometrically by monitoring the decomposition of Histological findings

H2O2 using the procedure of Aebi [1]. Briefly, 0.5 mL In the control group, the Purkinje cells of the of 30 mM H2O2 in 50 mM phosphate buffer (pH 7.0) ganglion cell layer were in a horizontal and fibrillar was added to 1 mL of tissue supernatant (diluted structure. In the glomerular areas, the Purkinje cells

1:10), and the consumption of H2O2 was followed were arranged regularly in a granular layer, and the spectrophotometrically at 240 nm for 2 min at 25oC. microglia and basket cells were scattered diffusely The molar extinction coefficient was 43.6 L/mol per in the molecular layer. In the granular layer of the 224 A. Çetin, E. Deveci, Evaluation of PECAM-1 and p38 MAPK expressions in cerebellar tissue

nuclei with chromatin-rich extensions protruding to the molecular layer. Diffused microglia and basket cells were scattered in the molecular layer.

Immunohistochemical findings In the control group, p38 MAPK expression was observed in some Purkinje cells of the ganglion cell layer and in the basket cells of the molecular layer. The TBI group had positive expression of p38 MAPK in inflammatory cells around the blood vessels in the granular layer and in the molecular layer boundary. p38 MAPK expression also was seen in the glial cells and in some apoptotic Purkinje cells. In the TBI + CAPE group, p38 MAPK was expressed in the Purkinje cells, but was not expressed in the small granular cells of the granular layer. There was no p38 MAPK expression in the microglia and basket cells. The control group had no PECAM-1 expression in the microglial cells and endothelial cells of the capillary vessels in the molecular layer; however, PECAM-1 expression was positive in the Purkinje cells. In the TBI group, PECAM-1 expression was positive in the endothelial cells of the dilated blood vessels between the ganglion layer and the granular layer, in inflammatory cells, as well as in some degenerate Purkinje cells. Expression of PECAM-1 was observed in the microglial and basket cells and in the degenerative fibres in the molecular layer. In the TBI + CAPE group, there was no PECAM-1 expression in the microglial and basket cells or in the endothelial cells of some dilated vessels in the molecular layer; however, PECAM-1 expression was positive in the Purkinje cells (Fig. 2).

DISCUSSION Traumatic brain injury usually results in severe shock to the head and/or body. A minor traumatic Figure 1. Biochemical results relevant to the study groups; A. MDA; B. GSH-Px; C. SOD; D. CAT. Abbreviations — see text. brain injury can lead to the transient dysfunction of brain cells. Previous studies have shown that a variety of pathological factors (e.g., oxidative stress, trauma group, there was dilatation and haemorrhage the inflammatory response, and apoptosis) are in- in the capillary vessels as well as inflammatory cell volved in secondary brain injury following traumatic infiltration around the periphery of the blood ves- brain injury. Early interventions to reduce the level of sels. Degenerative changes in the Purkinje cells (i.e., oxidative stress and the extent of the inflammatory pyknosis in some of the nuclei) were observed in the response can significantly reduce the extent of trau- ganglion cell layer. In the TBI + CAPE group, the small matic brain injury [19]. Activated microglia cells serve capillaries were slightly dilated in the white matter. as cellular cleaners to alleviate traumatic damage and There were no inflammatory cells, dense chromatin/ promote wound healing. It has been reported that granular cells were observed in the granular layer, and these numbers may increase in cases of Purkinje cell the Purkinje cells of the ganglion cell layer had ovoid damage [25].

225 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. A. Control group. Purkinje cells in horizontal view, fibrillar structure and granular cells regular, basket in the molecular layer and microglia cells diffuse appearance. H&E staining, bar 50 µm; B. Trauma group. Capillary dilatation of veins and haemorrhage and inflammatory cell infiltration around blood vessels (yellow arrow), Degenerative changes in pyknotic Purkinje cells in some nuclei of the ganglion cell layer (red arrow), H&E staining, bar 50 µm; C. Trauma + CAPE group. Small dilation of small capillaries in white matter (yellow arrow). There are no inflammatory cells, dense chromatin cells in the granular layer, Purkinje cell of the ganglion cell layer, nuclei oval and chromatin rich appearance, and microglia and basket cells (red arrow). H&E staining, bar 50 µm; D. Control group. p38 MAPK expression in some Purkinje cells of ganglion cell layer and in the basket cells of the molecular layer (yellow arrow). p38 MAPK immunostaining, bar 50 µm; E. Trauma group. Positive expression of p38 MAPK in inflammatory cells around the blood vessels in the granular layer and molecular layer (red arrow), also positive p38 MAPK expression in glial cells and some apoptotic Purkinje cells (yellow arrow). p38 MAPK immunostaining, bar 50 µm; F. Trauma + CAPE group. p38 MAPK expression in Purkinje cells (yellow arrow), but not expression in small granular cells of granular layer. Negative p38 MAPK expression in microglia and basket cells p38 MAPK (red arrow), p38 MAPK immunostaining, bar 50 µm; G. Control group. PECAM-1 expression negative in microglial cells and endothelial cells of the capillary vessels in the molecular layer (red arrow), positive PECAM-1 expression in Purkinje cells (yellow arrow). PECAM-1 immunostaining, bar 50 µm; H. Trauma group. Positive PECAM-1 expression in the endothelial cells of dilated blood vessels between the ganglion layer (red arrow), the granular layer, in inflammatory cells and in some degenerate Purkinje cells (yellow arrow), expression of PECAM-1 in microglial and basket cells and degenerative fibres in the molecular layer. PECAM-1 immunostaining, bar 50 µm; I. Trauma + CAPE group. Negative PECAM-1 expression in microglial and basket cells and endothelial cells (red arrow), positive PECAM-1 expression in Purkinje cells (yellow arrow). PECAM-1 immunostaining, bar 50 µm. Abbreviations — see text.

In TBI models, regional specific cerebellar Purkinje Traumatic brain injury causes glial cell activa- cell injury and/or loss are important features that are tion in the cerebellum, and changes in the Purkinje known to cause cerebellar damage, including a distinct cells are important for determining the extent of spatial regeneration of glia and traumatic axonal dam- damage. Following selective and diffuse TBI, the age [38]. Fukuda et al. [16] examined an injury caused cerebellum exhibits pathological changes, including by frontoparietal cortex lateral fluid percussion at selective cell loss, altered metabolism, and white 1 day and 7 days after injury. At both time points, they matter damage [38, 46]. Damage to the Purkinje reported death of Purkinje cells in the cerebellar vermis. cells is considered an indicator of TBI [25]. In our

226 A. Çetin, E. Deveci, Evaluation of PECAM-1 and p38 MAPK expressions in cerebellar tissue

current study, the TBI group had dilatation in the Vascular endothelial cells that cross the endothe- capillary vessels and inflammatory cell infiltration lial cell-cell junction during inflammatory processes around the haemorrhage in the granular layer. The have been shown to play an important role in vascular Purkinje cells revealed degenerative changes with regulation. These cells have the ability to modulate some pyknotic nuclei in the ganglion cells. In the PECAM-1 intracellular signalling processes [39, 60]. TBI + CAPE group, it was observed slightly dilated During recovery of ischaemic damage, PECAM-1 small capillaries (especially in white matter), and has been shown to regenerate angiogenic and anti- the granular cells were chromatin-rich with no in- apoptotic isoforms, as well as repair and remodel flammation. The Purkinje cells in the ganglion layer the damaged blood-brain barrier [12, 15]. PECAM- have chromatin-rich, oval-shaped nuclei, and their 1-mediated activation of the PI3K/Akt pathway up- cytoplasmic extensions extend into the molecular regulates NF-kB-mediated transcription to facilitate layer. In addition, it was seen small and diffuse angiogenesis, cell survival/growth, and the recovery microglia and basket cells. In the ganglion cell of the endothelial cell barrier [8, 9]. Hwang et al. [23] layer, the nuclei of the Purkinje cells were ovoid and observed that transient ischaemia produced in gerbils chromatin-rich, and their extensions protruded to by 5-min occlusion of both common carotid arteries the molecular layer. Microglia and basket cells were results in a significant increase of PECAM-1 immuno- diffuse through the molecular layer. reactivity and protein level in hippocampus proper. Caffeic acid phenethyl ester has been reported And, they reported that PECAM-1 immunoreactivity to block many inflammatory agents (primarily tu- in post-ischaemic group was significantly increased mour necrosis factor alpha) by blocking nuclear by 4 days after ischaemic insult. Rosenblum et al. factor kappa B (NF-kB) and oxygen radicals. It has [41] suggested that anti-PECAM-1 treatment may be also been shown that CAPE inhibits apoptosis in effective in inhibiting the in vivo inhibition of plate- inflammatory cells independent of glucocorticoid let adhesion/aggregation at the site of endothelial receptors [10]. To determine the molecular effects injury in brain arterioles. Vaporciyan et al. [51] have of CAPE, it was investigated changes in p38 MAPK shown that human PECAM-1 antibody cross-reacting and PECAM-1 expression in cerebellum tissue in with rat PECAM-1 can block the in vivo accumulation traumatic brain injured rats. Most mRNAs coding of rat neutrophils in the peritoneal cavity and the for inflammatory response genes are unstable and alveolar part of the lung. The authors of this trial stabilised by the p38 MAPK pathway [5, 54]. It is also suggested that PECAM-1 might be a potential known that microglia respond quickly (30 min) and therapeutic target. TBI may be expected to increase migrate immediately to the wound centre. MAPK the expression of PECAM-1 on vascular endothelium p38a helps to bind inflammatory stimuli in the mi- in the traumatised brain, which is constantly followed croglia in response to cellular activities. Sustained by an acute inflammatory response. However, Carlos activation of p38a MAPK was observed in activated et al. [7] showed that PECAM-1 expression remained microglia in the brain after cerebral brain ischaemia unchanged in a rat model. [24]. Furthermore, Haines et al. [20]. revealed that MAPK signalling pathways regulate myelination. In CONCLUSIONS our current study’s TBI group, p38 MAPK was ex- It was hypothesized that PECAM-1 expression pressed in the inflammatory cells around the blood contributes to the development of angiogenesis af- vessels in the granular layer and in the molecular ter endothelial damage. In addition, it was thought layer boundary. Additionally, positive p38 MAPK that CAPE administration inhibits the progression of expression was seen in the glial cells and in some the inflammatory reaction in endothelial cells, which apoptotic Purkinje cells. Results of the TBI + CAPE was initially induced by proinflammatory processes. group revealed that p38 MAPK was not expressed CAPE administration is thought to regulate inflam- in the small granular cells of the granular layer or mation, cell damage, and angiogenetic development in the microglial and basket cells. However, the by affecting the PECAM-1 and p38 MAPK proteins. Purkinje cells positively expressed p38 MAPK. It These proteins are key modulators of endothelial in- was hypothesized that CAPE regulates the common tegrity and neuroinflammation in vessels in response inflammatory response observed after cerebellum to endothelial damage, and are key regulators of the injury, thus leading to re-activation of the microglial proinflammatory response occurring in the cerebel- and Purkinje cells (Fig. 2E, F). lum in response to traumatic damage. 227 Folia Morphol., 2019, Vol. 78, No. 2

REFERENCES 17. Gahmberg CG, Tolvanen M, Kotovuori P. Leukocyte adhesion--structure and function of human leukocyte beta2- 1. Aebi HE. Catalase. In: Bergmeyer HU (eds). Methods of en- integrins and their cellular ligands. Eur J Biochem. 1997; zymatic analysis. 3rd edn. Verlag Chemie, Germany 1983: 245(2): 215–232, indexed in Pubmed: 9151947. 273–286. 18. Greenwood J, Etienne-Manneville S, Adamson P, et al. 2. Arumugam S, Girish Subbiah K, Kemparaju K, et al. Neutro- Lymphocyte migration into the central nervous system: phil extracellular traps in acrolein promoted hepatic ischemia implication of ICAM-1 signalling at the blood-brain bar- reperfusion injury: Therapeutic potential of NOX2 and rier. Vascul Pharmacol. 2002; 38(6): 315–322, indexed in p38MAPK inhibitors. J Cell Physiol. 2018; 233(4): 3244–3261, Pubmed: 12529926. doi: 10.1002/jcp.26167, indexed in Pubmed: 28884828. 19. Gyoneva S, Ransohoff RM. Inflammatory reaction after 3. Ayla Ş, Tunalı G, Bilgiç BE, et al. Antioxidant activity of traumatic brain injury: therapeutic potential of targeting CAPE (caffeic acid phenethyl ester) in vitro can protect hu- cell-cell communication by chemokines. Trends Pharmacol man sperm deoxyribonucleic acid from oxidative damage. Sci. 2015; 36(7): 471–480, doi: 10.1016/j.tips.2015.04.003, Acta Histochem. 2018; 120(2): 117–121, doi: 10.1016/j. indexed in Pubmed: 25979813. acthis.2018.01.001, indexed in Pubmed: 29325972. 20. Haines JD, Fragoso G, Hossain S, et al. p38 Mitogen-acti- 4. Barateiro A, Afonso V, Santos G, et al. S100B as a Potential vated protein kinase regulates myelination. J Mol Neurosci. Biomarker and Therapeutic Target in Multiple Sclerosis. Mol 2008; 35(1): 23–33, doi: 10.1007/s12031-007-9011-0, Neurobiol. 2016; 53(6): 3976–3991, doi: 10.1007/s12035- indexed in Pubmed: 17994198. 015-9336-6, indexed in Pubmed: 26184632. 21. Hu M, Zhang Y, Feng J, et al. Uterine progesterone signaling 5. Caput D, Beutler B, Hartog K, et al. Identification of a com- is a target for metformin therapy in PCOS-like rats. J Endo- mon nucleotide sequence in the 3’-untranslated region of crinol. 2018; 237(2): 123–137, doi: 10.1530/JOE-18-0086, mRNA molecules specifying inflammatory mediators. Proc indexed in Pubmed: 29535146. Natl Acad Sci U S A. 1986; 83(6): 1670–1674, indexed in 22. Huang C, Cao Z, Ma J, et al. AKR1B10 activates diacylglycerol Pubmed: 2419912. (DAG) second messenger in breast cancer cells. Mol Car- 6. Cargnello M, Roux PP. Activation and function of the MAPKs cinog. 2018; 57(10): 1300–1310, doi: 10.1002/mc.22844, and their substrates, the MAPK-activated protein kinases. indexed in Pubmed: 29846015. Microbiol Mol Biol Rev. 2011; 75(1): 50–83, doi: 10.1128/ 23. Hwang InK, Kim DW, Yoo KY, et al. Ischemia-induced MMBR.00031-10, indexed in Pubmed: 21372320. changes of platelet endothelial cell adhesion molecule-1 7. Carlos TM, Clark RS, Franicola-Higgins D, et al. Expression of in the hippocampal CA1 region in gerbils. Brain Res. 2005; endothelial adhesion molecules and recruitment of neutro- 1048(1-2): 251–257, doi: 10.1016/j.brainres.2005.04.049, phils after traumatic brain injury in rats. J Leukoc Biol. 1997; indexed in Pubmed: 15913570. 61(3): 279–285, indexed in Pubmed: 9060450. 24. Kaminska B, Gozdz A, Zawadzka M, et al. MAPK signal trans- 8. Chen J, Leskov IL, Yurdagul A, et al. Recruitment of the duction underlying brain inflammation and gliosis as thera- adaptor protein Nck to PECAM-1 couples oxidative stress peutic target. Anat Rec (Hoboken). 2009; 292(12): 1902–1913, to canonical NF-kB signaling and inflammation. Sci Signal. doi: 10.1002/ar.21047, indexed in Pubmed: 19943344. 2015; 8(365): ra20, doi: 10.1126/scisignal.2005648, indexed 25. Kaur C, Sivakumar V, Zou Z, et al. Microglia-derived pro- in Pubmed: 25714462. inflammatory cytokines tumor necrosis factor-alpha and 9. Chen Z, Tzima E. PECAM-1 is necessary for flow-induced interleukin-1beta induce Purkinje neuronal apoptosis via vascular remodeling. Arterioscler Thromb Vasc Biol. 2009; their receptors in hypoxic neonatal rat brain. Brain Struct 29(7): 1067–1073, doi: 10.1161/ATVBAHA.109.186692, Funct. 2014; 219(1): 151–170, doi: 10.1007/s00429-012- indexed in Pubmed: 19390054. 0491-5, indexed in Pubmed: 23262920. 10. Choi JH, Roh KH, Oh H, et al. Caffeic acid phenethyl ester lessens 26. Kerman M, Kanter M, Coşkun KK, et al. Neuroprotective disease symptoms in an experimental autoimmune uveoretinitis effects of caffeic acid phenethyl ester on experimental mouse model. Exp Eye Res. 2015; 134: 53–62, doi: 10.1016/j. traumatic brain injury in rats. J Mol Histol. 2012; 43(1): exer.2015.03.014, indexed in Pubmed: 25795054. 49–57, doi: 10.1007/s10735-011-9376-9, indexed in Pub- 11. Coso OA, Chiariello M, Yu JC, et al. The small GTP-binding med: 22124729. proteins Rac1 and Cdc42 regulate the activity of the JNK/SAPK 27. Kyosseva SV. Mitogen-activated protein kinase signaling. signaling pathway. Cell. 1995; 81(7): 1137–1146, indexed in Int Rev Neurobiol. 2004; 59: 201–220, doi: 10.1016/S0074- Pubmed: 7600581. 7742(04)59008-6, indexed in Pubmed: 15006489. 12. Deddens LH, van Tilborg GAF, van der Toorn A, et al. PECAM- 28. Lavoie JN, L’Allemain G, Brunet A, et al. Cyclin D1 expression 1-targeted micron-sized particles of iron oxide as MRI con- is regulated positively by the p42/p44MAPK and negatively trast agent for detection of vascular remodeling after cerebral by the p38/HOGMAPK pathway. J Biol Chem. 1996; 271(34): ischemia. Contrast Media Mol Imaging. 2013; 8(5): 393–401, 20608–20616, indexed in Pubmed: 8702807. doi: 10.1002/cmmi.1536, indexed in Pubmed: 23740809. 29. Lertkiatmongkol P, Liao D, Mei H, et al. Endothelial functions 13. Demeule M, Labelle M, Régina A, et al. Isolation of en- of platelet/endothelial cell adhesion molecule-1 (CD31). dothelial cells from brain, lung, and kidney: expression of Curr Opin Hematol. 2016; 23(3): 253–259, doi: 10.1097/ the multidrug resistance P-glycoprotein isoforms. Biochem MOH.0000000000000239, indexed in Pubmed: 27055047. Biophys Res Commun. 2001; 281(3): 827–834, doi: 10.1006/ 30. Mandyam CD, Villalpando EG, Steiner NL, et al. Platelet bbrc.2001.4312, indexed in Pubmed: 11237734. endothelial cell adhesion molecule-1 and oligodendro- 14. Draper HH, Hadley M. Malondialdehyde determination as in- genesis: significance in alcohol use disorders. Brain Sci. dex of lipid Peroxidation. Methods Enzymol. 1990: 421–431, 2017; 7(10), doi: 10.3390/brainsci7100131, indexed in doi: 10.1016/0076-6879(90)86135-i. Pubmed: 29035306. 15. Duan S, Shao G, Yu L, et al. Angiogenesis contributes to the 31. Mao LM, Wang JQ. Synaptically localized mitogen-activated neuroprotection induced by hyperbaric oxygen precondition- protein kinases: local substrates and regulation. Mol Neu- ing against focal cerebral ischemia in rats. Int J Neurosci. robiol. 2016; 53(9): 6309–6315, doi: 10.1007/s12035-015- 2015; 125(8): 625–634, doi: 10.3109/00207454.2014.956 9535-1, indexed in Pubmed: 26567109. 101, indexed in Pubmed: 25171223. 32. Marmarou A, Foda MA, van den Brink W, et al. A new 16. Fukuda K, Aihara N, Sagar SM, et al. Purkinje cell vulner- model of diffuse brain injury in rats. Part I: Patho- ability to mild traumatic brain injury. J Neurotrauma. 1996; physiology and biomechanics. J Neurosurg. 1994; 80(2): 13(5): 255–266, doi: 10.1089/neu.1996.13.255, indexed in 291–300, doi: 10.3171/jns.1994.80.2.0291, indexed in Pubmed: 8797175. Pubmed: 8283269.

228 A. Çetin, E. Deveci, Evaluation of PECAM-1 and p38 MAPK expressions in cerebellar tissue

33. Minden A, Lin A, Claret FX, et al. Selective activation of 50. Tan J, Liu D, Lv X, et al. MAPK mediates inflammatory the JNK signaling cascade and c-Jun transcriptional activity response and cell death in rat pulmonary microvascular by the small GTPases Rac and Cdc42Hs. Cell. 1995; 81(7): endothelial cells in an ischemia-reperfusion model of lung 1147–1157, indexed in Pubmed: 7600582. transplantation. J Heart Lung Transplant. 2013; 32(8): 34. Murohara T, Delyani JA, Albelda SM, et al. Blockade of platelet 823–831, doi: 10.1016/j.healun.2013.05.005, indexed in endothelial cell adhesion molecule-1 protects against myo- Pubmed: 23747218. cardial ischemia and reperfusion injury in cats. J Immunol. 51. Vaporciyan AA, DeLisser HM, Yan HC, et al. Involvement of 1996; 156(9): 3550–3557, indexed in Pubmed: 8617985. platelet-endothelial cell adhesion molecule-1 in neutrophil 35. Nguyen V, Sabeur K, Maltepe E, et al. Sonic hedgehog agonist recruitment in vivo. Science. 1993; 262(5139): 1580–1582, protects against complex neonatal cerebellar injury. Cerebel- indexed in Pubmed: 8248808. lum. 2018; 17(2): 213–227, doi: 10.1007/s12311-017-0895- 52. Wada T, Penninger JM. Mitogen-activated protein kinases in 0, indexed in Pubmed: 29134361. apoptosis regulation. Oncogene. 2004; 23(16): 2838–2849, 36. Pabón MM, Acosta S, Guedes VA, et al. Brain region-specific doi: 10.1038/sj.onc.1207556, indexed in Pubmed: 15077147. histopathological effects of varying trajectories of controlled 53. Waskiewicz AJ, Cooper JA. Mitogen and stress response cortical impact injury model of traumatic brain injury. pathways: MAP kinase cascades and phosphatase regula- CNS Neurosci Ther. 2016; 22(3): 200–211, doi: 10.1111/ tion in mammals and yeast. Curr Opin Cell Biol. 1995; 7(6): cns.12485, indexed in Pubmed: 26775604. 798–805, indexed in Pubmed: 8608010. 37. Paglia DE, Valentine WN. Studies on the quantitative and 54. Winzen R, Kracht M, Ritter B, et al. The p38 MAP kinase qualitative characterization of erythrocyte glutathione per- pathway signals for cytokine-induced mRNA stabilization oxidase. J Lab Clin Med. 1967; 70(1): 158–169, indexed in via MAP kinase-activated protein kinase 2 and an AU- Pubmed: 6066618. rich region-targeted mechanism. EMBO J. 1999; 18(18): 38. Potts MB, Adwanikar H, Noble-Haeusslein LJ. Models 4969–4980, doi: 10.1093/emboj/18.18.4969, indexed in of traumatic cerebellar injury. Cerebellum. 2009; 8(3): Pubmed: 10487749. 211–221, doi: 10.1007/s12311-009-0114-8, indexed in 55. Wolters TL, Netea MG, Hermus AR, et al. IGF1 potentiates Pubmed: 19495901. the pro-inflammatory response in human peripheral blood 39. Privratsky JR, Newman PJ. PECAM-1: regulator of en- mononuclear cells via MAPK. J Mol Endocrinol. 2017; dothelial junctional integrity. Cell Tissue Res. 2014; 355(3): 59(2): 129–139, doi: 10.1530/JME-17-0062, indexed in 607–619, doi: 10.1007/s00441-013-1779-3, indexed in Pubmed: 28611056. Pubmed: 24435645. 56. Wu N, Kurosu T, Oshikawa G, et al. PECAM-1 is involved in 40. Rose BA, Yokota T, Chintalgattu V, et al. Cardiac myocyte BCR/ABL signaling and may downregulate imatinib-induced p38a kinase regulates angiogenesis via myocyte-endothelial apoptosis of Philadelphia chromosome-positive leukemia cell cross-talk during stress-induced remodeling in the heart. cells. Int J Oncol. 2013; 42(2): 419–428, doi: 10.3892/ J Biol Chem. 2017; 292(31): 12787–12800, doi: 10.1074/jbc. ijo.2012.1729, indexed in Pubmed: 23233201. M117.784553, indexed in Pubmed: 28637870. 57. Xia Z, Dickens M, Raingeaud J, et al. Opposing effects of 41. Rosenblum WI, Murata S, Nelson GH, et al. Anti-CD31 delays ERK and JNK-p38 MAP kinases on apoptosis. Science. 1995; platelet adhesion/aggregation at sites of endothelial injury in 270(5240): 1326–1331, indexed in Pubmed: 7481820. mouse cerebral arterioles. Am J Pathol. 1994; 145(1): 33–36, 58. Xiao H, Liu X, Wang Y, et al. Angiotensin(17) prevents lipopol- indexed in Pubmed: 8030753. ysaccharideinduced hepatocellular inflammatory response by 42. Roy AR, Delgado-Olguin P. Visualizing the vascular network inhibiting the p38MAPK/AP1 signaling pathway. Mol Med in the mouse embryo and yolk sac. Methods Mol Biol. 2018; Rep. 2018; 17(4): 5492–5497, doi: 10.3892/mmr.2018.8527, 1752: 11–16, doi: 10.1007/978-1-4939-7714-7_2, indexed indexed in Pubmed: 29393446. in Pubmed: 29564758. 59. Xie Y, Muller WA. Molecular cloning and adhesive properties 43. Rubinfeld H, Seger R. The ERK cascade as a prototype of of murine platelet/endothelial cell adhesion molecule 1. Proc MAPK signaling pathways. Methods Mol Biol. 2004; 250: Natl Acad Sci U S A. 1993; 90(12): 5569–5573, indexed in 1–28, doi: 10.1385/1-59259-671-1:1, indexed in Pub- Pubmed: 8516303. med: 14755077. 60. Yang G, Li T, Xu J, et al. Mitogen-activated protein kinases 44. Sahu V, Nigam L, Agnihotri V, et al. Diagnostic significance regulate vascular reactivity after hemorrhagic shock through of p38 isoforms (p38a, p38b, p38g, p38d) in head and neck myosin light chain phosphorylation pathway. J Trauma squamous cell carcinoma: comparative serum level evalu- Acute Care Surg. 2013; 74(4): 1033–1043, doi: 10.1097/ ation and design of novel peptide inhibitor targeting the TA.0b013e31828586a2, indexed in Pubmed: 23511142. same. Cancer Res Treat. 2019; 51(1): 313–325, doi: 10.4143/ 61. Yang SH, Sharrocks AD, Whitmarsh AJ. Transcriptional regula- crt.2018.105, indexed in Pubmed: 29747487. tion by the MAP kinase signaling cascades. Gene. 2003; 320: 45. Sari E, Bakar B, Sarkarati B, et al. Effectiveness of dimethylsulfoxide 3–21, indexed in Pubmed: 14597384. on the survival and volume preservation of autologous fat graft 62. Zhang J, Zhao D, Na N, et al. Renoprotective effect of eryth- tissue: a preliminary study. Aesthet Surg J. 2016; 36(2): NP58– ropoietin via modulation of the STAT6/MAPK/NF-kB pathway NP67, doi: 10.1093/asj/sjv119, indexed in Pubmed: 26242854. in ischemia/reperfusion injury after renal transplantation. Int 46. Seo TB, Kim BK, Ko IG, et al. Effect of treadmill exercise on J Mol Med. 2018; 41(1): 25–32, doi: 10.3892/ijmm.2017.3204, Purkinje cell loss and astrocytic reaction in the cerebellum indexed in Pubmed: 29115389. after traumatic brain injury. Neurosci Lett. 2010; 481(3): 63. Zhang X, Liu J, Zhang P, et al. Silibinin induces G1 arrest, 178–182, doi: 10.1016/j.neulet.2010.06.087, indexed in apoptosis and JNK/SAPK upregulation in SW1990 human Pubmed: 20603186. pancreatic cancer cells. Oncol Lett. 2018; 15(6): 9868–9876, 47. Springer T. Adhesion receptors of the immune system. Na- doi: 10.3892/ol.2018.8541, indexed in Pubmed: 29805688. ture. 1990; 346(6283): 425–434, doi: 10.1038/346425a0. 64. Zhang YY, Kong LQ, Zhu XD, et al. CD31 regulates me- 48. Sun L, Zou LX, Wang J, et al. Mucin 4 gene silencing re- tastasis by inducing epithelial-mesenchymal transition in duces oxidative stress and calcium oxalate crystal formation hepatocellular carcinoma via the ITGB1-FAK-Akt signaling in renal tubular epithelial cells through the extracellular pathway. Cancer Lett. 2018; 429: 29–40, doi: 10.1016/j. signal-regulated kinase signaling pathway in nephrolithiasis canlet.2018.05.004, indexed in Pubmed: 29746931. rat model. Kidney Blood Press Res. 2018; 43(3): 820–835, 65. Zhu J, Ji Y, Yu Y, et al. Knockdown of serine/threonine protein doi: 10.1159/000490136, indexed in Pubmed: 29843125. phosphatase 5 enhances gemcitabine sensitivity by promot- 49. Sun Y, Oberley LW, Li Y. A simple method for clinical assay ing apoptosis in pancreatic cancer cells . Oncol Lett. 2018; of superoxide dismutase. Clin Chem. 1988; 34(3): 497–500, 15(6): 8761–8769, doi: 10.3892/ol.2018.8363, indexed in indexed in Pubmed: 3349599. Pubmed: 29805615.

The effect of prolonged formalin fixation on the staining characteristics of archival human brain tissue A. Alrafiah1, 3, R. Alshali2, 3

1Department of Laboratory Technology, Faculty of Applied Medical Sciences, King Abdulaziz University, Jeddah, Saudi Arabia 2Department of Anatomy, Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi Arabia 3Neuroscience Research Unit, Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi Arabia

[Received: 16 August 2018; Accepted: 3 October 2018]

Background: Neurodegenerative disorders include wide range of conditions, which affect millions of people worldwide. Unfortunately, they are incurable and irreversibly progressive. Immunohistochemical staining of paraffin-fixed tissues for both diagnostic and research purposes are widely used. However, large amount of brain tissues are fixed but little is known about whether they are suitable for retrospective studies. The study aimed at investigating the effects of prolonged formalin fixation time on immunohistochemical expression of some common neurodegenerative markers in archival brain specimens. Materials and methods: Twenty brain specimens were obtained from human cadavers in the Anatomy Department of King Abdulaziz University that were prefixed in 10% formalin. They were divided into two equal groups according to time of fixation: group 1 — less than 1 year, group 2 —up to 20 years. Histological examination of white and grey matter was done using haematoxylin and eosin, luxol fast blue (LFB) for myelin staining, Con- go red for amyloid plaques, CD 68 for microglial cells, tenascin-C (large extracellular matrix glycoprotein) and caspase 3 antibody for apoptotic cells. Results: For both groups, corpus callosum sections displayed myelination with LFB staining. The distribution of CD 68 positive microglial cells was evident in frontal and temporal grey matter, but not in corpus callosum sections. Strongly positive masses were seen in Congo red-stained frontal and temporal sections. Anti-caspase 3 immunostaining revealed positively stained neurons. Conclusions: Histological and immunohistochemical techniques yielded reproducible staining results when applied to human brain tissue stored in formalin for long periods; so they can be used in well preserved biobank material which are the most targeting research areas in neuropathology. (Folia Morphol 2019; 78, 2: 230–236)

Key words: human brain, immunohistochemistry, formalin fixed, neurodegenerative markers

Address for correspondence: Dr. A. Alrafiah, Department of Laboratory Technology, Faculty of Applied Medical Sciences, King Abdulaziz University, P.O. Box 80200, Jeddah 21589, Saudi Arabia, tel: +966 (01) 26401000, ext. 23495, fax: +966 (01) 26401000, ext. 21686, e-mail: [email protected]

230 A. Alrafiah, R. Alshali, Effect of prolonged fixation on neurodegenerative markers

INTRODUCTION tion and repair [18]. Tenascin-C is also considered Paraformaldehyde (PFA) is the most commonly a potential biomarker induced during inflammation used fixative material that preserves the gross anat- or cancer [17]. Caspase 3 is essential for normal brain omy of organs and the histological details of tissue development and cell apoptosis [21]. LFB stain, also compared to other fixatives like acetone and alcohol. known as (myelin stain), is commonly used to detect For formalin fixation process, tissue is soaked in 10% demyelination in the CNS [5, 12]. CR is the diagnos- neutral buffered formalin for 24–48 h [8]. tic dye for amyloidosis, which is the hallmark of Formalin fixation is a chemical process in which degenerative diseases such as Alzheimer’s disease formaldehyde forms covalent bonds and produces [6]. The CR staining is a reliable technique to detect cross-links. Thus, this fixation may limit antibodies amyloid deposits in tissues, especially if it is confused permeation into the tissue, decrease the accessibility with other materials when stained by haematoxylin of antigenic determinants and alter the structure of and eosin (H&E). the epitopes to be visualised by immunohistochem- Our aim in this study is to determine the effect of istry (IHC) [16, 20]. Moreover, prolonged time of for- prolonged formalin fixation on immunohistochemical malin fixation can modify the structure of the epitope expression of a panel of common neurodegenerative to be visualised by IHC [16] and hence, reduce the markers in archival human brain specimens. The study antigenicity and efficiency of IHC [20]. Despite that, of postmortem human brain tissue is a powerful researches up to date revealed that some antigens method to investigate various neurological diseases, are sensitive to the fixation time more than others thus providing better insight into their pathogenesis. [4, 16, 31]. However, only few studies showed that human postmortem brain tissue can be fixed for MATERIALS AND METHODS long periods up to several decades, and that some Twenty brain specimens were obtained from hu- epitopes could be preserved for years before declin- man cadavers in the Anatomy Department of King Ab- ing in their immunoreactivity [7]. In contrast, other dulaziz University that were prefixed in 10% formalin. studies indicated that prolonged fixation may affect They were divided into two equal groups according to the efficiency of IHC [1]. time of fixation: group 1 — less than 1 year, group 2 Neurodegenerative diseases (NDDs) refer to — up to 20 years. a wide range of conditions in which the neurons in the Five-micrometre (4 µm) sections of formalin-fixed, brain and spinal cord become injured, degenerated paraffin-embedded (FFPE) tissue from the frontal and or eventually die. Those neuron cells actually cannot temporal gyrus region were processed and sectioned. regenerate themselves nor be replaced by the body. Histological examination of white and grey matter Nowadays, NDDs became more common, espe- was done using H&E, LFB for myelin staining, CR for cially Parkinson’s, Alzheimer’s and multiple sclerosis. amyloid plaques, and IHC was performed to assess Unfortunately, they are incurable and irreversibly pro- the expression of common molecular markers for gressive conditions that lead to hopeless situations neurodegenerative diseases using antibodies against starting from mental malfunctioning (dementias) to CD 68 for microglial cells, tenascin-C (large extracel- more serious movement impairments (ataxias). Those lular matrix glycoprotein) and caspase 3 antibody for NDDs have some common markers such as CD 68, apoptotic cells (Table 1). tenascin-C and caspase 3 that can be identified by Immunohistochemistry was performed using fully IHC. Also, luxol fast blue (LFB) and Congo red (CR) automated Leica Bond machine. Epitope retrieval was are specific stains confirming degenerative changes. performed in all samples using Leica ER2 solution o CD 68 is a useful marker for microglia, which (AR9640) at 97 C. Sections were placed in H2O2 (3%) are specialised macrophage cells in the brain [11]. for 30 min in methanol. Microglia are activated and become phagocytes in response to central nervous system (CNS) infection RESULTS or damage [24], which occurs in multiple sclerosis, Histological examination of H&E-stained sections Alzheimer’s disease and Parkinson’s disease. Tenas- of frontal and temporal grey matter and corpus cal- cin-C is expressed transiently by astrocytes during losum revealed normal appearance. Examination of embryonic development of the CNS [29]. Fibroblasts LFB-stained corpus callosum sections of both groups and other tissues express it during cell differentia- revealed darkly stained blue myelinated areas in some

231 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. Antibodies and conditions Antibody Species Clone/company Dilution Antigen retrieval technique Anti-tenascin-C Mouse Polyclonal/Merck Millipore (AB19013) 1:200 Automated ER1 Anti-CD 68 Mouse Clone kp1/Dako Prediluted Automated ER2 Caspase 3 Rabbit Ab2302 1:100 Automated ER1

sections whereas hypo-myelinated sections showed pale blue staining (Fig. 1A–C). In CD 68-immunostained sections of both studied groups, positively stained microglia cells were seen in frontal and temporal grey matter sections (Fig. 2A, B, respectively) while in corpus callosum sections no positive immunostaining was detected (Fig. 2C). Examination of CR-stained sections of frontal and temporal grey matter of both groups revealed dull red staining of the neurons as well as the neuropil (Fig. 3A). A Strongly positive congophilic masses of variable sizes were also seen among neurons (Fig. 3B, 3C). In some sections, the blood vessels revealed amorphous congophilic amyloid deposits in their walls (Fig. 3D). In tenascin-C-stained sections, differences in the staining properties were observed among studied groups. In corpus callosum sections, there was positive staining of the extra cellular matrix in both groups (Fig. 4A, B), while in frontal and temporal grey matter sections positive intracytoplasmic as well as extracellular staining were seen (Fig. 4C, D). Frontal and temporal grey matter sections of both groups that were stained with anti-caspase 3 antibod- B ies showed multiple positively stained neurons as well as some positively stained endothelial cells lining blood vessels (Fig. 5A, B). The caspase 3-positive immunostaining was seen to be cytoplasmic (Fig. 5C).

DISCUSSION In this work, we evaluate the effect of prolonged formalin fixation on the efficiency of IHC staining of archival human brain specimens. This approach has been put to most practical use through the use of specific antibodies raised against these ‘molecular markers’. Several biomarkers and stains for neurodegenerative diseases were used. C Luxol fast blue stain was used to assess the fea- sibility of myelin staining in archival sections. Our results were similar to those of Sheaffer et al. [22] who Figure 1. Luxol fast blue stained — corpus callosum sections; A. Group 1, showing dark blue areas of myelination; B. Group 1, examined brain sections, which were fixed in formalin showing pale blue areas denoting hypomyelination; C. Group 2, for up to 53 years, for myelin loss and astrocytosis showing areas of hypomyelination. with satisfactory results. Others studied the relation

232 A. Alrafiah, R. Alshali, Effect of prolonged fixation on neurodegenerative markers

Figure 2. CD 68 immunostaining; A. Frontal lobe section of group 1, showing positively stained microglial cells (arrows); B. Temporal lobe section of group 2, showing positively stained microglial cells (arrows); C. Corpus callosum section of group 1, showing negative immunostaining for microglial cells. D

Figure 3. Congo red stained — frontal lobe sections; A. Group 1, showing dull red staining of neurons (arrows) as well as neuropil (arrowhead); B. Group 2, showing some small positively stained congophilic masses (arrows); C. Group 1, showing large positively congophilic masses (arrows) among positively stained neurons; D. Group 2, showing congophilic deposits (arrows) within the wall of some blood vessels.

233 Folia Morphol., 2019, Vol. 78, No. 2

A B

C D

Figure 4. Tenascin-C immunostaining; A. Corpus callosum section of group 1, showing positive staining for tenascin-C in the extra cellular matrix (arrowhead); B. Corpus callosum section of group 2, showing positive staining for tenascin-C in the extra cellular matrix (arrowhead); C. Frontal grey matter section of group 2, showing positive staining for tenascin-C intracytoplasmic (arrow) as well as in the extra cellular matrix (arrowhead); D. Temporal grey matter section of group 1, showing positive staining for tenascin-C intracytoplasmic as well as in the extra cellular matrix (arrowhead).

A B

Figure 5. Anti-caspase 3 immunostaining of frontal grey matter section; A. Group 1, showing some caspase 3 positive neurons (arrows); B. Group 2, showing caspase 3 positive immunostained neurons (arrows) as well as in some endothelial lining of a blood C vessel (encircled); C. Group 2, showing cytoplasmic positive caspase 3 immunostaining (arrow).

234 A. Alrafiah, R. Alshali, Effect of prolonged fixation on neurodegenerative markers

between brain stiffness and myelin content in brain Our results point to the preservation of the an- sections stained with LFB and quantified using image tigenicity of some epitopes in archival human brain analyser. They stated that brain stiffness increases tissue prefixedin formalin for long time. In agreement with myelin content [28]. with our results, Webster et al. [27] revealed negligi- According to Walker and Lue [26], microglia can ble effects of prolonged time of formalin fixation on be detected in human brain sections, particularly in antigen detection for many used antibodies in IHC diseased brains, using CD 68-antibody. This may be staining in domestic animals. All results collectively attributed to changes in the properties of microglia encourage the use of IHC as a tool in diagnostic and can be considered as a common feature of human pathology [19]. neurodegenerative diseases [26]. That was consistent According to another study, exogenous factors with our results of CD 68 immunostaining to detect such as premortem or post mortem conditions can microglia in studied sections. affect postmortem brain tissue preservation [2]. The Ando et al. [3] found that during life, the accuracy question asked by Grillo et al. [9] is whether IHC on of amyloid deposition in brain by clinical diagnosis is archival formalin-fixed paraffin blocks has an expiry only 65%; however, definite diagnosis is reached by date? Grillo et al. [9] studied the factors affecting im- examination of the brain at autopsy. Previous work munoreactivity in paraffin embedded sections after showed that the detection of condensed beta-amyloid long term storage of formalin fixation. They stated in a secondary structure in brain parenchyma (amyloid that loss of antigenicity was relational to the age of plaques) and blood vessels can be confirmed using tissue section and mode of storage. The cold storage CR stain [30]. In our studied sections, CR staining has the minimum effect and heat pretreatment for successfully detected amyloid deposits in the wall of antigen retrieval was needed for all antigens. blood vessels, and in support to our data, Smith [23] stated that cerebral amyloid angiopathy is considered Conclusions a morphological characteristic of Alzheimer disease. It is concluded that long-term formalin fixation of However, it may be also seen in the brains of elderly human brains does not affect their antigenicity. There- people who don’t have any neurological problems. fore there is increasing demand for brain banks with The extracellular matrix glycoprotein tenascin-C is proper selection and collection of postmortem brain recognised in embryogenesis and wound healing as tissue. Information on individual’s previous neurologi- well as in some neoplasia, such as gliomas, where it cal condition must be also provided. This represents can be used as prognostic marker for tumour recur- a unique archive for scientific research worldwide rence [10]. The immunostained sections in this study using different histological (histochemistry and IHC) revealed varied staining pattern similar to results of techniques to study neurodegenerative diseases. a study comparing the expression of tenascin-C in grey and white matter of control human brains; sig- Acknowledgements nificant differences was found by using Western blot Neuroscience research unit at King Abdulaziz analysis or IHC [14]. Lee and McKinnon [13] stated University, Jeddah, Saudi Arabia. that apoptosis occurring in the nervous system may be seen either during normal development, or can be References induced by disease or DNA damage. 1. Alafuzoff I, Pikkarainen M, Al-Sarraj S, et al. Interlaboratory It has been stated that caspase 3 protein is es- comparison of assessments of Alzheimer disease-related lesions: a study of the BrainNet Europe Consortium. J Neu- sential in mediating cell apoptosis, because it is af- ropathol Exp Neurol. 2006; 65(8): 740–757, doi: 10.1097/01. fected by the cell surface death receptor [25]. On this jnen.0000229986.17548.27, indexed in Pubmed: 16896308. basis, our studied sections showed variable caspase 2. Alturkistani HA, Tashkandi FM, Mohammedsaleh ZM. 3 positive immunostaining. In support to our figures, Histological Stains: A Literature Review and Case Study. Glob J Health Sci. 2015; 8(3): 72–79, doi: 10.5539/gjhs. Machaalani et al. [15] studied the effect of tissue v8n3p72, indexed in Pubmed: 26493433. fixation on caspase 3 immunostaining in brain tissue. 3. Ando Y, Haraoka K, Terazaki H, et al. A novel tool for They found variations in the efficiency of IHC staining detecting amyloid deposits in systemic amyloidosis in using commonly available antibodies for caspase 3, vitro and in vivo. Lab Invest. 2003; 83(12): 1751–1759, which was dependent on the animal species and tis- doi: 10.1097/01.lab.0000101701.87433.c5. 4. Arber D. Effect of prolonged formalin fixation on the sue fixation method. immunohistochemical reactivity of breast markers. Appl

235 Folia Morphol., 2019, Vol. 78, No. 2

Immunohistochem Mol Morphol. 2002; 10(2): 183–186, 18. Midwood KS, Hussenet T, Langlois B, et al. Advances doi: 10.1097/00129039-200206000-00015. in tenascin-C biology. Cell Mol Life Sci. 2011; 68(19): 5. Carriel V, Campos A, Alaminos M, et al. Staining Meth- 3175–3199, doi: 10.1007/s00018-011-0783-6, indexed ods for Normal and Regenerative Myelin in the Nerv- in Pubmed: 21818551. ous System. Methods Mol Biol. 2017; 1560: 207–218, 19. Pikkarainen M, Martikainen P, Alafuzoff I. The effect of doi: 10.1007/978-1-4939-6788-9_15, indexed in Pub- prolonged fixation time on immunohistochemical staining med: 28155156. of common neurodegenerative disease markers. J Neu- 6. Clement CG, Truong LD. An evaluation of congo red fluo- ropathol Exp Neurol. 2010; 69(1): 40–52, doi: 10.1097/ rescence for the diagnosis of amyloidosis. Hum Pathol. NEN.0b013e3181c6c13d, indexed in Pubmed: 20010304. 2014; 45(8): 1766–1772. 20. Ploeger S, Guldemond JM, Feirabend HK, et al. Acidifica- 7. Evers P, Uylings H. An optimal antigen retrieval method tion of human brains stored in fixatives. Eur J Morphol. suitable for different antibodies on human brain tis- 1993; 31(4): 286–290. sue stored for several years in formaldehyde fixative. 21. Porter AG, Jänicke RU. Emerging roles of caspase-3 in apop- J Neurosci Methods. 1997; 72(2): 197–207, doi: 10.1016/ tosis. Cell Death Differ. 1999; 6(2): 99–104, doi: 10.1038/ s0165-0270(96)02204-2. sj.cdd.4400476, indexed in Pubmed: 10200555. 8. Fox CH, Johnson FB, Whiting J, et al. Formaldehyde 22. Sheaffer S, Rosoklija G, Dwork AJ. Myelin staining of archi- fixation. J Histochem Cytochem. 1985; 33(8): 845–853, val brain tissue. Clin Neuropathol. 1999; 18(6): 313–317. doi: 10.1177/33.8.3894502, indexed in Pubmed: 3894502. 23. Smith EE. Cerebral amyloid angiopathy as a cause of 9. Grillo F, Pigozzi S, Ceriolo P, et al. Factors affecting im- neurodegeneration. J Neurochem. 2018; 144(5): 651–658, munoreactivity in long-term storage of formalin-fixed doi: 10.1111/jnc.14157, indexed in Pubmed: 28833176. paraffin-embedded tissue sections. Histochem Cell Biol. 24. Stankov A, Belakaposka-Srpanova V, Bitoljanu N, et al. 2015; 144(1): 93–99, doi: 10.1007/s00418-015-1316-4, Visualisation of Microglia with the use of Immunohisto- indexed in Pubmed: 25757745. chemical Double Staining Method for CD-68 and Iba-1 10. Herold-Mende C, Mueller MM, Bonsanto MM, et al. Clinical of Cerebral Tissue Samples in Cases of Brain Contusions. impact and functional aspects of tenascin-C expression Pril (Makedon Akad Nauk Umet Odd Med Nauki). 2015; during glioma progression. Int J Cancer. 2002; 98(3): 36(2): 141–145, doi: 10.1515/prilozi-2015-0062, indexed 362–369, indexed in Pubmed: 11920587. in Pubmed: 27442380. 11. Holness CL, Simmons DL. Molecular cloning of cd68, 25. Sun Y, Xu Y, Geng L. Caspase-3 inhibitor prevents the a human macrophage marker related to lysosomal glyco- apoptosis of brain tissue in rats with acute cerebral infarc- proteins. Blood. 1993; 81(6): 1607–1613. tion. Exp Ther Med. 2015; 10(1): 133–138, doi: 10.3892/ 12. Klüver H, Barrera E. A Method for the Combined Staining of etm.2015.2462, indexed in Pubmed: 26170924. Cells and Fibers in the Nervous System. J Neuropathol Exp 26. Walker DG, Lue LF. Immune phenotypes of microglia in Neurol. 1953; 12(4): 400–403, doi: 10.1097/00005072- human neurodegenerative disease: challenges to detect- 195312040-00008. ing microglial polarization in human brains. Alzheimers 13. Lee Y, McKinnon PJ. Detection of apoptosis in the central Res Ther. 2015; 7(1): 56, doi: 10.1186/s13195-015-0139-9, nervous system. Methods Mol Biol. 2009; 559: 273–282, indexed in Pubmed: 26286145. doi: 10.1007/978-1-60327-017-5_19, indexed in Pub- 27. Webster JD, Miller MA, Dusold D, et al. Effects of pro- med: 19609763. longed formalin fixation on diagnostic immunohistochem- 14. Leins A, Riva P, Lindstedt R, et al. Expression of tenascin-C in istry in domestic animals. J Histochem Cytochem. 2009; various human brain tumors and its relevance for survival in 57(8): 753–761, doi: 10.1369/jhc.2009.953877, indexed patients with astrocytoma. Cancer. 2003; 98(11): 2430–2439, in Pubmed: 19398606. doi: 10.1002/cncr.11796, indexed in Pubmed: 14635078. 28. Weickenmeier J, de Rooij R, Budday S, et al. Brain stiffness 15. Machaalani R, Radford JL, Waters KA. Tissue fixation effects increases with myelin content. Acta Biomater. 2016; 42: on immunohistochemical staining of caspase-3 in brain 265–272, doi: 10.1016/j.actbio.2016.07.040, indexed in tissue. Appl Immunohistochem Mol Morphol. 2007; 15(4): Pubmed: 27475531. 463–470, doi: 10.1097/01.pai.0000213150.56687.78, 29. Wiese S, Karus M, Faissner A. Astrocytes as a source for indexed in Pubmed: 18091392. extracellular matrix molecules and cytokines. Front Phar- 16. Mason JT, O’Leary TJ. Effects of formaldehyde fixation on macol. 2012; 3: 120, doi: 10.3389/fphar.2012.00120, protein secondary structure: a calorimetric and infrared indexed in Pubmed: 22740833. spectroscopic investigation. J Histochem Cytochem. 1991; 30. Wilcock DM, Gordon MN. Quantification of cerebral 39(2): 225–229, doi: 10.1177/39.2.1987266, indexed in amyloid angiopathy and parenchymal amyloid plaques Pubmed: 1987266. with congo red histochemical stain. Nat Protoc. 2006; 17. Midwood KS, Orend G. The role of tenascin-C in tissue 1(3): 1591–1595. injury and tumorigenesis. J Cell Commun Signal. 2009; 3(3- 31. Williams JH, Mepham BL, Wright DH. Tissue preparation 4): 287–310, doi: 10.1007/s12079-009-0075-1, indexed for immunocytochemistry. J Clin Pathol. 1997; 50(5): in Pubmed: 19838819. 422–428, doi: 10.1136/jcp.50.5.422.

Aging changes in the retina of male albino rat: a histological, ultrastructural and immunohistochemical study M.E.I. Mohamed, E.A.A. El-Shaarawy, M.F. Youakim, D.M.A. Shuaib, M.M. Ahmed

Department of Anatomy and Embryology, Faculty of Medicine, Cairo University, Cairo, Egypt

[Received: 9 June 2018; Accepted: 26 June 2018]

Background: Degenerative changes caused by aging may affect the eye, especially the retina. Such changes occur as a part of normal physiological process and may be irreversible. The aim of the study was to demonstrate the influence of aging on the morphology of the retina to provide a basis to explain the pathogenesis of age-associated decline in visual acuity, scotopic and photopic sensitivity. Materials and methods: Forty male albino rats were used and divided into four age groups (group I: age of cortical maturity, group II: middle-aged, group III: aged group and group IV: senile group). The rats were sacrificed, the eye balls were enucleated. Intra-vitreal injections of formalin for haematoxylin and eosin and immunohistochemical sections, glutaraldehyde for toluidine blue semithin and E/M ultra-thin sections were performed. Measurements and quantitative histomorphometric estimation of the layers of the retina were done. Results: Light microscopic examination revealed age-dependent attenuation of photoreceptor striations. Aged and senile groups presented pyknotic, widely- -spaced nuclei of the outer nuclear layer. The inner nuclear layer was thinned out to 2 or 3 cellular rows. Retinal capillaries showed progressive dilatation and congestion. Statistical analysis proved significant thinning of the retina with variable degrees of thinning of the constituting layers. Decreased arborisation with age was confirmed with quantification of synaptophysin-immunostained sections. Glial fibrillary acidic protein immunostaining revealed the picture of reactive gliosis. On the ultrastructural level, the retinal pigment epithelium exhibited major alterations with aging. Numerous phagosomes, lipofuscin and melanolipofusin granules appeared within the cells, together with exaggerated basal infoldings. The photoreceptor nuclei became degenerated and the perinuclear space was widened. Conclusions: Rat retinae clearly undergo age-related morphological changes. Such changes provide a cellular base for explanation of decreased vision in humans with aging other than reflection errors. Effect of aging was not only qualitative, but also quantitative. (Folia Morphol 2019; 78, 2: 237–258)

Key words: retina, ageing, ultrastructure, immunohistochemical, histology

Address for correspondence: Dr. E.A.A. Ahmed El-Shaarawy, Professor of Anatomy and Embryology, Department of Anatomy and Embryology, Faculty of Medicine, Cairo University, Kasr Al Aini Street, Cairo, Egypt, tel: 0112 01001582811, e-mail: [email protected]; [email protected]

237 Folia Morphol., 2019, Vol. 78, No. 2

INTRODUCTION lary acidic protein (GFAP) in glial cells [9]. The aim of The development of the visual system involves the present study was to determine the histological a sequence of neurochemical signals which regulate and immunohistochemical changes in the retina of synaptic connections and those processes modulate male albino rats that occur with aging and demon- the visual experience [23]. Older people may experi- strate the ultrastructural age-related alterations and ence alteration of vision, even in the absence of any their correlation with the histological changes. identifiable eye disease. Advancing age has been associated with decrease in visual acuity and reduction MATERIALS AND METHODS in scotopic sensitivity, contrast sensitivity and motion Animals sensitivity as well as poor performance on different The present study was carried out on 40 male visual discrimination tasks [19]. albino rats. The rats were selected to fall equally into The vertebrate retina, like other parts of the cen- four age groups, chosen in correspondence to human tral nervous system, is subjected to degenerative [33] as follows: changes caused by aging. In addition, the retina is — group I: 6 months old (age of cortical maturity); a site of diseases for which age is a major risk fac- — group II: 12 months old (middle-aged); tor. The retina is the best understood part of the — group III: 24 months old (aged); vertebrate central nervous system with regard to — group IV: 30 months old (senile). its cellular patterning, circuitry and function [32]. The rats were obtained from the Animal House, The retina is composed of five major neuron types: Research Institute of Ophthalmology. They were retinal ganglion cells, amacrine cells, bipolar cells, housed in cages, 5 rats/cage, under standard labo- horizontal cells and photoreceptors that integrate ratory and environmental conditions 2 weeks before visual information and send it to the brain [32]. The selection. The rats were sacrificed by cervical disloca- neuronal somas are located in three nuclear layers tion and the eye balls were enucleated. The upper whose axonal and dendritic processes form complex nasal part of the sclera was pierced 1.5 mm posterior and orderly networks of chemical and electrical syn- to the limbus, with a 30 gauge Hamilton needle fit- apses forming two plexiform layers; outer and inner. ted to a syringe. The needle tip was visualised during Synaptophysin is an integral membrane protein of the the procedure to avoid retinal injury, with intraocular synaptic vesicles. It possibly serves multiple functions injections in the posterior vitreous of the eye behind in synaptic vesicle formation and exocytosis, playing the lens [38] either with formalin for haematoxylin an important role in neurotransmitter delivery. It is and eosin (H&E) and immunohistochemical sections widely used as one of the synaptic function markers. or with glutaraldehyde for toluidine blue stain semi- The expression pattern of synaptophysin in retina in- thin and E/M ultrathin sections. dicates an important role in ribbon synapse formation and visual signal transmission [11]. Retinal macroglia Methodology (MG), consisting of astrocytes and Müller cells, play Light microscopy. The right eye balls were inject- key roles in the homeostasis of retinal neurons, keep ed with 10% formalin in the posterior chamber of the the retina healthy and functioning properly. Mamma- eye behind the lens and left for 24 h. The peripheral lian MG are quiescent and respond to retinal damage retinae were obtained and processed. by reactive gliosis, rarely with proliferation, and do Histological examination using H&E stain. not naturally regenerate neurons [18]. Immunohistochemical staining done for: Astrocytes, almost entirely restricted to the retinal — GFAP: was performed to detect glial cell reactivity nerve fibre layer, have close relationship with neurons [8]. Positive control (brown discoloration of the and the major blood vessels. They are commonly glial cells mainly astrocytes denoting increase glio- thought to play an important part in the proper de- sis) was IMR5 cells in brain. For negative controls, velopment and functioning of the vascular system in incubation was carried out with the omission of the retina, including blood flow and formation of the the primary antiserum. blood-retinal barrier [21]. Astrocyte activation and re- — Synaptophysin (SYN) was applied, to detect syn- active gliosis are common traits in neurodegenerative aptic functions [2]. SYN is a synaptic vesicle glyco- processes. A hallmark of gliosis is the upregulation of protein. It is present in almost all neuroendocrine intermediate filament proteins, including glial fibril- cells that participate in synaptic transmission.

238 M.E.I. Mohamed et al., Aging of retina

microscope in the Electron Microscope Research Labo- ratory, Faculty of Agriculture Research Centre. Morphometric study. Image J image analysis software was used to analyse the photomicrographs of the different groups taken at magnification¥ 400 using Leica ICC50 microscope equipped with digital camera. The following measurements were done (Figs. 1a–c): — The thickness of the retina, inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL) and outer nuclear layer (ONL) thickness: to quantify the thickness of the retina and its nuclear and synaptic layers, images were taken at equivalent retinal eccentricities from the optic nerve head. Layer thickness was measured in two to four areas from each retina. The measurements were done using semithin epoxy resin sections photomicrographs. — Measurement of the density of synapses: using the photomicrographs of SYN-immunostained sections. — Measurement of the ganglion cell count/300 µm ganglion cell layer (GCL): using the photomicrographs of H&E sections.

Statistical analysis Statistical measures were done using SPSS programme, version 20. One way analysis of variance (ANOVA) was employed to compare means between groups. Bonferroni post hoc test was used to detect significance between every two individual groups. The mean, standard deviation and the p value were calculated for each experimental group and compared among the different groups. Significance was consid- Figure 1. An image displayed on the monitor screen of a field in rat retina showing: a. Measuring of the thickness of different layers; ered when p-value was ≤ 0.05. The data were exam- b. Measuring of synaptophysin optical density; c. Ganglion cell ined by the Kolmogorov-Smirnov test for normality. counting; OLM — outer limiting membrane; ONL — outer nuclear layer; OPL — outer plexiform layer; INL — inner nuclear layer; IPL — inner plexiform layer. RESULTS Light microscopic results Group I (cortical maturity). The retina of group I featured the 10 layers, from outside inwards: retinal Its ubiquity at the synapse has led to the use of pigment epithelium (RPE), photoreceptor layer of rods SYN immunostaining for the quantification of and cones with outer lightly stained segment and in- synapses [36]. ner deeply stained segment, outer limiting membrane Electron microscopy. The retinae of the left eye (OLM), outer nuclear layer, outer plexiform layer, in- balls were prepared for ultrastructural sections for ner nuclear layer, inner plexiform layer, ganglion cell examination using transmission electron microscope. layer, nerve fibre layer and inner limiting membrane Parts of the peripheral retinae were obtained and (ILM) (Fig. 2a, b). The RPE appeared as a single layer processed. The sections were examined and pho- of cuboidal cells with ill-defined boundaries and pale, tographed by JEOL JEM 1400 transmission electron oval, vesicular nuclei. This layer was lying on Bruch’s

239 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. A micrograph of rat retina (group I) showing: a. The arrangement of the 10 layers, from choroid (C) inwards as: retinal pigment epithelium (RPE), photoreceptor layer (Ph) with outer lightly stained outer segment (OS) and inner deeply stained inner segment (IS), outer limiting membrane (OLM), outer nuclear layer (ONL), outer plexiform layer (OPL), inner nuclear layer (INL), inner plexiform layer (IPL), ganglion cell layer (GCL), nerve fibre layer (NFL) and inner limiting membrane (arrows); H&E ×400; b. A semithin section with 10 layers, from outside inwards; retinal pigment epithelium (RPE), photoreceptor layer (Ph) of rods and cones with outer segment (OS) and inner segment (IS), outer limiting membrane (OLM), outer nuclear layer (ONL), outer plexiform layer (OPL), inner nuclear layer (INL), inner plexiform layer (IPL), ganglion cell layer (GCL), nerve fibre layer (NFL) and inner limiting membrane (arrows); toluidine blue ×1250; c. Semithin section of rat retina (group I) showing the retinal pigment epithelium (RPE) layer with pale nuclei (N) lying on the Bruch’s membrane (arrowheads). Choriocapillaries (CC) are seen beneath the Bruch’s membrane. The outer nuclear layer (ONL) is formed of multiple rows of nuclei of the photoreceptors. The outer plexiform layer (OPL) forms a dense network of synapses; toluidine blue ×1250; d. Retinal ganglion cell layer (GCL), covered with the nerve fibre layer (NFL). The inner plexiform layer (IPL) is formed of network of synapses; H&E ×1000; e. Semithin section of rat retina (group I) showing the four types of cells identified in inner nuclear layer (INL); horizontal cells (H), bipolar cells (B), amacrine cells (A) and Müller cells (M). In ganglion cell layer (GCL), the cells exhibit pale nuclei (N) and their axons (arrowsheads) exit to form nerve fiber layer (NFL). The inner limiting mebrane (arrows) is seen; toluidine blue ×1250.

240 M.E.I. Mohamed et al., Aging of retina

Figure 3. A micrograph of rat retina (group II) showing: a. The ganglion cell layer (GCL) exhibiting a dilated blood vessel (arrow) extending into the inner plexiform layer (IPL). The choriocapillaries (CC) are congested; H&E ×400; b. Photoreceptors outer segments (OS) exhibiting fragmentation (short arrows). Exudate (arrowheads) on the inner limiting membrane (ILM) and dilated blood vessels (long arrows) in the outer plexiform layer are observed; toluidine blue ×1250; c. A higher magnification of the inset in the previous figure revealing attenuated striations in the outer (OS) and the inner (IS) segments of photoreceptors. The retinal pigment epithelium (RPE) is shown; toluidine blue ×1250.

membrane which separates it from the choroid (Fig. 2c). nuclei and angulated Müller cells in between the other The photoreceptor layer appeared as fibrillary striations, cell types (Fig. 2e). The ganglion cells was aligned as composed of an outer lightly stained segment and a single layer of nuclei (Fig. 2d, e). Their axons exit as the an inner, deeply stained segment. The outer nuclear nerve fibrelayer limited internally by the inner limiting layer was formed of nuclei of photoreceptors which membrane (Fig. 2e). form multiple deeply-stained rows (Fig. 2b, c). Their Group II (middle-aged). The rat retina of group II axons synapsed with dendrites of inner nuclear layer displayed mild dilatation of blood vessels in ganglion cells forming the dense outer plexiform layer (Fig. 2c). cell layer extending into the inner plexiform layer A thicker network of synapses, the inner plexiform layer, (Fig. 3a) and in the outer plexiform (Fig. 3b). Exuda- is formed by arborisation of axons of the inner nuclear tion of the inner limiting membrane was seen in layer cells, together with dendrites of the ganglion cell rare foci of this group (Fig. 3b), photoreceptors layer layer (Fig. 2d). Four types of cells could be identified in showed attenuation of normal striation and mild the inner nuclear layer: horizontal cells with large pale outer segment fragmentation (Fig. 3b, c). nuclei, bipolar cells with smaller, rounded or oval nuclei Group III (aged). Group III showed apparent de- and chromatin clumps, amacrine cells with indented crease in the whole retinal thickness (Fig. 4a) especial-

241 Folia Morphol., 2019, Vol. 78, No. 2

Figure 4. A micrograph of rat retina (group III) showing: a. Thinning out of the retina. A dilated blood vessel in ganglion cell layer and dilated congested choriocapillaries (CC) are observed. Cells of the ganglion cell layer (GCL) exihibit irregular distribution. The photoerceptor layer (Ph) shows attenuated striations; H&E ×400; b. Thinning of the inner plexiform layer (IPL) and the inner nuclear layer (INL); H&E ×1000; c. Dilated blood vessel (arrow) in the outer plexiform layer (OPL) which encroaches on the inner nuclear layer (INL). Spacing between cells of the outer nuclear layer (ONL) is seen; H&E ×1000; d. Fragmentation of photoreceptor outer segments (OS), low cell density is observed in both outer (ONL) and inner (INL) nuclear layers. Pyknotic nuclei (white arrows) and ghost cells (black arrows) ware seen. A dilated blood vessel with thickened wall (arrowhead) in ganglion cell layer is also seen; toluidine blue ×1250; e. Loss of striations in photoreceptor outer segments (OS). Dark irregular pyknotic nuclei (arrows) and ghost cells (arrowheads) are seen in the outer nuclear layer (ONL). Müller cell processes (Mp) are observed in inner plexiform layer (IPL); toluidine blue ×1000.

ly of the inner plexiform and the inner nuclear layers The inner retina was thinned out and the inner (Fig. 4b). The photoreceptor outer segments showed nuclear layer was formed of two to three cellular rows attenuated striations (Fig. 4a) while the outer nuclear (Fig. 4b–d). The ganglion cells displayed irregular layer showed spacing between the photoreceptor nu- distribution along the ganglion cell layer (Fig. 4a). clei (Fig. 4c). Semithin sections showed fragmentation Dilatation of retinal blood vessels was observed of the photoreceptor outer segments (Fig. 4d). Some (Fig. 4a, c, d) with thickening of their wall (Fig. 4d). photoreceptor nuclei appeared pyknotic; moreover, Group IV (senile). The rat retina of group IV ghost cells were encountered (Fig. 4d, e). presented thinning of different layers (Fig. 5a) and

242 M.E.I. Mohamed et al., Aging of retina

Figure 5. A micrograph of rat retina (group IV) showing: a. Thinning out of the retina particularly the inner nuclear layer (INL) and the inner plexiform layer (IPL). Congestion of choriocapillaries (CC), spacing between photoreceptor nuclei in the outer nuclear layer (ONL) and irregular distribution of cells (arrowheads) in the ganglion cell layer (GCL) are seen; H&E ×400; b. Two to three cellular rows forming the inner nuclear layer (INL) and marked spacing between outer nuclear layer (ONL) cells. Ganglion cells (GC) are irregularly distributed; H&E ×1000; c. Areas of severe dilatation and congestion of choriocapillaries (arrowheads); H&E ×1000; d. Dilated congested choriocapillaries (CC), foamy appearance (arrowheads) of photoreceptor outer segments (OS) and degenerating nuclei (arrows) in the outer nuclear layer (ONL); toluidine blue ×1250; e. Degenerating photoreceptor nuclei (arrows) in the outer nuclear layer (ONL), severely affected inner nuclear layer (INL) cells (arrowheads) and rarity of cells in the ganglion cell layer (GCL); toluidine blue ×1250.

the inner nuclear layer was formed only of two to Group II sections showed positive immune reaction three cellular rows (Fig. 5b). Choriocapillaries were for GFAP labelling the nerve fibre layer and Müller markedly dilated and congested (Fig. 5s, c, d). The cell processes extending into the inner portion of IPL photoreceptor layer exhibited a foamy appearance (Fig. 6c). SYN showed strong positive reaction in the (Fig. 5d). The outer nuclear layer showed spacing IPL and OPL (Fig. 6d). GFAP immunostaining revealed between cells (Fig. 5a, b) and contained many ir- strong positive reaction (in group III) at the nerve fibre regular, degenerating nuclei (Fig. 5d, e). The inner layer together with staining of Müller cells processes nuclear layer displayed severe damage in some throughout the retinal layers (Fig. 6e). SYN showed specimens (Fig. 5e). The ganglion cells exhibited ir- moderate reaction in the inner and outer plexiform regular distribution throughout this layer (Fig. 5a, b) layers (Fig. 6f). In group IV stained with GFAP im- with rarity of cells in some specimens (Fig. 5e). munostaining, numerous Müller cell processes were Regarding the GFAP staining of rat retina in con- stained positively and seen traversing the inner and trol group I showed positive immune reaction which outer plexiform layers; along with a strong positive is confined to the nerve fibre layer (Fig. 6a). SYN reaction of astrocytes at the nerve fibre layer (Fig. 6g). staining revealed strong positive immune reaction The inner and outer plexiform layers gave a weak in the two synaptic layers, IPL and OPL (Fig. 6b). positive reaction to synaptophysin (Fig. 6h).

243 Folia Morphol., 2019, Vol. 78, No. 2

Figure 6. Micrograph of rat retina showing: a. Group I with glial fibrillary acidic protein (GFAP) immunoreactivity is confined to nerve fibre layer (NFL); GFAP ×400); b. Group I strong positive brown reaction to synaptophysin (SYN) in both outer (OPL) and inner (IPL) plexiform layers; SYN ×400; c. Group II showing positive immune reaction GFAP in the nerve fibre layer (NFL) and Müller cell processes (arrows) extending into the inner portion of the inner plexiform layer (IPL); GFAP ×400; d. Group II revealing strong immunoreactivity to synaptophysin. in the outer (OPL) and inner (IPL) plexiform layers; SYN ×400; e. Group III showing strong positive reaction at the nerve fibre layer (NFL) and Müller cell processes (arrows) extending through the inner (IPL) and outer plexiform (OPL) layers; GFAP ×400; f. Group III revealing moderate immunoreaction in both inner (IPL) and outer plexiform layer (OPL); SYN ×400; g. Group IV showing numerous Müller cell processes (arrows) extending through the inner (IPL) and outer (OPL) plexiform layers. Also, strong positive reaction of astrocytes (arrowheads) at the nerve fibre layer (NFL) was seen; GFAP ×400; h. Group IV showing weak positive immune reaction for synaptophysin, in the inner (IPL) and the outer (OPL) plexiform layers; SYN ×400).

244 M.E.I. Mohamed et al., Aging of retina

Figure 7. An electron micrograph of rat retina (group I) showing: a. Retinal pigment epithelium (RPE)/Bruch’s membrane (BrM)/choriocapillaries (CC) complex. RPE cells have oval euchromatic nucleus (N) and contain melanin granules (m) in their cytoplasm. The BrM separates the pigment epithelium from the choroid and CC; b. The five layers of Bruch’s membrane: basal lamina (BL), inner collagenous layer (ICL), elastic lamina (EL), outer collagenous layer (OCL) and basement membrane (arrow) of CC. The retinal pigment epithelium (RPE) lies on the Brush’s membrane; c. Photoreceptor rod outer segments (ROS) and cone outer segment (COS) with packed membranous discs (arrows); d. A higher magnification of the inset of Figure 7c revealing the plasmalemmal invaginations (arrows) in the cone outer segment (COS) which are absent in rod outer segment (ROS); e. The photoreceptors outer (OS) and inner (IS) segments. Inner segments show longitudinal arrangement of mitochondria (Mit) and abundant rough endoplasmic reticulum (arrowheads); f. The outer nuclear layer (ONL) is separated from the photoreceptor layer (Ph) by the outer limiting membrane (arrowheads). Photoreceptor outer segments (OS) are shown; g. The outer limiting membrane (arrowheads) separating photoreceptor somas (s) and nuclei (N) from photoreceptor inner segments (IS). Photoreceptor nuclei (N) are located at different distance from the outer limiting membrane; h. The photoreceptor nuclei (N) forming the outer nuclear layer (ONL). Synap- tic ribbons (arrows) are seen within the outer plexiform layer (OPL) which lies between two nuclear layers; the outer (ONL) and the inner (INL) nuclear layers; i. A higher magnification of the inset of Figure 7h revealing the outer plexiform layer (OPL) contains rod spherules (RS) and the cone pedicles (CP). Both contain ribbon synapses (arrowheads); j. different population of cells of inner nuclear layer (INL); bipolar cells (B), amacrine cells (A) and horizontal cell (H). All are surrounded by Müller cell processes (Mp).

Electron microscopic results The photoreceptors’ inner segments contained Group I (cortical maturity). The retina of group longitudinal tubular mitochondria and abundant I showed that the RPE had oval euchromatic nuclei rough endoplasmic reticulum (Fig. 7e). The outer and the cytoplasm contained melanin granules in the limiting membrane appeared between the outer apical portion of the cell. Basal infoldings appeared nuclear layer and the photoreceptor layer (Fig. 7f) lying on Bruch’s membrane which separates the pig- separating photoreceptors somas and nuclei from ment epithelium from the choroid and choriocapillar- their inner segments (Fig. 7g). Nuclei of rods and ies (Fig. 7a). Bruch’s membrane featured its 5 layers; cones were located at different distances from basal lamina, inner collagenous layer, elastic lamina, the outer limiting membrane (Fig. 7g) forming the outer collagenous layer and basement membrane of outer nuclear layer (Fig. 7h). choriocapillaries (Fig. 7b). The outer plexiform layer revealed the synaptic The photoreceptor outer segments were attached terminals of photoreceptors with the globular rod to the RPE. They contained well organised, horizon- terminals (rod spherules) and the expanded cone tally aligned membranous discs (Fig. 7c). Outer seg- terminals (cone pedicles) (Fig. 7i). Synaptic terminals ments of cones were conical in shape while those of contained sharp synaptic ribbons (Fig. 7h, i). rods were straight (Fig. 7c, d). Cones outer segments The inner nuclear layer contained heterogeneous featured plasmalemmal invaginations at one side population of cells; bipolar cells with heterochromatic which were absent in rods (Fig. 7d). nuclei, horizontal cells with euchromatic nuclei and

245 Folia Morphol., 2019, Vol. 78, No. 2

Figure 8. An electron micrograph of rat retina (group II) showing: a. Detached part of membranous discs (arrowheads) within the retinal pigment epithelium (RPE). Lipofuscin granules (L) are seen in the RPE; b. Phagosomes (arrows) and melanolysosomes (MLL) inside a retinal pigment epithelial cell; BrM — Bruch’s membrane; c. Electron-dense deposits (arrows) are present in the outer collagenous layer (OCL) of Bruch’s membrane (BrM); d. Mild malalignment of membranous discs (arrows) of photoreceptors outer segment (OS); e. The photoreceptor inner segments (IS), containing longitudinally arranged mitochondria (Mit) and well-organized microtubules (arrows); f. Loss of mitochondrial cristae (arrowheads) within the inner segments (IS). Mit — mitochondria; g. Degeneration of few photoreceptor nuclei (arrow) within the outer nuclear layer (ONL); in addition to thickening of Müller cell processes (Mp). Outer limiting membrane (arrowheads) is shown; h. A higher magnification of the Figure 8g revealing the photoreceptor nuclei (N) with widening of the perinuclear space (*). Müller cell processes (MP) intervening the space between photoreceptor somas, with their apical processes form the outer limiting membrane (arrowheads); i. The rod spherules (RS) and the cone pedicles (CP). Synaptic ribbons (arrows) appear as sharp well-defined bars; j. Vacuolations (V) in the cells of inner nuclear layer (INL) with decreased electro-density of their axons (arrows) in the inner plexiform layer.

amacrine cells with heterochromatic nuclei and pale occasionally noticed (Fig. 8h). The synaptic ribbons indented cytoplasm. The Müller cell processes were appeared sharp and well defined inside the photore- filling the background between other cells (Fig. 7j). ceptors synaptic terminals in both rod spherules and Group II (middle-aged). Examination of the cone pedicles (Fig. 8i). Some cells of the inner nuclear different layers of the retinae of group II revealed layer exhibited vacuolations (Fig. 8j) with decrease in mild changes. Some of the RPE cells contained de- the electron density of their axons located in the inner tached parts of outer segment membranous discs plexiform layer (Fig. 8j). and lipofuscin granules (Fig. 8a). Phagosomes ap- Group III (aged). The RPE presented whirling peared within the cells, in addition to melanolyso- of basal folds, numerous phagosomes (Fig. 9a) as somes (Fig. 8b). well as lipofuscin and melanolipofuscin granules Bruch’s membrane showed electron-dense depos- (Fig. 9b). Filling of the subretinal space was encoun- its, especially in outer collagenous layer (Fig. 8c). The tered (Fig. 9b) together with poor attachment of membranous discs of photoreceptor outer segments outer segments to the RPE (Fig. 9b). Thickening of showed mild malalignment in some specimens (Fig. Bruch’s membrane was encountered, in addition to 8d); the inner segments showed apparently normal dilatation of choriocapillaries (Fig. 9c). Choriocapillar- arrangement of microtubules and longitudinally ies wall presented proliferation of the basal lamina, settled mitochondria (Fig. 8e) with occasional loss degeneration of the endothelial cells and extrusion of mitochondrial cristae (Fig. 8f). The outer nuclear of pericytes were observed in (Fig. 9d). The photo- layer displayed degeneration of few photoreceptor recerptor outer segment showed malalignment of the nuclei, concomitant with thickening of Müller cell membranous discs, in addition to areas of distorted processes which form the matrix background be- discs (Fig. 9e). The photoreceptor inner segments tween the photoreceptor nuclei (Fig. 8g). Widening contained long convoluted mitochondria (Fig. 9f). De- of the perinuclear space of photoreceptor nuclei was fective mitochondrial cristae, numerous phagosomes

246 M.E.I. Mohamed et al., Aging of retina

Figure 9. A photomicrograph of rat retina (group III) showing: a. Whirling of basal folds (black arrow), numerous phagolysosomes (white arrows) in the retinal pigment epithelium (RPE); b. That the subretinal space is filled (arrows). Retinal pigment epithelium (RPE) exhibits numer ous lipofuscin (LF) and melanolipofuscin (MLF) granules; c. Thickening of the Bruch’s membrane (BrM) and dilated choriocapillaries (CC). Retinal pigment epithelium (RPE) is seen; d. Dilatation of choriocapillaries (CC) with proliferation of the basement membrane (BM) and extrusion of pericytes (Pe). Projection inside the vessel lumen was observed (arrow); e. Distortion of the membranous discs (arrowheads); f. Numerous phagosomes (arrowheads), defective mitochondrial cisterns (Mit) and dilated congested blood vessel (arrow) are seen in inner segments (IS); g. Cells of the outer nuclear layer (ONL) are variable-sized with the intervening Müller cell processes (Mp) appear thicker. The photoreceptor nuclei show degeneration (arrows); h. Degeneration of photoreceptor nuclei (arrows), associated with low electron-density of rod terminals (RT). Dilated blood vessels (BV) are seen in outer plexiform layer (OPL) encroaching the inner nuclear layer (INL). Cells of inner nuclear layer show sever vacuolation (V); i. Swollen synaptic ribbons (arrows) in the rod terminals (RT) and fragmented ribbons (black arrow) in the cone pedicle (CP) were encountered. Lipofuscin (LF) granules and inclusion bodies (arrowheads) were seen in the rod terminals; j. Ganglion cells (GC) with cytoplasmic vacuolation (V). Müller cell processes (Mp) are highly electron-dense together with faint axons (arrows) in the inner plexiform layer (IPL). Nerve fibre layer (NF) is seen.

and dilated congested blood vessels were seen within pofuscin granules (Fig. 10a). The subretinal space inner segments (Fig. 9f). appeared obliterated with dense microvilli sent from Cells of the outer nuclear layer were of variable the apical aspect of the RPE (Fig. 10b). The photore- size separated with broadened Müller cell processes ceptor outer segments showed major disorganisa- (Fig. 9g). Some nuclei displayed degeneration (Fig. tion of the membranous discs (Fig. 10c) Moreover, 9g, h). The outer plexiform layer showed decreased empty portions within the plasmalemmal envelops electron-density of the rod terminals (Fig. 9h). Syn- were seen (Fig. 10a, c). aptic ribbons decreased in number and appeared The outer limiting membrane was thickened as the fragmented (Fig. 9i) and swollen (Fig. 9i). Rod synaptic apical portions of Müller cell processes were thickened terminals contained lipofuscin granules and inclu- (Fig. 10d); also breaks in the OLM were encountered sion bodies (Fig. 9i). Dilated blood vessels were seen (Fig. 10e). The photoreceptor nuclei exhibited exten- in the outer plexiform layer encroaching the inner sive degeneration (Fig. 10e, f, g) with widening of the nuclear layer. Their basement membranes displayed perinuclear space (Fig. 10f, g). The outer plexiform thickening and lamination (Fig. 9h). Inner nuclear layer appeared as a thin layer of low electron-dense layer cells exhibited severe vacuolations (Fig. 9h). synaptic terminals (Fig. 10g). The synaptic terminals Müller cell processes were highly electron-dense revealed evident loss of synaptic vesicles and frag- between faint axons in the inner plexiform layer mentation of the synaptic ribbons (Fig. 10h). Inner (Fig. 9j). Ganglion cells presented severe cytoplasmic nuclear layer cells exhibited severe vacuolations with vacuolation (Fig. 9j). poor cytoplasmic outlines in some cells (Fig. 10i, j). Group IV (senile). Retinal pigment epithelium The synaptic terminals of the bipolar and amacrine showed numerous large lipofuscin and melanoli- cells appeared as empty figures (Fig. 10j). Decreased

247 Folia Morphol., 2019, Vol. 78, No. 2

Figure 10. A photomicrograph of rat retina (group IV) showing: a. Numerous lipofuscin (LF) and melanolipofuscin (MLF) granules are seen in the retinal pigment epithelium (RPE); b. The photoreceptors/ retinal pigment epithelium (RPE)/Bruch’s membrane complex with empty plasmalemmal chambers (arrows) and obliterated subretinal space (arrowheads). The photoreceptors outer segments (OS) are seen; c. Empty potions (arrows) within the photoreceptor outer segments (OS) together with sever distortion of (arrowheads) the remaining discs; d. Thick, electro-dense Müller cell processes (arrowheads) causing thickening of the outer limiting membrane. Congested blood vessels (arrows) are seen; e. A higher magnification of the previous figure. The photoreceptor nuclei exhibit degeneration (arrow). Breaks in the outer limiting membrane (arrowheads) are encountered; f. Distorted nuclear envelop and increased perinuclear space (arrowheads) in addition to nuclear degeneration (arrow); g. Ghost-like synaptic terminals are seen in the outer plexiform layer (black arrows), vacuolations (V) in bipolar (B) and horizontal (H) cells, degeneration of photoreceptor nuclei (white arrow) and increased perinuclear space (arrowheads); h. Marked loss of synaptic vesicles (arrowheads) and fragmentation of the synaptic ribbons (arrows) inside the synaptic terminals (ST); i. That the cells of the inner nuclear layer (INL) exhibit cytoplasmic vacuolations (V). Müller cell processes (MP) are of high electron density. Synaptic terminals are seen within the outer plexiform layer (OPL); j. The inner nuclear layer (INL) with bipolar (B) and amacrine (A) cells show severe vacuolation (V) with poor cytoplasmic outlines of the cells (arrows). The synaptic terminals of both cells appear as empty-like figures in the inner plexiform layer (IPL); k. A dilated blood vessels (BV) in the ganglion cell layer. Together with the empty figures of bipolar axons and areas of decreased arborisation (arrows) in the inner plexiform layer (IPL); l. Disrupted cytoplasmic membrane (arrow) of retinal ganglion cell (GC) together with sever vacuolation (V). Nerve fibre (NF) bundles are apparently smaller; m. A dilated congested blood vessel (BV) with thick basement membrane (arrow). Thick detached inner limiting membrane (ILM) is seen; n. Thickening of the internal (white arrow) and external (black arrow) basement membranes. Pericytes (Pe) send cytoplasmic processes (arrowheads) within the thickened basement membrane. Endothelial cells are degenerated and can’t be seen; o. The previous figure at a higher magnification. Pericytes show electro-dense inclusions (arrows) and long cytoplasmic processes (arrowheads) within the vessel basement membrane. Marked proliferation of capillary basal lamina is demonstrated (thick arrow).

arborisation (Fig. 10k) were encountered in the inner blood vessels of different retinal layers, both in- plexiform layer. Disrupted cytoplasmic membrane of ternal and external basement membranes were retinal ganglion cells together with severe vacuolation thickened, along with endothelial cells degen- were characteristic. Nerve fibre bundles were appar- eration (Fig. 10n) and basal lamina proliferation. ently smaller (Fig. 10l). Thickening and detachment Pericytes sent cytoplasmic processes extending were encountered in other areas (Fig. 10m). through the thickened basement membrane The ganglion cell layer showed dilatation and (Fig. 10n, o) and exhibited electron-dense cyto- congestion of blood vessels (Fig. 10m). In the plasmic inclusions (Fig. 10o).

248 M.E.I. Mohamed et al., Aging of retina

Table 1. The mean thickness of different layers of the retina in different age groups Parameter Group I Group II Group III Group IV ILM-OLM thickness [µm] 152.25 ± 17.30 159.21 ± 21.62 112.99 ± 5.90 102.08 ± 23.44 IPL thickness [µm] 73.92 ± 6.29 76.41 ± 9.27 49.51 ± 3.27 43.95 ± 4.56 INL thickness [µm] 30.04 ± 3.03 29.63 ± 6.26 18.09 ± 2.36 11.78 ± 4.05 OPL thickness [µm] 12.44 ± 1.75 11.88 ± 3.19 9.43 ± 1.61 6.26 ± 1.65 ONL thickness [µm] 47.07 ± 2.08 38.11 ± 6.65 33.01 ± 4.94 33.98 ± 9.39

Data are shown as mean ± standard deviation. ILM-OLM — inner limiting membrane-outer limiting membrane; IPL — inner plexiform layer; INL — inner nuclear layer; OPL — outer plexiform layer; ONL — outer nuclear layer

Table 2. Comparison of the mean thickness of the inner limiting membrane-outer limiting membrane (ILM-OLM) among the different age groups Group ILM-OLM thickness [µm] (mean ± SD) Comparative groups P (ANOVA test) Group I 152.25 ± 17.30 Group II 1.000 Group III 0.000* Group IV 0.000* Group II 159.21 ± 21.62 Group I 1.000 Group III 0.000* Group IV 0.000* Group III 112.99 ± 5.90 Group I 0.000* Group II 0.000* Group IV 0.230 Group IV 102.08 ± 23.44 Group I 0.000* Group II 0.000* Group III 0.230

*p-value ≤ 0.05 statistically significant; SD — standard deviation

Statistical and histomorphometric results IPL thickness. The mean thickness of IPL was ILM-OLM thickness. The mean thickness of ILM- 73.92 ± 6.29 µm and it was non-significantly -OLM in group I was 152.25 ± 17.30 µm and in different from IPL thickness which was 76.41 ± group II was 159.21 ± 21.62 µm. The difference in ± 9.27 (Tables 1, 3; Fig. 11B). The mean thickness of the the mean thickness between the two groups was IPL significantlydecreased in group III to be 49.51 ± found to be statistically non-significant (p = 1.0) ± 3.27; as compared with group I and II (p = 0.000 (Tables 1, 2; Fig. 11A). The mean thickness of ILM- and p = 0.000). IPL mean thickness became 43.95 ± -OLM in group III was 112.99 ± 5.90 µm. The mean ± 4.56 in group IV. This found to be highly statistically thickness was less in group III as compared to group I significant as compared with groups I, II and III and group II and these difference were found to be (p = 0.000, p = 0.000 and p = 0.014, respectively). statistically significant(p = 0.000). In group IV, the INL thickness. Group I mean INL thickness was mean thickness of ILM-OLM was 102.08 ± 23.44 µm. 30.04 ± 3.03 and it was 29.63 ± 6.26 in group II There was age related statistical significant decrease (p = 1.0). INL of group III decreased significantly as com- in the thickness of ILM-OLM as compared to group pared with group I and II (p = 0.000) for both (Tables 1, 4; I and group II (p = 0.000) for both. The difference Fig. 11C). INL thickness in group IV was 11.78 ± 4.05. in the mean thickness of ILM-OLM between group There was age related significant difference between III and group IV was found to be statistically non- group IV and group I (p = 0.000) and significant dif- significant (p = 0.2). ference between group IV and group III (p = 0.000).

249 Folia Morphol., 2019, Vol. 78, No. 2

Figure 11. The mean thickness in the different groups (error bars: ± standard deviation); A. Inner limiting membrane-outer limiting membrane (ILM-OLM); B. Inner plexiform layer (IPL); C. Inner nuclear layer (INL); D. Outer plexiform layer (OPL); E. Outer nuclear layer (ONL).

OPL thickness. The mean thickness was 12.44 ± ONL thickness. The mean thickness of ONL was ± 1.75 in group I, 11.88 ± 3.19 in group II, 9.43 ± 47.07 ± 2.08 in group I and was 38.11 ± 6.65 in ± 1.61 in group III and 6.26 ± 1.65 in group IV. The group II. The mean thickness decreased to 33.01 ± difference between group I and II was statistically ± 4.94 in group III; this was statistically significant non-significant; while a statistically significant when compared with group I and II (p = 0.000 and decrease was calculated between group II and III 0.033 respectively) (Table 6, Fig. 11E). There was (p = 0.001) and between III and IV (p = 0.000) a highly significant decrease in ONL thickness be- (Table 5; Fig. 11D). tween group IV and group I (p = 0.000), while the

250 M.E.I. Mohamed et al., Aging of retina

Table 3. Comparison of the mean thickness of the inner plexiform layer (IPL) among the different age groups Group IPL thickness [µm] (mean ± SD) Comparative groups P (ANOVA test) Group I 73.92 ± 6.29 Group II 0.972 Group III 0.000* Group IV 0.000* Group II 76.41 ± 9.27 Group I 0.972 Group III 0.000* Group IV 0.000* Group III 49.51 ± 3.27 Group I 0.000* Group II 0.000* Group IV 0.014* Group IV 43.95 ± 4.56 Group I 0.000* Group II 0.000* Group III 0.014*

*p-value ≤ 0.05 statistically significant; SD — standard deviation

Table 4. Comparison of the mean thickness of the inner nuclear layer (INL) among the different age groups Group INL thickness [µm] (mean ± SD) Comparative groups P (ANOVA test) Group I 30.04 ± 3.03 Group II 1.000 Group III 0.000* Group IV 0.000* Group II 29.63 ± 6.26 Group I 1.000 Group III 0.000* Group IV 0.000* Group III 18.09 ± 2.36 Group I 0.000* Group II 0.000* Group IV 0.000* Group IV 11.78 ± 4.05 Group I 0.000* Group II 0.000* Group III 0.000*

*p-value ≤ 0.05 statistically significant; SD — standard deviation

Table 5. Comparison of the mean thickness of the outer plexiform layer (OPL) among the different age groups Group OPL thickness [µm] (mean ± SD) Comparative groups P (ANOVA test) Group I 12.44 ± 1.75 Group II 1.000 Group III 0.000* Group IV 0.000* Group II 11.88 ± 3.19 Group I 1.000 Group III 0.001* Group IV 0.000* Group III 9.43 ± 1.61 Group I 0.001* Group II 0.000* Group IV 0.000* Group IV 6.26 ± 1.65 Group I 0.000* Group II 0.000* Group III 0.000*

*p-value statistically significant; SD — standard deviation

251 Folia Morphol., 2019, Vol. 78, No. 2

Table 6. Comparison of the mean thickness of the outer nuclear layer (ONL) among the different age groups Group ONL thickness [µm] (mean ± SD) Comparative groups P (ANOVA test) Group I 47.07 ± 2.08 Group II 0.000* Group III 0.000* Group IV 0.000* Group II 38.11 ± 6.65 Group I 0.000* Group III 0.033* Group IV 0.140 Group III 33.01 ± 4.94 Group I 0.000* Group II 0.033* Group IV 1.000 Group IV 33.98 ± 9.39 Group I 0.000* Group II 0.140 Group III 1.000

*p-value statistically significant; SD — standard deviation

Table 7. Comparison of the mean optical density of synaptophysin among the different age groups Group Synaptophysin optical density (mean ± SD) Comparative groups P (ANOVA test) Group I 0.94 ± 0.03 Group II 0.000* Group III 0.000* Group IV 0.000* Group II 0.79 ± 0.03 Group I 0.000* Group III 0.000* Group IV 0.000* Group III 0.66 ± 0.03 Group I 0.000* Group II 0.000* Group IV 0.000* Group IV 0.40 ± 0.03 Group I 0.000* Group II 0.000* Group III 0.000*

*p ≤ 0.05 statistically significant; SD — standard deviation

Figure 12. The mean (error bars: ± standard deviation); A. Optical density of synaptophysin in the different groups; B. Number of the retinal ganglion cells (RGC)/300 µm in the different groups.

252 M.E.I. Mohamed et al., Aging of retina

Table 8. Comparison of the mean number of retinal ganglion cells (RGC) layer among the different age groups Group Number of RGC/300 µm (mean ± SD) Comparative groups P (ANOVA test) Group I 20.20 ± 3.19 Group II 0.000* Group III 0.000* Group IV 0.000* Group II 9.20 ± 1.76 Group I 0.000* Group III 1.000 Group IV 1.000 Group III 8.60 ± 1.04 Group I 0.000* Group II 1.000 Group IV 1.000 Group IV 8.60 ± 1.04 Group I 0.000* Group II 1.000 Group III 1.000

*p ≤ 0.05 statistically significant; SD — standard deviation

difference between group IV and group II (p = 0.14) epithelial layer during early phases of aging. Signs and between group IV and group III (p = 1.00) was included presence of numerous phagolysosomes and found to be statistically non-significant. melanolysosomes in addition to exaggerated basal Synaptophysin optical density. The mean den- infoldings. In the aged rat retina, whirling exten- sity of synapses was 0.94 ± 0.03 in group I, 0.79 ± sions of the basal membrane into the cytoplasm ± 0.03 in group II, 0.66 ± 0.03 in group III and were characteristic in the RPE cells. These findings 0.40 ± 0.03 in group IV. The difference among the were consistent with the data provided by Boya et different groups was found statistically significant al. [6] who observed progressive changes in the RPE/ (Table 7, Fig. 12A). Bruch’s complex in aged Fischer rats. In the present Ganglion cell count. The mean number of retinal work, changes in pigmentation were among the ganglion cells/300 µm in group I was 20.20 ± 3.19 most significant and noticeable changes to occur (Table 8, Fig. 12B). Group II presented a statistically in RPE cells with age. During normal aging, the ap- significant decrease in ganglion cell count tobe pearance of the RPE layer was altered; there was 9.20 ± 1.76 (p = 0.000) (Table 8, Fig. 12B). The a decrease in the number of melanosomes and an mean number of retinal ganglion cells in group III increase in the number of lipofuscin granules. These was 8.60 ± 1.04 and in group IV was 8.60 ± 1.04. results were in agreement with results of Boya et al. The difference between groups II, III and IV were [6] who reported accumulation of lipids in the RPE, found to be statistically non-significant. for which the autofluorescence was indicative of lipofuscin. Boulton [5] submitted that RPE lipofuscin DISCUSSION is a result of both the incomplete degradation of In this work, effect of age on the morphology phagocytosed photoreceptor outer segments and of retinal layers was evident at the light, electron autophagic removal of damaged organelles-protein microscopic and immunohistochemical levels. The aggregates. Keeling et al. [20] added that chronic changes were encountered both in RPE and retinal degeneration of the RPE is a precursor to pathologi- neurons. Retinal neurons exhibited numerous age- cal changes in the outer retina. Impaired handling related quantitative and qualitative alterations in cells and processing in the endocytic/phagosome and and synapses, some of which could underlie declines autophagy pathways lead to the accumulation of in visual acuity. lipofuscin and chemically modified compounds In the present study, although light histological within RPE. These contribute to increased proteo- findings were scarce, the ultrastructural examination lytic and oxidative stress, resulting in irreversible revealed prominent changes in the retinal pigment damage to post-mitotic RPE cells.

253 Folia Morphol., 2019, Vol. 78, No. 2

Heterogeneous deposits in the form of melano- branous discs in few outer segments that progressed lipofucin granules were found in the present work to presence membranous whorls in the aged group. to increase within the RPE with age. Lipofuscin and The affection in the senile group was profound so melanolipofuscin granules started to appear not only that areas devoid of the discs inside the plasmalem- in the aged group, but also in the middle-aged group. mal envelops were encountered. Disorganised discs This was consistent with a manuscript published by were also reported by Nag and Wadhwa [28], who Warrburton et al. [35] which detected the accumula- studied the structural alterations of aging human tion of melanolipofuscin in human RPE from different retina and linked these changes to the effect of oxida- decades of life and assessed their phototoxicity to tive stress, inflammation and chronic light exposure. RPE cultures in vitro. Furthermore, in the present work, photoreceptor in- In the present study, senile group sections revealed ner segments presented numerous phagosomes, long rarity of melanin granules and the remaining melanin convoluted mitochondria and defective mitochondrial granules were conjugated with lipofuscin granules. cisterns. Similar changes were reported by Litts et al. Bonilha [4] suggested that melanolipofuscin might [22]. Nag and Wadhwa [28] suggested that the occur- not originate from photoreceptor outer segments rence of aging changes of mitochondria might be due phagocytosis but melanolipofuscin accumulates as to increased mitochondrial vulnerability via oxidative a result of the melanosomal autophagocytosis of RPE stress, light or toxic substances. Such changes might cells. That hypothesis was supported by the absence cause energy depletion and photoreceptor loss in of photoreceptor-specific proteins on analysis of the human retina with aging. composition of melanolipofuscin. Studies have sug- Transmission electron microscopy in the present gested that various components of these heteroge- work revealed thickening of the OLM. In the senile neous lipofuscin deposits may drive immune dysregu- group, there was prominent thickening of the apical lation via monocyte and microglial activation [24]. part of Müller cell processes, which form a network The present work revealed thickening of the of adherent junction complexes that form the OLM. Bruch’s membrane that was obvious and progressive In addition, breaks and interruption of the OLM were with aging. Electro-dense deposits were noticed espe- encountered. Hippert et al. [17] gave enlightenment cially in the inner collagenous layer, starting from the that photoreceptor death is likely to disrupt the OLM 12 months group. This was also reported by Szabadfi and compromises the orientation and polarity of the et al. [34] who clarified that aged Bruch’s membrane photoreceptors. displays an increase in phospholipids, triglycerides, In the present work, the retinal (OLM-ILM) thick- fatty acids and free cholesterol content which called ness displayed an initial increase in group II which advanced glycation end products. proved to be statistically non-significant. Then, the In this work, dilated choriocapillaries with prolifer- retinal thickness of group III decreased significantly ation of the basal lamina, degeneration of endothelial by 25.7% and further thinned by 32.9% of its original cells and extrusion of pericytes were observed. Fibro- thickness in group I. Significant thinning of the retina sis of the choriocapillaries was profound especially in was also observed by Nadal-Nicolás et al. [27]. On the the senile group. Chirco et al. [10] elucidated that the contrary, Szabadfi et al. [34] found increased OLM- choriocapillaries are especially susceptible to hypoxia. -ILM distance and IPL thickness in 36-month-old degu This dense layer allows the passage of oxygen and rat retinae. The authors explained their findings by nutrients to the RPE and photoreceptor cells and also the loose retinal structure observed in aging retinae. removes waste products via systemic circulation. They In the present work, the outer nuclear layer added that loss of the lining endothelial cells is a key showed spacing between the photoreceptor nuclei. contributor to the development of age-related retinal Some nuclei appeared fragmented and pyknotic on degenerative changes. light microscopic examination and degenerated on The present study found that the photoreceptors electron microscopic examination with widening of were evidently affected as the age increased. The pho- the perinuclear space. These changes appeared lim- toreceptor outer segments showed attenuated stria- ited in group II and became more prevalent and tions on the light microscopic level. On the electron evident in older groups. Concomitantly, the Müller microscopic level, the middle-aged group featured cell processes which form the matrix between the loss of the proper alignment of photoreceptors mem- photoreceptor nuclei appeared thicker. Similar find-

254 M.E.I. Mohamed et al., Aging of retina

ings were observed by Rodriguez-Muela et al. [30] oles were merged in the senile group so that bipolar who stated that the signs of neurodegeneration, and amacrine cells appeared to have ill-defined cell including increased cell death in the retinal photore- boundaries. In this work, IPL presented decreased ar- ceptor nuclei, start from 12 months of age onwards. borisation. Moreover, the synaptic terminals showed The present work demonstrated that the outer patchy affection in the 12 months group and were nuclear layer thickness decreased significantly in more severely affected appearing as empty figures group II compared with group I and showed further in the 24 months and 30 months groups. Further- significant decrease in group III. However, the senile more, Fernàndez-Sánchez et al. [16] found that at group presented mild increase in the ONL thickness 16 months of age, connectivity between photorecep- compared with group III, although the difference was tors and horizontal and bipolar cell processes were statistically non-significant. This increase in thickness lost over some areas. In the present work, the mor- may be explained by the loosened structure that may phological findings were confirmed by immunohis- occur in senile retinae Szabadfi et al. [34]. tochemical examination and statistical analysis. SYN The present work revealed age-dependent struc- immunostaining showed significant decrease in the tural changes at the ribbon synapses of photoreceptor trabecular meshwork of synapses in the inner plexi- synaptic terminals especially in rod spherules. Electron form layer. The decrease was progressive and almost microscopic examination revealed decrease in the steady among the studied age groups. The optical density of synaptic ribbons. Swollen, floated and density of SYN in group IV was found to be 42% fragmented ribbons were also encountered. These of its density in group I. This work concluded 33% changes were progressive with age. In addition, pro- decrease in IPL thickness in the aged group that was gressive decrease of synaptic vesicles in rod spherules magnified to 50% in the senile group. Nieves-Moreno were present in senile retinae. These findings were in et al. [29] estimated linear regression of IPL thickness accordance with Dorfman et al. [14] who observed with aging using optical coherence tomography on that the initial decline of ribbon synapses and the loss volunteers of different age groups. of synaptic sites was not complete in the aging rats. The present work revealed ganglion cells displayed In the present work, synaptophysin was used to vacuolations at the 12 months group. The degenera- mark synapses. Immunohistochemical staining re- tive vacuoles became more severe and extensive to- vealed significant decrease in arborisation between gether with disrupted cytoplasmic membrane based photoreceptor axons and dendrites of bipolar cells on electron microscopy observations in the 24 and and horizontal cells. Not only the density of synapses 30 months groups. The nerve fibre bundles were decreased, but also the thickness of this layer. OPL apparently smaller in senile group. Inner limiting thickness was halved in group IV compared with membrane affection, in the form of thickening and group I. Meanwhile, a statistically non-significant dif- exudation, was only observed in 30 months group. ference was calculated between the two adult groups. A study by Nag and Wadhwa [28] documented that Samuel et al. [31] pointed out that the thickness of the cystic mitochondria and cystoid bodies appear in gan- retina decreases with age and such thinning signifi- glion cells and nerve fibres in the aging human retina. cantly affected both synaptic layers as well as the INL. An important finding in this study was the apparent The present study showed that the INL was decrease and the irregular distribution of ganglion cells thinned out and was formed only of two to three over the GCL, detected on light microscopic examina- cellular rows in the aged and senile groups. The mor- tion of the aged and senile specimens. Statistical analy- phological findings was confirmed by the statistical sis elucidated a significant decrease in ganglion cell analysis; the INL thickness in group III recorded 60% count in group II; while the decrease in the subsequent of its initial thickness in the adult group and fell down groups was found to be non-significant. Fernàndez- to about one third in group IV. These findings were -Sánchez et al. [16] stated that in rodents, age causes in accordance with Aggarwal et al. [1] who reported morphologic changes and/or loss of retinal ganglion that aging causes loss of retinal neurons, including cells. Several works reported retinal ganglion cells or rod photoreceptors and rod bipolar cells. axonal loss [7]; some reported degeneration but no Ultrastructural changes of the INL cells, in the loss [37]. Yet, other authors perceived no loss in the present work were in the form of cytoplasmic vacu- GCL in aged animals [26]. Although some works have olations which increased in severity with aging. Vacu- described that the retina thins with age, Feng et al. [15]

255 Folia Morphol., 2019, Vol. 78, No. 2

found that this thinning is not accompanied by Glial fibrillary acidic protein immunostaining in a reduction of the retinal volume. In the present study, our study, were dramatically increased in the MG. numerous dilated blood capillaries were encountered Astrocytes at the nerve fibre layer together with in the GCL and the two plexiform layers. These blood Müller cells processes presented strong positive re- capillaries became more dilated and more congested action in the aged retinae. With aging, the reaction in the old and senile groups. Nadal-Nicolás et al. [26] became stronger and more numerous and thicker. clarified that the compensatory changes occurring in Müller cells processes appeared traversing the IPL the capillaries have a negative impact on neuronal extending between ILM and OLM. Martinez-De Luna function. A decrease in capillary stability and new ves- et al. [25] elucidated that the MG play a role in the sels formation, culminate in vessel loss in response to extensive structural changes resulting in Müller cell ‘physiological hypoxia’ and likely further reduce the hypertrophy and glial scar formation. They also ex- ability of the microvasculature to meet the metabolic plained those finding as mammalian MG respond to needs of the parenchyma. retinal injury by age changes summarised as reactive The present work revealed that retinal capillaries of gliosis. Müller cell gliosis is a complex response that different layers of the retina displayed major thickening involves changes in cell physiology, gene expres- and lamination of the basement membrane with age. sion and morphology. A hallmark of gliosis is the This seemed to cause outward displacement of the upregulation of intermediate filament proteins in pericyte somas. Pericytes processes were embedded glial cells, including GFAP [12]. Dysfunction of glial within the thick basement membrane. The same results cells in different pathologies of the retina has been were obtained by Donato et al. [13] who suggested that linked to retinal swelling and blood-retinal barrier the observed age-dependent broadening of peripheral breakdown [21]. retinal capillaries may reflect the thickening of the capil- The present study provides a baseline to better lary basement membrane and/or an increase in capillary understanding of the implication of aging in rat luminal diameter similar to that demonstrated in regions models of human pathologies that are closely re- of the aging brain. In healthy humans, increased pro- lated to senescence. Elucidating the morphological inflammatory cytokines have been observed in the aging changes in retinal aging would facilitate the design vascular endothelial cells [13]. of new strategies to prevent visual dysfunction as- The present study showed that the retinal capillar- sociated with both the normal aging process and ies exhibited thickening of the internal and external age-related retinal pathologies. basement membranes. Pericytes presented cytoplasmic inclusions and long cytoplasmic processes that CONCLUSIONS extended and embedded in the thickened basement It could be concluded that rat retinae clearly un- membrane. These changes are found to increase pro- dergo age-related morphological changes on both gressively with age. In a study performed by Bianchi light microscopic and ultrastructural levels and as- et al. [3] on human eyes of elderly patients, similar similated with immunohistochemistry and statistical results were obtained with highlighting the exter- analysis. Such changes provide a cellular base for nal basal lamina thickening. Their findings assigned explanation of decreased vision in humans with a primary role to endothelial dysfunction as a cause of aging other than reflection errors. basement membrane thickening, while retinal altera- All retinal layers were affected by aging with tions were considered to be a secondary cause of either variable degrees. The changes in the retinal pigment ischaemia or exudation. In addition, the present work epithelium and photoreceptors were prominent. found that the endothelial cells lining the blood capil- Some changes started as early as the middle-aged laries were severely affected and displayed progressive group, while others appeared in the aged groups. degeneration with age. Marked proliferation of the Moreover, the alterations presented in the aged capillaries endothelial basal laminae was encountered. group were progressed and aggravated in the senile This was explained by Bianchi et al. [3] who illustrated group. Effect of aging was not only qualitative, but that endothelial tight junction integrity was disrupted also quantitative as proven by the statistical analysis. by various inflammatory cytokines and leukocyte–en- This was applied on the retinal neuronal synapses dothelial interaction. as well as somas.

256 M.E.I. Mohamed et al., Aging of retina

RECOMMENDATIONS 11. Dan C, Jian-Bin T, Hui W, et al. Synaptophysin expression in rat retina following acute high intraocular pressure. Acta It is recommended to do further studies with link- Histochem Cytochem. 2008; 41(6): 173–178, doi: 10.1267/ ing the morphological findings of this study to the ahc.08034, indexed in Pubmed: 19180202. functional alterations that occur with normal aging. It 12. de Pablo Y, Nilsson M, Pekna M, et al. Intermediate is also recommended to investigate genetic factors and filaments are important for astrocyte response to oxidative stress induced by oxygen-glucose deprivation concomitant ocular diseases that aggravate the histo- and reperfusion. Histochem Cell Biol. 2013; 140(1): logical alterations occurring during the aging process. 81–91, doi: 10.1007/s00418-013-1110-0, indexed in The results of this study are hopefully to be taken Pubmed: 23756782. in consideration to build strategies to decelerate or 13. Donato AJ, Black AD, Jablonski KL, et al. Aging is associ- reverse age-associated deterioration in ocular func- ated with greater nuclear NF kappa B, reduced I kappa B alpha, and increased expression of proinflammatory tion to avoid the psychosocial consequences of visual cytokines in vascular endothelial cells of healthy humans. impairment in elderly. Aging Cell. 2008; 7(6): 805–812, doi: 10.1111/j.1474- 9726.2008.00438.x, indexed in Pubmed: 18782346. References 14. Dorfman AL, Cuenca N, Pinilla I, et al. Immunohisto- 1. Aggarwal P, Nag TC, Wadhwa S. Age-related decrease in chemical evidence of synaptic retraction, cytoarchitectural rod bipolar cell density of the human retina: an immu- remodeling, and cell death in the inner retina of the rat nohistochemical study. J Biosci. 2007; 32(2): 293–298, model of oygen-induced retinopathy (OIR). Invest Oph- indexed in Pubmed: 17435321. thalmol Vis Sci. 2011; 52(3): 1693–1708, doi: 10.1167/ 2. Bancroft JD, Gamble M. Immunohistochemical technique. iovs.10-6197, indexed in Pubmed: 21071736. In: Theory and practice of histological techniques, 6th ed. 15. Feng L, Sun Z, Han H, et al. No age-related cell loss in three Churchill Livingstone, London 2008: 433–472. retinal nuclear layers of the Long-Evans rat. Vis Neurosci. 3. Bianchi E, Ripandelli G, Taurone S, et al. Age and 2007; 24(6): 799–803, doi: 10.1017/S0952523807070721, diabetes related changes of the retinal capillaries: indexed in Pubmed: 18093367. An ultrastructural and immunohistochemical study. 16. Fernández-Sánchez L, de Sevilla Müller LP, Brecha NC, et Int J Immunopathol Pharmacol. 2016; 29(1): 40–53, al. Loss of outer retinal neurons and circuitry alterations doi: 10.1177/0394632015615592, indexed in Pub- in the DBA/2J mouse. Invest Ophthalmol Vis Sci. 2014; med: 26604209. 55(9): 6059–6072, doi: 10.1167/iovs.14-14421, indexed 4. Bonilha VL. Age and disease-related structural changes in in Pubmed: 25118265. the retinal pigment epithelium. Clin Ophthalmol. 2008; 17. Hippert C, Graca AB, Barber AC, et al. Müller glia acti- 2(2): 413–424, indexed in Pubmed: 19668732. vation in response to inherited retinal degeneration is 5. Boulton ME. Studying melanin and lipofuscin in highly varied and disease-specific. PLoS One. 2015; 10(3): RPE cell culture models. Exp Eye Res. 2014; 126: e0120415, doi: 10.1371/journal.pone.0120415, indexed 61–67, doi: 10.1016/j.exer.2014.01.016, indexed in Pub- in Pubmed: 25793273. med: 25152361. 18. Joly S, Pernet V, Samardzija M, et al. Pax6-positive Müller 6. Boya P, Esteban-Martínez L, Serrano-Puebla A, et al. Au- glia cells express cell cycle markers but do not proliferate tophagy in the eye: Development, degeneration, and ag- after photoreceptor injury in the mouse retina. Glia. 2011; ing. Prog Retin Eye Res. 2016; 55: 206–245, doi: 10.1016/j. 59(7): 1033–1046, doi: 10.1002/glia.21174, indexed in preteyeres.2016.08.001, indexed in Pubmed: 27566190. Pubmed: 21500284. 7. Calkins DJ. Age-related changes in the visual pathways: 19. Karthaus M, Falkenstein M. Functional Changes and blame it on the axon. Invest Ophthalmol Vis Sci. 2013; Driving Performance in Older Drivers: Assessment and 54(14): ORSF37–ORSF41, doi: 10.1167/iovs.13-12784, Interventions. Geriatrics. 2016; 1(2): 12, doi: 10.3390/ indexed in Pubmed: 24335066. geriatrics1020012. 8. Cattoretti G, Pileri S, Parravicini C, et al. Antigen un- 20. Keeling E, Lotery AJ, Tumbarello DA, et al. Impaired masking on formalin-fixed, paraffin-embedded tissue cargo clearance in the retinal pigment epithelium sections. J Pathol. 1993; 171(2): 83–98, doi: 10.1002/ (RPE) underlies irreversible blinding diseases. Cells. path.1711710205, indexed in Pubmed: 7506771. 2018; 7(2), doi: 10.3390/cells7020016, indexed in Pub- 9. Chernoivanenko IS, Matveeva EA, Gelfand VI, et al. med: 29473871. Mitochondrial membrane potential is regulated by 21. Klaassen I, Van Noorden CJF, Schlingemann RO. Molecular vimentin intermediate filaments. FASEB J. 2015; 29(3): basis of the inner blood-retinal barrier and its breakdown 820–827, doi: 10.1096/fj.14-259903, indexed in Pub- in diabetic macular edema and other pathological condi- med: 25404709. tions. Prog Retin Eye Res. 2013; 34: 19–48, doi: 10.1016/j. 10. Chirco KR, Sohn EH, Stone EM, et al. Structural and preteyeres.2013.02.001, indexed in Pubmed: 23416119. molecular changes in the aging choroid: implications 22. Litts KM, Messinger JD, Freund KB, et al. Inner segment for age-related macular degeneration. Eye. 2017; 31(1): remodeling and mitochondrial translocation in cone 10–25, doi: 10.1038/eye.2016.216, indexed in Pub- photoreceptors in age-related macular degeneration with med: 27716746. outer retinal tubulation. Invest Ophthalmol Vis Sci. 2015;

257 Folia Morphol., 2019, Vol. 78, No. 2

56(4): 2243–2253, doi: 10.1167/iovs.14-15838, indexed stasis. Aging Cell. 2013; 12(3): 478–488, doi: 10.1111/ in Pubmed: 25758815. acel.12072, indexed in Pubmed: 23521856. 23. Luke MPS, LeVatte TL, O’Reilly AM, et al. Effect of 31. Samuel MA, Zhang Y, Meister M, et al. Age-related NCAM on aged-related deterioration in vision. Neuro- alterations in neurons of the mouse retina. J Neurosci. biol Aging. 2016; 41: 93–106, doi: 10.1016/j.neurobio- 2011; 31(44): 16033–16044, doi: 10.1523/JNEURO- laging.2016.02.003, indexed in Pubmed: 27103522. SCI.3580-11.2011, indexed in Pubmed: 22049445. 24. Ma W, Coon S, Zhao L, et al. A2E accumulation influ- 32. Sanes JR, Zipursky SL. Design principles of insect and vertebrate ences retinal microglial activation and complement visual systems. Neuron. 2010; 66(1): 15–36, doi: 10.1016/j. regulation. Neurobiol Aging. 2013; 34(3): 943–960, neuron.2010.01.018, indexed in Pubmed: 20399726. doi: 10.1016/j.neurobiolaging.2012.06.010, indexed in 33. Sengupta P. The laboratory rat: relating its age with hu- Pubmed: 22819137. man’s. Int J Prev Med. 2013; 4(6): 624–630, indexed in 25. Martinez-De Luna RI, Ku RY, Aruck AM, et al. Müller glia Pubmed: 23930179. reactivity follows retinal injury despite the absence of the 34. Szabadfi K, Estrada C, Fernandez-Villalba E, et al. Retinal glial fibrillary acidic protein gene in Xenopus. Dev Biol. aging in the diurnal Chilean rodent (Octodon degus): 2017; 426(2): 219–235, doi: 10.1016/j.ydbio.2016.03.005, histological, ultrastructural and neurochemical alterations indexed in Pubmed: 26996101. of the vertical information processing pathway. Front Cell 26. Nadal-Nicolás FM, Sobrado-Calvo P, Jiménez-López M, et Neurosc. 2015; 9, doi: 10.3389/fncel.2015.00126. al. Long-Term Effect of Optic Nerve Axotomy on the Reti- 35. Warburton S, Davis WE, Southwick K, et al. Proteomic and nal Ganglion Cell Layer. Invest Ophthalmol Vis Sci. 2015; phototoxic characterization of melanolipofuscin: correla- 56(10): 6095–6112, doi: 10.1167/iovs.15-17195, indexed tion to disease and model for its origin. Mol Vis. 2007; 13: in Pubmed: 26393669. 318–329, indexed in Pubmed: 17392682. 27. Nadal-Nicolás FM, Vidal-Sanz M, Agudo-Barriuso M. 36. Yao J, Jia L, Shelby SJ, et al. Circadian and noncircadian The aging rat retina: from function to anatomy. Neu- modulation of autophagy in photoreceptors and retinal robiol Aging. 2018; 61: 146–168, doi: 10.1016/j.neuro- pigment epithelium. Invest Ophthalmol Vis Sci. 2014; biolaging.2017.09.021, indexed in Pubmed: 29080498. 55(5): 3237–3246, doi: 10.1167/iovs.13-13336, indexed 28. Nag T, Wadhwa S. Ultrastructure of the human retina in Pubmed: 24781939. in aging and various pathological states. Micron. 2012; 37. Yassa HD. Age-related changes in the optic nerve of 43(7): 759–781, doi: 10.1016/j.micron.2012.01.011. Sprague-Dawley rats: an ultrastructural and immuno- 29. Nieves-Moreno M, Martínez-de-la-Casa JM, Morales- histochemical study. Acta Histochem. 2014; 116(6): Fernández L, et al. Impacts of age and sex on retinal layer 1085–1095, doi: 10.1016/j.acthis.2014.05.001, indexed thicknesses measured by spectral domain optical coher- in Pubmed: 24958340. ence tomography with Spectralis. PLoS One. 2018; 13(3): 38. Zhao JJ, Ouyang H, Luo J, et al. Induction of retinal e0194169, doi: 10.1371/journal.pone.0194169, indexed progenitors and neurons from mammalian Müller glia in Pubmed: 29522565. under defined conditions. J Biol Chem. 2014; 289(17): 30. Rodríguez-Muela N, Koga H, García-Ledo L, et al. Balance 11945–11951, doi: 10.1074/jbc.M113.532671, indexed between autophagic pathways preserves retinal homeo- in Pubmed: 24523410.

Anatomical variations and dimensions of arteries in the anterior part of the circle of Willis J. Shatri1, S. Cerkezi2, V. Ademi2, V. Reci2, S. Bexheti1

1Institute of Anatomy, Faculty of Medicine, University of Prishtina, Kosovo 2Institute of Anatomy, Faculty of Medicine, University of Tetovo, Macedonia

[Received: 3 September 2018; Accepted: 23 September 2018]

Background: The aim of this study was to investigate different anatomic variations and dimensions of anterior part of the circle of Willis (CW) and their prevalence on Kosovo’s population. Materials and methods: This is an observational descriptive and retrospective study performed at the University Clinical Centre, Clinic of Radiology. The three- -dimensional time-of-flight technique was used for magnetic resonance angiography imaging to evaluate the anatomy of the CW in 513 adults without clinical manifestations for cerebrovascular disease. The diameters of arteries of the CW were measured and variations were recorded. Results: The complete anterior part of CW was found in 64.3% of cases, more evident in female at about 66% than male 62.2%. Morphologic variations of the ACA1 are agenesis or hypoplasia in 5.65%, fusion of the anterior cerebral arteries (ACAs) on a short distance in 6.5% and fusion of the ACA on a long distance in 2.5%, median ACA is presented in 11 examined patients or 2.1%. Normal anterior communicating artery (ACoA) was seen in 68.2%, hypoplasia or absence 15.66%, double ACoA in 0.6% and fenestrations in 3.89%. The mean calibrations of the vessels were measured as 2.04 mm in right A1, 2.06 mm in left A1 and 1.16 mm of ACoA. While, 14.1 mm length of right A1, 13.87 mm of left A1 and 2.99 mm of ACoA. Conclusions: Knowledge of the variations and diameter of the anterior part of the CW has a great importance in determination of anatomical variations and diameter in general populations, also, in interventional radiology for various endovascular interventions as well as during anatomy lessons. (Folia Morphol 2019; 78, 2: 259–266)

Key words: variations of anterior part, dimensions of anterior cerebral artery, dimensions of anterior communicating artery, circle of Willis, anterior part of circle of Willis

INTRODUCTION ies (ICAs), posterior cerebral arteries (PCAs), and pos- The circle of Willis (CW) is an anastomotic polygon terior communicating arteries (PCoAs). ACA divides at the base of the brain that supplies blood to the into two parts; A1 and A2. The A1 segment is also brain structures. It consists of the following arteries known as the pre-communicating part of the ACA, including: anterior cerebral arteries (ACAs), anterior while, A2 segment is post-communicating part of the communicating artery (ACoA), internal carotid arter- ACA. The ACoA is a small blood vessel which bridges

Address for correspondence: Dr. S. Bexheti, Institute of Anatomy, Faculty of Medicine, University of Prishtina, 10000 Prishtina, Kosovo, e-mail: [email protected]

259 Folia Morphol., 2019, Vol. 78, No. 2

the two larger ACAs. A majority of anatomic varia- head with 40 mm saturation at the head end, field of tions have been observed in the ACoA section, char- view 180 × 158 and 256 matrix size. All MRA were acterised as anatomic complexity and diversity. The evaluated by using native source images and maxi- ACoA, A1 and A2 segments of ACA, and the recurrent mum intensity projections (MIP) images. These axial artery of Heubner, combined with their perforators source images were post processed by the MIP algo- and other branches, are often referred as the ACoA rithm to produce 8 projections rotating about the complex. There are very few case reports regarding section axis and 1 axial image (projection images). the variations encountered in the ACA and ACoA. All component vessels of the CW were accessed by Knowledge of these variations is important for brain measuring the diameter on the individual MIP im- surgeons considering intervention to the intracranial ages. Whenever there was a doubt in determining arteries and for interventional radiologists [6]. Vari- the diameter of one vessel due to overlapping vessels ous authors have studied the length, diameter and in the MIP images, the 3D time-of-flight source im- anomalies in the origin of these arteries and found ages are then reviewed on the advanced workstation that the parameters in different geographical limits (GE ADW 4.0 workstation). Occasionally, it was nec- were different. The most common morphological essary to cut off the unwanted branching vessel on changes in the blood vessels of the brain are encoun- the images by an experienced radiologist to better tered in their origins, calibre; often they are hypoplas- depict the target vessels and assess correct diameter. tic, duplicated or event absent, number, communi- So, each patient was positioned in the supine posi- cation, and branching. So, presence of anatomical tion, the head was immobilised by the coil for head; variations means deviation from the normal pattern noise protectors were placed in the ears of patients without any functional impairment to the individuals. to reduce noise. During the examination, patients’ Vascular variation has been examined using various vital signs were kept under control and monitored methods including autopsy, computed tomography all the time through the monitor. Vessels that were angiography and magnetic resonance angiography visualised as continuous segments of at least 0.8 mm (MRA). However, earlier studies are mainly based on in diameter were considered present; arteries when autopsy study, with numerous limitations on the seen as non-continuous segments those smaller than connections between morphology and physiological 0.8 mm in diameter were considered hypoplastic or changes in the haemodynamic system. Moreover, the absent. Arteries when seen as non-continuous seg- sample number was limited and their results were ments were considered as absent [12]. not comprehensive and did not reflect the normal The A1 segment of the ACA and ACoA, which physiological status. With the development of imag- are components of the CW, was studied for its length ing diagnostic methods, such as MRA, have been and diameters (Fig. 1). The following two points were progressed to study morphology of blood vessels. marked separately for ACA: at its origin from the ICA The present study was undertaken to study the origin, (taken as point ‘A’), at the proximal part of its junction course and termination of the ACA, to observe the with the ACoA (taken as point ‘B’). Then, the lengths of ACoA and to observe variations between the two the A1 segment of the ACA on the right and left sides sides. CW variations have been examined in various were measured between the two points (point ‘A’ and populations and age groups [3, 11, 12]. ‘B’). In cases with the tortuous course of the A1, with an- gulations, the length was measured in segments, which MATERIALS AND METHODS were then added and the total length was calculated. Five hundred and thirteen patients, without clini- The projections which showed the anterior part of the cal manifestations of cerebrovascular diseases, have CW with minimum overlapping were selected for taking been included in the study and are considered as measurements. The cases in which points ‘A’ and ‘B’ at healthy subjects. All patients (222 men and 291 which the measurements were to be taken were not women; mean age: 46 years) underwent the three- clear, either because of overlapping in the projections, -dimensional (3D) time of flight MRA of the CW, or because of any other reasons, were not included. using a 1.5T scanner (Avanti, Siemens, Germany). Also, the cases in which any of the projections of the A1 Following imaging parameters were used: repetition segment could not make the entire course clear were time/echo time 23/7.0, flip angle 25 degrees, slice not included in the study. The length of the ACoA was thickness 0.7 mm, number of slice 44/slab, number taken between two points of communication between of slabs 4, slice overlap 25%, flow direction feet to right A1 segments and the left A1 segment. 260 J. Shatri et al., Anatomical variations and dimensions of arteries in the anterior part of the circle of Willis

cance are shown in Table 2. The differences of measured values between individuals younger and older than 40 years are in Table 3, and the differences between results of the measurements in the male and female persons, and their statistical significances are shown in Table 4. Comparing the right and the left A1 segments, the average length was found to be larger on the right side than on the left side. The difference of the length on the two sides was not significant (p > 0.05). The average diameter of the ACA was larger on the left side than on right (p > 0.05; Table 2). In the individuals younger than 40 years, the average diameters of ACA and ACoA were larger than in older but not significantly, while the mean lengths Figure 1. Standardised method for vessel diameter measurement were longer in older than in younger persons, also with transversal cuts 5 mm from their origin and length of the not significantly (Table 3). arteries, axial maximum intensity projections reconstruction As show in Table 4, the mean diameter of ACA in three-dimensional time-of-flight magnetic resonance angiography. the female was larger than in male, but not significantly, while the average diameter of ACoA tended to be larger in female (p < 0.05). There were no notable differences among the ACA and ACoA between male and female (p > 0.05).

Variations of the anterior part of the CW Morphologic variations of the A1 segment of ACA are agenesis or hypoplasia, fenestration, fusion of the ACAs on a short distance and fusion of the ACA on Figure 2. Schematic representation of arteries in anterior part of a long distance. Normal structure, when presented circle of Willis. a single ACoA and ICA which bifurcates into the A1 segment of the ACA and the middle cerebral artery, was found in 64.3% of cases, more in female 66% RESULTS than male 62.2%, shown in Figure 3. Unilaterally The present study focuses on the distribution of di- agenesis of A1 was found in 29 (5.65%) cases, more ameter and length among the subjects studied including in male than female (Fig. 4). Fusion of the ACAs for age and gender. Most of the patients were female 291 a short distance was presented in 6.5%, while for (56.72%) while 222 (43.27%) male. About 61% of the a long distance in 2.55% (Figs. 4, 5). patients were ≥ 40 years age. The patients were ranged Median ACA is presented in 11 examined pain ages from 11 to 83; the mean age was 46 years. tients or 2.1%, in women 2.4% and in men 1.8%. The inner diameters and lengths of the ACAs and Normal ACoA was seen in 68.2% (350/513) of the ACoAs are shown in Table 1 and Figure 2. The differences cases, presented in Figure 3. Hypoplasia or absence of between right and left sides and their statistical signifi- an anterior communication was seen in 15.66%; while

Table 1. Diameter and length (mm) of A1 of anterior cerebral artery (ACA) and anterior communicating artery (ACoA) Artery Diameter [mm] Length [mm] AM SD SE Min. Max. AM SD SE Min. Max. ACA A1 2.05 0.27 0.01 1 3.8 14.01 1.4 0.08 8.9 21 ACoA 1.16 0.17 0.01 0.9 2.1 2.99 0.62 0.05 1 5.6

AM — arithmetic mean; Max. — maximum value; Min. — minimum value; SD — standard deviation; SE — standard error

261 Folia Morphol., 2019, Vol. 78, No. 2

Table 2. The average diameters and lenghts (mm) of the right and left A1 segments, and statistical significances between left and right sides values A1 segment [mm] Right Left T-test P-value SE AM SD Min. Max. AM SD Min. Max. Diameter 2.04 0.28 1 3.8 2.06 0.26 1 2.93 1.14 0.2 0.17 Length 14.1 1.51 8.9 20.4 13.87 1.3 9.6 21 0.11 0.9 1.9

AM — arithmetic mean; Max. — maximum value; Min. — minimum value; P-value: p > 0.05 (not significant), p < 0.05 (significant), p < 0.01 (highly significant); SD — standard deviation; SE — standard error; T-test — Student’s t-test

Table 3. The average diameters and lengths (mm) of the A1 segment of anterior cerebral artery (ACA) and anterior communicating artery (ACoA) with statistical significance between two groups of different geda persons Arteries [mm] > 40 year < 40 year T-test P-value SE AM SD Min. Max. AM SD Min. Max. ACA A1 (d) 2.05 0.2 1.1 2.8 2.06 0.24 1 2.93 0.5 0.5 0.01 ACoA (d) 1.16 0.15 0.89 2.1 1.17 0.18 0.9 1.7 1.08 0.2 0.009 ACA A1 (l) 14 1.29 8.9 21 13.9 1.22 10 19.7 0.8 0.3 0.1 ACoA (l) 3.2 0.6 1 5 3.15 0.6 1.8 5.19 0.9 0.3 0.05

AM — arithmetic mean; d — diameter; l — length; Max. — maximum value; Min. — minimum value; P-value: p > 0.05 (not significant), p < 0.05 (significant), p < 0.01 (highly significant); SD — standard deviation; SE — standard error; T-test — Student’s t-test

Table 4. The average diameters and lenghts (mm) of the of the A1 segment of anterior cerebral artery (ACA) and anterior communicating artery (ACoA) with statistical significance between two groups of patients ithw different gender Arteries [mm] Male (n = 222) Female (n = 291) T-test P-value SE AM SD Min. Max. AM SD Min. Max. ACA A1 (d) 2.04 0.23 1 2.7 2.06 0.32 1 3.8 0.97 0.33 0.02 ACoA (d) 1.15 0.14 0.9 2.1 1.18 0.2 0.89 2.1 1.9 0.05 0.01 ACA A1 (l) 14.1 1.1 10 21 13.9 1.6 8.9 13.8 1.6 0.1 0.1 ACoA (l) 3.3 0.63 2 5.1 3.31 0.27 2.5 4 0.2 0.8 0.04

double ACoA was present in 3 (0.6%) cases (Fig. 6). of intracranial arteries. They are typically located at the Fenestrations of ACoA existed in 3.89 (20/513). arterial branching points near the skull base (Fig. 7A, B). Morphologic variations of ACoA were present in Fenestration of the anterior cerebral artery has 35.5% of cases while anterior part of CW has a com- been known in the anatomical field but it has rarely plete structure in 77.7%, meaning that the arteries been reported in the radiological literature. The pur- constituting the front of the CW have origin, diameter pose of this paper is to present cases with fenestration and regular extension. of the ACA; in our study it was present in 3 (0.58%) Our results show that the more completed an- cases, shown in Figure 7C. terior part of CW is presented in female (78.57%), comparing to male (76%). A single ACoA is the most DISCUSSION common type (64.5%) and hypoplasia or absence of The CW is the main source of supply to the brain ACoA (15.6%), in all age and both sex groups. and many authors have studied the diameter and In 9 specimens or 1.75% an aneurysm of the proximal length of the blood vessels that form the CW in segment of the ACA and the ACoA was observed on its cadaver-brain. Thus, Ghazali et al. [4] tested digital right side. Intracranial aneurysms are acquired dilatations subtraction angiography with 3D time-of-flight MRA

262 J. Shatri et al., Anatomical variations and dimensions of arteries in the anterior part of the circle of Willis

A B C

Figure 3. A–C. Image by three-dimensional time-of-flight magnetic resonance angiography — a single anterior communicating artery. The internal carotid artery bifurcates into the pre-communicating segment of the anterior cerebral artery and the middle cerebral artery.

A B C D

Figure 4. A–D. Image by three-dimensional time-of-flight magnetic resonance angiography — one pre-communicating segment of an anterior cerebral artery is hypoplastic or absent, the other pre-communicating segment gives rise to both post-communicating segments of the anterior cerebral arteries.

A B C D

Figure 5. Image by three-dimensional time-of-flight magnetic resonance angiography; A, B. Fusion of the anterior cerebral arteries occurs for a short distance; C, D. Anterior cerebral artery forms a common trunk and split distally into two post-communicating segments.

in the comparison of the morphology of CW arteries. use of contrast substance, it does not use ionising With the exception of PCoA, MRA was found to be radiation, the processing of data is easy and there is a more sensitive method in the evaluation of the other not the visualisation of pure bone artefacts as seen CW arteries. In this respect, MRA seems to be more on computed tomography in the posterior fosse and advantageous than CTA, as it does not require the particularly in the base of the skull.

263 Folia Morphol., 2019, Vol. 78, No. 2

A B C

Figure 6. Image by three-dimensional time-of-flight magnetic resonance angiography;A, B. Hypoplasia or absence of an anterior communication; C. Two or more anterior communicating artery.

A B C

Figure 7. Image of three-dimensional time-of-flight magnetic resonance angiography; A, B. Aneurysm on right side of anterior cerebral artery; C. Fenestration of anterior cerebral artery.

The proximal segment of anterior cerebral artery (A1) al. [13]. In our study the mean diameter of ACoA was supplies the basal surface of the cerebral hemisphere. 1.16 mm and length 2.99 mm compared with the stud- The diameter of A1 according to Kamath et al. [10], ies in other people we noticed no greater differences 2.3 mm for diameter and 14.25 mm length, Iqbal et al. [8] in the values obtained in Kosovo. The largest diameter diameter 2.3 mm and length 12.4 mm, on Turkish popu- is observed in individuals younger than 40 years, but lation length of A1 is 14.4 mm [11] while diameter without significant difference, while, the length in older 1.58 mm in the right A1, 1.64 mm in the left A1 [18], in persons was larger, but with no significance. Kosovo’s population the diameters of A1 was 2.09 mm The absence of ACA-A1 of one side is present in while length 13.96 mm [16]. In our study the diameter 5.6% or 29 examined cases, almost proportionally in of ACA was 2.05 and length 14.01 mm. There were women 6.75% and in male 4.8% comparing to 4% no significant differences in diameter and length of [7], 4.76% [9] and 5% [2]. Median ACA is presented A1 segments based on gender, age or side. The ACoA in 11 examined patients or 2.1%, in women 2.4% represents an important anastomosis between the left and in men 1.8%, similar results are 2.5% and 5% and right ACA. Also, demarcates the junction between If the artery on one side is narrowed, the vascular the A1 and A2 segment. ACoA diameter ranged from insufficiency is compensated by crossing over by op- 0.8 to 2.3 mm and from 1.9 to 2.5 mm by Kamath et posite side artery, or by giving branches that cross al. [10], 2.1 mm and 2.5 mm in by Krishnamurthy et over to the other side. It indicates that the CW offers 264 J. Shatri et al., Anatomical variations and dimensions of arteries in the anterior part of the circle of Willis

a potential shunt in abnormal conditions such as are found in about 2% of the general adult popula- occlusions and spasms. In normal circumstances it is tion, and are considered to be acquired lesions. not an equalizer and distributer of blood from differ- The A1 fenestration was found in: 0.1% of cases, ent sources [15]. Cerebral infarct due to occlusion of respectively, 0.58% was seen in the distal half of the ACA is common in stroke and has grave morbidity. A1 segment of ACA. Therefore variation in anterior circulation of the brain Also, anatomic parameters of the anterior cerebral is of great importance particularly in the surgery of artery can be used to plan and design equipment the region. Some authors associate ACA-A1 deficiency such as angiographic micro catheter and guide used with the presence of aneurysm in ACoA. The absence in endovascular procedures [17]. of proximal segment of ACA-A1 is a common finding in ACoA aneurysm patients and can be considered as CONCLUSIONS a risk factor for the formation of aneurysm. The morphological variations and diameter differ- Data’s from this study about ACA dimensions and ences demonstrated in our study providing an impor- variations can provide accurate micro anatomic infor- tant reference value for the 3D time-of-flight MRA. Our mation for surgical treatment of aneurysms or dur- findings show that the configuration of the CW may ing surgical procedures about reconstructing blood vary largely in general population. The prevalence of vessels in circulus arteriosus cerebri. Also, anatomic complete configuration of the circle is higher in younger parameters of the ACA can be used to plan and design individuals as well as in female. Some of these variations equipment such as angiographic micro catheter and may be associated with certain risks like aneurismal guide used in endovascular procedures. The vascular development or watershed infractions. anatomy of the region of ACoA is generally complex Knowledge of the variations and diameter of the due to its development. In the 24 mm embryo the anterior part of the CW has a great importance in ACoA is still a plexiform structure connecting both determination of anatomical variations and diameter ACA. Incomplete fusion of this anastomosis may lead in general populations especially in Kosovo’s popu- to fenestration, doubling or tripling of the ACoA [15]. lation, also, in interventional radiology for various Fenestration can be a protective mechanism for ste- endovascular interventions and anatomy teaching. nosis, injury to, or occlusion of one of the duplicated limbs. Such anomalies may become important in the References planning of interventional procedures [5]. Fenestra- 1. Chen HW, et al. Magnetic resonance angiographic evaluation can pose unexpected difficulties in the surgical tion of circle of Willis in general population: a morphologic study in 507 cases. Chinese J Radiol. 2004. treatment, as unilateral or bilateral fenestration often 2. De Silva KR, Silva R, Gunasekera WSL, et al. Prevalence accompanies aneurysms or other anomalies. The most of typical circle of Willis and the variation in the an- common variant is when ACoA is absent and ap- terior communicating artery: A study of a Sri Lankan pears in 80 patients or 15.6%, most common among population. Ann Indian Acad Neurol. 2009; 12(3): 157–161, doi: 10.4103/0972-2327.56314, indexed in women 9.75% than men 5.45%. The frequency of lack Pubmed: 20174495. of ACoA is related to other studies such Jalali Kondori 3. El-Barhoun EN, Gledhill SR, Pitman AG. Circle of Wil- et al. [9], Chen et al. [1], respectively 11.6, 14.28 and lis artery diameters on MR angiography: an Australian 11.44. Also in cases when ACoA is missing both ACAs reference database. J Med Imaging Radiat Oncol. 2009; join in a trunk or run as united for a certain length. 53(3): 248–260, doi: 10.1111/j.1754-9485.2009.02056.x, indexed in Pubmed: 19624291. The first variant was present in 6.8% of patients, in 4. Ghazali RM, Shuaib IL. Comparison Between 3D TOF equal proportion between females and males, while Magnetic Resonance Angiography and Intraarterial Digital the next is less common and appears in 2.55% of Subtraction Angiography in Imaging the Circle of Willis. examined cases, 2.4% in female and 1.8% in male. Malays. J. Med. Sci. 2003; 10(1): 37–42. 5. Goddard AJ, Annesley-Williams D, Guthrie JA, et al. Du- Our results are similar with findings of Jalali Kondori plication of the vertebral artery: report of two cases and et al. [9] 5.7% and 0.95%, Chen et al. [1] 5.92% with review of the literature. Neuroradiology. 2001; 43(6): 1.38%. Duplication of ACoA occurred in 0.6%, or 477–480, indexed in Pubmed: 11465760. 3 patients, 0.3% in female and 0.9% in men, com- 6. Hoksbergen AWJ, Majoie CBL, Hulsmans FJH, et al. As- sessment of the collateral function of the circle of Willis: paring with similar results of 0.95% [9] or 0.6% [14]. Three-dimensional time-of-flight MR angiography com- When an intracranial aneurysm ruptures, it causes pared with transcranial color-coded duplex sonography. subarachnoid haemorrhage. Intracranial aneurysms Am J Neuroradiol. 2003; 24: 456–462.

265 Folia Morphol., 2019, Vol. 78, No. 2

7. Iqbal S. A comprehensive study of the anatomical cal importance. Clin Ter. 2010; 161(3): 231–234, indexed in variations of the circle of willis in adult human brains. Pubmed: 20589352. J Clin Diagn Res. 2013; 7(11): 2423–2427, doi: 10.7860/ 14. Naveen SR, Bhat V, Karthik GA. Magnetic resonance angio- JCDR/2013/6580.3563, indexed in Pubmed: 24392362. graphic evaluation of circle of Willis: A morphologic study 8. Iqbal S. Average dimensions of the vessels at the base of in a tertiary hospital set up. Ann Indian Acad Neurol. 2015; the brain and the embryological basis of its variations. 18(4): 391–397, doi: 10.4103/0972-2327.165453, indexed National J Clin Anat. 2013; 2: 180–189. in Pubmed: 26713008. 9. Jalali Kondori B, Azemati F, Dadseresht S. Magnetic reso- 15. Puchades-Orts A, Nombela-Gomez M, Ortuño-Pacheco nance angiographic study of anatomic variations of the G. Variation in form of circle of Willis: some anatomi- circle of Willis in a population in Tehran. Arch Iran Med. cal and embryological considerations. Anat Rec. 1976; 2017; 20(4): 235–239, doi: 0172004/AIM.009, indexed in 185(1): 119–123, doi: 10.1002/ar.1091850112, indexed Pubmed: 28412828. in Pubmed: 1267194. 10. Kamath S. Observations on the length and diameter of 16. Shatri J, Bexheti D, Bexheti S, et al. Influence of gender vessels forming the circle of Willis. J Anat. 1981; 133(Pt 3): and age on average dimensions of arteries forming the 419–423, indexed in Pubmed: 7328048. circle of Willis study by magnetic resonance angiography 11. Karatas A, Yilmaz H, Coban G, et al. The anatomy of circulus on kosovo’s population. Open Access Maced J Med Sci. arteriosus cerebri (circle of Willis): a study in Turkish popula- 2017; 5(6): 714–719, doi: 10.3889/oamjms.2017.160, tion. Turk Neurosurg. 2016; 26(1): 54–61, doi: 10.5137/1019- indexed in Pubmed: 29104678. 5149.JTN.13281-14.1, indexed in Pubmed: 26768869. 17. Stefani M, Schneider F, Marrone A, et al. Anatomic variations of 12. Krabbe-Hartkamp MJ, van der Grond J, de Leeuw FE, et al. anterior cerebral artery cortical branches. Clin Anat. 2000; 13(4): Circle of Willis: morphologic variation on three-dimension- 231–236, doi: 10.1002/1098-2353(2000)13:4<231::aid- al time-of-flight MR angiograms. Radiology. 1998; 207(1): ca1>3.0.co;2-t. 103–111, doi: 10.1148/radiology.207.1.9530305, indexed 18. Yeniçeri IÖ, Çullu N, Deveer M, et al. Circle of Willis variations in Pubmed: 9530305. and artery diameter measurements in the Turkish popula- 13. Krishnamurthy A, Nayak SR, Bagoji IB, et al. Morphometry tion. Folia Morphol. 2017; 76(3): 420–425, doi: 10.5603/ of A1 segment of the anterior cerebral artery and its clini- FM.a2017.0004, indexed in Pubmed: 28150270.

Morphological study of myelinated and unmyelinated fibres in the sacrococcygeal dorsal roots of the rat J.-C. Lee1, C.-H. Cheng2, C.-T. Yen1

1Department of Life Science, National Taiwan University, Taipei, Taiwan 2Department of Veterinary Medicine, National Taiwan University, Taipei, Taiwan

[Received: 22 August 2018; Accepted: 3 October 2018]

Background: The number and calibre of myelinated and unmyelinated fibres of the sacrococcygeal dorsal roots innervating the tail of rats were studied by means of light and electron microscopy. Materials and methods: There were an estimated total of 12,500 myelinated and 25,500 unmyelinated dorsal root fibres innervating the tail of a rat. Results: The results showed that from the second sacral (S2) to the fourth sacral (S4) segment, the fibre diameter spectrum of myelinated fibres within each dorsal root was bimodal with two peaks at 5 microns and 10 microns, respectively. The first sacral (S1) segment was composed of numerous smaller-size myelinated fibres, thus forming a right-skewed distribution. The coccygeal (Co) segments showed a unimodal distribution peaking at 10 microns for the first (Co1) segment and gradually shifting to 7 microns for the third (Co3) segment. Overall, there was a continuous relative increase of the larger vs. the smaller myelinated fibres from the sacral to coccygeal segments. The fibre diameter of unmyelinated fibres of all these roots was unimodal with a single peak at 0.5 microns. The ratio of unmyelinated to myelinated fibre numbers was on average 2.83 for the S1–S2 roots, 1.66 for the S3–S4 roots, and 1.24 for the coccygeal roots. Conclusions: The comparison of the left- and right-side nerve fibres show that there was no significant difference, thus implying a symmetrical sensory innervation of the rat’s tail. (Folia Morphol 2019; 78, 2: 267–273)

Key words: tail, dorsal root, axon diameter, frequency distribution

INTRODUCTION or in a region. The evaluation of nerves’ fibre popu- Axonal size is the most important determinant lations is frequently performed by light microscopy of the nerve fibres’ functional parameters, including with osmium- or silver-impregnated preparations. their conduction velocity, refractory period, and sen- This technique is suitable to examine myelinated fi- sory and motor functions [27, 28]. Close correlations bres. On the other hand, electron microscopy allows between conduction velocity and nerve fibre diameter a correct evaluation of all fibre populations [17]; but, have been revealed in myelinated [4, 10, 26, 31] and its high magnification sometimes makes it impractical unmyelinated fibres [20]. Therefore, it is important to perform a total count of the nerve population of to ascertain the nerve fibre composition in a nerve the whole nerve.

Address for correspondence: Dr. C.-T. Yen, Department of Life Science, National Taiwan University, Taipei, Taiwan 106, tel: +886-2-3366-2451, fax: +886-2-2363-6837, e-mail: [email protected]

267 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. The numbers of unmyelinated (UMY) and myelinated (MY) fibres in sacrococcygeal dorsal roots of one exemplar rat Roots UMY Sampling ratio MY C/A ratio LRLRLRLR Average S1 5049 5654 34.6% 31.5% 1795 1657 2.81 3.41 3.11 S2 1734 2248 34.9% 45.9% 1235 1287 1.4 1.75 1.58 S3 2028 2353 47.2% 42.8% 1037 1006 1.96 2.34 2.15 S4 1281 1739 27.8% 21.8% 838 840 1.53 2.07 1.8 Co1 832 1023 50.5% 30.9% 748 782 1.11 1.31 1.21 Co2 460 735 54.3% 56.6% 426 394 1.08 1.87 1.47 Co3 181 163 98.8% 97.6% 103 144 1.76 1.13 1.44 Sum 11,565 13,915 6182 6110 L — left; R — right

The present study is an electron and light micro- vations. Light microscopic photographs of nerve sec- scopic analysis of the number of dorsal root axons for tions were obtained at a final magnification of 400×. all segments of the sacrococcygeal cord innervating the From these photographs, we counted the number tail of the rat [33]. We estimated that a total of 12,500 and obtained the morphometric histograms of myelin- myelinated and 25,500 unmyelinated dorsal root fibres ated fibres. The myelinated fibres with calibre equal innervate the tail of a rat. The ratio of unmyelinated or smaller than 6.5 micron were assigned as A-delta fibres vs. myelinated fibres ranges from 3.4 to 1.1 in group, whereas those larger than 6.5 micron were the individual roots. The frequency distribution of my- assigned as A-beta group. The ultrathin (gold) sections elinated fibre diameter was bimodal with a continued were placed on formvar-coated copper grids. These shift from a smaller calibre (A delta fibre) to larger cal- sections were stained with 0.2% lead citrate and 1% ibre (A beta fibre), from sacral to coccygeal segments. uranyl acetate and photomicrographic montages of This will be discussed in relation to the differential the nerve at 4000× magnification were made. Using functions of the segment organisation of the tail. the transmission electron microscope (JEOL: JEM-1200 EXII, USA), we photographed the nerve’s total surface. MATERIALS AND METHODS The photographs were then overlapped partly in order Sample preparation to reconstruct the whole nerve. The montaged pho- Thirteen male Wistar rats weighing 250–400 g tography of the whole nerve was used in counting the were used in this study. Rats were anesthetised by in- total number of unmyelinated fibres and to identify traperitoneal injections of sodium pentobarbital (100 the location of each and every Remak bundle. All Re- mg/kg). When anaesthesia was deep, the right atrium mak bundles were further photographed at 7500× for was opened, and the animal was perfused through morphometric analysis. Since the copper grid blocked a needle with buffered saline followed by 500 mL part of the sample, the counting of unmyelinated fibres of fixative which contained 4% paraformaldehyde was carried out on the observable region only. The total and 1% glutaraldehyde in 0.1 M phosphate buffer, number of unmyelinated fibres in the whole roots was pH 7.4. After perfusion ceased, the laminectomy was estimated using the ratio of the sampling area (Table 1). performed from segments L6 to Co3. Dorsal roots were then identified and removed from the animal. Morphometric analysis A 5 mm-long piece of central process was cut about The contact print photographs were measured 2 mm away from the dorsal root ganglion. The roots with a computer-operated planimeter (Jandel corp., were then rinsed with buffer and placed in fresh fixa- USA). The fibres were traced one by one, and the cali- tive for 2 h and then placed in 1% osmium tetroxide in bre areas were measured and written to a worksheet. 0.1 M phosphate buffer, pH 7.4. The tissue was again Diameters were computed by taking areal measure- rinsed, dehydrated in ascending concentrations of eth- ments and calculated assuming a perfect round shape. anol then acetone, and embedded in the Spurr plastic. Therefore, the diameters are abstraction numbers.

Histological methods Statistical methods From each specimen, we prepared semithin sections The data was expressed in terms of mean values stained with toluidine blue for light microscope obser- ± standard deviations (SD). The ratio of A-delta vs.

268 J.-C. Lee et al., Sacrococcygeal dorsal roots of the rat

Figure 2. An electron micrograph of a typical group of unmyelinated axons (a Remak bundle) in a rat dorsal root. Note the large prominent nucleus of the Schwann cell in the left-hand side of the picture. The unmyelinated axons are smaller, irregularly round, or oval profiles embedded in troughs of Schwann cell cytoplasm. Edges of larger myelinated axons can be seen at the upper-right corner of this micrograph. M — myelinated axon; UM — unmyelinated axon; N — nucleus of the Schwann cell. Calibration bar is 1 micrometre.

in troughs of Schwann cell cytoplasm (Fig. 2). The major Figure 1. Representative examples of photomicrographs of my organelles in these axons are neurofilaments and micro- elinated fibres in several sacrococcygeal dorsal roots S1 (A, B), tubules but mitochondria and cisternae of the agranular S3 (C, D), Co1 (E, F) and Co3 (G, H). The whole roots are shown on the left (A, C, E, G), and their enlarged parts are shown on the right endoplasmic reticulum can occasionally be seen. (B, D, F, H). The myelinated axons and blood vessels are the most Myelinated axon numbers were obtained by total obvious components. The ventral root adheres to the dorsal root in counting, while unmyelinated axon numbers by par- the first coccygeal sample (E). S1 contains a greater proportion of smaller myelinated axons compared to the other segments. tial sampling. The sampling ratio varied from 21.8% to 98.8% (Table 1). We estimated that a total of 12,500 myelinated and 25,500 unmyelinated dorsal root A-beta fibre was analysed by one-way ANOVA on fibres innervate the tail of a rat. The ratio of unmy- rank followed with post-hoc multiple comparison elinated fibres vs. myelinated fibres (C/A ratio) ranges (Dunn’s Method, Sigmaplot v.11) where appropriate. from 3.4 to 1.1. Note the ratio in the S1 dorsal root The significant level was p < 0.05. was much larger than in other segments (Table 1). The data pooled from 11 rats are shown in Table 2. RESULTS The fibrediameter frequency distribution of mye- Dorsal roots consist of myelinated and unmyelinated linated fibre (Fig. 3) was bimodal and unmyelinated axons embedded in a collagenous connective tissue. fibre(Fig. 4) was unimodal. It can be seen that there Perineurial cells line each root and blood vessels are are two or three peaks in the S1 dorsal roots. Note prominent within them. Figure 1 illustrates the mor- that the S1 roots had a high proportion of small fibres phology of exemplar S1, S3, Co1, and Co3 dorsal roots whereas those of S2–S4 have a greater proportion of in semi-thin sections. The myelinated axons and blood medium and large size fibres with peaks at 5 microns vessels are the most obvious components. Note that the and 10 microns. myelinated axons in the S1 contain a greater proportion The myelinated fibre histogram indicates a decline of smaller axons compared to the other segments. The change from S1 to Co2 in the proportion of small myelinated axons can be counted with both the light and calibre A-delta (≤ 6.5 micron) vs. large calibre A-beta electron microscopes, because they are surrounded by an (> 6.5 micron) fibres. The S1 segment was composed easily visible sheath. The unmyelinated axons are round of a significantly larger portion of small A-delta fibres or irregularly oval on cross section and are embedded than the S4, Co1, and Co2 segments (Fig. 5). Note

269 Folia Morphol., 2019, Vol. 78, No. 2

Table 2. The numbers of unmyelinated (UMY, C fibre) and myelinated (MY, A fibre) fibres in sacrococcygeal dorsal roots of 11 rats Root UMY MY C/A ratio Mean SD N Mean SD N S1 4959 899 6 1659 184 10 2.99 S2 2984 1023 6 1147 106 7 2.60 S3 1828 484 6 931 142 9 1.96 S4 1305 472 4 642 234 4 2.03 Co1 813 165 4 568 176 11 1.43 Co2 577 121 4 504 230 9 1.14 Co3 172 13 2 189 101 7 0.91 Sum 12,637 5640

SD — standard deviation

Figure 3. Examples of histograms of myelinated axon fibre size of Figure 4. Examples of histograms of unmyelinated axon fibre size sacrococcygeal dorsal roots in one rat. Note bimodal distribution of of sacrococcygeal dorsal roots in one rat. Note unimodal distribu- fibre size in the S2 to S4 dorsal roots. The A-delta fibres were most tion of the fibre size in all segments. abundant in the S1 dorsal roots, whereas their peaks were much lower in the coccygeal segments. the Co3 segment also contained a larger portion S1 was much larger than other roots, although it did of A-delta fibres, but not reaching significant level. not reach significant level by ANOVA analysis. The fibre spectrum of unmyelinated fibres of these roots was unimodal with a single peak at 0.5 microns DISCUSSION (Fig. 4). The fibre number ratio of unmyelinated vs. The tail of the rat is a prominent organ with mul- myelinated fibres (C/A ratio) was 2.83 for the S1–S2 tiple functions. In an adult rat, the tail is over 20 roots, 1.66 for the S3–S4 roots, and 1.24 for the cm. It serves as the fifth limb of the rat for reaching, coccygeal roots, with a decreasing trend from sacral holding, balancing, and climbing. The tail of the rat is to coccygeal segments. Note the C/A ratio value of innervated at its base by the first segment of the sacral

270 J.-C. Lee et al., Sacrococcygeal dorsal roots of the rat

3.0 * our data showed an estimation of 21% less unmyelinated fibres and 39% less myelinated fibres in Wistar 2.5 rats. The different species of the rats may contribute

2.0 to the differences in nerve fibre population. However, Langford and Coggeshall’s study [18] was performed 1.5 5–7 days after cutting the ventral root and sympathectomy surgery whereas our study kept all components A-delta / A-beta 1.0 intact. The injury of the neural tissue would trigger 0.5 the development of the adjacent intact component to compensate for the loss of function and even to cause 0.0

S1 S2 S3 S4 Co1 Co2 Co3 neuropathic outcomes. The sprouting of nerve branch- Segment es caused by injury has been reported in DRG [6, 21], spinal cord [5], and dorsal roots [15]. The composition Figure 5. Ratio of A-delta vs. A-beta myelinated fibres. The S1 of the sciatic nerve was determined with more myeli- segment contained a significantly larger portion of A-delta fibres compared with S4, Co1, and Co2 segments. Note the Co3 segment nated fibres in the sympathectomy group compared was also composed of a large portion of A-delta fibres, but did not with control group [29]. We propose that the injury of show significant differences compared with other segments. One- ventral roots and sympathetic nerve may cause neuro- -way ANOVA on rank with Dunn’s post hoc comparison; *p < 0.05. plasticity and affect the morphometric analysis. The ratio of unmyelinated to myelinated fibres cord (S1). There is a regularly organised dermatome is at least 1.1 for the total fibre population at the from S1 to S4 followed by the coccygeal segments, coccygeal level, but it can be greater than 3 for the from rostral tail to the most caudal tip of the tail [33]. sacral component. It should be emphasized that this There were remarkable differences in the propor- C/A ratio is comparable to the situation observed in tions of axons of different size categories in the roots somatic nerves; Serratrice et al. [30] found the ratio we examined. For example, the S1 dorsal roots had of 4 for the sensory population of the sural nerve. a large proportion of small myelinated fibres and un- At the thoracic level of the vagus nerve, this pro- myelinated fibres. The meaning of this phenomenon portion is much higher since the vagi that enter the is not clear, but one possible correlation is that the abdominal cavity consist almost entirely of C fibres. S1 roots in the rat are an important supplier of the This has been demonstrated both by histological sensory innervation of the pelvic viscera. This might [1, 12, 14] and electrophysiological studies [16, 19, imply that the sensory visceral innervation is predom- 22, 25]. On the other hand, we find at this level inantly by small cells. The axon-calibre histogram of a clear overlapping of spectra of the unmyelinated dorsal roots differ on a segmental basis, which may and myelinated fibres. This shows that the classifica- provide clues for experiments designed to provide tion of the smallest myelinated fibres and the largest insights into the organisation of segmental sensory C fibresis difficult using only diameter as a criterion. input in the spinal cord. Nevertheless, the unmyelinated fibres are usually in It is now possible to estimate precisely the total a unimodal distribution, as in somatic nerves [23]. sensory and motor population of nerve fibres in the There is no clear consensus in the literature regard- dorsal and ventral roots of the rat, including unmyeli- ing the actual shape of nerve fibres in the transverse nated fibres using electron microscopy. The number section. Some researchers suggest that the minimum of myelinated fibres established in the present study is diameter represents the true diameter. Others claim comparable to that obtained with electron microscopy. means of two orthogonal diameters. Still others ac- The data of the unmyelinated component is regularly cept every irregular shape as the true shape and underestimated by light microscopy. This is true for calculate the diameter from traced circumferences or certain somatic nerves [30] and also for the dorsal areas [13]. Orgel (1980) [24] has noted that as long roots. We estimated a total of 25,500 unmyelinated as one measurement method is used consistently and 12,500 myelinated fibres in the sacrococcygeal throughout a study, this issue can be disregarded. (S1–Co3) dorsal roots of adult Wistar rats. Compared A central dogma of vertebrate anatomy is that ven- with a previous study on counting nerve fibres of tral spinal roots contain motor fibres, the cell bodies S1–S3 dorsal roots of adult Sprague-Dawley rats [18], of which are in the spinal cord, and dorsal roots con-

271 Folia Morphol., 2019, Vol. 78, No. 2

tain sensory fibres, with cell bodies in the dorsal root the heart, lungs and abdominal viscera in the cat. J Physiol. ganglion. However, several investigators have shown 1957; 135(1): 182–205, indexed in Pubmed: 13398974. 2. Applebaum ML, Clifton GL, Coggeshall RE, et al. Unmy- by both anatomical and physiological methods that elinated fibres in the sacral 3 and caudal 1 ventral roots sensory fibres are also present in the ventral roots [7]. of the cat. J Physiol. 1976; 256(3): 557–572, indexed in In spinal segments S3 and caudal Co1 of the cat, which Pubmed: 1271293. contain part of the parasympathetic outflow, 29% of 3. Besson JM, Chaouch A. Peripheral and spinal mechanisms of the axons are unmyelinated [2]. Approximately half of nociception. Physiol Rev. 1987; 67(1): 67–186, doi: 10.1152/ physrev.1987.67.1.67, indexed in Pubmed: 3543978. these unmyelinated fibres are preganglionic efferent 4. Boyd IA, Kalu KU. Scaling factor relating conduction ve- axons, and the rest are sensory [11]. In the rat, 30% locity and diameter for myelinated afferent nerve fibres in of the axons of ventral roots L6 and S1 are unmyeli- the cat hind limb. J Physiol. 1979; 289: 277–297, indexed nated [9]. Thirty per cent of these unmyelinated fibres in Pubmed: 458657. 5. Carlton S, Coggeshall R. Sprouting and reorganization in arise from dorsal root ganglion cells. The other 70% the spinal cord after nerve injury. Mechanisms and Media- arise from the spinal cord and are presumed to be tors of Neuropathic Pain. 2002: 89–106, doi: 10.1007/978- preganglionic parasympathetic axons. The variable 3-0348-8129-6_6. distribution also occurs in the rat [8] where ventral 6. Chung K, Lee BH, Yoon YW, et al. Sympathetic sprouting in the dorsal root ganglia of the injured peripheral nerve in a rat neu- roots T1–L2, L6, and S1 contain many unmyelinated ropathic pain model. J Comp Neurol. 1996; 376(2): 241–252, and small myelinated axons, and segments C1–C8 and doi: 10.1002/(SICI)1096-9861(19961209)376:2<241::AID- L3–L5 contain few. During this study, we observed CNE6>3.0.CO;2-3, indexed in Pubmed: 8951640. several of the coccygeal ventral roots (e.g., Fig. 1E), 7. Coggeshall RE, Coulter JD, Willis WD. Unmyelinated axons and found that the unmyelinated fibre composed in the ventral roots of the cat lumbosacral enlargement. J Comp Neurol. 1974; 153(1): 39–58, doi: 10.1002/ less than 10% of the total axon number. The number cne.901530105, indexed in Pubmed: 4817345. decreased from Co1 to Co3, and was much less than 8. Coggeshall RE, Ito H. Sensory fibres in ventral roots L7 and that of the sacral ventral roots. This observation im- Si in the cat. J Physiol. 1977; 267(1): 215–235, indexed plies that our investigation of the tail might have been in Pubmed: 559756. 9. Coggeshall RE. Law of separation of function of the spinal slightly underestimated for around 10%. roots. Physiol Rev. 1980; 60(3): 716–755, doi: 10.1152/ There has been extensive study of peripheral and physrev.1980.60.3.716, indexed in Pubmed: 6994143. spinal nociceptive neurons [3, 32]. Nociceptive infor- 10. Cragg BG, Thomas PK. The relationships between conduc- mation is carried by two populations of peripheral tion velocity and the diameter and internodal length of fibres: thinly myelinated A-delta fibres and unmyeli- peripheral nerve fibres. J Physiol. 1957; 136(3): 606–614, indexed in Pubmed: 13429525. nated C fibres. A decrease in the ratio of the thinly 11. Emery DG, Ito H, Coggeshall RE. Unmyelinated axons in myelinated and the unmyelinated fibres in the caudal thoracic ventral roots of the cat. J Comp Neurol. 1977; segments of the coccygeal level, which innervate the 172(1): 37–47, doi: 10.1002/cne.901720103, indexed in tip of the tail, implies that the sensitivity for pain is Pubmed: 838878. 12. Evans DH, Murray JG. Histological and functional stud- lower in the tip of the tail. This fits well with our own ies on the fibre composition of the vagus nerve of the observation that the laser heat sensitivity of the tail- rabbit. J Anat. 1954; 88(3): 320–337, indexed in Pu- tip is the lowest compared to that of the base or the bmed: 13192020. middle part of the tail. 13. Ewart DP, Kuzon WM, Fish JS, et al. Nerve fibre morphometry: a comparison of techniques. J Neurosci Methods. 1989; 29(2): 143–150, indexed in Pubmed: 2770338. CONCLUSIONS 14. Gabella G, Pease HL. Number of axons in the abdominal A holistic picture of an organ’s total innervation vagus of the rat. Brain Res. 1973; 58(2): 465–469, indexed is important for understanding the function of that in Pubmed: 4756138. organ. In this study, we revealed the afferent organ- 15. Hulsebosch CE, Coggeshall RE. Quantitation of sprouting of dorsal root axons. Science. 1981; 213(4511): 1020– isation of the tail and demonstrated the continually 1021, indexed in Pubmed: 7268404. changed calibre profile of nerve fibres. The description 16. Iggo A. The electrophysiological identification of single of the tail’s whole sensory innervation can be a useful nerve fibres, with particular reference to the slowest-con- data point for comparison with other organs or limbs ducting vagal afferent fibres in the cat. J Physiol. 1958; 142(1): 110–126, indexed in Pubmed: 13564422. of the rat, as well as those of other species. 17. Kissin I, Freitas CF, Mulhern HL, et al. Sciatic nerve block with resiniferatoxin: an electron microscopic study of un- REFERENCES myelinated fibers in the rat. Anesth Analg. 2007; 105(3): 1. Agostoni E, Chinnock JE, Daly MB, et al. Functional and 825–831, doi: 10.1213/01.ane.0000277491.40055.47, histological studies of the vagus nerve and its branches to indexed in Pubmed: 17717246.

272 J.-C. Lee et al., Sacrococcygeal dorsal roots of the rat

18. Langford LA, Coggeshall RE. Branching of sensory axons 26. Ritchie JM. On the relation between fibre diameter and in the peripheral nerve of the rat. J Comp Neurol. 1981; conduction velocity in myelinated nerve fibres. Proc R Soc 203(4): 745–750, doi: 10.1002/cne.902030411, indexed Lond B Biol Sci. 1982; 217(1206): 29–35, doi: 10.1098/ in Pubmed: 7328204. rspb.1982.0092, indexed in Pubmed: 6131421. 19. Leek BF. Abdominal and pelvic visceral receptors. Br Med 27. Rushton WAH. A theory of the effects of fibre size in Bull. 1977; 33(2): 163–168, indexed in Pubmed: 324560. medullated nerve. J Physiol. 1951; 115(1): 101–122, in- 20. Matsumoto G, Tasaki I. A study of conduction velocity dexed in Pubmed: 14889433. in nonmyelinated nerve fibers. Biophys J. 1977; 20(1): 28. Schalow G, Zäch GA, Warzok R. Classification of human 1–13, doi: 10.1016/S0006-3495(77)85532-X, indexed in peripheral nerve fibre groups by conduction velocity and Pubmed: 901899. nerve fibre diameter is preserved following spinal cord 21. McLachlan EM, Jänig W, Devor M, et al. Peripheral lesion. J Auton Nerv Syst. 1995; 52(2-3): 125–150, indexed nerve injury triggers noradrenergic sprouting within in Pubmed: 7615895. dorsal root ganglia. Nature. 1993; 363(6429): 543–546, 29. Schmalbruch H. Fiber composition of the rat sciatic doi: 10.1038/363543a0, indexed in Pubmed: 8505981. nerve. Anat Rec. 1986; 215(1): 71–81, doi: 10.1002/ 22. Mei N. Mecanorecepteurs vagaux digestifs chez le chat. ar.1092150111, indexed in Pubmed: 3706794. Exp Brain Res. 1970; 11(5), doi: 10.1007/bf00233971, 30. Serratrice G, Mei N, Pelissier JF, et al. Cutaneous, mus- indexed in Pubmed: 5490689. cular and visceral unmyelinated afferent fibres: com- 23. Ochoa J, Mair WG. The normal sural nerve in man. parative study. Peripheral Neuropathies. 1978: 69–82, I. Ultrastructure and numbers of fibres and cells. Acta doi: 10.1016/b978-0-444-80079-4.50010-7. Neuropathol. 1969; 13(3): 197–216, indexed in Pu- 31. Waxman SG. Determinants of conduction velocity in my- bmed: 5805973. elinated nerve fibers. Muscle Nerve. 1980; 3(2): 141–150, 24. Orgel MGA. Critical review of histological methods used doi: 10.1002/mus.880030207, indexed in Pubmed: 6245357. in the study of nerve regeneration. Nerve Repair and Re- 32. Willis WD. The pain system. The neural basis of nocicep- generation: Its clinical and experimental basis. St Louis: tive transmission in the mammalian nervous system. Pain Mosby. 1980: 141–148. Headache. 1985; 8: 1–346, indexed in Pubmed: 2983301. 25. Paintal AS. Vagal Afferent Fibres. Ergebnisse der Physiolo- 33. Yen CT, Chen RS. Tail region of the primary somatosensory cor- gie Biologischen Chemie und Experimentellen Pharmakol- tex and its relation to pain function. Novel Trends in Brain Sci- ogie. 1963: 74–156, doi: 10.1007/978-3-642-49896-1_3. ence. 2018: 233–252, doi: 10.1007/978-4-431-73242-6_14.

Combination of vitamin E and L-carnitine is superior in protection against isoproterenol-induced cardiac injury: histopathological evidence E.A. Huwait

Department of Biology, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia

[Received: 16 April 2018; Accepted: 23 May 2018]

Background: L-carnitine and vitamin E have antioxidant properties. This study aimed to assess the effectiveness of L-carnitine, vitamin E and their combination in protection against isoproterenol (ISO)-induced biochemical and histopathological changes in rat heart. Materials and methods: Fifty male Wistar rats assigned to five groups; control, ISO-treated group (100 mg/kg), ISO + vitamin E-treated group (100 IU/kg), ISO + L-carnitine (100 mg/kg) and ISO + vitamin E + L-carnitine treated group. At the end of the experiment, serum cardiac enzyme as well as the cardiac level malondialdehyde (MDA), antioxidant enzymes and inflammatory cytokines interleukin-6, tumour necrosis factor alpha (TNF-a) were assessed. Histopathological changes in the left ventricle wall were assessed using the light and electron microscopy. Results: Treating rats with vitamin E and L-carnitine could alleviate ISO-induced changes as it significantly reduced the serum level cardiac enzymes, MDA and IL-6, TNF-a and improved the antioxidants enzymes (SOD, GSPxase and GSRase). Histopathologically, they improved cardiac fibres atrophy, haemorrhages between cardiac fibres, lost striations, and disturbed sarcomere structure. The combined effect of vitamin E and L-carnitine was more superior compared to the other groups. Conclusions: Combined administration of vitamin E, L-carnitine ameliorated the biochemical and histopathological cardiac injury induced by ISO. The effect seemed to be mediated through the antioxidant and anti-inflammatory effect of vitamin E, L-carnitine. Administration of these two elements is recommended for patient at risk for myocardial infarction. (Folia Morphol 2019; 78, 2: 274–282)

Key words: vitamin E, L-carnitine, isoproterenol, heart, histology, antioxidant, anti-inflammatory

INTRODUCTION conditions [25]. It was reported that “Carnitine plays L-carnitine, L-trimethy l-3-hydroxy ammoni- a pivotal role in myocardial energy metabolism obetanoate, is a natural endogenous water soluble through transport of long-chain fatty acyl interme- antioxidant located on the mitochondrial membrane diates, across the inner mitochondrial membrane for and is found in all mammals. It reduces the intra- subsequent oxidation, and regulation of carbohydrate cellular buildup of toxic metabolites in ischaemic metabolism by modulation of the intramitochondrial

Address for correspondence: Dr. E.A. Huwait, Associate Professor, Department of Biology, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia, e-mail: [email protected]; [email protected]

274 E.A. Huwait, Vitamin E and L-carnitine reduce isoproterenol-induced injury?

acetyl-CoA:CoA ratio” [26]. As such, it is used in the and L-carnitine were obtained from GNC Store at prevention and treatment of oxidative stress and Jeddah, Saudi Arabia. related health problems [1]. It exerts its antioxidant potential by reducing reactive oxygen species (ROS) Animals production and scavenging free radicals [17]. Fifty male Westar albino rats weighing from 150 A link was reported between consumption of to 200 g and were purchased from the experiment carnitine and cardiovascular diseases [27]. However, animal unit at the KFMRC. They were kept in large contradictory data were found regarding beneficial cages at 25°C with a 12-h dark/light cycle with free effects of L-carnitine. Liepinsh et al. [18] reported mobility and allowed free access to water and stand- that long-term decrease in L-carnitine is important ard rat diet ad libitum. They were left to acclimatise for the energy metabolism regulation and treatment for 2 weeks in their cages before start the experiment. of atherosclerosis and heart diseases. On the other hand, Zambrano et al. [35] reported that arterial Experimental protocols hypertension-related cardiac fibrosis could be inhib- Rats were randomly assigned to five groups (n = 10). ited by L-carnitine through modulation of peroxisome — Group I — served as a control group. Rats of this proliferator activated receptor-g expression. group were injected subcutaneously with 2 mL of Vitamin E, a fat soluble vitamin, possesses saline daily for 4 weeks; a powerful antioxidant activity [2]. Administration of — Group II — in which rats were subcutaneously in- vitamin E was found to decrease the cardiovascular jected with isoproterenol (100 mg/kg body weight events in diabetic patients [20]. Contradictory data (BW) in 2 mL of saline daily for 3 successive days were found in literature regarding the role of vitamin E according to Panda and Kar [23]; in protection against cardiac disease. Hu et al. [13] — Group III — in which rats were subcutaneously found that vitamin E exerted harmful effect in young injected with isoproterenol with same dose and female mice, as it increased cardiomyocyte apoptosis duration plus vitamin E (100 IU/kg BW) intraperi- after induction of myocardial infarction (MI) while it toneal injection daily for 4 weeks according to is protective in aged male mice. Gayathri et al. [8]; Myocardial infarction is an ischaemia-induced — Group IV — in which rats were subcutaneously cardiac muscle necrosis which results from “imbal- injected with isoproterenol plus intraperitoneal ance between coronary blood supply and myocardial daily injection of L-carnitine (100 mg/kg BW) for demand” [32]. MI induced using isoproterenol was 4 weeks according to Wong et al. [33]; adequately reproduced by many researchers [11]. It — Group V — in which rats were subcutaneously was used to test therapeutic or cardioprotective ef- injected with isoproterenol plus vitamin E and fect of any new medication or adjuvant natural sup- L-carnitine at the same dose for 4 weeks. plementation [7]. Therefore, this study was designed to evaluate the effectiveness of L-carnitine, vitamin E Biochemical assessment and a combination of both in protection against At the end of the experiment, blood sample was isoproterenol-induced biochemical histopathological collected from the retro-orbital vein under after thio- changes in rat heart. pental anaesthesia. Blood was centrifuged at 4000 g at 4°C for 15 min and the serum was stored at –80°C MATERIALS AND METHODS till the time of the biochemical assessment of serum Ethical approval aspartate transaminase (AST), lactate dehydrogenase This study was conducted after obtaining the (LDH), and creatine kinase (CK-MB) by colorimetric ethical approval of the Biomedical Research Ethics method using commercial KIT from Biomedical [30]. Committee, Faculty of Medicine, King Abdulaziz Uni- Rats were then sacrificed with cervical dislocation, versity, Jeddah Saudi Arabia and was done at the the chest was open and the heart was immediately King Fahad Medical Research Centre (KFMRC), King dissected out. Abdulaziz University. Part of the heart wall was homogenised in 9 volumes of ice-cold saline and centrifuged at 8000 g at Drug 4°C for 20 min. Malondialdehyde (MDA) level was as- Isoproterenol hydrochloride was obtained from sessed in the homogenate as was described according Sigma chemical company, St. Louis, MO, USA. Vitamin E to Moran [21]. Superoxide dismutase (SOD) activity 275 Folia Morphol., 2019, Vol. 78, No. 2

was assessed in the homogenate by the nitro blue The levels of SOD, GSPxase and GSRase, the car- tetrazolium reduction method according to Habig diac antioxidant enzymes, were assessed in the ho- et al. [12]. Glutathione peroxidase (GSPxase) was mogenate of the cardiac tissue. It was observed that assessed as described by Pagila and Valentine [22] administration of ISO induced a significant increase in and glutathione reductase (GSRase) was estimated the level of SOD and a significant decrease in the level according to Deore et al. [6]. Levels of inflammatory of GSPxase and GSRase compared with the control cytokines interleukin 6 (IL-6), tumour necrosis fac- group. Administration of vitamin E, L-carnitine or both tor alpha (TNF-a) were measured by ELISA kits from along with ISO significantly reduce the level of SOD BIORAD (England). and significantly increased the levels of GSPxase and GSRase when compared with the untreated group. Histopathological assessment Although none of these antioxidants reached the The dissected heart was washed with saline solu- level observed in the control group, the combined tion, cut transversely and longitudinally and fixed effect of both vitamin E and L-carnitine was more in 10% buffered neutral formalin solution, and pro- significant on the level of GSRase compared with cessed for obtaining paraffin blocks. These blocks other enzymes (Fig. 1C). When it came to the levels were sectioned serially at 4–6 µ and routinely stained of the inflammatory cytokines, it was found that the with haematoxylin and eosin (H&E) [3]. levels of IL-6 and TNF-a showed a significant increase Specimens measured about 2 × 2 mm from left in the cardiac tissue after administration of ISO while ventricle were fixed in 3% glutaraldehyde in phos- their levels were significantly reduced in rats receiving phate buffer at pH 7.4 for 24 h, and post-fixed in both vitamin E and L-carnitine, compared with the 1% osmic acid for 1 h and further processed to be untreated rats (Fig. 1D, E). examined with transmission electron microscopy (TEM) (JEM-100 Cx11; Jeol) at Assuit University, As- Histopathological results suit, Egypt. In order to determine the orientation of Light microscopic findings.The light microscopic the ultrathin sections, semithin ones with thickness inspection of the left ventricle of the control rat re- 1 mm were prepared and stained using toluidine blue vealed intact cylindrical cardiac muscle fibres, with to be examined by the light microscope. The ultrathin acidophilic cytoplasm and ill-defined transverse stria- sections (prepared at the thickness of 500–800 Å) tions. The nuclei of the cardiac fibres were vesicular, were stained with uranyl acetate and lead citrate. oval or rounded and central in position. Connective tissue between the fibres was scanty and contained RESULTS thin wall blood capillaries and few fibroblasts identi- Biochemical findings fied by their flat nuclei. The blood vessels, which were In this study, rats that received isoproterenol (ISO) branches of the coronary vessels between the cardiac showed a significant increase in the serum level of muscle fibres, had intact wall with normal structure cardiac enzymes including AST, LDH and CK-MB com- and perivascular tissue (Fig. 2A, B). pared to the control rats. Treating rats with vitamin E, Isoproterenol administration induced histological L-carnitine or both along with ISO resulted in a sig- alterations including atrophy of some cardiac fibres and nificant reduction in these enzymes levels compared degeneration of some fibres, which was recognised by to the untreated rats. However, the level of these the dark cytoplasm and small pyknotic nuclei. Some enzymes did not reach that of the control group. The capillaries appeared congested or ruptured, resulting combined effect of both vitamin E and L-carnitine was in haemorrhagic foci between the fibres. The blood more significant on the level of CK-MB compared with vessels between the fibres appeared damaged with LDH and AST (Fig. 1A). The level of MDA, a marker of perivascular inflammatory cells around them (Fig. 2C, D). oxidative stress, was assessed in the cardiac tissue ho- Sections in the heart wall of groups treated with mogenate. It was observed that ISO significantly ele- vitamin E, L-carnitine or their combination showed that vated MDA level in the cardiac homogenate compared most of the cardiac muscle were intact while few fibres to the control group. Administration of vitamin E, were degenerated and the blood vessels appeared nor- L-carnitine or both along with ISO significantly de- mal similar to control (Fig. 2E, J). creased in MDA in the cardiac tissue compared to Semithin sections prepared from the same sam- the untreated group (Fig. 1B). ples confirmed what was observed by routine histo-

276 E.A. Huwait, Vitamin E and L-carnitine reduce isoproterenol-induced injury?

Figure 1. Changes in cardiac enzyme (lactate dehydrogenase [LDH], aspartate transaminase [AST], creatine kinase [CK-MB]) (A), malondialdehyde (MDA) (B), antioxidant (C), tumour necrosis factor alpha (TNF-a) and interleukin 6 (IL-6) levels (D) in the studied groups; GI — control group; GII — isoproterenol group; GIII — isoproterenol + vitamin E group, GIV — isoproterenol + L-carnitine group; GV — isoproterenol + vitamin E + L-carnitine group; GSPxase — glutathione peroxidase; GSRase — glutathione reductase; SOD — superoxide dismutase.

Figure 2. Transverse (A) and longitudinal (B) sections in control rat heart (GI) showed intact cardiac muscle fibres (black arrow) and intact blood vessels (white arrow). Sections in heart of GII rats (C, D) show atrophy and decreased size of cardiac fibres (thick arrows) with some haemorrhages in between them (asterisk). Some muscle fibres appear degenerated with dark cytoplasm and small pyknotic nuclei (interrupted arrow). The blood vessel (white arrow) appears damaged with perivascular inflammatory cells around. Sections in the heart of GIII rats (E, F) and those of GIV rats (G, H) showing few degenerated muscle fibres (interrupted arrows) and the blood vessels appear intact. Sections in the heart of GV rats (I, J) showing almost intact muscle fibres and blood vessels (H&E, ×400); GI — control group; GII — isoproterenol group; GIII — isoproterenol + vitamin E group, GIV — isoproterenol + L-carnitine group; GV — isoproterenol + vitamin E + L-carnitine group; LS — longitudinal section; TS — transverse section.

277 Folia Morphol., 2019, Vol. 78, No. 2

Figure 3. Semithin sections in the heart of GI (A) showing intact cardiac muscle fibres with vesicular central nuclei (arrows) and intact striation. Note, the presence of thin-walled capillaries (white arrow). Section in the heart of GII (B) showing reduced diameter of muscle fibres; some of these fibres have dark small nuclei (thick arrow) and some lost their striations (asterisk). Note that most of blood capillaries are dilated and congested (white arrow). Sections in hearts of GIII (C), GIV (D) and GV (E) show that most of muscle fibres are intact with vesicular nuclei (arrow). Striations are preserved in most of the muscle fibres while few lost them (toluidine blue stain, ×600); GI — control group; GII — isoproterenol group; GIII — isoproterenol + vitamin E group, GIV — isoproterenol + L-carnitine group; GV — isoproterenol + vitamin E + L-carnitine group.

logical stains. The cardiac muscle fibres of the control euchromatic nuclei, intact, well-defined sarcomeres rats appeared intact with preserved striation and and intact mitochondria with normal size and shape. intact capillaries in between. Evident focal damage On the other hand, cardiac fibres of rats receiving of cardiac fibres was observed in rats receiving ISO ISO appeared atrophied with occasional ill-defined as some fibres lost their striation compared to the sarcomeres, smaller mitochondria with disrupted control. Most of blood capillaries were dilated and cristae. Although the cardiac fibres of rats treated congested. On the other hand, most of the cardiac with vitamin E, L-carnitine or a combination of both muscle fibres in the hearts of the treated groups possessed preserved sarcomeres, the mitochondria showed preserved striations while few fibres had lost still appeared smaller in size and less frequently seen them. The rats receiving both vitamin E and L-carnitine in the fibres compared to the control rats (Figs. 4, 5). showed the best effect on preserving the structure of the cardiac muscle (Fig. 3). DISCUSSION Electron microscopic findings.When the ultras- “Myocardial infarction is a condition that results tructure of the cardiac muscle fibres was assessed us- from the interruption of coronary blood supply ing TEM, it was noticed that the control rats showed needed to satisfy myocardial demand, leading to

278 E.A. Huwait, Vitamin E and L-carnitine reduce isoproterenol-induced injury?

Figure 4. Electron microscope micrograph of rat cardiac muscle from the left ventricle of control group (GI) (A, B) showing normal structure of the cardiac fibres with open face nuclei (N), well defined sarcomeres (black arrows) and normal size and site of mitochondria (thick white arrows). Cardiac muscle of isoproterenol-treated group (GII) (C, D) showing reduced diameter (atrophy) of cardiac muscles (interrupted arrow) with occasional ill-defined sarcomeres (arrows). Mitochondria appear smaller in size with disrupted cristae (thick white arrows).

oxygen and nutrient deprivation of the heart, even- rats. In addition, it significantly increased SOD activity tually destroying cardiac tissues” [5]. “Isoproterenol, and significantly decreased GSPxase and GSRase activ- a synthetic catecholamine and adrenergic agonist ity indicating its harmful effect on the cardiac muscle. that is documented to produce in large dose due to These findingswere supported by those of Panda et generation of highly cytotoxic free radicals through al. [24] observed in rats after administration of ISO. its auto-oxidation”. These free radicals initiate lipid Free radicals generated by ISO stimulate lipid peroxidation and induce irreversible damage of the peroxidation of polyunsaturated fatty acid of the myocardial membrane [9]. This study was planned membrane, with subsequent loss of the structure and to assess the possible protective effect of vitamin E, function of the myocardium. When the metabolic L-carnitine, and a combination of both on ISO-in- function of the myocardium is disturbed, it releases duced changes in rat cardiac muscle. its lysosomal enzymes into the blood. Therefore, the The diagnostic cardiac enzymes CK-MB, LDH and serum levels of AST, LDH and CK-MB were found to AST are considered sensitive markers that are used be significantly increased [19]. to assess the severity of MI [32]. In this study, ISO Isoproterenol induced structural alterations in car- significantly increased the levels of diagnostic cardiac diac muscle. Among these alternations were the ap- enzymes (LDH, AST and CK-MB) in the serum and MDA pearance of some muscle fibres with dark cytoplasm level in the cardiac tissue compared with the control and small pyknotic nuclei, indicating apoptosis and

279 Folia Morphol., 2019, Vol. 78, No. 2

Figure 5. Electron microscope micrograph of rat cardiac muscle from the left ventricle of isoproterenol + vitamin E group (GIII) (A, B), isoproterenol + L-carnitine group (GIV) (C, D) and isoproterenol + vitamin E + L-carnitine group (GV) (E, F) showing increased diameter of most of the cardiac fibres, defined sarcomeres (black arrows) while the mitochondria (thick white arrows) still have small size and fewer number compared to the control group.

degeneration of these fibres. These finding was sup- ated with vascular changes, oedema and leucocyte ported by those previous reported by Shukla et al. [28] infiltration.Marked focal myonecrosis where among who stated that ISO initiates apoptosis evidenced by the histopathological changes induced by ISO in up-regulation of TUNEL and Bax expression and down rats as described by Shukla et al. [28]. In this study, regulation of Bcl-2 expression. Atrophy of cardiac ISO administration affected also the sarcomere, the muscle fibres as well as haemorrhage and conges- functional unit of muscle fibre, and mitochondria, tion of some blood vessel were observed following which are the powerhouses of the fibre. Damage ISO administration in this study. These finding were to membranous organelles, such as mitochondria, in agreement with what was reported by Panda et is attributed to lipid peroxidation induced by ISO al. [24] following administration of ISO. Gayathri et [10]. Tappia et al. [29] reported that “oxidative stress al. [8] reported that ISO administration was associ- increases cAMP levels by exhausting ATP, depresses

280 E.A. Huwait, Vitamin E and L-carnitine reduce isoproterenol-induced injury?

sarcolemmal Ca2+ transport resulting in intracellular CONCLUSIONS calcium overload, leading to ventricular dysfunction AND CLINICAL IMPLICATIONS and contractile failure in rat heart”. This biochemical and histological studies dem- In the present study, administration of L-carnitine onstrated that administration of vitamin E and was found to protect rat heart from ISO-induced L-carnitine ameliorated cardiac alteration induced changes observed in untreated group. It improved the by ISO with the superior effect of the combination oxidative stress induced by ISO as evidenced by the of both agents. Based on these findings it is likely significant reduction in MDA. In addition, L-carnitine that administration of these two elements is rec- increased antioxidant capacity of the cardiac muscle ommended for patient at risk of MI as they exert evidenced by increased GSPxase and GSRase activity. a cardio protective effect by stabilizing the myocardial In another model of MI induced by amethopterin, membrane. The membrane-stabilising activity may L-carnitine was found to improve the biochemical, be due to an increase in endogenous antioxidants, histopathological, and immunohistochemical alter- which may increase the myocardial antioxidant re- ations followed amethopterin administration [31]. serve and strengthen the defence mechanism(s) in Previously L-carnitine administration was reported the myocardium. to have a cardioprotective role in cardiomyopathy and prevention of MI [14]. Lee et al. [15] found that References L-carnitine supplementation increased antioxidant 1. Ahmed MM, Ibrahim ZS, Alkafafy M, et al. L-carnitine enzymes, improved lipid profile and decreased oxida- protects against testicular dysfunction caused by gamma irradiation in mice. Acta Histochem. 2014; 116(6): tive stress in patients with coronary artery diseases. 1046–1055, doi: 10.1016/j.acthis.2014.04.010, indexed Vitamin E in the present study was also found to in Pubmed: 24925768. ameliorate degenerative changes and other cardiac 2. Alshiek JA, Dayan L, Asleh R, et al. Anti-oxidative treatment fibre alterations induced by ISO administration. This with vitamin E improves peripheral vascular function in pa- protection was observed at both light and electron tients with diabetes mellitus and Haptoglobin 2-2 genotype: A double-blinded cross-over study. Diabetes Res Clin Pract. microscopic level. In addition, vitamin E reduced ISO- 2017; 131: 200–207, doi: 10.1016/j.diabres.2017.06.026, -induced oxidative stress as evidenced by reduction indexed in Pubmed: 28759833. in MDA. Vitamin E also increased cardiac muscle 3. Bancroft JD, Gamble M. Theory and practice of histological antioxidant capacity evidenced by increased GSPxase techniques. 6th ed. Churchill Livingstone, Elsevier, China. 2008. 4. Boaz M, Smetana S, Weinstein T, et al. Secondary preven- and GSRase activity. The anti-inflammatory effect of tion with antioxidants of cardiovascular disease in endstage vitamin E was evident as it significantly reduced the renal disease (SPACE): randomised placebo-controlled elevated inflammatory cytokines IL-6 and TNF-a in the trial. Lancet. 2000; 356(9237): 1213–1218, indexed in cardiac tissue after administration of ISO. These find- Pubmed: 11072938. ing were supported by the study of Boaz et al. [4] and 5. De Bono DP, Boon NA. Diseases of the cardiovascular system. In: Edwards CRW, Bouchier IAS, eds. Davidson’s Lee et al. [16], who stated that vitamin E was known Principles of Practice and Medicine. Hong Kong: Churchill to be a potent antioxidant with proven effectiveness Livingstone. 1992: 249–340. in prevention of cardiovascular diseases. It has been 6. Deore AB, Vinayak D. Sapakal and Naikwade NS: Antioxi- proposed that the cardio protective effect of L-car- dant and hepatoprotective activity of Garcinia indica Linn fruit rind. J Comprehensive Pharmacy. 6.08. nitine and vitamin E is due to anti-inflammatory and 7. Dhivya V, Priya LB, Chirayil HT, et al. Piperine modulates antioxidant activities on coronary endothelial function isoproterenol induced myocardial ischemia through an- and vasoreactivity [34]. This could explain the better tioxidant and anti-dyslipidemic effect in male Wistar rats. effect induced by the combination of L-carnitine and Biomed Pharmacother. 2017; 87: 705–713, doi: 10.1016/j. biopha.2017.01.002, indexed in Pubmed: 28088738. vitamin E in protection against ISO-induced cardiac 8. Gayathri K, Jayachandran KS, Vasanthi HR, et al. Cardiopro- injury that was documented in this study. tective effect of lemon grass as evidenced by biochemical Among the limitations of this study was the in- and histopathological changes in experimentally induced ability to assess changes in gene expression of the cardiotoxicity. Hum Exp Toxicol. 2011; 30(8): 1073–1082, antioxidant profile in the cardiac muscle in all stud- doi: 10.1177/0960327110386391, indexed in Pub- med: 20952451. ied groups in order to prove the mechanism behind 9. Goyal S, Siddiqui MK, Siddiqui KM, et al. Cardioprotective the augmentation of action between vitamin E and effect of ‘Khamira Abresham Hakim Arshad Wala’ a unani L-carnitine. formulation in isoproterenol-induced myocardial necrosis in

281 Folia Morphol., 2019, Vol. 78, No. 2

rats. Exp Toxicol Pathol. 2010; 62(1): 61–74, doi: 10.1016/j. 23. Panda S, Kar A. Combined effects of vincristine and quercetin etp.2009.02.115, indexed in Pubmed: 19285846. in reducing isoproterenol-induced cardiac necrosis in rats. 10. Goyal SN, Arora S, Sharma AK, et al. Preventive effect of Cardiovasc Toxicol. 2015; 15(4): 291–299, doi: 10.1007/ crocin of Crocus sativus on hemodynamic, biochemical, s12012-014-9291-x, indexed in Pubmed: 25537132. histopathological and ultrastuctural alterations in isopro- 24. Panda V, Kamble S, Desai Y, et al. Antioxidant and cardioprotec- terenol-induced cardiotoxicity in rats. Phytomedicine. 2010; tive effects of Garcinia indica (kokoberry), an Indian super fruit 17(3-4): 227–232, doi: 10.1016/j.phymed.2009.08.009, in isoproterenol induced myocardial necrosis in rats. J Berry indexed in Pubmed: 19747807. Res. 2014; 4(3): 159–174, doi: 10.3233/jbr-140078. 11. Goyal SN, Sharma C, Mahajan UB, et al. Protective 25. Salama A, Kasem S, Tousson E, et al. L-carnitine and vitamin E effects of cardamom in isoproterenol-induced myo- alleviate reproductive toxicity caused by triton WR 1339 in cardial infarction in rats. Int J Mol Sci. 2015; 16(11): male albino rats. Toxicol Ind Health. 2015; 31(5): 467–474. 27457–27469, doi: 10.3390/ijms161126040, indexed in 26. Schönekess B, Lopaschuk G. The effects of carnitine on Pubmed: 26593900. myocardial carbohydrate metabolism. The Carnitine Sys- 12. Habig WH, Pabst MJ, Jakoby WB. Glutathione S-transferas- tem. A New Therapeutical Approach to Cardiovascular es. The first enzymatic step in mercapturic acid formation. Diseases, edited by De Jong JW, Ferrari R, Dordrecht, Kluwer J Biol Chem. 1974; 249(22): 7130–7139, indexed in Pub- Academic Publishers. 1995: 39–52, doi: 10.1007/978-94- med: 4436300. 011-0275-9_4. 13. Hu XX, Fu Li, Li Y, et al. The cardioprotective effect of vitamin 27. Shang R, Sun Z, Li H. Effective dosing of L-carnitine in the e (alpha-tocopherol) is strongly related to age and gender secondary prevention of cardiovascular disease: a system- in mice. PLoS One. 2015; 10(9): e0137405, doi: 10.1371/ atic review and meta-analysis. BMC Cardiovasc Disord. journal.pone.0137405, indexed in Pubmed: 26331272. 2014; 14: 88, doi: 10.1186/1471-2261-14-88, indexed in 14. Lango R, Smolenski RT, Narkiewicz M, et al. Influence of Pubmed: 25044037. L-carnitine and its derivatives on myocardial metabolism 28. Shukla SK, Sharma SB, Singh UR, et al. Eugenia jambolana and function in ischemic heart disease and during cardio- pretreatment prevents isoproterenol-induced myocardial pulmonary bypass. Cardiovasc Res. 2001; 51(1): 21–29, damage in rats: evidence from biochemical, molecular, indexed in Pubmed: 11399244. and histopathological studies. J Med Food. 2014; 17(2): 15. Lee BJ, Lin JS, Lin YC, et al. Effects of L-carnitine supple- 244–253, doi: 10.1089/jmf.2013.2795, indexed in Pub- mentation on lipid profiles in patients with coronary artery med: 24325453. disease. Lipids Health Dis. 2016; 15: 107, doi: 10.1186/ 29. Tappia PS, Hata T, Hozaima L, et al. Role of oxidative stress s12944-016-0277-5, indexed in Pubmed: 27317162. in catecholamine-induced changes in cardiac sarcolemmal 16. Lee IM, Cook N, Gaziano J, et al. Vitamin E in the pri- Ca2+ transport. Arch Biochem Biophys. 2001; 387(1): mary prevention of cardiovascular disease and cancer. 85–92, doi: 10.1006/abbi.2000.2234, indexed in Pub- The Women’s Health Study: a randomized controlled trial. med: 11368187. JAMA. 2005; 294(1): 56–65, doi: 10.1001/jama.294.1.56. 30. Thygesen K, Mair J, Katus H, et al. Acute Cardiac Care. 17. Li Li, Zhu Di, Sun X. [Cellular uptake of TPS-L-carnitine Recomendations for the use of cardiac troponin meas- synthesised as transporter-based renal targeting prodrug]. urement in acute cardiac care. Eur Heart J. 2010; 31(18): Sichuan Da Xue Xue Bao Yi Xue Ban. 2012; 43(6): 936–940, 2197–2204, doi: 10.1093/eurheartj/ehq251, indexed in indexed in Pubmed: 23387232. Pubmed: 20685679. 18. Liepinsh E, Makrecka-Kuka M, Kuka J, et al. Inhibition of L-car- 31. Tousson E, Hafez E, Zaki S, et al. The cardioprotective effects nitine biosynthesis and transport by methyl-g-butyrobetaine of L-carnitine on rat cardiac injury, apoptosis, and oxidative decreases fatty acid oxidation and protects against myocar- stress caused by amethopterin. Environ Sci Pollut Res Int. dial infarction. Br J Pharmacol. 2015; 172(5): 1319–1332, 2016; 23(20): 20600–20608, doi: 10.1007/s11356-016- doi: 10.1111/bph.13004, indexed in Pubmed: 25363063. 7220-1, indexed in Pubmed: 27464663. 19. Panteghini M. Role and importance of biochemical markers 32. Whellan D. Heart failure disease management. Implemen- in clinical cardiology. Eur Heart J. 2004; 25(14): 1187–1196, tation and outcomes. Cardiol Rev. 2005; 13(5): 231–239, doi: 10.1016/j.ehj.2004.04.026. doi: 10.1097/01.crd.0000135765.60824.2f. 20. Milman U, Blum S, Shapira C, et al. Vitamin E supplementa- 33. Wong GK, Pehora C, Crawford MW. L-carnitine reduces tion reduces cardiovascular events in a subgroup of middle- susceptibility to bupivacaine-induced cardiotoxicity: an aged individuals with both type 2 diabetes mellitus and the experimental study in rats. Can J Anaesth. 2017; 64(3): haptoglobin 2-2 genotype: a prospective double-blinded 270–279, doi: 10.1007/s12630-016-0797-5, indexed in clinical trial. Arterioscler Thromb Vasc Biol. 2008; 28(2): Pubmed: 28032304. 341–347, doi: 10.1161/ATVBAHA.107.153965, indexed in 34. Yüncü M, Bükücü N, Bayat N, et al. The effect of vitamin E Pubmed: 18032779. and L-carnitine against methotrexate-induced injury in rat 21. Moran MS. Levels of glutathione, glutathione reductase testis. Turk J Med Sci. 2015; 45(3): 517–525, indexed in and glutathione-S-transferase activity in rat lung and liver. Pubmed: 26281314. Biochim et Biophys. 1979; 582: 67–78. 35. Zambrano S, Blanca AJ, Ruiz-Armenta MV, et al. L-carnitine 22. Paglia DE, Valentine WN. Studies on the quantitative and protects against arterial hypertension-related cardiac qualitative characterization of erythrocyte glutathione fibrosis through modulation of PPAR-g expression. Bio- peroxidase. J Lab Clin Med. 1967; 70(1): 158–169, indexed chem Pharmacol. 2013; 85(7): 937–944, doi: 10.1016/j. in Pubmed: 6066618. bcp.2012.12.021, indexed in Pubmed: 23295156.

Observations of foetal heart veins draining directly into the left and right atria J.H. Kim1, O.H. Chai1, C.H. Song1, Z.W. Jin2, G. Murakami3, H. Abe4

1Department of Anatomy and Institute of Medical Sciences, Chonbuk National University Medical School, Jeonju, Korea 2Department of Anatomy, Wuxi Medical School, Jiangnan University, Wuxi, China 3Division of Internal Medicine, Jikou-kai Clinic of Home Visits, Sapporo, Japan 4Department of Anatomy, Akita University School of Medicine, Akita, Japan

[Received: 19 June 2018; Accepted: 8 August 2018]

Evaluation of semiserial sections of 14 normal hearts from human foetuses of gestational age 25–33 weeks showed that all of these hearts contained thin veins draining directly into the atria (maximum, 10 veins per heart). Of the 75 veins in these 14 hearts, 55 emptied into the right atrium and 20 into the left atrium. These veins were not accompanied by nerves, in contrast to tributaries of the great cardiac vein, and were negative for both smooth muscle actin (SMA) and CD34. However, the epithelium and venous wall of the anterior cardiac vein, the thickest of the direct draining veins, were strongly positive for SMA and CD34, respectively. In general, developing fibres in the vascular wall were positive for CD34, while the endothelium of the arteries and veins was strongly positive for the present DAKO antibody of SMA. The small cardiac vein, a thin but permanent tributary of the terminal portion of the great cardiac vein, was also positive for SMA and CD34. A few S100 protein-positive nerves were observed along both the anterior and small cardiac veins, but no nerves accompanied the direct drainage veins. These findings suggested that the latter did not develop from the early epicardiac vascular plexus but from a gulfing of the intratrabecular space or sinus of the atria. However, the immunoreactivity of the anterior cardiac vein suggests that it originated from the vascular plexus, similar to tributaries of the great cardiac vein. (Folia Morphol 2019; 78, 2: 283–289)

Key words: great cardiac vein, anterior cardiac vein, venae cordis minimae, smooth muscle actin, CD34, human foetuses

Introduction veins, drain directly into the atrial and ventricular Most cardiac veins drain into the right atrium via cavities, although the definition of the venae cordis the coronary sinus. Of these veins, the great cardiac minimae has varied among researchers (reviewed by vein is always thickest and has the largest area of von Lüdinghausen et al. [17]). Of the latter group of drainage [13]. The small cardiac vein is a tributary of veins, the anterior cardiac vein is always thickest and the great cardiac vein or the coronary sinus. These drains directly into the right atrium. The terminal to veins develop from the epicardiac vascular plexus the atrium exhibits individual variations, independent [15, 16]. In contrast, the other thin veins, including of tributaries of the great cardiac vein, or one of the the venae cordis minimae or the smallest cardiac branches joins the peripheral portion of the great

Address for correspondence: J.H. Kim, BS, PhD, Department of Anatomy, Chonbuk National University Medical School, 20 Geunji-ro, Deokjin-gu, Jeonju, 54907, Republic of Korea, tel: 82-63-270-3097, fax: 82-63-274-9880, e-mail: [email protected]

283 Folia Morphol., 2019, Vol. 78, No. 2

cardiac vein [9, 18]. These veins, which are independent of the coronary sinus, are believed to develop from the embryonic intertrabecular spaces along the internal aspects of the developing cardiac walls [3, 4]. The intertrabecular spaces correspond to the intramural venous sinuses or tunnels in adults [12, 18]. Alternatively, Mechanik [8] stated that all cardiac veins may develop from the epicardiac vascular plexus, although no line- drawings or photographic evidence was presented. The present study was therefore designed to determine whether foetal thin veins drain directly into the left and right atria. For comparison, the anterior cardiac vein and small cardiac vein were visualised in late-stage foetuses.

Materials and methods The study was performed in accordance with the provisions of the Declaration of Helsinki 1995 (as revised in 2013). Sagittal sections were obtained from 14 normal hearts of human foetuses of gestational age 25–33 weeks (crown-rump length [CRL], 200–280 mm). These foetuses were parts of a collection in the Department of Anatomy of Akita University, Akita, Japan. The foetuses had been donated by their families to the Department from 1975 to 1985 and had been preserved in 10% w/w neutral formalin solution for more than 30 years. The available data were limited to the date of donation and the gestational age; family name, the names of the obstetrician and hospital and the reason for abor- tion were not available. The use of these samples for research was approved by the Akita University Ethics Committee (No. 1428). From 200 to 300 semiserial sections, 5 microns thick at 50–100 micron intervals, covering the entire heart, were prepared from each sample. One of every 5–10 sections was stained with haematoxylin and eosin (H&E), whereas the other sections were used for immunohistochemistry. Antibodies used were 1) mouse monoclonal Figure 1. A–C. Topographical anatomy of veins directly draining into the atria. A heart from a foetus of crown-rump length (CRL) anti-human CD34 (dilution, 1:100; Dako, Glostrup, Den- 275 mm. Panel A is 1.2 mm medial to panel B. Panel A contains mark; IR632); 2) mouse monoclonal anti-human alpha a wide sinus of the right atrium (RA). Panel B contains a large gulf- smooth muscle actin (SMA; 1:100; Dako M0851); and -like sinus of the left atrium (LA), with the latter becoming narrower in panel C. The squares in panels A–C are shown at higher magni- 3) rabbit polyclonal anti-human S100 protein (dilution, fication in Figures 2 and 3. All panels were prepared at the same 1:100; Dako N1573). The sections were subsequently magnification (scale bar in panel A, 1 mm). An insert (line-drawing) incubated for 30 min in Histofine Simple Stain Max-PO between panels B and C exhibits one of sagittal sectional planes used in the present study: this is similar or corresponds to sectional (Nichirei, Tokyo, Japan) for the diaminobenzidine (DAB) planes for panels A and C. However, the heart morphology in the reaction with horseradish peroxidase (HRP) and coun- insert (posterior view) is not foetal but adult-like for simplification; terstained with haematoxylin. Sections were monitored AO — ascending aorta; CS — coronary sinus; FO valve — valve of the foramen ovale; LA — left atrium; LSPV — left superior pulmo- and photographed with a Nikon Eclipse 80, whereas nary vein; RA — right atrium; RAU — right auricle; RCA — right photos at ultra-low magnification (objective lens less coronary artery; RV — right ventricle.

284 J.H. Kim et al., Foetal heart veins draining into the atria

Figure 2. A–H. A vein draining directly into the right atrium. A heart from a foetus of crown-rump length (CRL) 275 mm (the same specimen as in Fig. 1). Sections were stained with antibodies to S100 (A), CD34 (C) and smooth muscle actin (SMA) (E). Intervals between panels are 0.5 mm. Panel D is a higher magnification view of a square in Figure 1A. The arrowhead and star in each panel display the same site of a vein or sinus, respectively. An opening to the right atrium (RA) is wide in panels D and E, but the peripheral part (i.e., a directly draining vein) is narrow in the myocardium (A). Panels F–H are higher magnification views of veins and a thin artery indicated in panel E: SMA-positive structures are limited to the endothelium (F, G). A part of endothelium of the direct drainage vein also expresses reactivity of SMA (arrows in panel H). Panels A–E or F–H were prepared at the same magnification (scale bars: 1 mm in panel A: 0.1 mm in panel F); RCA — right coronary artery; RV — right ventricle. than ×2) were taken using a high-grade flat scanner showed 75 veins draining directly into the atria. with translucent illumination (Epson scanner GTX970). All hearts were positive for direct drainage, with Although CD34 is a marker of developing blood a maximum of 10 veins per heart. Of these 75 veins, vessels (reviewed by Chang et al. [2]), this study aimed 55 emptied into the right atrium and 20 into the to identify fibrous structures in and around the vascular left atrium, with some hearts containing both right wall [1, 6]. The antibody to SMA used in this study (Dako and left terminals (Figs. 1–3). These veins tended to M0851) stained all smooth muscle cells, but in this be found near the auricle or the atrial septum and study, it was used to detect venous endothelium [5, 10]. few were seen near orifices of the caval veins and pulmonary veins. The atrial openings of the direct Results draining veins were narrow; however, when traced Semiserial sagittal sections at 50–100 micron in- along semiserial sections, they became wider and tervals were obtained from 14 hearts of gestational continuous with an intratrabecular sinus or space age 25–33 weeks. H&E staining of these sections of the atrial wall (Figs. 1B, C; 3A, C). During a wavy

285 Folia Morphol., 2019, Vol. 78, No. 2

Figure 3. A–G. A vein draining directly into the left atrium. A heart from a foetus of crown-rump length (CRL) 275 mm (the same specimen as in Fig. 1). Sections were stained with antibodies to S100 (A, E), CD34 (F) and smooth muscle actin (SMA) (G). Intervals between panels are 0.35 mm (A–B), 0.2 mm (B–C, E–F), 0.05 mm (C–D), and 0.1 mm (D–E, F–G). Panels A and C are higher magnification views of the squares in Figures 1B and C, respectively. The star in each panel displays the same site of a vein or sinus. An opening of the sinus to the left atrium (LA) is wide in panel A, but becomes narrower in panel C and the peripheral part (i.e., a directly draining vein) is very thin in panel G (arrows). The arrowheads in panels F and G indicate another directly draining vein. All panels were prepared at the same magnification (scale bar in panel A, 1 mm)

course less than 1 mm peripheral to the terminal, observed communicating between the anterior and the direct draining veins became much thinner and great cardiac veins in the present sections although disappeared within the myocardium (Figs. 2A; 3G). there was a limitation of observations using semise- We did not identify direct draining veins to the left rial sections. or right ventricle in the present sections. All of the samples were immunohistochemically The anterior cardiac vein ended in a deep in- positive for S100 protein. S100-positive nerves were tratrabecular space or sinus (Fig. 4), while the short always present along the thin branches of cardiac cardiac vein drained into the coronary sinus (Fig. 5). arteries, with a few nerves also present along the The anterior cardiac vein was observed near the tributaries of the epicardiac and intramuscular veins. peripheral course of the great cardiac vein. In one The directly draining veins were not accompanied specimen, a branch of the great cardiac vein emptied by nerves within a distance of 50 microns from the directly into the right atrium near the terminal of wall. In 8 of the 14 heart specimens tributaries of the the anterior cardiac vein (Fig. 4G–I). No veins were great cardiac vein were weakly and non-specifically

286 J.H. Kim et al., Foetal heart veins draining into the atria

Figure 4. A–I. An anterior cardiac vein draining into the right atrium. A heart from a foetus of crown-rump length (CRL) 250 mm. Panel A displays the topographical anatomy, including the anterior cardiac vein (ACV). Sections were stained with antibodies to S100 (F), CD34 (D) and smooth muscle actin (SMA) (B). Panel E is a higher magnification view of the square in panel A. Intervals between panels are 0.05 mm (B–C, C–D, D–E, E–F), 0.25 mm (F–G), 0.3 mm (G–H) and 0.2 mm (H–I), respectively. Panels G–I show a branch (star) of the great cardiac vein (GCV) draining directly into the right atrium (RA) near the opening of the anterior cardiac vein. Panels B–I were prepared at the same magnification (scale bars in panels A and B: 1 mm); RAU — right auricle; RCA — right coronary artery; RV — right ventricle.

positive for CD34 and SMA, possibly due to long numbers of CD34-positive fibres were observed in the preservation of these specimens in formalin solution. venous walls of these two veins (Figs. 4D; 5F). Figures 2–5 show some of our observations in In spite of the present long preservation, the 6 of these hearts. Direct draining veins were negative positive reaction of SMA in the venous endothelia for SMA (Figs. 2E; 3G) except for a limited short part was just same as in our previous studies using expressing reactivity of SMA (Fig. 2H). In contrast, the short-preserved foetuses [10]. In addition, SMA expres- positive reaction of the endothelia was seen in the other sion was seen also in media of thick arteries (e.g., the veins in the myocardium (Fig. 2G) as well as tributaries right coronary artery; Figs. 2E; 4B). The endocardium of the great cardiac vein (e.g., the middle cardiac vein; and a thin tissue layer beneath the endocardium were Fig. 5B). No or few CD34-positive fibres were seen in, negative or weakly positive for both SMA and CD34. along and around the walls of the directly draining The myocardium contained a meshwork of abundant veins (Figs. 2C; 3F), in contrast to tributaries of the great CD34-positive fibres. Beneath the endocardium, a thick cardiac vein (e.g., the middle cardiac vein; Fig. 5E). The CD34-positive plate was seen, possibly corresponding to endothelia of the small (Fig. 5B) and anterior (Fig. 4B) a border between the epicardium and myocardium (Fig. cardiac veins were immunohistochemically positive for 2C; 3F). The smooth muscle layer of the coronary artery SMA expression, except at the terminal portion near had not yet been established (e.g., the right coronary the coronary sinus or right atrium. Moreover, abundant artery; Figs. 2E; 4B).

287 Folia Morphol., 2019, Vol. 78, No. 2

Figure 5. A small cardiac vein draining into the coronary sinus. A heart from a foetus of crown-rump length (CRL) 265 mm. Panel A displays the topographical anatomy, including the short cardiac veins (SCV). Sections were stained with antibodies to S100 (D), CD34 (F) and smooth muscle actin (SMA) (B). Panel E is a higher magnification view of the square in panel A. Intervals between panels are 0.05 mm (B–C, C–D, D–E, E–F). The immunoreactivity of the middle cardiac vein (MCV) to anti-SMA and anti-CD34 was typical of veins. Panels B–F were prepared at the same magnification (scale bars in panels A and B, 1 mm); CS — coronary sinus; RA — right atrium; RCA — right coronary artery; RV — right ventricle.

Discussion tion may make the gulf even narrower. Notably, the Veins originating from the left ventricular wall thin vein (i.e., a developing direct drainage vein) “cross” the great cardiac vein superficially without was negative for both CD34 and SMA, in contrast to any communication and empty directly into the tributaries of the great cardiac vein. Because of the adjacent left atrium [18]. These morphologic find- thinness and short course of the direct veins, they ings strongly suggest that the origin of these direct might not connect with the doubly negative veins drainage veins differ from the origin of the great at least in the late stage foetuses. This negativity cardiac vein and its tributaries. The present study suggests that the vein was immature and developed demonstrated that, in late-stage human foetuses, much later than the tributaries of the great cardiac a gulf-like intertrabecular sinus or space in the vein. It seemed unlikely that these directly drain- atrium narrowed at the margins and continued to ing veins had developed from the early epicardiac a thin vein originating from the myocardium. The vascular plexus connecting to the coronary sinus. In developing myocardium and possibly its contrac- addition, although the cardiac vein was not always

288 J.H. Kim et al., Foetal heart veins draining into the atria

accompanied by nerves, the directly draining veins 2. Chang H, Cho KHo, Hayashi S, et al. Site- and stage- were distant from nerve twigs in the myocardium. dependent differences in vascular density of the human People may suspect that the present direct drain- fetal brain. Childs Nerv Syst. 2014; 30(3): 399–409, age veins were likely to be peripheral parts of the doi: 10.1007/s00381-013-2272-8, indexed in Pub- med: 24005801. intertrabecular sinus or spaces. However, the di- 3. Grant RT. Development of the cardiac coronary vessels rect veins reached the myocardium according to in the rabbit. Heart. 1926; 13: 261–271. a mesh-like configuration of CD34-positive fibres. 4. Grant RT, Regnier M. The comparative anatomy of the According to interpretations by von Lüdinghausen cardiac coronary vessels. Heart. 1926; 13: 285–317. 5. Hayashi S, Murakami G, Ohtsuka A, et al. Connective [18], the direct veins (his smallest cardiac vessels) tissue configuration in the human liver hilar region with represent the primitive, embryonic sinusoidal circula- special reference to the liver capsule and vascular sheath. tion. However, we did not found such a morphology J Hep Bil Pancr Surg. 2008; 15(6): 640–647, doi: 10.1007/ in previous studies of the right and left atria using s00534-008-1336-8, indexed in Pubmed: 18987936. 6. Katori Y, Kiyokawa H, Kawase T, et al. CD34-positive many midterm foetuses [7, 11]. Other than primitive primitive vessels and other structures in human fetuses: and embryonic, the intertrabecular sinus might be an immunohistochemical study. Acta Otolaryngol. 2011; involved into the myocardium after midterm. We did 131(10): 1086–1090, doi: 10.3109/00016489.2011.590 not find SMA-positive endothelium in most parts of 152, indexed in Pubmed: 21651317. the direct vein as well as the atrial endocardium. As 7. Kim JiH, Hwang SiE, Rodríguez-Vázquez JF, et al. Upper terminal of the inferior vena cava and development Robb [14] concluded, the endothelia of the direct of the heart atriums: a study using human embryos. vein seemed to be originated and extended from Anat Cell Biol. 2014; 47(4): 236–243, doi: 10.5115/ the atrial endocardium. acb.2014.47.4.236, indexed in Pubmed: 25548721. The anterior cardiac veins, usually 2 or 3 but 8. Mechanik N. Das Venensystem der Herzwände. Zeitschrift für Anatomie und Entwicklungsgeschichte. 1934; 103(6): sometimes as many as 5 in number, drain directly 813–843, doi: 10.1007/bf02118755. into the (right) atrium. The epicardial course of the 9. Mierzwa J, Kozielec T. Variation of the anterior cardiac anterior vein is long, crossing the right coronary veins. Folia Morphol. 1975; 34(2): 125–133, indexed in artery in the atrioventricular sulcus. In contrast, the Pubmed: 1080721. course of the direct veins from the myocardium to 10. Miyake N, Hayashi S, Kawase T, et al. Fetal anatomy of the human carotid sheath and structures in and the atria was short. Moreover, the anterior veins around it. Anat Rec (Hoboken). 2010; 293(3): 438–445, were positive for both CD34 and SMA. Therefore, doi: 10.1002/ar.21089, indexed in Pubmed: 20169562. like tributaries of the great cardiac vein, the anterior 11. Naito M, Yu HC, Kim JiH, et al. Topographic anatomy cardiac veins may develop from the epicardiac vas- of the fetal inferior vena cava, coronary sinus, and pulmonary veins: Variations in Chiari’s network. Clin Anat. cular plexus. One of the anterior cardiac veins often 2015; 28(5): 627–637, doi: 10.1002/ca.22489, indexed joins the great cardiac vein in variable combinations in Pubmed: 25425533. [9]. We found that a branch of the great cardiac vein 12. Ortale JR, Marquez CQ. Anatomy of the intramural emptied into a deep intertrabecular sinus of the right venous sinuses of the right atrium and their tributaries. Surg Radiol Anat. 1998; 20(1): 23–29, indexed in atrium near the terminal of the anterior cardiac vein Pubmed: 9574485. (Fig. 4G–I). An anastomosis between the anterior 13. Ortale JR, Gabriel EA, Iost C, et al. The anatomy of the and great cardiac veins may develop secondarily coronary sinus and its tributaries. Surg Radiol Anat. 2001; through the involvement of the sinus with the vein. 23(1): 15–21, indexed in Pubmed: 11370136. 14. Robb JS. Comparative basic cardiology. Grune & Stratton, New York. 1965: 123–140. Acknowledgements 15. Steding G, Seidl W. Cardiovaskuläres System. In Human This study was supported in part by a Grant-in-Aid Embryologie. Hinrichsen KV (ed). Springer, Berlin. 1990: for Scientific Research (JSPS KAKENHI No. 16K08435) 205–294. from the Ministry of Education, Culture, Sports, 16. Vobořil Z, Schiebler TH. Über die Entwicklung der Gefäßversorgung des Rattenherzens. Z Anat EntwGesch. Science and Technology in Japan. 1969; 129(1): 24–40, doi: 10.1007/bf00521953. 17. von Lüdinghausen M, Ohmachi N, Besch S, et al. References Atrial veins of the human heart. Clin Anat. 1995; 8(3): 1. Abe Si, Suzuki M, Cho KHo, et al. CD34-positive developing 169–189, doi: 10.1002/ca.980080302, indexed in Pub- vessels and other structures in human fetuses: an immu- med: 7606591. nohistochemical study. Surg Radiol Anat. 2011; 33(10): 18. von Lüdinghausen M. The venous drainage of the human 919–927, doi: 10.1007/s00276-011-0854-2, indexed in myocardium. Adv Anat Embryol Cell Biol. 2003; 168(I-VIII): Pubmed: 21789504. 1–104, indexed in Pubmed: 12645157.

Renal vascularisation anomalies in the Polish population. Coexistence of arterial and venous anomalies in the vascular pedicle of the kidney H. Sośnik, K. Sośnik

Department of Pathomorphology, Regional Specialist Hospital, Wrocław, Poland

[Received: 22 April 2018; Accepted: 19 July 2018]

Background: The aim of the study was to determine the coexistence of arterial and venous anomalies in the vascular pedicle of the kidney in the Polish population. Materials and methods: The study group comprised 550 corpses, including 281 male (mean age 52 ± 22 years), and 269 female corpses (mean age 56.4 ± ± 23 years) (p = 0.02). The vascular bundle was removed together with the kidney, “en bloc”; the vessels were subject to radiological contrasting and preparation. The obtained results were subject to statistical analysis (the exact Fisher test, c2 test of independence for cross-tabulation 2 × 2 tables, and the odds ratio with the confidence interval at a level of 0.95, calculated on the basis of the contingency dependence). Statistica 12 and Microsoft Excel were used for calculations. Results: Variations in the structure of the renal venous system were observed in 33.8%; insignificantly more often in male (35.9%), as compared to female patients (31.6%). Deviations in the structure of the renal arteries were observed significantly more often in patients with renal venous system developmental disorders (p = 0.0071). In patients with a normal renal venous system, arterial kidney structure deviations were observed in 34.9% of cases, significantly more often in male (40%), as compared to female (29.7%) patients (p = 0.043). On the other hand, in venous system pathologies, the above-mentioned deviations were observed in 46.8% of cases (p = 0.002). Amongst developmental anomalies of the venous system, right-sided venous excess was observed in 20.4% of cases, including 48% of renal artery developmental pathologies, significantly more often in male (61.3%), as compared to female (32%) patients (p = 0.002). Conclusions: Renal artery anomalies of the kidney vascular pedicle are significantly correlated with the coexistence of venous system variations. This is especially true for male patients, which favours female kidneys for transplantation. (Folia Morphol 2019; 78, 2: 290–296)

Key words: renal vascular pedicle, arterial and venous anomalies, correlations

INTRODUCTION aneurysms [2, 26], during retroperitoneal procedures Variations in the structure of the renal vascular [3, 12], in the treatment of renovascular hyperten- pedicle occur relatively frequently [11, 29], which sion [5], as well as during kidney transplantation should be considered in surgical treatment of aortic [4, 6, 7, 14, 20].

Address for correspondence: H. Sośnik, MD, PhD, Department of Pathomorphology, Regional Specialist Hospital, ul. St. Jaracza 82B/4, 50–305 Wrocław, Poland, tel: +48 71 791 41 29, e-mail: [email protected]

290 H. Sośnik, K. Sośnik, Coexistence of renal arterial and venous anomalies

In studies performed on a large sectional material, excess arteries ranged between 19.1% [23] and 32.3% [21]. In our own material the above-mentioned amounted to 38.3% [27]. The percentage of variations in the structure of the venous system was also high. Pick and Anson [21] observed 14.4% of such cases, while in our previous investigations the above- mentioned amounted to 33.8% [28]. Developmental disorders of the renal arteries are often accompanied by deviations of the venous structure [16, 30]. It is suggested that both anomalies go hand-in-hand [3, 21]. We decided to investigate the problem in the Figure 1. Vasculorenal sample in the antero-posterior position. Polish population. Two left renal arteries (arrows) running parallel with left renal vein (arrowhead) to the left kidney hilum. Three renal arteries (arrows) and two renal veins (arrowheads) on the right side. Two accessory MATERIALS AND METHODS renal arteries running precaval and parallel with an accessory renal The study group comprised 550 deceased pa- vein to the inferior pole of right kidney. Right renal artery and vein in normal position. tients, including 281 men, aged between 0.2 and 91 years (mean age 52 ± 22.6 years), and 269 women, aged between 0.1 and 92 years (mean age 56.4 ± ± 23 years) (p = 0.02) (approval of the Bioethical Committee No. 2/ BOPD/2017 DIL). The renovascular samples were removed from the deceased bodies, “en bloc”. In 377 (68.5%) cases post- mortal aortonephrography was performed, while in 145 (26.4%) — cavonephrography. After 2-week fixation in a 10% formalin solution, each prepared sample was subjected to description (Figs. 1, 2). The methodology was mentioned in our previous studies [27, 28]. The obtained statistical data was subjected to analysis by means of the exact Fisher test, and c2 test of independence for cross-tabulation 2 × 2 tables. The odds ratio (OR) with a confidence interval (CI) at a level of 0.95 Figure 2. Vasculorenal sample in the antero-posterior position. On the left side two renal arteries (arrows). One of them wrapping was calculated in the case of significant contingency around the left renal vein (arrowhead), running to the left kidney 2 (dependence). The c test was used to verify the uniform hilum. Normal arterio-venous position on the right side. distribution of subgroups of bodies with different renal vessel anomalies. Statistica 12 software and Microsoft Excel were used for calculation.

RESULTS Results are presented in two stages: — the occurrence of arterial variations in relation to the number of renal veins (1100 kidneys); — occurrence of variations in the structure of arterial system in relation to particular types of developmental disorders of the venous system (7 types in our material; 550 cases). The percentage of excess arteries significantly -in creased with the number of renal veins (Fig. 3, Table 1). Figure 3. Percentage of supernumerary arteries depending on Considering the side of the body, the number of arterial number of veins.

291 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. Arterial-venous coincidence in the vascular bundle of kidney (1100 kidneys) Number of Number of arteries Total Supernumerary arteries veins 1 2 3 4 Sum Percentage 1 738 172 9 4 923 185 20.0% 2 103 48 10 2 163 60 36.8% 3 3 4 4 1 12 9 75.0% 4 0 0 1 1 2 2 100.0% Total 844 224 24 8 1100 256 23.3%

P-values: 1–2: 0.0000, 1–3: 0.0000, 1–4: 0.005, 2–3: 0.009, 2–4: 0.067, 3–4: 0.42

Table 2. Arterial-venous coincidence in the vascular bundle of kidney depending on the side of body (n = 550) Right kidney P Left kidney Number Number of Number of Per Per cent of super- Number of Number of Number of Per Per cent of super of veins arteries kidneys cent numerary arteries veins arteries kidney cent numerary arteries 1 1 341 82.8% 1 1 397 77.7% 2 66 16.0% 2 106 20.7% 17.2% 0.056 22.3% 3 4 1.0% 3 5 1.0% 4 1 0.2% 4 3 0.6% 2 1 85 67.5% 2 1 18 50.0% 2 34 27.0% 2 14 38.9% 32.5% 0.055 50.0% 3 6 4.8% 3 4 11.1% 4 1 0.8% 4 0 0.0% 3 1 3 30.0% 3 1 1 33.3% 2 3 30.0% 2 1 33.3% 70.0% 0.71 66.7% 3 3 30.0% 3 1 33.3% 4 1 10.0% 4 0 0.0% 4 1 0 0.0% 4 1 0 – 2 0 0.0% 2 0 – 100.0% – – 3 1 50.0% 3 0 – 4 1 50.0% 4 0 –

variations was significantly greater on the left side, es- In patients with “normal” venous system structure pecially in kidneys with one or two renal veins (Table 2). (both-sided, single vein), 34.9% of arterial kidney Gender also had significantly influence on the oc- structure variations were observed, significantly more currence of arterial system deviations, in relation to often in male (40%), as compared to female patients the number of renal veins, although only in kidneys (29.89%; p = 0.043). with a single renal vein. The above-mentioned ratio Rows 2–7 of Table 6, present the particular types was lower in female (14.8%), as compared to male of venous system structure abnormalities. Amongst patients (25.3%) (p = 0.0001) (Table 3). the above-mentioned, isolated right-sided venous In male patients the percentage of excess arteries excess (112/550; 20.4%) was observed significantly was significantly greater on the left side in the case more often (p = 0.005), including 48.2% of arterial of kidneys with one or two renal veins (Table 4). Such system variation, significantly more often in men differences were not seen in women (Table 5). (38/62; 61.3%), as compared to women (16/50; 32%; Amongst the 7 types of venous renal system struc- p = 0.002). Amongst the remaining 74 venous sys- ture variations, considering 550 bodies, 214 were tem, structure variations (13.5%), the retroaortic diagnosed with arterial kidneys structure variations course of the left kidney vein accounted for 31.1%, (38.9%), significantly more often in men (46.6%), as while the periaortic venous ring — 28.4%. compared to women (30.9%; p = 0.0002).

292 H. Sośnik, K. Sośnik, Coexistence of renal arterial and venous anomalies

Table 3. Arterial-venous coincidence in the vascular bundle of kidney depending on the sex of body (n = 550) Male (n = 281) P Female (n = 269) Number Number of Number of Per Per cent of super- Number Number of Number of Per Per cent of super- of veins arteries kidneys cent numerary arteries of veins arteries kidneys cent numerary arteries 1 1 345 74.7% 1 1 393 85.2% 2 107 23.2% 2 65 14.1% 25.3% 0.0001 14.8% 3 6 1.3% 3 3 0.7% 4 4 0.9% 4 0 0.0% 2 1 54 59.3% 2 1 49 68.1% 2 30 33.0% 2 18 25.0% 40.7% 0.163 31.9% 3 7 7.7% 3 3 4.2% 4 0 0.0% 4 2 2.8% 3 1 2 28.6% 3 1 1 20.0% 2 2 28.6% 2 2 40.0% 71.4% 0.64 80.0% 3 2 28.6% 3 2 40.0% 4 1 14.3% 4 0 0.0% 4 1 0 0.0% 4 1 0 – 2 0 0.0% 2 0 – 100.0% – – 3 1 50.0% 3 0 – 4 1 50.0% 4 0 –

Table 4. Arterial-venous coincidence in the vascular bundle of kidney depending on the side of male body (n = 281) Right kidney P Left kidney Number Number Number Per Per cent of super- Number Number of Number Per Per cent of super- of veins of arteries of kidneys cent numerary arteries of veins arteries of kidneys cent numerary arteries 1 1 162 79.0% 1 1 183 71.2% 2 40 19.5% 2 67 26.1% 21.0% 0.035 28.8% 3 2 1.0% 3 4 1.6% 4 1 0.5% 4 3 1.2% 2 1 45 65.2% 2 1 9 40.9% 2 20 29.0% 2 10 45.5% 34.8% 0.039 59.1% 3 4 5.8% 3 3 13.6% 4 0 0.0% 4 0 0.0% 3 1 2 40.0% 3 1 0 0.0% 2 1 20.0% 2 1 50.0% 60.0% 0.67 100.0% 3 1 20.0% 3 1 50.0% 4 1 20.0% 4 0 0.0% 4 1 0 0.0% 4 1 0 – 2 0 0.0% 2 0 – 100.0% – – 3 1 50.0% 3 0 – 4 1 50.0% 4 0 –

Generalising our results, in the case of both-sided (OR: 1.6399, 95% CI :1.1444–2.500, p = 0.007). The single venous system presence, arterial variations above mentioned occurred more often in men, as were observed in 34.9% of cases, while in the case of compared to women (OR: 1.566, 95% CI: 1.0129– venous system developmental disorders — in 46.8% –2.413, p = 0.0436).

293 Folia Morphol., 2019, Vol. 78, No. 2

Table 5. Arterial-venous coincidence in the vascular bundle of kidney depending on the side of female body (n = 269) Right kidney P Left kidney Number Number of Number of Per Per cent of super- Number Number of Number Per Per cent of super of veins arteries kidneys cent numerary arteries of veins arteries of kidneys cent numerary arteries 1 1 179 86.5% 1 1 214 84.3% 2 26 12.6% 2 39 15.4% 13.5% 0.297 15.7% 3 2 1.0% 3 1 0.4% 4 0 0.0% 4 0 0.0% 2 1 40 70.2% 2 1 9 64.3% 2 14 24.6% 2 4 28.6% 29.8% 0.45 35.7% 3 2 3.5% 3 1 7.1% 4 1 1.8% 4 0 0.0% 3 1 1 20.0% 3 1 1 100.0% 2 2 40.0% 2 0 0.0% 80.0% 0.333 0.0% 3 2 40.0% 3 0 0.0% 4 0 0.0% 4 0 0.0% 4 1 0 – 4 1 0 – 2 0 – 2 0 – – – – 3 0 – 3 0 – 4 0 – 4 0 –

Table 6. Correlation of kidney venous anomalies with other kidney arterial anomalies in 550 cadaver bodies with p-value of chi-square test between men and women No. Type of venous Venous/other kidney arterial anomalies P anomalies Men Women Total N % N%N % 1 180/72 40.00% 184/55 29.89% 364/127 34.89% 0.043*

2 62/38 61.29% 50/16 32.00% 112/54 48.21% 0.002**

3 7/3 42.85% 5/1 20.00% 12/4 33.33% 0.408

4 6/6 100.00% 6/4 66.66% 12/10 83.33% 0.121

5 13/4 30.76% 10/1 23.50% 23/5 21.73% 0.699

6 2/1 50.00% 4/2 50.00% 6/3 50.00% 1.000

7 11/7 63.63% 10/4 40.00% 21/11 52.38% 0.279

Total 281/131 46.61% 269/83 30.85% 550/214 38.90% 0.0002*** *p < 0.05, **p < 0.01, ***p < 0.001

Considering the clinical aspect, we presented the presented with a precaval course, while in female percentage of additional renal artery topographic patients — 3% (Fig. 1). Apart from the above-men- variations, in relation to the veins. Considering male tioned, in 3.2% of male, and in 0.7% of female, the patients, 4.3% of right-sided additional arteries right additional artery wrapped around the renal vein.

294 H. Sośnik, K. Sośnik, Coexistence of renal arterial and venous anomalies

On the left side, the above-mentioned phenomenon to 13.5% [1, 2, 4, 8, 25]. Amongst them, the most im- was observed in 2.13% of male and 0.36% of female portant include rupture of arterial anastomoses, which (Fig. 2). might lead to transplant rejection, and even lower leg loss [18]. Bleeding was observed in 2.8% and sten- DISCUSSION otic vascular complications in 0.9% [19], 1.5% [5], and Our investigations showed significant coexistence 3.17%, [6], respectively. Vascular complications occurred of venous and arterial developmental anomalies, significantly more often in the case of a transplant considering the renal vascular bundle. Based on the with many vessels, as compared to a single renal artery performed angiographies, Baptista-Silva et al. [3] transplant. The above-mentioned was associated with observed more than one renal artery, being convinced prolonged surgery and transplant ischaemia [14, 15, 19, of finding two or more additional renal veins. The 24, 25], which had, however, no negative influence on existence of such anatomical variants should be taken its future functioning. The described complications were under consideration, in the prevention of bleeding reduced with the development of operative techniques during retroperitoneal space and abdominal aortic and acquired surgical knowledge of the operating team surgery [2, 3]. The mentioned anomalies result from [4, 9, 18]. the complicated development of the vasculorenal The mentioned problems concerning arterial com- system, which was thoroughly described in our previ- plications may also apply to venous damage compli- ous studies [27, 28]. Generally speaking, an embryo cations. Variants in their structure were observed in 14–16 mm in length (6–8 weeks of development), is 33.8%, mainly on the right side (20.4%) [28]. Pick subjected to mesonephros ascend from the pelvis to and Anson [21] noted such complications in 27.8% the lumbar area with simultaneous atrophy of lower of cases. Bleeding was observed in 25.4%, and ve- arteries and the development of novel abdominal nous anastomoses stenosis in 8.2% [19]. Veins do aortic vessels [21, 29]. At the same time, the venous not pulsate, are subject to rupture, and are difficult system develops from three successive and degenera- to distinguish from adhesions [21]. tive, cardinal veins. One may observe many transverse Considering patients operated for abdominal and anterodorsal connections between them. Thus, aortic aneurysms, knowledge of venous anomalies, the aorta and renal artery branches are connected in especially on the left side, seems very important. the network of numerous venous vessels, and blood The above-mentioned veins are often hidden in the initially flowing to the heart symmetrically, is shifted adhesions, and overlooked [2, 26]. Understanding to the right side. At the same time, the inferior caval the possible existence of particular venous anomalies, vein develops from four separate fragments. These and especially precise computed tomography image complicated processes favour the development of analysis before surgery, as well as effective surgical the mentioned anomalies. We do not know the final technique are conditions, which minimalise the oc- cause of their development. In our research, consid- currence of these complications [13]. ering a normal venous structure (both-sided, single The retroaortic course of the left renal vein was vein), arterial variants were observed in 34.9%, while observed in only 1.4% of cases [3]. Considering our in the case of developmental disorders — in 46.8%. material, these figures amounted to 12.4%. The com- Pick and Anson [21] showed the existence of a single pression of the vein by the aorta led to the clinical artery and vein on both sides in only 38.5%. diagnosis of haematuria in a 13-year-old boy [10]. For transplantation purposes, cases with one renal Considering the presented material, 4.3% male artery are desired [6]. However, due to the increased and 3% female cases showed a precaval course of demand for transplantation material and ageing popu- right-sided additional renal arteries. These cases do lation, no material is discarded. Thus, knowledge of not exclude kidney transplantation [22]. the described anomalies, as well as introduction of dif- Compression of renal veins by surrounding ad- ferent surgical techniques seems especially important ditional renal arteries was observed in 3.2% of male [7]. Transplantations with the presence of the above- and 0.74% female cases, on the right side. On the mentioned anomalies are safe, if performed by an ex- other hand, 2.13% of male and 0.36% of female cases perienced team. Fast diagnosis and proper treatment on the left side. Nathan came to similar conclusions, protect both the transplanted organ and the patient. considering a study material of 200 analysed corpses; However, vascular complications are observed in 2.55% but he observed 6% of such cases on the left side,

295 Folia Morphol., 2019, Vol. 78, No. 2

and only 1% on the right one [17]. It is generally ac- 14. Keller JE, Dolce CJ, Griffin D, et al. Maximizing the donor pool: use of right kidneys and kidneys with multiple arteries cepted that increased kidney venous system pressure for live donor transplantation. Surg Endosc. 2009; 23(10): might lead to orthostatic proteinuria. In search of the 2327–2331, doi: 10.1007/s00464-009-0330-9, indexed in cause of proteinuria one should not forget about the Pubmed: 19263162. 15. Lim YMJ, Han X, Raman L, et al. Outcome of living do- mentioned possibility. nor transplant kidneys with multiple arteries. Transplant Proc. 2016; 48(3): 848–851, doi: 10.1016/j.transproceed.2015.12.104, indexed in Pubmed: 27234750. CONCLUSIONS 16. Madhyastha S, Suresh R, Rao R. Multiple variations of renal The coexistence of renal, venous and arterial, devel- vessels and ureter. Indian J Urol. 2001; 17: 164. 17. Nathan H. Observations on aberrant renal arteries curving opmental anomalies in the vascular bundle is significant- around and compressing the renal vein; possible relation- ly correlated. It occurs especially often in men; therefore, ship to orthostatic proteinuria and to orthostatic hypertension. Circulation. 1958; 18(6): 1131–1134, indexed in the female kidney is preferred for transplantation. Pubmed: 13608842. 18. Orlic P, Vukas D, Drescik I, et al. Vascular complications after 725 kidney transplantations during 3 decades. References Transplantation Proceedings. 2003; 35(4): 1381–1384, 1. Aktas S, Boyvat F, Sevmis S, et al. Analysis of vascular com- doi: 10.1016/s0041-1345(03)00506-2. plications after renal transplantation. Transplant Proc. 2011; 19. Osman Y, Shokeir A, Ali-el-Dein B, et al. Vascular com- 43(2): 557–561, doi: 10.1016/j.transproceed.2011.01.007, plications after live donor renal transplantation: study indexed in Pubmed: 21440760. of risk factors and effects on graft and patient sur- 2. Aljabri B, MacDonald PS, Satin R, et al. Incidence of major vival. J Urol. 2003; 169(3): 859–862, doi: 10.1097/01. venous and renal anomalies relevant to aortoiliac surgery ju.0000050225.74647.5a, indexed in Pubmed: 12576799. as demonstrated by computed tomography. Ann Vasc Surg. 20. Paragi PR, Klaassen Z, Fletcher HS, et al. Vascular con- 2001; 15(6): 615–618, doi: 10.1007/s10016-001-0095-7, straints in laparoscopic renal allograft: comparative indexed in Pubmed: 11769141. analysis of multiple and single renal arteries in 976 lapa- 3. Baptista-Silva J, Veríssimo M, Castro M, et al. Anatomi- roscopic donor nephrectomies. World J Surg. 2011; 35(9): cal study of the renal veins observed during 342 living- 2159–2166, doi: 10.1007/s00268-011-1168-6, indexed in donor nephrectomies. Sao Paulo Med J. 1997; 115(3): Pubmed: 21713578. 1456–1459, doi: 10.1590/s1516-31801997000300011. 21. Pick J, Anson B. The renal vascular pedicle: an anatomical 4. Bessede T, Droupy S, Hammoudi Y, et al. Surgical prevention and study of 430 body-halves. J Urol. 1940; 44(4): 411–434, management of vascular complications of kidney transplanta- doi: 10.1016/s0022-5347(17)71289-x. tion. Transpl Int. 2012; 25(9): 994–1001, doi: 10.1111/j.1432- 22. Radolinski B, Diner EK, Ghasemian SR. Precaval right renal 2277.2012.01533.x, indexed in Pubmed: 22816523. artery during laparoscopic donor nephrectomy. Trans- 5. Bude RO, Forauer AR, Caoili EM, et al. Is it necessary to plantation. 2006; 82(11): 1554–1555, doi: 10.1097/01. study accessory arteries when screening the renal arteries tp.0000243738.44921.e2, indexed in Pubmed: 17164737. for renovascular hypertension? Radiology. 2003; 226(2): 23. Reis RH, Esenther G. Variations in the pattern of renal ves- 411–416, doi: 10.1148/radiol.2263011576, indexed in sels and their relation to the type of posterior vena cava Pubmed: 12563134. in man. Am J Anat. 1959; 104: 295–318, doi: 10.1002/ 6. Carter JT, Freise CE, McTaggart RA, et al. Laparoscopic pro- aja.1001040206, indexed in Pubmed: 14437208. curement of kidneys with multiple renal arteries is associated 24. Saidi R, Kawai T, Kennealey P, et al. Living donor kidney with increased ureteral complications in the recipient. Am transplantation with multiple arteries: recent increase J Transplant. 2005; 5(6): 1312–1318, doi: 10.1111/j.1600- in modern era of laparoscopic donor nephrectomy. 6143.2005.00859.x, indexed in Pubmed: 15888035. Arch Surg. 2009; 144(5): 472–475, doi: 10.1001/arch- 7. De Rosa P, Santangelo M, Scala A, et al. Difficult vas- surg.2009.49, indexed in Pubmed: 19451491. cular conditions in kidney transplantation. Transplant 25. Salehipour M, Salahi H, Jalaeian H, et al. Vascular compli- Proc. 2006; 38(4): 1040–1043, doi: 10.1016/j.transpro- cations following 1500 consecutive living and cadaveric ceed.2006.03.073. donor renal transplantations: a single center study. Saudi 8. Eufrásio P, Parada B, Moreira P, et al. Surgical complications J Kidney Dis Transpl. 2009; 20(4): 570–572, indexed in in 2000 renal transplants. Transplant Proc. 2011; 43(1): Pubmed: 19587495. 142–144, doi: 10.1016/j.transproceed.2010.12.009. 26. Shindo S, Kubota K, Kojima A, et al. Anomalies of infe- 9. Gentil Govantes MA, Rodriguez-Benot A, Sola E, et al. rior vena cava and left renal vein: risks in aortic surgery. Trends in kidney transplantation outcome: the Andalu- Ann Vasc Surg. 2000; 14(4): 393–396, indexed in Pub- sian Kidney Transplant Registry, 1984-2007. Transplant med: 10943793. Proc. 2009; 41(5): 1583–1585, doi: 10.1016/j.transpro- 27. Sośnik H, Sośnik K. Investigations on renal vascularisa- ceed.2009.01.094, indexed in Pubmed: 19545684. tion pathology in the Polish population. 1. Incidence 10. Gibo M, Onitsuka H. Retroaortic left renal vein with renal of multiple kidney arteries. Folia Morphol. 2017; 76(2): vein hypertension causing hematuria. Clin Imaging. 1998; 226–231, doi: 10.5603/FM.a2016.0073, indexed in Pub- 22(6): 422–424, indexed in Pubmed: 9876912. med: 28026854. 11. Gupta A, Gupta R, Singal R. Congenital variations of renal 28. Sośnik H, Sośnik K. Renal vascularization anomalies veins: embryological background and clinical implications. in the Polish population. Pol Przegl Chir. 2017; 89(6): J Clin Diagn Res. 2011; 5(Suppl 1): 1140. 26–30, doi: 10.5604/01.3001.0010.6760, indexed in Pub- 12. Hennayake S. Prone retroperitoneoscopy in treating com- med: 29335391. plex renal vascular malformations. J Pediatr Urol. 2010; 29. Watson CJE, Harper SJF. Anatomical variation and its 6(2): 192–197, doi: 10.1016/j.jpurol.2009.06.014, indexed management in transplantation. Am J Transplant. 2015; in Pubmed: 19664958. 15(6): 1459–1471, doi: 10.1111/ajt.13310, indexed in 13. Karkos CD, Bruce IA, Thomson GJ, et al. Retroaortic left Pubmed: 25981150. renal vein and its implications in abdominal aortic surgery. 30. Yahiro J, Miyoshi S. Double left renal veins and multiple right Ann Vasc Surg. 2001; 15(6): 703–708, doi: 10.1007/s10016- renal veins found in Japanese adults. Okajimas Folia Anat 001-0022-y, indexed in Pubmed: 11769156. Jpn. 1993; 70(5): 237–245, indexed in Pubmed: 8202308.

Can constant light exposure affect the thyroid gland in prepubertal male albino rats? Histological and ultrastructural study F.A. Abdel Gawad, E.A.A. El-Shaarawy, S.F. Arsanyos, T.I. Abd El-Galil, G.N. Awes

Department of Anatomy and Embryology, Faculty of Medicine, Cairo University, Cairo, Egypt

[Received: 25 April 2018; Accepted: 24 June 2018]

Background: Through scientific literature, there is evidence that light affects thyroid function in human, mice and rabbits. Constant light and sleep deprivation is also used as a form of human torture, as it has impact on cognitive performances. The present work was conducted to study the effect of constant light for short and long periods on the thyroid gland in the prepubertal male albino rats. Materials and methods: A total of 30 prepubertal male albino rats were used. The rats separated into three groups: group I (control); group II were those rats put under steady encompassing light (24 h/day, light intensity of 600 lux) for 4 weeks; and group III were the rats maintained in constant light for 3 months. The rat thyroid gland was subjected to histological and ultrastructural examination. Results: The rats exposed to light for long durations showed disturbed architecture; the follicles exhibited back to back arrangement (signs of hypertrophy with hyperplasia), lined by multiple layers of follicular cells or were lined by vacuolated cells. Few thyroid follicles exhibited cystic hyperplasia. Congested blood capillaries were demonstrated between the follicles. Conclusions: It can be concluded that the short-term exposure to constant light for 1 month had no apparent effect on thyroid gland tissues while longer exposure to light for 3 months had detrimental effects on the thyroid gland structure of male albino rats. (Folia Morphol 2019; 78, 2: 297–306)

Key words: constant light, thyroid gland histological, prepubertal rats, thyroid gland ultrastructural

INTRODUCTION thalamus. Through this nucleus directly or through The environmental 24-h light/dark cycle is involved pineal hormone melatonin coordinates the timing of in modulation of circadian rhythm in mammals. It circadian rhythms, including daily control of hormone regulates various biological activities including energy secretion [21, 22]. Melatonin is one of the most stud- metabolism, body temperature, feeding, locomotion, ied biomarkers of the human physiological response and other types of behaviour [4, 8]. to light [14]. The production of melatonin takes place Hormone secretion is sensitive to time, which is almost exclusively at night. It has given rise to the necessary for achieving biological functioning. In concept that melatonin is the chemical expression mammals, the principal circadian mechanism is lo- of darkness. Continuous exposure to light leads to cated in the suprachiasmatic nucleus of the hypo- a complete suppression of melatonin release [9, 30].

297 Folia Morphol., 2019, Vol. 78, No. 2

Melatonin supplementation inhibits thyroid growth, of constant light for short and long periods on the while pinealectomy stimulates it [3]. thyroid gland in the prepubertal male albino rats using There is evidence in the scientific literature that histological and ultrastructural examinations. light affects thyroid function in human [19], mice [2], and rabbits [40]. MATERIALS AND METHODS Vinogradova et al. [39] reported that exposure Animals to constant light shortened average rat lifespan, A total of 30 prepubertal male albino rats were increased incidence of metabolic syndrome devel- used in this study. The body weight of the prepubertal opment, accelerated spontaneous tumorigeneses rats (4 weeks old) ranged from 80 to 100 g. Animal and contribute to accelerated age [42]. Thyroid dys- care was provided by laboratory animal house unit functions contributed to development of various of Kasr Al-Ainy, Faculty of Medicine, Cairo Universi- pathological states. Studies demonstrated that thy- ty. The rats were acclimatised in the laboratory for roid hormones (TH) and its receptors play a role in a period of 2 weeks before carrying out the exper- tumorigenesis and tumour progression [7, 16]. In line iment. They were provided with ordinary rat chow with hypothesis of circadian disruption, light expo- and were housed in wire mesh cages (4 rats/cage) sure at night disrupts the daily endogenous rhythm. at a controlled temperature (24 ± 1°C), with normal Besides melatonin suppression, other mechanisms, light-dark cycle. The animals were given food and including autonomic nervous system adjustment of water ad libitum. The experiment was conducted in different endocrine organ functions, may contribute accordance with the guidelines of the committee of to adaptation to constant light. laboratory animals at Kasr El-Ainy School of Medicine. Constant light (CL) and sleep deprivation is also used as a form of human torture, as it has impact on Experimental design cognitive performances [34]. All control rats were housed in standard stainless Light at night became an essential part of modern steel cages (4 rats per cage) with galvanised iron lifestyle todays. Pollution with light, defined as excess wires. The size of the control cages was 41 × 28 × artificial lighting during night-time can have detri- × 19 cm allowing the animals to move freely. The mental effect on human health [29]. Thyroid gland custom-designed light experiment cage had a length, is specialised for production, storage and release of width, and height of 45 cm, 180 cm, and 45 cm; it was thyroid hormones thyroxine (T4) and triiodothyronine vertically divided into four equal chambers (length, (T3). T4 is the dominant hormone released from the width, and height of 45 cm, 45 cm, and 45 cm for thyroid gland, while T3 is biologically more active each chamber). A fluorescent lamp (light intensity: and originates mainly from peripheral deiodination 600 lux) was installed in every chamber and had lights of T4 [5]. Thyroid hormones play essential role in controlled by a switch that allowed free adjustment regulation of metabolic rate, regulation of growth, of the illumination time. The rats were divided into maturation of many organs and systems, including three groups. brain, skeleton, reproductive organs and energy ex- penditure [24, 37]. Group I (normal control) Rats prefer low light intensity and a well-controlled Ten prepubertal rats, 4 weeks old [27]. The rats photoperiod will certainly contribute to stable circadi- were kept under regular laboratory lighting condi- an rhythms. Animal’s activity and behaviour are also tions (12-h light–12-h dark) during the same period. influencedby environmental intense light. Intense light This was achieved by switching on the light at 7.00 conditions markedly suppress social play behaviour of in the morning and off at 7.00 in the evening. The Wistar rats, which is important for the animals’ de- rats were sacrificed according to the experimental velopment [6]. Data on the effects of CL exposure on schedule. pituitary thyroid axis are incomplete and mostly limited to determinations of serum hormone concentrations Group II (exposed to constant light for short period) [39]. Little is known about the effect of light on the Ten prepubertal rats, 4 weeks old. The rats thyroid in humans. The normal thyroid-stimulating hor- were placed into constant ambient light conditions mone (TSH) rhythm can be reset by a pulse of light [1]. (24 h/day, light intensity of 600 lux). The animals The aim of the present work was to study the effect were maintained in constant light for 4 weeks [11].

298 F.A. Abdel Gawad et al., Effect of continuous light on the thyroid gland

Group III (exposed to constant light for long period) peripheral vacuolations (Fig. 1c). Some fields showed Ten prepubertal rats, 4 weeks old. The rats dilated congested blood vessels (Fig. 1d–f). Some were placed into constant ambient light conditions of the follicles exhibited several cellular layers while (24 h/day, light intensity of 600 lux). The animals others were lined by flat cells (Fig. 1e). were maintained in constant light for 3 months [27]. Other fields confirmed widening inside the interfollicular spaces. A number of the follicles exhibited The rats of groups II (prepubertal, light exposure: vacuolation of their cytoplasm (Fig. 1f). up to 4 weeks) and III (light exposure: up to 12 week) Group III. The prepubertal rats exposed to con- nearly reaching the adulthood were exposed to light stant light for long period demonstrated micro fol- during the prepubertal period. The rats of group I licular thyroid follicles with small amount of colloid. were controls for group II and group III. The follicles were lined by tall columnar follicular cells with dark nuclei (Fig. 2a, b). Other fields showed dis- Sacrification and thyroid extraction turbed architecture; the follicles exhibited back to back To reduce variance in the physiological parameters arrangement (signs of hypertrophy with hyperplasia), due to daily rhythms, all animals were sacrificed at lined by multiple layers of follicular cells or were lined the same time point in the circadian cycle, between by vacuolated cells. Some follicles showed vacuola- 9:00 and 11:00 am. On the designated day, the rats tions of colloid (Fig. 2b, c). Few thyroid follicles were were sacrificed by the end of 4 weeks for group II, cystically dilated, some of them showed ingrowth pro- 3 months for group III while the half of the control jecting into the lumen and nearly obliterating it (Cystic group were sacrificed after 4 weeks and the rest hyperplasia) (Fig. 2c). Occasionally the thyroid follicles after 3 months. The animals were sacrificed under became adherent to each other with focal hyperpla- non stress conditions by rapid cervical decapitation. sia of the follicular cells that nearly obliterated their Their necks were opened through a ventral midline lumen (Fig. 2d, e). Congested blood capillaries were incision and the thyroid gland was rapidly dissected demonstrated between the follicles (Fig. 2a, b, c, e). out and immediately weighed. Toluidine blue semithin stained sections Methods Group I. Thyroid sections belonging to control Thyroid specimens were obtained from all groups; prepubertal rats revealed thyroid follicles lined by one one lobe was allocated for light microscopic examina- layer of cubical follicular cells with vesicular spherical tion and stained with haematoxylin and eosin (H&E) nuclei, parafollicular cells with large pale nuclei and and other stained with toluidine blue. The other lobe pale cytoplasm (Fig. 3a). was allocated for ultrastructural examination using Group II. Thyroid sections belonging to group II transmission electron microscope. rats revealed parts of thyroid follicles with increase in the height of cells lining the follicles. Most of these RESULTS cells showed cytoplasmic vacuolations with either Light microscopic results pale nuclei or pyknotic ones, whereas other follicular Haematoxylin and eosin stained sections cells were ballooned with dark nuclei (Fig. 3b). Group I. Histological examination of thyroid sec- Group III. Thyroid sections belonging to group III tions revealed thyroid follicles of different sizes; their revealed degenerated thyroid follicles containing des- cavities contained acidophilic colloid with peripheral quamated epithelial cells in their lumen. Dark nuclei, vacuolations (Fig. 1a). Close observation revealed that vacuolated cytoplasm and dark flattened elongated the follicles were lined by cubical follicular cells that nuclei of some follicular cells with colloidal vacuola- exhibited rounded nuclei. The follicles were separated tions were noticed (Fig. 3c). from each other by narrow interfollicular spaces that contained minute blood capillaries (Fig. 1b). Electron microscopic results Group II. The prepubertal rats exposed to con- Group I. Ultrastructural examination of the thy- stant light for short period demonstrated disturbed roid gland showed the follicular cells with euchromat- architecture of the thyroid lobule with large periph- ic rounded or oval nucleus and prominent nucleolus. eral and small central thyroid follicles. Most of the It was surrounded by regular nuclear membrane with follicles were filled with cracked colloid. Few follicles nuclear pores. The cytoplasm contained organelles as were filled completely with colloid and others showed mitochondria, rough endoplasmic reticulum, Golgi 299 Folia Morphol., 2019, Vol. 78, No. 2

Figure 1. Micrograph of a section of the thyroid gland: a. Control prepubertal rat (group I) showing thyroid follicles (F) of variable sizes, filled with vacuolated acidophilic colloid (C) and surrounded by interfollicular tissue containing blood vessels (BV); ×200; b. Control prepubertal rat (group I) showing thyroid follicles (F) lined by a single layer of cubical follicular cells exhibiting spherical vesicular nuclei (black arrow) and contain acidophilic colloid (C). The follicles are separated by narrow interfollicular spaces that contain blood vessels (BV); ×400; c. Prepubertal rat (group II) showing disturbed architecture of the thyroid lobule with large peripheral (F) and small central thyroid follicles (black arrow). Most of the follicles are filled with cracked colloid (c). Few follicles are filled completely with colloid (arrow head) and others show peripheral vacuolations (curved arrow). Note the parathyroid gland (p); ×200; d. Prepubertal rat (group II) showing partial disruption of the thyroid lobule in the form of scalloping of the colloid in the thyroid follicles of the peripheral part of the thyroid lobule (red curved arrow). Note the dilated congested blood vessels in the centre of the thyroid lobule (black arrow); ×200; e. Prepubertal rat (group II) showing the thyroid follicles (F). Most of them are lined by a single layer of cubical cells. Some of the follicles exhibit several cellular layers (black arrow) while others are lined by flat cells (yellow arrow). Note a dilated congested blood vessel (BV) in the interfollicular space; ×400; f. Prepubertal rat (group II) showing widening in the interfollicular spaces. Most of the follicles are lined by single layer of cubical cells. Some of the follicles exhibit several cellular layers (black arrow) while others show vacuolation in the cytoplasm of the follicular cell (thin arrow). Some of the follicles show scalloping of colloid (arrow head). Note the ruptured blood vessels (BV). ×400.

apparatus and electron dense granules. The follicular surrounded by irregular nuclear membrane. The lumen showed fine granular colloid material with mitochondria were swollen with disrupted cristae. microvilli in the free surface of the follicular cells Bundles of collagen fibres were demonstrated in (Fig. 4a, b). the stroma (Fig. 5c). Some fields showed corru- Group II. Ultrastructural examination of the thy- gated nucleus with clumping of heterochromatin. roid sections belonging to group II rats revealed the The follicular lumen contained fine granular colloid follicular cells with undulation of nuclear membrane material with almost loss of microvilli on the lumi- and margination of heterochromatin. The cytoplasm nal surface of cells (Fig. 5d). Other fields showed showed electron dense granules as compared to follicular cells with corrugated nucleus with thin electron lucent granules, dilated rough endoplas- rim of heterochromatin, dilated rough endoplasmic mic reticulum and mitochondria with rupture of reticulum and colloid droplets in their cytoplasm. their outer membrane. The luminal surface showed The luminal surface demonstrated elongated and multiple detached microvilli (Fig. 5a, b). Other fields detached microvilli (Fig. 5e). showed heterochromatic nucleus with widely dis- Group III. Ultrastructural examination of the thy- persed chromatin and prominent nucleolus. It was roid sections belonging to group III rats revealed

300 F.A. Abdel Gawad et al., Effect of continuous light on the thyroid gland

Figure 2. Micrograph of a section of the thyroid gland of a prepubertal rat of (group III) showing: a. Microfollicular thyroid follicles (F) with small amount of colloid. Note the congested blood capillaries between the follicles (BV). Some follicles show collapsed lumen (black arrow); b. Dis- turbed architecture of the gland; the follicles are lined by multiple layers of follicular cells in some parts (black arrow) and in other parts are lined by vacuolated cells (red arrow). Some follicles show vacuolations of colloid (v). Congested blood capillaries are seen (BV); c. Some thyroid follicles are cystically dilated, some of them show ingrowth projecting into the lumen (red arrow) and others exhibit multiple follicular cells with dark nuclei (N); they nearly obliterate its lumen. Note colloid resorption droplets (V), vacuolated cytoplasm (black arrow) and congested blood vessel (BV); d. Thyroid follicles become adherent to each other with focal hyperplasia of the follicular cells (white arrow). Note the vacuolated colloid (black arrow); e. Aggregate of small follicles lined by dark nuclei and having obliterated lumen (black arrow). Note congested blood vessels (BV) and retracted colloid (red arrow); ×400.

Figure 3. Micrograph of a semithin section in the thyroid gland of a prepubertal rat of (group I) demonstrating: a. Thyroid follicle (F) lined by one layer of cubical follicular cells with vesicular spherical nuclei (black arrows), parafollicular cells with large pale nuclei and pale cytoplasm (yellow arrow) are seen. Note the variable densities of colloid (C); b. Degenerated thyroid follicles containing desquamated epithelial cells in their lumen (D). Dark nuclei (arrow head), vacuolated cytoplasm (thin arrows) and dark flattened elongated nuclei of some follicular cells are seen (curved arrow). Note the colloidal vacuolations (V); c. Parts of thyroid follicles (F). An apparent increase in the height of cells (red arrow) lining the follicles is seen. Most of these cells show cytoplasmic vacuolations (black arrow) with either pale nuclei (arrow head) or pyknotic ones (curved arrow), other follicular cells are ballooned with dark nuclei (B). ×1000. follicular cells with variably shaped hyperchromatic 6d). Other fields showed disrupted follicular cells with nuclei, swollen degenerated mitochondria and frag- shrunken, condensed nuclei. Cytoplasmic vacuola- mented rough endoplasmic reticulum. Interfollicu- tions, fragmented rough endoplasmic reticulum and lar capillaries containing RBCs in the stroma were detached microvilli were exfoliated in the follicular detected (Fig. 6a). Some areas showed the follicular lumen (Fig. 6e, f). Few fields showed exfoliation of cells with rarefied cytoplasm (Fig. 6b). Electron-dense some cells in the fine granular colloid material. The granules similar to lysosomes and fragmented dilated exfoliated cells showed condensation of the nuclear rough endoplasmic reticulum were demonstrated chromatin with distorted, variable shapes of nuclei (Fig. 6c). Some fields showed detached apical mi- and marked reduction of cell organelles (Fig. 6g). Thin crovilli in the lumen (Fig. 6a–c); others showed the bundles of collagen fibres were demonstrated in the follicular lumen with disrupted apical membrane (Fig. stroma (Fig. 6f, g).

301 Folia Morphol., 2019, Vol. 78, No. 2

Figure 4. An electron micrograph of the thyroid gland of control prepubertal rat group I showing: a. A follicular cell with euchromatic nucleus (N) and prominent nucleolus. It is surrounded by regular nuclear membrane with nuclear pores (black arrow). The cytoplasm contains organelles as mitochondria (M), rough endoplasmic reticulum (rER) and electron-dense granules (d); ×4000; b. Follicular cells with euchromatic nucleus (N) and prominent nucleolus. The follicular lumen shows fine granular colloid (C) material. Note Golgi apparatus (G) and the microvilli (arrow head) in the free surface; ×6000.

Figure 5. An electron micrograph of the thyroid gland of a prepubertal rat of group II showing: a. Electron dense granules (white arrow) as compared to electron lucent granules (black arrow). The luminal surface shows multiple detached microvilli (arrow heads). The mitochondria (M) are relatively swollen with rupture of their outer membrane. Note the dilated rough endoplasmic reticulum (rER), undulation of nuclear membrane and margination of heterochromatin (N); ×6000; b. Electron dense granules (white arrow) as compared to electron lucent granules (black arrow) and swollen mitochondria with ruptured cristae are seen (M). The luminal surface demonstrates numerous detached microvilli in the lumen (arrow head); ×10,000; c. Heterochromatic nucleus (N) with widely dispersed chromatin and prominent nucleolus (n). It is surrounded by irregular nuclear membrane (arrow head). The mitochondria are swollen with disrupted cristae (M), electron lucent (black arrow) and electron dense (white arrow) secretory granules are shown. Note bundle of collagen fibres (curved arrow); ×10,000; d. Corrugated nucleus (N) with clumping and margination of heterochromatin. The mitochondria are swollen with disrupted cristae. The follicular lumen contains fine granular colloid material (C) with almost loss of microvilli on the luminal surface of cells (black arrow); ×6000; e. A follicular cell with colloid droplet (c), dilated rough endoplasmic reticulum (rER) and corrugated nucleus with thin rim of heterochromatin (white arrow). The luminal surface demonstrates elongated and detached microvilli (black arrow); ×10,000.

DISCUSSION functions [13]. Also it is considered a model of stress, Light at night became an essential part of modern because it increases serum concentration of corti- lifestyle. This causes a significant disruption of the costerone [18]. The thyroid gland is very sensitive to daily circadian rhythm [29]. Disruption of circadian stress [20]. All metabolically active cells require TH rhythm affects many physiological and behavioural for proper functioning. Therefore, thyroid dysfunc-

302 F.A. Abdel Gawad et al., Effect of continuous light on the thyroid gland

Figure 6. An electron micrograph of the thyroid gland of a prepubertal rat of group III showing: a. Follicular cells with variably shaped hyper chromatic nuclei (N), swollen degenerated mitochondria (M), and fragmented rough endoplasmic reticulum (rER). The luminal surface shows numerous, detached microvilli (arrow head). Note the capillary containing RBCs in the interfollicular stroma (white arrow); ×3000; b. Fol- licular cells with rarefied cytoplasm (star), fragmented rough endoplasmic reticulum (rER) and hyperchromatic nuclei (N). Note detached apical microvilli in the lumen (black arrow); ×6000; c. Large follicular cell nucleus with dense chromatin at the periphery (N). Electron dense granules similar to lysosomes (white arrow) and fragmented and dilated rough endoplasmic reticulum (rER) are also seen. Note the interfollicular capillary containing RBCs (arrow head) and detached microvilli from the luminal surface (black arrows); ×5000; d. Follicular cells with rarefied cytoplasm (white arrow), swollen degenerated mitochondria (M), fragmented rough endoplasmic reticulum (rER) (arrow head) and hyperchromatic nuclei with marginated condensed chromatin (N). Note the follicular lumen (L) and disrupted apical membrane (black arrow); ×4000; e. Disrupted follicular cells with shrunken, condensed, variably shaped nuclei (N). Cytoplasmic vacuolation (V), fragmented rough endoplasmic reticulum and detached microvilli (black arrow) are exfoliated in the follicular lumen; ×6000; f. Ruptured follicular cells with shrunken, hyper chromatic nucleus (N). Note cytoplasmic vacuolation (V), fragmented rough endoplasmic reticulum (rER), electron dense granules (d) similar to lysosomes and detached microvilli are exfoliated in the follicular lumen (white arrow). Thin bundles of collagen fibres (black arrow) are seen;×4000; g. Exfoliation (EX) of some cells in the fine granular colloid material. The exfoliated cells show condensation of the nuclear chromatin (black arrow) with distorted, variably shaped nuclei (N). Note marked reduction of cell organelles (star) and thin bundles of collagen fibres (white arrow). ×4000. tion can have a wide range of effects on most of exposure (as in the present work) and inhibited in body systems [31]. There is a functional connection dark condition. between pineal gland, melatonin and the thyroid In the current study, rats exposed to constant gland. Melatonin had a strong inhibitory effect on light for long period (3 months) had overstimulated cell proliferation and TH synthesis [25]. This explains glands, some of which exhibited hyperplastic chang- the stimulation of thyroid gland under exposure to es. The later findings were novel in this work. The constant light as it is relieved from the inhibitory follicles exhibited back to back arrangement (signs effect of melatonin [3]. of hypertrophy with hyperplasia). Occasionally, the In the present work, light microscopic findings follicles were lined by multiple layers of follicular in rats exposed to constant light for short period cells. Diffusely hyperplastic thyroid was detected in (1 month) revealed that the thyroid gland was stimu- the form of follicles lined by tall columnar epitheli- lated, as evidenced by microfollicular thyroid structure um and the crowded epithelial cells projected into and little or no colloid in the thyroid follicles. Similar the lumen. These cells were actively resorbing the results were detected by Olatunji and Sofola [30]. colloid. This result was comparable with Milosevic Belviranli and Baltaci [3] reported that the activity of et al. [28] who reported that prolonged exposure to the pineal gland increased and thyroid growth was light of 600 lux (the same intensity was used in the inhibited in male rats under light restriction. This present study) activated stress system and stimulated might be explained by increased melatonin secretion growth and hypertrophy of adrenal zona fasciculata in response to dark condition. These results ensured cells in female rats. In this regard, Miler et al. [27] that the thyroid gland is stimulated in constant light mentioned that exposure to CL diminished nocturnal

303 Folia Morphol., 2019, Vol. 78, No. 2

increase in pituitary deiodinase type 2 which con- to increased lipoperoxide levels and reduced anti-oxi- verted T4 to biologically active T3. The decrease in dative stress biomarkers. In agreement, Rifaai and Abd this enzyme lowered circulating T3 levels. Therefore, El Baky [35] demonstrated that 1-month period of con- we can conclude that the compensatory increase in stant light caused structural changes in the rat thymus. secretion of TSH caused follicular cell hypertrophy These changes might cause autoimmune response. and hyperplasia. Therefore, we can hypothesize that In the present work, the electron microscopic the function of thyroid gland was stimulated in the results of group II confirmed the results obtained short duration group (group II). This was consistent by the light microscope. The follicular cells had hy- with Gooley et al. [15] who reported that thyroid was perchromatic nuclei, fragmented and dilated rough relieved from the inhibitory effect of melatonin, as endoplasmic reticulum (rER), swollen mitochondria exposure to constant light suppressed its secretion. It with disrupted cristae and electron dense secretory might affect thyroid directly or indirectly by affecting granules. Vacuolated or rarefied cytoplasm has been hypothalamus-pituitary regulation [12]. attributed by Rajkovic et al. [33] to the presence On the other hand after prolonged exposure to of dilated cisternae of rER. These markedly dilated constant light in the long duration group (group III), cisternae of rER in combination with other cellular thyroid tissue became sensitised to stimulation with changes might be an evidence of disruption to pro- TSH. In a contradictory study, Kursawe and Paschke tein synthesis. [23] demonstrated that after prolonged exposure Swollen mitochondria with disrupted membrane to constant light conditions thyroid tissue became were observed in both group II and group III in the desensitised to stimulation with TSH, due to reduced present study, which might be explained by oxidative activity and binding capacity of the thyroid tissue stress induced by constant light exposure. This may TSH receptor. This discrepancy might be attributed to lead to mitochondrial dysfunction, disruption of their the period and intensity of constant light. Regarding inner membrane with increased permeability which the shape of follicular cells in groups II and III, the allowed the solute to enter the matrix and lead to its majority of cells were high cuboidal or columnar and swelling, followed by rupture of the outer mitochondri- this finding was in concomitance with Qureshi and al membrane and release of the pre-apoptotic proteins. Mahmood [32]. This might be attributed to the stimu- El Mahalawy et al. [10] reported the same finding in latory effect of TSH on the follicular cells to synthesize renal tubular cells. Zhang et al. [43] added that mela- and secrete more hormones into the circulation due tonin counteracted oxidative damage of mitochondrial to the increased demand for the TH in an attempt DNA and restored the mitochondrial respiratory control to compensate for TH decrease. Degeneration and system. So melatonin deficiency due to constant light apoptosis of follicular cells were noticed in groups II in this work might be an explanation. and III, as evidenced histologically by dark nuclei and The electron microscopic results of group III ruptured follicles in some parts with desquamation showed that some follicles were lined by high cuboi- of cells inside the lumen. This was in accordance with dal cells with clumping of peripheral heterochroma- Escribano et al. [11] who reported that exposure tin. Their cytoplasm contained several vacuoles, col- to constant light increases cellular oxidative stress, loid vesicles, and dense lysosomes. Massively dilated decreases the activity of antioxidants and increases cisternae of rER and loss of their lamellar arrangement apoptosis inside the gland. In this regard, Tunez and were also observed. Their apical borders showed de- Montilla [38] added that melatonin had anti-oxida- tached aggregated microvilli. In some cells, nuclei ap- tive stress and anti-apoptotic effects on rats. Similar peared irregular in shape, shrunken with condensed findings were reported by Talaber et al. [36] who heterochromatin. Also, follicular hyperplasia may be detected degenerated thymic epithelial cells after connected with a cascade of cellular events involving constant light exposure. The latter suggested anoth- oxidative stress, genomic DNA damage, and modula- er explanation: accelerated thymic epithelial aging tion of apoptotic regulatory gene after exposure to caused by elevated glucocorticoids. constant light [43]. In the current study, most of the follicular cells in In the current study, exfoliated follicular cells with group III exhibited cytoplasmic vacuolations, which indistinct organelles in the lumen were observed in could be attributed to apoptosis. Similar finding was group III. These ultrastructural findings confirmed the reported by Escribano et al. [11] who attributed this light microscopic results and indicated that prolonged

304 F.A. Abdel Gawad et al., Effect of continuous light on the thyroid gland

exposure to constant light could induce follicular cell 6. Castelhano C, Baumans V. The impact of light, noise, cage damage and follicular disruption. This result could be cleaning and in-house transport on welfare and stress of laboratory rats. J Sci. 2009; 43: 311–327. attributed to apoptosis and thyroid toxicity explained 7. Chi HC, Chen SL, Liao CJ. Thyroid hormone receptors pro- by Escribano et al. [11]. Congested blood vessels mote metastasis of human hepatoma cells via regulation of observed in this study were in accordance with that regulation of trail. J Cell Death Differ. 2012; 19: 1802–1814. reported by Yu et al. [41] and Hassanin et al. [17]. They 8. Coomans CP, van den Berg SAA, Houben T, et al. Detrimental explained that oxidative stress and lipid peroxidation effects of constant light exposure and high-fat diet on circadian energy metabolism and insulin sensitivity. FASEB J. might affect vascular walls, leading to their dilatation 2013; 27(4): 1721–1732, doi: 10.1096/fj.12-210898, and congestion. indexed in Pubmed: 23303208. The present work revealed histological changes in 9. Dauchy RT, Dauchy EM, Tirrell RP, et al. Dark-phase light the parafollicular (C cells) cells in group III that were contamination disrupts circadian rhythms in plasma measures of endocrine physiology and metabolism in evident at the ultrastructural level. The cells showed rats. Comp Med. 2010; 60(5): 348–356, indexed in Pu- swollen mitochondria, few ribosomes, hyperchro- bmed: 21262119. matic nucleus and few lysosomes. These observations 10. El-Mahalaway A, Salem M, Mousa A. The effect of po- indicated the possibility of a relationship between the tassium dichromate on convoluted tubules of the kidney functional state of the thyroid gland and the activity of adult male albino rats and the possible protective role of ginseng. Egypt J Histol. 2015; 38(2): 157–167, of C cells. It was suggested that changes in C cells doi: 10.1097/01.ehx.0000464738.41270.06. were in line with the alterations in the follicular cells. 11. Escribano BM, Díaz-Moreno A, Moreno A, et al. Impact Considering that; the elevated TSH level might directly of light/dark cycle patterns on oxidative stress in an adri- regulate C cells [26]. amycin-induced nephropathy model in rats. PLoS One. 2014; 9(5): e97713, doi: 10.1371/journal.pone.0097713, indexed in Pubmed: 24852173. CONCLUSIONS 12. García-Marín R, de Miguel M, Fernández-Santos JM, In the current work, it can be concluded that et al. Melatonin-synthesizing enzymes and melatonin short-term exposure to constant light for 1 month receptor in rat thyroid cells. Histol Histopathol. 2012; had minimal effect on thyroid gland tissues (micro- 27(11): 1429–1438, doi: 10.14670/HH-27.1429, indexed in Pubmed: 23018242. follicular structure with decreased luminal colloid and 13. Ghadiri E, Ahmadi R, Nargesi A. Effects of Darkness Stress hormonal picture of early hypothyroidism). While on Thyroid Function. International Conference on Chemical, longer exposure to light for 3 months had detrimen- Agricultural and Medical Sciences (CAMS-2014) May 2-3, tal effects on the thyroid gland structure of male 2014 Antalya (Turkey). 2014, doi: 10.15242/iicbe.c514077. albino rats (hyperplastic and degenerative changes). 14. Gibbons R, Hankey J. Influence of vertical illuminance on pedestrian visibility in crosswalks. Trans Res Record. 2009; To the best of our knowledge, this is the first report 1973(1): 105–112, doi: 10.1177/0361198106197300113. demonstrating that CL induced hyperplastic changes 15. Gooley J, Chamberlain K, Smith K, et al. Exposure to room in the thyroid gland. These effects were more evident light before bedtime suppresses melatonin onset and in the adult groups. shortens melatonin duration in humans. Endocrinology. 2011; 152(2): 742–742, doi: 10.1210/endo.152.2.zee742. 16. Hall LC, Salazar EP, Kane SR, et al. Effects of thyroid hor- References mones on human breast cancer cell proliferation. J Steroid 1. Allan JS, Czeisler CA. Persistence of the circadian thy- Biochem Mol Biol. 2008; 109(1-2): 57–66, doi: 10.1016/j. rotropin rhythm under constant conditions and after jsbmb.2007.12.008, indexed in Pubmed: 18328691. light-induced shifts of circadian phase. J Clin Endocrionol 17. Hassanin KMA, Abd El-Kawi SH, Hashem KS. The prospec- Metab. 2010; 79(2): 508–512, doi: 10.1210/jc.79.2.508. tive protective effect of selenium nanoparticles against 2. Azevedo LH, Aranha AC, Stolf SF, et al. Evaluation of low chromium-induced oxidative and cellular damage in intensity laser effects on the thyroid gland of male mice. rat thyroid. Int J Nanomedicine. 2013; 8: 1713–1720, Photomed Laser Surg. 2005; 23(6): 567–570, doi: 10.1089/ doi: 10.2147/IJN.S42736, indexed in Pubmed: 23658489. pho.2005.23.567, indexed in Pubmed: 16356148. 18. Haugen BR. Drugs that suppress TSH or cause central 3. Belviranli M, Baltaci KA. The relation between reduced hypothyroidism. Best Pract Res Clin Endocrinol Metab. serum melatonin levels and zinc in rats with induced 2009; 23(6): 793–800, doi: 10.1016/j.beem.2009.08.003, hypothyroidism. Cell Biochem Funct. 2008; 26: 19–23. indexed in Pubmed: 19942154. 4. Benstaali C, Mailloux A, Bogdan A, et al. Circadian rhythms 19. Höfling DB, Chavantes MC, Juliano AG, et al. Low-level of body temperature and motor activity in rodents their laser in the treatment of patients with hypothyroidism relationships with the light-dark cycle. Life Sci. 2001; induced by chronic autoimmune thyroiditis: a rand- 68(24): 2645–2656, indexed in Pubmed: 11400908. omized, placebo-controlled clinical trial. Lasers Med Sci. 5. Boelaert K, Franklyn JA. Thyroid hormone in health and 2013; 28(3): 743–753, doi: 10.1007/s10103-012-1129-9, disease. J Endocrinol. 2005; 187(1): 1–15 indexed in Pubmed: 22718472.

305 Folia Morphol., 2019, Vol. 78, No. 2

20. Iglesias P, Díez JJ. Thyroid dysfunction and kidney disease. 33. Rajkovic V. Light and electron microscopic study of the Eur J Endocrinol. 2009; 160(4): 503–515, doi: 10.1530/ thyroid gland in rats exposed to power-frequency elec- EJE-08-0837, indexed in Pubmed: 19095779. tromagnetic fields. J Exp Biol. 2006; 209(17): 3322–3328, 21. Kalsbeek A. Functional connections between the supra- doi: 10.1242/jeb.02375. chiasmatic nucleus and the thyroid gland as revealed 34. Ratcliff R, Van Dongen HPA. Sleep deprivation affects by lesioning and viral tracing techniques in the rat. En- multiple distinct cognitive processes. Psychon Bull Rev. docrinology. 2000; 141(10): 3832–3841, doi: 10.1210/ 2009; 16(4): 742–751, doi: 10.3758/PBR.16.4.742, indexed en.141.10.3832. in Pubmed: 19648462. 22. Kriegsfeld LJ, Silver R. The regulation of neuroendocrine 35. Rifaai AR, Abd El Baky FA. Histological and immunohisto- function: Timing is everything. Horm Behav. 2006; 49(5): chemical study on the effect of constant light exposure 557–574, doi: 10.1016/j.yhbeh.2005.12.011, indexed in on T lymphocyte subsets in the thymus and lymph node Pubmed: 16497305. of male albino rats. J Histol. 2017; 40(1): 1110–1159. 23. Kursawe R, Paschke R. Modulation of TSHR signaling by 36. Talaber G, Kvell K, Varecza Z, et al. Wnt-4 Protects posttranslational modifications. Trends Endocrinol Metab. Thymic Epithelial Cells Against Dexamethasone-Induced 2007; 18(5): 199–207, doi: 10.1016/j.tem.2007.05.002, Senescence. Rejuvenation Res. 2011; 14(3): 241–248, indexed in Pubmed: 17524661. doi: 10.1089/rej.2010.1110. 24. Mann DR, Bhat G, Stah CD, et al. Induction of a hypothy- 37. Thakkar BP, Zala VM, Ramachandran AV. Simultaneous roid state during juvenile develop constant light affects pi- melatonin administration effectively deprograms the tuitary thyroid axis. J Morphol Monkey. 2006; 18: 662–671. negative influence of neonatal hypothyroidism on im- 25. Marin GR, Fernandez JM. Melatonin in the thyroid gland: mature follicles but not on mature follicles and body and regulation by thyroid-stimulating hormone and role in ovarian weights. J Endocrinol Metab. 2011; 1: 220–226, thyroglobulin gene expression. J Physiol Pharma. 2015; doi: 10.4021/jem54w. 66(5): 643–652. 38. Tunez I, Montilla P. Neuroprotective role of melatonin 26. Martín-Lacave I, Borrero MJ, Utrilla JC, et al. C cells evolve against Alzheimer’s disease, Huntington’s disease and at the same rhythm as follicular cells when thyroidal other cerebral disorders. In: Tu´nez I; Montilla P, editors. status changes in rats. J Anat. 2009; 214(3): 301–309, Melatonin present and future. Nova Science Publishers doi: 10.1111/j.1469-7580.2008.01044.x, indexed in Pu- Inc., New York 2007: 315–347. bmed: 19245497. 39. Vinogradova IA, Anisimov VN, Bukalev AV, et al. Circadian 27. Miler M, Branka S, Natasa N. Constant light affects pitui- disruption induced by light-at-night accelerates aging and tary-thyroid axis. J Morph. 2014; 275: 1161–1172. promotes tumorigenesis in rats. Aging (Albany NY). 2009; 28. Milosević V, Trifunović S, Sekulić M, et al. Chronic expo- 1(10): 855–865, doi: 10.18632/aging.100092, indexed in sure to constant light affects morphology and secretion Pubmed: 20157558. of adrenal zona fasciculata cells in female rats. Gen 40. Weber JB, Mayer L, Cenci RA, et al. Effect of three dif- Physiol Biophys. 2005; 24(3): 299–309, indexed in Pu- ferent protocols of low-level laser therapy on thyroid bmed: 16308425. hormone production after dental implant placement in an 29. Navara KJ, Nelson RJ. The dark side of light at night: phys- experimental rabbit model. Photomed Laser Surg. 2014; iological, epidemiological, and ecological consequences. 32(11): 612–617, doi: 10.1089/pho.2014.3756, indexed J Pineal Res. 2007; 43(3): 215–224, doi: 10.1111/j.1600- in Pubmed: 25265487. 079X.2007.00473.x, indexed in Pubmed: 17803517. 41. Yu Fu, Wang Z, Ju B, et al. Apoptotic effect of organophos- 30. Olatunji BI, Sofola AO. Effect of continuous light and dark- phorus insecticide chlorpyrifos on mouse retina in vivo via ness exposures on the pituitary gonadal axis and thyroid oxidative stress and protection of combination of vitamins activity in male rats. Afr J Biomed Res. 2001; 4: 119–122. C and E. Exp Toxicol Pathol. 2008; 59(6): 415–423, doi: 31. Patil, V, Dhurvey, V. Exposure to sodium fluoride affects 10.1016/j.etp.2007.11.007, indexed in Pubmed: 18222074. thyroid follicular cells in albino rats. IJPAES. 2015; 5: 42. Zambrano A, García-Carpizo V, Gallardo ME, et al. The 56–61. thyroid hormone receptor b induces DNA damage 32. Qureshi IZ, Mahmood T. Prospective role of ascorbic and premature senescence. J Cell Biol. 2014; 204(1): acid (vitamin C) in attenuating hexavalent chromi- 129–146, doi: 10.1083/jcb.201305084, indexed in Pu- um-induced functional and cellular damage in rat bmed: 24395638. thyroid. Toxicol Ind Health. 2010; 26(6): 349–359, 43. Zhang J, Wang X, Vikash V, et al. ROS and ROS-Mediated doi: 10.1177/0748233710371109, indexed in Pu- Cellular Signaling. Oxidative Med Cell Longev. 2016; 2016: bmed: 20504825. 1–18, doi: 10.1155/2016/4350965.

The effect of diclofenac sodium on neural tube development in the early stage of chick embryos T. Ertekin1, A. Bilir1, E. Aslan2, B. Koca3, O. Turamanlar1, A. Ertekin4, S. Albay5

1Department of Anatomy, University of Afyon Kocatepe, School of Medicine, Afyonkarahisar, Turkey 2Department of Histology and Embryology, University of Afyon Kocatepe, School of Medicine, Afyonkarahisar, Turkey 3Department of Biochemistry, University of Afyon Kocatepe, School of Medicine, Afyonkarahisar, Turkey 4Emergency Medicine, Afyonkarahisar State Hospital, Afyonkarahisar, Turkey 5Department of Anatomy, University of Süleyman Demirel, School of Medicine, Isparta, Turkey

[Received: 4 June 2018; Accepted: 9 July 2018]

Background: Neural tube defects are congenital malformations of the central nervous system. Genetic predisposition and some environmental factors play an important role in the development of neural tube defects. This study aimed to investigate the effects of diclofenac sodium on the neural tube development in a chick embryo model that corresponds to the first month of vertebral development in mammals. Materials and methods: Seventy-five fertile, specific pathogen-free eggs were incubated for 28 h and were divided into five groups of 15 eggs each. Diclofenac sodium was administered via the sub-blastodermic route at this stage. Incubation was continued till the end of the 48th h. All eggs were then opened and embryos were dissected from embryonic membranes and evaluated morphologically and histopathologically. Results: It was determined that the use of increasing doses of diclofenac sodium led to defects of midline closure in early chicken embryos. There were statistically significant differences in neural tube positions (open or close) among the groups. In addition; crown–rump length, somite number were significantly decreased in high dose experimental groups compared with control group. Conclusions: This study showed that development of neurons is affected in chicken embryos after administration of diclofenac sodium. The exact teratogenic mechanism of diclofenac sodium is not clear; therefore it should be investigated. (Folia Morphol 2019; 78, 2: 307–313)

Key words: diclofenac sodium, chick embryo, neural tube defect, development

INTRODUCTION co-administered with quinolone-derivative antibacte- Non-steroidal anti-inflammatory drugs (NSAIDs) rial drugs [11, 36, 42]. Diclofenac sodium (sodium-(O- are important chemicals in the human body and have ((2,6-dichlorophenyl)-amino)-phenyl)-acetate) (DS) is been widely used for alleviation of pain, inflamma- a NSAID characterised by a relatively low molecular tion, myocardial infarction and stroke [22]. NSAIDs weight, having potent anti-inflammatory, analgesic, have different side effects, such as gastrointestinal and antipyretic effects on tissues [35, 37] and used damage, platelet dysfunction and convulsions when commonly by women of reproductive age for treat-

Address for correspondence: T. Ertekin, PhD, Associate Professor, Department of Anatomy, University of Afyon Kocatepe, School of Medicine, Afyonkarahisar, Turkey, tel: 00 90 505 6482854, fax: 00 90 0 272 246 33 00, e-mail: [email protected]

307 Folia Morphol., 2019, Vol. 78, No. 2

ment of dysmenorrhoea and menorrhagia [12]. Up to MATERIALS AND METHODS 20% of women consume prescribed medicine during This study was conducted in Afyon Kocatepe Uni- the first trimester of pregnancy; of these medications, versity, Department of Anatomy Laboratory. All the 3% are NSAIDs [28]. DS acts by inhibiting the enzyme experiments were performed following ethical guide- cyclooxygenase (COX), reducing the arachidonic acid lines established for animal usage by Institutional release and enhancing its uptake. The COX reaction Animal Ethics Committee (IAEC) Afyon Kocatepe Uni- is the rate-limiting step in the formation of prosta- versity. The 75 eggs were procured from the Veteri- glandins from arachidonic acid. COX-2, the inducible narian Control and Research Institute, Bornova, Izmir, isoform of COX family, is selectively expressed in Turkey, 65 ± 5 g in weight, specific pathogen-free neurons of the cerebral cortex, hippocampus, and and day 0 fertilised eggs of white chickens. The eggs amygdale. Prostaglandins are important chemical were placed in the incubator with sharp ends pointing mediators in the human body, being involved in both down in order to ensure the continuity of the embryos normal and abnormal function of virtually every organ and to have them available at the times desired. The and system [2, 37]. incubator was kept at a constant temperature range It was determined that DS crosses from the hu- of 37.8 ± 0.2ºC and humidity of 60–70%.The eggs man placenta during the first and second trimesters were automatically rolled every 2 h at a 45° angle to to foetus [30, 37]. The fact that DS cross the placental the vertical axis during the incubation period. barrier to prevent the biosynthesis of prostanoids and The eggs were opened using window procedure passes into the foetal circulation causes important after 28-h incubation (the embryos reached stage 8 side effects and sometimes malformations in new- of development according to Hamburger and Ham- borns [12, 29, 45]. Some assays determined that the ilton), and were divided into five groups of 15 eggs DS using throughout the perinatal period may cause each (with one control and four experimental groups) teratogenic effects on some organs [3, 6, 16, 19]. In [20]. Firstly; the egg shell was sterilised with povi- addition; the use of DS during embryonic develop- done-iodine (10%) and then ethanol (70%). A small ment of the central nervous system (CNS) can pro- window (1–2 cm) was made aseptically with a spe- duce a wide array of neurological dysfunctions and cific technique in the eggshell. The round-shaped neuroanatomical anomalies in animal models [9, 22]. embryonic disc became visible after the surrounding However, very little information is available respecting membranes were cut. At this stage; under sterile con- DS effects on neurological structures especially with ditions, DS (Voltaren, 75 mg/3 mL ampoule, Novartis, regard to prenatal development. Kartal, Istanbul, Turkey) was diluted in saline and Congenital CNS anomalies are the second most prepared in the selected dosages. In humans; the prevalent anomalies following congenital cardiovas- recommended daily dose is 50 mg/day. The daily dose cular anomalies [24]. Neural tube defects (NTD) are an can be increased to a total of 200 mg/day according important area of congenital malformations. Genetic to the severity of the symptoms. The maximum daily predisposition and some environmental factors play dose is 200 mg/day [14, 41]. For animals, 1–18 mg/kg a significant role in the development of NTDs. NTDs, DS doses were used in teratogenic studies related which occur in approximately 6/10000 newborns, to pregnancy [3, 6, 19, 44]. A toxic dose of DS was create a heterogeneous and complex group of con- 5 mg/kg in avian species [1, 33]. Four doses were genital anomalies [43]. tested: first dose, 1/20 of the toxic dose = 0.25 mg/kg; Neuronal and spinal development stages of chick second dose, 1/10 of the toxic dose = 0.5 mg/kg; third embryo are closely similar to the development stages dose, 1/5 of the toxic dose = 1 mg/kg; and fourth of human embryos. To the best of our knowledge, dose, 1/2.5 of the toxic dose = 2 mg/kg. there was no study in the literature that investi- Diclofenac sodium was administered via the sub- gates the toxic effects of DS on neural tube (NT) blastodermic route in a volume of 10 µL in groups development by using chick embryo model. This B, C, D and E by Hamilton microinjector (0.25 mg/kg, study was designed to fill this gap in knowledge 0.5 mg/kg, 1 mg/kg and 2 mg/kg in groups B, C, D and E, and to determine the probable toxic effect of DS respectively). Group A served as the control group with respect to different doses. Thereby, it was also and was administered 10 µL 0.9% NaCl via the sub- aimed to determine potential results with its use blastodermic route. After treatment, small window during pregnancy. was sealed with cellophane tape. Then the eggs were

308 T. Ertekin et al., Effect of diclofenac sodium on embryonic development

A B

Figure 1. A. Normal appearance of post-incubation 2-day chick embryo of group A; B. Histopathologic views of normal chicken embryos under light microscope (H&E, ×20); nt — neural tube; n — notochord; s — somites; h — heart.

hand-turned 180° and placed in the incubator. All the light microscope after staining with H&E were eggs were opened at the 48th h (Hamburger–Hamilton found to be consistent with the stereomicroscopic stage 12), and the morphological features of each examination (Fig. 1). embryo were evaluated under stereomicroscope to Group B: 7 embryos (46.7%) had NTD and 8 em- assess any gross developmental abnormalities [20]. bryos (53.3%) were intact and these embryos sustained Embryos were classified as with defect, normal or their normal development and were in the embryonic undeveloped. The samples were then transferred to stage 12 where they were expected to be according to Petri dishes containing 10% formaldehyde solutions Hamburger–Hamilton classification (Fig. 2). for histopathological study. The embryos from each Group C: 8 embryos (53.3%) had NTD, 7 embryos group were fixed in 10% formaldehyde, dehydrated (46.7%) were intact and their NTs were closed (Fig. 3). through a graded alcohol series, cleaned in xylene and Group D: 13 embryos had NTD (86%) and 2 em- then embedded in paraffin wax. Serial sections of four bryos (14%) were intact and their NTs were closed. micron thickness were taken from the paraffin blocks Group E: 15 embryos (100%) had NTD and 1 of and stained with haematoxylin and eosin (H&E) dye. these embryos was undeveloped and its development The sections were examined using a light microscope. stage was 9 according to Hamburger–Hamilton embryonic classification (Fig. 4). Statistical analysis There were statistically significant differences Analysis of all findings was performed using the (p < 0.001) in NT positions (open or close) among Statistical Package for the Social Sciences (SPSS) the groups. The crown–rump length and mean somite 22.0 programme. The data related to NT (open or numbers were diminished in experimental groups ac- closed) were analysed by using c2 test. The somite cording to used dosages compared to control group. number, crown–rump length and protein contents These decreases were determined statistically signifi- were analysed by using non-parametric Kruskal-Wallis cant (p < 0.001, Table 1). tests. Dunn test were employed as post-hoc tests and p < 0.001 were considered significant. DISCUSSION Neural tube defects are serious birth defects of RESULTS the CNS that occur during embryonic development In our study, we investigated the effect of DS at when the NT fails to close completely, leading to different dosages on NT development. brain and spine anomalies that can lead to death or Group A: 15 embryos (100%) in group A were lifelong disability [5]. The period from appearance expected to be according to Hamburger–Hamilton of neural plaque up to closure of palate, i.e. the embryonic classification stage 12 and their NTs were period between 18th day and 60th day of pregnancy, closed. No malformation or developmental retarda- is the period where the possibility of congenital tion was observed. The tissue samples observed under anomaly is highest, often before women are aware

309 Folia Morphol., 2019, Vol. 78, No. 2

A B

Figure 2. A. The opening in the neural tube is seen in the light microscope view of group B; B. A histologic section was obtained and stained with H&E. The neural tube is open in the light microscope image of the same group (H&E, ×20); ont — open neural tube; n — notochord; s — somites; h — heart.

A B

Figure 3. A. The neural tube is not closed in the light microscope image of the group C; B. A histologic section was obtained and stained with H&E. The neural tube is open in the light microscope image of the same group (H&E, ×20); ont — open neural tube; n — notochord; s — somites; h — heart.

A B

Figure 4. A. Developmental retardation seen in the light microscope view of group E; B. The opening in the neural tube is seen in the cervical and caudal region of the chick embryo of group E; ont — open neural tube.

310 T. Ertekin et al., Effect of diclofenac sodium on embryonic development

Table 1. The statistical analyses embryonic development in control and experimental groups (with diclofenac sodium) Parameters Group A Group B Group C Group D Group E P Open NT /Close NT 0/15 7/8 8/7 13/2 15/0 < 0.001* Crown–rump length [µm] 639.06 ± 105.48 559.42 ± 118.61 540.02 ± 113.86 498.02 ± 83.94 497.62 ± 76.52 < 0.001**, # Somite number 17.73 ± 1.98 14.73 ± 3.03 14.60 ± 3.73 12.86 ± 3.22 12.86 ± 2.55 < 0.001**, #

*Chi-square test; **Kruskal-Wallis test; #The difference was determined between group A and group D, group A and group E; NT — neural tube

they are pregnant. These anomalies originate from bryos. All embryos had NTD in the high dose group. insufficiency in NT formation or re-opening after In addition; crown–rump length, somite number were formation of NT [15, 40]. Genetic and environmen- significantly decreased in high dose experimental tal factors (geographical factors, socio-economic groups (groups C, D) compared with control group. factors, alcohol and drug use) can be specified in Assays related to DS have reported that its tera- aetiology of NT closure defects [21, 26, 38]. togenic effect risk is low in case it is used in preg- The primary experimental methods of NTD in- nancy, but they are limited. However, the mecha- clude amphibian, mammalian, poultry, and computer nism of embryonic damage related to DS use is modelling. These models have advantages and dis- uncertain [6, 9]. advantages compared to each other. The early chick Researchers showed that diclofenac inhibited im- embryo model that corresponds to the first month plantation and embryonic development in rats when of embryonal development in mammals is an ideal given on gestation day 5. In this study, rat blastocysts modelling [13]. were cultured in diclofenac in vitro, then implanted In addition, the studies have demonstrated that to host mothers on day 5 of pseudopregnancy. Large cytochalasins, papaverine, diazepam, caffeine, etha- doses of diclofenac (75 µg/mL) in culture were toxic. nol, folic acid antagonists such as methotrexate and Smaller doses (40 µg/mL) had a profound effect on aminopterin, antiepileptic drugs such as phenytoin and implantation. Another group of host mothers received lacosamide, and local anaesthetics cause NT closure diclofenac i.p., 1 h prior to transfer of untreated blas- defects in early stage chick embryos [4, 18, 23, 25, 39]. tocysts. Control animals had a 72% implantation rate, It is known that NSAIDs are one of the commonest whereas there was only a 35–41% implantation rate drugs prescribed by general practitioners worldwide, after in vitro diclofenac treatment. They determined and they have been taken during the early concep- that prostaglandin appears to be essential during the tion period [37]. DS may trigger alterations in the process of implantation and placentation. When this CNS morphology having long-term teratogenic ef- process is disturbed by DS the number of growth- fects on neuronal development [10]. The developing retarded embryos increases. If DS is such a potent CNS is also the most vulnerable to insufficient and inhibitor of prostaglandin synthetase it should be harmful conditions such as drug exposure during the expected that decidualisation would also be defec- gestational stage [17]; since different parts of the tive. In the treated host mothers only 7% of embryos CNS form at different stages of development, there were normal, while 34% were growth-retarded [7]. is not one critical period but many critical periods. Similar to above study; another assay was reported Some neurons are formed around the time of closure a positive correlation between use of NSAID during of the NT [34]. pregnancy and miscarriages [27]. In this study, possible adverse effect of DS on The effect of DS on the developing embryo during the NT was investigated in chick embryo model. All the critical period of organogenesis was investigated embryos in the control group that were monitored by using a whole rat embryo culture model. They ex- under the microscope after 48 h had reached the posed to embryos various concentrations of DS and appropriate embryonic stage according to the Ham- scored for growth and differentiation at the end of burger–Hamilton method. The effects of DS on the the culture period. It was found that although caudal embryo were correlated with the dose of DS. It was NT, flexion and hind limb were significantly lower determined that the use of increasing doses of DS in embryos exposed high concentration of DS (7.5 led to defects of midline closure in early chicken em- and 15.0 µg/mL); there were no effect low doses of

311 Folia Morphol., 2019, Vol. 78, No. 2

it. Based on their results it was suggested that high tion of potentially hazardous substances directly on concentration of DS has a teratogenic effect [9]. the embryo. In the present study, we did not use Same researchers thought that the teratogenic effects specific markers of neurons to see the toxicity induced of NSAIDs might be mediated through free oxygen by diclofenac. We interpreted our results based on radical production. The influence of DS on cellular the light of histopathological findings. It is clear that reactive oxygen species production in embryos was improved technical materials and studies with larger evaluated by measuring 8 isoprostaglandin F2a level. sample sizes would be useful to confirm the toxic ef- Their results showed that 8-isoprostaglandin F2a level fects of DS in prenatal period. Further investigation on was significantly elevated in embryos exposed to high the exact mechanism of diclofenac toxicity would be concentration of DS (7.5 and 15.0 µg/mL) but no sig- valuable. Present findings cannot serve as definitive nificant difference was detected between the control evidence of the use of DS for embryo-toxicity but it and low concentration group (1.5 µg/mL). Embryos does provide an experimental basis for the caution exposed to the high concentration of DS (15.0 µg/ of DS use prescription in pregnancy. /mL) had a significantly lower total morphology scores for caudal NT, hind limb, flexion, and brain. There REFERENCES was no significant difference in yolk sac diameter, 1. Akter R, Sarker M. Effect of diclofenac sodium in broilers. Bangl J Vet Med. 2015; 13(1): 19–24, doi: 10.3329/bjvm. crown–rump length or number of somite between v13i1.23710. experimental and control groups [8]. Free oxygen 2. Andreasson KI, Savonenko A, Vidensky S, et al. Age- radicals are highly reactive and unstable. In the event -dependent cognitive deficits and neuronal apoptosis in cyclooxygenase-2 transgenic mice. J Neurosci. 2001; 21(20): of oxidative stress, free radicals are in excess, resulting 8198–8209, indexed in Pubmed: 11588192. in cellular damage. Embryotoxic effects of reactive 3. Arslan H, Aktaş A, Elibol E, et al. Effects of prenatal diclofenac oxygen species may be related to oxidative damage sodium exposure on newborn testis: a histomorphometric study. Biotech Histochem. 2016; 91(4): 277–282, doi: 10.310 to DNA and other cellular macromolecules [31]. 9/10520295.2016.1151551, indexed in Pubmed: 26984645. Prenatally exposed DS had a neurotoxicity effect 4. Barutcuoglu M, Selcuki M, Vatansever S, et al. The effects of in the CNS, namely the pyramidal and granular cells ethanol on neural tube development in early stage neural tube development chick embryos. Türk Nöroşir Derg. 2001; of hippocampus [17] and Purkinje cells of cerebel- 11: 32–36. lum [32]. NSAIDs suppress cell proliferation in the 5. Botto LD, Moore CA, Khoury MJ, et al. Neural-tube defects. N Engl J Med. 1999; 341(20): 1509–1519, doi: 10.1056/ spinal cord and dorsal root ganglia by affecting cell NEJM199911113412006, indexed in Pubmed: 10559453. cycle regulators since it has been shown that DS, in 6. Canan S, Aktaş A, Ulkay MB, et al. Prenatal exposure to contrast to other NSAIDs including aspirin, naproxen, a non-steroidal anti-inflammatory drug or saline solution impairs sciatic nerve morphology: a stereological indomethacin and ibuprofen, restrains the differ- and histological study. Int J Dev Neurosci. 2008; 26(7): entiation of neuronal stem cells into neurons and 733–738, doi: 10.1016/j.ijdevneu.2008.07.005, indexed in also suppresses cell proliferation via the induction of Pubmed: 18678241. 7. Carp HJ, Fein A, Nebel L. Effect of diclofenac on implanta- apoptosis. Therefore, DS appears to have some nega- tion and embryonic development in the rat. Eur J Obstet tive effect on both development and differentiation Gynecol Reprod Biol. 1988; 28(3): 273–277, indexed in nerve cells [2, 22]. Pubmed: 3208969. 8. Chan LYS, Chiu PYu, Siu NSS, et al. Diclofenac-induced em- bryotoxicity is associated with increased embryonic 8-isopros- CONCLUSIONS taglandin F2alpha level in rat whole embryo culture. Reprod Consequently, our study has demonstrated that Toxicol. 2002; 16(6): 841–844, indexed in Pubmed: 12401513. 9. Chan LY, Chiu PY, Siu SS, et al. A study of diclofenac-induced DS exerts direct teratogenic effect on the process of teratogenicity during organogenesis using a whole rat em- NT formation of chick embryo in a dose-dependent bryo culture model. Hum Reprod. 2001; 16(11): 2390–2393, manner. It has also been shown that DS significantly indexed in Pubmed: 11679526. 10. Chen LC, Ashcroft DM. Do selective COX-2 inhibitors increase decreases the crown–rump length and somite number the risk of cerebrovascular events? A meta-analysis of ran- in high dose experimental groups compared with domized controlled trials. J Clin Pharm Ther. 2006; 31(6): 565–576, doi: 10.1111/j.1365-2710.2006.00774.x, indexed control group. The chosen model does not directly in Pubmed: 17176361. reflect the environment and conditions of a devel- 11. Davey PG. Overview of drug interactions with the quinolo- oping human embryo. Therefore, it is not possible nes. J Antimicrob Chemother. 1988; 22 Suppl C: 97–107, indexed in Pubmed: 3053579. to simply extend and apply the results observed in 12. Dawood MY. Nonsteroidal antiinflammatory drugs and re- chick embryos to humans. However, the chick embryo production. Am J Obstet Gynecol. 1993; 169(5): 1255–1265, model has the advantage of allowing the investiga- doi: 10.1016/0002-9378(93)90292-q.

312 T. Ertekin et al., Effect of diclofenac sodium on embryonic development

13. Drake VJ, Koprowski SL, Lough JW, et al. Gastrulating chick safest treatment options? Drug Saf. 1998; 19(5): 389–410, embryo as a model for evaluating teratogenicity: a compari- doi: 10.2165/00002018-199819050-00006, indexed in son of three approaches. Birth Defects Res A Clin Mol Teratol. Pubmed: 9825952. 2006; 76(1): 66–71, doi: 10.1002/bdra.20202, indexed in 30. Ostensen M. Nonsteroidal anti-inflammatory drugs during Pubmed: 16333841. pregnancy. Scand J Rheumatol Suppl. 1998; 107: 128–132, 14. Fortun PJ, Hawkey CJ, Fortun PJ, et al. Nonsteroidal antiinflam- indexed in Pubmed: 9759151. matory drugs and the small intestine. Curr Opin Gastroenterol. 31. Parman T, Chen G, Wells PG. Free radical intermediates 2005; 21(2): 169–175, indexed in Pubmed: 15711208. of phenytoin and related teratogens. Prostaglandin H 15. Gardner WJ. Myelomeningocele, the result of rupture of synthase-catalyzed bioactivation, electron paramagnetic the embryonic neural tube. Cleve Clin Q. 1960; 27: 88–100, resonance spectrometry, and photochemical product analy- indexed in Pubmed: 13826548. sis. J Biol Chem. 1998; 273(39): 25079–25088, indexed in 16. Gevrek F, Kara M, Ragbetli M, et al. Effects of prenatally ex- Pubmed: 9737965. posed diclofenac sodium on rat heart tissue:a stereological 32. Ragbetli MC, Ozyurt B, Aslan H, et al. Effect of prenatal and histological study. Turk J Med Sci. 2015; 45: 474–480, exposure to diclofenac sodium on Purkinje cell numbers in doi: 10.3906/sag-1404-173. rat cerebellum: a stereological study. Brain Res. 2007; 1174: 17. Gokcimen A, Rağbetli MC, Baş O, et al. Effect of prenatal 130–135, doi: 10.1016/j.brainres.2007.08.025, indexed in exposure to an anti-inflammatory drug on neuron num- Pubmed: 17868656. ber in cornu ammonis and dentate gyrus of the rat hip- 33. Reddy NC, Anjaneyulu Y, Sivasankari B, et al. Comparative pocampus: a stereological study. Brain Res. 2007; 1127(1): toxicity studies in birds using nimesulide and diclofenac 185–192, doi: 10.1016/j.brainres.2006.10.026, indexed in sodium. Environ Toxicol Pharmacol. 2006; 22(2): 142– Pubmed: 17123485. –147, doi: 10.1016/j.etap.2006.02.004, indexed in Pub- 18. Güney O, Canbilen A, Konak A, et al. The effects of fo- med: 21783701. lic acid in the prevention of neural tube development 34. Rodier PM. Chronology of neuron development: animal defects caused by phenytoin in early chick embryos. Spine studies and their clinical implications. Dev Med Child Neurol. (Phila Pa 1976). 2003; 28(5): 442–445, doi: 10.1097/01. 1980; 22(4): 525–545, indexed in Pubmed: 7409345. BRS.0000048647.17577.13, indexed in Pubmed: 12616154. 35. Savaşer A, Ozkan Y, Işimer A. Preparation and in vitro evalu- 19. Güven D, Altunkaynak BZ, Ayranci E, et al. Stereological and ation of sustained release tablet formulations of diclofenac histopathological evaluation of ovary and uterine horns of sodium. Farmaco. 2005; 60(2): 171–177, doi: 10.1016/j. female rats prenatally exposed to diclofenac sodium. J Obstet farmac.2004.10.001, indexed in Pubmed: 15752476. Gynaecol. 2013; 33(3): 258–263, doi: 10.3109/01443615.2 36. Segev S, Rehavi M, Rubinstein E. Quinolones, theophylline, 012.761185, indexed in Pubmed: 23550853. and diclofenac interactions with the gamma-aminobutyric 20. Hamburger V, Hamilton HL. A series of normal stages in the acid receptor. Antimicrob Agents Chemother. 1988; 32(11): development of the chick embryo. J Morphol. 1951; 88(1): 1624–1626, indexed in Pubmed: 2855295. 49–92, indexed in Pubmed: 24539719. 37. Siu SS, Yeung JH, Lau TK. A study on placental transfer of di- 21. Honein MA, Paulozzi LJ, Mathews TJ, et al. Impact of folic clofenac in first trimester of human pregnancy. Hum Reprod. acid fortification of the US food supply on the occurrence 2000; 15(11): 2423–2425, indexed in Pubmed: 11056146. of neural tube defects. JAMA. 2001; 285(23): 2981–2986, 38. Tunçbilek E, Boduroğlu K, Alikaşifoğlu M. Neural tube indexed in Pubmed: 11410096. defects in Turkey: prevalence, distribution and risk factors. 22. Kudo C, Kori M, Matsuzaki K, et al. Diclofenac inhibits prolif- Turk J Pediatr. 1999; 41(3): 299–305, indexed in Pub- eration and differentiation of neural stem cells. Biochem Phar- med: 10770089. macol. 2003; 66(2): 289–295, indexed in Pubmed: 12826271. 39. Vatansever HS, Umur AS, Inan VS, et al. The effects of 23. Lee H, Nagele RG, Pietrolungo JF. Toxic and teratologic effects methotrexate on the development of neural tube defects in of caffeine on explanted early chick embryos. Teratology. the chick embryo. Turk J Vet Anim Sci. 2003; 27: 1119–1125. 1982; 25(1): 19–25, doi: 10.1002/tera.1420250104, indexed 40. Von Recklinghausen F. Untersuchungen über die Spi- in Pubmed: 7064111. na bifida. Arch Pathol Anat. 1886; 105: 243–373, 24. Manning N, Archer N. Treatment and outcome of serious doi: 10.1515/9783111485584. structural congenital heart disease. Semin Neonatol. 2001; 41. Wallace JL, Del Soldato P. The therapeutic potential of 6(1): 37–47, doi: 10.1053/siny.2000.0033, indexed in Pub- NO-NSAIDs. Fundam Clin Pharmacol. 2003; 17(1): 11–20, med: 11162284. indexed in Pubmed: 12588626. 25. Mete M, Gurcu B, Collu F, et al. Effects of lacosamide 42. Yakushiji T, Shirasaki T, Akaike N. Non-competitive inhibi- “a novel antiepileptic drug” in the early stages of chicken tion of GABAA responses by a new class of quinolones and embryo development. Childs Nerv Syst. 2016; 32(9): non-steroidal anti-inflammatoriesin dissociated frog sensory 1715–1719, doi: 10.1007/s00381-016-3181-4, indexed in neurones. Br J Pharmacol. 1992; 105(1): 13–18, indexed in Pubmed: 27473858. Pubmed: 1317734. 26. Mortimer EA. The puzzling epidemiology of neural tube defects. 43. Yerby MS. Clinical care of pregnant women with epilepsy: Pediatrics. 1980; 65(3): 636–638, indexed in Pubmed: 7360557. neural tube defects and folic acid supplementation. Epilep- 27. Nielsen GL, Sørensen HT, Larsen H, et al. Risk of adverse birth sia. 2003; 44 Suppl 3: 33–40, indexed in Pubmed: 12790884. outcome and miscarriage in pregnant users of non-steroidal 44. Yurt KK, Kaplan S, Kıvrak EG. The neuroprotective effect of anti-inflammatory drugs: population based observational melatonin on the hippocampus exposed to diclofenac so- study and case-control study. BMJ. 2001; 322(7281): 266– dium during the prenatal period. J Chem Neuroanat. 2018; –270, indexed in Pubmed: 11157526. 87: 37–48, doi: 10.1016/j.jchemneu.2017.05.006, indexed 28. Olesen C, Steffensen FH, Nielsen GL, et al. Drug use in first in Pubmed: 28576559. pregnancy and lactation: a population-based survey among 45. Zenker M, Klinge J, Krüger C, et al. Severe pulmonary hyper- Danish women. The EUROMAP group. Eur J Clin Pharmacol. tension in a neonate caused by premature closure of the duc- 1999; 55(2): 139–144, indexed in Pubmed: 10335909. tus arteriosus following maternal treatment with diclofenac: 29. Ostensen M, Ramsey-Goldman R. Treatment of inflam- a case report. J Perinat Med. 1998; 26(3): 231–234, indexed matory rheumatic disorders in pregnancy: what are the in Pubmed: 9773385.

The effects of leptin on F-actin remodelling in type 1 diabetes C. Guven1, E. Taskin1, H. Akcakaya2, R. Nurten2

1Nigde Omer Halisdemir University, Turkey 2Istanbul University, Turkey

[Received: 27 July 2018; Accepted: 11 September 2018]

Background: The aim of the current study is to investigate the effect of leptin on cytoskeleton structures in both in vivo and in vitro model of diabetes. Materials and methods: For in vivo studies, leptin in different doses (240, and 480 mg/kg) was injected to the diabetic rats after 1-week of streptozotocin (STZ, 55 mg/kg) treatment. Leptin levels were analysed in serum, liver, and pancreas samples. Hepatic and pancreatic F- and G-actin expressions were determined by Western blotting. For in vitro studies, hepatic and pancreatic primary cell lines were obtained from the control rats. To these cultures, STZ (15 and 30 mM), leptin (50, 60 and 100 ng/mL), and their combinations were applied for 1, 3, and 4 weeks. After the treatment period, F-actin was visualised by the Alexa-fluor fluorescent dye. Results: Streptozotocin decreased the G-actin in both tissues in vivo. However, leptin caused a dose-dependent increase in G-actin levels while F-actin decreased in both tissues. Moreover, leptin caused the perimembranous condensation of actin filaments and amelioration of F-actin structures in vivo. A dose-dependent corruption of F-actin filament structures was observed in leptin-treated primary cells in vitro, while STZ also caused corruption of these filaments. Co-exposure of STZ and leptin caused the amelioration of F-actin filaments, while the perimembranous condensation was also observed as was in vivo study. Conclusions: Leptin therapy could be a candidate for diabetes, but it should not be ruled out as being important the severity of diabetes and leptin doses. (Folia Morphol 2019; 78, 2: 314–324)

Key words: cytoskeleton, diabetes, F-actin, G-actin, hepatic cells, pancreatic b-cells, leptin therapy

INTRODUCTION called type 1 associated with the increased blood glu- Diabetes mellitus (DM) is one of common disease cose due to insufficient production of insulin in the worldwide. In 2014, it was reported that 7.5% of body. Insulin insensitivity of the cells causes type 2, women and 5% of men face to DM. The prevalence and the body produces sufficient insulin in this cases. is expected to reach 439 million by 2030 meaning Therefore, it is essential to find effective strate- of total 7.7% adult population between 20 and 79 gies to manage or control the prevalence of DM. years [43]. There are two types of DM. One of this Herein, leptin takes great attention in the control

Address for correspondence: Dr. C. Guven, Niğde Ömer Halisdemir University, Medicine Faculty, Biophysics Department, Niğde, Turkey, e-mail: [email protected]

314 C. Guven et al., F-actin remodelling by diabetes and leptin treatment

of hyperglycaemia and/or hyperglycemia-induced affected by leptin therapy in streptozotocin (STZ)- dysfunctions. -induced diabetic rats in both in vivo and in vitro Leptin, a protein hormone, affects the body weight, conditions. To our knowledge, it is the first study energy metabolism, reproductive system, and the neu- to investigate the effects of leptin on cytoskeleton roendocrine system [24]. Furthermore, the hormone structures of the STZ-treated cells and animals. plays a significant role in insulin-glucose homeostasis in the pancreas, liver, muscle, and hypothalamus as MATERIALS AND METHODS well. Insulin stimulates leptin secretion [38]. Leptin has Experimental animals acted centrally in the hypothalamus and peripherally Male Wistar rats weighing between 140 and in the pancreas, muscle, and liver. Leptin also improves 250 g were purchased from the Faculty of Veterinary glucose-insulin metabolism via attenuating glycemia, Medicine, Department of Reproduction and Artificial insulinemia, and insulin resistance Insemination, Istanbul University. The animals were To develop leptin-based therapies against diabe- individually housed in stainless cages at 20 ± 2°C, tes, and other insulin resistance syndromes, leptin’ 60–70% humidity with 12:12 h photoperiod. Water effect on the glucose-insulin homeostasis should be and food were given ad libitum. All processes were understood well [38]. There is a complicated rela- approved by the Istanbul University Animal Welfare tionship between insulin resistance and leptin. For Committee. example, leptin may increase insulin sensitivity and improve the glucose metabolism [38]. Also, a previous Induction of experimental diabetes study indicated that insulin resistance was amelio- Diabetes was induced by STZ. Streptozotocin was rated by administration of leptin in mice [6]. Hence, freshly prepared in a citrate buffer (50 mM, pH 4.5). it was proposed that the leptin plays a possible role Control group received only citrate buffer as a vehicle. in insulin resistance and it can improve insulin pro- 55 mg/kg of STZ was injected the intraperitoneally to duction [6]. Insulin resistance can be described as rats [42]. One week after STZ injection, animal having insufficiency in the insulin pathway at target tissues blood glucose levels ≥ 250 mg/dL was accepted as which generally response, like adipose, liver, skeletal, diabetic [27]. The blood glucose levels were measured and heart muscle [2]. by using a glucometer (OneTouch Ultra; LifeScan, The cytoskeleton is essential to keep cell shape Milpitas, CA, USA) from orbital vein blood samples. and signalling pathways for the beta (b) cells. The perturbation of actin structures might give rise to In vivo experimental design enhance b cell dysfunction, including inflammation and Four groups were randomly carried out as fibrosis [1]. Actin plays a vital function in morphology, (I) control (C group): non-diabetic rats treated with migration, adhesion, and permeability [33]. Actin is the vehicle, (II) diabetes (D group): diabetic rats abundant and highly conserved proteins among species. treated with the vehicle, (III) diabetes+leptin 240 It is a member of structural superfamily sugar kinases (DL240 group): diabetic rats treated with 240 ng/ hexokinases and Hsp70 proteins [10]. Actin is one of the kg dose of leptin, (IV) diabetes+leptin 480 (DL480 cytoskeleton element and has many essential roles in group): diabetic rats treated with 480 ng/kg dose of cells such as cell shape, polarity, endocytosis, contractil- leptin (modified from [47]). Both doses are physio- ity, motility, intracellular trafficking, etc. Actin has two logic doses of leptin [31]. Experimental animals were form of monomeric G-actin and filamentous polymeric treated with the leptin every other day for 1 week. F-actin [46]. G-actin can be converted to F-actin by the C group received intraperitoneally physiological presence of adenosine three phosphate (ATP), ions, saline as the same volume of DL group. Four ex- actin-binding proteins. F-actin and myosin interaction perimental animals were used for each group. After cause muscle contraction. The primary control factor of treatments periods, liver and pancreas were removed transformation is nucleotide hydrolysis [10]. Conversion from experimental animals for F- and G-actin analysis of G-actin to the filamentous form participates in patho- by Western blotting. Body weights were measured logical processes, such as cancer metastasis as well [50] at two time points: at the beginning of the first in- There is little knowledge available on the relation- jection and 24 h after the last injection. Then, body ship between insulin and leptin. The current study weight gain was calculated by subtracting from the was aimed to evaluate how F-actin structure can be last day to the first day of the experiment.

315 Folia Morphol., 2019, Vol. 78, No. 2

In vitro experimental design F-actin organisation For in vitro studies, primary hepatic (modified Immunofluorescence microscopy was used to from [28]) and pancreatic cells (modified from [21]) determine F-actin organisation in both in vitro (the were isolated from the corresponding tissues of the primary cell cultures) and in vivo (cryosections of C group. Hepatic primary cells were recognised as liver and pancreas) [17]. Briefly, primary culture cells parenchyma and Kupffer. The liver was removed were fixed with 4% paraformaldehyde in phosphate- aseptically and washed with Hanks’ balanced salt buffered saline (PBS) for 10 min at room temperature solution (HBSS). Then, the tissues were minced into and then permeabilised and blocked in PBS con- 1-mm3 pieces in HBSS. The cells were suspended taining 0.1% Triton X-100 and 5% foetal calf serum in William’s Medium E (WME) and centrifuged for for 30 min. Actin filaments were visualised using 5 min at 3000 g and the supernatant discarded. rhodamine-phalloidin or Alexa 497, and nuclei were Cells were dissolved in WME and transferred into labeled with 4’,6-diamidino–2–phenylindole dichlo- an incubator with 5% CO2 atmosphere at 37°C. For ride (DAPI) (Invitro Molecular Probes, OR, USA). All pancreatic acinar primary cell culture, each pancreas images were obtained using an Olympus BX51 Mi- was minced into 1 mm3 piece and incubated with croscope equipped with a DP72 camera controlled 2 mL of HBSS with 2 M EDTA at 37°C for 15 min. by Olympus DP2-TWAIN software. The chelated mixture was centrifuged for 5 min at 3000 g, and the supernatant was discarded. The Western blotting pellet was rinsed with 10 mL Ham’s F-12 medium, In vivo, F- and G-actin proteins were isolated centrifuged for 5 min at 3000 g, and the supernatant from the pancreas and liver tissues. Tissues were ho- discarded. The tissue pellet digested with 10 mL of mogenised by a MagNA lyser (Roche Diagnostics). 1 mg/mL collagenase type II, 1 mg/mL hyaluronidase Aliquots of liver and pancreas homogenates were and 20% heat-inactivated calf serum in Ham’s F-12 centrifuged at 3000 g for 15 min and 4500 g for medium at 37°C for 30 min in a shaking water bath. 20 min, respectively. F- and G-actin were separated Following centrifugation at 3000 g for 5 min, the by a kit (Cytoskeleton, Cat #BK037) according to supernatant discarded, pellet rinsed with 10 mL of the manufacturer’s instruction [26]. Then, 50 µg HBSS, and this procedure was replicated. The cells F- or G-actin were loaded to SDS–PAGE and trans- were rinsed with 5% heat-inactivated calf serum in ferred to a PVDF membrane (Millipore Corporation, Ham’s F-12 medium and centrifuged at 3000 g for Bedford, MA, USA) at 100 V for 45 min. The mem- 10 min. The pellet rinsed with 10 mL of 5% heat- branes were blocked by using 5% bovine serum inactivated calf serum in Ham’s F-12. Cells allocated albumin. Membranes were incubated with anti- to a well containing 2 mL of Waymouth’s serum-free actin rabbit polyclonal antibody (1/500) at +4 for medium and incubated at 37°C with 5% humidified overnight (Abcam, USA), and then incubated with ℃ CO2 for 48 h. a horseradish peroxidase-conjugated goat anti- The hepatic and pancreatic cells from non-dia- rabbit IgG antibody for 1 h at room temperature. betic control rats were divided into four groups as 5-bromo-chloro-3-indol phosphate (BCIP) was used control (c group), STZ (S group), leptin (L group), for visualisation for protein band. After membranes and STZ+leptin (SL group) for 1 and 4 weeks. Like were photographed with Canon camera, ImageJ hepatic cells, the pancreatic cells were also divided was used to analyse the band. into four groups. To observe the effect of severity of hyperglycaemia, the cell lines were treated with two Leptin analysing different doses of STZ as 15 or 30 mM. To determine Blood and homogenates of liver and pancreas the effect of different doses of leptin on cell lines, were centrifuged at 10,000 g for 15 min at +4°C. 50, 60 and 100 ng/mg of leptin were injected into Leptin levels were measured by ELISA using a com- the cells. The hepatic cells were treated with the mercial assay kit according to the manufacturer’s leptin for 1 and 4 weeks. The pancreatic cells were directions (Millipore, USA) [7]. treated with the leptin for 3 weeks. The data from DL groups in the pancreatic cells were not obtained Statistical analysis because of lower cell numbers although the medium The analyses of statistical were performed by SPSS was replaced 2 times a week. version 23.0. Data are expressed as mean ± standard

316 C. Guven et al., F-actin remodelling by diabetes and leptin treatment

error of the mean (SEM). One-way analysis of variance (one-way ANOVA) with post hoc Tukey for equal variance assumed and post hoc Dunnett’T3 for equal variance not assumed were used to compare mean between the groups. P < 0.05 was accepted as the statistical significance.

RESULTS Effect of leptin on body weight and blood Figure 1. Body weight gain; C — control; D — diabetes; glucose level DL240 — diabetic rats treated with 240 ng/kg dose of leptin; Diabetes decreased significantly body weight gain DL480 — diabetic rats treated with 480 ng/kg dose of leptin; *p < 0.01; different from C group. in D groups when compared to the C group (p < 0.05). The treatment of leptin failed to control body weight gain in diabetic rats. Interestingly, 240 ng/kg of leptin non-significantly worsen the body weight gain in DL240 group compared to the D group. But, there were no significant differences in body weight gain among D, DL240, and DL480 groups (Fig. 1). Blood glucose level is a primary parameter to determine diabetes. The delta was calculated by subtraction of glucose levels before and after experiments. As expected, all diabetes groups had high blood glucose level and the delta compared to C group Figure 2. Blood glucose level; C — control; D — diabetes; after STZ injection (p < 0.05). Leptin treatment did DL240 — diabetic rats treated with 240 ng/kg dose of leptin; not improve the high blood glucose when compared DL480 — diabetic rats treated with 480 ng/kg dose of leptin; *p < 0.01; different from C group. to D group (Fig. 2). The leptin levels were measured in serum, liver and pancreas samples. Diabetes did not influence the leptin levels in serum, pancreas, and liver in contrast to C group. Also, leptin treatment in all dose significantly increased the serum leptin levels compared to D group as expected (p < 0.01). The doses of leptin higher than 240 ng/kg significantly increased the serum leptin levels in contrast to D and DL240 groups (p < 0.01). However, the treatment of diabetic rats with leptin did not affect the level of leptin in liver Figure 3. Leptin levels in serum, pancreas, and liver; C — control; and pancreas when compared to D group (Fig. 3). D — diabetes; DL240 — diabetic rats treated with 240 ng/kg dose of leptin; DL480 — diabetic rats treated with 480 ng/kg dose of leptin; *p < 0.01; different from D group. The effect of leptin on actin structure The actin structure is so crucial for glucose metabolism. Therefore, it was evaluated the F-actin distribution of liver and pancreas tissues in vivo and on F-actin (Fig. 4A–D). In pancreatic tissue, while pancreas primary acinar, Kupffer and parenchymal diabetes also has a destructive impact on the F-actin also called as hepatocyte cells in vitro, as well. structure in D group compared to C group, the leptin For in vivo studies, diabetes induction gave rise treatment in all doses had the protective effect on to disrupt F-actin structure in liver tissues. The treat- F-actin structure compared to D group (Fig. 4E–H). ment of diabetic rats with 240 ng/kg dose of leptin To evaluate the effect of leptin treatment on differ- improved F-actin in DL240 group when compared to ent cell types, parenchymal and Kupffer cells were D group; however, as the dose of leptin (480 ng/kg, also isolated from rats in C, D and DL240 groups. DL480) increases, it worsens the effect of diabetes Based on cell types, the effects of leptin on F-actin

317 Folia Morphol., 2019, Vol. 78, No. 2

A B C D

E F G H

Figure 4. The leptin effect on F-actin distribution from liver and pancreas tissues; liver tissues were as from A to D, although pancreas tissues were from E to H; A and E. Control; B and F. Diabetes; C and G. Diabetic rats treated with 240 ng/kg dose of leptin; D and H. Diabetic rats treated with 480 ng/kg dose of leptin.

A B C

D E F

Figure 5. F-actin distribution of primary liver cells at parenchymal (upper line) and Kupffer cells (below line); red colour represents F-actin, blue represents the nucleus by DAPI at ×100 magnification. Parenchymal cells: A and D. Control; B and E. Diabetes; C and F. Diabetic rats treated with 240 ng/kg dose of leptin.

showed differences. In the parenchymal cells, F-actin structure in the isolated liver cells. In pancreas cells, disappeared in the cytosol (Fig. 5B, C) whereas slight the detrimental effect of leptin on F-actin was only destruction was observed in Kupffer cells (Fig. 5E, F). observed at the dose of 60 ng/mL (Fig. 7C, D). In the Isolated primary cell cultures were preferred to present study, the effect of leptin on F-actin structure reduce the number of animals used in the experiment. in STZ-induced pancreatic cells (SL group) were not In primary liver cells, only 100 ng/mL dose of leptin evaluated due to high toxicity of STZ on pancreatic (higher dose) was destructive on F-actin structure cells (Fig. 7E, F) because the cells were exposed to (Fig. 6G–I); therefore, the treatment of the cell with STZ at the beginning of the study and F-actin de- the leptin was not continued after 3 weeks. When struction was measured after three and/or 4 weeks. compared to 100 ng/mL dose of leptin, 50 ng/mL That indicates the highly destructive effects of STZ dose of leptin was not so destructive to the F-actin on the pancreas.

318 C. Guven et al., F-actin remodelling by diabetes and leptin treatment

A B C

D E F

G H I

Figure 6. The leptin effect on F-actin distribution after 3 weeks treated on primary liver cells; A, B, C. Control; D, E, F. Cells treated with 50 ng/mL dose of leptin; G, H, I. Cells treated with 100 ng/mL dose of leptin (Panels C, F, I is at 20 magnification, and the other is at 100 magnification).

A B

C D

E F

Figure 7. The leptin effect on F-actin distribution at primary pancreas cells at 100 magnification; A, B. Control; C, D. Cells treated with 60 ng/mL dose of leptin; E , F. Cells treated with 30 doses of mM streptozotocin (magnification × 100).

319 Folia Morphol., 2019, Vol. 78, No. 2

Also, the effects of leptin on F-actin structures in the primary hepatic cells were studied based on the severity of hyperglycaemia or diabetic condition. At the mild diabetic condition (15 mM STZ), filopodia formation by hyperglycaemia (Fig. 8B1, 8B2) was reversed by leptin treatment without any protection on the F-actin structure after 4 weeks in both at parenchymal and Kupffer (Fig. 8D1, 8D2). Leptin failed to improve F destruction and filopodia formation (Fig. 8E1, 8E2); both induced by severe diabetic condition (30 mM STZ; Fig. 8C1, 8C2).

The effect of leptin treatment on F- and G-actin protein expression Leptin treatment showed a protective effect on disproportional F- and G-actin protein expressions compared to diabetes groups in both liver and pancreas tissues in vivo (Fig. 9A–D).

DISCUSSION Cytoskeleton plays a vital role in the cell morphology and the elasticity [14]. Insulin can tonically regulate actin filament network and microtubule as well [35, 51]. Actin network mediates glucose transport through insulin action [51]. Actin phosphorylation alters cytoskeleton re-orientation via polymerisation and depolymerisation [14]. Insulin stimulates leptin secretion [38]. Leptin is considered as an anti-diabetic hormone. It plays a significant role in glucose- ho meostasis in the pancreas, liver, muscle, and hypo- Figure 8. The leptin effect on F-actin distribution at parenchymal thalamus. Also, it blocks glucose production in the

(A1, B1, C1, D1, E1) and Kupffer (A2, B2, C2, D2, E2) cells after liver and enhances glucose consumption in the tissue 4 weeks treatment; A1, A2. Control; B1, B2. Cells treated with [37]. It was reported that the effects of leptin were 15 mM of streptozotocin; C , C . Cells treated with 30 mM of 1 2 mediated via the JAK/STAT signalling pathways. b-cells streptozotocin, D1 and D2) 15 mM exposed cells treated with 60 ng leptin E1 and E2) 30 mM exposed cells treated with 60 ng leptin. are hyperpolarised via ATP sensitive potassium chan-

Figure 9. The alternation of F- and G-actin protein expression in pancreas and liver tissue; 1. Control; 2. Diabetes; 3. Diabetic rats treated with 240 ng/kg dose of leptin; 4. Diabetic rats treated with 480 ng/kg dose of leptin; 5. Diabetic rats treated with 720 ng/kg dose of leptin; 6. Diabetic rats treated with 960 ng/kg dose of leptin; A. F-actin of the liver; B. G-actin of the liver; C. F-actin of the pancreas; D. G-actin of the pancreas.

320 C. Guven et al., F-actin remodelling by diabetes and leptin treatment

nel (KATP) opening by leptin; actin cytoskeleton reor- depolarisation [8]. Finally, insulin secretion occurs. ganises through the PI3K pathway [34]. STZ causes The second phase is essential to maintain the insulin hyperphagia, hyperleptinaemia, hypoinsulinaemia, level in plasma [8]. This phase is associated with the and hyperglycaemia [40]. Insulin can secrete leptin reserved pool, which is far away from the plasma from adipose tissue, controversy; leptin suppresses membrane [12]. Reserved insulin granules must be insulin secretion from pancreatic b-cells by activation transported to the plasma membrane [8]. F-actin is of KATP. Leptin is reported to increase KATP trafficking vital to maintaining basal plasma insulin levels [22]. to enhance the channel density on the membrane by Therefore, actin cytoskeleton plays a significant role strengthening F-actin depolymerisation. The traffick- in both phases of insulin secretion [12]. In the current ing of KATP by leptin is mediated through AMPK and study, diabetes caused a decrease of G-actin, while protein kinases-A (PKA) [4]. any effect was observed on F-actin in vivo. This effect In the present study, diabetes did not affect the can be explained that glucose enhances F- or G-actin serum, liver and pancreatic leptin levels. However, ratio which means that hyperglycaemia can inhibit leptin treatment elevated only the serum leptin levels G-actin disassociation by partial stimulation of cofilin in diabetic rats without altering the pancreatic and phosphorylation. Cofilin belongs to one of actin- liver levels. Our results are an inconsistency with the depolymerising factor family and can be inactivated previous study, which indicates that increased plasma with phosphorylation by LIM kinase, causing F-actin insulin levels due to increased insulin resistance could stabilisation [19]. Rho-A could participate in this path- elevate the pancreatic leptin levels [41]. The result of way [48]. Diabetes is well known as an autoimmune the present study demonstrates that leptin did not im- disease. Macrophages inhibit Rho pathways caused prove blood glucose level in diabetes. Inconsistently, to disrupt of the cytoskeleton from subcortical actin it was reported that a high dose of leptin ameliorated ring to near the cell periphery [13]. G-actin is needed the both of hyperlipidaemia and hyperglycaemia [11, for F-actin resembling. Therefore; F-actin maintains its 20]. Co-treatment of leptin and insulin could be more proper function. It was proposed that attenuation of useful to prevent hyperglycaemia than insulin only G-actin participate in the development of diabetes. therapy, which reduces blood glucose to undesirable Furthermore, F-actin reorientation was observed levels. One advantage of leptin is without to gain the at the plasma membrane surrounding in the weight in the treatment of type 1 DM while making current study. F-actin can be depolymerised and scav- a balance of blood glucose. An extensive research enged by the action of gelsolin, which is one of the report that type 1 DM reduces body weight [11, 23] free calcium regulators, resulting in blood glucose at- as reported in the present study. It is probably that tenuation via the blockage of insulin vesicle transport type 1 DM causes dehydration, declining glucose uti- to the membrane [10]. It was suggested that a high lisation and protein anabolism [10]. Leptin provides plasma glucose level could induce actin cytoskeleton a link between body weight and food intake [5, 49]. disruption [22]. This is asserted to relate apoptotic A correlation between serum leptin levels, plasma cell death [25]. That is why F-actin could release from insulin, and body mass index has been reported [49]. cell to plasma when cell injury is occurred [10]. Leptin Leptin in high concentrations causes insulin resistance is also reported to activate LIM kinase for phospho- [49]. In the present study, diabetic rats lose the body rylation of cofilin, resulting in accumulation F-actin weight as stated in the previous study [9]. While leptin and depletion of G-actin. The other effect of leptin treatment seems to be protective, it was considered on actin is up-regulation of alpha-skeletal actin and that weight gain in leptin-treated groups looks like myosin light chain-2 (MLC-2) [16]. developing type 2 DM as was explained in below. Lep- Insulin and leptin have anti-apoptotic and prolif- tin antagonizes the effect of insulin on glucokinase erative effects [39]. Furthermore, leptin was shown and phosphoenolpyruvate carboxylase enzymes to to have a positive impact on glucose metabolism, decrease blood glucose levels in the liver. A decrease and it was an excellent adjuvant candidate for insulin in leptin efficiency is related to insulin resistance therapy [44]. Leptin enhanced the insulin sensitivity in type 2 DM [36]. Insulin secretion occurs in two in rats with type 2 DM and patients with type 2 as phases. One phase called fast secretion requires glu- well. A single administration of leptin ameliorated the cose entry to the b-cells, and this leads to the closure mortality related to high blood glucose and insulin of ATP sensitive potassium channel (KATP) resulting in independent insulin deficiency [15]. Leptin is shown

321 Folia Morphol., 2019, Vol. 78, No. 2

its anti-diabetic effect via increasing insulin receptor has been presented in the current study. However, sensitivity. However, the involvement of F-actin in leptin could not significantly improve F-actin orienta- the mechanism of leptin on type-1 DM is not clearly tion in vitro in different cell types. A slight improve- evaluated. Diabetes and leptin therapy showed their ment of F-actin orientation was observed in Kupffer destructive effect on F-actin according to cell type in cells compared to parenchymas. Leptin at the higher the current study in vitro. F-actin disappeared from doses can contribute to the destructive effects of dia- cytosol center and accumulated within the periphery betes. Filopodia formation can be reversed by leptin of the cell membrane in primary parenchymal cells. treatment without changing the F-actin destruction On the contrary, F-actin clustered within the cyto- according to the severity of diabetic conditions. In sol center as multicentered radiant groups in Kupper conclusion, leptin treatment could be useful when primary cells. In primary pancreas cells, the destruc- the severity of diabetes and leptin doses are taken tion process in actin filament is similar to Kuppfer into consideration. primary cell. It was shown that leptin binding to its receptor (Ob-Rb) phosphorylates JAK2 and transforms Acknowledgements

PIP2 into PIP3. This destroys the F-actin into the G-actin This study was supported by the Istanbul Univer- [18]. One study is reported that RhoA and ROCK are sity Research Fund for (Project No: 8903). mediated leptins effect on F-actin distribution. Leptin is suggested to enhance the expression and activities REFERENCES of GLUT2 and GLUT5 transporters [30]. Leptin has 1. Arous C, Halban P. The skeleton in the closet: actin cy- been reported to improve insulin sensitivity in both toskeletal remodeling in b-cell function. Am J Physiol En- healthy and diabetic animals. However, the hormone docrinol Metab. 2015; 309(7): E611–E620, doi: 10.1152/ ajpendo.00268.2015. is also said to destroy insulin effect on hepatocytes. 2. Beale EG. Insulin signaling and insulin resistance. Leptin infusion is an essential factor for disclosing its J Investig Med. 2013; 61(1): 11–14, doi: 10.2310/ effect on tissue. Leptin is suggested to decrease blood JIM.0b013e3182746f95, indexed in Pubmed: 23111650. glucose and insulin level without altering body weight 3. Ceddia RB, Koistinen HA, Zierath JR. Analysis of paradoxical observations on the association between leptin and insulin of animals by subcutan infusion. Leptin intravenously resistance. Faseb J. 2002; 16: 1163–1176. injection boosts inulin sensitivity in healthy animals, 4. Chen PC, Kryukova YN, Shyng SL. Leptin regulates KATP increase glucose consumption in diabetic animals. channel trafficking in pancreatic b-cells by a signaling STZ destroy pancreatic b-cell and decrease adipose mechanism involving AMP-activated protein kinase (AMPK) and cAMP-dependent protein kinase (PKA). J Biol and plasma leptin levels. Therefore, leptin therapy can Chem. 2013; 288(47): 34098–34109, doi: 10.1074/jbc. mimic insulin action by insulin-independent and an M113.516880, indexed in Pubmed: 24100028. insulin-sensitising mechanism in STZ treated animals. 5. Considine R, Sinha M, Heiman M, et al. Serum immu- noreactive-leptin concentrations in normal-weight and Leptin has been shown to activate KATP, resulting in decreasing insulin secretion [3]. obese humans. N Engl J Med. 1996; 334(5): 292–295, doi: 10.1056/nejm199602013340503. Cell skeleton is a dynamic and sensitive system. It is 6. Cortés VA, Cautivo KM, Rong S. Leptin ameliorates insulin pointed out as an important morphologic parameter resistance and hepatic steatosis in Agpat2(−/−) lipodys- in the study of changes that took place in the cell trophic mice independent of hepatocyte leptin receptors. after getting into reaction with substance [29]. Some J Lipid Res. 2014; 55: 276–288. 7. Coupe B, Grit I, Hulin P. Postnatal growth after intrauterine vesication is formed within the whole cell during the growth restriction alters central leptin signal and energy process of cell skeleton’s restructuring. This incidence homeostasis. PLoS One. 2012; 7: e30616. of vesication that took place outside the cell goes 8. Dehghany J, Hoboth P, Ivanova A, et al. A Spatial Model along both with the distribution of microtubules [45] of Insulin-Granule Dynamics in Pancreatic b-Cells. Traffic. and actin filaments’ restructuring. Actin polymerisa- 2015; 16(8): 797–813, doi: 10.1111/tra.12286, indexed in Pubmed: 25809669. tion occurs at G1 and G2/M phases of the cell cycle 9. Denroche H, Quong W, Bruin J, et al. Leptin administra- [32]. In actin polymerisation, F- or G-actin rates show tion enhances islet transplant performance in diabetic cellular separation (differentiation) [32]. mice. Diabetes. 2013; 62(8): 2738–2746, doi: 10.2337/ db12-1684. CONCLUSIONS 10. Dominguez R, Holmes KC. Actin structure and function. Annu Rev Biophys. 2011; 40: 169–186, doi: 10.1146/ The in vivo protective effect of leptin treatment on annurev-biophys-042910-155359, indexed in Pub- the F-actin structure in both liver and pancreas tissues med: 21314430.

322 C. Guven et al., F-actin remodelling by diabetes and leptin treatment

11. Emekli-Alturfan E, Kasikci E, Yarat A. Tissue factor activi- Res. 2014; 2014: 152075, doi: 10.1155/2014/152075, ties of streptozotocin induced diabetic rat tissues and the indexed in Pubmed: 25478578. effect of peanut consumption. Diabetes Metab Res Rev. 26. Kim HR, Gallant C, Leavis PC. Cytoskeletal remodeling in 2007; 23(8): 653–658, doi: 10.1002/dmrr.757, indexed in differentiated vascular smooth muscle is actin isoform Pubmed: 17595667. dependent and stimulus dependent. Am J Physiol Cell 12. Fan F, Ji C, Wu Y, et al. Dynamin 2 regulates biphasic insulin Physiol. 2008; 295: C768–C778. secretion and plasma glucose homeostasis. J Clin Invest. 27. King A. The use of animal models in diabetes research. Br 2015; 125(11): 4026–4041, doi: 10.1172/jci80652. J Pharmacol. 2012; 166(3): 877–894, doi: 10.1111/j.1476- 13. Fan H, Patel VA, Longacre A. J.S. Levine, Abnormal regula- 5381.2012.01911.x. tion of the cytoskeletal regulator Rho typifies macrophages 28. Klaunig J, Goldblatt P, Hinton D, et al. Mouse liver cell culture. of the major murine models of spontaneous autoimmun- In Vitro. 1981; 17(10): 926–934, doi: 10.1007/bf02618289. ity, J Leukoc Biol. 2006; 79: 155–165. 29. Kohler M, Aufderheide M, Ramm D. Method for the de- 14. Fofana B, Yao XH, Rampitsch C, et al. Prenatal alcohol scription of differences in the filamentous structure of the exposure alters phosphorylation and glycosylation of cytoskeleton in cultured cells. Toxicol Lett. 1994; 72(1-3): proteins in rat offspring liver. Proteomics. 2010; 10(3): 33–42, indexed in Pubmed: 8202948. 417–434, doi: 10.1002/pmic.200800969. 30. Le Drean G, Segain JP. Connecting metabolism to intestinal 15. Fujikawa T, Chuang JC, Sakata I, et al. Leptin therapy barrier function: The role of leptin. Tissue Barriers. 2014; improves insulin-deficient type 1 diabetes by CNS- 2: e970940. dependent mechanisms in mice. Proc National Academy 31. Londoño M, Tellez L, Gutierrez R. Physiological low doses Scien. 2010; 107(40): 17391–17396, doi: 10.1073/ of leptin and cholecystokinin induces body weight-loss pnas.1008025107. in juvenile and lean, but not in adult-obese rats. Int 16. Ghantous CM, Azrak Z, Hanache S. Differential Role of J Peptide Res Therap. 2011; 18(1): 77–88, doi: 10.1007/ Leptin and Adiponectin in Cardiovascular System. Int s10989-011-9281-9. J Endocrinol. 2015; 2015: 534320. 32. Lu QY, Jin YS, Zhang Q, et al. Ganoderma lucidum extracts 17. Guven C. The effect of Diazoxide on norepinephrine- inhibit growth and induce actin polymerization in blad- induced cardiac hypertrophy, in vitro. Cell Mol Biol der cancer cells in vitro. Cancer Lett. 2004; 216(1): 9–20, (Noisy-le-grand). 2018; 64(10): 50–54, doi: 10.14715/ doi: 10.1016/j.canlet.2004.06.022, indexed in Pubmed: cmb/2018.64.10.8. 15500944. 18. Harvey J, Ashford MLJ. Leptin in the CNS: much more 33. Maria Z, Yin W, Rubenstein DA. Combined effects of than a satiety signal. Neuropharmacology. 2003; 44(7): physiologically relevant disturbed wall shear stress and 845–854, indexed in Pubmed: 12726816. glycated albumin on endothelial cell functions associated 19. Hien TT, Turczyńska KM, Dahan D, et al. Elevated glucose with inflammation, thrombosis and cytoskeletal dynamics. levels promote contractile and cytoskeletal gene expres- J Diabetes Investig. 2014; 5(4): 372–381, doi: 10.1111/ sion in vascular smooth muscle via rho/protein kinase jdi.12162, indexed in Pubmed: 25411596. c and actin polymerization. J Biol Chem. 2016; 291(7): 34. Marroqui L, Gonzalez A, Neco P. Role of leptin in the pan- 3552–3568, doi: 10.1074/jbc.M115.654384, indexed in creatic beta-cell: effects and signaling pathways. J Molec Pubmed: 26683376. Endocrinol. 2012; 49: R9–17. 20. Ishii S, Kamegai J, Tamura H, et al. Role of ghrelin in strep- 35. Millioni R, Iori E, Puricelli L, et al. Abnormal cytoskel- tozotocin-induced diabetic hyperphagia. Endocrinology. etal protein expression in cultured skin fibroblasts from 2002; 143(12): 4934–4937, doi: 10.1210/en.2002-220612. type 1 diabetes mellitus patients with nephropathy: 21. Justice JD, Brannon PM. Synthesis of amylase by cultured a proteomic approach. Proteomics Clin Appl. 2008; 2(4): rat pancreatic acinar cells: effects of antecedent diet. 492–503, doi: 10.1002/prca.200780112, indexed in Pub- J Nutr. 1989; 119(5): 805–812, doi: 10.1093/jn/119.5.805, med: 21136853. indexed in Pubmed: 2470881. 36. Moon HS, Dalamaga M, Kim SY, et al. Leptin’s role in 22. Kalwat MA, Thurmond DC. Signaling mechanisms of glu- lipodystrophic and nonlipodystrophic insulin-resistant and cose-induced F-actin remodeling in pancreatic islet b cells. diabetic individuals. Endocrine Rev. 2013; 34(3): 377–412, Exp Mol Med. 2013; 45: e37, doi: 10.1038/emm.2013.73, doi: 10.1210/er.2012-1053. indexed in Pubmed: 23969997. 37. Naito M, Fujikura J, Ebihara K, et al. Therapeutic impact 23. Kaya ST, Bozdogan O, Ozarslan TO, et al. The protec- of leptin on diabetes, diabetic complications, and lon- tion of resveratrol and its combination with glibencla- gevity in insulin-deficient diabetic mice. Diabetes. 2011; mide, but not berberine on the diabetic hearts against 60(9): 2265–2273, doi: 10.2337/db10-1795, indexed in reperfusion-induced arrhythmias: the role of myocardial Pubmed: 21810600. K channel. Arch Physiol Biochem. 2019; 125(2): 114–121, 38. Paz-Filho G, Mastronardi C, Wong ML, et al. Leptin therapy, doi: 10.1080/13813455.2018.1440409, indexed in Pub- insulin sensitivity, and glucose homeostasis. Indian J Endo- med: 29457517. crinol Metab. 2012; 16(9): 549–555, doi: 10.4103/2230- 24. Kelesidis T, Kelesidis I, Chou S. Narrative review: the role of 8210.105571. leptin in human physiology: emerging clinical applications. 39. Rubenstein JH, Morgenstern H, McConell D, et al. Associa- Ann Internal Med. 2010; 152: 93–100. tions of diabetes mellitus, insulin, leptin, and ghrelin with 25. Khatri N, Sagar A, Peddada N, et al. Plasma gelsolin levels gastroesophageal reflux and Barrett’s esophagus. Gastro- decrease in diabetic state and increase upon treatment enterology. 2013; 145(6): 1237–12344.e1, doi: 10.1053/j. with F-actin depolymerizing versions of gelsolin. J Diabetes gastro.2013.08.052, indexed in Pubmed: 23999171.

323 Folia Morphol., 2019, Vol. 78, No. 2

40. Sah S, Singh B, Choudhary S, et al. Animal models of in- 46. Visa N, Percipalle P. Nuclear functions of actin. Cold Spring sulin resistance: A review. Pharmacological Reports. 2016; Harb Perspect Biol. 2010; 2(4): a000620, doi: 10.1101/ 68(6): 1165–1177, doi: 10.1016/j.pharep.2016.07.010. cshperspect.a000620, indexed in Pubmed: 20452941. 41. Shoji T, Nishizawa Y, Emoto M, et al. Renal function and 47. Wagoner B, Hausman DB, Harris RBS. Direct and indirect insulin resistance as determinants of plasma leptin lev- effects of leptin on preadipocyte proliferation and differ- els in patients with NIDDM. Diabetologia. 1997; 40(6): entiation. Am J Physiol Regul Integr Comp Physiol. 2006; 676–679, doi: 10.1007/s001250050733, indexed in 290(6): R1557–R1564, doi: 10.1152/ajpregu.00860.2005, Pubmed: 9222647. indexed in Pubmed: 16424081. 42. Soetikno V, Sari FR, Veeraveedu PT, et al. Curcumin 48. Wang X, Tang P, Guo F, et al. RhoA regulates Activin ameliorates macrophage infiltration by inhibiting NF-kB B-induced stress fiber formation and migration of bone activation and proinflammatory cytokines in streptozo- marrow-derived mesenchymal stromal cell through distinct tocin induced-diabetic nephropathy. Nutr Metab (Lond). signaling. Biochim Biophys Acta Gen Subj. 2017; 1861(1 2011; 8(1): 35, doi: 10.1186/1743-7075-8-35, indexed in Pt A): 3011–3018, doi: 10.1016/j.bbagen.2016.09.027, Pubmed: 21663638. indexed in Pubmed: 27693126. 43. Trevelin SC, Carlos D, Beretta M, et al. Diabetes mellitus 49. Widjaja A, Stratton IM, Horn R, et al. UKPDS 20: plasma and sepsis: A Challenging Association. Shock. 2017; 47(3): leptin, obesity, and plasma insulin in type 2 diabetic 276–287, doi: 10.1097/SHK.0000000000000778, indexed subjects. J Clin Endocrinol Metab. 1997; 82(2): 654–657, in Pubmed: 27787406. doi: 10.1210/jc.82.2.654. 44. Tucholski K, Otto-Buczkowska E. The role of leptin in the 50. Yu M, Yuan X, Lu C, et al. mDia1 senses both force and regulation of carbohydrate metabolism. Endokrynol Pol. torque during F-actin filament polymerization. Nature 2011; 62: 258–262. Communications. 2017; 8: 1650. 45. Uzdensky A, Kolpakova E, Juzeniene A, et al. The effect of 51. Zhu X, Hu R, Brissova M, et al. Microtubules negatively sub-lethal ALA-PDT on the cytoskeleton and adhesion of regulate insulin secretion in pancreatic b cells. Devel- cultured human cancer cells. Biochim Biophys Acta. 2005; opmental Cell. 2015; 34(6): 656–668, doi: 10.1016/j. 1722(1): 43–50, doi: 10.1016/j.bbagen.2004.11.011. devcel.2015.08.020.

Evaluation of the factors associated with sublingual varices: a descriptive clinical study N. Akkaya1, D. Ölmez1, G. Özkan2

1Department of Dentomaxillofacial Radiology, Faculty of Dentistry, Hacettepe University, Ankara, Turkey 2Department of Dentomaxillofacial Radiology, Faculty of Dentistry, Aydin Adman Menderes University, Aydin, Turkey

[Received: 2 September 2018; Accepted: 28 September 2018]

Background: Age is considered as an important factor for the development of sublingual varices (SV). It has been suggested that some other conditions such as gender, systemic diseases, smoking, denture wearing may play a role in aetiology. However, there are a limited number of studies investigating these associations. This study was perform to evaluate association between SV and the conditions which are described as possible risk factors previously. Materials and methods: A total of 691 patients (470 females, 221 males) who attended for comprehensive clinical examination were included in the study. Age, gender, systemic health, venous varix of the lower extremities, smoking status, denture wearing were recorded during the history taking. SV were classified into two categories: stage 0 (few or none visible) and stage 1 (moderate or severe). Tongue photographs were taken from a group of these patients. For the evaluation of intra-observer reliability, 60 photographs of tongue were re-evaluated by the same observer. Intra-observer reliability was evaluated using Kappa statistics. Pearson c2 test and Fisher’s exact test were used to assess SV in relation to each variable, and variables showing associations with p < 0.05 were selected for the multivariable analysis, then logistic regression analysis was applied. Results: Kappa value of intra-observer reliability was 0.91. SV were significantly associated with age (odds ratio [OR] = 1.08, p = 0.000), hypertension (OR = 2.3, p = 0.007) and denture wearing (OR = 2.17, p = 0.02). Conclusions: The presence of SV is associated with hypertension and denture wearing as well as aging. More detailed studies are needed to prove causative relations between SV and systemic diseases. (Folia Morphol 2019; 78, 2: 325–330)

Key words: varicose veins/aetiology, age factor, dentures/adverse effect, hypertension/complications, tongue/blood supply, odds ratio

INTRODUCTION sublingual varicosities, phlebectasia linguae, caviar Varices are abnormally dilated tortuous veins or tongue, spots or lesions [18]. Although oral varices venules, artery or lymphatic vessels; however, the are most frequently seen on the ventral surface of term refers venous lesions for the oral cavity [10, 18]. the tongue, they may also occur on the lips and floor Sublingual varices (SV) involving the lateral branches of the mouth. Less frequently a varix, which is a fo- of vena profunda linguae are common type of ac- cal dilatation of vein, may occur on buccal mucosa, quired vascular lesions [8, 26]. This benign condition labial commissures, and sometimes soft and hard has been given various names including lingual or palate [18].

Address for correspondence: N. Akkaya, Assist. Prof., Hacettepe University, Faculty of Dentistry, Department of Dentomaxillofacial Radiology, Sıhhiye Campus, Ankara, Turkey, tel: +90 312 305 22 05, e-mail: [email protected]

325 Folia Morphol., 2019, Vol. 78, No. 2

The clinical features of SV are usually multiple, 05) and was performed in accordance with the Helsin- irregular, blue-purple, elevated or papular blebs on ki Declaration. This cross-sectional clinical study was the ventral and lateral border of the tongue. They are conducted on patients who attended the Department usually distributed from the posterior part bilaterally of Dentomaxillofacial Radiology, Hacettepe Universi- to the apex of tongue. The lesions are usually asymp- ty for comprehensive clinical examination between tomatic and frequently encountered during routine April 2016 and October 2016. Before starting the clinical examination [22]. study, tongue examination of 50 patients was made Since incidence increases with aging, age is con- by 2 observers simultaneously as an exercise for the sidered an important factor. Kleinman [16] stated that differentiation between few visible and moderate the condition is related to the aging process, and prior lesions as classified in the previous study [12]. All of to the fifth decade, varices may be an indication of the patients who agreed to participate in the study premature aging. Chronic cough [3], venous insuffi- were examined by an investigator. Patients who could ciency [8], cardiovascular/cardiopulmonary diseases not clearly answered the questions about medical [11, 12, 19], portal hypertension [14], diabetes mel- history were excluded from the study. Prior to clinical litus [29], chronic vitamin C deficiency [2], smoking examination, demographic characteristics and his- [12], denture wearing [1] have been suggested as torical information including age, gender, systemic predisposing factors for SV. There are limited number health (hypertension, other CVD, diabetes mellitus), of studies which have been examined the association smoking status (never, current, former), and denture between SV and conditions listed above [1–3, 8, wearing were recorded for all subjects. At the end of 11, 12, 16, 19]. The role of the systemic health on the clinical examination, SV were classified. In order the development of SV is controversial. Ettinger and to examine ventral surface and lateral borders of the Manderson [8] stated that a person with varicose tongue and floor of the mouth, patients were asked veins of the legs is likely to have lingual varicosities, to move the tongue upward and turn it to the left but reverse is not necessarily true. Bean [3] reported and then to the right. Using the criteria described that there was no connection between SV and sys- by Hedström and Bergh [12], SV classified into two temic diseases. Lynge Pedersen et al. [19] found an categories: grade 0 (few or none visible) (Fig. 1A) and association between sublingual varicosities and car- grade 1 (moderate or severe) (Fig. 1B). To evaluate diovascular diseases (CVD) and medication. Recently, intra-observer reliability, tongue photographs were Hedström et al. [11] found an association between taken from a group of the patients. Sixty photo- SV and hypertension; Al-Shayyab and Baqain [1] de- graphs were randomly selected by other investigator scribed that old age, female gender, denture wearing and re-examined by the same investigator 6 months and CVD are the risk factors for SV. These results arises after the initial examination. Before analysing data, a question whether cardiovascular system diseases as intra-observer reliability was evaluated using Kappa a whole or only hypertensive disease is related to SV. statistic to ensure that there was no significant meas- Lazos et al. [18] emphasized that more detailed stud- urement error. Kappa value of intra-observer reliability ies are needed to explain relations between varices was 0.91. The agreement of measurements was found and systemic diseases, which may contribute to early to be almost perfect according to the categorisation diagnosis of possible medical conditions during in- of Kappa statistic proposed by Fleiss [9]. traoral examination. Therefore, the purpose of this study was to investigate some potential risk factors Statistical analysis suggested previously, including hypertensive disease, Statistical analysis was performed by IBM SPSS Sta- other diseases of the circulatory system, diabetes tistics, Version 20.0 (Armonk, NY: IBM Corp.). Prelimi- mellitus, varicose veins of lower extremities, smoking, nary analysis was carried out using Pearson c2 test and and denture wearing for the development of SV in Fisher’s exact test to assess SV in relation to each varia- patients through a wide age range. ble, and variables showing associations with p < 0.05 were selected for the multivariable analysis. Logistic Materials and methods regression was used to assess the association of SV with The investigational protocol for the study was ap- age, smoking, systemic diseases (yes/no), hypertension, proved by the Non-Interventional Clinical Researches other CVD, diabetes mellitus, varicose veins of lower Ethics Board at the Hacettepe University (GO 16/647- extremities, and denture wearing. Odds ratios (OR) were

326 N. Akkaya et al., Factors associated with sublingual varices

A B

Figure 1. Sublingual varices classified into two categories using the criteria described by Hedström and Bergh [12]. Clinical appearances of the stages are shown in panel A, grade 0 (few or none visible) and in panel B, grade 1 (moderate or severe).

Table 1. Distribution of sublingual varices (SV) by genders; *p > 0.05 SV Female (n = 470) Male (n = 221) Total (n = 691) Grade 0 345 (73.4%) 160 (72.4%) 505 (73.1%) Grade 1 125 (26.6%) 61 (27.6%) 186 (26.9%)

used to report the size of the associations. P-values less than 0.05 were considered to be statistically significant.

RESULTS Seven hundred and fourteen patients were invited to participate in the study. The patients who refused to participate in the study (n = 7), and have unclear medical history (n = 16) were excluded. A total of 691 patients (470 females, 221 males) were included in the study. The patients were 18–88 years of age (mean 38.88 ± 16 years). The percentage of females (68%) was higher than that of males (p < 0.001). The prevalence of SV was 26.9%. Distribution of SV according to gender is presented in Table 1. There was no statistically significant gender difference (p > 0.05). Figure 2 shows distribution of SV according to the age groups. There was a significant relation Figure 2. Distribution of sublingual varices according to age groups. between age and SV (p < 0.001). The prevalence increases apparently for the patients older than 50. The Logistic regression model was constructed to an- prevalence of SV according to risk factors evaluated alyse the variables associated with higher probability and their significance values are presented in Table 2. of having SV. The independent variables included in

327 Folia Morphol., 2019, Vol. 78, No. 2

Table 2. Prevalence of sublingual varices according to risk factors

Variable N Sublingual varices P Systemic diseases 244 114 (46.7%) < 0.001 Cardiovascular diseases 62 35 (56.5%) < 0.001 Hypertension 86 63 (73.3%) < 0.001 Diabetes mellitus 45 27 (60.0%) < 0.001 Venous varix of the lower extremities 152 61 (40.1%) < 0.001 Denture wearing: 74 54 (73.0%) < 0.001 Smoking status: < 0.001 Current 173 38 (22.0%) Former 84 40 (47.6%) Never 434 108 (24.9%)

*Pearson c2 test with Fisher’s exact test

Table 3. The influence of risk factors for sublingual varices by multivariablelogistic regression model Variable B SE OR P 95% CI Age 0.077 0.008 1.080 < 0.001 1.063–1.098 Hypertension 0.837 0.312 2.309 0.007 1.254–4.252 Denture wearing 0.777 0.335 2.175 0.020 1.128–4.193

CI — confidence interval; OR — odds ratio; SE — standard error

the multivariable analysis were: age, smoking, den- accepted as SV in the previous studies [13, 19]. This ture wearing, systemic diseases, diabetes mellitus, classification is not in detail and it may be difficult hypertension, CVD, and venous varix of the lower to differentiate between minimum and moderate extremities. The results of logistic regression analysis dilatations. For this reason, an exercise was made are presented in Table 3. Age (OR: 1.08, p < 0.001), before starting the study and the tongue images of hypertension (OR: 2.30, p < 0.05), and denture wear- participants were re-examined by same investigator ing (OR: 2.17, p < 0.05) were statistically significant to evaluate intra-observer reliability. Intra-observer risk factors for SV. reliability was found 0.91. Hedström et al. [11] reported 0.87 inter-observer agreement using the same DISCUSSION diagnostic criteria. These intra- and inter-observer The prevalence of SV was 26.9% in this study. values shows that reproducibility and repeatability This result is consistent with a recent study using of the classification system is satisfactory. the same diagnostic criteria (26.5%) [1]. However, In the present study, there was no significant the prevalence of SV varies from 4% to 68% in sev- difference in both genders. This finding is consist- eral studies [6, 8, 13, 15, 17, 22]. These variations ent with the previous studies [12, 21]. Contrary to may be due to age differences of the study samples. this finding, Al-Shayyab and Baqain [1] found that Some of the earlier studies have been conducted on females were more likely to have sublingual varies. geriatric patients [13, 19, 20, 25]. In addition, there They proposed that SV may be influenced by hormo- may be differences in diagnostic criteria, methodol- nal conditions. ogy and sampling procedures between the studies. In this study, the prevalence of SV showed a sig- In this study, previously described diagnostic criteria nificant increase with advancing age, especially after by Hedström et al. [12] were used. In this two-grade the 50 years old. This finding is consistent with the classification, minimal dilatation of veins has included previous studies [8, 12, 16]. It has been thought in grade 0, so that these lesions are not accepted as that loss of supporting connective tissue of vessels SV. It is not clear whether minimal dilatations were with aging may lead to venous dilatation [23]. How-

328 N. Akkaya et al., Factors associated with sublingual varices

ever, SV may also occur in individuals younger than ferent results may be due to the fact that quality of 50 years. Kleinman [16] suggested that this finding dentures has not been evaluated during the studies. may be related to premature aging. However, it is Ill-fitting dentures may contribute to the development unclear which mechanism triggers the onset of this of SV in the tissues that become susceptible to trau- condition in younger people. Perhaps the main factor matic effects with age. In this study, denture quality may be an underlying systemic disease, and the in- was not evaluated; this is one of the limitations of creased incidence with advanced age may be due to the study. In the future studies, the effect of denture an increase in the incidence of systemic diseases with quality on the development of SV may be examined. age. Wolff et al. [28] found that healthy, non-medicat- The prevalence of SV in ex-smokers was higher than ed persons, ranging in age between 20 and 95 years, that of current smokers in this study. Therefore, smok- had no oral mucosal changes. These authors noted ing was not found as a contributory factor. This finding that aging alone has not a significant effect on the is not consistent with the previous observations [1, 12]. oral mucosa and its protective defence mechanisms, This result might be related to the fact that smoking when individuals are healthy. They emphasized that has a preventive effect on the development of varicose age-related oral mucosal changes may be associated veins due to vasoconstrictive effect of nicotine [4]. with systemic diseases, malnutrition, medication, and It has been suggested that chronic coughing could wearing of ill-fitting dentures. These results may also cause a recurrent increase in venous pressure because be valid for SV. of avalvular venous system of the tongue; so that it The present study showed that there was an as- might be a contributing factor [8]. In this study this sociation between SV and hypertension (OR: 2.30, variable could not be investigated because a sufficient p < 0.05) and denture wearing (OR: 2.17, p < 0.05). number of subjects were not found. Chronic vitamin Hypertension and denture wearing raised the prob- C deficiency has also been previously reported as an ability of having SV approximately two times. How- associated factor [2]. Because of non-invasive study ever, aging has less influence on development of SV design, we did not include this variable. (OR: 1.08, p < 0.001). This finding is consistent with The limitation of our study is that patients’ med- the previous study reported that SV may be used as ical health information was based on their own an indicator of hypertension (OR: 2.25, p = 0.021) statements. Occult diseases that may be present in [11]. Lynge Pedersen et al. [19] found a relationship the participants could affect the results. Because of between SV and CVD and medication. The association cross-sectional data, the results reflect only associa- between SV and CVD has also stated in another study tions; therefore, cannot be generalised. More detailed [12]. In the light of these previous observations, in our studies are needed to prove causative relations be- study, CVD were classified as hypertension and other tween SV and systemic diseases. CVD. This classification has shown that hypertension may be predisposing factor for the development of CONCLUSIONS SV, as well as aging. If this association reflects true The findings of this study revealed that presence relationship, it will be important for undiagnosed of SV is associated with hypertension and denture hypertensive cases. wearing. The definition of these possible implications Denture wearing may play a role as a contributing is important for the early recognition of the disease factor for the development of SV. It has been well and referral for treatment of the patients. known that especially ill-fitting dentures are related to oral mucosal lesions. Since SV are considered as References normal age-related degenerative changes of the tis- 1. Al-Shayyab MH, Baqain ZH. Sublingual varices in relation sue by several investigators, they are usually exclud- to smoking, cardiovascular diseases, denture wearing, and consuming vitamin rich foods. Saudi Med J. 2015; 36(3): ed from the studies on oral mucosal lesions [7, 24, 310–315, doi: 10.15537/smj.2015.3.10429, indexed in 27]. Therefore, limited information is present about Pubmed: 25737173. pathogenesis of this condition in the literature. No 2. Andrews J, Letcher M, Brook M. Vitamin C supplementation relation has been found between denture wearing in the elderly: a 17-month trial in an old persons’ home. Br Med J. 1969; 2(5654): 416–418, indexed in Pubmed: 5781487. and SV in some previous studies [5, 15, 22]. On the 3. Bean WB. The caviar lesion under the tongue. Trans Am other hand, denture wearing has been reported as Clin Climatol Assoc. 1952; 64: 40–51, indexed in Pu- a risk factor for SV in a recent study [1]. These dif- bmed: 13136210.

329 Folia Morphol., 2019, Vol. 78, No. 2

4. Chalon S, Moreno H, Benowitz NL, et al. Nicotine impairs 18. Lazos JP, Piemonte ED, Panico RL. Oral varix: a review. Ger- endothelium-dependent dilatation in human veins in vivo. odontology. 2015; 32(2): 82–89, doi: 10.1111/ger.12074, Clin Pharmacol Ther. 2000; 67(4): 391–397, doi: 10.1067/ indexed in Pubmed: 23992044. mcp.2000.105153, indexed in Pubmed: 10801248. 19. Lynge Pedersen AM, Nauntofte B, Smidt D, et al. Oral mucosal 5. Corbet EF, Holmgren CJ, Phillipsen HP. Oral mucosal lesions lesions in older people: relation to salivary secretion, systemic in 65-74-year-old Hong Kong Chinese. Community Dent diseases and medications. Oral Dis. 2015; 21(6): 721–729, Oral Epidemiol. 1994; 22(5 Pt 2): 392–395, indexed in doi: 10.1111/odi.12337, indexed in Pubmed: 25754744. Pubmed: 7835037. 20. Mozafari PM, Dalirsani Z, Delavarian Z, et al. Prevalence 6. Corrêa PH, Nunes LC, Johann AC, et al. Prevalence of of oral mucosal lesions in institutionalized elderly people oral hemangioma, vascular malformation and varix in in Mashhad, Northeast Iran. Gerodontology. 2012; 29(2): a Brazilian population. Braz Oral Res. 2007; 21(1): 40–45, e930–e934, doi: 10.1111/j.1741-2358.2011.00588.x, indexed in Pubmed: 17384854. indexed in Pubmed: 22136071. 7. Espinoza I, Rojas R, Aranda W, et al. Prevalence of oral mu- 21. Mumcu G, Cimilli H, Sur H, et al. Prevalence and distri- cosal lesions in elderly people in Santiago, Chile. J Oral Pathol bution of oral lesions: a cross-sectional study in Turkey. Med. 2003; 32(10): 571–575, indexed in Pubmed: 14632931. Oral Dis. 2005; 11(2): 81–87, doi: 10.1111/j.1601- 8. Ettinger RL, Manderson RD. A clinical study of sublingual 0825.2004.01062.x, indexed in Pubmed: 15752080. varices. Oral Surg Oral Med Oral Pathol. 1974; 38(4): 22. Nevalainen MJ, Närhi TO, Ainamo A. Oral mucosal le- 540–545, indexed in Pubmed: 4528384. sions and oral hygiene habits in the home-living elderly. 9. Fleiss, JL. Statistical methods for rates and proportions. J Oral Rehabil. 1997; 24(5): 332–337, indexed in Pu- 2nd ed. New York: John Wiley. 1981. bmed: 9183025. 10. Ghom AG, Ghom SA. Textbook of Oral Medicine. 3th ed. 23. Neville BW, Damm DD, Allen CM, et al. Oral & Maxillofacial New Delhi: Jaypee Brothers. 2014. Pathology, 4th ed. Philadelphia, USA: WB Saunders. 2002. 11. Hedström L, Albrektsson M, Bergh H. Is there a connection 24. Pentenero M, Broccoletti R, Carbone M, et al. The prev- between sublingual varices and hypertension? BMC Oral alence of oral mucosal lesions in adults from the Turin Health. 2015; 15: 78, doi: 10.1186/s12903-015-0054-2, area. Oral Dis. 2008; 14(4): 356–366, doi: 10.1111/j.1601- indexed in Pubmed: 26163474. 0825.2007.01391.x, indexed in Pubmed: 18410578. 12. Hedström L, Bergh H. Sublingual varices in relation to 25. Rabiei M, Kasemnezhad E, Masoudi rad H, et al. Prevalence smoking and cardiovascular diseases. Br J Oral Max- of oral and dental disorders in institutionalised elderly illofac Surg. 2010; 48(2): 136–138, doi: 10.1016/j. people in Rasht, Iran. Gerodontology. 2010; 27(3): 174– bjoms.2009.05.005, indexed in Pubmed: 19540027. –177, doi: 10.1111/j.1741-2358.2009.00313.x, indexed 13. Jainkittivong A, Aneksuk V, Langlais RP. Oral mucosal in Pubmed: 19702672. conditions in elderly dental patients. Oral Dis. 2002; 8(4): 26. Southam JC, Ettinger RL. A histologic study of sublingual 218–223, indexed in Pubmed: 12206403. varices. Oral Surg Oral Med Oral Pathol. 1974; 38(6): 14. Jassar P, Jaramillo M, Nunez DA. Base of tongue 879–886, indexed in Pubmed: 4531618. varices associated with portal hypertension. Post- 27. Triantos D. Intra-oral findings and general health condi- grad Med J. 2000; 76(899): 576–577, indexed in Pu- tions among institutionalized and non-institutionalized bmed: 10964125. elderly in Greece. J Oral Pathol Med. 2005; 34(10): 577– 15. Kaplan I, Moskona D. A clinical survey of oral soft –582, doi: 10.1111/j.1600-0714.2005.00356.x, indexed tissue lesions in institutionalized geriatric patients in in Pubmed: 16202076. Israel. Gerodontology. 1990; 9(2): 59–62, indexed in 28. Wolff A, Ship JA, Tylenda CA, et al. Oral mucosal appear- Pubmed: 2133459. ance is unchanged in healthy, different-aged persons. Oral 16. Kleinman HZ. Lingual varicosities. Oral Surg Oral Med Oral Surg Oral Med Oral Pathol. 1991; 71(5): 569–572, indexed Pathol. 1967; 23(4): 546–548, indexed in Pubmed: 5227407. in Pubmed: 2047098. 17. Kovac-Kovacic M, Skaleric U. The prevalence of oral 29. Vasconcelos BC, Novaes M, Sandrini FA, et al. Prevalence mucosal lesions in a population in Ljubljana, Slovenia. of oral mucosa lesions in diabetic patients: a preliminary J Oral Pathol Med. 2000; 29(7): 331–335, indexed in study. Braz J Otorhinolaryngol. 2008; 74(3): 423–428, Pubmed: 10947249. indexed in Pubmed: 18661018.

Morphometric evaluation and surgical implications of the infraorbital groove, canal and foramen on cone-beam computed tomography and a review of literature İ. Bahşi1, M. Orhan1, P. Kervancioğlu1, E.D. Yalçin2

1Department of Anatomy, Faculty of Medicine, Gaziantep University, Gaziantep, Turkey 2Department of Dentomaxillofacial Radiology, Faculty of Dentistry, Gaziantep University, Gaziantep, Turkey

[Received: 25 June 2018; Accepted: 8 August 2018]

Background: The purpose of this study is to evaluate the anatomy, morphometry, and variations of infraorbital groove (IOG), infraorbital canal (IOC) and infraorbital foramen (IOF) on the cone-beam computed tomography (CBCT) images and to investigate their relations with surrounding structures. Methods: IOG, IOC and IOF were evaluated retrospectively in CBCT images of 75 female (F) and 75 male (M) cases with a range of 18–65 years (F: 37.62 ± ± 13.55, M: 37.53 ± 15.87) by Planmeca Romexis programme. IOG, IOC and IOF were examined bilaterally (300 sides) in the cases. The 13 parameters were measured on these images in axial, sagittal and coronal planes. Results: There was a very weak positive correlation between the age and the angle between IOC and IOG (p = 0.015, r = 0.198), there was a weak positive correlation between the age and skin thickness (p = 0.001, r = 0.281), and there was no correlation between the age and other parameters. A total of 21 (7%) IOCs were detected in maxillary sinus, bilaterally in 6 cases and unilaterally in 9 cases (5 on the left, 4 on the right). In 1 case, bilaterally, IOC was separated 2 canals while running anteriorly in the maxillary sinus. The larger one was directed to IOF in its normal course and the smaller one was directed to lateral wall of nasal cavity and opened to the inferior nasal meatus in front of the opening of nasolacrimal duct. Conclusions: We suggest that the parameters found in the present study may facilitate prediction of the location of the infraorbital nerve. Knowledge of this exact position in relation to easily measurable parameters may decrease the risk of infraorbital nerve injury during surgical approaches directed to this region and might serve as a guide during local anaesthetic interventions for dentistry, ophthalmology, plastic surgery, rhinology, neurosurgery and dermatology. (Folia Morphol 2019; 78, 2: 331–343)

Key words: cone-beam computed tomography, infraorbital foramen, infraorbital canal, infraorbital groove, infraorbital nerve blockage

Address for correspondence: Dr. İ. Bahşi, Department of Anatomy, Faculty of Medicine, Gaziantep University, Gaziantep, Turkey, tel: +90 342 360 60 60/74070, fax: +90 342 472 07 18, e-mail: [email protected]

331 Folia Morphol., 2018, Vol. 78, No. 2

Introduction CBCT images with no pathology of 75 female and Infraorbital groove (IOG), located on the inferior 75 male subjects aged 18–65 years were selected wall of the orbit and at the orbital surface of the randomly. The images of patients who were admitted body of maxilla, begin from the inferior orbital fissure to Gaziantep University Faculty of Dentistry for any posteriorly and continues as infraorbital canal (IOC) reason were evaluated retrospectively by Planmeca anteriorly and IOC opens to the middle of the face Romexis (Planmeca, Helsinki, Finland) programme. through the infraorbital foramen (IOF). Infraorbital On the images with artefact that prevents detection nerve (ION) and vessels pass through IOG, IOC and and measurement of reference points; maxillary sinus spread from IOF [70, 74]. ION is responsible for the pathologies that may affect maxilla, IOG, IOC and sensory innervation of upper cheek skin, maxillary IOF size; cases with developmental, metabolic or sinus mucosa, maxillary incisor, canine and premolar inflammatory jaw disease were not included in the teeth, occasionally mesiobuccal root of first molar study. IOG, IOC and IOF were examined bilaterally tooth and adjacent buccolabial gingiva and perios- (300 sides) in the cases. The following 13 parameters tium, the skin and conjunctiva of the inferior eyelid, were measured on these images in axial, sagittal and part of the nose, and the skin and mucosa of the coronal planes (Fig. 1). upper lip [52]. Anaesthesia of the ION may be performed on Axial: interventions such as surgical procedure of the facial — 1: The angle between the IOC and the sagittal region involving nose soft tissues, cheek, lower eyelid plane (Fig. 1A) and maxillary premolar, canine and incisive teeth, Sagittal: orbital floor fracture treatment, nasal bone fracture — 2: The angle between the IOC and the axial plane reduction, scar revisions, cosmetic cutaneous proce- (Fig. 1B) dures and polypectomy [51, 55, 74]. ION blockage — 3: The angle between IOC and IOG (Fig. 1C) may be performed for trigeminal neuralgia and post- — 4: IOC length (Fig. 1C) operative pain [2, 27]. It is known that ION paraes- — 5: IOG length (Fig. 1C) thesia may occur after surgical interventions [4, 54]. — 6: Vertical diameter of IOF (VD-IOF) (Fig. 1C) Bilateral ION blockage is a preferred local method for — 7: The distance between IOF and infraorbital mar- early repair of cleft lip, facial lacerations, rhinoplasty gin (IOF-IOM) (Fig. 1D) repair and endoscopic endonasal maxillary sinus sur- — 8: Skin thickness over IOF (skin thickness) (ST) geries [13, 31, 35, 48, 56]. It is important to know (Fig. 1D) the IOF neighbourhood for reducing complications in Coronal: placement of malar, submalar, or paranasal implants — 9: Transverse diameter of IOF (TD-IOF) (Fig. 1E) [61]. The risk of ION’s injury is high because it is dif- — 10: The distance between IOF and the mid-sagittal ficult to determine the localisation and the course of plane (IOF-ML) (Fig. 1E) the ION [4]. Although IOC is important, it is not clearly — 11: The distance between IOF and the lateral wall defined [79]. The pathway of ION is important for of the nasal cavity (IOF-LNW) (Fig. 1F) dentistry, ophthalmology, plastic surgery, rhinology, — 12: The distance between IOF and the occlusal neurosurgery and dermatology. For this reason, it is plane of the second premolar tooth (IOF-PM). substantial to know the morphometric properties of The distance between the transverse axis passing IOG, IOC, and IOF and their relations with surround- through the second premolar tooth level and the ing structures. IOF was measured in the images where the IOF The purpose of this study is to evaluate the anat- and the second premolar tooth were not in the omy, morphometry, and variations of IOG, IOC and same section (Fig. 1F) IOF on the cone-beam computed tomography (CBCT) — 13: The distance between IOF and vertical axis of images and to investigate their relations with sur- lateral rim of frontozygomatic suture (IOF-FZS) rounding structures. (Fig. 1G, H)

Materials and methods Statistical analysis Before the study, permission was received from The data were evaluated statistically. The normality the Clinical Trials Ethics Committee. A total of 150 of data distribution was tested by the Shapiro-Wilk test.

332 İ. Bahşi et al., Evaluation of infraorbital groove, canal and foramen

Figure 1. Thirteen parameters (abbreviations — see text) were measured on cone-beam computed tomography images in axial (A), sagittal (B, C, D) and coronal (E, F, G, H) planes.

Student t test was used for comparison of variables with and the smaller one was directed to lateral wall of normal distribution in two independent groups, and nasal cavity and opened to the inferior nasal meatus paired t test was used for comparison of two depend- in front of the opening of nasolacrimal duct (Fig. 3). ent measures. The Pearson correlation coefficient was used when the relations between numerical variables Discussion were tested. The SPSS 22.0 software package was used There are many bone [2, 12, 18, 20, 38, 45, 53, 57, in the analyses and p < 0.05 was accepted significant. 58, 60, 66, 68] and cadaver [6, 14, 25, 39, 44, 59, 66, 71] studies in which the relationship between morpho- Results metry of IOG, IOC, and IOF and their relations to the In this study, IOG, IOC and IOF were evaluated bilat- surrounding structures have been examined. Kazkayasi erally in CBCT images of 75 female (F) and 75 male (M) et al. [38] evaluated 35 dry skulls both in bones and cases with a range of 18–65 years (F: 37.62 ± 13.55, cephalometric radiography. Lee et al. [43] evaluated the M: 37.53 ± 15.87). No significant mean age differ- computed tomography (CT) of the 42 dry skulls. In the ence existed between genders (p > 0.956). Thirteen literature, there are publications in which IOG, IOC or parameters were evaluated (Table 1). The parameters IOF are assessed with CT in the living person [34, 36, 42, were compared by gender; significant differences were 61, 67, 72, 78, 79]. There are few publications in which found in the IOC length (right [R]: p = 0.029, left [L]: these three structures have been evaluated like in this p = 0.021), in the IOF-IOM (R: p = 0.005, L: p = 0.012) study [34]. CT can show localisation and morphometry and in the VD-IOF (R: p = 0.001, L: p = 0.002). of the structures in three dimensions. Hwang et al. There was a very weak positive correlation be- [34] stated that CT mediated measurements were as tween the age and the angle between IOC and IOG effective as dry skull and cadaver measurements. The (p = 0.015, r = 0.198), there was also a weak posi- most important advantage of CT measurements in the tive correlation between the age and ST (p = 0.001, living person compared to dry skull and cadaver is that r = 0.281), and there was no correlation between the the gender and age of the cases are known precisely. age and other parameters. In recent years, the use of CBCT with cheap cost and A total of 21 (7%) IOCs were detected in maxillary low radiation doses has increased for dentomaxillofacial sinus, bilaterally in 6 cases, and unilaterally in 9 cases diagnosis and treatment [7, 30]. (5 on the left, 4 on the right) (Fig. 2). Eppley [24] stated that at the ION injury may result In 1 case, bilaterally, IOC was divided into 2 canals in dysesthesia. Vriens et al. [77] stated long term sen- while running anteriorly in the maxillary sinus. The sory disturbances vary between 24% and 50% after larger one was directed to IOF in its normal course orbitozygomatic complex traumas. ION localisation

333 Folia Morphol., 2018, Vol. 78, No. 2

Table 1. The measurements of 13 parameters about infraorbital groove (IOG), infraorbital canal (IOC) and infraorbital foramen (IOF) Parameter Gender Right Left P F 35.38 ± 7.77 35.42 ± 8.49 0.839 1. The angle between the IOC and the sagittal plane [°] M 34.23 ± 8.60 33.44 ± 9.07 0.154 Total 34.81 ± 8.19 34.44 ± 8.81 0.215 F 46.81 ± 6.82 46.41 ± 7.54 0.580 2. The angle between the IOC and the axial plane [°] M 47.58 ± 7.34 47.22 ± 7.71 0.665 Total 47.20 ± 7.07 46.81 ± 7.61 0.489 F 144.77 ± 7.29 144.82 ± 7.93 0.961 3. The angle between IOC and IOG [°] M 143.40 ± 6.41 143.22 ± 6.88 0.856 Total 144.08 ± 6.41 144.02 ± 7.44 0.925 F 8.20 ± 1.60 8.45 ± 1.94 0.244 4. IOC length [mm] M 8.37 ± 1.78 8.45 ± 1.80 0.654 Total 8.28 ± 1.69 8.45 ± 1.87 0.233 F 21.90 ± 3.57 20.49 ± 3.49 0.001* 5. IOG length [mm] M 23.28 ± 4.07 21.97 ± 4.29 0.004* Total 22.59 ± 3.88 21.23 ± 3.97 0.001* F 3.21 ± 0.45 3.11 ± 0.55 0.194 6. Vertical diameter of IOF (VD-IOF) [mm] M 3.77 ± 1.30 3.46 ± 0.80 0.034* Total 3.49 ± 1.01 3.29 ± 0.71 0.013* F 7.15 ± 1.22 7.10 ± 1.30 0.714 7. The distance between IOF and infraorbital margin (IOF-IOM) [mm] M 7.79 ± 1.49 7.67 ± 1.47 0.368 Total 7.47 ± 1.40 7.39 ± 1.41 0.384 F 9.99 ± 2.48 10.17 ± 2.38 0.303 8. Skin thickness over IOF (skin thickness) (ST) [mm] M 9.31 ± 1.64 9.45 ± 1.95 0.454 Total 9.65 ± 2.13 9.81 ± 2.19 0.208 F 3.32 ± 0.47 3.15 ± 0.45 0.001* 9. Transvers diameter of IOF (TD-IOF) [mm] M 3.42 ± 0.56 3.29 ± 0.66 0.147 Total 3.37 ± 0.52 3.22 ± 0.57 0.004* F 22.65 ± 2.17 22.67 ± 2.39 0.922 10. The distance between IOF and the mid-sagittal plane (IOF-ML) [mm] M 24.39 ± 2.00 24.19 ± 2.15 0.466 Total 23.52 ± 2.26 23.43 ± 2.39 0.640 F 9.24 ± 2.34 8.77 ± 2.19 0.026* 11. The distance between IOF and the lateral wall of the nasal cavity M 9.89 ± 2.42 9.90 ± 2.41 0.966 (IOF-LNW) [mm] Total 9.57 ± 2.39 9.34 ± 2.36 0.154 F 38.04 ± 1.75 37.94 ± 1.96 0.122 12. The distance between IOF and the occlusal plane of the second M 39.33 ± 2.99 39.94 ± 3.06 0.110 premolar tooth (IOF-PM) [mm] Total 38.68 ± 2.53 38.94 ± 2.75 0.102 F 24.68 ± 1.45 24.59 ± 1.38 0.161 13. The distance between IOF and vertical axis of lateral rim of M 24.40 ± 1.22 24.47 ± 1.37 0.612 frontozygomatic suture (IOF-FZS) [mm] Total 24.54 ± 1.34 24.53 ± 1.34 0.829 *Significant difference; F — female; M — male

Figure 2. Infraorbital canal was detected in the maxillary sinus; A. Axial plane; B. Coronal plane; C. Sagittal plane.

334 İ. Bahşi et al., Evaluation of infraorbital groove, canal and foramen

Figure 3. Infraorbital canal (IOC) was divided into two canals while running from posterior to anterior in the maxillary sinus (red arrow: the larger one, yellow arrow: the smaller one).

should be kept in mind in maxilla, zygoma, or deep The angle between the IOC and the axial plane. cheek injury procedures [33]. A common descrip- It has been noted that in studies evaluating CT images tion of the radiofrequency neurotomy procedure [34, 78] this angle gives higher values than dry skull for ION is not available. The angle between the IOC [2, 3, 43] and cadaver studies [59] (Table 2). Hwang and the sagittal plane, the angle between the IOC et al. [34] stated that this angle was statistically sig- and the axial plane, IOF-ML and IOF-IOM distances nificantly different between the genders (p = 0.038). should be known for targeting the needle to IOF In our study, there was no statistically significant dif- and IOC applying the radiofrequency neurotomy ference between the genders (Table 2). [59]. ST is also important during the intervention. The angle between IOC and IOG. Only Hwang et In addition, advancing the needle behind the IOC al. [34] reported measuring this angle in the literature. may damage structures in the orbital cavity or the Similar result was found in our study (Table 2). pterygopalatine fossa [34]. Saeedi et al. [65] and IOC and IOG lengths. Frequently examined in dry Chan et al. [15] reported cases of blunt eye injury skull [32, 38, 58, 63] and cadaver studies [49, 59], as a complication of ION blockade. The shorter and less frequently in CT images [34, 78]. In studies and steeper the IOC is, the more likely it is that the evaluating CT images including the present study, eyeball will be injured. Therefore, it is important to IOC length was shorter than in dry skull and cadaver know morphometry of IOG, IOC and IOF that the studies and IOG length was longer. The bony part ION passes through. which forms the upper wall of the back of the IOC The angle between the IOC and the sagittal is so thin that cannot be clearly detected in the CT plane. In the studies performed [2, 34, 43, 59, 78], images, which may be the possible reason of these results were found in a wide range between 12º differences. Present study and Hwang et al. [34] found and 69º and in our study this angle was found as R: a significant difference in IOG length between the 34.81º, L: 34.44º (Table 2). genders (p = 0.001, p = 0.012, respectively) (Table 3).

335 Folia Morphol., 2018, Vol. 78, No. 2

Table 2. Comparison of the angle between infraorbital canal (IOC) and the sagittal plane, IOC and the axial plane, IOC and infraorbital groove (IOG) with the literature Study Specimen Side The angle between the IOC The angle between the IOC The angle between IOC and the sagittal plane [°] and the axial plane [°] and IOG [°] F M Total F M Total FM Total Aggarwal et al. [2] Dry skull R 21.48 ± 10.41 21.14 ± 31.43 ± 7.97 31.79 ± L 20.81 ± 9.92 10.10 32.10 ± 7.43 7.68 Rahman et al. [59] Cadaver R 22 30 L Lee et al. [43] Dry skull R 12 ± 6.5 44 ± 7.9 L Agthong et al. [3] Dry skull R 24.4 ± 0.7 25.4 ± 0.5 25.1 ± 0.4 L 26.3 ± 0.6 27.0 ± 0.6 26.8 ± 0.4 Hwang et al. [34] HRCT R 13.0 ± 13.5 ± 13.2 ± 47.6 ± 45.3 ± 46.7 ± 144.8 ± 146.8 ± 145.5 ± L 6.5 6.3 6.4 7.6 * 7.4 * 7.6 * 9.0 7.4 8.5 Xu et al. [78] CT R 67.96 ± 69.22 ± 65.72 ± 68.25 ± 10.01 8.52 6.82 6.97 L 67.30 ± 68.71 ± 66.19 ± 68.22 ± 10.08 9.19 6.28 7.15 Present study CBCT R 35.38 ± 34.23 ± 34.81 ± 46.81 ± 47.58 ± 47.20 ± 144.77 143.40 144.08 7.77 8.60 8.19 6.82 7.34 7.07 ± 7.29 ± 6.41 ± 6.41 L 35.42 ± 33.44 ± 34.44 ± 46.41 ± 47.22 ± 46.81 ± 144.82 143.22 144.02 8.49 9.07 8.81 7.54 7.71 7.61 ± 7.93 ± 6.88 ± 7.44

*Significant difference; L — left; R — right; F — female; M — male; CBCT — cone-beam computed tomography; CT — computed tomography; HRCT — high resolution computed tomography

VD-IOF and TD-IOF. These are measured in a num- black females and 6.90 ± 0.30 mm in black males. ber of publications in the literature [2, 5, 6, 10, 12, Apinhasmit et al. [5] reported the significant dif- 18–20, 36, 40, 68, 69, 71–73]. Kara et al. [36] found ference in IOF-IOM distance between the genders. the mean diameter of the IOF and found a signifi- Elias et al. [22] and Macedo et al. [46] reported cant difference between the genders (p = 0.001). there was a significant difference in IOF-IOM dis- Cisneiros de Oliveira et al. [20] reported a significant tance between the both sides. In this study, it was difference in left VD-IOF, right TD-IOF and left TD-IOF observed that the distance between IOF-IOM was measurements between the genders, as well as VD- similar to other studies and there was no significant IOF measurements between the both sides. In our difference between the sides and genders (Table 5). study, there was a significant difference in VD-IOF ST length. Hwang et al. [34] determined the ST and TD-IOF measurements between the both sides length as 11.4 mm. In this study, length was de- (p = 0.013, p = 0.004, respectively) (Table 4). termined as R: 9.65 mm and L: 9.81 mm (Table 6). Distance between IOF-IOM. The most meas- ST length is important in extraoral ION blockade ured distance [1–3, 5, 6, 12, 18, 20–22, 25, 28, 29, approach and radiofrequency neurotomy. For 34, 37, 38, 42, 45, 46, 50, 59, 61, 62, 66, 68, 72, this reason, it is appropriate to move the needle 73, 75, 78]. This distance is used for the detection of 9–11 mm from the skin surface to reach the IOF. IOF localisation in extraoral ION blockade approach The distance between IOF-ML. It is determined and radiofrequency neurotomy. CT measurements between 25.20–30.30 mm [2, 5, 18, 25, 29, 34, 59, show relatively higher values than dry skull meas- 72, 78] (Table 5). In our study, the measurement was urements. Kazkayasi et al. [38] measured 7.19 ± relatively shorter when compared to the literature ± 1.39 mm in the dry skull, and 7.45 ± 0.95 mm (R: 23.52 mm, L: 23.43 mm) (Table 5). in the X-ray graph. By a cadaveric study, Cutright Distance between IOF-LNW or IOF-piriform ap- et al. [21] found 5.80 ± 0.30 mm in white females, erture (IOF-PA). The distance between the IOF-PA 7.10 ± 0.30 mm in white males, 5.70 ± 0.20 mm in examined on spiral CT images [61], skull [2, 18, 22, 38,

336 İ. Bahşi et al., Evaluation of infraorbital groove, canal and foramen

Table 3. Comparison of infraorbital canal (IOC), infraorbital groove (IOG) and IOC + IOG lengths with the literature Study Specimen Side IOC length [mm] IOG length [mm] IOC + IOG length [mm] FM Total FM Total Total Hwang et al. HRCT R 11.7 ± 1.9 16.7 ± 2.4 * [34] L Kazkayası et al. Dry skull R 22.95 ± 5.43 5.95 ± 4.90 [38] L Pryzgocka et al. Dry skull R 14.23 ± 4.68 13.49 ± 3.87 27.71 ± 3.54 [58] L 13.71 ± 4.62 14.14 ± 4.36 28.11 ± 3.22 Huanmanop et al. Dry skull R 13.2 ± 4.4 11.3 ± 3.6 12.3 ± 4.1 [32] L 12.5 ± 3.5 12.3 ± 3.4 12.4 ± 3.4 McQueen et al. Cadaver R 17.08 ± 3.64 [49] L Rontal et al. Dry skull R [63] 14 (8-28) L Xu et al. CT R 9.52 ± 3.00 10.29 ± 2.57 9.98 ± 2.75 [78] L 9.48 ± 2.71 10.18 ± 2.31 9.92 ± 2.47 Rahman et al. Cadaver R 14 13 [59] L Berge and Dry skull R 27.70 Bergman [10] L Karakas et al. Dry skull R 31.9 ± 3.9 [37] L Abed et al. Cadaver R 25.4 ± 2.7 [1] L Present study CBCT R 8.20 ± 8.37 ± 8.28 ± 21.90 ± 23.28 ± 22.59 ± 1.60 1.78 1.69 3.57* 4.07* 3.88* L 8.45 ± 8.45 ± 8.45 ± 20.49 ± 21.97 ± 21.23 ± 1.94 1.80 1.87 3.49* 4.29* 3.97*

*Significant difference; L — left; R — right; F — female; M — male; CBCT — cone-beam computed tomography; CT — computed tomography; HRCT — high resolution computed tomography

45, 46, 68, 75] and cadaver [59]. IOF-LNW examined of the second premolar teeth (46%, 43%, respec- on cephalometric radiographs [38], MD-CT images tively). For this reason, IOF-PM distance in the coronal [72]. This is the first study, in which the distance plane was measured in this study. Raschke et al. [61] between IOF-LNW was evaluated with CBCT. Elias et determined the IOF-PM distance 41.81 ± 1.07 mm al. [22] pointed out that the distance between IOF-PA in males and 37.23 ± 1.58. in females. In the present was statistically significant between the both sides. In study, similar measurements were found with Raschke the present study, the IOF-LNW distance was relatively et al. [61], 38.68 ± 2.53 mm on the right side and shorter (R: 9.57 ± 2.39 mm, L: 9.34 ± 2.36 mm) than 38.94 ± 2.75 mm on the left side (Table 6). reported in the previous studies (Table 7). Distance between IOF-FZS. Raschke et al. [61] Distance between IOF-PM. Taşpınar [72] meas- assessed the distance between the lateral orbital rim ured the distance between the root of second inci- and IOF (Table 7). In this study, the distance between sive tooth and IOF (R: 29.97 ± 3.10 mm, L: 30.20 ± IOF and vertical axis of lateral rim of frontozygomatic ± 3.33 mm), Kara et al. [36] measured the distance be- suture was measured, because it was more precisely tween the IOF and the lateral process of canine tooth detected than the lateral orbital rim on the CBCT im- (F: 37.10 ± 5.30 mm, M: 34.50 ± 4.60 mm). Kaz- ages. The results obtained in both studies are similar kayasi et al. [38] stated IOF to the lateral process of to each other (24–25 mm) (Table 6). This distance will the canine tooth in the vertical direction was 33.94 ± be clinically useful for identifying IOF localisation. ± 3.15 mm. Ukoha et al. [75] and Gour et al. [28] There is little in literature about variation related stated IOF was detected most frequently at the level to IOC. Chandra and Kennedy [16], Elnil et al. [23]

337 Folia Morphol., 2018, Vol. 78, No. 2

Table 4. Comparison of diameters of infraorbital foramen (IOF) with the literature Study Specimen Side Vertical diameter of IOF Transvers diameter of IOF Average diameter of IOF The largest di- (VD-IOF) [mm] (TD-IOF) [mm] ameter of IOF FM Total F M Total FM Total Total Kara et al. CT R 1.51 ± 2.13 ± [36] L 0.49* 0.75* Berge and Dry skull R 3.42 ¥ 2.40 Bergman [10] L Lang [40] Dry skull R 4.57 (7.9–2.5) L 4.56 (7.5–2.6) Song et al. Cadaver R 5 ± 1 [69] L Tezer [73] Dry skull R 4.10 ± 0.87 L 4.33 ± 0.94 Taşpınar [72] MDBT R 1.70 ± 0.43 1.71 ± 1.89 ± 0.44 1.88 ± 0.44 L 1.72 ± 0.40 0.42 1.88 ± 0.44 Chung et al. Dry skull R 4.5 ± 1.1 4.9 ± 1.5 [19] L 5.0 ± 1.4 4.9 ± 1.5 Apinhasmit et al. Dry skull R 3.35 ± 0.62 [5] L Boopathi et al. Dry skull R 2.79 ± 0.79 2.82 ± 2.73 ± 0.73 2.87 ± 0.78 [12] L 2.85 ± 0.80 0.79 3.00 ± 0.81 Aziz et al. Cadaver R 4.4 ± 1.0 4.8 ± 1.0 [6] L 4.5 ± 1.0 4.6 ± 1.2 Cisneiros de Dry skull R 4.0 4.5 4.0 4.0 4.0 4.0 (3.5–4.5) Oliveira et al. (4.0–5.0) (4.0–5.0) (4.0–5.0)* (3.0–4.0)* (3.5–4.5)* [20] L 4.0 4.0 4.0 3.5 4.0 4.0 (3.5–4.5) (3.5– (4.0– (3.5– (3.0– (3.5– 4.12)* 5.0)* 4.5)* 4.0)* 4.5)* Takahashi et al. Cadaver R 5.10 5.70 5.50 [71] L Chrcanovic et al. Dry skull R 3.20 ± 3.31 ± 3.23 ± 0.81 [18] L 0.81 0.81 Singh [68] Dry skull R 3.39 ± 0.96 3.57 ± 3.19 ± 1.18 3.35 ± 1.3 L 3.75 ± 1.07 1.0 3.52 ± 1.35 Aggarwal et al. Dry skull R 3.50 ± 1.16 3.54 ± 2.65 ± 0.93 2.72 ± 0.95 [2] L 3.58 ± 1.07 1.11 2.80 ± 0.98 Sinanoğlu et al. CBCT R 1.85 ± 2.83 ± 3.82 ± 4.83 ± [67] 0.2 0.1 0.2 0.1 L 1.83 ± 2.63 ± 3.97 ± 4.74 ± 0.2 0.2 0.2 0.3 Present study CBCT R 3.21 ± 3.77 ± 3.49 ± 3.32 ± 3.42 ± 3.37 ± 0.52* 0.45 1.30* 1.01* 0.47* 0.56 L 3.11 ± 3.46 ± 3.29 ± 3.15 ± 3.29 ± 3.22 ± 0.57* 0.55 0.80* 0.71* 0.45* 0.66

*Significant difference; L — left; R — right; F — female; M — male; CBCT — cone-beam computed tomography; CT — computed tomography; VD-IOF — vertical diameter of infraorbital foramen; TD-IOF — transvers diameter of infraorbital foramen

and Mailleux et al. [47] reported cases of IOC pass- 150 cases (Fig. 2). Functional endoscopic sinus surgery ing through in the maxillary sinus. Ference et al. [26] (FESS) is the standard surgical treatment for chronic reported this variation in 12.5% of 200 cases, Lantos inflammatory and allergic synonasal disease [23]. FESS et al. [41] reported it in 10.8% of 500 cases and Yeni- procedure is based on the theory that the opening of gun et al. [79] reported in 12.3% of 750 cases. In the the obstructed pathway of osteomeatal drainage com- present study, this variation was found to be 7% of plexes will restore the normal ventilation of the sinuses

338 İ. Bahşi et al., Evaluation of infraorbital groove, canal and foramen

Table 5. Comparison of IOF-IOM and IOF-ML with the literature Study Specimen Side IOF-IOM [mm] IOF-ML [mm] F M Total F M Total Aggarwal et al. [2] Dry skull R 6.37 ± 1.52 25.63 ± 2.27 6.33 ± 1.39 25.69 ± 2.37 L 6.28 ± 1.25 25.74 ± 2.50 Kazkayasi et al. [38] Dry skull R 7.19 ± 1.39 L Kazkayasi et al. [38] X-ray R 7.45 ± 0.95 L Hwang et al. [34] HRCT R 9.4 ± 1.6 9.7 ± 1.7 9.6 ± 1.7 29.1 ± 1.8 26.9 ± 1.9 26.5 ± 1.9 L Cutright et al. [21] Cadaver (white) R 5.8 ± 0.3 7.1 ± 0.3 24.5 ± 0.3* 27.4 ± 0.5 L Cutright et al. [21] Cadaver (black) R 5.7 ± 0.2 6.9 ± 0.3 26.4 ± 0.4* 29.5 ± 0.5 L Raschke et al. [61] Spiral CT R 8.25 ± 0.54 8.61 ± 0.64 L Ercikti et al. [25] Cadaver R 7.8 ± 1.0 9.6 ± 1.2 27.5 ± 2.7 32.0 ± 2.0 8.8 ± 1.0 30.3 ± 2.7 L 7.9 ± 0.8 9.8 ± 0.9 31.2 ± 3.1 30.5 ± 2.9 Lokanayaki [25] Dry skull R 6.12 ± 1.43 L 6.53 ± 1.53 Saylam et al. [66] Cadaver and R 8,3 10,9 dry skull L Rahman et al. [59] Cadaver R 8 26 L Aziz et al. [6] Cadaver R 8.1 ± 1.6 8.5 ± 2.1 25.5 ± 3.6 27.9 ± 4.9 L 7.6 ± 1.6 8.5 ± 2.3 26.9 ± 2.7 27.5 ± 3.7 Apinhasmit et al. [5] Dry skull R 8.71 ± 1.51* 9.53 ± 2.23* 9.23 ± 2.03 27.29 ± 2.12* 29.10 ± 2.13* 28.43 ± 2.29 L Boopathi et al. [12] Dry skull R 6.49 ± 1.26 6.57 ± 1.28 L 6.65 ± 1.30 Cisneiros de Oliveira Dry skull R 8.0 (7.0–9.0) 8.0 (7.0–9.0) et al. [20] L 8.0 (7.0–9.0) 8.0 (6.5–9.0) Chrcanovic et al. [18] Dry skull R 6.35 ± 1.67 6.63 ± 1.75 6.41 ± 1.69 24.67 ± 2.41 26.48 ± 2.58 25.26 ± 2.60 L Singh [68] Dry skull R 6.12 ± 1.79 6.16 ± 1.8 L 6.19 ± 1.81 Taşpınar [72] MDCT R 7.44 ± 1.34 27.16 ± 3.80* 7.50 ± 1.36 26.81 ± 3.19 L 7.55 ± 1.37 26.47 ± 2.40* Tezer [73] Dry skull R 6.98 ± 1.68 28.27 ± 2.38 L 6.98 ± 1.75 28.67 ± 2.59 Karakas et al. [37] Dry skull R 6.7 ± 1.9 L Elias et al. [22] Dry skull R 6.71 ± 1.70* L 6.83 ± 1.83* Agthong et al. [3] Dry skull R 7.5 ± 0.3 8.0 ± 0.3 7.8 ± 0.2 L 7.8 ± 0.2 8.2 ± 0.3 8.0 ± 0.2 Gupta [29] Dry skull R 6.8 ± 1.6 28.0 ± 2.8 7.0 ± 1.6 28.5 ± 2.6 L 7.0 ± 1.7 28.7 ± 2.5 Macedo et al. [46] Dry skull R 6.28 ± 1.79* 6.37 ± 1.69 L 6.45 ± 1.76* Ukoha et al. [75] Dry skull R 6.94 ± 2.57 7.38 ± 2.28 L 7.83 ± 1.86 Gour et al. [28] Dry skull R 6.20 ± 1.859 6.55 ± 1.667 6.50 ± 1.744 L Xu et al. [78] CT R 9.23 ± 1.90 9.30 ± 1.55 9.26 ± 1.68 26.03 ± 1.94 27.21 ± 2.25 26.75 ± 2.20 L 8.81 ± 1.76 9.21 ± 1.33 9.04 ± 1.52 26.19 ± 1.84 26.97 ± 2.28 26.69 ± 2.15 Robinson and Wormald [62] Cadaver R 6,8 L Abed et al. [1] Cadaver R 8.95 ± 1.53 L Lee et al. [42] CT R 8.33 ± 1.50 8.49 ± 1.50 8.42 L 8.34 ± 1.80 8.50 ± 1.60 8.42 Michalek et al. [50] Cadaver R 7.60 ± 1.30 (ultrasound) L Michalek et al. [50] Cadaver (Direk R 6.70 ± 0.90 measurement) L Present study CBCT R 7.15 ± 1.22 7.79 ± 1.49 7.47 ± 1.40 22.65 ± 2.17 24.39 ± 2.00 23.52 ± 2.26 L 7.10 ± 1.30 7.67 ± 1.47 7.39 ± 1.41 22.67 ± 2.39 24.19 ± 2.15 23.43 ± 2.39 *Significant difference; L — left; R — right; F — female; M — male; CBCT — cone-beam computed tomography; CT — computed tomography; HRCT — high resolution computed tomography; MDCT — multiple detector computed tomography; IOF-IOM — the distance between IOF and infraorbital margin; IOF-ML — the distance between IOF and the mid-sagittal plane

339 Folia Morphol., 2018, Vol. 78, No. 2

Table 6. Comparison of IOF-PM, IOF-LOR, IOF-FZS and ST with the literature Study Specimen Side IOF-PM IOF-LOR ST [mm] FM Total FM Total FM Total Raschke et al. Spiral CT R 37.23 ± 41.81 ± 24.21 ± 25.93 ± [61] L 1.58 1.07 1.68 1.59 Hwang et al. CT R 11.4 ± 1.4 11.5 ± 1.9 11.4 ± 1.9 [34] L IOF-FZS Present study CBCT R 38.04 ± 39.33 ± 38.68 ± 24.68 ± 24.40 ± 24.54 ± 9.99 ± 9.31 ± 9.65 ± 1.75 2.99 2.53 1.45 1.22 1.34 2.48 1.64 2.13 L 37.94 ± 39.94 ± 38.94 ± 24.59 ± 24.47 ± 24.53 ± 10.17 ± 9.45 ± 9.81 ± 1.96 3.06 2.75 1.38 1.37 1.34 2.38 1.95 2.19 L — left; R — right; F — female; M — male; CBCT — cone-beam computed tomography; CT — computed tomography; IOF-PM — the distance between IOF and the occlusal plane of the second premolar tooth; IOF-LOR — the distance between IOF and lateral orbital rim; IOF-FZS — the distance between IOF and vertical axis of lateral rim of frontozygomatic suture; ST — skin thickness over IOF (skin thickness)

Table 7. Comparison of IOF-LNW and IOF-PA with the literature Study Specimen Side IOF-LNW [mm] IOF-PA [mm] F M Total F M Total Aggarwal et al. [2] Dry skull R 15.51 ± 1.63 15.19 ± 1.70 L 14.87 ± 1.73 Kazkayasi et al. [38] Dry skull R 17.23 ± 2.64 L Kazkayasi et al. [38] X-ray R 14.31 ± 1.96 L Raschke et al. [61] Spiral CT R 15.69 ± 0.76 17.43 ± 1.19 L Lokanayaki [45] Dry skull R 16.58 ± 2.37 L 16.38 ± 2.25 Rahman et al. [59] Cadaver R 17 L Chrcanovic et al. [18] Dry skull R 14.37 ± 2.04 15.44 ± 1.79 14.72 ± 2.02 L Singh [68] Dry skull R 15.31 ± 1.77 15.56 ± 2.6 L 15.80 ± 2.86 Taşpınar [72] MDCT R 16.58 ± 1.85 16.63 ± 1.81 L 16.68 ± 1.77 Elias et al. [22] Dry skull R 13.28 ± 2.17* L 13.31 ± 2.19* Macedo et al. [46] Dry skull R 17.75 ± 2.10 17.67 ± 1.95 L 17.60 ± 2.04 Ukoha et al. [75] Dry skull R 19.36 ± 3.54 18.82 ± 3.29 L 18.27 ± 2.94 Present study CBCT R 9.24 ± 2.34 9.89 ± 2.42 9.57 ± 2.39 L 8.77 ± 2.19 9.90 ± 2.41 9.34 ± 2.36

*Significant difference; L — left; R — right; F — female; M — male; CBCT — cone-beam computed tomography; MDCT — multidetector computed tomography; IOF-LNW — the distance between infraorbital foramen and the lateral wall of the nasal cavity; IOF-PA — the distance between infraorbital foramen and preform aperture

340 İ. Bahşi et al., Evaluation of infraorbital groove, canal and foramen

in cases of chronic sinusitis [17]. Prior to FESS, Mailleux tee of Gaziantep University and has, therefore, been et al. [47] and Elnil et al. [23] proposed CT imaging performed in accordance with the ethical standards as a preoperative procedure to reduce complications. laid down in the 1964 Declaration of Helsinki and its Lantos et al. [41] reported a high risk of ION injury in later amendments. maxillary sinus with chronic inflammation, neoplasms, and patients with these variations in resection of an in- Acknowledgements verted papilloma require an antral punch to introduce The authors are greatful to Associate Professor a balloon dilation catheter to approach the maxillary Dr. Seval KUL, Department of Biostatistics, School of infundibulum. A window is opened through the gin- Medicine, Gaziantep University for the support. givobuccal sulcus in the anterior wall of the maxilla to reach maxillary sinus in Caldwell Luc operation, which References is one of the preferred methods in various situations 1. Abed SF, Shams PN, Shen S, et al. Morphometric and geometric anatomy of the caucasian orbital floor. Orbit. 2011; 30(5): such as chronic rhinosinusitis, paranasal papilloma and 214–220, doi: 10.3109/01676830.2010.539768, indexed in foreign body removal [8, 9, 11, 56]. For these reasons, Pubmed: 21812531. 2. Aggarwal A, Kaur H, Gupta T, et al. Anatomical study of the variations of IOC in the maxillary sinus should be the infraorbital foramen: A basis for successful infraorbital considered in the surgical procedures. Although for the nerve block. Clin Anat. 2015; 28(6): 753–760, doi: 10.1002/ cases that IOC in maxillary sinus is less likely to have ca.22558, indexed in Pubmed: 26119635. 3. Agthong S, Huanmanop T, Chentanez V. Anatomical varia- infraorbital neurovascular bundle injury in orbital floor tions of the supraorbital, infraorbital, and mental foramina fractures as they move away from the orbital floor. related to gender and side. J Oral Maxillofac Surg. 2005; 63(6): 800–804, doi: 10.1016/j.joms.2005.02.016, indexed Leo et al. [44] identified two separate IOC with in Pubmed: 15944977. ION that opened to the anterior surface of maxilla 4. Aitasalo K, Kinnunen I, Palmgren J, et al. Repair of orbital in a cadaver. Von Arx et al. [76] defined canalis sin- floor fractures with bioactive glass implants. J Oral Maxillofac Surg. 2001; 59(12): 1390–5; discussion 1395, doi: 10.1053/ uosus, as a tortuous bone channel originating from joms.2001.27524, indexed in Pubmed: 11732018. the IOC slightly posterior to the IOF and coursing in 5. Apinhasmit W, Chompoopong S, Methathrathip D, et al. Supraorbital Notch/Foramen, Infraorbital Foramen and an anteromedial direction to the anterior wall of the Mental Foramen in Thais: anthropometric measurements and nasal antrum below the orbital margin. Rusu et al. surgical relevance. J Med Assoc Thai. 2006; 89(5): 675–682, [64] identified a case in which the IOC was absent indexed in Pubmed: 16756055. 6. Aziz SR, Marchena JM, Puran A. Anatomic characteristics of and the lateroantral duct was present. In this study the infraorbital foramen: a cadaver study. J Oral Maxillofac a very rare case was detected; IOC was divided into Surg. 2000; 58(9): 992–996, doi: 10.1053/joms.2000.8742, indexed in Pubmed: 10981979. two canals in the maxillary sinus and the smaller one 7. Bahşi I, Orhan M, Kervancıoğlu P, et al. Anatomical evaluation was directed to the lateral wall of the nasal cavity and of nasopalatine canal on cone beam computed tomography opened to the inferior nasal meatus in front of the images. Folia Morphol. 2019; 78(1): 153–162, doi: 10.5603/ FM.a2018.0062, indexed in Pubmed: 30009362. opening of the nasolacrimal duct (Fig. 3). 8. Balıkçı HH, Özbay İ, Karakaş M. Etmoid sinüste saçma tanesi: olgu sunumu. ODÜ Tıp Dergisi. 2015; 2(3): 126–129. 9. Becker SS, Roberts DM, Beddow PA, et al. Comparison of Conclusions maxillary sinus specimens removed during Caldwell-Luc In conclusion, we suggest that the parameters procedures and traditional maxillary sinus antrostomies. Ear Nose Throat J. 2011; 90(6): 262–266, doi: 10.1177/0145561 found in the present study may facilitate prediction 31109000607, indexed in Pubmed: 21674469. of the location of the ION. Knowledge of this exact 10. Berge JK, Bergman RA. Variations in size and in symmetry of position in relation to easily measurable parameters foramina of the human skull. Clin Anat. 2001; 14(6): 406– –413, doi: 10.1002/ca.1075, indexed in Pubmed: 11754234. may decrease the risk of ION injury during surgical 11. Berkiten G, Bal M, Kumral TL, et al. Schneiderian papilloma: approaches directed to this region and might serve our experience over 12 years and literature review. Nobel Medicus. 2015; 11(2): 19–23. as a guide during local anaesthetic interventions for 12. Boopathi S, Chakravarthy Marx S, Dhalapathy SL, et al. An- dentistry, ophthalmology, plastic surgery, rhinology, thropometric analysis of the infraorbital foramen in a South neurosurgery and dermatology. Indian population. Singapore Med J. 2010; 51(9): 730–735, indexed in Pubmed: 20938615. 13. Bösenberg AT, Kimble FW. Infraorbital nerve block in neo- Ethical standards nates for cleft lip repair: anatomical study and clinical application. Br J Anaesth. 1995; 74(5): 506–508, indexed in This study was approved by the ethics committee Pubmed: 7772421. of Gaziantep University (approval date and number: 14. Brandão FH, Machado M, Aquino J, et al. The foramen and 07 March 2016; 2016/72). We declare that this hu- infraorbital nerve relating to the surgery for external access to the maxillary sinus (Caldwell-Luc). Int Arch Otorhinolarnygol. man study has been approved by the ethics commit- 2008; 12(3): 342–346.

341 Folia Morphol., 2018, Vol. 78, No. 2

15. Chan BJ, Koushan K, Liszauer A, et al. Iatrogenic globe pen- Assoc Thai. 2007; 90(11): 2383–2391, indexed in Pub- etration in a case of infraorbital nerve block. Can J Ophthal- med: 18181324. mol. 2011; 46(3): 290–291, doi: 10.1016/j.jcjo.2011.05.012, 33. Hwang K, Suh MS, Chung InH. Cutaneous distribution of indexed in Pubmed: 21784222. infraorbital nerve. J Craniofac Surg. 2004; 15(1): 3–5; discus- 16. Chandra RK, Kennedy DW. Surgical implications of an unusual sion 5, indexed in Pubmed: 14704552. anomaly of the infraorbital nerve. Ear Nose Throat J. 2004; 34. Hwang SeH, Kim SW, Park CS, et al. Morphometric analysis 83(11): 766–767, indexed in Pubmed: 15628633. of the infraorbital groove, canal, and foramen on three- 17. Chong VF, Fan YF, Lau D, et al. Functional endoscopic sinus dimensional reconstruction of computed tomography scans. surgery (FESS): what radiologists need to know. Clin Radiol. Surg Radiol Anat. 2013; 35(7): 565–571, doi: 10.1007/ 1998; 53(9): 650–658, indexed in Pubmed: 9766718. s00276-013-1077-5, indexed in Pubmed: 23404562. 18. Chrcanovic BR, Abreu MH, Custódio AL. A morphometric 35. Jonnavithula N, Durga P, Kulkarni DK, et al. Bilateral intra- analysis of supraorbital and infraorbital foramina rela- oral, infra-orbital nerve block for postoperative analgesia tive to surgical landmarks. Surg Radiol Anat. 2011; 33(4): following cleft lip repair in paediatric patients: comparison 329–335, doi: 10.1007/s00276-010-0698-1, indexed in of bupivacaine vs bupivacaine-pethidine combination. Pubmed: 20625730. Anaesthesia. 2007; 62(6): 581–585, doi: 10.1111/j.1365- 19. Chung MS, Kim HJ, Kang HS, et al. Locational relationship 2044.2007.05028.x, indexed in Pubmed: 17506736. of the supraorbital notch or foramen and infraorbital and 36. Kara SA, Ünal B, Erdal H, et al. İnfraorbital Foramen Anato- mental foramina in Koreans. Acta Anat (Basel). 1995; 154(2): misinin Radyolojik Analizi. KBB ve BBC Dergisi. 2003; 11(1): 162–166, indexed in Pubmed: 8722516. 17–21. 20. Cisneiros de Oliveira LC, Silveira MP, de Almeida Júnior E, et al. 37. Karakaş P, Bozkir MG, Oguz O. Morphometric measurements Morphometric study on the infraorbital foramen in relation from various reference points in the orbit of male Caucasians. to sex and side of the cranium in northeastern Brazil. Anat Surg Radiol Anat. 2003; 24(6): 358–362, doi: 10.1007/ Cell Biol. 2016; 49(1): 73–77, doi: 10.5115/acb.2016.49.1.73, s00276-002-0071-0, indexed in Pubmed: 12652362. indexed in Pubmed: 27051570. 38. Kazkayasi M, Ergin A, Ersoy M, et al. Certain anatomical relations 21. Cutright B, Quillopa N, Schubert W. An anthropometric and the precise morphometry of the infraorbital foramen--canal analysis of the key foramina for maxillofacial surgery. and groove: an anatomical and cephalometric study. Laryngo- J Oral Maxillofac Surg. 2003; 61(3): 354–357, doi: 10.1053/ scope. 2001; 111(4 Pt 1): 609–614, doi: 10.1097/00005537- joms.2003.50070, indexed in Pubmed: 12618976. 200104000-00010, indexed in Pubmed: 11359128. 22. Elias MG, Silva RB, Pimentel ML, et al. Morphometric analy- 39. Kazkayasi M, Ergin A, Ersoy M, et al. Microscopic anatomy sis of the infraorbital foramen and acessories foraminas in of the infraorbital canal, nerve, and foramen. Otolaryngol brazilian skulls. Int J Morphol. 2004; 22(4), doi: 10.4067/ Head Neck Surg. 2003; 129(6): 692–697, doi: 10.1016/S0194- s0717-95022004000400006. 59980301575-4, indexed in Pubmed: 14663437. 23. Elnil H, Al-Tubaikh JA, El Beltagi AH. Into the septum I go, 40. Lang J. Clinical Anatomy of the Head: Neurocranium· Orbit· a case of bilateral ectopic infraorbital nerves: a not-to-miss Craniocervical Regions. Berlin: Springer Science & Business preoperative sinonasal CT variant. Neuroradiol J. 2014; Media. 2012. 27(2): 146–149, doi: 10.15274/NRJ-2014-10033, indexed in 41. Lantos JE, Pearlman AN, Gupta A, et al. Protrusion of the Pubmed: 24750699. Infraorbital Nerve into the Maxillary Sinus on CT: Preva- 24. Eppley B. Re: cutaneous distribution of infraorbital nerve. lence, Proposed Grading Method, and Suggested Clinical J Craniofac Surg. 2004; 15(1): 5, doi: 10.1097/00001665- Implications. AJNR Am J Neuroradiol. 2016; 37(2): 349–353, 200401000-00003. doi: 10.3174/ajnr.A4588, indexed in Pubmed: 26564432. 25. Ercikti N, Apaydin N, Kirici Y. Location of the infraorbital 42. Lee T, Lee H, Baek S. A three-dimensional computed tomograph- foramen with reference to soft tissue landmarks. Surg Radiol ic measurement of the location of infraorbital foramen in East Anat. 2017; 39(1): 11–15, doi: 10.1007/s00276-016-1683-0, Asians. J Craniofac Surg. 2012; 23(4): 1169–1173, doi: 10.1097/ indexed in Pubmed: 27146295. SCS.0b013e31824f658a, indexed in Pubmed: 22801117. 26. Ference EH, Smith SS, Conley D, et al. Surgical anatomy and 43. Lee UY, Nam SH, Han SH, et al. Morphological characteris- variations of the infraorbital nerve. Laryngoscope. 2015; tics of the infraorbital foramen and infraorbital canal using 125(6): 1296–1300, doi: 10.1002/lary.25089, indexed in three-dimensional models. Surg Radiol Anat. 2006; 28(2): Pubmed: 25992806. 115–120, doi: 10.1007/s00276-005-0071-y, indexed in 27. Feriani G, Hatanaka E, Torloni MR, et al. Infraorbital nerve Pubmed: 16432643. block for postoperative pain following cleft lip repair in 44. Leo JT, Cassell MD, Bergman RA. Variation in human infraor- children. Cochrane Database Syst Rev. 2016; 4: CD011131, bital nerve, canal and foramen. Ann Anat. 1995; 177(1): doi: 10.1002/14651858.CD011131.pub2, indexed in Pub- 93–95, doi: 10.1016/S0940-9602(11)80139-1, indexed in med: 27074283. Pubmed: 7872502. 28. Gour K, Nair S, Trivedi G, et al. Anthropometric measurements 45. Lokanayaki V. Anatomic variations of infra orbital foramen. of infraorbital foramen in dried human skulls. Int J Biol Med J Surg. 2013; 2(2): 30–36. Res. 2012; 3(3): 2003–2006. 46. Macedo V, Cabrini R, Faig-Leite H. Infraorbital foramen location 29. Gupta T. Localization of important facial foramina encoun- in dry human skulls. Braz J Morphol Sci. 2009; 26(1): 35–38. tered in maxillo-facial surgery. Clin Anat. 2008; 21(7): 633– 47. Mailleux P, Desgain O, Ingabire MI. Ectopic infraorbital nerve –640, doi: 10.1002/ca.20688, indexed in Pubmed: 18773483. in a maxillary sinus septum: another potentially dangerous 30. Harris D, Buser D, Dula K, et al. E.A.O. Guidelines for the use of variant for sinus surgery. JBR-BTR. 2010; 93(6): 308–309, Diagnostic Imaging in Implant Dentistry. A consensus work- indexed in Pubmed: 21381529. shop organized by the European Association for Osseointe- 48. Maravolo M, Guerra-Narducci V, Mucciolo A, et al. [Local gration in Trinity College Dublin. Clin Oral Impl Res. 2002; analgesia and infraorbital block in rhinoplasty operations. 13(5): 566–570, doi: 10.1034/j.1600-0501.2002.130518.x. Our experience]. Minerva Anestesiol. 1991; 57(9): 526–527, 31. Higashizawa T, Koga Y. Effect of infraorbital nerve block indexed in Pubmed: 1798469. under general anesthesia on consumption of isoflurane 49. McQueen CT, DiRuggiero DC, Campbell JP, et al. Orbital oste- and postoperative pain in endoscopic endonasal maxillary ology: a study of the surgical landmarks. Laryngoscope. 1995; sinus surgery. J Anesth. 2001; 15(3): 136–138, doi: 10.1007/ 105(8 Pt 1): 783–788, doi: 10.1288/00005537-199508000- s005400170014, indexed in Pubmed: 14566510. 00003, indexed in Pubmed: 7630287. 32. Huanmanop T, Agthong S, Chentanez V. Surgical anatomy 50. Michalek P, Donaldson W, McAleavey F, et al. Ultrasound of fissures and foramina in the orbits of Thai adults. J Med imaging of the infraorbital foramen and simulation of the

342 İ. Bahşi et al., Evaluation of infraorbital groove, canal and foramen

ultrasound-guided infraorbital nerve block using a skull mod- 65. Saeedi OJ, Wang H, Blomquist PH. Penetrating globe el. Surg Radiol Anat. 2013; 35(4): 319–322, doi: 10.1007/ injury during infraorbital nerve block. Arch Otolaryngol s00276-012-1039-3, indexed in Pubmed: 23129265. Head Neck Surg. 2011; 137(4): 396–397, doi: 10.1001/ 51. Molliex S, Navez M, Baylot D, et al. Regional anaesthesia for archoto.2010.239, indexed in Pubmed: 21242530. outpatient nasal surgery. Br J Anaesth. 1996; 76(1): 151–153, 66. Saylam C, Özer MA, Bilge O, et al. Anatomic variations of the indexed in Pubmed: 8672358. infraorbital foramen. Ann Plast Surg. 1999; 43(6): 613–617. 52. Moore KL, Dalley AF, Agur AM. Clinically oriented anatomy. 67. Sinanoglu A, Orhan K, Kursun S, et al. Evaluation of optic 7th Ed. Philadelphia, Lippincott Williams & Wilkins. 2013. canal and surrounding structures using cone beam com- 53. Ozer MA, Govsa F, Kazak Z, et al. Redesign and treatment puted tomography: considerations for maxillofacial surgery. planning orbital floor reconstruction using computer analysis J Craniofac Surg. 2016; 27(5): 1327–1330, doi: 10.1097/ anatomical landmarks. Eur Arch Otorhinolaryngol. 2016; SCS.0000000000002726, indexed in Pubmed: 27391501. 273(8): 2185–2191, doi: 10.1007/s00405-015-3741-3, in- 68. Singh R. Morphometric analysis of infraorbital fora- dexed in Pubmed: 26242253. men in Indian dry skulls. Anat Cell Biol. 2011; 44(1): 54. Peltomaa J, Rihkanen H. Infraorbital nerve recovery after 79–83, doi: 10.5115/acb.2011.44.1.79, indexed in Pub- minimally dislocated facial fractures. Eur Arch Otorhinolaryn- med: 21519552. gol. 2000; 257(8): 449–452, indexed in Pubmed: 11073197. 69. Song WC, Kim SH, Paik DJ, et al. Location of the infraorbi- 55. Prabhu KP, Wig J, Grewal S. Bilateral infraorbital nerve block tal and mental foramen with reference to the soft-tissue is superior to peri-incisional infiltration for analgesia after landmarks. Plast Reconstr Surg. 2007; 120(5): 1343–1347, repair of cleft lip. Scand J Plast Reconstr Surg Hand Surg. doi: 10.1097/01.prs.0000279558.86727.5a, indexed in 1999; 33(1): 83–87, indexed in Pubmed: 10207969. Pubmed: 17898610. 56. Przygocka A, Jędrzejewski K, Szymański J, et al. Infraorbital 70. Standring S. Gray’s anatomy: the anatomical basis of clinical groove localisation for the endoscopic decompression of the practice. 41th Ed. Edinburgh, Elsevier Health Sciences. 2016. orbit in Graves’ disease. Folia Morphol. 2015; 74(1): 78–83, 71. Takahashi Y, Kakizaki H, Nakano T. Infraorbital foramen: doi: 10.5603/FM.2015.0013, indexed in Pubmed: 25792400. horizontal location in relation to ala nasi. Ophthalmic 57. Przygocka A, Podgórski M, Jędrzejewski K, et al. The loca- Plast Reconstr Surg. 2011; 27(4): 295–297, doi: 10.1097/ tion of the infraorbital foramen in human skulls, to be used IOP.0b013e3182078e72, indexed in Pubmed: 21386744. as new anthropometric landmarks as a useful method for 72. Taşpınar Ç. Multidetektör Bilgisayarlı Tomografi İle maxillofacial surgery. Folia Morphol. 2012; 71(3): 198–204, Değerlendirilen Canlı Kişilerde Foramen Infraorbitale indexed in Pubmed: 22936558. Varyasyonları: Afyon Kocatepe Üniversitesi, Sağlık Bilimleri 58. Przygocka A, Szymański J, Jakubczyk E, et al. Variations in Enstitüsü, Yüksek Lisans Tezi, 46 Sayfa, Afyon. (Prof. Dr. the topography of the infraorbital canal/groove complex: Ahmet Songur): 2013. a proposal for classification and its potential usefulness in 73. Tezer M. Foramen infraorbitale’nin anatomik ve morfometrik orbital floor surgery. Folia Morphol. 2013; 72(4): 311–317, özellikleri: İstanbul Üniversitesi, İstanbul Tıp Fakültesi, indexed in Pubmed: 24402752. Uzmanlık Tezi, 44 Sayfa, İstanbul, (Prof. Dr. Adnan Öztürk. 59. Rahman M, Richter EO, Osawa S, et al. Anatomic study of 2008. the infraorbital foramen for radiofrequency neurotomy of 74. Tomaszewska IM, Zwinczewska H, Gładysz T, et al. Anatomy the infraorbital nerve. Neurosurgery. 2009; 64(5 Suppl 2): and clinical significance of the maxillary nerve: a literature 423–427, doi: 10.1227/01.NEU.0000336327.10368.79, review. Folia Morphol. 2015; 74(2): 150–156, doi: 10.5603/ indexed in Pubmed: 19404120. FM.2015.0025, indexed in Pubmed: 26050800. 60. Rai AR, Rai R, Vadgaonkar R, et al. Anatomical and mor- 75. Ukoha UU, Umeasalugo KE, Udemezue OO, et al. Anthropo- phometric analysis of accessory infraorbital foramen. metric measurement of infraorbital foramen in south-east J Craniofac Surg. 2013; 24(6): 2124–2126, doi: 10.1097/ and south-south Nigeria. Natl J Med Res. 2014; 4: 225–227. SCS.0b013e31828f2fa6, indexed in Pubmed: 24220421. 76. von Arx T, Lozanoff S, Sendi P, et al. Assessment of bone 61. Raschke R, Hazani R, Yaremchuk MJ. Identifying a safe channels other than the nasopalatine canal in the anterior zone for midface augmentation using anatomic landmarks maxilla using limited cone beam computed tomography. Surg for the infraorbital foramen. Aesthet Surg J. 2013; 33(1): Radiol Anat. 2013; 35(9): 783–790, doi: 10.1007/s00276- 13–18, doi: 10.1177/1090820X12468752, indexed in Pub- 013-1110-8, indexed in Pubmed: 23539212. med: 23220876. 77. Vriens JP, van der Glas HW, Bosman F, et al. Information on 62. Robinson S, Wormald PJ. Patterns of innervation of the an- infraorbital nerve damage from multitesting of sensory func- terior maxilla: a cadaver study with relevance to canine fossa tion. Int J Oral Maxillofac Surg. 1998; 27(1): 20–26, indexed puncture of the maxillary sinus. Laryngoscope. 2005; 115(10): in Pubmed: 9506294. 1785–1788, doi: 10.1097/01.mlg.0000176544.72657.a6, 78. Xu H, Guo Y, Lv D, et al. Morphological structure of the indexed in Pubmed: 16222195. infraorbital canal using three-dimensional reconstruction. 63. Rontal E, Rontal M, Guilford FT. Surgical anatomy of the orbit. J Craniofac Surg. 2012; 23(4): 1166–1168, doi: 10.1097/ Ann Otol Rhinol Laryngol. 1979; 88(3 Pt 1): 382–386, doi: 10 SCS.0b013e31824dfcfd, indexed in Pubmed: 22801116. .1177/000348947908800315, indexed in Pubmed: 464532. 79. Yenigun A, Gun C, Uysal II, et al. Radiological classification 64. Rusu MC, Săndulescu M, Ilie OC. Infraorbital canal bilaterally of the infraorbital canal and correlation with variants of replaced by a lateroantral canal. Surg Radiol Anat. 2015; neighboring structures. Eur Arch Otorhinolaryngol. 2016; 37(9): 1149–1153, doi: 10.1007/s00276-015-1468-x, indexed 273(1): 139–144, doi: 10.1007/s00405-015-3550-8, indexed in Pubmed: 25823691. in Pubmed: 25673024.

Assessment of paranasal sinus parameters according to ancient skulls’ gender and age by using cone-beam computed tomography K.O. Demiralp1, S. Kursun Cakmak1, S. Aksoy2, S. Bayrak3, K. Orhan4, P. Demir5

1Ministry of Health, Ankara, Turkey 2Department of Dentomaxillofacial Radiology, Dentistry Faculty, Near East University, Nicosia, Cyprus 3Department of Dentomaxillofacial Radiology, Dentistry Faculty, Abant Izzet Baysal University, Bolu, Turkey 4Department of Dentomaxillofacial Radiology, Dentistry Faculty, Ankara University, Ankara, Turkey 5Biostatistics Department, Yıldırm Beyazıt University, Ankara, Turkey

[Received: 11 July 2018; Accepted: 12 September 2018]

Background: The aim of this study was to determine whether paranasal sinus dimensions and volume can be useful to identify gender and age estimation for ancient skulls using cone-beam computed tomography (CBCT) images. Materials and methods: CBCT scans of 32 ancient skulls of approximately 1000 years of age were included in this retrospective study. The gender and age estimation of the skulls were made by an independent anthropologist, which was considered as the gold standard. Paranasal sinuses’ dimensions (width and height) and volumes of each sinus were measured from the CBCT data set that was linked to the three-dimensional rendering software (Anatomage, Invivo 5.2). All measurements were performed by an independent observer. Intra-observer analysis was made. Mann-Whitney and Kruskal-Wallis tests were used to compare paranasal sinus parameters in terms of age estimation and gender (p < 0.05). Results: The results demonstrated no statistically significant difference between measurements (p < 0.05). The measurements were found to be highly reproducible. The mean volumes of frontal and sphenoid sinus were found to be higher in males. The distance from anterior-posterior wall of sphenoid sinus in axial sections is larger in males (p > 0.05). The frontal sinus width and volume increased statistically with age above 60 years of age (p > 0.05). Conclusions: The paranasal volume and dimensions’ measurements from CBCT data can be a promising technique to determine gender and age of ancient skulls because of its lower voxel sizes and higher resolution. (Folia Morphol 2019; 78, 2: 344–350)

Key words: age estimation, anthropology, gender, paranasal sinus, cone-beam computed tomography

INTRODUCTION method indispensable in personal identification [8]. Fingerprints, dental comparison and biological By means of unrecognised skeletons epiphysis and methods such as DNA profiling are essential tools metaphysis’ evaluation, gender has been estimated in identification of human beings. However soft tis- from pelvis, skull, and long bones [20]. sues can be perished as are in ancient remains, and In gender determination, detection from skeleton, this makes skeletal examination and anthropological from both pelvis and skull, from pelvis only or the

Address for correspondence: Dr. S. Bayrak, Abant Izzet Baysal University, Faculty of Dentistry, Dentomaxillofacial Radiology Department, Bolu, Turkey, tel: +90 374 253 8361, fax: +90 374 254 0066, e-mail: [email protected]

344 K.O. Demiralp et al., Paranasal sinus parameters in ancient skulls

Figure 1. Cone beam computed tomography scan of the skull. pelvis and long bones, from both the skull and long for imaging of the skull base because of reduced ra- bones, from long bones only exhibits 100%, 98%, 95%, diation dose, higher spatial resolution, smaller voxel 90–95%, 80–90% precise results, respectively [11, 14]. sizes along with smaller thickness size on CBCT images In age-at-death estimation, the most commonly compared with MDCT images [15, 27]. used indicators are briefly: skeletal maturation, pu- The aim of this study was to determine whether bic symphysis, sternal end of ribs, auricular surface, paranasal sinus dimensions and volume can be useful acetabulum, suture closure, pathology and cartilage to identify gender and age estimation for ancient ossification. Other methods include microscopic, mo- skulls using CBCT images. lecular and chemical assessments [17]. Bone structures such as paranasal sinuses are typi- MATERIALS AND METHODS cal and specific because of their unique nature and ir- Thirty-two ancient dry skulls that were unearthed regular shape. The durable features of these osteolog- from certain archaeological excavations from Ana- ical structures make them indispensable for forensic tolia, Turkey dated approximately back 1000 years purposes [2]. It has been suggested that the frontal si- ago from period of 800–1000 C.E. were used for this nus (FS) has the potential to be used for personal iden- study. The sex and the age estimations of each skull tification, age estimation, and sexual dimorphism [25]. were determined by an independent anthropologist. Also for sinuses, remaining intact after hard condition In this study the symphyseal surfaces of the pubic and and protecting their structure makes them useful for the auricular surfaces of the coxa, the aspect of the identity purposes. Considering the complex structure spongious tissue in humerus and femur, the closure of paranasal sinuses, computed tomography (CT) is of the cranial sutures and the tooth wear were con- a gold standard method to assess the exact anatomy sidered when estimating the age of the adults. For of sinuses [16]. Disadvantage of high dose and high sexual diagnosis, the following complex of characters cost makes its use to be limited. X-ray and CT analyses were considered: general shape of the pelvis, size of have importance for observing some gender identifica- the greater sciatic notch, the shape of the sacrum, tion of facial traits of the skull; furthermore, these radi- robustness of the skeleton, development of muscular ological modalities are significant in order to estimate joints and insertions, cranial relief, forehead shape, the age at death [16]. Medical CT (MDCT) has been robustness and shape of the mandible [3, 6, 14, 24]. shown to yield accurate and reliable assessments for There were 18 ancient male skulls with a mean age skull evaluations. A cone-beam CT (CBCT), a technique estimated at 41.4 ± 10.2 years and 14 ancient female that has been proposed in the last two decades, uses skulls with a mean age estimated at 39.6 ± 9.2 years. a different type of acquisition than MDCT [4, 10, 19]. Rather than capturing an image as separate slices, as Imaging using CBCT in MDCT, CBCT produces a cone-shaped X-ray beam Cone beam CT images were taken with Planmeca that makes it possible to capture the image in a single ProMax 3D Max CBCT (PlanmecaOy, Helsinki, Finland) shot. CBCT recently becoming an alternative of MDCT (Fig. 1). All CBCT scans of skulls were made according to

345 Folia Morphol., 2019, Vol. 78, No. 2

A B

Figure 3. Three-dimensional rendered cone beam computed tomo graphy images showing paranasal sinus volumes; A. Maxillary sinus; B. Frontal sinus; C. Sphenoid sinus. C

matrix and were imported to software Invivo 5.1.2® Figure 2. Measurement sites related to paranasal sinuses on the cone beam computed tomography images; A. Coronal and axial (Anatomage Inc., San Jose, CA, USA) for measure- view of the frontal sinus; B. Coronal and axial view of the sphenoid ment. The anatomical volumetric measurements were sinus; C. Coronal and axial view of the maxillary sinus. done by means of this software in hand tracing. The software uses an inverse present method the pro- a strict standardised scanning protocol; stabilised with gramme reconstructs a 3D model of the sinus from head band and monitored to ensure that they remained the DICOM image sequence on which the volume was motionless throughout the duration of the scan. All selected by cutting out the complementary areas of constructions and measurements were performed on the air-filled area in the three dimensions manually, a 21.3-inch flat-panel colour-active matrix TFT medical then volume measurement was calculated by the display (NEC MultiSync MD215MG, Munchen, Germa- software (Fig. 3). ny) with a resolution of 2048 × 2560 at 75 Hz and All constructions and measurements were per- 0.17-mm dot pitch operated at 11.9 bits. In the mean- formed on a 21.3-inch flat-panel colour-active ma- time, CBCT images of ancient skulls were also obtained trix TFT medical display (NEC MultiSync MD215MG, in same protocol using the same CBCT machines. Munchen, Germany) with a resolution of 2048 × × 2560 at 75 Hz and 0.17-mm dot pitch operat- Image evaluation ed at 11.9 bits. The examiner was also permitted Paranasal sinuses dimensions (width and height) to use enhancements and orientation tools such as and volumes of each sinus were measured from the magnification, brightness, and contrast to improve CBCT data. Coronal and axial images were used for visualisation. measuring the dimensions. The height and width were measured from coronal images whereas the Statistical analysis distance between anterior and posterior (Fig. 2). SPSS 17.0.1 (SPSS, Chicago, ILL) software pro- Paranasal sinus volumes were calculated using gramme used to carry out the statistical analysis. three-dimensional (3D) software. Axial images were In order to determine intra-observer variability, ob- exported in a DICOM file format with a 512 × 512 server performed the analysis twice with an inter-

346 K.O. Demiralp et al., Paranasal sinus parameters in ancient skulls

val of 2 weeks. To assess intra-observer reliability, Table 1. The descriptive analysis of skulls the Wilcoxon matched-pairs signed rank test was N Median (minimum; Mean ± standard used for repeat measurements. Mann-Whitney and maximum) deviation Kruskal-Wallis tests were used to compare paranasal Frontal sinus volume 32 2.77 (0.55; 9.00) 3.18 ± 2.27 sinus parameters in terms of age estimation and Maxillary sinus volume 32 4.72 (1.71; 10.63) 5.13 ± 2.21 gender. Age was divided into three groups (21–40, (right) 41–60, 61–80 years). A p value of less than or equal Maxillary sinus volume 32 5.46 (2.23; 9.73) 5.44 ± 1.92 to 0.05 was considered statistically significant. (left) Sphenoid sinus volume 32 3.47 (0.95; 7.14) 3.46 ± 1.36 RESULTS Maxillary sinus height 32 28.22 (19.59;40.9) 28.57 ± 4.77 (right) Repeated CBCT evaluation and measurements Maxillary sinus width 32 22.18 (14.69; 35.14) 22.99 ± 5.20 indicated no significant intra-examiner difference (right) for the examiner (p > 0.05). Overall intra-observer Maxillary sinus ant- 32 35.14 (23.62; 43.49) 34.77 ± 4.99 consistency was rated at 92% and 95%. All measure- -post distance (right) ments were found to be highly reproducible for the Maxillary sinus height 32 28.51 (17.86; 39.17) 28.41 ± 5.27 examiner and no significant difference was obtained (left) from two measurements (p > 0.05). Maxillary sinus width 32 23.62 (14.40; 35.71) 23.43 ± 5.44 Table 1 shows the descriptive analysis of skulls. The (left) mean FS volume was 2.77 cubic centimetre (cc) (min: Maxillary sinus ant- 32 34.85 (24.77; 43.2) 34.62 ± 4.82 0.55; max: 9.00). The sphenoid sinus volume 3.47 cc -post distance (left) (min: 0.95; max: 7.14). The maxillary sinus left and right Sphenoid sinus height 32 22.75 (14.40; 30.24) 22.86 ± 3.25 volumes were found to be 4.72 cc (min: 1.71; max: Sphenoid sinus width 32 36.00 (19.87; 46.94) 35.07 ± 5.50 10.63) and 5.46 cc (min: 2.23; max: 9.73), respective- Sphenoid sinus 32 22.18 (12.67; 35.71) 22.54 ± 5.52 ly. No significant difference was found between left ant-post distance and right for maxillary sinus volume and dimensions Frontal sinus height 32 22.04 (9.79; 31.15) 22.56 ± 6.15 (p < 0.05). Table 2 shows the measurements accord- Frontal sinus width 32 38.88 (11.04; 64.51) 38.72 ± 13.68 ing to gender. The mean volumes for frontal and Frontal sinus 32 8.93 (4.32; 14.69) 8.92 ± 2.45 ant-post distance sphenoid sinus were found to be higher in males.

Table 2. Measurements according to gender groups Male Female Test statistics N Median (minimum; Mean ± standard N Median (minimum; Mean ± standard Z p maximum) deviation maximum) deviation Frontal sinus volume 14 2.84 (0.55; 8.01) 3.34 ± 2.18 10 1.86 (1.00; 9.00) 2.98 ± 2.49 0.744 0.483 Maxillary sinus volume (right) 13 4.60 (1.71; 10.63) 5.20 ± 2.74 15 4.87 (2.61; 8.44) 5.06 ± 1.72 0.345 0.751 Maxillary sinus volume (left) 14 5.64 (2.23; 9.73) 5.60 ± 2.09 17 5.46 (2.78; 8.86) 5.30 ± 1.82 0.397 0.710 Sphenoid sinus volume 14 4.09 (1.82; 5.13) 3.24 ± 1.23 17 3.61 (1.45; 7.14) 3.65 ± 1.47 0.635 0.050 Maxillary sinus height (right) 14 29.24 (19.59; 40.9) 29.62 ± 5.24 17 27.65 (21.89; 40.61) 27.70 ± 4.31 1.351 0.186 Maxillary sinus width (right) 14 22.61 (18.14; 34.27) 23.41 ± 4.36 17 20.74 (14.69; 35.14) 22.65 ± 5.91 0.596 0.570 Maxillary sinus ant-post distance (right) 14 34.85 (23.62; 43.49) 34.23 ± 6.04 17 35.14 (27.36; 42.05) 35.22 ± 4.07 0.338 0.739 Maxillary sinus height (left) 14 28.51 (17.86; 37.73) 29.07 ± 5.52 17 26.50 (21.02; 39.17) 27.87 ± 5.15 0.695 0.493 Maxillary sinus width (left) 14 22.61 (14.4; 35.71) 23.64 ± 5.02 17 24.77 (14.69; 33.99) 23.26 ± 5.90 0.139 0.891 Maxillary sinus ant-post distance (left) 14 35.14 (24.77; 43.20) 34.68 ± 5.46 17 34.85 (26.21; 41.78) 34.56 ± 4.39 0.099 0.922 Sphenoid sinus height 14 24.92 (14.4; 30.24) 23.62 ± 3.97 17 22.18 (19.01; 27.08) 22.23 ± 2.47 1.451 0.149 Sphenoid sinus width 14 35.28 (28.22; 40.04) 35.04 ± 3.91 17 36.29 (19.87; 46.94) 35.10 ± 6.65 0.238 0.830 Sphenoid sinus ant-post distance 14 23.33 (16.13; 35.71) 24.73 ± 5.47 17 19.58 (12.67; 29.66) 20.74 ± 5.02 1.985 0.048 Frontal sinus height 14 26.07 (10.66; 31.11) 23.62 ± 5.90 16 20.60 (9.79; 31.15) 21.64 ± 6.41 0.894 0.377 Frontal sinus width 14 40.22 (11.04; 61.48) 38.51 ± 14.06 16 38.46 (16.65; 64.51) 38.91 ± 13.79 0.187 0.854 Frontal sinus ant-post distance 14 8.93 (4.32; 10.08) 8.16 ± 1.98 16 9.08 (5.47; 14.69) 9.58 ± 2.69 1.270 0.208

347 Folia Morphol., 2019, Vol. 78, No. 2

Table 3. Measurements results according to age 21–40 years 41–60 years 61–80 years Test statistics N Median Mean ± standard N Median Mean ± standard N Median Mean ± standard χ2 p (minimum; deviation (minimum; deviation (minimum; deviation maximum) maximum) maximum) Frontal sinus 10 2.77 2.50 ± 1.52 9 2.83 2.68 ± 2.33 4 5.51 5.71 ± 2.54 5.841 0.050 volume (0.57; 5.65) (0.55; 8.01) (2.84; 9.00) Maxillary 12 4.72 4.87 ± 1.47 9 4.60 4.90 ± 2.43 7 4.08 4.81 ± 2.26 0.100 0.951 sinus volume (2.61; 7.62) (1.71; 9.34) (2.22; 8.44) (right) Maxillary 13 5.20 5.08 ± 1.40 10 4.87 5.05 ± 2.34 8 5.66 5.91 ± 1.96 1.389 0.499 sinus volume (3.08; 7.68) (2.23; 9.73) (3.24; 8.86) (left) Sphenoid 13 3.61 3.43 ± 1.13 10 2.51 2.94 ± 1.35 8 3.67 4.08 ± 1.68 2.485 0.289 sinus volume (1.45; 4.71) (0.95; 5.13) (2.35; 7.14) Maxillary 13 27.65 27.94 ± 3.25 10 27.51 27.36 ± 4.40 8 29.24 29.05 ± 5.63 0.252 0.882 sinus height (24.48; 34.27) (19.59; 35.71) (21.89; 40.61) (right) Maxillary 13 20.74 21.39 ± 3.29 10 23.19 23.21 ± 6.43 8 21.75 23.69 ± 5.93 0.627 0.731 sinus width (14.69; 25.63) (14.69; 34.27) (17.28; 35.14) (right) Maxillary 13 35.14 35.03 ± 3.73 10 35.14 34.07 ± 6.38 8 32.55 34.02 ± 4.68 0.623 0.732 sinus ant- (27.36; 41.47) (23.62; 42.34) (29.09; 41.19) -post distance (right) Maxillary 13 26.50 27.91 ± 5.12 10 28.08 27.39 ± 5.61 8 28.95 29.59 ± 5.66 0.615 0.735 sinus height (21.02; 36.58) (17.86; 37.73) (23.04; 39.17) (left) Maxillary 13 21.60 21.65 ± 3.99 10 23.62 23.59 ± 6.76 8 22.76 24.41 ± 5.62 0.975 0.614 sinus width (14.69; 26.78) (14.40; 35.71) (16.70; 33.99) (left) Maxillary 13 36.29 35.79 ± 3.80 10 32.40 33.67 ± 6.29 8 32.69 33.48 ± 4.46 1.893 0.388 sinus ant- (27.07; 42.05) (24.77; 43.20) (26.21; 40.03) -post distance (left) Sphenoid 13 21.89 21.78 ± 4.11 10 22.04 22.06 ± 2.46 8 25.20 24.66±2.20 5.086 0.079 sinus height (14.4; 30.24) (18.43; 25.63) (21.60; 27.08) Sphenoid 13 33.70 32.91±6.16 10 37.16 36.61 ± 6.13 8 36.72 36.94 ± 2.84 2.593 0.273 sinus width (19.87; 40.04) (27.65; 46.94) (33.12; 40.91) Sphenoid 13 18.72 20.61 ± 6.29 10 22.61 23.41 ± 5.23 8 22.90 22.9 ± 4.67 2.224 0.329 sinus ant-post (12.67; 31.11) (16.42; 35.71) (16.42; 29.66) distance Frontal sinus 13 22.18 21.88 ± 7.14 10 21.46 21.52 ± 6.37 7 27.36 24.87 ± 5.04 0.788 0.674 height (9.79; 31.15) (10.66; 31.11) (19.30; 30.82) Frontal sinus 13 38.03 34.61 ± 13.19 10 36.98 36.63 ± 11.15 7 55.01 50.48 ± 12.82 5.591 0.050 width (11.04; 54.94) (19.43; 54.94) (33.12; 64.51) Frontal sinus 13 8.64 9.16 ± 3.08 10 8.64 7.98 ± 2.16 7 9.80 9.88 ± 1.83 3.179 0.204 ant-post (4.61; 14.69) (4.32; 10.66) (6.97; 13.25) distance

The distance from anterior-posterior wall of sphenoid DISCUSSION sinus in axial sections was larger in males compared to Forensic anthropology is a key constituent of an females (p > 0.05). Table 3 shows the measurement individual’s antemortem background from skeletal re- according to age groups. The results showed that the mains. Gender, race, age and stature constitute the FS width and volume increased statically. antemortem profile. The durable structure and intact

348 K.O. Demiralp et al., Paranasal sinus parameters in ancient skulls

of paranasal sinuses make them useful for forensic Kawarai et al. [13] stated that paranasal sinuses were anthropology [1, 18]. clearly larger in males than in females after studying in 20 Michel et al. [18] worked to find out whether it Japanese’s 3D CT scans. Amin and Hassan [1] measured was possible to predict the age and gender of an maxillary sinuses with multidetector CT and stated that individual by using FS volume. Sixty-nine anonymised cephalocaudal and size of left maxillary sinus showed CT scans were studied according to FS volume in mm3. significant significance for gender determination. Cor- While sex determination accuracy was found 72.5%, rect predictive accuracy was 70.8% in males and 62.5% there was no correlation between age and frontal in females. In our study, the mean volumes for frontal sinus. Male left side FS volume was stated higher and sphenoid sinus were found to be higher in males. without significant difference. In accordance with In a study done by Teke et al. [23] width, height the literature Michel et al. [18] revealed a significant and length of maxillary sinus were measured. It was difference in FS volume between females and males. stated that all measurements were higher in males Ponde et al. [9] carried out a study on 100 macer- than in females. The mean estimated rate of gender ated skulls and adjudged that there was a significant was detected at 69.3%. However, in this study, no difference regarding sex and location of sinus with significant difference was found between left and predominance of the left side in males. Tatlısumak right for maxillary sinus volume and dimensions. et al. [22] conducted a study on 300 paranasal CT The influence of voxel size on image resolution has scans and estimated that maximum size of FS was been widely confirmed in the literature — small voxel reached between 30 and 40 years of age and that FS sizes generate images with high diagnostic power. volume decreased thereafter. Similarly, in this study, Differences have been observed in the quality of im- the results showed that the FS width and volume ages obtained from different devices when the field increased statically with ageing. of view (FOV) or voxel size is changed. Smaller voxel Uthman et al. [25] found 79.7% accuracy in sizes make CBCT better diagnostic tool than conven- gender estimation by using frontal sinus CT scans, tional radiography and CT scans. When considering whereas this ratio increased to 85.9% when they the bone, measurements made by CBCT images and combined skull measurements and FS measurements. comparing them to digital calliper measurements In another study done by Uthman et al. [26] maxillary (as a gold standard), Sun et al. [21] indicated (by sinus height was the best indicative factor for gender evaluating bone thickness) that bone measurements estimation with overall accuracy of 71.6%. obtained with voxel size of 0.25 mm3 were closer to Buckland-Wright [5] was one of the earliest to the results of direct measurement than images of report sex differences, stating that frontal sinuses in 0.4 mm3 voxels. males were approximately twice as large as in females; however, Yoshino et al. [28] evaluated antero-pos- CONCLUSIONS terior radiographs and found no significant sexual Gender and age estimation are crucial factors in dimorphism through the application of univariate personal identification. The result of the current study statistics. Cox et al. [7] undertook computer-based showed that the paranasal sinus can be combined assessments of radiograph-traced sinus outlines, but in forensic anthropological studies as well as gender obtained no statistically significant sexual variation. predilection. According to Belaldavar et al. [2], the height and REFERENCES area of left FS were better regressors for sex determination among other individual variables with the ratio 1. Amin MF, Hassan EI. Sex identification in Egyptian population using Multidetector Computed Tomography of 64.6% and 63.2%, respectively. Three systemic factors, the maxillary sinus. J Forensic Leg Med. 2012; 19(2): that is the craniofacial configuration, the thickness of 65–69, doi: 10.1016/j.jflm.2011.10.005, indexed in Pu- the frontal bone and growth hormone levels, influence bmed: 22281213. the FS morphology within each population [2]. 2. Belaldavar C, Kotrashetti VS, Hallikerimath SR, et al. Assessment of frontal sinus dimensions to determine In a study done by Kanthem et al. [12] the dimen- sexual dimorphism among Indian adults. J Forensic Dent sions and volume of maxillary sinuses of right and left Sci. 2014; 6(1): 25–30, doi: 10.4103/0975-1475.127766, side were much larger in males than in females. Sexual indexed in Pubmed: 24695810. dimorphism according to volume was estimated at 3. Brothwell DR. Diggingupbones: theexcavation, treatment, andstudy of humanskeletalremains: Cornell University 85.46% for right side and 78.38% for left side. Press. 1981.

349 Folia Morphol., 2019, Vol. 78, No. 2

4. Brown AA, Scarfe WC, Scheetz JP, et al. Linear accuracy of tions. Ann Hum Biol. 2015; 42(4): 308–322, doi: 10.3109/ cone beam CT derived 3D images. Angle Orthod. 2009; 03014460.2015.1048288, indexed in Pubmed: 26366990. 79(1): 150–157, doi: 10.2319/122407-599.1, indexed in 18. Michel J, Paganelli A, Varoquaux A, et al. Determina- Pubmed: 19123719. tion of sex: interest of frontal sinus 3D reconstructions. 5. Buckland-Wright JC. A radiographic examination of frontal J Forensic Sci. 2015; 60(2): 269–273, doi: 10.1111/1556- sinuses in early British populations. Man. 1970; 5(3): 512, 4029.12630, indexed in Pubmed: 25676659. doi: 10.2307/2798956. 19. Moreira CR, Sales MAO, Lopes PML, et al. Assessment of 6. Buikstra JE, Ubelaker DH. Standards for data collection linear and angular measurements on three-dimensional from human skeletal remains: proceedings of a seminar cone-beam computed tomographic images. Oral Surg at the Field Museum of Natural History. 1994. Oral Med Oral Pathol Oral Radiol Endod. 2009; 108(3): 7. Cox M, Malcolm M, Fairgrieve SI. A new digital meth- 430–436, doi: 10.1016/j.tripleo.2009.01.032, indexed in od for the objective comparison of frontal sinuses for Pubmed: 19386521. identification. J Forensic Sci. 2009; 54(4): 761–772, 20. Prabhat M, Rai S, Kaur M, et al. Computed tomography doi: 10.1111/j.1556-4029.2009.01075.x, indexed in Pu- based forensic gender determination by measuring the bmed: 19486246. size and volume of the maxillary sinuses. J Forensic Dent 8. David MP, Saxena R. Use of frontal sinus and nasal septum Sci. 2016; 8(1): 40–46, doi: 10.4103/0975-1475.176950, patterns as an aid in personal identification: A digital indexed in Pubmed: 27051222. radiographic pilot study. J Forensic Dent Sci. 2010; 2(2): 21. Sun Z, Smith T, Kortam S, et al. Effect of bone thickness 77–80, doi: 10.4103/0975-1475.81286, indexed in Pu- on alveolar bone-height measurements from cone-beam bmed: 21731344. computed tomography images. Am J Orthod Dentofacial 9. del SenoFrontal VA Anatomical variations of the frontal Orthop. 2011; 139(2): e117–e127, doi: 10.1016/j.ajo- sinus. Int J Morphol. 2008; 26(4): 803–808. do.2010.08.016, indexed in Pubmed: 21300222. 10. Farman AG, Scarfe WC. Development of imaging selection 22. Tatlisumak E, Ovali GY, Asirdizer M, et al. CT study on mor- criteria and procedures should precede cephalometric assess- phometry of frontal sinus. Clin Anat. 2008; 21(4): 287–293, ment with cone-beam computed tomography. Am J Orthod doi: 10.1002/ca.20617, indexed in Pubmed: 18428994. Dentofacial Orthop. 2006; 130(2): 257–265, doi: 10.1016/j. 23. Teke HY, Duran S, Canturk N, et al. Determination of ajodo.2005.10.021, indexed in Pubmed: 16905073. gender by measuring the size of the maxillary sinuses in 11. Günay Y, Altinkök M. The value of the size of foramen computerized tomography scans. Surg Radiol Anat. 2007; magnum in sex determination. J Clin Forensic Med. 2000; 29(1): 9–13, doi: 10.1007/s00276-006-0157-1, indexed in 7(3): 147–149, doi: 10.1054/jcfm.2000.0430, indexed in Pubmed: 17171233. Pubmed: 16083665. 24. Ubelaker D, Grant L. Human skeletal remains: Preservation 12. Kanthem RK, Guttikonda VR, Yeluri S, et al. Sex determi- or reburial? Am J Physical Anthropology. 1989; 32(S10): nation using maxillary sinus. J Forensic Dent Sci. 2015; 249–287, doi: 10.1002/ajpa.1330320511. 7(2): 163–167, doi: 10.4103/0975-1475.154595, indexed 25. Uthman AT, Al-Rawi NH, Al-Naaimi AS, et al. Evaluation in Pubmed: 26005308. of frontal sinus and skull measurements using spiral CT 13. Kawarai Y, Fukushima K, Ogawa T, et al. Volume quanti- scanning: an aid in unknown person identification. Foren- fication of healthy paranasal cavity by three-dimensional sic Sci Int. 2010; 197(1-3): 124.e1–124.e7, doi: 10.1016/j. CT imaging. Acta Otolaryngol Suppl. 1999; 540: 45–49, forsciint.2009.12.064, indexed in Pubmed: 20097024. indexed in Pubmed: 10445079. 26. Uthman A, Al-Rawi N, Al-Naaimi A, et al. Evaluation of 14. Krogman WM. Determination of sexandparturition. maxillary sinus dimensions in gender determination using Thehumanskeleton in forensicmedicine. 1986: 208–259. helical CT scanning. J Forensic Scien. 2011; 56(2): 403–408, 15. Lascala CA, Panella J, Marques MM. Analysis of the accu- doi: 10.1111/j.1556-4029.2010.01642.x. racy of linear measurements obtained by cone beam com- 27. van Vlijmen OJC, Bergé SJ, Swennen GRJ, et al. Com- puted tomography (CBCT-NewTom). Dentomaxillofac Ra- parison of cephalometric radiographs obtained from diol. 2004; 33(5): 291–294, doi: 10.1259/dmfr/25500850, cone-beam computed tomography scans and conven- indexed in Pubmed: 15585804. tional radiographs. J Oral Maxillofac Surg. 2009; 67(1): 16. Licata M, Borgo M, Armocida G, et al. New paleoradi- 92–97, doi: 10.1016/j.joms.2008.04.025, indexed in ological investigations of ancient human remains from Pubmed: 19070753. North West Lombardy archaeological excavations. Skeletal 28. Yoshino M, Miyasaka S, Sato H, et al. Classification system Radiol. 2016; 45(3): 323–331, doi: 10.1007/s00256-015- of frontal sinus patterns by radiography. Its application 2266-6, indexed in Pubmed: 26481783. to identification of unknown skeletal remains. Forensic 17. Márquez-Grant N. An overview of age estimation in fo- Sci Int. 1987; 34(4): 289–299, doi: 10.1016/0379- rensic anthropology: perspectives and practical considera- 0738(87)90041-7, indexed in Pubmed: 3623370.

Prevalence of second canal in the mesiobuccal root of permanent maxillary molars from a Turkish subpopulation: a cone-beam computed tomography study G. Magat, S. Hakbilen

Department of Oral and Maxillofacial Radiology, Faculty of Dentistry, Necmettin Erbakan University, Konya, Turkey

[Received: 19 July 2018; Accepted: 19 September 2018]

Background: Permanent maxillary molar teeth have the most complex root and canal systems. This study aimed to determine the frequency and the localisation of secondary mesiobuccal canal (MB2) in the mesiobuccal roots of permanent maxillary molars in a Turkish subpopulation using cone-beam computed tomography (CBCT) and compare them by different age groups and gender. Materials and methods: In this retrospective study, a total of 800 maxillary molar teeth in 200 subjects were investigated. The distances between canals were calculated from the centre point of the MB2 canal (PMB2) to the centre point of the mesiobuccal canal (PMB1) and the centre point of the palatal canal (PP). Results: The incidence of MB2 canals was found to be 33.5% (23.2% in females, 43.6% in males) in 200 subjects. A significant difference was found between age groups and genders in terms of the frequency of MB2 (p < 0.05). The distances between PMB1- -PMB2 were 2.95 ± 0.58 mm, 3.08 ± 0.67 mm for 1MM and 2MM, respectively. For 1MM and 2MM, the averages of PMB2-PP distances were 5.81 ± 1.09 mm and 5.55 ± 1.09 mm, respectively. The contralateral formation of the MB2 canals was 90.5% (16, 26) in the first molars and 93.9% in the second molars (17, 27). The contingency rate in the first quadrant was 86.0% (16, 17) and 83.5% (26, 27) in the second quadrant. Conclusions: It is important for the clinician to know the high probability of finding the MB2 canal in the maxillary first and second molars. CBCT imaging can facilitate the clinician to detect complex and variable root canal anatomy. (Folia Morphol 2019; 78, 2: 351–358)

Key words: maxillary molars, secondary mesiobuccal canal, cone-beam computed tomography

INTRODUCTION broad individual, genetic and ethnic diversity, clini- The success of endodontic treatment relies on the cians should seek extra canals [29, 47]. complete cleaning, shaping, and filling of the entire Permanent maxillary molar teeth have the most root canal system to prevent recontamination. One complex root and canal systems [3, 52]. Their canal of the endodontic treatment failures is the presence system is very diverse [7]. The high failure rate stems of an additional canal in the teeth [1, 13]. For this from the impossibility of locating the presence and reason, detailed information of the morphology of location of the secondary mesiobuccal canal (MB2) the root canals is of great importance. Given the in the mesiobuccal roots of the permanent maxillary

Address for correspondence: Assist. Prof. Dr. G. Magat, Necmettin Erbakan University Dentistry Faculty Oral Radiology Department, Konya, Turkey, tel: +90 505 945 61 57, e-mail: [email protected]

351 Folia Morphol., 2019, Vol. 78, No. 2

molars [9]. Due to excessive dentin accumulation in Necmettin Erbakan University, Konya, Turkey (decision opening the canal and difficulty of visualisation of no: 2018.01). In this retrospective study, CBCT images molar teeth, its location in clinical practice is quite from 200 subjects were acquired from the archive of complicated [7]. Throughout the literature, the pres- Oral and Maxillofacial Radiology Department, Faculty ence of the MB2 canal ranges from 18.6% to 96.1% of Dentistry, Necmettin Erbakan University, Konya, [20, 22, 31]. When the MB2 canal cannot be located Turkey. The CBCT images were obtained from Janu- or properly treated, it may contribute to continued ary 2014 to December 2017 for diagnosis and dental patient pain or root canal failure [52]. treatment. For this study, we investigated CBCT im- The MB2 canals are visualised by histological sec- ages of 625 subjects. However, only 200 CBCT images tions [44], diaphanisation [17], magnifying loupes were evaluated because of our inclusion criteria. The [41], endodontic surgical microscope [26], scanning CBCT images of 101 men and 99 women between electron microscope [44], periapical radiography [40], 13 and 67 years old were enrolled according to follow- micro-computed tomographic analysis [27, 51], and ing criteria; (i) maxillary permanent molars with fully cone-beam computed tomography (CBCT) [7–9, 57, erupted and matured apices; and (ii) no root canal 58]. However, some of these techniques are invasive, fillings, caries, posts, crown restorations, apical peri- disadvantageous and can only be used to investigate odontitis, resorption, calcification or any pathology. extracted teeth [50]. Nowadays, CBCT has become The 3D Accuitomo CBCT machine (J Morita Manu- more widely used because it can provide more reli- facturing Corp, Kyoto, Japan) was operated at 90 kVp able diagnostics in the field of endodontics [11, 23, and 5 mA with a 17.5-second exposure time with 34] and can perform morphological analysis of the a 100 × 100 mm field of view, and with a 0.25 × root canal system [9, 30, 32, 37, 58]. The American × 0.25 mm voxel size. All images were taken by Association of Endodontists (AAE) and the European oral and maxillofacial radiologists with 5+ years of Society of Endodontology (ESE) have released posi- experience according to the manufacturer recom- tion statements stating that limited field of view mendations. A total of 400 maxillary permanent first CBCT should be considered the imaging modality of molars and 400 maxillary permanent second molars choice for initial endodontic treatment of teeth with were analysed using the CBCT machine software pro- the potential for extra canals and suspected complex gramme (i-Dixel, J Morita Manufacturing Corp, Kyoto, morphology (AAE and AAOMR Joint Position State- Japan) on the axial plane at 0.5 mm intervals and ment 2015, ESE 2014). The studies confirm that the a 1 mm thickness. The examination was made from CBCT images are as accurate as the modified staining the coronal to apical. To standardise the observation technique to describe the root canal system and far and measurements, the MB2 was searched in a 1 mm more accurate than periapical radiographs [37]. apical of the pulp chamber floor (Fig. 1). In the literature, there were several studies using The distances between canals were calculated in the CBCT to investigate the prevalence and morphol- millimetres from the centre point of the MB2 canal ogy of the MB2 canal of permanent maxillary molars (PMB2) to the centre point of the mesiobuccal canal in various ethnicities [16, 19, 25, 45, 49, 58]. However, (PMB1) and the centre point of the palatal canal there were not sufficient studies that evaluated the (PP) (Fig. 2) according to the protocol described by MB2 canal and its geometric localisation in Turkish Betancourt et al. [7, 8]. population using CBCT. In addition, there are no re- The images were analysed by licensed oral and ports about the symmetry in the MB2 canal between maxillofacial radiologist (S.H.) with 4 years’ experi- contralateral and adjacent molars in a Turkish popula- ence and previous training, and using consensus. The tion. In this regards, this study aimed to determine same investigator examined the radiographs twice in (i) the frequency and (ii) the geometric localisation a period of 1 month and the intra-observer variability of MB2 canal (distances between the canals) in the was calculated. The data collected in terms of gender, mesiobuccal roots of permanent maxillary molars age, side, and distances between the different points from a Turkish subpopulation using CBCT and (iii) were recorded. compare them by different age groups and gender. Statistical analysis MATERIALS AND METHODS The software SPSS/PC + v. 21.0 (SPSS, Chicago, The present study was approved by the Human IL) was used for statistical analysis. The descriptive Research Ethics Committee of the Faculty of Dentistry, statistics (mean ± standard deviation), the c2 test 352 G. Magat, S. Hakbilen, Prevalence of secondary mesiobuccal canals of permanent maxillary molars

and the Spearman correlation tests were used. The reliability of data was analysed using the kappa test. A value of p < 0.05 was considered significant.

Results In this retrospective CBCT-based study, 400 maxillary first molar and 400 maxillary second molar teeth were examined in 200 subjects (400 men and 99 women). The age range of the individuals was 13–67 years and the mean age was 29.45 ± 11.73 years. The kappa values for intra-observer reliability were ranged from 0.90 to 0.96. The incidence of MB2 canals was found to be 33.5% (23.2% in females, Figure 1. Axial view of cone-beam computed tomography (CBCT) images of maxillary first molars with secondary mesiobuccal 43.6% in males) in 200 subjects. The prevalence of canals (MB2) (arrows). the MB2 canal was 19.65% (n = 79) and 17.70% (n = 71) in the first maxillary molars (1MM) and the second maxillary molars (2MM), respectively. MB2 canals had a homogeneous distribution of 49.5% on the right and 50.5% on the left. The total MB2 prevalence was 18.7% if we assessed on tooth basis. These MB2 canals were in 44 males and 23 females. There was a statistically significant difference according to the gender (p = 0.002; Fig. 3). Males had more prevalence of MB2 than females (Table 1). The average age of the individuals with MB2 canal was 25.38 ± 10.16 years. MB2 canal prevalence of permanent maxillary molars according to age groups was shown in Table 2. Pearson c2 test revealed that as the age progressed, the prevalence of MB2 was decreased (p = 0.001). Figure 2. Axial view of left maxillary first molar. The distance in The average distances between PMB1-PMB2 were the lines drawn between the points was measured in millimetres; 2.95 ± 0.58 mm, 3.08 ± 0.67 mm for 1MM and PMB1 — centre of mesiobuccal canal; PMB2 — centre of secondary mesiobuccal canal; PP — centre palatal canal. 2MM, respectively. For 1MM and 2MM, the averages of PMB2-PP distances were 5.81 ± 1.09 mm and 5.55 ± 1.09 mm, respectively. The contralateral and adjacent occurrence probability for each tooth position was calculated between patients with MB2 canals. The contralateral formation of the MB2 canals was 90.5% (16, 26) in the first permanent maxillary molars and 93.9% in the second permanent maxillary molars (17, 27). The contingency rate in the first quadrant was 86.0% (16, 17) and 83.5% (26, 27) in the second quadrant. Table 3 shows the validity rates of an existing MB2 canal to estimate the MB2 canal on the contralateral side. The possibility of contralateral occurrence of MB2 canals in permanent first maxillary molar teeth (16, 26) was considerably better than for other tooth Figure 3. The frequency of secondary mesiobuccal canals according to gender. positions. Their sensitivity was 93.9% and positive

353 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. Prevalence of secondary mesiobuccal canals in the first and second permanent maxillary molars according to gender Gender Tooth 16 Tooth 26 Tooth 17 Tooth 27 Female (n = 100) 9.1% 12.1% 13.1% 14.1% Male (n = 100) 28.7% 28.7% 21.8% 21.8%

Table 2. Prevalence of secondary mesiobuccal canals in the first and second permanent maxillary molars according to age Age [years] Tooth 16 Tooth 26 Tooth 17 Tooth 27 13–25 (n = 88) 27.3% 30.7% 22.7% 25.0% 26–40 (n = 62) 14.5% 16.1% 17.7% 16.1% > 40 (n = 50) 10.0% 8.05 8.0% 8.0%

Table 3. Probability accuracy assessment of secondary mesiobuccal canals (MB2) in different tooth positions Probability of an MB2 canal in the contralateral tooth Probability of an MB2 canal in the adjacent tooth Test values 16/26 17/27 16/17 26/27 Sensitivity 90.5% 93.9% 86.0% 83.5% Specificity 56.0% 69.4% 42.8% 38.9% Positive predictive value 88.8% 93.3% 87.6% 86.2% Negative predictive value 60.5% 71.4% 39.5% 34.1% False-positive result 11.2% 6.7% 12.4% 13.8% False-negative result 9.5% 6.1% 14.0% 16.5%

predictive value was 93.3%. Their false-positive and but the detection can be difficult in many patients false-negative results were lowest. [21, 46]. The intraclass correlation coefficient scores of the The permanent molars may have more than one intra-observer agreement were found to be 0.862, or two canals of the first and second molar- me 0.879, 0902, 0.880, 0.895 and 0.901 for distances siobuccal root; can also be separated from various and absent-present of MB2. sides and may have lateral openings [14]. Vertucci [52] proposed a classification involving eight differ- DISCUSSION ent types, while Weine et al. [53] studied root canal A comprehensive knowledge of root canal mor- anatomy by separating one or two channels into four phology and possible variations is a prerequisite for groupings. However, we have not classified the teeth successful endodontic treatment. Anatomical varia- according to any of these classification systems. Only, tions should be examined both clinically and radio- this retrospective study provides MB2 prevalence of logically. Endodontic therapy should be initiated with maxillary molar teeth using CBCT images based on proper preparation to allow access to the cavity; this age and gender. can facilitate the investigation and successful detec- There are too many techniques in literature to visu- tion of all root canal holes [10]. alise the accessory canals [17, 26, 41, 44]. Although Detection of all channels in the root canal sys- the enlargement systems are useful for finding the tem and adequate clearing of these channels affect MB2 canal, they show only the average orifice of the the ultimate success of treatment. For this reason, MB2 canal, not the entire root canal system. How- each tooth should be considered to have an extra ever, if the pulp cavity is not opened sufficiently, the channel. If possible, all extra canals should be de- MB2 channel cannot be detected by magnification. tected [41]. MB2 canal was chosen as a parameter For inclined or rotated molar teeth, the magnifica- for this study because it is widely common [31], tion is less effective because of a sharp and moder-

354 G. Magat, S. Hakbilen, Prevalence of secondary mesiobuccal canals of permanent maxillary molars

ate angle of the tooth [8]. Stopko [46] stated that detection rate of MB2 in females can be explained these microsurgical devices alone are inadequate by demineralisation and loss of bone mass, which to find and measure the MB2 canal. Moreover, the is three times higher in females than in males [6]. most commonly used method for the detection of This will prevent accurate tracking of the canal by accessory canals in everyday practice was periapical computed tomography due to lack of contrast [4]. radiographs. They are important for endodontic pre- Neaverth et al. [36] stated that the older patients operative diagnosis. However, they can only provide had fewer MB2 canal than younger ones because of two-dimensional information limits the diagnostic an increase in the canal calcification, tertiary dentin efficiency. In addition, anatomical constructions are formation and porosity of the cortical bone. Zhang et difficult to interpret in terms of factors such as super- al. [57] said that secondary dentin deposition sepa- imposed anatomical structures, increased zygomatic rated a flattened canal into two canals or even create bone density, or embedded teeth [39]. Barton et al. a calcifiedimperforate canal. In the present study, it [5] and Abuabara et al. [2] reported the frequency was found that as the age progressed, the prevalence of MB2 canals in molar teeth was 39.2% and 8%, of MB2 decreased (p < 0.01), which is consistent with respectively. They noted that the effectiveness of the results of Neaverth et al. [36], Thomas et al. [48], conventional periapical X-rays was low. Nattress and Betancourt et al. [7] and Zhang et al. [57]. For this rea- Martin [35] stated that buccolingual radiographs son, we recommend that the maxillary molar teeth be alone were not reliable for detecting multiple canals. directed more attention to seeking and finding MB2 The use of CBCT imaging is not recommended for all canals in the MB roots, especially in young patients. endodontic treatment situations. At the same time, It would be very difficult to identify additional canal recent studies have shown that CBCT imaging shows in the CBCT image because older people had more excellent accuracy values compared to periapical ra- calcified canals and the diameter of the additional diographs [54]. For this reason, it is very important canals was smaller than the MB1 canal. to know and use additional tools to help diagnose The MB2 canal is commonly located mesiopala- MB2 canal detection. tally to the MB1 canal [44]. The location of the MB2 The incidence of MB2 canals in 1MMs and 2MMs canal is often indicated by in vitro studies [17, 18]. varies from 8.0% to 96.1% [4, 35, 47, 55–57]. In this There were only two studies that demonstrated the study, the prevalence of MB2 was 18.75%, which is MB2 canal location using in vivo CBCT [7, 8]. These lower than some studies [4, 7, 12, 31, 38, 42]. Stude- studies were conducted by the same authors. They baker et al. [47] found that an additional canal was found that the MB2 canal was located in the 1MM 11.7% in CBCT images compared to other techniques. 2.68 ± 0.49 mm palatally and 1.25 ± 0.34 mesially Silva et al. [45] reported that the MB2 canal incidence to the MB1 canal. In the 2MM it was located 2.41 ± was higher than in first molars than in second molars ± 0.64 mm palatally and 0.98 ± 0.33 mm mesially. (42–34%). These differences may result from sample In this study, the average distance of MB2 to the size, methodology and ethnic differences. Alexander MB1 canal of the 1MM was 2.95 ± 0.58 mm and its et al. [4] stated that the reason of differences between distance to the PP1 canal was 5.81 ± 1.09 mm. In the results obtained from CBCT may be CBCT parameter 2MM, while the average distance to the MB1 canal settings and software differences. In addition, detec- was 3.08 ± 0.67 mm, the distance to the PP1 was tion of the MB2 canals is probably difficult due to 5.55 ± 1.09 mm. These distances were greater than the dentin cap covering the orifice [15, 24]. Another reported by Betancourt et al. [8], and Gorduysus et difficulty is that some of these canals can be curved. al. [18]. Our results were similar to study conducted Their coronal part may contain one or two sudden by Gilles et al. [17]. They found that the MB2 canal folds [22, 31]. This may explain the fact that MB2 was located mesially to the MB1 canal at a distance canals are detected more often in vitro studies than of 2.31 mm in the 1MM and 2.06 mm in the 2MM in vivo studies [3]. by scanning electronic microscopy. The differences In this study, it was found that males had more between the results may be due to differences in the prevalence of MB2 than females. These results are techniques used, measurement techniques, or sample similar to some studies in the literature [7, 15, 33, sizes. We believe that the inter-orifice distances can 42]. However, there were also studies which found no differ at different measurement levels. We measured differences between genders [19, 43, 58]. The smaller at 1 mm below the pulpal floor level. Zhang et al.

355 Folia Morphol., 2019, Vol. 78, No. 2

[57] reported that a larger distance ratio of MB1-PP mittee on human experimentation (institutional and to distobuccal (DB)-PP (> 1.26) indicated a highly national) and with the Helsinki Declaration of 1975, probable existence of an MB2 canal. They stated that as revised in 2008. CBCT imaging with a small voxel size is particularly important in determining the presence of an MB2 CONCLUSIONS canal if clinicians fail to locate it. Since CBCT imaging In conclusion, the results of the present study in- is not used routinely in the diagnosis of endodontic dicate that the MB2 canal is found in 19.75% of the patients, it is expected that our results will be useful 1MM and 17.75% of the 2MM. the average distance for clinicians to treat teeth with MB2 canal or complex of MB2 to the MB1 canal of the 1MM was 2.95 ± morphology potential during initial therapy. ± 0.58 mm and its distance to the PP1 canal was In the literature, only two studies have reported 5.81 ± 1.09 mm. In the 2MM, while the average dis- the adjacent or contralateral formation of MB2 canals tance to the MB1 canal was 3.08 ± 0.67 mm, the dis- [4, 28]. In this study, bilateral symmetry formation for tance to the PP1 was 5.55 ± 1.09 mm. It is important for the maxillary first and second molar teeth was 88% clinicians to know the localisation and the probability of and 93%, respectively. These results were similar to finding the MB2 canal in the first and second permanent that reported by Alexander et al. [4]. However, these maxilla. The occurrence possibility of these canals was results were higher than the study of Kim et al. [28]. also found high in the contralateral and adjacent molar We found that the adjacent concurrence was 87% and teeth. CBCT imaging can facilitate the clinician to detect 82% for the maxillary first and second molar teeth on complex and variable root canal anatomy. the same side, respectively, which is approximately consistent with the results of a previous study [4]. REFERENCES However, the results of the present study were higher 1. Abarca J, Gómez B, Zaror C, et al. Assessment of mesial than that of Kim et al. [28] where the finding was root morphology and frequency of MB2 canals in maxil- 64%. While our sample size (n = 200) was equal to lary molars using cone beam computed tomography. Int J Morphol. 2015; 33(4): 1333–1337, doi: 10.4067/s0717- that in the study by Alexander et al. [4]; it was lower 95022015000400023. than that in the study by Kim et al. [28] (n = 351). 2. Abuabara A, Baratto-Filho F, Aguiar Anele J, et al. Efficacy Their study was carried out about 6 years before ours. of clinical and radiological methods to identify second These differences may be due to improvements in mesiobuccal canals in maxillary first molars. Acta Odontol CBCT software. These improvements can increase im- Scand. 2013; 71(1): 205–209, doi: 10.3109/00016357.20 11.654262, indexed in Pubmed: 22320229. age resolution and improve diagnostic accuracy and 3. Alaçam T, Tinaz AC, Genç O, et al. Second mesiobuccal sensitivity. To our knowledge, this is the first study canal detection in maxillary first molars using microscopy to report the symmetry of the MB2 canal between and ultrasonics. Aust Endod J. 2008; 34(3): 106–109, contralateral and adjacent molars in a Turkish popula- doi: 10.1111/j.1747-4477.2007.00090.x, indexed in Pubmed: 19032644. tion. If an MB2 canal is found on a molar, clinicians 4. Alexander FN, Dirk H, Corne PT, et al. The prevalence of should consider the possibility of additional canals second canals in the mesiobuccal root of maxillary molars: in adjacent molars and more in contralateral molars. A cone beam computed tomography study. Aust Endod However, this study had some limitations. First, J. 2018 [Epub ahead of print], doi: 10.1111/aej.12263, our sample consisted of subjects who referred to indexed in Pubmed: 29573065. 5. Barton DJ, Clark SJ, Eleazer PD, et al. Tuned-aperture com- our faculty for a dental examination. This implies puted tomography versus parallax analog and digital ra- that our sample was not representative of Turkish diographic images in detecting second mesiobuccal canals adults in general. Therefore, there is a need for stud- in maxillary first molars. Oral Surg Oral Med Oral Pathol ies conducted throughout our country. Second, the Oral Radiol Endod. 2003; 96(2): 223–228, doi: 10.1016/ S1079210403000611, indexed in Pubmed: 12931097. study had a retrospective design. Hence, it was not 6. Benson BW, Prihoda TJ, Glass BJ. Variations in adult cortical possible to obtain the patient’s entire history. Some bone mass as measured by a panoramic mandibular index. cultural and individual characteristics may affect the Oral Surg Oral Med Oral Pathol. 1991; 71(3): 349–356, prevalence of the MB2 canal. Third, only one radiolo- indexed in Pubmed: 2011361. 7. Betancourt P, Navarro P, Cantín M, et al. Cone-beam com- gist has evaluated the measurements. However, the puted tomography study of prevalence and location of intra-observer agreement values were perfect. MB2 canal in the mesiobuccal root of the maxillary second All procedures performed were in accordance molar. Int J Clin Exp Med. 2015; 8(6): 9128–9134, indexed with the ethical standards of the responsible com- in Pubmed: 26309568.

356 G. Magat, S. Hakbilen, Prevalence of secondary mesiobuccal canals of permanent maxillary molars

8. Betancourt P, Navarro P, Muñoz G, et al. Prevalence and 21. Henry BM. The fourth canal: its incidence in maxillary first location of the secondary mesiobuccal canal in 1,100 molars. J Can Dent Assoc. 1993; 59(12): 995–996, indexed maxillary molars using cone beam computed tomography. in Pubmed: 8275393. BMC Med Imaging. 2016; 16(1): 66, doi: 10.1186/s12880- 22. Ibarrola JL, Knowles KI, Ludlow MO, et al. Factors affecting 016-0168-2, indexed in Pubmed: 27908285. the negotiability of second mesiobuccal canals in maxil- 9. Blattner TC, George N, Lee CC, et al. Efficacy of cone- lary molars. J Endod. 1997; 23(4): 236–238, doi: 10.1016/ beam computed tomography as a modality to accurately S0099-2399(97)80054-4, indexed in Pubmed: 9594773. identify the presence of second mesiobuccal canals in 23. Ilgüy D, Ilgüy M, Fisekcioglu E, et al. Detection of jaw maxillary first and second molars: a pilot study. J Endod. and root fractures using cone beam computed tomogra- 2010; 36(5): 867–870, doi: 10.1016/j.joen.2009.12.023, phy: a case report. Dentomaxillofac Radiol. 2009; 38(3): indexed in Pubmed: 20416435. 169–173, doi: 10.1259/dmfr/54020205, indexed in Pub- 10. Christie WH, Thompson GK. The importance of endodontic med: 19225088. access in locating maxillary and mandibular molar canals. 24. Imura N, Hata GI, Toda T, et al. Two canals in mesiobuc- J Can Dent Assoc. 1994; 60(6): 527–32, 535, indexed in cal roots of maxillary molars. Int Endod J. 1998; 31(6): Pubmed: 8032994. 410–414, indexed in Pubmed: 15551608. 11. Cotton TP, Geisler TM, Holden DT, et al. Endodontic ap- 25. Parker J, Mol A, Rivera EM, et al. CBCT uses in clinical plications of cone-beam volumetric tomography. J Endod. endodontics: the effect of CBCT on the ability to locate 2007; 33(9): 1121–1132, doi: 10.1016/j.joen.2007.06.011, MB2 canals in maxillary molars. Int Endod J. 2017; indexed in Pubmed: 17931947. 50(12): 1109–1115, doi: 10.1111/iej.12736, indexed in 12. Degerness RA, Bowles WR. Dimension, anatomy and mor- Pubmed: 27977863. phology of the mesiobuccal root canal system in maxillary 26. Karaman GT, Onay EO, Ungor M, et al. Evaluating the molars. J Endod. 2010; 36(6): 985–989, doi: 10.1016/j. potential key factors in assessing the morphology of joen.2010.02.017, indexed in Pubmed: 20478451. mesiobuccal canal in maxillary first and second molars. 13. Del Fabbro M, Taschieri S, Testori T, et al. Surgical versus Aust Endod J. 2011; 37(3): 134–140, doi: 10.1111/j.1747- non-surgical endodontic re-treatment for periradicular 4477.2010.00240.x, indexed in Pubmed: 22117721. lesions. Cochrane Database Syst Rev. 2007(3): CD005511, 27. Kim Y, Chang SW, Lee JK, et al. A micro-computed tomog- doi: 10.1002/14651858.CD005511.pub2, indexed in raphy study of canal configuration of multiple-canalled Pubmed: 17636803. mesiobuccal root of maxillary first molar. Clin Oral Investig. 14. Eder A, Kantor M, Nell A, et al. Root canal system in 2013; 17(6): 1541–1546, doi: 10.1007/s00784-012-0852- the mesiobuccal root of the maxillary first molar: an 8, indexed in Pubmed: 23053704. in vitro comparison study of computed tomography 28. Kim Y, Lee SJ, Woo J. Morphology of maxillary first and and histology. Dentomaxillofac Radiol. 2006; 35(3): second molars analyzed by cone-beam computed tomog- 175–177, doi: 10.1259/dmfr/73909470, indexed in Pub- raphy in a korean population: variations in the number med: 16618851. of roots and canals and the incidence of fusion. J Endod. 15. Fogel HM, Peikoff MD, Christie WH. Canal configura- 2012; 38(8): 1063–1068, doi: 10.1016/j.joen.2012.04.025, tion in the mesiobuccal root of the maxillary first indexed in Pubmed: 22794206. molar: a clinical study. J Endod. 1994; 20(3): 135–137, 29. Kontakiotis EG, Tzanetakis GN. Four canals in the mesial doi: 10.1016/S0099-2399(06)80059-2, indexed in Pub- root of a mandibular first molar. A case report under med: 7996086. the operating microscope. Aust Endod J. 2007; 33(2): 16. Ghobashy AM, Nagy MM, Bayoumi AA. Evaluation of root 84–88, doi: 10.1111/j.1747-4477.2007.00068.x, indexed and canal morphology of maxillary permanent molars in in Pubmed: 17697250. an egyptian population by cone-beam computed tomog- 30. Kottoor J, Velmurugan N, Sudha R, et al. Maxillary first raphy. J Endod. 2017; 43(7): 1089–1092, doi: 10.1016/j. molar with seven root canals diagnosed with cone-beam joen.2017.02.014, indexed in Pubmed: 28476465. computed tomography scanning: a case report. J Endod. 17. Gilles J, Reader A. An SEM investigation of the mesiolin- 2010; 36(5): 915–921, doi: 10.1016/j.joen.2009.12.015, gual canal in human maxillary first and second molars. indexed in Pubmed: 20416446. Oral Surg Oral Med Oral Pathol. 1990; 70(5): 638–643, 31. Kulild JC, Peters DD. Incidence and configuration of canal indexed in Pubmed: 2234885. systems in the mesiobuccal root of maxillary first and 18. Görduysus MO, Görduysus M, Friedman S. Operating second molars. J Endod. 1990; 16(7): 311–317, indexed microscope improves negotiation of second mesiobuc- in Pubmed: 2081944. cal canals in maxillary molars. J Endod. 2001; 27(11): 32. La SH, Jung DH, Kim EC, et al. Identification of independ- 683–686, indexed in Pubmed: 11716081. ent middle mesial canal in mandibular first molar using 19. Guo J, Vahidnia A, Sedghizadeh P, et al. Evaluation of root cone-beam computed tomography imaging. J Endod. and canal morphology of maxillary permanent first molars 2010; 36(3): 542–545, doi: 10.1016/j.joen.2009.11.008, in a North American population by cone-beam computed indexed in Pubmed: 20171380. tomography. J Endod. 2014; 40(5): 635–639, doi: 10.1016/j. 33. Lee JH, Kim KD, Lee JK, et al. Mesiobuccal root canal joen.2014.02.002, indexed in Pubmed: 24767556. anatomy of Korean maxillary first and second molars by 20. Hartwell G, Bellizzi R. Clinical investigation of in vivo cone-beam computed tomography. Oral Surg Oral Med endodontically treated mandibular and maxillary molars. Oral Pathol Oral Radiol Endod. 2011; 111(6): 785–791, J Endod. 1982; 8(12): 555–557, doi: 10.1016/S0099- doi: 10.1016/j.tripleo.2010.11.026, indexed in Pub- 2399(82)80016-2, indexed in Pubmed: 6962275. med: 21439860.

357 Folia Morphol., 2019, Vol. 78, No. 2

34. Lofthag-Hansen S, Huumonen S, Gröndahl K, et al. Limited 446–450, doi: 10.1016/S0099-2399(99)80276-3, indexed cone-beam CT and intraoral radiography for the diagnosis in Pubmed: 10530248. of periapical pathology. Oral Surg Oral Med Oral Pathol 47. Studebaker B, Hollender L, Mancl L, et al. The Incidence Oral Radiol Endod. 2007; 103(1): 114–119, doi: 10.1016/j. of Second Mesiobuccal Canals Located in Maxillary tripleo.2006.01.001, indexed in Pubmed: 17178504. Molars with the Aid of Cone-beam Computed Tomog- 35. Nattress BR, Martin DM. Predictability of radiographic raphy. J Endod. 2018; 44(4): 565–570, doi: 10.1016/j. diagnosis of variations in root canal anatomy in mandibu- joen.2017.08.026, indexed in Pubmed: 29153734. lar incisor and premolar teeth. Int Endod J. 1991; 24(2): 48. Thomas RP, Moule AJ, Bryant R. Root canal morphol- 58–62, indexed in Pubmed: 1917090. ogy of maxillary permanent first molar teeth at various 36. Neaverth EJ, Kotler LM, Kaltenbach RF. Clinical investiga- ages. Int Endod J. 1993; 26(5): 257–267, indexed in tion (in vivo) of endodontically treated maxillary first Pubmed: 8300257. molars. J Endod. 1987; 13(10): 506–512, doi: 10.1016/ 49. Tian XM, Yang XW, Qian L, et al. Analysis of the root and S0099-2399(87)80018-3, indexed in Pubmed: 3482228. canal morphologies in maxillary first and second molars 37. Neelakantan P, Subbarao C, Ahuja R, et al. Cone-beam in a chinese population using cone-beam computed to- computed tomography study of root and canal morphol- mography. J Endod. 2016; 42(5): 696–701, doi: 10.1016/j. ogy of maxillary first and second molars in an Indian popu- joen.2016.01.017, indexed in Pubmed: 26994598. lation. J Endod. 2010; 36(10): 1622–1627, doi: 10.1016/j. 50. Vasundhara V, Lashkari KP. An study to find the incidence joen.2010.07.006, indexed in Pubmed: 20850665. of mesiobuccal 2 canal in permanent maxillary first mo- 38. Parker J, Mol A, Rivera EM, et al. CBCT uses in clinical lars using three different methods. J Conserv Dent. 2017; endodontics: the effect of CBCT on the ability to locate 20(3): 190–193, doi: 10.4103/0972-0707.218308, indexed MB2 canals in maxillary molars. Int Endod J. 2017; in Pubmed: 29279624. 50(12): 1109–1115, doi: 10.1111/iej.12736, indexed in 51. Verma P, Love RM. A Micro CT study of the mesiobuccal Pubmed: 27977863. root canal morphology of the maxillary first molar tooth. 39. Patel S, Dawood A, Whaites E, et al. New dimensions in Int Endod J. 2011; 44(3): 210–217, doi: 10.1111/j.1365- endodontic imaging: part 1. Conventional and alternative 2591.2010.01800.x, indexed in Pubmed: 20880136. radiographic systems. Int Endod J. 2009; 42(6): 447–462, 52. Vertucci FJ. Root canal anatomy of the human perma- doi: 10.1111/j.1365-2591.2008.01530.x, indexed in Pub- nent teeth. Oral Surg Oral Med Oral Pathol. 1984; 58(5): med: 19298577. 589–599, indexed in Pubmed: 6595621. 40. Pattanshetti N, Gaidhane M, Al Kandari AM. Root and canal 53. Weine FS, Healey HJ, Gerstein H, et al. Canal configuration morphology of the mesiobuccal and distal roots of perma- in the mesiobuccal root of the maxillary first molar and nent first molars in a Kuwait population--a clinical study. its endodontic significance. 1969. J Endod. 2012; 38(10): Int Endod J. 2008; 41(9): 755–762, doi: 10.1111/j.1365- 1305–1308, doi: 10.1016/j.joen.2012.08.005, indexed in 2591.2008.01427.x, indexed in Pubmed: 18637850. Pubmed: 22980167. 41. Peeters HH, Suardita K, Setijanto D. Prevalence of a second 54. Weissman J, Johnson JD, Anderson M, et al. Association canal in the mesiobuccal root of permanent maxillary first between the presence of apical periodontitis and clini- molars from an Indonesian population. J Oral Sci. 2011; cal symptoms in endodontic patients using cone-beam 53(4): 489–494, indexed in Pubmed: 22167035. computed tomography and periapical radiographs. 42. Ratanajirasut R, Panichuttra A, Panmekiate S. A cone-beam J Endod. 2015; 41(11): 1824–1829, doi: 10.1016/j. computed tomographic study of root and canal morpholo- joen.2015.06.004, indexed in Pubmed: 26349581. gy of maxillary first and second permanent molars in a thai 55. Weng XL, Yu SB, Zhao SL, et al. Root canal morphology of population. J Endod. 2018; 44(1): 56–61, doi: 10.1016/j. permanent maxillary teeth in the Han nationality in Chi- joen.2017.08.020, indexed in Pubmed: 29061352. nese Guanzhong area: a new modified root canal staining 43. Reis AG, Grazziotin-Soares R, Barletta FB, et al. Second canal technique. J Endod. 2009; 35(5): 651–656, doi: 10.1016/j. in mesiobuccal root of maxillary molars is correlated with joen.2009.02.010, indexed in Pubmed: 19410077. root third and patient age: a cone-beam computed tomo- 56. Zhang R, Yang H, Yu X, et al. Use of CBCT to identify the graphic study. J Endod. 2013; 39(5): 588–592, doi: 10.1016/j. morphology of maxillary permanent molar teeth in a Chi- joen.2013.01.003, indexed in Pubmed: 23611373. nese subpopulation. Int Endod J. 2011; 44(2): 162–169, 44. Schwarze T, Baethge C, Stecher T, et al. Identification of doi: 10.1111/j.1365-2591.2010.01826.x, indexed in second canals in the mesiobuccal root of maxillary first and Pubmed: 21091495. second molars using magnifying loupes or an operating 57. Zhang Y, Xu H, Wang D, et al. Assessment of the second microscope. Aust Endod J. 2002; 28(2): 57–60, indexed in mesiobuccal root canal in maxillary first molars: a cone- Pubmed: 12360670. beam computed tomographic study. J Endod. 2017; 45. Silva EJ, Nejaim Y, Silva AIV, et al. Evaluation of root canal 43(12): 1990–1996, doi: 10.1016/j.joen.2017.06.021, configuration of maxillary molars in a Brazilian population indexed in Pubmed: 29032819. using cone-beam computed tomographic imaging: an in 58. Zheng Qh, Wang Y, Zhou Xd, et al. A cone-beam computed vivo study. J Endod. 2014; 40(2): 173–176, doi: 10.1016/j. tomography study of maxillary first permanent molar root joen.2013.10.002, indexed in Pubmed: 24461399. and canal morphology in a Chinese population. J Endod. 46. Stropko JJ. Canal morphology of maxillary molars: clinical 2010; 36(9): 1480–1484, doi: 10.1016/j.joen.2010.06.018, observations of canal configurations. J Endod. 1999; 25(6): indexed in Pubmed: 20728713.

Incidence, number and topography of Wormian bones in Greek adult dry skulls K. Natsis1, M. Piagkou2, N. Lazaridis1, N. Anastasopoulos1, G. Nousios1, G. Piagkos2, M. Loukas3

1Department of Anatomy, Faculty of Health and Sciences, Medical School, Aristotle University of Thessaloniki, Greece 2Department of Anatomy, Medical School, National and Kapodistrian University of Athens, Greece 3Department of Anatomical Sciences, School of Medicine, St. George’s University, Grenada, West Indies

[Received: 19 January 2018; Accepted: 7 March 2018]

Background: Wormian bones (WBs) are irregularly shaped bones formed from independent ossification centres found along cranial sutures and fontanelles. Their incidence varies among different populations and they constitute an anthropological marker. Precise mechanism of formation is unknown and being under the control of genetic background and environmental factors. The aim of the current study is to investigate the incidence of WBs presence, number and topographical distribution according to gender and side in Greek adult dry skulls. Materials and methods: All sutures and fontanelles of 166 Greek adult dry skulls were examined for the presence, topography and number of WBs. One hundred and nineteen intact and 47 horizontally craniotomised skulls were examined for WBs presence on either side of the cranium, both exocranially and intracranially. Results: One hundred and twenty-four (74.7%) skulls had WBs. No difference was detected between the incidence of WBs, gender and age. Sutures and fontanelles located in neurocranium showed a higher incidence of WBs, contrariwise to orbital sutures that indicated a low incidence. WBs most commonly located in the lambdoid suture (44.6%), followed in order of frequency by the coronal suture (39.8%), asterion (21% on the left and 15.3% on the right side) and parietomastoid suture (15.1% on the left and 13.9% on the right side). Other sutures with WBs were the occipitomastoid, sagittal, squamosal, zygomaticosphenoid, metopic, frontonasal and frontozygomatic. Regarding the skull fontanelles, WBs were found at pterion, posterior and anterior fontanelles. Conclusions: The current study highlights a high incidence of WBs in a Greek population, indicating racial variation. The in depth knowledge of exact location, frequency and number of WBs is essential for clinicians intervening in the skull area, anthropologists and forensic surgeons investigating child abuse cases. (Folia Morphol 2019; 78, 2: 359–370)

Key words: Wormian bones, sutural bones, fontanelle, skull asymmetry, variation, syndrome, deformation

Address for correspondence: Assistant Professor M. Piagkou, Department of Anatomy, School of Medicine, Faculty of Health Sciences, National and Kapodistrian University of Athens, Greece, e-mail: [email protected]

359 Folia Morphol., 2019, Vol. 78, No. 2

INTRODUCTION pathological conditions (trauma or bone disease) and Wormian bones (WBs) or sutural bones or super- craniosynostosis were excluded. Our sample comprised numerary bones or ossicles, are irregularly shaped 119 intact and 47 horizontally craniotomised skulls bones formed from independent ossification centres which were examined for WBs presence on either found along cranial sutures and fontanelles [17]. side of the cranium, both exocranially and intracrani- These ossicles may appear on both the outer and in- ally. From an embryological point of view, the skull is ner tables of the skull, or exclusively on either table. composed of the neurocranium and viscerocranium They joined with the surrounding bones by particu- which derive from the mesoderm and neural crest [2]. larly complex sutures and appear with a variable size, The orbit is formed from both of them. Thus, we ob- number and shape in different locations [3]. served all sutures and fontanelles in viscerocranium, Although two main hypotheses have been proposed orbit and neurocranium and recorded WBs presence, to explain their formation, the precise mechanism still topographic location and number of WBs, by visual remains unknown. The first hypothesis considers that inspection, in order to determine which part of the WBs are under genetic influence [13]. Bennett [4] sup- skull presented the higher incidence of WBs. In our ported that WBs are inherited as dominant traits, Finkel study, only the bones surrounded by apparent sutures [13] suggested that their formation is the result of were recorded as WBs and only the obvious ones were a single gene expression and Mao et al. [21] mentioned counted. Inca bones were ignored, since their devel- that their formation is under the epigenetic control opmental background is different. Correlation of WBs of traits. The second hypothesis considers mechanical presence with gender and age was investigated using 2 stress (artificial cranial deformation or craniosynostosis), the c test. Side asymmetry was further examined with as the main reason for WBs formation [13, 26, 31]. McNemar test. Statistical analysis was carried out using Another hypothesis [11, 26, 31] suggests that genetic IBM SPSS Statistics for Windows version 21.0. factors influence WBs appearance, while the mechanical stress has an impact on their number. Ethical approval Although WBs can be found in healthy individuals, The performed investigation in dry human skulls of they may also be observed in patients with a variety Greek population was in accordance with the ethical of congenital disorders, like osteogenesis imperfecta, standards of the Ethical Committee of our Institutions cretinism, cleidocranial dysostosis and enlarged parietal and with the 1964 Helsinki declaration and its later foramina [23, 24, 28], as well as in patients with central amendments or comparable ethical standards. nervous system abnormalities [29]. Their incidence varies among different populations and therefore these bones RESULTS constitute an anthropological marker. One hundred and twenty-four (74.7%) dry skulls The aim of the current study is to investigate the had WBs. No significant difference was detected be- incidence of WBs, number and topographical distribu- tween the incidence of WBs, gender and age. Sutures tion according to gender and side in a population of and fontanelles located in neurocranium showed Greek adult skulls, both intracranially and exocranially. a higher incidence of WBs (Tables 1–4); while a low incidence of WBs was found in the majority of the MATERIALS AND METHODS orbital sutures (Tables 2, 4). WBs most commonly lo- One hundred and sixty-six (92 males and 74 females) cated in the lambdoid suture (LS) (44.6%) (Figs. 1, 2), Greek dry human skulls of known age from the osteo- followed, in order of frequency, by the coronal suture logical collection of the Department of Anatomy and (CS) (39.8%) (Fig. 3), asterion (21% on the left and Surgical Anatomy (Medical School of Aristotle University of Thessaloniki) and the Department of Anatomy (Medi- cal School of National and Kapodistrian University of Table 1. Wormian bones (WBs) presence and absence in Athens) were examined. The skulls, which belonged to Greek skulls according to gender body donators after informed consent, were separated into three age subgroups: 20–39 years old (n = 41), Gender N (%) WBs presence Absence 40–59 (n = 30) and over 60 years (n = 95). Skulls of Male 92 (55.4%) 67 (72.8%) 25 (27.2%) children, those with deformities (occipital flattening, Female 74 (44.6%) 57 (77%) 17 (23%) syndromic background and asymmetries), obvious Total 166 124 (74.7%) 42 (25.3%)

360 K. Natsis et al., Wormian bones’ variability

Table 2. Exocranial topographical distribution and frequency of Wormian bones (WBs) according to side WBs located at Range of Side Number of Frequency over Frequency over observed WBs skulls the total number the number of (min–max) of skulls skulls with WBs Lambdoid suture (LS) 1–18 74 74/166 (44.6%) 74/124 (59.7%) Parietomastoid suture (PMS) 1–6 23 R (18.5%) 25 L (20.2%) 48 48/166 (28.9%) 48/124 (38.7%) Occipitomastoid suture (OMS) 1–7 11 R (8.9%) 8 L (6.5%) 16 16/166 (9.63%) 16/124 (12.9%) Sagittal suture (SS) 1–5 and multiple 18 18/166 (10.8%) 18/124 (14.5%) Coronal suture (CS) 1–21 and multiple 66 66 (39.8%) 66/124 (53.2%) Squamosal suture (SQS) 1–5 12 R (9.7%) 9 L (7.3%) 21 21/166 (12.7%) 21/124 (16.9%) Metopic suture (MS) 1 1 1/166 (0.6%) 1/124 (0.8%) Frontonasal suture (FNS) 1 1 1/166 (0.6%) 1/124 (0.8%) Asterion (Ast) 1–8 19 R (15.3%) 26 L (21%) 38 38/166 (22.9%) 38/124 (30.6%) Pterion (Pt) 1–2 6 R (4.8%) 4 L (3.2%) 9 9/166 (5.4%) 9/124 (7.25%) Anterior fontanelle (AF) 1–2 2 2/166 (1.2%) 2/124 (1.6%) Posterior fontanelle (PF) 1–2 3 3/166 (1.8%) 3/124 (2.4%) Frontal bone (FB) 1 2 2/166 (1.2%) 2/124 (1.6%) Frontonasal suture (FNS) 1 1 1/166 (0.6%) 1/124 (0.8%) Sphenoid bone (SB) 1 2 2/166 (1.2%) 2/124 (1.6%) Zygomatosphenoid suture* (ZMSS) 1–4 4 R (3.2%) 2 L (1.6%) 4 4/166 (2.4%) 4/124 (3.2%) Frontozygomatic suture (FZMS) 1–4 intraorbital 4 R (3.2%) 1 L (0.8%) 3 3/166 (1.8%) 3/124 (2.4%) 1 extraorbital 1 R – 1 1/166 (0.6%) 1/124 (0.8%) Sphenofrontal suture (SFS) 2 intraorbital – 2 L (1.6%) 2 2/166 (1.2%) 2/124 (1.6%) Frontolacrimal suture* (FLS) 1 – 1 L (0.8%) 1 1/166 (0.6%) 1/124 (0.8%) Zygomatomaxillary suture* (ZMMS) 1–2 1 R (0.8%) 2 L (1.6%) 1 1/166 (0.6%) 1/124 (0.8%) Sphenomaxillary suture* (SMS) 1 1 L (0.8%) 1 1/166 (0.6%) 1/124 (0.8%) Lacrimomaxillary suture* (LMS) 1 1 1/166 (0.6%) 1/124 (0.8%) Sphenozygomatic suture* (SZMS) 1–3 3 R (2.4%) 1 L (0.8%) 4 4/166 (2.4%) 4/124 (3.2%)

*Intraorbital location; L — left side; R — right side

15.3% on the right side) (Fig. 4) and parietomastoid The number of WBs in each suture demonstrated suture (PMS) (15.1% on the left and 13.9% on the a wide range. LS and CS presented the higher number right side) (Fig. 5). WBs were also detected in the oc- of WBs (1–18) (Figs. 3, 14, 15) and (1–21 and multi- cipitomastoid (OCMS) (Fig. 6), sagittal (SS) (Fig. 7), ple), respectively. An interesting observation was that squamosal (SQS) (Fig. 8), zygomaticosphenoid (ZMSS), all WBs in the LS were larger than 1 cm, while in the metopic (MS), frontonasal (FNS) (Fig. 9) and fron- CS were smaller than 0.5 cm in 95.32% of the skulls. tozygomatic suture (FZMS) (Fig. 10). As regards the Taking into account the laterality — side asymmetry of skull fontanelles, WBs appeared at pterion (Fig. 11), WBs presence, our analysis concluded that a statisti- posterior and anterior fontanelle (PF and AF) (Fig. 12; cally significant difference existed only in the asterion Tables 1–4). Another remarkable point was the pres- (p = 0.04) with a right side dominance. ence of a single WB in the frontal bone (FB) (2 skulls) and sphenoid bone (SB) (2 skulls) (Fig. 13). Regarding DISCUSSION the craniotomised skulls (44 out of 47), an incidence Wormian bones are joined with the surrounding of 93.6% had WBs and 21 out of 47, an incidence of bones by particularly complex sutures, especially on 44.7% were detected with WBs intracranially. WBs the outer table of the skull. They are located along were located into the LS, asterion, CS, PMS, SS, SQS, cranial sutures and fontanelles, originating from in- FB, SB (greater wing), OCMS, pterion, SPS, SFS, AF dependent ossification centres [1, 17]. These ossicles and PF (Figs. 13–15). possess a variable size and shape, being ectopic is-

361 Folia Morphol., 2019, Vol. 78, No. 2

Table 3. Intracranial topographical distribution and frequency of Wormian bones (WBs) according to side WBs located at Range of Side N Frequency over the Frequency over the observed WBs total number number of skulls (min–max) of skulls with WBs Lambdoid suture (LS) 1–22 11 11/166 (6.62%) 11/124 (8.9%) Parietomastoid suture (PMS) 1–5 6 R 2 L 1 B 9 9/166 (5.42%) 9/124 (7.25%) Occipitomastoid suture (OMS) 1–2 2 R 1 L 3 3/166 (1.8%) 3/124 (2.4%) Sagittal suture (SS) 2 1 1/166 (0.6%) 1/124 (0.8%) Coronal suture (CS) 1 1 1/166 (0.6%) 1/124 (0.8%) Squamosal suture (SQS) 1–2 5 R – 5 5/166 (3%) 5/124 (4%) Metopic suture (MS) 1 1 1/166 (0.6%) 1/124 (0.8%) Frontonasal suture (FNS) 1 1 1/166 (0.6%) 1/124 (0.8%) Asterion (Ast) 1–6 8 R 11 L 19 19/166 (11.44%) 19/124 (15.3%) Pterion (Pt) 1 2 R – 2 2/166 (1.2%) 2/124 (1.6%) Anterior fontanelle (AF) 1 1 1/166 (0.6%) 1/124 (0.8%) Posterior fontanelle (PF) 1 1 1/166 (0.6%) 1/124 (0.8%) Frontal bone (FB) 2–13 6 6/166 (3.6%) 6/124 (4.83%) Sphenoid bone (SB) 1–8 5 5/166 (3%) 5/124 (4%) Sphenofrontal suture (SFS) (intra) 2 1 R 1 L 2 2/166 (1.2%) 2/124 (1.6%) Sphenoparietal suture (SPS) 2 – 1 L 1 1/166 (0.6%) 1/124 (0.8%)

B — bilaterally; L — left side; N — number of skulls; R — right side

Table 4. Exocranial topographical distribution and frequency of Wormian bones (WBs) according to gender (males and females) and sides (right [R] and left [L]) WBs located at Range of observed WBs Number of skulls and frequency (%) (min–max) Males Females Total Lambdoid suture (LS) 1–18 40 (43.5%) 34 (45.9%) 74 (44.6%) Parietomastoid suture (PMS) 1–6 16 R 16 L 7 R 9 L 23 R 25 L Occipitomastoid suture (OMS) 1–7 10 R 6 L 1 R 2 L 11 R 8 L Sagittal suture (SS) 1–5 and multiple 9 (9.8%) 9 (12.2%) 18 (10.8%) Coronal suture (CS) 1–12 and multiple 39 (42.4%) 27 (36.5%) 66 (39.8%) Squamosal suture (SQS) 1–5 7 R 4 L 5 R 5 L 12 R 9 L Frontonasal suture (FNS) 1 – 1 (1.4%) 1 (0.8%) Metopic suture (MS) (1WB) 1 1(1.1%) – 1 (0.8%) Asterion (Ast) 1–10 15 R 17 L 4 R 9 L 19 R 26 L Pterion (Pt) 1–2 2 R 2 L 4 R 2 L 6 R 4 L Anterior fontanelle (AF) 1–2 2 (2.2%) – 2 (1.6%) Posterior fontanelle (PF) 1–2 1(1.1%) 2 (2.8%) 3 (2.4%) Frontal bone (FB) (extra) 1 1(1.1%) 1 (1.4%) 2 (1.6%) Sphenoid bone (SB) (extra) 1 (exocranial) 2 (2.2%) – 2 (1.6%) Zygomatosphenoid suture (ZMSS) 1–4 2 R 1 L 2 R 1 L 4 R 2 L Frontozygomatic suture (FZMS) intraorbitally 1–4 4 R 1 L –– 4 R 1 L Sphenofrontal suture (SFS) intracranially 1 ––– 2 L 2 L Frontolacrimal suture (FLS) 1 – 1 L –– 1 L Zygomatomaxillary suture (ZMMS) 1–2 – 1 L 1 R 1 L 1 R 2 L Sphenomaxillary suture (SMS) 1 ––– 1 L 1 L Lacrimomaxillary suture (LMS) 1 1 (1.4%) 1 (0.8%) Sphenozygomatic suture (SZMS) 1–3 1 R – 2 R 1 L 3 R 1 L

362 K. Natsis et al., Wormian bones’ variability

Figure 1. Wormian bones (WBs) in lambdoid suture; A. On the left side and asterion; B. On the right side, a WB of triangular shape; C. Multiple right-sided WBs; D. Various shapes of left-sided WBs.

Figure 2. Multiple Wormian bones (WBs) in lambdoid suture; A. With asterisks are depicted 10 WBs symmetrically located and an inca bone (IB); B. Black (*) left-sided WBs and white (*) posterior fontanellar bones; C. WB in the occipitomastoid suture.

lands of intramembranous ossification. In foetuses, craniosynostosis affects the frequency and location they are composed of a single layer of compact bone of certain types of WBs [1, 26, 31]. When WBs ap- on the dural side, since diploe is not formed yet [19]. pear as a normal variant, they have usually smaller Mechanical factors increasing sutural width may trig- dimensions and number [19]. ger WBs formation [6]. The study of El-Najjar and In our study, conducted on Greek skulls, the inci- Dawson [11] confirms the hypothesis of genetic in- dence of WBs presence (74.7%) is remarkably high. fluence in WBs development, since they appeared in A similar high incidence (80.32%) was reported in foetal skulls, with reduced environmental stressing Chinese [5], whereas in West Anatolian and Eastern factors. Cranial deformation either cultural or after Indian skulls, the reported incidences were 59.3%

363 Folia Morphol., 2019, Vol. 78, No. 2

Figure 3. A–I. Wormian bones located into the coronal suture.

Figure 4. Wormian bones (WBs) in the area of asterion (black arrows) (A–D); A. Enlarged WB (asterisk).

and 45%, respectively [8, 14]. The lowest incidence most common location was the CS followed by aste- (9%) was reported by Govsa et al. [15] in Turkish skulls. rion, PMS, SQS and pterion. Several authors [23, 24, Cirpan et al. [8] along with current study highlighted 27] found the asterion, as the second most common a stronger number of multiple WBs found in skulls location, while Nayak [25] reported the epipteric bone, (44.6% and 39.3%, respectively). A wide variability in as the second most common location. Cirpan et al. [8] the occurrence of WBs among different ethnic groups mentioned that WBs appear in order of frequency in LS is a reality [3]. On the other hand, WBs most commonly (40.7% on the left and 37.3% on the right side), pterion located is the LS, among different populations. In the (8% on the left side), asterion, SQS, lambda, SS, CS, present study, WBs most commonly appear in the LS OCMS and bregma. In another study [31], WBs were (44.6%), similarly to Edwards and co-authors’ study identified in LS (50%), CS (25%), within fontanelles in [10]. Murlimanju et al. [23, 24] reported that WBs are asterion, PF and AF and into the orbit. In skulls with found along the LS in 56.4%. In our study, the second sagittal synostosis, the incidence of WBs in the AF was

364 K. Natsis et al., Wormian bones’ variability

Figure 5. A, B. Wormian bones into the parietomastoid and lambdoid sutures (black arrows).

Figure 6. A–D. Wormian bones in the occipitomastoid suture (black arrows).

reported in 4% [1]. Sanchez-Lara et al. [31] and Jeanty outer one. Specifically, even though intracranial and et al. [18] showed that WBs were more frequently exocranial WBs appeared at the same sutures, their observed on the right side of the skulls, contrariwise exact location and number mismatch. Also in the cur- to Cirpan et al. [8] who reported that WBs appear rent study, no gender dimorphism was observed with more often on the left side. Both genders had a similar respect to overall incidence, but regarding location, incidence of WBs [16]. a male predominance was found in LS and OCMS. This In the present study, it is worth mentioning that result differs from Patil and Sheelavant [27] findings, the shape, number and even location of WBs differ which reported a female predominance in LS and SS between outer and inner tables of the skull. In all in the area of lambda. Deformation induced by certain examined cases, WBs were smaller and sutures were customs directly affecting the cranial shape, are termed less complicated intracranially. It is suggested that WBs as ‘‘cultural cranial deformation’’ and ultimately influ- are isolated bones formed by an inner and outer plate. ence the number of WBs [11]. A probable explanation An interesting point of the current study was that the of deformation is the continuously exerted pressure on inner plate of the bones did not correspond to the the front and back of the infant’s cranium, resulting

365 Folia Morphol., 2019, Vol. 78, No. 2

Figure 7. A–C. Wormian bones into the sagittal suture (thick and thin arrows).

Figure 8. A–C. Wormian bones in the squamosal suture (black arrows).

in a pronounced frontal and occipital flattening and Wormian bones may be clinically important signs an increase in cranial height [33]. El-Najjar and Daw- for silent pathologies. They are most commonly small- son [11] concluded that the occurrence of lambdoid er and less numerous in normal than pathologic vari- WBs was genetically controlled. Bennett [4] correlated ant [19]. Bennet [4] explained that WBs may develop WBs in posterior vault with basi-occipital length and due to rapid cerebral expansion, and this is why concluded that stresses connected with higher meas- they are found in higher numbers in hydrocephalic urements caused WBs formation. Barberini et al. [2] patients. Nowadays, plagiocephaly is correlated with supported that WBs formation may be controlled by asymmetry due to pressure and a higher number of a number of genes with additive action. WBs. It is demonstrated that the deformed crania have

366 K. Natsis et al., Wormian bones’ variability

Figure 9. A–C. Wormian bones in the frontonasal suture (black arrows).

Figure 10. A, B. Wormian bones in sphenofrontal suture (top of the pterygopalatine fossa) and into the orbit (black arrows).

a higher number of WBs than the undeformed, and early identification and treatment in affected paediatric most WBs have been located in an antero-posteriorly population [29]. Moreover, their presence is particularly oriented deformation [31]. WBs in more posteriorly higher in non-symmetrical skulls with metopism [8]. placed sutures are more prone to environmental factors Cirpan et al. [9] reported a strong correlation between in comparison to their more anteriorly placed counter- WBs coexistence and metopism, explaining that the parts which are under stronger genetic control [31]. WBs factors leading to metopism may also form WBs. presence is associated with abnormal central nervous Wormian bones are clinically important markers system development (microcephaly, macrocephaly, hy- for many pathological entities. Though common in drocephalus, craniosynostosis, cerebral palsy, epilepsy adults, their occurrence in children can be associ- and learning difficulties) and this may be a sign of ated with a plethora of pathological entities such as

367 Folia Morphol., 2019, Vol. 78, No. 2

Figure 11. Wormian bones in the area of pterion (right-sided: A, B, D and left-sided: C, E) (black arrows).

Figure 12. Wormian bones in bregma (B, D, E) and posterior fontanelle (A, C, F, G) (asterisks).

Rickets syndrome, “Kinky-hair” Menke’s syndrome, hypoparathyroidism, hypophosphatasia and osteo- Otopalatodigital syndrome, Hajdu-Cheney syndrome genesis imperfecta [20, 22, 30, 32]. Mutations of (platybasia and WBs, osteoporosis with fractures, and stable proteins may lead to a skeletal homeostasis acro-osteolysis), acrocallosal syndrome (facial dys- disorder and WBs appearance [7]. morphism and midline abdominal defects), Down’s Wormian bones are of diagnostic value, since syndrome, cleidocranial dysostosis, pycnodysostosis, their presence may be occasionally misinterpreted

368 K. Natsis et al., Wormian bones’ variability

Figure 13. Wormian bones intracranially at the frontal (A) and sphenoid bones (B).

Figure 14. Wormian bones intracranially; A. Lambdoid, occipitomastoid and parietomastoid sutures; B. In between greater wing of sphenoid bone, parietal and squamous part of temporal bone; C. Lambdoid suture.

of view in forensic investigation of non-accidental skull injuries in order to rule out physical abuse and brittle bones [15]. In some cases, the traumatic lesion of a gunshot entry wound of the skull, may have similarities with the orifice of a WB, thus emphasis must be given on differential diagnosis during routine anthropological examination. Neurosurgeons should be cautious when performing burr holes over the pterion, since presence of WBs may lead to complications [12]. Figure 15. Multiple Wormian bones intracranially at the lambdoid Limitations of the study were: (i) The sample did suture. not include infant skulls and skulls with syndromic background and asymmetries, or pathological skulls so we could not investigate the effect of these vari- as fractures during radiological examination [11]. ants in WBs appearance; (ii) Due to the small number WBs morphological and topographical details are of of skulls with inca bones we could not detect any paramount importance from the medico-legal point possible correlation between WBs and inca bones;

369 Folia Morphol., 2019, Vol. 78, No. 2

(iii) WBs shape was not investigated; (iv) A small 14. Ghosh SK, Biswas S, Sharma S, et al. An anatomical study of wormian bones from the eastern part of India: is genetic influ- number of skulls was craniotomised, as all bones were ence a primary determinant of their morphogenesis? Anat Sci part of the osteological collection of our Institutions. Int. 2017; 92(3): 373–382, doi: 10.1007/s12565-016-0342-1, indexed in Pubmed: 27038026. 15. Govsa F, Ozer MA, Bayraktaroglu S, et al. Anatomoradiological CONCLUSIONS identification of intrasutural bones for importance of cranial frac- The current study highlights a high incidence of ture. Turk Neurosurg. 2014; 24(3): 357–362, doi: 10.5137/1019- 5149.JTN.8380-13.2, indexed in Pubmed: 24848174. WBs in a Greek population, indicating ethnical varia- 16. Hanihara T, Ishida H. Frequency variations of discrete cranial tion. The knowledge of frequency, number and exact traits in major human populations. I. Supernumerary ossicle variations. J Anat. 2001; 198(Pt 6): 689–706, indexed in location of WBs is essential for anthropologists, any Pubmed: 11465862. physician involved in diagnosis and treatment of skull 17. Hauser G, De St. Epigenetic variants of the human skull. Stutt- gart: Schweizerbart. Actas del XVII Congreso Internacional de area pathologies and forensic specialists investigating Americanistas, Buenos Aires. 1989. child abuse suspicious cases. 18. Jeanty P, Silva SR, Turner C. Prenatal diagnosis of wormian bones. J Ultrasound Med. 2000; 19(12): 863–869, indexed in Pubmed: 11127012. Acknowledgements 19. Kaplan SB, Kemp SS, Oh KS. Radiographic manifestations of congenital anomalies of the skull. Radiol Clin North Am. 1991; This study has not received any funding support. 29(2): 195–218, indexed in Pubmed: 1998047. Miss Elpida Repousi contributed to this work. 20. Kumar A, Rajesh B, Arumugam K, et al. Sutural bones associated with lambdoid suture of human skull: presence, variations and References clinical importance. Int J Anat Res. 2016; 4(2.2): 2331–2336, doi: 10.16965/ijar.2016.208. 1. Agrawal D, Steinbok P, Cochrane DD. Pseudoclosure of 21. Mao JJ, Wang X, Mooney MP, et al. Strain induced osteo- anterior fontanelle by wormian bone in isolated sagittal genesis of the craniofacial suture upon controlled delivery of craniosynostosis. Pediatr Neurosurg. 2006; 42(3): 135–137, low-frequency cyclic forces. Front Biosci. 2003; 8: a10–a17, doi: 10.1159/000091854, indexed in Pubmed: 16636612. indexed in Pubmed: 12456317. 2. Barberini F, Bruner E, Cartolari R, et al. An unusually-wide human 22. Marti B, Sirinelli D, Maurin L, et al. Wormian bones in bregmatic Wormian bone: anatomy, tomographic description, a general paediatric population. Diagn Interv Imaging. 2013; and possible significance. Surg Radiol Anat. 2008; 30(8): 683–687, 94(4): 428–432, doi: 10.1016/j.diii.2013.01.001, indexed in doi: 10.1007/s00276-008-0371-0, indexed in Pubmed: 18523715. Pubmed: 23352712. Muralimanju BV, Prabhu LV, Ashraf CM, 3. Bellary SS, Steinberg A, Mirzayan N, et al. Wormian bones: et al. Morphological and topographical study of Wormian a review. Clin Anat. 2013; 26(8): 922–927, doi: 10.1002/ca.22262, bones in cadaver dry skulls. J Morphol Sci. 2011; 28: 176–179. indexed in Pubmed: 23959948. 23. Murlimanju BV, Gupta C, Samiullah D, et al. Morphological 4. Bennett KA. The etiology and genetics of Wormian bones. investigation of cranial sutures in Indian human adult skulls. Am J Phys Anthropol. 1965; 23(3): 255–260, indexed in Pub- Rom J Morphol Embryol. 2011; 52(3 Suppl): 1097–1100, med: 5861223. indexed in Pubmed: 22119831. 5. Brothwell DR. The use of non-metrical characters of the 24. Nayak S. Presence of Wormian bone at bregma and paired skull in differentiating populations. Dt Ges Anthrop. 1958; frontal bone in an Indian skull. Neuroanatomy. 2006; 5: 42–43. 6: 103–109. 25. O’Loughlin VD. Effects of different kinds of cranial deforma- 6. Burrows AM, Caruso KA, Mooney MP, et al. Sutural bone tion on the incidence of wormian bones. Am J Phys Anthropol. frequency in synostotic rabbit crania. Am J Phys Anthro- 2004; 123(2): 146–155, doi: 10.1002/ajpa.10304, indexed in pol. 1997; 102(4): 555–563, doi: 10.1002/(SICI)1096- Pubmed: 14730648. 8644(199704)102:4<555::AID-AJPA9>3.0.CO;2-Q, indexed 26. Patil M, Sheelavant S. Sexual Dimorphism among the Wormian in Pubmed: 9140544. Bones In Adult Human Skulls. J Indian Acad Forensic Med. 7. Canalis E, Zanotti S. Hajdu-Cheney syndrome, a disease as- 2012(2): 124–127. sociated with NOTCH2 mutations. Curr Osteoporos Rep. 2016; 27. Piagkou M, Skotsimara G, Repousi E, et al. Enlarged parietal 14(4): 126–131, doi: 10.1007/s11914-016-0311-6, indexed in foramina: a rare finding in a female Greek skull with unusual Pubmed: 27241678. multiple Wormian bones and a rich parietal vascular network. 8. Cirpan S, Aksu F, Mas N. The incidence and topographic distri- Anat Sci Int. 2013; 88(3): 175–180, doi: 10.1007/s12565-013- bution of sutures including wormian bones in human skulls. 0173-2, indexed in Pubmed: 23543411. J Craniofac Surg. 2015; 26(5): 1687–1690, doi: 10.1097/ 28. Pryles CV, Khan AJ. Wormian bones. A marker of CNS abnor- SCS.0000000000001933, indexed in Pubmed: 26114515. mality? Am J Dis Child. 1979; 133(4): 380–382, indexed in 9. Cirpan S, Aksu F, Mas N, et al. Coexistence of Wormian bones Pubmed: 433853. with metopism, and vice versa, in adult skulls. J Craniofac Surg. 29. Reid TH, Tam A, Antoniou G, et al. Anterior fontanelle Wormian 2016; 27(2): 493–495, doi: 10.1097/SCS.0000000000002370, bone with exomphalos major and dysmorphic facial features: indexed in Pubmed: 26845093. a previously unseen association? J Craniofac Surg. 2016; 27(7): 10. Edwards B, Wang JMh, Iwanaga J, et al. Hiding within the 1799–1801, doi: 10.1097/SCS.0000000000002962, indexed cracks: case report of rare sutural bone found at the nasion. in Pubmed: 27513767. Cureus. 2017; 9(6): e1333, doi: 10.7759/cureus.1333, indexed 30. Sanchez-Lara PA, Graham JM, Hing AV, et al. The morpho- in Pubmed: 28698833. genesis of wormian bones: a study of craniosynostosis and 11. El-Najjar M, Dawson GL. The effect of artificial cranial deformation purposeful cranial deformation. Am J Med Genet A. 2007; on the incidence of Wormian bones in the lambdoidal suture. 143A(24): 3243–3251, doi: 10.1002/ajmg.a.32073, indexed Am J Phys Anthropol. 1977; 46(1): 155–160, doi: 10.1002/ in Pubmed: 18000970. ajpa.1330460119, indexed in Pubmed: 835687. 31. Shah AB, Tisano BK, Elattar O, et al. Foot deformities in hajdu- 12. Ersoy M, Evliyaoglu C, Bozkurt MC, et al. Epipteric bones in the cheney syndrome: a rare case report and review of the litera- pterion may be a surgical pitfall. Minim Invasive Neurosurg. ture. J Orthop Case Rep. 2017; 7(5): 11–15, doi: 10.13107/ 2003; 46(6): 363–365, doi: 10.1055/s-2003-812434, indexed in jocr.2250-0685.876, indexed in Pubmed: 29242787. Pubmed: 14968406. 32. Tubbs RS, Salter EG, Oakes WJ. Artificial deformation of the 13. Finkel DJ. Wormian bones — a study of environmental stress. Am human skull: a review. Clin Anat. 2006; 19(4): 372–377, J Physical Anthropol. 1971; 35: 278. doi: 10.1002/ca.20177, indexed in Pubmed: 16092127.

Greater palatine foramen: assessment with palatal index, shape, number and gender A. Ortug1, M. Uzel2

1Department of Anatomy, School of Medicine, Istanbul Medipol University, Istanbul, Turkey 2Departmeny of Anatomy, Cerrahpasa Faculty of Medicine, Istanbul University-Cerrahpaşa, Istanbul, Turkey

[Received: 21 May 2018; Accepted: 27 July 2018]

Background: Accurate knowledge of location and relation with different parameters of the greater palatine foramen (GPF) is a crucial necessity in performing a variety of anaesthesiological, dental and surgical procedures. The main aim of this study was to identify the GPF’s locations, numbers and shapes via associating with gender and palatal indices and compare with literature results. Materials and methods: This study was held on the cranium collections of the many anatomy departments in Turkey. Various metric assessments were observed on sexed, dry, total of 97 craniums. Results: Thirty-eight male cranium observed and mean values of palatal indices was 86.28 ± 10.75 and for 48 female craniums mean value for palatal indices was 81.06 ± 10.56. Location of GPF observed bilaterally and mostly found near the third upper molar in either of both males and females. It was 62.7% (right), 60.9% (left) for male craniums and 49.0% (right) and 47.1% (left) for female craniums observed near the lateral border of upper third molar. GPF found oval shape for male craniums 62.8% (right) and 61.0% (left) and for female craniums 66.0% (right) and 66.0% (left). Conclusions: These results were compared with already existing anatomical data in other races and populations. These results would have great clinical influence in localising the palatine foramina toward better palatal area surgical approaches to and anaesthetise this area. (Folia Morphol 2019; 78, 2: 371–377)

Key words: greater palatine foramen, lesser palatine foramen, hard palate, palatal index, palatal anaesthesia

INTRODUCTION termine exact location of greater and lesser palatine Hard and soft palate innervation is supplied by great- foramens so giving insufficient anaesthesia solution er and lesser palatine nerves. Those nerves approach [8]. Previous research indicates that success for palatal related area inside the canals known by same as their anaesthesia only possible with the correct identifica- name and reaches to roof of oral cavity via greater pala- tion of greater palatine nerves location [8]. tine foramen (GPF) and lesser palatine foramen (LPF). In addition to the distress associated with the Anaesthetic block of greater palatine nerve was emerging location of greater palatine nerve in palate, firstly described in 1927 [24] and recommended for possible problems can be experienced where greater surgical practices involving upper molars, maxillary palatine arteries emergence is important such as free sinus and nasal region. However, common problem vascular flap, cleft palate or maxillary sinus surgery reported for this application is being not easy to de- [8, 9, 18]. On the other hand; the detailed clinical anat-

Address for correspondence: Dr. A. Ortug, Istanbul Medipol University, School of Medicine, Department of Anatomy, Istanbul, Turkey, tel: +90 532 468 46 80, e-mail: [email protected]

371 Folia Morphol., 2019, Vol. 78, No. 2

omy knowledge of greater palatine nerve could be beneficial for research related with craniofacial growth [30]. Yet, the developmental distances of maxillary molars and premolars are obtained by the development process of the transverse palatine suture [14]. Tomaszewska et al. [34] state that, general description resides in classical anatomy lecture books about the location of greater palatine nerve; however, there is not always a consensus on this information. For example, previous researches reported different information as lateral border of posterolateral margin [11], medial to last molar teeth [23] or in front of last molar teeth [27]. In addition, anaesthesia books state the location little more detailed but variable as Figure 1. Image indicates the measurement parameters: M1, M2, M3 — maxillary molar I, II, III; A — measurement of greater pala- in front of second maxillary molar teeth [31], third tine foramen (GPF) to closest maxillary molar teeth; B — measure- maxillary molar teeth or anywhere close to second ment GPF to sagittal plane; C — measurement of GPF to incisive and third molars [32]. foramen (IF); D — measurement of GPF to posterior nasal spine (PNS). It is obvious that many researchers evaluated the location of GPF for surgical approaches planned to be performed at palatal area. Besides, it is suggested blinded and in case of different value, average re- that location and number of this foramen could corded. The parameters below were chosen in light differ because of the difference among the ethnic of previous research as clear and exact locations, and groups. Studies revealed variable results in Mongo- direct measurements were done (Fig. 1). lian [2, 3], Caucasians [1] and African population Parameters were: [15]. In our population, except a few studies about — calculation of palatal index (PI) and classification the location of GPF, information and comparison of of palatal shape; the location, number, ethnicity and gender is miss- — distance of GPF to closest maxillary molar (MM); ing [7, 10, 28]. — distance of GPF to sagittal plane (SP); In this respect, the aim of this study is to deter- — distance of GPF to posterior nasal spine (PNS); mine the location of GPF evaluate its shape, number — distance of GPF to incisive foramen (IF); and distribution across genders and observe relation — shape and number of GPF bilaterally; with palatal index. Then, the data obtained will be — number of LPF bilaterally. compared with different populational results. Palatal index is the ratio of palatal breadth to It is believed obtained results would have great palatal length (Fig. 2). Types of palate are classified contribution for clinical success in maxillofacial and according to the formula given below: oral surgeries regional anaesthesia and establish a basic index for antropomorphological studies. Palatal index (PI) x = (palatal breadth) / (palatal length) × 100 Palates were classified as leptostaphyline for MATERIALS AND METHODS (x < 80), mesostaphyline for (80 ≤ x < 85), and brachy- Skull collections staphyline for (85 ≤ x) the obtained rational value. This study was conducted on sexed, intact, without developmental malformations, 97 dry craniums Statistical analysis belong to Turkish population. Cranium collections Fitness of our right and left GPF measurements to of different medical and dentistry faculties in Turkey normal distribution was analysed by the Shapiro-Wilk were used upon their permission. test. As the measurements were distributed normally, they were expressed as values of average ± standard Measurements deviation (AVG ± SD) and minimum–maximum (min– Digital calliper (Altas 905, 150 mm), were used max). Categorical variables such as sex and type of for measurements. Palates were photographed in palate were shown numerically (percentage). multiple views for multiple times (LUMIX Panasonic An unpaired t test was used analyse the variance DMC-T25). Each parameter was measured double of right and left GPF measurements by sex, and a one- 372 A. Ortug, M. Uzel, Evaluation of greater palatine foramen

The individuals in the study were grouped according to PI as leptostaphyline, mesostaphyline or brachystaphyline. As for the distribution of these types of palate by sex, the brachystaphyline type was observed in 50.0% (n = 19) of males, and the leptostaphyline type palate was observed in 43.8% (n = 21) of females. No difference was observed in the distribution of palate types by sex (c2 = 3.964, p = 0.138). The distribution of the closest maxillary molar at right side by sex and type of palate was determined accordingly that the third molar was the closest MM Figure 2. Calculation of palatal index; A — length of palate; B — at right in 49.0% (n = 26) and 62.7% (n = 27) of breadth of palate. females and males, respectively. It was found that the rate of those for whom third molar was the closest MM at right was the highest among all palate types -way ANOVA test was used to analyse their variance (leptostaphyline: 53.1%, mesostaphyline: 65.0%, by palate type. As regards types of palate, variance brachystaphyline: 48.4%). of the variables were homogenous and therefore As for the distribution of the closest MM at left side ANOVA F-test statistics were conducted, and paired by sex and type of palate, the number of those whose comparisons of the variables involving statistically sig- third molar was the closest MM at left was calculated 2 nificant variance were made with the Tukey test. A c to be 47.1% (n = 25) and 60.9% (n = 25) for females analysis was applied to examine the distribution of and males, respectively. Among those with the mes- types of palate by sex, and the distribution of right and ostaphyline type of palate, no one had the second molar left GPF shapes by types of palate, and the Pearson’s as the closest MM at left. 2 c value was given. The distribution of the number of At right, the distance of GPF to the posterior nasal right and left lesser palatine foramina by sex and type spine was measured to be 15.84 ± 2.05 mm (n = 38) of palate, and the distribution of the closest maxillary in males and 14.91 ± 1.93 (n = 47) mm in females, 2 molar at right and left were expressed. No c analysis averaging to 15.33 ± 2.02 mm (n = 85). And at the was made due to the insufficient number in cells. The left side, the values obtained were 16.18 ± 1.80 mm values of all the individuals in the study whose GPF (n = 38) for males and 15.16 ± 1.82 mm (n = 48) and LPF numbers were measurable were found to be for females, averaging to 15.66 ± 1.87 mm (n = 86). 1. These values were not included in statistical analyses Distance to incisive foramen from GPF, at the as there was no variance in GPF and LPF numbers. The right side found 38.27 ± 3.61 (n = 43) in males and statistical significance level was taken as p < 0.05. 35.99 ± 3.11 (n = 53) in females, with an average total IBM SPSS Statistics 21.0 (IBM Corp. Released 2012. of 37.01 ± 3.51 (n = 96). And the left side values were IBM SPSS Statistics for Windows, Version 21.0. Ar- 38.54 ± 4.17 (n = 42) for males and 36.81 ± 3.06 monk, NY: IBM Corp.) software was utilised for the (n = 54), with the average 37.57 ± 3.67 (n = 96). statistical analyses and calculations in this study. As for distance to sagittal plane, at the right side found 14.99 ± 2.45 (n = 39) in males and 14.35 ± RESULTS ± 1.95 (n = 47) in females, with an average total of The palatal index for the 43 male and 54 female 14.64 ± 2.20 (n = 86). And the left side values were adults who were included in the study, and the dis- 15.07 ± 2.39 (n = 38) for males and 14.47 ± 2.06 tributions of the values observed in the right and left (n = 47), with the average 14.74 ± 2.22 (n = 86). GPF measurements are given in Table 1. Accordingly, Except for one case in our study, the number of GPF the average PI of the 38 measurable adult males was was found to be one each at two sides. In the excep- 86.28 ± 10.75, whereas the PI of the 48 measurable tional case, no GPF was found at the right side (Fig. 3). adult females was calculated to be 81.06 ± 10.56. Our data demonstrated 64.94% oval/round It was found that the PI values of adult male were shape and an antero-posterior extension. A compar- statistically significantly higher than those of adult ison according to sex and types of palate showed female (t = 2.260, p = 0.026). that the shape of GPF was oval in 62.8% (n = 27) of 373 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. Palatal index, right-left greater palatine foramen (GPF) distribution according to genders and overall samples Male Female All Test statistics P Average ± SD Average ± SD Average ± SD (min–max) (min–max) (min–max) PI N = 38 N = 48 N = 86 86.28 ± 10.75 81.06 ± 10.56 83.37 ± 10.90 2.260 0.026 (66.13–114.00) (58.67–112.10) Right GPF-MM N = 43 N = 53 N = 96 4.77 ± 1.75 4.18 ± 1.15 4.45 ± 1.47 1.892 0.063 (1.32–9.00) (1.92–7.49) Left GPF-MM N = 42 N = 53 N = 95 5.40 ± 1.86 4.86 ± 1.38 5.10 ± 1.62 1.624 0.108 (1.46–9.48) (1.73–7.85) Right GPF-SP N = 39 N = 47 N = 86 14.99 ± 2.45 14.35 ± 1.95 14.64 ± 2.20 1.354 0.179 (11.32–23.11) (10.96–18.57) Left GPF-SP N = 39 N = 48 N = 87 15.07 ± 2.39 14.47 ± 2.06 14.74 ± 2.22 1.255 0.213 (10.28–23.63) (10.56–18.46) Right GPF-PNS N = 38 N = 47 N = 85 15.84 ± 2.05 14.91 ± 1.93 15.33 ± 2.02 2.143 0.035 (11.68–21.42) (10.67–19.25) Left GPF-PNS N = 38 N = 48 N = 86 16.18 ± 1.80 15.16 ± 1.82 15.66 ± 1.87 2.594 0.011 (12.95–22.85) (11.52–19.74) Right GPF-IF N = 43 N = 53 N = 96 38.27 ± 3.61 35.99 ± 3.11 37.01 ± 3.51 3.318 0.001 (30.76–45.47) (30.06–41.55) Left GPF-IF N = 42 N = 54 N = 96 38.54 ± 4.17 36.81 ± 3.06 37.57 ± 3.67 2.259 0.027 (30.86–48.92) (29.63–43.63)

IF — incisive foramen; LPF — lesser palatine foramen; MM — maxillary molar; PI — palatal index; PNS — posterior nasal spine; SD — standard deviation; SP — sagittal plane

males, and round and 34.0% (n = 18) of females at right. No statistically significant difference was found in the distribution of the right GPF shape by sex. As for the left side, GPF’s shape was oval in 61.0% (n = 25) of males and 66.0% (n = 35) of females. No statistically significant difference was determined in the distribution of the left GPF shape by sex. Lesser palatine foramen associated with sex and type of palate, and one left-side LPF was found in 40.5% (n = 17) of males and 63.0% (n = 34) of females. In adults with the leptostaphyline type of palate, two left-side LPF was observed at the most.

Figure 3. One exceptional sample includes just one greater palatine In a great majority of the cases, one LPF was found foramen on left side and missing lesser palatine foramen on both sides. regardless of sex and type of palate. 374 A. Ortug, M. Uzel, Evaluation of greater palatine foramen

DISCUSSION that the difference of measurement method might Classical anatomy and surgery books provide some account for the index difference with other research- general information about the localisation of GPF. ers. Hassanali and Mwaniki [15] and Gözil et al. [13] Some problems with surgical interventions at this measured palatal length from the oral point, which region were reported due to the lack of detailed corresponds to the centre of the line that connects the information about some parameters [20, 25, 33]. posterior margins of the alveoli of maxillary central In recent years, studies were carried out attract- incisors, from the staphylon point, which corresponds ing attention to the fact that GPF’s localisation is to the centre of the line that connects the foremost a clinically important anatomic point [25, 33]. The points of the posterior margins of both sides of the possibility of stimulating the pterygopalatine gan- hard palate. In the current study, which used Dave’s glion through GPF also resulted in getting more at- method, the length from the oral point to the pos- tention [16, 26, 34]. It was reported upon these terior nasal spine was measured. developments that GPF could be utilised for alleviat- The literature on the subject shows no consensus on ing the effects of paralysis in paralytic patients, and whether the location of GPF is influenced by ethnic dif- also for patient interventions in the cases of cerebral ferences. Wang et al. [35] supported the idea of ethnic vasospasms or cluster and migraine headaches [26]. effects, while Jaffar and Hamadah rejected this theory. In the current study, the subjects were also classi- Although homogeneous results were obtained in the fied according to their palate types. The participants studies carried out in Europe, significant variations were were grouped as leptostaphyline (narrow), mes- found by Indian researchers in their studies that were ostaphyline (medium) and brachystaphyline (wide) made on the same population group [5, 6, 19, 29]. according to their PI. In previous palatal morphometry This suggests that broad anatomic variations can occur studies carried out in Turkey, no distinction was made even in the same population group, and a comparison according to sex. In Turkey, Gözil et al. [13] identified of various studies is presented in Table 3. We had no 58.1% leptostaphyline, 17.4% mesostaphyline and opportunity for making a comparison for Turkey due 24.4% brachystaphyline with an average value of to the lack of elaborate studies on the subject matter. 77.94 ± 9.54. In Kenya, Hassanali and Mwaniki [15] In the current study, a bilateral approach was taken to found 43.2% leptostaphyline, 23.7% mesostaphyline, the question of whether there is a little relationship and 33.1% brachystaphyline with an average value of between sex and type of palate and localisation. No 82.0 ± 7.84. In India, Dave et al. [6] found the lep- sex-based difference was determined according to tostaphyline type of palate in 61.5% of females and the results. Tomaszewska et al. [34], drew attention 63.3% of males, the mesostaphyline type in 30.8% to the fact that there were significant differences of females and 20.0% of males, and the brachy- between females and males and the measurements staphyline type in 7.7% of females and 16.7% of relating to GPF, and stated that this could be utilised males (Table 2). Dave’s findings demonstrated 63% even in forensic medicine applications. Our findings, leptostaphyline in the evaluation of all crania. In our however, do not support this theory as we determined study, the general PI was 83.37 ± 10.90 (n = 86). it to be generally aligned with the third molar across The average PI of the 38 measurable adult males was the population regardless of sex or type of palate. 86.28 ± 10.75, whereas the PI of the 48 measurable A comparison of the studies previously performed on adult females was calculated to be 81.06 ± 10.56. the localisation of GPF is given in Table 3. As for the distribution of these palate types by sex, We also evaluated metric values for positioning the brachystaphyline type was observed in 50.0% GPF as distance to sagittal plane, posterior nasal (n = 19) of males, and the leptostaphyline type palate spine and incisive foramen. Gibelli et al. [12] showed was observed in 43.8% (n = 21) of females. In their all these values were significantly influenced by sex respective studies, Hassanali and Mwaniki [15], Gözil at their study conducted on Italian crania. Our re- et al. [13] and Dave et al. [6] found a great majority sults for distance to sagittal plane did not show any of their cases to belong to the leptostaphyline pal- statistically significant difference according to sex, ate group. In our study, however, this only applied side or palate type. However, our other two values to female crania. A majority of male crania were in were statistically significant according to sex and the brachystaphyline palate group. As a result, the side. All measurements Gibelli et al. [12]; indicated palatal indexes of males were found to be significantly were slightly higher than our results which could be higher than those of females. We are of the opinion a possible result of palate type. However, they did not 375 Folia Morphol., 2019, Vol. 78, No. 2

Table 2. Comparison of studies including palatal index Studies Leptostaphyline Mesostaphyline Brachystaphyline Overall Female Male Overall Female Male Overall Female Male Our study (2017) 37.5% 43.8% 28.9% 24.4% 27.0% 21.1% 38.3% 29.2% 50.0% Dave et al. (2013) [6] 63% 61.5% 63.3% 24% 30.%8 20.0% 13% 7.7% 16.7% Gözil et al. (1999) [13] 58.1% –– 17.4% –– 24.4% –– Hassanali and Mwaniki (1984) 43.2% –– 23.7% –– 33.1% –– [15]

Table 3. Comparison of previous studies Studies GPF-SP [mm] GPF-posterior Relation with maxillary molar (%) Right/Left margin [mm] MM2 Between MM2 MM3 Distal to MM3 and MM3 Westmoreland and Blanton 1982 [36] 14.8/15.0 1.9 9.70 33.60 50.70 6.00 Langenegger et al. 1983 [21] – – 1.00 3.00 62.00 34.00 Hassanali and Mwaniki 1984 [15] – – 10.40 13.60 76.00 0.00 Ajmani 1994 (Nigerian Crania) [1] 15.4* 3.5 13.07 38.46 48.46 0.00 Ajmani 1994 (Indian Crania) [1] 14.7/14.6 3.7 0.00 32.35 64.69 2.94 Jaffar and Hamadah 2003 [17] 15.7* 4.86 12.00 19.00 55.00 14.00 Methathrathip et al. 2005 [22] 16.2* 2.1 7.00 14.10 71.90 7.00 Saralaya and Nayak 2007 [29] 14.7/14.7 4.2 0.40 24.20 74.60 0.80 Chrcanovic and Custódio 2010 14.68/14.44 3.39 0.00 6.19 54.87 38.94 (Brazilian Crania) [4] Gibelli et al. 2017 (Italian Crania) [12] 16.4/16.8 3.8 –

*No distinction between right and left sides; GPF — greater palatine foramen; SP — sagital plane; MM1, MM2, MM3 — maxillary molar I, II, III

indicate the exact type of their crania. Our obtained CONCLUSIONS values for distance to incisive foramen paralleled Greater palatine foramen might appear like an Kumar’s study regardless of sex [20]. anatomic obstruction in all the interventions involving Generally, GPF is considered as symmetrical at a maxillary division blockage of the trigeminal nerve each side at least one. The literature includes one especially in oral and maxillofacial surgery. We carried study which reports a bilateral lack of both GPF out this study on Turkish population in the belief that and LPF [33]. In our study, only at one sample, both an understanding of GPF-related variations and assess- GPF and LPF were lacking at right, whereas only LPF ment with different parameters would benefit clinicians were lacking at left (Fig. 3). Variance of the number in this sense. Regardless of sex and type of palate, GPF of LPF bilaterally regardless of palate type and sex mostly had an oval shape and aligned with the third was parallel with other studies [15, 33]. molar, and one each GPF changing in size was observed Greater palatine foramen showed variability in terms bilaterally in all the crania measured except for one. We of shape and size. In some cases, its diameter was less hope that the data we obtained will provide also an than 1 mm, close to LPF in size, while it was observed to index for future anthropological studies. be rather large in some other cases. As for shape, studies reported that it could be round or an oval structure Acknowledgements with its longitudinal axis extending antero-posteriorly This study was conducted as a MSc Thesis at [15, 21]. Previous studies carried out on different popu- Istanbul University Cerrahpasa School of Medicine lations reported it to often possess an oval shape with Anatomy Department during 2014 and 2015 Se- an antero-posterior extension. Our data demonstrated mester. Authors would like to thank chief of depart- these results are paralleled by our findings. ments of the Istanbul University, Istanbul School of

376 A. Ortug, M. Uzel, Evaluation of greater palatine foramen

Medicine and School of Dentistry Anatomy Depart- 17. Jaffar AA, Hamadah HJ. An analysis of the position of ments, Ankara University School of Medicine Anatomy the greater palatine foramen. J Basic Med Sci. 2003; 3(1): 24–32. Department, Istanbul Medipol University, School of 18. Jones RG. The reduction of bleeding in hare-lip and cleft- Medicine Anatomy Department and Bahcesehir Uni- palate surgery. Br J Anaesth. 1962; 34: 481–488, indexed in Pubmed: 14452322. versity School of Medicine Anatomy Department for 19. Jotania B, Patel SV, Patel SM, et al. Morphometric analysis their generous help as opening cranium collections of hard palate. Int J Res Med. 2013; 2(2): 72–75. 20. Kumar A, Anu S, Poonam S. Assessment of the relative belongs to their departments. location of greater palatine foramen in adult Indian skulls: Consideration for maxillary nerve block. Eur J Anat. 2011; REFERENCES 15(3): 150–154. 21. Langenegger JJ, Lownie JF, Cleaton-Jones PE. The relation- 1. Ajmani ML. Anatomical variation in position of the greater ship of the greater palatine foramen to the molar teeth palatine foramen in the adult human skull. J Anat. 1994; 184 and pterygoid hamulus in human skulls. J Dent. 1983; (Pt 3): 635–637, indexed in Pubmed: 7928651. 11(3): 249–256, indexed in Pubmed: 6580311. 2. Canter SR, Slavkin HC, Canter MR. Anatomical study of 22. Methathrathip D, Apinhasmit W, Chompoopong S, pterygopalatine fossa and canal: considerations applicable et al. Anatomy of greater palatine foramen and canal to the anesthetization of the second division of the fifth and pterygopalatine fossa in Thais: considerations for cranial nerve. J Oral Surg Anesth Hosp Dent Serv. 1964; 22: maxillary nerve block. Surg Radiol Anat. 2005; 27(6): 318–323, indexed in Pubmed: 14155475. 511–516, doi: 10.1007/s00276-005-0016-5, indexed in 3. Chentanez V, Kaweewongprasert S, Thunvarachorn P, et al. Pubmed: 16228112. Position of greater palatine foramens, length and direction of 23. Moore KL. Clinically Oriented Anatomy. 5th ed. Williams greater palatine canals: anatomic study of 120 adult human and Wilkins, Baltimore 2006: 886–996. skulls. Chula Med J. 1985; 29: 1187–97. 24. Nevin M, Puterbaugh PG. Conduction, Infiltration, and 4. Chrcanovic BR, Custódio ALN. Anatomical variation in the General Anesthesia in Dentistry. 4th ed. Brooklyn: Dental position of the greater palatine foramen. J Oral Sci. 2010; Items Interest Pub. 1938. 52(1): 109–113, indexed in Pubmed: 20339241. 25. Nimigean V, Nimigean VR, Buţincu L, et al. Anatomical 5. D’Souza AS, Mamatha H, Nayak J. Morphometric analysis and clinical considerations regarding the greater palatine of hard palate in south Indian skulls. Biomed Res. 2012; foramen. Rom J Morphol Embryol. 2013; 54(3 Suppl): 23: 173–175. 779–783, indexed in Pubmed: 24322027. 6. Dave MR, Gupta S, Vyas KK, et al. study of palatal indices 26. Piagkou M, Xanthos T, Anagnostopoulou S, et al. Anatomi- and bony prominences and grooves in the hard palate of cal variation and morphology in the position of the pala- adult human skulls. NJIRM. 2013; 4(1): 7–11. tine foramina in adult human skulls from Greece. J Cranio- 7. Diyarbakırlı S, Keleç P. Foramen palatinum majus varyasyonları. maxillofac Surg. 2012; 40(7): e206–e210, doi: 10.1016/j. Atatürk Üniv Diş Hek Fak Derg. 1996; 6(1): 37–39. jcms.2011.10.011, indexed in Pubmed: 22055651. 8. Douglas R, Wormald PJ. Pterygopalatine fossa infiltration 27. Romanes GJ. Cunningham Textbook ofAnatomy. 12th ed. through the greater palatine foramen: where to bend New York: Oxford University Press. 1981. the needle. Laryngoscope. 2006; 116(7): 1255–1257, 28. Sakul BU, Turkaslan AK. and Kose S.K. The examination of doi: 10.1097/01.mlg.0000226005.43817.a2, indexed in the asymmetry on the human dry skull. X. International Pubmed: 16826070. Symposium on Morphological Sciences, Toronto, Ontario, 9. Ducic Y, Herford AS. The use of palatal island flaps as an Canada, July 1-5. 1991(Poster No 88). adjunct to microvascular free tissue transfer for reconstruc- 29. Saralaya V, Nayak SR. The relative position of the greater tion of complex oromandibular defects. Laryngoscope. 2001; palatine foramen in dry Indian skulls. Singapore Med J. 111(9): 1666–1669, doi: 10.1097/00005537-200109000- 2007; 48(12): 1143–1146, indexed in Pubmed: 18043845. 00033, indexed in Pubmed: 11568625. 30. Sejrsen B, Kjaer I, Jakobsen J. Human palatal growth evalu- 10. Erimoğlu C, Özbayrak T, Kökten G. Foramen palatinum ated on medieval crania using nerve canal openings as majus ve canalis palatinus majorun anatoma-morfolojik references. Am J Phys Anthropol. 1996; 99(4): 603–611, varyasyonları. JIUFD 1983;17(3-4. : 144–157. doi: 10.1002/(SICI)1096-8644(199604)99:4<603::AID- 11. Gardner E, Gray DJ, Rahilly R. Anatomy: a regional study of hu- AJPA6>3.0.CO;2-U, indexed in Pubmed: 8779341. man structure. 4th edition. Philadelphia: WB Saunders. 1975. 31. Seldin H. Practical Anesthesia for Dental and Oral Surgery, 12. Gibelli D, Borlando A, Dolci C, et al. Anatomical character- Local and General. Philadelphia: Lea & Febiger. 1942. istics of greater palatine foramen: a novel point of view. 32. Shane SME. Principles of Sedation, Local, and General Surg Radiol Anat. 2017; 39(12): 1359–1368, doi: 10.1007/ Anesthesia in Dentistry. Springfield: Thomas. 1975. s00276-017-1899-7, indexed in Pubmed: 28752360. 33. Sharma NA, Garud RS. Greater palatine foramen--key to 13. Gözil R, Şakul BU, Çalgüner E, et al. Sert damak morfometrisi successful hemimaxillary anaesthesia: a morphometric ve torus palatinus, sutura palatina transversa şekilleri. Turkiye study and report of a rare aberration. Singapore Med J. Klinikleri J Dental Sci. 1999; 5(3): 149–153. 2013; 54(3): 152–159, indexed in Pubmed: 23546029. 14. Harnet JC, Lombardi T, Lutz JC, et al. Sagittal craniofacial 34. Tomaszewska IM, Tomaszewski KA, Kmiotek EK, et al. growth evaluated on children dry skulls using V2 and V3 Anatomical landmarks for the localization of the greater canal openings as references. Surg Radiol Anat. 2007; palatine foramen--a study of 1200 head CTs, 150 dry skulls, 29(7): 589–594, doi: 10.1007/s00276-007-0237-x, indexed systematic review of literature and meta-analysis. J Anat. in Pubmed: 17653618. 2014; 225(4): 419–435, doi: 10.1111/joa.12221, indexed in 15. Hassanali J, Mwaniki D. Palatal analysis and osteology of the Pubmed: 25131842. hard palate of the Kenyan African skulls. Anat Rec. 1984; 35. Wang TM, Kuo KJ, Shih C, et al. Assessment of the relative 209(2): 273–280, doi: 10.1002/ar.1092090213, indexed in locations of the greater palatine foramen in adult Chinese Pubmed: 6465536. skulls. Acta Anat (Basel). 1988; 132(3): 182–186, indexed in 16. Ikuta CR, Cardoso CL, Ferreira-Júnior O, et al. Position of the Pubmed: 3414365. greater palatine foramen: an anatomical study through cone 36. Westmoreland EE, Blanton PL. An analysis of the variations beam computed tomography images. Surg Radiol Anat. in position of the greater palatine foramen in the adult hu- 2013; 35(9): 837–842, doi: 10.1007/s00276-013-1151-z, man skull. Anat Rec. 1982; 204(4): 383–388, doi: 10.1002/ indexed in Pubmed: 23811933. ar.1092040412, indexed in Pubmed: 7181144. 377 Folia Morphol. Vol. 78, No. 2, pp. 378–388 DOI: 10.5603/FM.a2018.0087 O R I G I N A L A R T I C L E Copyright © 2019 Via Medica ISSN 0015–5659 journals.viamedica.pl

Investigation of the styloid process length in a Greek population N. Zokaris1, I. Siska2, K. Natsis3, M. Piagkou4, N. Lazaridis3, A. Skolka5, E. Piehslinger5

1Prosthodontics Department, 251 Hellenic Air Force General and VA Hospital, Athens, Greece 2Oral Radiology Department, 251 Hellenic Air Force General and VA Hospital, Athens, Greece 3Department of Anatomy and Surgical Anatomy, School of Medicine, Faculty of Health and Sciences, Aristotle University of Thessaloniki, Greece 4Department of Anatomy, Faculty of Health and Sciences, National and Kapodistrian University of Athens, Greece 5Prosthodontics Department, Dental School, Medical University of Vienna, Austria

[Received: 12 July 2018; Accepted: 15 August 2018]

Background: The styloid process (SP) is a slender cylindrical bony projection of the temporal bone with 2 ligaments and 3 muscles attached to it. Symptomatic SP elongation is also referred to, as Eagle’s syndrome. The aim of the present study is to investigate the distribution of the SP length in a young adult Greek population. Materials and methods: Moreover, we provide a comparison of the results by using two different methods for assessing SP elongation, as described in the literature. Finally, we explore the possibility of using orthopantomograms (OPGs), as a diagnostic aid by investigating inter-examiner, intra-examiner and inter-examination variability and we propose a limit for SP elongation measurable in OPGs. Results: The sample comprised 805 digital OPGs, taken from student pilots and engineers entering the Hellenic Air Force Academy, from 2008 onwards. Two measuring approaches were selected, one using the temporal bone, as a cranial landmark and the other, using the external auditory meatus. The end tip of the process was always the caudal landmark. The mean SP length was 28.42 ± ± 8.48 mm in males and 26.04 ± 7.69 mm in females, when measured from the temporal bone. The mean SP length was 38.35 ± 8.90 mm in males and 34.24 ± ± 8.63 mm in females, when measured from the external auditory meatus. The length of 30 mm is most commonly used as a starting point for SP elongation. In the total sample, 30.6% of the measured SPs exceeded the length of 30 mm. In males, 33.12% of the SPs were elongated; while in females the corresponding incidence was 20%. One hundred and nineteen (14.8%) SPs were not traceable. Conclusions: The SP is typically detectable and measurable in OPGs. An elongated SP should be kept in mind, since symptoms of elongation may overlap with clinical manifestations of temporomandibular joint disorders. (Folia Morphol 2019; 78, 2: 378–388)

Key words: temporomandibular disorders, Eagle’s syndrome, styloid syndrome, styloid process, elongation

Address for correspondence: M. Piagkou, DDS, MD, MSc, PhD, Assistant Professor of Anatomy, School of Medicine, Faculty of Health Sciences, National and Kapodistrian University of Athens, 75 Mikras Asias Street, 11527 Athens, Greece, e-mail: [email protected]

378 N. Zokaris et al., Styloid process length

INTRODUCTION hyperactivity in the patients’ group attributed to the The styloid process (SP) is a long osseous pro- SP interference. Krennmair and Piehslinger [34] sug- jection of the petrous part of the temporal bone gested that bone growth apposition or ligaments’ projecting forward, downward and slightly medial. ossificationdue to SP traction caused by malocclusion The functional unit, which is composed of 3 muscles may lead to a decrease in the hyoid bone mobility, (styloglossus, stylohyoid and stylopharyngeus) and affecting mandibular movement. Slavicek [60] con- 2 ligaments (stylohyoid and stylomandibular) attached sidered Eagle’s syndrome, as a cause for the reduced to it, is called styloid apparatus and is derived from mandibular mobility. There is however no unique the second branchial arch [11]. The symptomatic SP statement on what seems to be the normal SP length elongation is also referred to as Eagle’s syndrome and consequently on a proposed limit for elongation. [12–15] and causes cervicofacial pain, tinnitus and Eagle [13] reported that the typical SP length is 25 mm. otalgia. This definition can be found under the head- According to Stafne and Hollinshead [61], the SP var- ings of heterotopic ossification and/or temporal bone ies widely in length (5–50 mm), thickness, form and abnormalities. Alternative synonyms are “elongated shape. Langlais et al. [37] defined the typical SP length styloid process syndrome” [13, 19, 32], “styloid–stylo- varying between 25 and 32 mm and Moffat et al. [41] hyoid syndrome” [25], “carotid artery syndrome” [14] demonstrated values between 1.52 and 4.77 cm. Jung and stylalgia [27]. et al. [29] and Monsour and Young [42] proposed The stylohyoid ligament seems to be mainly in- 45 and 40 mm, as the limit for SP elongation. volved in Eagle’s syndrome pathophysiology. The In order to assess SP length, various imaging mo- ligament may be partially represented by bone, most dalities, such as orthopantomograms (OPGs), lateral often at its proximal extremity, or in some instances cephalometric radiographs [2, 47], anteroposterior may be completely ossified [61]. The topographic skull radiographs, computed tomography scanning anatomy of the parapharyngeal space emphasizes the [27, 44] and anatomical methods utilising measure- importance of SP length and angulation, since pos- ments in dry skulls [10, 11, 18, 38, 39, 41, 45, 50] sible compression and impingement of surrounding have been used. The advantage of metric studies in structures may appear in cases of SP elongation and/ dry skulls is lack of distortion or superimpositions of or ligaments’ ossification [9, 20, 51]. On the other structures. On the other hand, it is rather hard to find hand, muscles and ligaments attached to the SP reg- numerous intact SPs, since SP is a particularly fragile ulate the movements of the mandible, hyoid bone, part of the cranium [10]. tongue and pharynx [45]. Mineralisation/ossification Many radiological studies have used OPGs to eval- of the stylohyoid-stylomandibular ligament complex uate SP length (Table 1), the pattern of the stylohyoid is relative common [17] and may affect function of chain or the percentage of elongated complexes, the masticatory organ [60] (Fig. 1). however defined (Table 2), due to availability,- di Several of the symptoms attributed to an elongat- agnostic performance, low cost, less radiation and ed SP are common clinical manifestations of the tem- the particularly high sensitivity of the method in de- poromandibular joint disorders (TMD) and because tecting SP elongation [62, 63]. Three basic methods of this overlap, a potential of misdiagnosis always have been described regarding the assessment of exists [65]. Palesy et al. [49] concluded that the clinical the stylohyoid complex elongation in OPGs. In the examination of patients with cervicofacial pain and first method, Ferrario et al. [16] used the anterior mandibular dysfunction should include the SP investi- nasal spine and the mastoid process, as reference gation. Detailed knowledge of variations and possible landmarks. Those landmarks are connected with a line effects on suprahyoid structures is important for an and if the complex exceeded the line, it is considered accurate diagnosis of TMD and orofacial pain [34]. elongated. Therefore, SPs until 25 mm were normal in Vague facial pain, especially upon swallowing, turning the OPG, if they projected above the line connecting the head or opening the mouth, dysphagia, otalgia, the anterior nasal spine to the mastoid process. In headache and dizziness have been associated with the second method, Goldstein and Scopp [24] used elongated SPs or mineralised stylohyoid ligaments [8]. the height of the posterior border of the mandibular Siessere et al. [59] after a comparison of functional ramus to define SP elongation. If the SP exceeded electromyographic patterns in patients with Eagle’s the 1/3 of the height, it was considered elongated. syndrome and healthy individuals reported muscular Prasad et al. [52] used this approach before deciding

379 Folia Morphol., 2019, Vol. 78, No. 2

Topographic Compression/impingement Symptoms (facial pain, tinnitus, otalgia, anatomy of surrounding structures pain upon turning the head, swallowing, and mouth opening) Elongation

mineralisation Ligaments/muscles Regulate movements of mandible Possible impact on functional pattern attached hyoid, tongue, pharynx of the masticatory organ

Figure 1. Impact of styloid process elongation. Possible mechanisms and results.

Table 1. Limits and mean values for styloid process (SP) elongation in papers using orthopantomograms (OPGs) Study Year Limit for SP elongation Sample/OPGs Mean length Kaufman et al. [31] 1970 30 484 29.49 L/29.92 R Correl et al. [8] 1979 30 1771 43.6* Monsour and Young [42] 1986 30 1200 29.2 Keur et al. [32] 1986 30 1135 47.9 M/44.5 F Camarda et al. [5, 6] 1989 25 150 10 Zaki et al. [65] 1996 30 100 31.9 Bozkir et al. [4] 1999 30 200 53* Sokler and Sandev [62] 2001 < 21/21–30/30 > 296 21.6 R/21.2 L Yamashita [64] 2002 – 389 30.08 Krennmair and Piehslinger [35] 2003 25 420 12.6–23.2 Jung et al. [29] 2004 45 1000 31 Ilguy et al. [30] 2005 30 860 34 ± 8 L*/35 ± 15 R* Gokce et al. [23] 2008 30 698 38.1 M*/36.6 F* More and Asrani [43] 2010 30 500 25.41 L, 25.53 R de Andrade et al. [2] 2012 30 50 32.98 R/33.50 L Sancio-Goncalves et al. [57] 2013 30 78 TMD, 93 control 27.2, 28.2 Reddy et al. [55] 2013 30 260 36.7*

*Mean length of elongated SPs; F — females; L — left; M — males; R — right; TMD — temporomandibular joint disorder to perform surgical removal. The third method is not MATERIALS AND METHODS comparative. It is a metric approach utilising anatomic The sample comprised student pilots and engineers landmarks corresponding to the beginning and end entering the Hellenic Air Force Academy, from 2008 tip of the SP. Jung et al. [29] specified the starting onwards. A list of age specific candidates (17–21 years point of measurement, as the centre of a circle encir- old) was provided from the records of the Centre cling the cleft between SP and tympanic plate, while of Aviation Medicine in Athens, located in the 251 Monsour and Young [42] and Zaki et al. [65] used the Hellenic Air Force and VA Hospital. Upon acceptance point of SP emersion from the temporal bone. The in the Academy, all students had a thorough medical end tip of the SP or the ossified part of the ligament examination. An OPG combined with an intraoral and is obviously the other landmark. extraoral examination were part of this process. Gener- The aim of the current study is to investigate the al consent to anonymously use data for possible future distribution of SP length in a young adult Greek pop- research purposes is typically required from all patients ulation. Moreover we provide a comparison of the attending the hospital. All the dental records are kept results by using two different methods for assessing in the Dental Department of the Centre of Aviation SP elongation, as described in the literature. Finally we Medicine in Athens. Their digital copy is kept in the Oral explore the possibility of using OPGs, as a diagnostic Radiology Department of 251 Hellenic Air Force and VA aid by investigating inter-examiner, intra-examiner Hospital, where the standardised OPG was taken (Plan- and inter-examination variability and we propose mecaPromax 2002, 84 KV max/Total filtration 2.5 mm a limit for SP elongation measurable in OPGs. Al, Planmeca, 00880 Helsinki, Finland). Search strategy

380 N. Zokaris et al., Styloid process length

Table 2. Prevalence of styloid process (SP) elongation from papers using orthopantomograms (OPGs) Study Year Sample/OPGs SP elongation [%] Kaufman et al. [31] 1970 484 28% Goldstein and Scopp [24] 1973 554 22.2% Gossman and Tarsitano [25] 1977 4200 4% Correl et al. [8] 1979 1771 18.2% O’Carroll [46] 1984 479 35.3% Monsour and Young [42] 1986 1200 21.1% Keur et al. [32] 1986 1135 30% Ferrario et al. [16] 1990 286 84.4% Zaki et al. [65] 1996 100 27% Bozkir et al. [4] 1999 200 4% MacDonald-Jankowski [40] 2001 1662 7.8–8.6% Scaf et al. [58] 2003 166 12.6% Kursoglu et al. [36] 2005 55 83.6% Ilguy et al. [30] 2005 860 3.7% Radfar et al. [53] 2008 1000 22% Gokce et al. [23] 2008 698 7.7% More and Asrani [43] 2010 500 19.4% Bagga et al. [3] 2012 2706 52.1% Roopashri et al. [56] 2012 300 35.6% Alpoz et al. [1] 2014 1600 28.8%

required the use of the software “PlanmecaDimaxis orthopantomograph to orientate a standardised head Pro 4.3.2”. The programme required either the entry position. of the name or the birth date of an individual in order In order to ensure the measurements’ accuracy, to search the database for the corresponding OPG. If an attempt to calculate the OPG distortion in SP area more than one OPG happen to have the same name was made. For this purpose, 16 skulls were kindly pro- or birth date, then all of them are presented in a list. vided by the Departments of Anatomy and Surgical In the present study the variable “birth date” was used Anatomy of the Schools of Medicine of National and intentionally, for anonymity purposes and in order to Kapodistrian University of Athens and Aristotle Uni- include people randomly attending the department versity of Thessaloniki. Two radiological indexes were for an OPG and therefore widening the sample, with- constructed using wooden prefabricated sticks and out affecting age or geographic distribution of the metal balls/markers of 1.5 mm diameter each. Wood sample. Those individuals, that had the same birth was selected to ensure that no scattering would im- date with our initial list, were labelled duplicates and pede the measurements. The wooden rod was marked were included if the examination date did not exceed and cut with a disc to 22 mm and the metal balls 6/2015. Foreigners were excluded. were attached to each end with universal adhesive A dentist specialised in maxillofacial radiology glue. From end to end, the total distance was 25 mm performed panoramic imaging. For every OPG taken (Fig. 2A) and from ball centre to ball centre 23.5 mm. the exposure settings (KV and mA) were set to au- Each index was placed randomly to a right or left SP, tomatic. The patient was asked to stand up straight parallel to its longitudinal axis and orthodontic wax into the orthopantomograph and his maxillary and was used to keep it stable (Fig. 2B). Each skull was ir- mandibular anterior teeth were in the correspond- radiated twice. The process was repeated at different ing notch on the bite block device. Three references time intervals, repositioning the markers and setting laser lights (Frankfort horizontal plane light, canine the skull position according to the reference lines. light and mid-sagittal plane light) appeared on the Thirty-two OPGs were taken and 64 SPs/markers of

381 Folia Morphol., 2019, Vol. 78, No. 2

A B

Figure 2. A. Radiological index and digital calliper; B. Index attached with orthodontic wax.

Figure 3. A panoramic radiograph of a skull with the index positioned.

25 mm standard value were measured (Fig. 3). In the tory meatus (EAM) and the end of the SP tip. For both few cases that the ball (always the cranial one) was methods (Fig. 4), if the SP was nodular or radiopaque, not included in the image, the value was left blank. then the end tip of the last caudal radiopaque part Both sides were combined to give 58 measurements was defined as the end tip. All pictures, in which the in total and 6 blank. The test for the null hypothesis SP was not present or its total length was not clearly μ = 25 was not rejected, so the distortion could be depicted were marked as not traceable. considered negligible. In the present study 805 OPGs were selected, For each OPG selected, the described standardised obtained from 655 male (mean age 19.03 years) and steps were followed to measure SP length: 150 female patients (mean age 19.5 years). Each SP 1. Digital image opening; was assessed separately since SP elongation may be 2. Instant magnification; unilateral or bilateral. In 119 OPGs, SPs were charac- 3. Landmarks’ location at the right side (starting terised bilaterally as non-traceable. In the remaining point of measurement from a landmark caudal 686 OPGs, at least one SP was traceable (Table 3). For to cranial) and measurements’ registration; each identifiable SP four measurements were made: 4. Landmarks’ location at the left side and measure- measurement according to Jung et al. [29] method ments’ registration. by the main researcher, repeated twice, measurement To assess SP length, two approaches were used according to Jung et al. [29] method by the second [29, 45]. In the first approach [29], the landmarks examiner and measurement according to Natsis et al. used were: the centre of a circle encircling the cleft [45] method by the main researcher. between SP and tympanic plate and the end of the SP During statistical analysis, an extra value (e) was cre- tip. In the second approach [45] the used landmarks ated — by the combination of (a) and (b) values — the were: the lowest-posterior point of the external audi- mean value for each SP by the two measurements of the

382 N. Zokaris et al., Styloid process length

Figure 4. Measuring the styloid process length. The measured value is displayed in a drop up box. External acoustic meatus is also visible.

Table 3. Traceability of the styloid process in the study sample and between males and females according to the side of occurrence Side of occurrence Males Females Both genders Bilaterally not traceable 86 (13.1%) 33 (22%) 119 (14.8%) Unilaterally traceable: Left side 24 (3.7%) 9 (6%) 33 (4.1%) Right side 44 (6.7%) 15 (10%) 59 (7.3%) Bilaterally traceable 501 (76.5%) 93 (62%) 594 (73.8) Total 655 (100%) 150 (100%) 805 (100%)

The total number of traceable styloid processes — 686 main researcher. Measuring a structure twice or more responds to high validity of the utilised examination. and combining the measurements to provide a mean Inter-examiner and intra-examiner variability were adds to the accuracy of the process and reduces the tested by means of the non-parametric Wilcoxon error between measurements [2]. Comparison between signed rank test. The null hypothesis, that the two a–b describes the intra-examiner variability, between measurements were not different, was also retained a–c, the inter-examiner variability and between e–d at significance level 5%. Differences between meas- the relationship of the two methods of measurement. urements of the two examiners were also calculated. The non-parametric Mann-Whitney U test was Inter-OPG variability was a very interesting as- used to compare male and female values. Kolmogor- pect of this study. Out of the 805 individuals, 138 ov-Smirnov test was used to determine the normality had taken a second OPG within the time frame of of the distributions. The non-parametric Mann-Whitney the current study. These OPGs were also assessed, U test was used in order to compare two independent separately, creating a subsample of 138 images (109 samples, right and left. Non-parametric Mann-Whitney males, 19 females), which could be compared with tests comparing male versus female samples were their corresponding previous ones to evaluate how found to be significant, therefore the two samples reliable the OPG can determine SP length. The time were not combined. The extent of agreement among frame could certainly not justify a differentation in repeated measurements performed by an examiner length, hence the same structure was measured in is described by the intra-examiner variability and the two different OPGs. In such a case, the subsample extent of agreement between examiners is described could be unified regarding gender, since the point of by the inter-examiner variability. Low variability cor- interest was to directly check upon the accuracy of

383 Folia Morphol., 2019, Vol. 78, No. 2

Table 4. Descriptive statistics for the length of the styloid process (SPL), as it measured from the temporal bone (TB) and the external acoustic meatus (EAM) in males and females and in the total sample Metric variables Males Females SPL-R SPL-L SPL-R SPL-L SPL-R SPL-L SPL-R SPL-L from TB from TB from EAM from EAM from TB from TB from EAM from EAM N 545 525 520 497 108 102 108 103 Total 1070 1017 210 211 Mean ± SD 28.14 ± 8.35 28.60 ± 8.48 37.64 ± 8.62 38.98 ± 8.90 25.79 ± 7.48 26.27 ± 7.86 33.72 ± 8.23 34.78 ± 9.03 Mean ± SD (in total) 28.42 ± 8.48 38.35 ± 8.90 26.04 ± 7.69 34.24 ± 8.63 Median 26.32 26.83 35.58 37.32 25.27 25.11 32.60 34.17 Median (in total) 26.56 36.42 25.218 33.34 Minimum–maximum 10.33–76.60 9.68–70.20 12.94–84.34 15.38–83.94 9.86–54.80 9.65–60.77 14.34–66.68 10.89–72.44 Percentiles: 25 22.83 23.15 32.42 33.27 20.71 22.47 29.32 31.07 25 in total 23.07 32.79 22.09 30.31 50 26.32 26.83 35.58 37.32 25.27 25.11 32.60 34.17 50 in total 26.56 36.42 25.21 33.34 75 30.84 31.54 40.96 42.16 29.28 28.34 35.75 36.82 75 in total 31.50 41.47 28.75 36.49 90 38.19 39.65 48.65 51.89 32.81 33.54 41.96 43.24 90 in total 39.01 49.70 32.52 41.75

L — left; R — right; SD — standard deviation

OPG depicting the same structure, which definetely is values for females were 0.0006 and 0.09, respectively. not gender-related. Inter OPG variability was checked Differences between the two measurements by the for both methods — measurements from the tem- same examiner were also calculated. The mean value poral bone and from the EAM — for the former the for the calculated differences was 0.09 for the left SP mean values were compared with the measurements and 0.11 for the right SP in men. The corresponding in the second OPG. The protocol was approved by the values for females were –0.11 and –0.16, respectively. Ethics Committee, Educational and Scientific Board The proximity of the measurements is clearly depicted. of 251 Hellenic Air Force General and VA Hospital — The correlation coefficient between the two meas- Protocol Number 313/August 2015. uring methods was very high for genders, 0.931 for males and 0.952 for females. A correlation coefficient RESULTS near unity means that there is a linear relationship For the SP length assessment from the temporal between the two measuring methods. bone and the EAM, descriptive statistics are pre- The length of 30 mm is most commonly used as sented in Table 4. The mean SP length was 28.42 ± a starting point for SP elongation. In the total sample, ± 8.48 mm in males and 26.04 ± 7.69 mm in females, 30.6% of the measured SPs exceeded the length of when the SP length was measured from the temporal 30 mm (16.14% bilaterally and 14.53% unilaterally). bone. The mean SP length was 38.35 ± 8.90 mm in In males, 33.12% of the SPs were elongated (18.16% males and 34.24 ± 8.63 mm in females, when the SP bilaterally and 14.65% unilaterally), while in females length was measured from the EAM. SP elongation the corresponding incidence was 20% (12.6% unilat- was not side-related (p = 0.957 and 0.175 for both erally and 7.3% bilaterally). One hundred and nine- measurements in males, and p = 0.555 and 0.359 in teen SPs (14.8%) were not traceable (Table 5). females). Differences between measurements of the two examiners were also calculated. The mean value DISCUSSION of the calculated differences was 0.08 for the right SP In the present study, a particularly large sample and –0.08 for the left SP in men. The corresponding of 805 OPGs was examined. In the literature sample

384 N. Zokaris et al., Styloid process length

Table 5. Prevalence of styloid process (SP) elongation for pensated in the group of duplicates, leading finally to measurements from the temporal bone. Limit was set on 30 mm a total of 150 females. In the study by Correl et al. [8], SP Males Females Total which was also carried in a VA centre in the United Both > 30 mm 119 11 130 States, only 52 out of 1771 individuals were females Right only > 30 mm 47 13 60 and gender related comparisons were only descriptive. Regarding the correlation between gender and Left only > 30 mm 51 6 57 SP elongation, some papers reveal a male predomi- Both < 30 mm 352 87 439 nance, like in our study [26, 27, 43], others describe Both not traceable 86 33 119 a female predominance [7, 17, 30, 46, 52, 63], while Total 655 150 805 a third group supports a non-existing correlation [1, 16, 32, 36, 42, 45, 47, 55, 58]. sizes vary from small (50 patients in Kursoglu et al. As regards to the side asymmetry between left and [36] study) to extremely wide (4200 in Gossman and right SP, no statistical significant differences existed Tarsitano [25] study) depending on the applied meth- for both genders. According to Stafne and Hollins- odology and mainly on the tested variables. In the head [61], usually there is symmetry in SP length current investigation, the study population represents between the two sides, but sometimes there may be a random sample from Greece for three reasons: noteworthy differences concerning the SP position, a) candidates came from all over the country; b) 251 shape and size. SP in the present study was bilaterally Hellenic Air Force and VA Hospital is the only military traceable in 76.5% of the males and 62% of the fe- hospital of the Air Force in the country serving the males. From all cases of elongation, 54.8% were bilat- Veterans, the personnel of the Air Force (military and eral for males and 36.6% for females. The symmetrical administrative) and also members of their families; nature of this anatomic feature has been supported c) people working for the Armed Forces are tempo- by several authors [1, 40]. Regarding the unilateral rally transferred more often compared to other em- and bilateral SP elongation, Ferrario et al. [16] report- ployees. Geographic factors mainly connected with ed that the SP is often bilaterally elongated. Correl et dietary and daily habits have been correlated with al. [8] reported an incidence of bilateral elongation the variable prevalence of elongated SP in Indians of 16%, while the unilateral one was 2.25%. Zaki et [3, 54] and Turks [23]. al. [65] reported that from 27 out of 100 cases with The mean age of the participants in our sample at least one SP elongation, 16 cases were bilateral was 19.03 years for males and 19.5 years for females. and 11 unilateral. Ghosh and Dubey [21] mentioned The study is focused on young individuals in order to that an incidence of 57.1% of surgically treated pa- exclude aging as a possible factor for SP elongation tients showed bilateral SP elongation. Bagga et al. or ligaments’ ossification. A review of data literature [3] and More and Asrani [43] reported a bilateral SP concerning the relation between the stylohyoid appa- elongation in 79.5% and 68%. O’Carrol [46] reported ratus’ length and the age concludes to contradicted a bilateral calcification in 61% and a unilateral one results. This relation is in a way dependent on the in 17.53%. Side asymmetry was reported by Natsis etiologic theories utilised for explaining SP elonga- et al. [45] with the right side exhibiting higher values tion. Moreover, the fact that the plethora of papers of elongated SP. Andrade et al. [2] reported a slight subdivides their sample into different age groups tendency for the occurrence of elongation in the left makes impossible to draw conclusions. However, side. Scaf et al. [58] found unilateral elongation to many papers [5, 6, 11, 35, 46, 47, 61] seem to agree be the rule, with the majority of the cases (76.5%) that by the age of 20, the length and ossification in the right side and only 9.5% being bilateral. Bilat- have already been established, without excluding the eral elongation is usually not involved with bilateral potential for later stage ossification or elongation. symptoms [33, 52]. Prevalence of SP elongation and mean values were Styloid process elongation is the main reason for significantly higher in males, revealing a gender specif- clinical symptoms. An OPG still remains in the front ic dependency. The extremely higher ratio of males op- line, to detect SP elongation if Eagle’s syndrome is posed to females, especially in the students’ subgroup, suspected [28, 58, 63], as it has the advantage to dis- can easily be attributed to the military nature of the cern distinctly the entire SP length and its angulation academic school selected. However, the ratio is com- [21, 22]. Andrade et al. [2] also refer to the sensitivity

385 Folia Morphol., 2019, Vol. 78, No. 2

of OPGs due to head positioning. In the current study, difference in values is due to the part of the SP hidden three lines were used to orientate the head position in by the vaginal process. The typical SP length reaches relation to the Frankfurt horizontal plane, the midline the value of 30 mm [20]. Eagle [13] considered the plane and the canine plane. Distortions, magnifica- normal SP length to be 25 mm. According to Stafne tions and overlapping must always be kept in mind. and Hollinshead [61], the SP widely varies in length An attempt to calculate distortion in the present study (5–50 mm), thickness, form and shape. Moffat et al. was made to argue that measurements were as exact [41] regarded the normal range as being from 1.52 as possible. The measured SP length may vary with to 4.77 cm. Ianneti [28] considered normal values the process angulation itself [1, 22] and that is why varying from 20 to 35 mm and Reddy et al. [55] from the radiological index constructed to calculate the 20 to 30 mm. The normal length varies considerably distortion was adjusted across the longitudinal axis [23] and an attempt for grouping existing data is of the examined skulls. Jung et al. [29] used only two presented in Table 4. skulls to calibrate their measurements, while Monsour The use of different landmarks or even different and Young [42] used three, all placed in position by methods for assessing SP length is the main reason three different radiographers. In the present study for the diversities found in the literature regarding 16 skulls were used and irradiated twice ending up elongation (Table 5). Several authors use comparative with 64 possible measurements. Distortion could be approaches, while others proceed with measuring of considered negligible since the constructed index was the stylohyoid chain, either from the lower margin of 25 mm and the mean of calibration measurements the temporal bone or the EAM. In comparative ap- was 24.7 mm. proaches, other landmarks such as the mastoid process Several authors [16, 24, 40] utilised comparative [16], the posterior border of the mandibular ramus [24] approaches to estimate SP elongation, supporting or even the mandibular foramen [40] are used to form that measuring in OPGs is inaccurate. However, re- imaginary planes that represent limits for SP elonga- searchers utilising a metric approach measure what is tion. In the current study two methods were used to depicted as a SP, without claiming that the measure- compare the results and the accuracy of each method ment is the exact actual length of the SP. The desired and to end up with a possible correlation model be- is the identification of an elongated SP in an OPG, tween the used two methods that could be applied to thus considered in the differential diagnosis for the transform findings of other studies and make results orofacial pain of the patient. OPG is a valuable adjunct comparable. The selected methods by Jung et al. [29] in diagnostic procedure and clinicians should be able and Natsis et al. [45] were those selected due to the to recognise SP elongation. detailed description of the measuring process. A rather unique feature of the present study was Jung et al. [29] suggested the use of the 90th the comparison of the same structure in different percentile, as the limit for SP elongation (length’s OPGs. In the literature only Omnell et al. [47] com- distribution was not normal). The length distribution pared cephalometric radiographs of the same individ- of the current study was not normal and the values uals. That was actually a way to verify the preciseness for the 90th percentile were 39 mm for males and of the OPGs, as a diagnostic aid in measuring SP 32.5 mm for females — lower than Jung et al. [29] elongation since in the current study the time frame which were 47 mm and 45 mm, respectively. For the was not wide enough to support differentiation in EAM measurements no proposed limit in the literature length of the same individual. In the present study, exists but if we also consider the 90th percentile, then of the whole sample, 138 individuals had been irra- the corresponding values are 49.7 mm for males and diated twice. The hypothesis of equality was tested 41.8 mm for females. Mean values were 38 mm and by means of Wilcoxon signed rank test and was al- 34 mm, respectively. De Andrade et al. [2], who also ways retained — either for measurements from the used a very similar measuring approach, reported temporal bone or the EAM. The existed differences in mean values around 33 mm. Natsis et al. [45] meas- the number of measurements in males and females ured the SP emersion (floor of EAM) until its tip, after between the two methods is attributed to the fact drilling the entire part of the vaginal process and the that sometimes the EAM was not clearly depicted in outer part of the EAM. It is not easy to explain the the field of the OPG, while the point of SP emersion differences between populations if methodological from the temporal bone was clear, or inversely. The differences are excluded. Diet, habits, geographical

386 N. Zokaris et al., Styloid process length

parameters or even systematic conditions [3, 23, 48] 5. Camarda AJ, Deschamps C, Forest D, et al. I. Stylohyoid chain ossification: a discussion of etiology. Oral Surg Oral Med Oral Pathol. have been shown to interfere with the SP length, 1989; 67(5): 508–514, indexed in Pubmed: 2497419. while the theory one adopts for elongation may also 6. Camarda AJ, Deschamps C, Forest D, et al. II. Stylohyoid chain ossification: a discussion of etiology. Oral Surg Oral Med Oral Pathol. justify variability. Obviously, the lack of a clear formu- 1989; 67(5): 515–520, indexed in Pubmed: 2497420. lated and verified hypothesis regarding SP elonga- 7. Ceylan A, Köybaşioğlu A, Celenk F, et al. Surgical treatment of elongated styloid process: experience of 61 cases. Skull Base. tion explains the difficulty in attributing differences 2008; 18(5): 289–295, doi: 10.1055/s-0028-1086057, indexed in between samples to specific reasons with certainty. Pubmed: 19240828. 8. Correll RW, Jensen JL, Taylor JB, et al. Mineralization of the stylohyoid-stylomandibular ligament complex. A radiographic incidence CONCLUSIONS study. Oral Surg Oral Med Oral Pathol. 1979; 48(4): 286–291, indexed in Pubmed: 116179. Within the limitations of the current study the 9. Costantinides F, Vidoni G, Bodin C, et al. Eagle’s syndrome: following conclusions can be drawn: signs and symptoms. Cranio. 2013; 31(1): 56–60, doi: 10.1179/ crn.2013.008, indexed in Pubmed: 23461263. — Orthopantomograms are a valuable and reliable 10. De Paz FJ, Rueda C, Barbosa M, et al. Biometry and statistical aid in Eagle’s syndrome diagnosis. Inter and in- analysis of the styloid process. Anat Rec (Hoboken). 2012; 295(5): 742–747, doi: 10.1002/ar.22452, indexed in Pubmed: 22467546. tra-examiner agreement, as well as inter-exami- 11. Dwight T. Stylo-hyoid ossification. Ann Surg. 1907; 46(5): 721–735, nation reproducibility was high; indexed in Pubmed: 17862068. 12. Eagle WW. Elongated Styloid Processes: report of two cases. Arch — Both EAM and point of SP emersion from the tem- Otolaryngol Head Neck Surg. 1937; 25(5): 584–587, doi: 10.1001/ poral bone can be used as anatomical landmarks archotol.1937.00650010656008. 13. Eagle WW. Elongated styloid process; further observations and a in stylohyoid complex measurements. Difference new syndrome. Arch Otolaryngol. 1948; 47(5): 630–640, indexed corresponds to the hidden part by the vaginal in Pubmed: 18894764. 14. Eagle WW. Symptomatic elongated styloid process; report of process. The correlation coefficient between the two cases of styloid process-carotid artery syndrome with op- two methods was very high for both genders; eration. Arch Otolaryngol. 1949; 49(5): 490–503, indexed in Pubmed: 18131173. — The study concluded that a gender specific de- 15. Eagle WW. Elongated styloid process; symptoms and treatment. pendency of length and SP elongation had sig- AMA Arch Otolaryngol. 1958; 67(2): 172–176, indexed in Pu- bmed: 13497389. nificantly higher values in males; 16. Ferrario VF, Sigurtá D, Daddona A, et al. Calcification of the stylo- — Styloid process elongation was not side-related; hyoid ligament: incidence and morphoquantitative evaluations. Oral Surg Oral Med Oral Pathol. 1990; 69(4): 524–529, indexed — Almost 31% of the examined OPGs revealed at in Pubmed: 2326043. least one elongated SP. 17. Fini G, Gasparini G, Filippini F, et al. The long styloid process syndrome or Eagle’s syndrome. J Craniomaxillofac Surg. 2000; Styloid process is normally both detectable and 28(2): 123–127, doi: 10.1054/jcms.2000.0128, indexed in Pu- measurable in OPGs. Even if the measurement may bmed: 10958426. 18. Frommer J. Anatomic variations in the stylohyoid chain and their lack the absolute preciseness, an elongated SP is usu- possible clinical significance. Oral Surg Oral Med Oral Pathol. 1974; ally clearly depicted and must be kept in mind. Since 38(5): 659–667, indexed in Pubmed: 4530957. 19. Fritz M. Elongated styloid process: a cause of obscure throat symp- symptoms of elongation may overlap with symptoms toms. Arch Otolaryngol Head Neck Surg. 1940; 31(6): 911–918, of TMD and since OPGs are routinely used, dental doi: 10.1001/archotol.1940.00660010926002. 20. Fusco DJ, Asteraki S, Spetzler RF. Eagle’s syndrome: embryology, professionals must be familiar with this clinical and anatomy, and clinical management. Acta Neurochir (Wien). 2012; radiological entity. 154(7): 1119–1126, doi: 10.1007/s00701-012-1385-2, indexed in Pubmed: 22638594. 21. Ghosh LM, Dubey SP. The syndrome of elongated styloid pro- Acknowledgements cess. Auris Nasus Larynx. 1999; 26(2): 169–175, indexed in Pubmed: 10214896. We thank the Statistician Prof. Machaira-Kolyva Fo- 22. Ghosh S, Vengal M, Pai KM, et al. Remodeling of the antegonial tini for her valuable assistance in the statistical analysis. angle region in the human mandible: a panoramic radiographic cross-sectional study. Med Oral Patol Oral Cir Bucal. 2010; 15(5): e802–e807, indexed in Pubmed: 20383108. REFERENCES 23. Gokce C, Sisman Y, Ertas ET, et al. Prevalence of styloid process elongation on panoramic radiography in the Turkey population 1. Alpoz E, Akar GC, Celik S, et al. Prevalence and pattern of stylohy- from cappadocia region. Eur J Dent. 2008; 2(1): 18–22, indexed oid chain complex patterns detected by panoramic radiographs in Pubmed: 19212504. among Turkish population. Surg Radiol Anat. 2014; 36(1): 39–46, 24. Goldstein GR, Scopp IW. Radiographic interpretation of calcified doi: 10.1007/s00276-013-1137-x, indexed in Pubmed: 23689822. stylomandibular and stylohyoid ligaments. J Prosthet Dent. 1973; 2. de Andrade KM, Rodrigues CA, Watanabe PC, et al. Styloid process 30(3): 330–334, indexed in Pubmed: 4516436. elongation and calcification in subjects with TMD: clinical and 25. Gossman JR, Tarsitano JJ. The styloid-stylohyoid syndrome. J Oral radiographic aspects. Braz Dent J. 2012; 23(4): 443–450, indexed Surg. 1977; 35(7): 555–560, indexed in Pubmed: 406372. in Pubmed: 23207864. 26. Gözil R, Yener N, Calgüner E, et al. Morphological characteristics of 3. Bagga MB, Kumar CA, Yeluri G. Clinicoradiologic evaluation of sty- styloid process evaluated by computerized axial tomography. Ann loid process calcification. Imaging Sci Dent. 2012; 42(3): 155–161, Anat. 2001; 183(6): 527–535, doi: 10.1016/S0940-9602(01)80060- doi: 10.5624/isd.2012.42.3.155, indexed in Pubmed: 23071965. 1, indexed in Pubmed: 11766524. 4. Bozkir MG, Boğa H, Dere F. The evaluation of elongated styloid 27. Härmä R. Stylalgia: clinical experiences of 52 cases. Acta Otolar- process in panoramic radiographs in edentulous patients. Turkish yngol. 1967; Suppl 224: 149–155, indexed in Pubmed: 5992663. J Med Sci. 1999; 29: 481–486.

387 Folia Morphol., 2019, Vol. 78, No. 2

28. Iannetti G. Clinical-Radiographic analysis of eagle syndrome 46. O’Carroll MK. Calcification in the stylohyoid ligament. Oral atheroma of the carotid artery detected on a lateral radiograph. Surg Oral Med Oral Pathol. 1984; 58(5): 617–621, indexed in J Craniofac Surg. 2014; 25: 1578–1579. Pubmed: 6595625. 29. Jung T, Tschernitschek H, Hippen H, et al. Elongated styloid 47. Omnell KA, Gandhi C, Omnell ML. Ossification of the human stylohy- process: when is it really elongated? Dentomaxillofac Radiol. oid ligament: a longitudinal study. Oral Surg Oral Med Oral Pathol Oral 2004; 33(2): 119–124, doi: 10.1259/dmfr/13491574, indexed in Radiol Endod. 1998; 85(2): 226–232, indexed in Pubmed: 9503461. Pubmed: 15314005. 48. Öztaş B, Orhan K. Investigation of the incidence of stylohyoid 30. Ilgüy M, Ilgüy D, Güler N, et al. Incidence of the type and calcifi- ligament calcifications with panoramic radiographs. J Investig Clin cation patterns in patients with elongated styloid process. J Int Dent. 2012; 3(1): 30–35, doi: 10.1111/j.2041-1626.2011.00081.x, Med Res. 2005; 33(1): 96–102, doi: 10.1177/147323000503300 indexed in Pubmed: 22298518. 110, indexed in Pubmed: 15651721. 49. Palesy P, Murray GM, De Boever J, et al. The involvement of the 31. Kaufman SM, Elzay RP, Irish EF. Styloid process variation. Radiologic styloid process in head and neck pain: a preliminary study. J Oral and clinical study. Arch Otolaryngol. 1970; 91(5): 460–463, indexed Rehabil. 2000; 27(4): 275–287, indexed in Pubmed: 10792587. in Pubmed: 5442737. 50. Patil S. Morphometric study of the styloid process of tem- 32. Keur JJ, Campbell JP, McCarthy JF, et al. The clinical significance poral bone. J Clin Diagn Res. 2014; 8: 4–6, doi: 10.7860/ of the elongated styloid process. Oral Surg Oral Med Oral Pathol. jcdr/2014/9419.4867. 1986; 61(4): 399–404, indexed in Pubmed: 3458151. 51. Piagkou M, Anagnostopoulou S, Kouladouros K, et al. Eagle’s 33. Koivumäki A, Marinescu-Gava M, Järnstedt J, et al. Trauma induced syndrome: a review of the literature. Clin Anat. 2009; 22(5): eagle syndrome. Int J Oral Maxillofac Surg. 2012; 41(3): 350–353, 545–558, doi: 10.1002/ca.20804, indexed in Pubmed: 19418452. doi: 10.1016/j.ijom.2011.12.031, indexed in Pubmed: 22244077. 52. Prasad KC, Kamath MP, Reddy KJ, et al. Elongated styloid process 34. Krennmair G, Piehslinger E. The incidence and influence of ab- (Eagle’s syndrome): a clinical study. J Oral Maxillofac Surg. 2002; normal styloid conditions on the etiology of craniomandibular 60(2): 171–175, indexed in Pubmed: 11815916. functional disorders. Cranio. 1999; 17(4): 247–253, indexed in 53. Radfar L, Amjadi N, Aslani N, et al. Prevalence and clinical significance Pubmed: 10650396. of elongated calcified styloid processes in panoramic radiographs. 35. Krennmair G, Piehslinger E. Variants of ossification in the stylohyoid Gen Dent. 2008; 56(6): e29–e32, indexed in Pubmed: 21444269. chain. Cranio. 2003; 21(1): 31–37, indexed in Pubmed: 12555929. 54. Rath G, Anand C. Abnormal styloid process in a human skull. Surg 36. Kursoglu P, Unalan F, Erdem T. Radiological evaluation of the styloid Radiol Anat. 1991; 13(3): 227–229, indexed in Pubmed: 1754958. process in young adults resident in Turkey’s Yeditepe University 55. Reddy RS, Sai Kiran Ch, Sai Madhavi N, et al. Prevalence of elon- faculty of dentistry. Oral Surg Oral Med Oral Pathol Oral Radiol gation and calcification patterns of elongated styloid process in Endod. 2005; 100(4): 491–494, doi: 10.1016/j.tripleo.2005.05.061, south India. J Clin Exp Dent. 2013; 5(1): e30–e35, doi: 10.4317/ indexed in Pubmed: 16182171. jced.50981, indexed in Pubmed: 24455048. 37. Langlais RP, Miles DA, Van Dis ML. Elongated and mineralized 56. Roopashri G, Vaishali MR, David MP, et al. Evaluation of elongated stylohyoid ligament complex: a proposed classification and report styloid process on digital panoramic radiographs. J Contemp Dent of a case of Eagle’s syndrome. Oral Surg Oral Med Oral Pathol. Pract. 2012; 13(5): 618–622, indexed in Pubmed: 23250163. 1986; 61(5): 527–532, indexed in Pubmed: 3459129. 57. Sancio-Gonçalves FC, de Abreu MH, Netto Soares JM, et al. Stylohy- 38. Leite HF, Niccoli Filho WD, Liberti EA, et al. [Prevalence of elongated oid complex ossification in temporomandibular disorder: a case-con- styloid process in human skulls]. Rev Odontol UNESP. 1988; 17(1): trol study. J Prosthet Dent. 2013; 109(2): 79–82, doi: 10.1016/ 145–151, indexed in Pubmed: 3151022. S0022-3913(13)60019-0, indexed in Pubmed: 23395332. 39. Lengelé BG, Dhem AJ. Length of the styloid process of the temporal 58. Scaf G, Freitas DQ, De Loffredo LDCM. Diagnostic reproducibility bone. Arch Otolaryngol Head Neck Surg. 1988; 114(9): 1003–1006, of the elongated styloid process. J Appl Oral Sci. 2003; 11(2): indexed in Pubmed: 3408565. 120–124, indexed in Pubmed: 21409324. 40. MacDonald-Jankowski DS. Calcification of the stylohyoid complex 59. Siéssere S, Vitti M, de Sousa LG, et al. Changes in masticatory mus- in Londoners and Hong Kong Chinese. Dentomaxillofac Radiol. cles activity associated with Eagle’s syndrome. Electromyogr Clin 2001; 30(1): 35–39, doi: 10.1038/sj/dmfr/4600574, indexed in Neurophysiol. 2006; 46(4): 223–228, indexed in Pubmed: 16929628. Pubmed: 11175271. 60. Slavicek R. Clinical functional analysis with the aid of instrumental devic- 41. Moffat DA, Ramsden RT, Shaw HJ. The styloid process syndrome: es.The Masticatory Organ. English Ed. GAMMA Medizinisch. 2002: 384. aetiological factors and surgical management. J Laryngol Otol. 61. Stafne EC, Hollinshead WH. Roentgenographic observations on 1977; 91(4): 279–294, indexed in Pubmed: 856922. the stylohyoid chain. Oral Surg Oral Med Oral Pathol. 1962; 15: 42. Monsour PA, Young WG. Variability of the styloid process and 1195–1200, indexed in Pubmed: 13978697. stylohyoid ligament in panoramic radiographs. Oral Surg Oral Med 62. Sokler K, Sandev S. New classification of the styloid process Oral Pathol. 1986; 61(5): 522–526, indexed in Pubmed: 3086788. length-clinical application on the biological base. Coll Antropol. 43. More CB, Asrani MK. Evaluation of the styloid process on dig- 2001; 25(2): 627–632, indexed in Pubmed: 11811294. ital panoramic radiographs. Indian J Radiol Imaging. 2010; 63. Vieira EM, Guedes OA, Morais SDe, et al. Prevalence of elongated 20(4): 261–265, doi: 10.4103/0971-3026.73537, indexed in styloid process in a central brazilian population. J Clin Diagn Res. Pubmed: 21423900. 2015; 9(9): ZC90–ZC92, doi: 10.7860/JCDR/2015/14599.6567, 44. Murtagh RD, Caracciolo JT, Fernandez G. CT findings associated indexed in Pubmed: 26501021. with Eagle syndrome. AJNR Am J Neuroradiol. 2001; 22(7): 64. Yamashita M. Radiographic images of the styloid process. J Osaka 1401–1402, indexed in Pubmed: 11498437. Odontol Soc. 2002; 65: 165–180. 45. Natsis K, Repousi E, Noussios G, et al. The styloid process in 65. Zaki HS, Greco CM, Rudy TE, et al. Elongated styloid process in a Greek population: an anatomical study with clinical implications. a temporomandibular disorder sample: prevalence and treatment Anat Sci Int. 2015; 90(2): 67–74, doi: 10.1007/s12565-014-0232-3, outcome. J Prosthet Dent. 1996; 75(4): 399–405, indexed in indexed in Pubmed: 24664363. Pubmed: 8642526.

Studies on clinical anatomy of the maxillofacial and mandibular regions of the Madras Red sheep (Ovis aries) in India V. Sundaram1, P. Dharani2, R. Gnanadevi3, R. Kavya3

1Department of Basic Veterinary Sciences, School of Veterinary Medicine, Faculty of Medical Sciences, The University of the West Indies, Trinidad and Tobago 2College of Veterinary Science, SriVenkateswara Veterinary University, Proddatur, Andra Pradesh, India 3Department of Veterinary Anatomy, Madras Veterinary College, Tamilnadu Veterinary and Animal Sciences University, Chennai, India

[Received: 28 August 2018; Accepted: 3 October 2018]

Background: The present study aimed to study the clinical anatomy of the maxillofacial and mandibular regions of the Madras Red sheep through the morphometric analysis of some clinically important parameters pertaining to the regional anaesthesia. Materials and methods: The study was conducted in 20 male and 20 female adult Madras Red sheep around 20–25 kg body weight. Eleven parameters related to the clinical regional anaesthesia of the maxillofacial and the mandibular regions were recorded. Results: The distance between the facial tuberosity to the infraorbital canal was 2.25 ± 0.14 cm in males and 2.14 ± 0.15 cm in females. The distance from the infraorbital canal to the root of the alveolar tooth was 1.73 ± 0.28 cm in males and 1.73 ± 0.13 cm in females. The length and height of the mandibles were 15.37 ± 0.56 cm, 9.00 ± 0.24 cm and in males 14.53 ± 0.24 cm, 9.68 ± 0.16 cm in females, respectively. The maximum height of the mandible to the coronoid process, width or height of the mandibular foramen and the distance from the condyloid process to the base of the mandible were recorded as 9.0 ± 24, 0.82 ± 0.05, 6.68 ± 0.19 cm in males and 9.68 ± 0.16, 0.83 ± 0.04, 6.25 ± ± 0.28 cm in females, respectively. Conclusions: The present study revealed that most of the parameters showed statistically significant difference between the sexes i.e. the males had higher values. However, from the practical point of view, these differences were meagre. The results were discussed with regard to their clinical applications in various regional anaesthesia performed in maxillofacial and mandibular regions of the Madras Red sheep. (Folia Morphol 2019; 78, 2: 389–393)

Key words: clinical anatomy, Madras Red sheep, maxillofacial region, mandible

Address for correspondence: Dr. V. Sundaram, Department of Basic Veterinary Sciences, School of Veterinary Medicine, Faculty of Medical Sciences, The University of the West Indies, Trinidad and Tobago, tel: 1(868) 645-2640, ext. 4206, fax: 1(868) 662-9341, e-mail: [email protected]

389 Folia Morphol., 2019, Vol. 78, No. 2

INTRODUCTION The regional anatomy is the important branch of anatomy that deals with the form and relationships of various anatomic structures present in that area. It is one of the major foundations of clinical and surgical practice as it enables the clinician/surgeon to visualise the details of the structures relevant to the case at hand [1]. The knowledge of the regional anatomy of the head is crucial due to the presence of the vital organs and structures such as brain, tongue, eye, ears, nose, lips, horn and skull. Due to the presence of these structures, the head function is to coordinate the body, Figure 1. Lateral view of the skull of the Madras Red sheep; A — deglutition, olfaction and defence [1]. Numerous in- distance between the facial tuberosity to the orbital canal; B — vestigations were done on the regional anatomy of distance from the infraorbital canal to the root of the alveolar tooth; the head of the domestic animals including ox, horse, a — infraorbital foramen; b — facial tuberosity. sheep, goat and dog [1, 3, 4, 8, 10]. The morphological variations exhibited between the breeds in the same species and between the sexes within the same breed the dentition. All the selected animals were between will always be a hindering factor when the general data 18 to 24 months old and their weight ranged from of the species is extrapolated to all breeds and sexes 20 to 25 kg body weight. The heads of these selected of the species. Therefore, the breed- and sex-related animals were severed at the atlanto-occipital joint studies on the regional anatomy would be more use- after the slaughter and the head was collected and ful for the accurate understanding of the regional stored in a freezer at 20ºC. Since the animal heads anatomy of the head. were collected after the slaughter, the ethical approval The Madras Red sheep is a native breed of Tamil was not considered as necessary. The frozen heads Nadu in India, which contributes tremendously to the were processed by hot maceration techniques [13]. rural economy of the state by virtue of their meat and Eleven measurements on the maxillofacial and wool production. They are medium-sized, well-built, mandibular regions as adapted by Sarma and Devi hair sheep, well known for their known for high- [12] were measured by using metric ruler and vernier quality meat, skin, heat tolerance and adaptability calliper. The landmarks and methodology of each to the local agro-ecological conditions. The standard value obtained are described below and shown in anatomical details of the sheep are extrapolated to Figures 1–3. The data obtained were analysed by Madras Red sheep since the preliminary anatomi- the Student t test through the SPSS (Version 11.0) cal work is very scanty in this breed. The breed and software and are presented (Table 1). sex related craniometric data of the skulls in this Measurements of the maxillofacial and mandibu- animal would be very useful in clinical practice and lar regions: regional anaesthesia. Therefore, the present study — A: Distance between the facial tuberosity to the aimed to study the clinical anatomy of the maxil- infraorbital canal. It was measured from the level lofacial and mandibular regions of the Madras Red of the most lateral bulging of the facial tuberosity sheep through the morphometric analysis of some to mid-level of the infraorbital canal; clinically important parameters pertaining to the re- — B: Distance from the infraorbital canal to the root gional anaesthesia. of alveolar tooth. The measurement was taken mid-level of the infraorbital canal to the root of MATERIALS AND METHODS the alveolar tooth; A total number of 40 live apparently healthy adult — C: Length of the mandible. It was estimated by Madras Red sheep (20 males and 20 females) without measuring from the cranial extremity of the alveo- any skeletal deformation were first selected during lar root of the lower incisors to the caudal border ante-mortem examination at Chennai City Corpora- of the mandible; tion Slaughter House, Perambur, Chennai. The animals — D: Distance from the lateral alveolar root to the were weighed and the age was estimated through mental foramen. It was estimated as the shortest 390 V. Sundaram et al., Clinical anatomy of maxillofacial and mandibular regions of Madras Red sheep

Table 1. The measurements of the maxilla-facial and the mandibular regions of Madras Red sheep in cm (mean ± SE) Parameters* Male Female A 2.25 ± 0.15NS 2.14 ± 0.14NS B 1.73 ± 0.28NS 1.73 ± 0.13NS C 15.37 ± 0.56*** 14.53 ± 0.24*** D 2.20 ± 0.34** 1.87 ± 0.21** E 12.30 ± 0.46*** 11.53 ± 0.14*** F 3.63 ± 0.14*** 3.48 ± 0.12*** G 1.85 ± 0.16*** 1.60 ± 0.10*** H 2.88 ± 0.24** 2.68 ± 0.11** I 0.83 ± 0.05NS 0.82 ± 0.04NS Figure 2. Lateral view of the mandible of the Madras Red sheep; J 6.68 ± 0.19*** 6.25 ± 0.28*** C — length of the mandible; D — distance from the lateral root to the mental foramen; E — distance from the mental foramen to the K 9.00 ± 0.24*** 9.68 ± 0.16*** caudal border of the mandible. *As described in alphabetical order in materials and methods; NS — not significant (p ≥ 0.05); **Significant (p ≥ 0.05); ***Highly significant (p≤ 0.01); SE — standard error

the mandible is estimated from a horizontal line drawn from the level of convexity present at the ventral margin of the mandible; — G: Distance from the caudal border of the mandible to the middle of the vertical line produced by F as described above; — H: Height of the mandible from the condyloid process to the coronoid process; — I: Width of the mandibular foramen; — J: Distance from the condyloid process to the base of the mandible; Figure 3. Medial view of the mandible of the Madras Red sheep; F — — K: Maximum height of the mandible taken from distance of the mandibular foramen to the base of the mandible. The the basal level of the mandible to the highest level base of the mandible is estimated from a horizontal line drawn from the level of convexity present at the ventral margin of the mandible; of the coronoid process. G — distance from the caudal border of the mandible to the middle of the vertical line produced by the F as described above; H — height of RESULTS AND DISCUSSION the mandible to the condyloid process; I — width of the mandibular foramen; J — distance from the condyloid process to the base of the The several parameters showed statistically sig- mandible; K — maximum height of the mandible. nificant differences between the sexes as shown in the Table 1. The distance between the most lateral bulging distance from the mental foramen to the lateral of the facial tuberosity and the infraorbital canal end of the alveolar root of the lower incisor; was 2.25 ± 0.14 cm in males and 2.14 ± 0.15 cm in — E: Distance from the mental foramen to the caudal females. Distance from the infraorbital canal to the border of the mandible. It was taken from the root of the alveolar tooth was recorded as 1.73 ± level of the mental foramen to the extreme caudal ± 0.28 cm in males and 1.73 ± 0.13 cm in females. border of the mandible; Both the measurements did not shown any difference — F: Distance of the mandibular foramen to the base statistically between the sexes in the present study. of the mandible. It was estimated as the vertical The facial tuberosity of the Madras Red sheep was line from the ventral limit of the mandibular fo- prominent and placed at the level of 5th cheek tooth ramen to the base of the mandible. The base of as in other sheep breeds [7], whereas it was located at

391 Folia Morphol., 2019, Vol. 78, No. 2

the junction of the 4th and 5th cheek teeth in Kagani would be useful to the practitioners in confirmation of goats [11] and above the 3rd cheek tooth in ox [2]. The the mental nerve. All the measurements stated above distance from the facial tuberosity to the infraorbital regarding mandibles showed statistically significant canal and the distance from the infraorbital canal to differences between the males and females i.e. the the root of the alveolar tooth in the present study males showed higher values. However, the differences were 2.25 ± 0.15 cm and 2.14 ± 0.14 cm and 1.73 ± will be meagre when considering the practical point ± 0.28 cm and 1.73 ± 0.13 cm, respectively, whereas of view. The anaesthetic drugs can be injected in the values reported in Mehraban sheep [4] were 1.87 ± anterior aspect of the mandibular canal to desensi- ± 0.09 cm and 1.63 ± 0.04 cm and 2.36 ± 0.03 cm tise mental aspect of the mandibular nerve. This will and 2.06 ± 0.26 cm in Bakerwali goats [11]. The ensure the loss of sensation of the lower incisors, above data are of clinical importance because the premolar and lower lip on that side [3]. facial tuberosity is very prominent even in live ani- The mandibular length of 15.37 ± 0.56 cm in mals and acts as a guide for tracking the infraorbital males and 14.53 ± 0.24 cm in females and the maxi- nerve, and necessary for the desensitisation which mum mandibular height of 9.0 ± 0.24 cm, males and affects the skin of the upper lip, nostril and face on 9.68 ± 0.16 cm in females were less than the values that side of the level of the foramen. The infraorbital reported in the Mehraban sheep 15.76 ± 2.25 cm nerve is a continuation of the maxillary division of the and 9.57 ± 2.71 cm, respectively [5]. However, the trigeminal nerve passes through infraorbital canal; it values were higher than West African Dwarfs goats emerges out at the level of the infraorbital foramen, of Nigeria, which were 12.00 ± 1.89 cm and 6.90 ± and supplies the sensory innervation to the skin of 1.09 cm, respectively [8] and the black Bengal goat the upper lip, nostril and face on that side of the level 14.21 ± 0.98 cm and 8.83 ± 0.57 cm [14]. of the infraorbital foramen. The infraorbital canal is The mandibular nerve block is used to examine measured 5–6 cm along a line passing forward and the mandibular nerve during the clinical examinations downward from the anterior end of the facial crest and surgical procedures involving the alveoli and to the infraorbital foramen in horses [3], whereas teeth of the lower jaw in animals [6]. The width/height the infraorbital foramen is located 3 cm above the ratio of mental foramen in the present study can be gum line of the first cheek tooth and slightly rostral used to determine the exact area of the mandibular to it in cattle [6]. The injection of local anaesthetic nerve trunk. The distance from the mandibular fora- agents within the canal via the infraorbital foramen men to the base of the mandible was measured as will lead to analgesia of the incisor, canine and first 3.63 ± 0.14 cm males and 3.48 ± 0.12 cm in females. two premolars. Moreover, the infra-orbital foramen The distance from the caudal border of the mandible was located directly dorsal to the second or junction to the middle of the vertical line drawn downwards of the first and second upper premolar in Madras Red from the middle of the mandibular foramen to the sheep as in Maradi goats [9]. This information coupled base of the mandible was recorded as 1.85 ± 0.16 with the observed distance of 1.73 cm in between cm in males and 1.60 ± 0.10 cm in females. The same the root of the teeth and the foramen in the present values were recorded as 1.74 ± 0.33 cm in Mehraban study would prove a vital guide to administering the sheep [5] and 1.47 ± 0.25 cm in black Bengal goats infra-orbital nerve block in the Madras Red sheep. [14]. The same data was recorded in horses as 3.0 cm The distance from the lateral alveolar root to the [3], in dogs 1.5–2 cm [3] and in goats as 1.44 cm [12]. mental foramen was recorded as 2.20 ± 0.34 cm in The above data provided in this study will be useful males and 1.87 ± 0.21 cm in females (Fig. 2, Table 1) for the surgeons to locate the site for infiltration of whereas it was 1.6 ± 0.22 cm and 2.0 ± 0.3 cm in the anaesthetic drugs for the mandibular nerve block West African Dwarf goats and Maradi goats of Nige- in this particular Madras Red sheep. ria, respectively [8, 9]. It is an important landmark for the locating the mental nerve for mental nerve block CONCLUSIONS in Madras Red sheep. The measurements related to There is no previous information about these pa- mandible, namely, mandibular length, distance from rameters in Madras Red sheep, nor in any other sheep the lateral alveolar root to the mental foramen and breeds in India, to make comparisons. Therefore, the distance from the mental foramen to the caudal data presented in this study will form a vital baseline border of the mandible recorded in the present study for clinical manoeuvres around the head regional an-

392 V. Sundaram et al., Clinical anatomy of maxillofacial and mandibular regions of Madras Red sheep

aesthesia in the Madras Red sheep. Further, it will be west african dwarf goat in nigeria. Int J Morphol. 2005; very useful in further applied research works towards 23(1), doi: 10.4067/s0717-95022005000100006. 9. Olopade O, Onwuka SK. Osteometric studies of the red massive improvement in the livestock sector of the sokoto (Maradi) goats (Capra hircus): implication for re- international economy. gional anesthesia of the head. Int J Morphol. 2007; 25(2): 407–410, doi: 10.4067/S0717-95022007000200027. REFERENCES 10. Onar V, Ozcan S, Pazvant G. Skull typology of adult male 1. Dyce KM, Sack WO, Wensing CJG. Textbook of Veterinary Kangal dogs. Anat Histol Embryol. 2001; 30(1): 41–48, Anatomy. 2nd edn. Elsevier, Philadelphia 1996. doi: 10.1046/j.1439-0264.2001. 00292.x, indexed in 2. Getty R. The Anatomy of the Domestic Animals. Vol. I. Pubmed: 11284162. Saunders, Philadelphia 1975. 11. Sarma K. Morphological and craniometrical studies 3. Hall LW, Clarke KW. Trim, CM. Wright’s Veterinary Anes- on the skull of kagani goat (Capra hircus) of jammu thesia. 10th edn. ELBS and Baillierre Tindall, London 2000. region. Int J Morphol. 2006; 24(3), doi: 10.4067/s0717- 4. Karimi I, Onar V, Pazvant G, et al. The Cranial Morphomet- 95022006000400025. ric and Morphologic Characteristics of Mehraban Sheep 12. Sarma K, Devi J. Studies on clinical anatomy of the man- in Western Iran. Global Veterinaria. 2011; 6(2): 111–117. dibular and maxillofacial regions of the bakerwali goat 5. Karimi I, Hadipour M, Nikbakht P, et al. The Lower aw- (Capra hircus). Folia Vet. 2012; 56(2): 8–11. bone of Mehraban Sheep: A descriptive morphometric 13. Sundaram V, Leon K, Rao S, et al. Hind limb skeleton of approach. World’s Vet J. 2012; 2(4): 57–60. the orange rumped agouti (dasyprocta leporina linnaeus, 6. Lahunta ADE, Habel RE. Applied Veterinary Anatomy. WB 1758): structural and functional perspective. Ann. Res. Rev. Saunders Company; , Philadelphia 1986. Biol. 2017; 12(2): 1–12, doi: 10.9734/arrb/2017/30949. 7. May NDS. Anatomy of the Sheep – A dissection manual 14. Uddin M, Ahmed S, Islam K, et al. Clinical Anatomy of the 3rd ed. University of Queensland Press, St. Lucia 1970. Head Region of the Black Bengal Goat in Bangladesh. Int 8. Olopade J, Onwuka S. Some aspects of the clinical anat- J Morphol. 2009; 27(4): 1269–1273, doi: 10.4067/s0717- omy of the mandibular and maxillofacial regions of the 95022009000400048.

Accessory head of the flexor pollicis longus muscle: anatomical study and clinical significance D.R. Ballesteros1, P.L. Forero2, L.E. Ballesteros1

1Department of Basic Sciences, Industrial University of Santander. Bucaramanga, Colombia 2Department of Pathology, Industrial University of Santander. Bucaramanga, Colombia

[Received: 6 May 2018; Accepted: 24 June 2018]

Background: The accessory head of the flexor pollicis longus (AHFPL) has an oblique trajectory from medial to lateral aspect of the forearm below the flexor digitorum superficialis muscle and then joins the flexor pollicis longus muscle. When the anterior interosseous nerve (AIN) courses underneath the muscle belly of the AHFPL an entrapment neuropathy may occur, known as anterior interosseous nerve syndrome (AINS). Materials and methods: This descriptive cross-sectional study evaluated 106 fresh upper extremities. When the AHFPL was present, its fascicle was traced up to evaluate the origin site. The morphometric variables were measured using a digital micrometre (Mitutoyo, Japan). The relationship between the AHFLP and the AIN was evaluated. Results: The AHFPL was found in 34 (32.1%) of the 106 forearms. The AHFPL arose from the flexor digitorum superficialis muscle in 16 (47.1%) forearms, the medial epicondyle of the humerus in 10 (29.4%) forearms and the coronoid process of ulna in 8 (23.5%) forearms. The average total length of the AHFPL was 94.11 ± ± 10.33 mm. The AIN was located lateral to the AHFPL in 3 (8.8%) forearms, posterolateral in 7 (20.6%) forearms and posterior in 24 (70.6%) forearms. Conclusions: This study performed in a South American population sample revealed a prevalence of the AHFPL in a lower range compared to previous studies in North Americans and Asians. The AIN coursed more frequently underneath the muscle belly of AHFPL. This finding has clinical significance in the onset of the AINS and the subsequent surgical procedure for the AIN decompression. (Folia Morphol 2019; 78, 2: 394–400)

Key words: flexor pollicis longus, anterior interosseous nerve, nerve compression syndromes, anatomic variation, orthopaedic procedures

INTRODUCTION medial to lateral aspect of the forearm below the The accessory head of the flexor pollicis longus FDS and then joins the flexor pollicis longus muscle (AHFPL) was first described in 1813 by Karl Friedrich (FPL) [10, 21, 27]. Moreover, the AHFPL may also give Gantzer, a German anatomist [8]. This muscular struc- rise to a second tendinous contribution to the flexor ture may originate from the medial epicondyle of the digitorum profundus muscle (FDP) [3, 11, 16, 22]. humerus, the coronoid process of the ulna or the The AHFPL has been reported in different popula- flexor digitorum superficialis muscle (FDS) [7, 10, 14, tion groups with frequencies ranging between 25% 24, 27]. The AHFPL has an oblique trajectory from and 73.6%. Traditionally it has been considered as an

Address for correspondence: Dr. D.R. Ballesteros, Department of Basic Sciences, Faculty of Medicine, Industrial University of Santander. Cra. 32 #29-31, Bucaramanga, Colombia, tel: +573012352259, e-mail: [email protected]

394 D.R. Ballesteros et al., Accessory head of the flexor pollicis longus

anatomical variation, but in some population groups arm, the median and interosseous nerves were dis- it could be considered a normal pattern. Studies in sected. When the AHFPL was present, its fibres were Americans and Asians have reported frequencies of traced up doing a meticulous dissection to evaluate the AHFPL greater than 50% [10, 13, 14, 22]. When the site of origin. Tendons of the flexor carpi radialis this accessory muscular belly is present, it is more and palmaris longus muscles were identified and likely to be bilateral than unilateral [9, 11, 13, 14, 16]. reflected to expose the FPL and locate the insertion The occurrence of AHFPL could be explained dur- level of the AHFPL. ing the embryological development of the common The morphology was evaluated according to Pai flexormass, which differentiates into two layers: su- et al. [17] classification in three groups: strap-like perficial and deep [13, 17]. The deep layer in turn dif- (slender), fusiform and voluminous. The length of ferentiates into the FDP, the flexor pollicis longus and the muscle belly, the length of the tendon and the the pronator quadratus muscle [4, 24]. An incomplete width of the widest segment of the muscle belly were cleavage of the deep layer gives origin to the AHFPL measured using a digital micrometre (Mitutoyo, Ja- [11]. During the evolution, the Human acquired the pan). The relationship between the AHFLP and the AIN FPL. It is primitive or absent in primates, in which was evaluated. The statistical analysis was performed the inclusion of the FPL into the muscle mass of the using IBM® SPSS®. Continuous quantitative variables flexors is responsible for the loss of the functional were described with their averages and standard de- independence of the thumb flexor [7, 14]. viations. A Student’s t test was carried out accepting The AHFPL may lead to an entrapment neuropa- an alpha error of up to 5%. thy of the anterior interosseous nerve (AIN), known as Kiloh-Nevin syndrome or anterior interosseous RESULTS nerve syndrome (AINS) [5, 15, 25]. This syndrome is The accessory head of the flexor pollicis longus manifested clinically as a weakness in flexion of the was found in 34 (32.1%) of the 106 forearms. Of interphalangeal joint of the thumb and the distal the 53 cadaveric specimens, 23 (43.4%) presented interphalangeal joints of the index and middle finger AHFPL. In 11 (47.8%) individuals the AHFPL was found [18, 23, 26]. The present study was conducted to bilaterally and in 12 (52.2%) individuals was found evaluate the prevalence of this accessory muscle in a unilaterally. The AHFPL was distributed in 18 (53%) South American population sample, its morphometry right forearms and in 16 (47%) left forearms. No sta- and relationship with the AIN, constituting valuable tistical significant differences were found in relation information for clinical diagnosis and surgical ap- to the sides and the presence of the AHFPL (p > 0.05). proaches to the forearm. The origin of the AHFPL varied greatly. It arose from the flexor digitorum superficialis muscle in MATERIALS AND METHODS 16 (47.1%) forearms, the medial epicondyle of the This descriptive cross-sectional study evaluated humerus in 10 (29.4%) forearms and the coronoid 106 fresh upper extremities of 53 individuals who process of ulna in 8 (23.5%) forearms. No cases of were undergoing necropsy at the National Institute dual origin between the coronoid process and the of Forensic Medicine in Bucaramanga, Colombia. The medial epicondyle of the humerus were found. In all sample met the following inclusion criteria: male indi- the specimens, the insertion was on the ulnar border viduals over 18 years of age. Subjects with evidence of the FPL. In 82.1% of the cases the AHFPL inserted of direct trauma or scars in upper extremities were into the proximal third of the FPL, 14.2% into mid- excluded. The research protocol was developed in dle third, and 3.8% into the distal third. No cases of strict accordance with the Ethical Guidelines for Medi- second tendon to the FDP were found. cal and Health Research Involving Human Subjects The AHFPL was classified in three different mor- and was approved by the Ethics Committee of the phological groups: slender, fusiform and voluminous. Industrial University of Santander. The most common type was fusiform in 17 (50%) A longitudinal incision along the midaxillary line forearms (Fig. 1), then slender in 11 (32.4%) forearms was made extending from the antecubital fossa to (Figs. 2, 3) and voluminous in 6 (17.6%) forearms (Fig. 4). the wrist, which involved skin, subcutaneous tissue The average total length of the AHFPL was 94.11 ± and antebrachial fascia. Subsequently, the muscular ± 10.33 mm. The average length of the muscle belly structures of the anterior compartment of the fore- was 84.42 ± 9.27 mm and the tendon 9.68 ± 1.86 mm.

395 Folia Morphol., 2019, Vol. 78, No. 2

Figure 1. Anterior compartment of the right forearm. Fusiform mor- Figure 2. Anterior compartment of the right forearm. Slender morphology of the accessory head of the flexor pollicis longus (AHFPL) phology of the accessory head of the flexor pollicis longus (AHFPL) (arrow). The anterior interosseous nerve is located posterior to the (arrow); PT — pronator teres muscle; PL — palmaris longus mus- muscle belly of the AHFPL (asterisk); PT — pronator teres muscle; cle; FPL — flexor pollicis longus muscle; FDS — flexor digitorum FM — flexor mass; FPL — flexor pollicis longus muscle; FDS — superficialis muscle. flexor digitorum superficialis muscle; — median nerve.

The average width in the widest segment was 7.62 ± ing to that, in North-Americans and Asians the AHFPL ± 1.11 mm. The AIN was crossed by the belly of the could be considered as a normal pattern and not as AHFLP in 63.2% of the cases and by the tendon of an anatomical variation. the muscle in 36.8%. The AIN was located lateral to In South American population, earlier reports are the AHFPL in 3 (8.8%) forearms (Fig. 3), posterolateral contradictory. Studies by Caetano et al. [3] reveal in 7 (20.6%) forearms (Fig. 2) and posterior in 24 a frequency of 67.5% and studies by Riveros et al. (70.6%) forearms (Figs. 1, 4). No statistical significant [20] reveal a frequency of 10%. Including the present differences were found between the sides and the study, the consolidated prevalence of this group is morphometric variables (p > 0.05). 42.1%. In African population, studies by El Domiaty et al. [7] reveal a higher frequency (61.9%) but further DISCUSSION studies in this population group are needed to be rep- The frequencies of the AHFPL vary in different resentative. We found that the AHFPL occurred more population groups (Table 1), with frequencies rang- frequently bilaterally, agreeing with the findings of ing between 25% and 73.6%. The frequency of the Mangini [14], Uyaroglu et al. [27], Mahakkanukrauh AHFPL observed in the present study is in a low range et al. [13], Oh et al. [16]), Jones et al. [11], and Gun- compared to previous studies. In North-American nal et al. [9]. population, previous studies reveal a consolidated Regarding the origin of the AHFPL muscle, some frequency of 53.82% and in Asians 52.12%. Accord- authors have reported the medial epicondyle of the

396 D.R. Ballesteros et al., Accessory head of the flexor pollicis longus

Figure 3. Anterior compartment of the left forearm. Slender mor- Figure 4. Anterior compartment of the right forearm. Voluminous phology of the accessory head of the flexor pollicis longus (AHFPL) morphology of the accessory head of the flexor pollicis longus (arrow). The anterior interosseous nerve is located lateral to the (AHFPL) (arrow). The anterior interosseous nerve is located pos- muscle belly of the AHFPL (asterisk); uPT — ulnar head of the terior to the muscle belly of the AHFPL (asterisk); — median pronator teres muscle; Br — brachioradialis muscle; FPL — flexor nerve. pollicis longus muscle; — median nerve.

humerus as the most frequent site [1, 6, 10, 13, 14, to adequately distinguish in which of these three 22], others have reported that is the coronoid process structures the AHFPL arises from. [9, 16, 27]. In our series, the most frequent origin was Regarding the insertion, earlier reports [1, 7, 11, the flexor digitorum superficialis muscle, similarly -re 13] reveal that the AHFPL muscle inserts into the ported by Jones et al. [11], El Domiaty et al. [7], Kara ulnar border of the FPL. In our series, the most fre- et al. [12] and Caetano et al. [3]. However, a recent quent insertion level was the proximal third of the meta-analysis of 24 cadaveric studies found that the FPL tendon, similarly reported by Jones et al. [11] most common origin of the AHFPL was the medial and El Domiaty et al. [7]. There were no instances of epicondyle of the humerus [19]. a second tendon to the FDP. This discrepancy can be explained by the fact Jones et al. [11] were the first to classify the mor- that in its origin the AHFPL is intimately related with phology of the AHFPL in four groups: slender, volu- other structures of the flexor mass, leading to a pos- minous, triangular and fusiform. Subsequently Oh et sible misidentification by authors [3, 10, 16, 19]. It is al. [16] described three different morphological groups: necessary to do a meticulous dissection to determine papillary, spindle and band type. Afterwards Pai et al. the origin appropriately, because the FDS has fasci- [17] classified the morphology in three groups: strap- cles originated in the coronoid process and medial like (slender), fusiform and voluminous. In our series, epicondyle of the humerus. Therefore, it is important the most frequent type was fusiform, consistent to the

397 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. Prevalence of the accessory head of the flexor pollicis longus (AHFPL) reported in previous studies and organized by population groups Authors Nationality Sample Cases Population Prevalence Mangini (1960) American 76 56 Shirali et al. (1998) American 60 33 Malhotra, Sing and Tewari (1982) American 240 130 North American 53.82% Dykes and Anson (1944) American 150 80 Dellon and Mackinnon (1987) Canadian 43 14 Tubbs et al. (2006) American 20 4 Hemmady, Subramanya and Mehta (1993) Indian 54 36 Oh, Chung and Koh (2000) Korean 72 48 Kida (1988) Japanese 132 82 Mahakkanukrauh et al. (2004) Thai 240 149 Al-Qattan (1996) Saudi 25 13 Gunnal et al. (2013) Indian 180 92 Asian 52.12% Mori (1964) Japanese 205 103 Pai et al. (2008) Indian 126 58 Sharma et al (2008) Indian 60 24 Uyaroglu, Kayalioglu and Erturk (2006) Turkish (Anatolian) 52 27 Kara et al. (2012) Turkish (Anatolian) 142 49 Bilecenoglu, Uz and Karalezli (2005) Turkish (Anatolian) 30 6 Jones et al. (1997) British 80 36 Dolderer et al. (2011) German 19 5 European 49.71% Wood (1868) British 72 44 Caetano et al. (2015) Brazilian 80 54 Riveros, Olave and Sousa-Rodrigues (2015) Brazilian 30 3 South American 42.13% Present study (2018) Colombian 106 34 El Domiaty, Zoair and Sheta (2008) Egyptian 42 26 African 61.90% Overall – 2336 1206 – 51.6%

reported by Kara et al. [12], Pai et al. [17], El Domiaty ports found that the right side was wider than the et al. [7], Caetano et al. [3], Gunnal et al. [9]. left [10, 13], with a statistically significant differ- Regarding the morphometric data, the length of ence, explained possibly since most people are right the muscle belly and the tendon have been reported handed [13]. However, our findings did not show to range between 60 and 80.4 mm and between any statistically significant difference between sides 8 and 20.7 mm. respectively [7, 9, 11, 12, 16, 17, 26]. and width (p > 0.05). Our findings reveal greater lengths compared to stud- The relationship between the AHFPL and the AIN ies by Pai et al. [17], Kara et al. [12] and Oh et al. [16] was described originally by Mangini [14] who stated but similar to the reported by Uyaroglu et al. [27] and that the muscle always laid between the median nerve Gunnal et al. [9]. anteriorly and the AIN posteriorly. Subsequently, Del- The width has been reported to range between lon and Mackinnon [6] and Al-Qattan [1] stated that 6.9 and 7.8 mm [9, 12, 16, 26]. In our series, the the muscle always laid posterior to both the median width of the muscle belly was 7.62 ± 1.11 mm, nerve and the anterior interosseous nerve. In our se- consistent to the previously reported. Earlier re- ries, the median nerve was always located anterior to

398 D.R. Ballesteros et al., Accessory head of the flexor pollicis longus

the AHFPL. Moreover, the AIN was located posterior, References posterolateral or lateral to the AHFPL, similarly to 1. Al-Qattan MM. Gantzer’s muscle. An anatomical study of earlier reports [12, 13, 26]. the accessory head of the flexor pollicis longus muscle. In our series, the AIN was crossed by the muscle J Hand Surg Br. 1996; 21(2): 269–270, indexed in Pub- belly of the AHFLP in most of the cases, in concor- med: 8732417. 2. Azar FM, Canale ST, Beaty JH. Campbell’s Operative Or- dance with the results of the meta-analysis by Roy et thopaedics. 13th ed. Philadelphia: Elsevier. 2016: 3213. al. [19]. This finding is clinically relevant considering 3. Caetano EB, Sabongi JJ, Vieira LÂ, et al. Gantzer muscle. that the AINS may occur when the AIN courses un- An anatomical study. Acta Ortop Bras. 2015; 23(2): 72–75, derneath the muscle belly of the AHFPL originating doi: 10.1590/1413-78522015230200955, indexed in an entrapment neuropathy [2, 5, 24]. There are two Pubmed: 27069404. 4. Cihák R. Ontogenesis of the skeleton and intrinsic muscles characteristics of the AHFPL that could be predispos- of the human hand and foot. Ergeb Anat Entwicklungs- ing factors in the onset of the AINS: a hypertrophy gesch. 1972; 46(1): 5–194, indexed in Pubmed: 5043313. of the muscle belly [3, 6, 22] and the morphology by 5. Degreef I, De Smet L. Anterior interosseous nerve paralysis itself [19], since a greater contact surface with the due to Gantzer’s muscle. Acta Orthop Belg. 2004; 70(5): 482–484, indexed in Pubmed: 15587039. AIN may lead to an entrapment neuropathy. 6. Dellon AL, Mackinnon SE. Musculoaponeurotic variations The AIN is purely a motor nerve. It innervates the along the course of the median nerve in the proximal pronator quadratus, the FDP of the index and mid- forearm. J Hand Surg Br. 1987; 12(3): 359–363, indexed dle fingers, and the flexor pollicis longus [25]. The in Pubmed: 3437205. 7. El Domiaty MA, Zoair MM, Sheta AA. The prevalence of AINS is manifested clinically as weakness in flexion accessory heads of the flexor pollicis longus and the flexor of the interphalangeal joint of the thumb and the digitorum profundus muscles in Egyptians and their relations distal interphalangeal joints of the index and middle to median and anterior interosseous nerves. Folia Morphol. finger [2,18, 22]. During the physical examination, the 2008; 67(1): 63–71, indexed in Pubmed: 18335416. physician may ask the patient to join the thumb and 8. Gantzer KF. Dissertatio anatomica musculorum varietates sistens: quam consensu gratiosi medicorum ordinis. Bero- index fingers together forming a circle. The patient lini: Typis Joannis Friderici Starckii. 1813; 13: 14. with AINS will be unable to do an active flexion of 9. Gunnal SA, Siddiqui AU, Daimi SR, et al. A Study on the interphalangeal joint of the thumb and the distal the Accessory Head of the Flexor Pollicis Longus Muscle interphalangeal joint of the index finger [5, 18, 22, (Gantzer’s Muscle). J Clin Diagnostic Res. 2013; 7(3): 418–421, doi: 10.7860/jcdr/2013/4545.2788. 24]. During the surgical exploration, if the AHFPL is 10. Hemmady MV, Subramanya AV, Mehta IM. Occasional encountered, it is mandatory to resect any fibrous head of flexor pollicis longus muscle: a study of its mor- bands that may be compressing the AIN [2, 5, 24]. phology and clinical significance. J Postgrad Med. 1993; 39(1): 14–16, indexed in Pubmed: 8295137. CONCLUSIONS 11. Jones M, Abrahams PH, Sañudo JR, et al. Incidence and morphology of accessory heads of flexor pollicis longus The AHFPL has traditionally been considered and flexor digitorum profundus (Gantzer’s muscles). J Anat. a variant, however, in North Americans and Asians 1997; 191 (Pt 3): 451–455, indexed in Pubmed: 9419002. could be considered as a normal pattern. This study 12. Kara A, Elvan O, Yildiz S, et al. Accessory head of flexor performed in a South American population sample, pollicis longus muscle in fetuses and adult cadavers and its relation to anterior interosseous nerve. Clin Anat. revealed a prevalence of the AHFPL in a lower range 2012; 25(5): 601–608, doi: 10.1002/ca.21296, indexed compared to previous studies in North Americans in Pubmed: 22038878. and Asians. The results of the morphometric vari- 13. Mahakkanukrauh P, Surin P, Ongkana N, et al. Prevalence ables in our series are consistent with the findings in of accessory head of flexor pollicis longus muscle and its relation to anterior interosseous nerve in Thai population. the other population groups. The AIN coursed more Clin Anat. 2004; 17(8): 631–635, doi: 10.1002/ca.20016, frequently underneath the muscle belly of AHFPL. indexed in Pubmed: 15495169. This finding has clinical significance in the onset of 14. Mangini U. Flexor pollicis lungus muscle. Its morphology the AINS and the subsequent surgical procedures to and clinical significance. J Bone Joint Surg Am. 1960; 42-A: decompress the AIN. 467–470, indexed in Pubmed: 13853793. 15. Nagano A. Spontaneous anterior interosseous nerve palsy. J Bone Joint Surg. British volume. 2003; 85-B(3): 313–318, Acknowledgements doi: 10.1302/0301-620x.85b3.14147. We would like to recognise the National Institute 16. Oh CS, Chung IH, Koh KS. Anatomical study of the acces- of Forensic Medicine (Instituto Nacional de Medicina sory head of the flexor pollicis longus and the anterior interosseous nerve in Asians. Clin Anat. 2000; 13(6): Legal y Ciencias Forenses) for supplying the anatomi- 434–438, doi: 10.1002/1098-2353(2000)13:6<434::AID- cal specimens used in this study. CA7>3.0.CO;2-4, indexed in Pubmed: 11111895. 399 Folia Morphol., 2019, Vol. 78, No. 2

17. Pai MM, Nayak SR, Krishnamurthy A, et al. The acces- 22. Shirali S, Hanson M, Branovacki G, et al. The flexor pol- sory heads of flexor pollicis longus and flexor digitorum licis longus and its relation to the anterior and posterior profundus: Incidence and morphology. Clin Anat. 2008; interosseous nerves. J Hand Surg Br. 1998; 23(2): 170–172, 21(3): 252–258, doi: 10.1002/ca.20612, indexed in Pub- indexed in Pubmed: 9607652. med: 18351652. 23. Spinner M. The anterior interosseous-nerve syndrome, 18. Park IJ, Roh YT, Jeong C, et al. Spontaneous anterior with special attention to its variations. J Bone Joint Surg interosseous nerve syndrome: clinical analysis of eleven Am. 1970; 52(1): 84–94, indexed in Pubmed: 5411776. surgical cases. J Plast Surg Hand Surg. 2013; 47(6): 24. Standring S. Gray’s Anatomy: The Anatomical Basis of 519–523, doi: 10.3109/2000656X.2013.791624, indexed Clinical Practice. 41st ed. Elsevier; 2016: 837–861. in Pubmed: 23627594. 25. Tabib W, Aboufarah F, Asselineau A. Compression du nerf 19. Roy J, Henry BM, Pękala PA, et al. The prevalence and interosseux antérieur par le muscle de Gantzer. Chirurgie anatomical characteristics of the accessory head of the de la Main. 2001; 20(3): 241–246, doi: 10.1016/s1297- flexor pollicis longus muscle: a meta-analysis. Peer J. 3203(01)00041-5. 2015; 3: e1255, doi: 10.7717/peerj.1255, indexed in 26. Ulrich D, Piatkowski A, Pallua N. Anterior interosseous Pubmed: 26557419. nerve syndrome: retrospective analysis of 14 patients. 20. Riveros A, Olave E, Sousa-Rodrigues C. Estudio Anatómico Arch Orthop Trauma Surg. 2011; 131(11): 1561–1565, del Fascículo Accesorio del Músculo Flexor Largo del doi: 10.1007/s00402-011-1322-5, indexed in Pub- Pulgar y su Relación con el Nervio Interóseo Anterior en med: 21611763. Individuos Brasileños. Int J Morphol. 2015; 33(1): 31–35, 27. Uyaroglu FG, Kayalioglu G, Erturk M. Incidence and doi: 10.4067/s0717-95022015000100004. morphology of the accessory head of the flexor pollicis 21. Sharma M, Chabbra U, Kaushal S, et al. Accessory head longus muscle (Gantzer`s muscle) in a Turkish population. of flexor pollicis longus muscle. J Exerc Sci Physiother. Neurosciences (Riyadh). 2006; 11(3): 171–174, indexed in 2008; 4(1): 15–18. Pubmed: 22266615.

Age-related structural-mechanical property changes in human peroneus longus muscle A. Niyomchan1, B. Panichareon1, A. Siriphorn2, T. Wongtawatchai3

1Department of Anatomy, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand 2Department of Physical Therapy, Faculty of Allied Health Sciences, Chulalongkorn University, Bangkok, Thailand 3Department of Anatomy, Faculty of Science, Prince of Songkla University, Hat Yai, Songkhla, Thailand

[Received: 31 August 2018; Accepted: 8 October 2018]

Background: Functional impairment of the muscle-tendon unit is one of the most remarkable effects of aging. The function of the peroneus longus muscle is to stabilise the foot and ankle joint. A deterioration of the structural and mechanical properties of this muscle can potentially lead to foot problems in older adults. This study aimed to investigate the effects of age on structural, histological, and mechanical features in peroneus longus muscle samples taken from embalmed cadavers of two different age groups; young adult (30–60 years) and old adult (over 60 years). Materials and methods: The mechanical property was analysed through the results of cross-sectional area, tensile, tensile stress, and modulus of elasticity. The arrangement of the collagen in the perimysium and tendon was examined by scanning electron microscopy. Fatty infiltration within the musculotendinous junction was evaluated by Masson’ trichrome stained muscle sections. Results: This study thus provides evidence that there are indeed age-induced mechanical property changes in the peroneus longus muscle, which include reductions in the tensile force, tensile stress, and modulus of elasticity, and is related to the malformation of collagen fibrils and the massive fat accumulation in the musculotendinous junction. Conclusions: These alterations may further result in a reduction of muscle strength and quality in an older person. (Folia Morphol 2019; 78, 2: 401–407)

Key words: aging, collagen, musculotendinous junction, tendon

INTRODUCTION unit can cause foot problems, a common concern for Aging, a gradual deterioration of the physiologi- older adults [24]. Indeed, among the many muscles and cal processes associated with a dramatic decrease in tendons in the foot, the tibialis posterior and peroneus muscle-tendon performance [27, 28]. A reduction in longus have essential roles. The tibialis posterior is the muscle strength and mechanical properties of tendons strong supporter of the medial longitudinal arch and often leads to disability and loss of mobility in older assists in plantar flexion and inversion [11, 13], while adults [10]. One of the most extensively utilised parts of the peroneus longus produces eversion and plantar the body is the foot, which is primarily responsible for flexion. Furthermore, it is also responsible for stabilising weight-bearing activities. Muscles and tendons, aside two longitudinal arches (medial and lateral arches) [14, from bones, are responsible for controlling the normal 25]. As such, both of these muscles affect any dynamic position and function of the foot. Changes in structural changes involved in foot alignment [14]. Tibialis poste- and mechanical properties of the musculotendinous rior dysfunction is well recognised as a cause of flatfeet

Address for correspondence: Dr. A. Niyomchan, Department of Anatomy, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, tel: +662-419-5170, e-mail: [email protected]

401 Folia Morphol., 2019, Vol. 78, No. 2

Figure 1. The schematic drawing representing the set up used for testing mechanical properties of peroneus longus in adult and older adult cadavers through the use of a multi-functional digital force gauge.

in both adults and the elderly [13]. Thus, many studies junction was identified and marked with a permanent often investigate the alterations of this muscle during marker. The specimen was then cut into a 90-mm long the development of aging [3, 34]. However, there has section with a 70-mm long tendon side and 20-mm long been no comprehensive evaluation of degenerative muscle fibre side. Both free ends of the muscles were changes in structural and mechanical properties of the then clamped with adjustable C-clamps. The specimen peroneus longus muscle to date. Accordingly, the pur- was then placed in a multi-functional digital force gauge pose of this study was thus to examine aging-induced IPX-800 (INSPEX, Bowers Group, United Kingdom) with alterations involved in the structural and mechanical the side of the tendon facing the force gauge (Fig. 1). properties of the peroneus longus which can control The validity and reliability were analysed by four known foot posture along with the other foot muscles. While weights and test-retest values of four specimens, re- the morphological alterations observed in the present spectively. This system was found to be valid (per cent study may be one mechanism in the age-related struc- accuracy: ± 1.9%) and reliable (ICC3,1 = 0.93). Prior to tural, functional changes of muscle-tendon complex and mechanical testing, the specimen was tightened to allow new information will grant a clear awareness of the ef- for viscoelastic stress relaxation. The force gauge was fects of peroneus longus degradation in an older person. then calibrated to zero. Subsequently, an axial tensile force was applied to the muscle fibre side until it reached MATERIALS AND METHODS a 1-mm displacement of the musculotendinous junction. Specimen collection A maximum force in a unit of Newtons (N) was recorded The lower extremities of 10 adult (30–60 years) for each specimen. and 10 older adult (over 60 years) embalmed cadav- After that, the specimen was removed from the ers consisting of both genders were obtained for the clamps and then cross-sectionally dissected at the mus- dissection of the peroneus longus, including muscular culotendinous junction. Photographs of the cross-sec- part, musculotendinous junction and tendon portion tional area (CSA) of each specimen were obtained using from both sides of legs. Peroneus longus muscles from a digital camera. The UTHSCSA ImageTool for Windows the right legs were used for the mechanical testing, version 3.0 (University of Texas Health Science Centre in while the muscle from the left side were collected for San Antonio, USA) was used for measuring the CSA in histological and structural evaluations through light and a unit of square millimeter (mm2). The following formula scanning electron microscopic (SEM) studies. The cadav- was then used to calculate the tensile stress in a unit of ers had been donated to the Department of Anatomy, Megapascals (MPa): Tensile stress (MPa) = Force (N) / CSA Faculty of Medicine Siriraj Hospital, Mahidol University (mm2). The modulus of elasticity (Young’s modulus) was for use in anatomical studies. The embalming was done also calculated in a unit of MPa through the following for- within 24 h of death using a formalin that was mixed mula: modulus of elasticity (MPa) = Tensile stress (MPa)/ with adjuvants. The embalmed cadavers were kept and / [amount of length change (DL) / original length (L0)]. preserved for approximately 1 year prior used for analysis. Ethical approval for this study (COA no. Si074/2561) Histological study was granted by Institutional Review Board, Faculty of The musculotendinous parts of the peroneus longus Medicine Siriraj Hospital, Mahidol University. muscles were fixed in an immersion of 10% formalin, dehydrated in ascending grades of alcohol, cleared in Mechanical property testing xylene, infiltrated and embedded in paraffin. The paraffin After the peroneus longus muscle was dissected out block was cut by a rotatory microtome (Leica RM2035, from the cadaver, the highest point of musculotendinous Nussloch, Germany) with 6 µm thickness. Sections were

402 A. Niyomchan et al., Muscle changes in aging

Table 1. Mean ± standard deviation of the mechanical properties of peroneus longus under loading condition for the displacement of the musculotendinous junction for 1-mm length in adult and older adult Properties Adult (n = 10) Older adult (n = 10) P Tensile (N) 45.60 ± 17.69 28.13 ± 7.62* 0.010a Cross-sectional area [mm2] 106.80 ± 47.34 94.71 ± 14.35 0.450b Tensile stress (MPa) 0.47 ± 0.19 0.30 ± 0.10* 0.025b Modulus of elasticity (MPa) 32.64 ± 12.97 21.35 ± 6.89* 0.026b

*Significant difference (p < 0.05);a p-value was calculated by the Mann-Whitney test; bp-value was calculated by unpaired t-test stained with Masson’s trichrome and analysed under an longus in older adult were significantly lower than the Olympus Bx43 microscope. The pictures were taken by younger ones. a DP73 digital camera attached to the CellSens Standard software (Olympus Optical, Co. Ltd., Tokyo, Japan). Effect of aging on muscle-tendon histostructural changes Scanning electron microscopy sample preparation The SEM analysis of the peroneus longus muscle The muscular and tendon portion of the peroneus fibres revealed that there were long cylindrical forms longus muscles were cut into small cubes with a size with cross-striations of myofibrils on the muscle surface. of 0.5 × 0.5 × 0.2 cm. The specimens were washed in The fibres ran parallel to each other, and they were not a 0.1 M phosphate buffer solution at pH 7.4 and post- branched. However, there were no differences in the fixed with 1% osmium tetroxide for 1 h. After washing muscular morphologic features such as diameter and it with distilled water, the specimens were dehydrated shape between both groups of cadavers (Fig. 2). in a graded series of alcohol. The samples were then Considerable alterations in the collagen bundles of dried in the critical point dryer using liquid CO2 and the perimysium that were overlying muscle fibres were mounted on metal stubs. Finally, they were coated with noted in the older adult cadaver. Usually, perimysium gold-palladium in the sputter-coater and analysed in the comprised of the dense weave collagen bundles. How- JSM-6510LV (Jeol Ltd, Tokyo, Japan) SEM at 15 kV. ever, the peroneus longus muscle of the older adult cadaver contained thin and loosely packed perimysial Statistical analysis collagen bundles, which was comparatively different to Data analysis was performed using the GraphPad the thicker and more compact surface seen in the adult Prism programme for Windows version 6.05 (GraphPad cadaver’s perimysial collagen (Fig. 3). software Inc. CA, USA). All the data were presented as Within the tendon, the normal collagen fibres in the mean ± standard deviation. Prior to performing any peroneus longus tendon of the young cadavers showed between-group comparisons, all properties of the per- long, straight and tightly packed collagen fibres in oneus longus under loading condition were tested for a parallel arrangement along the long axis of the tendon normality using the Shapiro-Wilk test. Three parameters, (Fig. 4A, C). At higher magnification, most of the col- i.e., CSA, tensile stress, and modulus of elasticity, passed lagen fibres appeared to be more crimp and of a helical the normality test. Thus, an unpaired t-test was used to pattern (Fig. 4E). In an older adult cadaver, the tendon compare between the groups for these parameters. In consisted of disorientated and frayed collagen fibres contrast, the Mann-Whitney test was used for the tensile (Fig. 4B). In some tendons, disorganisation and frag- parameter, which did not distribute normally. A p-value mentation of collagen fibrils were observed (Fig. 4D). level of less than 0.05 was considered to be significant. They were flat, and the fracture surfaces appeared to look like tattered rope. Furthermore, the collagen fibrils RESULTS appeared straight without helical coils (Fig. 4F). Effect of aging on muscle-tendon mechanical Light microscopy of the muscle-tendon junction properties of the peroneus longus muscle showed remarkable Table 1 shows the results of the between-group changes in the aging process, particularly at the end of comparisons of all parameters. No significant differ- muscle inserted into the tendon via the myotendinous ence in the CSA of peroneus longus musculotendinous junction. In this region, the abundant tendon fibroblasts junction was found between the adult and older adult. were embedded in the collagen fibres, and the fibres of Interestingly, it was apparent from Table 1 that the ten- collagen were continuously anchored to the terminal ex- sile, tensile stress and elastic modulus of the peroneus tensions of muscle fibres, as shown in the adult cadaver

403 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. Scanning electron micrographs of the peroneus longus muscle fibres in an adult cadaver (A) and older adult cadaver (B). No differences were observed between younger and older cadavers in their muscle fibre morphology. The muscle fibres were arranged in a simple parallel fashion, with striations running perpendicular to the long axis of the muscle fibre.

Figure 3. Scanning electron micrographs of the perimysial collagen bundles of peroneus longus muscle in adult cadaver (A) and older adult cadaver (B). A looser arrangement of small perimysial collagen bands was present in the older adult cadaver when compared to the large and tight collagen bundles of perimysium found in the adult cadaver.

(Fig. 5A). With age, it was seen that fat accumulation findings of this study were as follows: 1) mechanical occurred in the myotendinous junction, and more fats properties including tensile, tensile stress and modulus were deposited closer to this region (Fig. 5B). of elasticity were lower, without a change in CSA in old adult cadavers when compared with the younger DISCUSSION subjects, 2) collagen fibre disorganisation appeared in The peroneus longus, which is comprised of muscular the aged tendon and perimysium, and 3) accumulation and tendinous tissues, helps to stabilise the arch and of fat was noted within the musculotendinous junction. assists in stabilising the ankle and foot [23]. It is predomi- Mechanical property testing in the present study nantly active during the stance phase of the gait [19, 21], demonstrated a decline in tensile, tensile stress, and which produces plantarflexed first ray of the foot and modulus of elasticity of the peroneus longus in old adult also acts as a secondary plantarflexor of the ankle [20]. cadavers. These changes were linked to alterations that Thus, deteriorated physical ability of the peroneus longus were sustained by the structure of the tendon portion can contribute to the development of foot deformity of this muscle. An abnormal architectures of collagen, [2, 29]. Age-related changes in the structural and me- the chief component in the tendon, were found with chanical properties of the muscle-tendon unit further aging. The tendon section of the peroneus longus in reflectthe alterations in functional competencies. Conse- the aged specimens appeared to be disorganised and quently this study investigated the potential architectural fragmented, along with having elongation without and mechanical changes that occur with age in the crimp. The reasons for these results were age-related peroneus longus musculotendinous unit, which could be alterations in the matrix contents, including an increase associated with the increased rate of foot problem cases in the degradation of collagens and proteoglycans via found clinically in older adults. We examined three differ- an activation of matrix metalloproteinases [5, 36] and ent parts of the peroneus longus, including the muscular a reduction in intra-molecular cross-links involving in part, the musculotendinous junction, and tendon por- crimp-like structure, a result of advanced glycation end- tion of young and old adult cadavers. The three main product accumulation in advancing age [1, 4]. 404 A. Niyomchan et al., Muscle changes in aging

Figure 4. Scanning electron micrographs of the peroneus longus tendon of an adult cadaver (A, C, E) and elderly cadaver (B, D, F). High-density collagen fibres in a parallel alignment (A), thick linear collagen bundles (C) and coiled collagen fibrils in the tendon (E) were shown in the adult cadaver, whereas the collagen found in the tendon of the elderly appeared disorganised (B), fragmented (D) and had straight collagen fibrils (F).

Figure 5. Light micrographs of the peroneus longus muscle at the myotendinous junction in the adult cadaver (A) and elderly cadaver (B). Fats intervened between the terminal end of each muscle fibre and the collagen of the tendon at the myotendinous junction, and a large amount of fat was found in this area in the older adult cadaver, a stark contrast when compared to the closely packed muscle-tendon interdigitation in the adult cadaver. Masson’s trichrome staining: blue collagen, red muscle fibre, and black cell nuclei.

405 Folia Morphol., 2019, Vol. 78, No. 2

Since each extracellular matrix layer (epimysium, per- ous studies have only reported an accumulation of fat imysium, endomysium) was primarily dominated by col- within the muscle and near the myotendinous junction lagen type I and III, which was similar to that in the tendon in an advanced age [22, 35]. An altered lipid metabo- [12]. The alterations of perimysial collagen arrangement lism with increasing age could be used to explain the reported in this study were thought to be due to the same fatty changes in aging muscles through an elevation of mechanism of having increased metalloproteinases regula- cholesterol and triglyceride levels [15], an increase in tory pathway that were used to break down the collagen lipogenesis and an impairment of lipolysis [31] as well fibrils,leading to small and loose arranged perimysial col- as changes in fat distribution from subcutaneous loca- lagen bundles in elderly cadavers. tions to intra-abdominal visceral depots and to harmful Although the morphology of the muscular part of ectopic sites [8]. Thus, an ectopic fat deposition within the peroneus longus remained stable during aging with the musculotendinous junction, which was found in this unchanged shape and size, fat infiltration in the mus- study, might impair muscle-tendon junction, especially culotendinous junction was found in the present study. the force transmission from muscle fibres to bone via Many previous studies have been reported to be age- the myotendinous junction and tendon. Moreover, both related, but do not appear to factor in sex-related loss the perimysium and endomysium play a crucial role in of muscle mass, as evidenced by a decrease in muscle the transmission of force via the action of collagen [16]. fibre CSA and fibre number determining in the various Therefore, an accumulation of fat in the musculoten- muscles from humans and cadaveric specimens [26, dinous junction and structural changes of collagen in 30]. However, our findings contradict these previous the tendon and perimysium during aging caused the reports. We found that there was no statistical difference peroneus longus muscle to alter its mechanical func- in CSA (measure of muscle fibre size) of the peroneus tion. Changes in structural-mechanical properties of the longus between young and old adult cadavers and the peroneus longus muscle might increase the risk for an ab- muscular architecture did not change substantially with normal foot shape and foot discomfort in elderly people. age. The differences between these findings may be Although the embalmed cadavers can be used for due to diversity in the range and distribution of muscle mechanical testing in both soft and hard tissues [32, fibre size in the aging, based on types of muscle fibres. 33]. A limitation of our study was the use of embalmed Atrophy of type II, fast-twitch or white muscle fibres cadavers that would have the effect of sample preserva- and hypertrophy in type I, slow-contracting or red fibres tion. In future, it would be interesting to confirm these can occur in the elderly [7]. Furthermore, a decline in findings using soft and fresh frozen cadavers. protein synthesis and an enhancement of overall protein degradation rates could also cause the atrophy of CONCLUSIONS muscles [17], while muscle hypertrophy takes place as a In summary, this study suggested age-related compensatory response to the reduction in muscle mass changes in the peroneus longus muscle, particularly that occurs with aging [7, 18]. Hepple et al. [7] indicated structural abnormalities in the collagen fibres and a variable alteration in the mean muscle fibre CSA of fatty infiltration in the musculotendinous junction. the gastrocnemius muscle. They exhibited a decrease in Consequently, the tensile, tensile stress, and modulus fibre size in the white region of the gastrocnemius, but of elasticity were decreased. These changes might increased fibre size in the red region of aged animals. also predict a decrease in the strength and poorer Similar to the animal study, muscle biopsies of the vastus functional outcomes of this muscle. lateralis from humans revealed a greater reduction in muscle fibre size of type II fibres than type I fibres [6]. Acknowledgements However, the current study did not find a difference in This work was partially supported by Chalermprakiat the mean CSA of the peroneus longus, which is a hybrid Grant, Faculty of Medicine Siriraj Hospital, Mahidol Uni- muscle fibres containing 62.5% slow-twitch type I fibres versity to A. Niyomchan and B. Panichareon. and 37.5% type II fast-twitch fibres, between young The authors gratefully acknowledge Mr. Kasem Koed- and old cadaveric specimens. Kano et al. [9] similarly puech, Mrs. Nusara Chomanee and Miss Hataichanok reported no statistical difference in the mean CSA of Pimpui for the skilful technical help of light and scanning soleus that consisted mainly of type I fibres between electron microscopy. young and old rats. The current study was the first finding to - dem References onstrate fat infiltration within the musculotendinous 1. Avery NC, Bailey AJ. Enzymic and non-enzymic cross-linking junction of the old age muscle tissues, as of the previ- mechanisms in relation to turnover of collagen: relevance

406 A. Niyomchan et al., Muscle changes in aging

to aging and exercise. Scand J Med Sci Sports. 2005; 15(4): 19. Louwerens JW, van Linge B, de Klerk LW, et al. Peroneus longus 231–240, doi: 10.1111/j.1600-0838.2005.00464.x, indexed in and tibialis anterior muscle activity in the stance phase. A quan- Pubmed: 15998340. tified electromyographic study of 10 controls and 25 patients 2. Brandes CB, Smith RW. Characterization of patients with primary with chronic ankle instability. Acta Orthop Scand. 1995; 66(6): peroneus longus tendinopathy: a review of twenty-two cases. 517–523, indexed in Pubmed: 8553819. Foot Ankle Int. 2000; 21(6): 462–468, doi: 10.1177/107110070 20. Lui TH, Hau WW. Peroneus longus tendoscopy at the sole. Arthro- 002100602, indexed in Pubmed: 10884103. sc Tech. 2018; 7(5): e479–e483, doi: 10.1016/j.eats.2017.12.001, 3. Bubra PS, Keighley G, Rateesh S, et al. Posterior tibial tendon indexed in Pubmed: 29868422. dysfunction: an overlooked cause of foot deformity. J Family Med 21. Lui TH. Stabilization of medial longitudinal foot arch by peroneus Prim Care. 2015; 4(1): 26–29, doi: 10.4103/2249-4863.152245, longus transfer. Foot (Edinb). 2016; 27: 22–26, doi: 10.1016/j. indexed in Pubmed: 25810985. foot.2016.03.001, indexed in Pubmed: 27058034. 4. Couppé C, Hansen P, Kongsgaard M, et al. Mechanical properties 22. Marcus RL, Addison O, Kidde JP, et al. Skeletal muscle fat infiltra- and collagen cross-linking of the patellar tendon in old and young tion: impact of age, inactivity, and exercise. J Nutr Health Aging. men. J Appl Physiol (1985). 2009; 107(3): 880–886, doi: 10.1152/ 2010; 14(5): 362–366, indexed in Pubmed: 20424803. japplphysiol.00291.2009, indexed in Pubmed: 19556458. 23. Mendicino RW, Orsini RC, Whitman SE, et al. Fibular groove 5. Dudhia J, Scott CM, Draper ERC, et al. Aging enhances deepening for recurrent peroneal subluxation. J Foot Ankle Surg. a mechanically-induced reduction in tendon strength by an active 2001; 40(4): 252–263, indexed in Pubmed: 11924687. process involving matrix metalloproteinase activity. Aging Cell. 24. Menz HB. Biomechanics of the Ageing Foot and An- 2007; 6(4): 547–556, doi: 10.1111/j.1474-9726.2007.00307.x, kle: A Mini-Review. Gerontology. 2015; 61(4): 381–388, indexed in Pubmed: 17578513. doi: 10.1159/000368357, indexed in Pubmed: 25402236. 6. Grimby G, Danneskiold-Samsøe B, Hvid K, et al. Morphology and en- 25. Mizel MS, Temple HT, Scranton PE, et al. Role of the peroneal ten- zymatic capacity in arm and leg muscles in 78-81 year old men and dons in the production of the deformed foot with posterior tibial women. Acta Physiol Scand. 1982; 115(1): 125–134, doi: 10.1111/ tendon deficiency. Foot Ankle Int. 1999; 20(5): 285–289, doi: j.1748-1716.1982.tb07054.x, indexed in Pubmed: 7136797. 10.1177/107110079902000502, indexed in Pubmed: 10353763. 7. Hepple RT, Ross KD, Rempfer AB. Fiber atrophy and hypertrophy 26. Narici MV, Maganaris CN, Reeves ND, et al. Effect of aging on in skeletal muscles of late middle-aged Fischer 344 x Brown human muscle architecture. J Appl Physiol (1985). 2003; 95(6): Norway F1-hybrid rats. J Gerontol A Biol Sci Med Sci. 2004; 59(2): 2229–2234, doi: 10.1152/japplphysiol.00433.2003, indexed in 108–117, indexed in Pubmed: 14999023. Pubmed: 12844499. 8. Hughes VA, Roubenoff R, Wood M, et al. Anthropometric assess- 27. Narici MV, Maganaris CN. Plasticity of the muscle-tendon com- ment of 10-y changes in body composition in the elderly. Am plex with disuse and aging. Exerc Sport Sci Rev. 2007; 35(3): J Clin Nutr. 2004; 80(2): 475–482, doi: 10.1093/ajcn/80.2.475, 126–134, doi: 10.1097/jes.0b013e3180a030ec, indexed in indexed in Pubmed: 15277173. Pubmed: 17620931. 9. Kano Y, Shimegi S, Furukawa H, et al. Effects of aging on capil- 28. Onambele GL, Narici MV, Maganaris CN. Calf muscle-tendon lary number and luminal size in rat soleus and plantaris muscles. properties and postural balance in old age. J Appl Physiol J Gerontol A Biol Sci Med Sci. 2002; 57(12): B422–B427, indexed (1985). 2006; 100(6): 2048–2056, doi: 10.1152/japplphysi- in Pubmed: 12456732. ol.01442.2005, indexed in Pubmed: 16455811. 10. Keller K, Engelhardt M. Strength and muscle mass loss with aging 29. Perera AM, Mason L, Stephens MM. The pathogenesis of hal- process. Age and strength loss. Muscles Ligaments Tendons J. lux valgus. J Bone Joint Surg Am. 2011; 93(17): 1650–1661, 2013; 3(4): 346–350, indexed in Pubmed: 24596700. doi: 10.2106/JBJS.H.01630, indexed in Pubmed: 21915581. 11. Kitaoka HB, Luo ZP, An KN. Effect of the posterior tibial tendon on 30. Ruggiero M, Cless D, Infantolino B. Upper and Lower Limb the arch of the foot during simulated weightbearing: biomechani- Muscle Architecture of a 104 Year-Old Cadaver. PLoS One. 2016; cal analysis. Foot Ankle Int. 1997; 18(1): 43–46, doi: 10.1177/107 11(12): e0162963, doi: 10.1371/journal.pone.0162963, indexed 110079701800109, indexed in Pubmed: 9013114. in Pubmed: 28033339. 12. Kjaer M. Role of extracellular matrix in adaptation of tendon 31. Seah NE, de Magalhaes Filho CD, Petrashen AP, et al. Autophagy- and skeletal muscle to mechanical loading. Physiol Rev. 2004; mediated longevity is modulated by lipoprotein biogenesis. 84(2): 649–698, doi: 10.1152/physrev.00031.2003, indexed in Autophagy. 2016; 12(2): 261–272, doi: 10.1080/15548627.20 Pubmed: 15044685. 15.1127464, indexed in Pubmed: 26671266. 13. Kohls-Gatzoulis J, Angel JC, Singh D, et al. Tibialis posterior 32. Ugbolue UC, Gislason MK, Carter M, et al. Tensile properties of dysfunction: a common and treatable cause of adult acquired the transverse carpal ligament and carpal tunnel complex. Clin flatfoot. BMJ. 2004; 329(7478): 1328–1333, doi: 10.1136/ Biomech (Bristol, Avon). 2015; 30(7): 649–656, doi: 10.1016/j. bmj.329.7478.1328, indexed in Pubmed: 15576744. clinbiomech.2015.05.010, indexed in Pubmed: 26048241. 14. Kokubo T, Hashimoto T, Nagura T, et al. Effect of the posterior 33. van Haaren EH, van der Zwaard BC, van der Veen AJ, et al. tibial and peroneal longus on the mechanical properties of the Effect of long-term preservation on the mechanical proper- foot arch. Foot Ankle Int. 2012; 33(4): 320–325, doi: 10.3113/ ties of cortical bone in goats. Acta Orthop. 2008; 79(5): FAI.2012.0320, indexed in Pubmed: 22735204. 708–716, doi: 10.1080/17453670810016759, indexed in Pub- 15. Kolovou G, Katsiki N, Pavlidis A, et al. Ageing mechanisms and med: 18839380. associated lipid changes. Curr Vasc Pharmacol. 2014; 12(5): 34. Watanabe K, Kitaoka HB, Fujii T, et al. Posterior tibial tendon dys- 682–689, indexed in Pubmed: 24350931. function and flatfoot: analysis with simulated walking. Gait Pos- 16. Kurose T, Asai Y, Mori E, et al. Distribution and change of collagen ture. 2013; 37(2): 264–268, doi: 10.1016/j.gaitpost.2012.07.015, types I and III and elastin in developing leg muscle in rat. Hiroshi- indexed in Pubmed: 22939754. ma J Med Sci. 2006; 55(3): 85–91, indexed in Pubmed: 16995494. 35. Yoshida Y, Marcus RL, Lastayo PC. Intramuscular adipose tissue 17. Lecker SH, Goldberg AL, Mitch WE. Protein degradation by the and central activation in older adults. Muscle Nerve. 2012; ubiquitin-proteasome pathway in normal and disease states. 46(5): 813–816, doi: 10.1002/mus.23506, indexed in Pub- J Am Soc Nephrol. 2006; 17(7): 1807–1819, doi: 10.1681/ med: 23055318. ASN.2006010083, indexed in Pubmed: 16738015. 36. Yu TY, Pang JHS, Wu KPH, et al. Aging is associated with increased 18. Lexell J, Taylor CC. Variability in muscle fibre areas in whole human activities of matrix metalloproteinase-2 and -9 in tenocytes. BMC quadriceps muscle: effects of increasing age. J Anat. 1991; 174: Musculoskelet Disord. 2013; 14: 2, doi: 10.1186/1471-2474-14-2, 239–249, indexed in Pubmed: 2032938. indexed in Pubmed: 23281803.

Differences in foetal topographical anatomy between insertion sites of the iliopsoas and gluteus medius muscles into the proximal femur: a consideration of femoral torsion P. Zhao1, Z.W. Jin1, J.H. Kim2, H. Abe3, G. Murakami4, J.F. Rodríguez-Vázquez5

1Department of Anatomy, Wuxi School of Medicine, Jiangnan University, Wuxi, China 2Department of Anatomy, Chonbuk National University Medical School, Jeonju, Republic of Korea 3Department of Anatomy, Akita University School of Medicine, Akita, Japan 4Division of Internal Medicine, Asuka Hospital, Sapporo, Japan 5Department of Anatomy and Human Embryology, Institute of Embryology, Faculty of Medicine, Complutense University, Madrid, Spain

[Received: 3 July 2018; Accepted: 24 August 2018]

Background: Prenatal twisting of the femoral neck seems to result in an angle of anteversion or torsion, but the underlying process has not been elucidated. Materials and methods: This study analysed sagittal, frontal and horizontal sections of 34 embryo and foetal specimens of gestational age (GA) 6–16 weeks (crown-rump length 21–130 mm). At GA 6–7 weeks, the iliopsoas (IP) and gluteus medius (GME) muscles were inserted into the anterior and posterior aspects of the femur, respectively, allowing both insertions to be viewed in a single sagittal section. Results: At GA 8 weeks, the greater trochanter and the femoral neck angle became evident, and the GME tendon was inserted into the upper tip of the trochanter. At GA 9 weeks, the location of IP insertion was to the medial side of the GME insertion. After 9 weeks, the IP insertion consisted of a wavy, tendinous part of the psoas muscle and another part of the iliacus muscle, with many fibresof the latter muscle attached to the joint capsule. After GA 12 weeks, the IP was inserted into the anteromedial side of the greater trochanter, while the aponeurotic insertion of the GME wrapped around the trochanter. At GA 15–16 weeks, a deep flexion at the hip joint caused an alteration in the relative heights of the lesser and greater trochanter, with the former migrating from the inferior to the slightly superior side. Conclusions: These findings indicate that twisting of the femoral neck started at GA 8–9 weeks. (Folia Morphol 2019; 78, 2: 408–418)

Key words: torsion, femur, neck angle, hip joint rotation, trochanter, iliopsoas muscle, gluteus medius muscle, human foetus

Address for correspondence: Dr. Z.W. Jin, Department of Anatomy, Wuxi School of Medicine, Jiangnan University, Wuxi, Jiangsu, 214122, China, e-mail: [email protected]

408 P. Zhao et al., Foetal hip joint rotation

INTRODUCTION section. The topographical relationship between the Torsion or twisting of the femur is defined mor- greater and lesser trochanters seems therefore to be phologically as a medial (internal) tilting of the left- altered after GA 9 weeks. right axis of the femoral condyles at an angle of Previous studies of the foetal development of 10–15 degrees (angle of anteversion) relative to the femoral torsion have utilised hard tissue specimens long axis of the femoral neck, with some variations from human foetuses [7, 12]. However, hard tissue among human populations [12]. Thus, the effect of specimens are difficult to obtain from early foetuses femoral torsion is identical to that resulting from an and they never contain muscle insertions. The present internal rotation at the hip joint. Evaluations of the study was designed to assess the topographic altera- effects of torsion on the segmental configuration of tions in the sites of insertion of the IP and GME ten- the cutaneous nerve supply in the lower extremities dons into the proximal femur. Although torsion, twist- have shown that the segments or dermatomes are ing and rotation of the femur are three-dimensional not arranged transversely but obliquely or even lon- events, we hypothesized that systematic evaluations gitudinally relative to the long axis of the extremity. of a set of sagittal, frontal and horizontal sections These findings suggested that an initial transverse would provide critical information for morphologic arrangement of the dermatome in embryos is twisted evaluations. internally [3, 9]. It remains unclear, however, whether torsion of the foetal femur results in twisting of the MATERIALS AND METHODS femoral skin, as little information is available on the This study was performed in accordance with initial development of torsion. the provisions of the Declaration of Helsinki 1995 In contrast to the internal rotation, the lower (as revised in 2013) and was approved by the ethics extremities of human foetuses of gestational age committee of the university. Serial sections of the (GA) > 15 weeks are highly flexed, abducted and hip and thigh from 34 embryos and foetuses of GA laterally (externally) rotated at the hip joint because 6–16 weeks (crown-rump length 21–130 mm) were of the increased size of the abdomen sandwiched evaluated, including 6 specimens each of GA 6–7, 8, by the two growing thighs [4]. Moreover, crossing 9–10, and 12 weeks; and 10 specimens of GA 15–16 of foetal legs with flexed knees (i.e., strong external weeks. The topographical relationships between the rotation) can save intrauterine space and result in greater and lesser trochanters and between the sites easier delivery. Meticulous measurement of the lower of insertion of the IP and GME muscles into the proxi- extremities of 500 children of various ages showed mal femur were compared in sagittal sections of 10 that this intrauterine posture usually moulds the fe- specimens, frontal sections of 6, horizontal sections mur by external rotation, with this moulding effect of 8 and tilted horizontal sections (i.e., intermediate usually resolving spontaneously during infancy [11]. planes between the horizontal and frontal sections) Therefore, external rotation of the femur is more likely of 10. Although horizontal sections may be most to be maintained from mid gestation to birth, with suitable for assessing the anteroposterior and medi- any prenatal torsion or internal rotation likely masked olateral relationships between the greater and lesser by the forced external rotation at the hip joint after trochanters, these sections had the disadvantage that mid-gestation. they could not show the entire shapes of the IP and In adults, the greater trochanter of the femur is GME muscles. located superior and lateral to and on the posterior All sections were from the large collection kept at side of the lesser trochanter. A band-like tendon from university, with the specimens derived from miscar- the gluteus medius (GME) muscle is inserted into the riages and ectopic pregnancies at the Department tip of the greater trochanter, and the iliopsoas (IP) of Obstetrics of the University. All specimens were tendon is inserted into the lesser trochanter. In human fixed with 10% formalin solution for 1–2 weeks. Most foetuses of GA 6–8 weeks, however, the IP is inserted sections were stained with haematoxylin and eosin immediately anterior to the site of GME insertion [10]. (H&E), with other sections stained with azan, orange Thus, these two insertions and future greater and G or silver stain. A Nikon Eclipse 80 camera was used lesser trochanters can be viewed in a single sagittal to visualise and photograph the sections.

409 Folia Morphol., 2019, Vol. 78, No. 2

RESULTS The IP insertion was composed of a tendinous part Examination of 6 specimens of GA 6–7 weeks from the psoas and a muscular part from the iliacus showed that the IP was inserted into the anterior (Fig. 2B). Muscle fibres of the iliacus were attached aspect of the femur and the GME into the posterior to and inserted into a loose and thick perichondri- aspect. Both insertions were visible on a single sagittal um of the femur near the future lesser trochanter or nearby section (Fig. 1A–C). Moreover, the supero- (Fig. 2A, B). The neck angle reached 120 degrees inferior levels of these muscle insertions were almost and the femoral head was 0.5–1.0 mm medial and equal. The greater trochanter consistently started to posterior to the greater trochanter (cf., diameter of develop at the attachment of the GME tendon (Fig. 1A), the head: 0.7–1.0 mm). However, the femoral head while a cartilaginous lesser trochanter was not pre- was still located at almost the same supero-inferior sent. The head, neck and shaft of the femur were level as the greater trochanter. arranged in an almost straight line without a neck Assessment of the 6 specimens of GA 12 weeks angle. Thus, the proximal femur appeared to push the showed that the femoral head was large (diameter acetabulum in an upward direction. However, a plate 1.8–2.5 mm) and pushed the tilted plate of the ilium of the ilium was tilted at various angles, depending superomedially (Figs. 3, 4). As the femoral neck in- on the specimens. At the location of IP insertion, the creased in length, the head was 2.5–3.5 mm medial psoas muscle provided a thick tendon associated with to the greater trochanter, but at the same supero- muscle fibres of the iliacus part of the IP. inferior level (Figs. 3F; 4C, F), possibly because of the Examination of the 6 specimens of GA 8 weeks excess neck angle of almost 90 degrees. The GME showed that the anteroposterior relationship of the was inserted aponeurotically, wrapping around the IP and GME insertions was generally maintained, but lateral aspect of the trochanter to reach its distal end. that both insertions could not be included in a single The GME aponeurosis was fused with the insertion of sagittal section (Fig. 1D–F). A gap of 0.2–0.4 mm in the gluteus minimus muscle (Figs. 3D, 4C). Muscular the anteroposterior axis was observed between the insertion of the GME was still evident (Fig. 4D), and sites of insertion (cf., diameter of the femoral head: the IP tendon was wavy near the femur (Figs. 3E, 4G). 0.5–0.7 mm), making the site of IP insertion slightly Muscle fibres of the iliacus were attached to the fe- medial to the site of GME insertion. The tip of the mur (Figs. 3D, 4G) and inserted into the joint capsule greater trochanter was consistently located on the (Fig. 4D). The obturator externus tendon ran across superior side of the IP insertion. A plate of the ilium a large intertrochanteric fossa (Fig. 4B). After 12 weeks, was erect, with an almost straight line from the ilium the near frontal sections did not contain the head, neck to the femoral shaft (Fig. 1E). The greater trochanter and proximal shaft of the femur because the femoral appeared as a cartilaginous upward protrusion, with head was directed posteriorly by torsion (Figs. 4, 5). a developing neck angle (Fig. 1F). GME fibres con- Examination of the 10 specimens of GA 15–16 verged to form a thick tendon reaching the tip of the weeks showed that, because of deep flexion at the greater trochanter. The psoas part of the IP converged hip joint, the location of the lesser trochanter rela- to form a long tendon that wrapped around the femo- tive to the greater trochanter had changed, from the ral head (Fig. 1D). The iliacus part of the IP also con- inferior to the superior side (Figs. 5, 6). The lesser verged into the psoas tendon, but the muscle fibres trochanter became evident as a cartilaginous projec- of the former ran along the tendon toward the femur. tion (Figs. 5A, 6D), whereas the size of the femoral The supero-inferior levels of the IP and GME insertions head (diameter 2.2–3.0 mm) was relatively stable. were almost equal, as were the heights of the femoral The IP tendon was wavy near the femur (Figs. 5A, C; head and the upper tip of the greater trochanter. 6D). Iliacus muscle fibres were attached to the femur Evaluation of the 6 specimens of GA 9–10 weeks (Figs. 5A, B; 6D) and inserted into the joint capsule showed that the location of IP insertion had changed (Fig. 5D). The GME muscle was inserted into the su- to the inferomedial side of the greater trochanter, and perficial side of the aponeurotic insertion in 4 of the that the lesser trochanter appeared as a cartilaginous 10 specimens of GA 15–16 weeks (Fig. 5F). Thus, the tuberosity (Fig. 2). In contrast to earlier specimens, GME insertion consisted of (1) a muscular part at- the GME tendon became unclear and the muscle tached to the trochanter with an excess muscle mass fibres appeared to be directly inserted into the supe- or a “slack” reaching the anterior side of the greater rolateral aspect of the greater trochanter (Fig. 2D). trochanter (Figs. 3A, 6A) and (2) an aponeurotic part

410 P. Zhao et al., Foetal hip joint rotation

Figure 1. A–F. Insertion of an iliopsoas muscle (IP) into the anterior side of the gluteus medius muscle (GME) insertion in foetuses of gestational age (GA) 6–8 weeks. Sagittal sections of specimens of crown-rump length 20.5 mm (GA 6 weeks; panels A–C) and crown-rump length 35 mm (GA 8 weeks; panels D–F). The left-hand side of each panel corresponds to the anterior side of the pelvis. Panels A and D (C and F) show the most medial (lateral) sites in the specimens. All intervals between panels are 0.1 mm. In the smaller specimen, a single sagittal section (panel B) contains both an IP tendon and another insertion of the GME. However, in the larger specimen, the IP tendon wrapped around the femoral head (panel D) 0.2 mm medial to the site of GME insertion (panel F). The head and neck of the femur are slightly tilted with respect to the shaft (panel F). In both specimens, the GME insertion appeared to be tendinous (panels C and F) and the femur head is small and directed upward. Panels A–C (and D–E) were prepared at the same magnification (scale bars in panels A and D: 1 mm). Common abbreviations: AM — adductor magnus muscle; F — femur; FN — femoral nerve; GME — gluteus medius muscle; GMI — gluteus minimus muscle; GMX — gluteus maximus muscle; GT — greater trochanter; IL — ilium; IP — iliopsoas muscle; LT — lesser trochanter; OE — obturator externus muscle; OI — obturator internus muscle; RF — rectus femoris muscle; TFL — tensor fasciae latae muscle; SA — sartorius muscle; SN — sciatic nerve; VI — vastus intermedius muscle; VL — vastus lateralis muscle.

411 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. A–D. A slightly wavy tendon of the iliopsoas muscle (IP) on the inferomedial side of the gluteus medius muscle (GME). Tilted horizontal sections of a specimen of crown-rump length 40 mm (gestational age 9 weeks). The left-hand side of each panel corresponds to the medial side of the pelvis. Panels A and D show the most inferior and superior sites in the figure, respectively. Intervals between panels are 0.1 mm (A–B), 0.3 mm (B–C) and 0.5 mm (C–D). The insertion of the IP is composed of a tendinous part from the psoas and a muscular part from the iliacus (panels A and B). The GME does not have a definite tendon but a muscular insertion along the superior aspect of the greater trochanter (GT in panel D). The femoral head is pointed posteromedially. All panels were prepared at the same magnification (scale bar in panel A: 1 mm). Abbreviations — see Figure 1.

wrapped around the entire surface of the greater DISCUSSION trochanter. However, a lower magnification view The present study showed that the drastic changes (Fig. 6B) provided the erroneous impression that the in morphology of the proximal femur in foetuses GME was simply inserted into a tip of the trochanter, altered the topographical relationship between inser- as in samples of GA 6–7 weeks. In addition, as seen tions of the IP and GME. Figure 7 summarises these in the iliacus part of the IP, the gluteus minimus mus- observations. At the earliest stage, GA 6–7 weeks, cle belly was attached to a large area of the femur a very short neck along a straight line from the shaft (Fig. 5C, D). Consequently, after 12 weeks, the IP to the head of the femur provided no neck angle or insertion or lesser trochanter was shifted to the an- torsion. The IP and GME were inserted into the fe- teromedial side of the greater trochanter. mur at the same supero-inferior level (Fig. 7A). This

412 P. Zhao et al., Foetal hip joint rotation

Figure 3. A–F. A slightly wavy tendon of the iliopsoas muscle on the almost medial side of the gluteus medius muscle (GME). Horizontal sections of a specimen of crown-rump length 82 mm (gestational age 12 weeks). The upper side of each panel corresponds to the posterior side of the pelvis. Panels A–C display topographical anatomy at lower magnification, while panels D–F are higher magnifications of the squares in panels A–C, respectively. Panels A and D (C and F) show the most inferior (superior) sites in the figure. Intervals between panels are 0.3 mm (D–E) and 0.6 mm (E–F). The psoas tendon exhibits a slightly wavy course (panel E), while the iliacus muscle fibres (asterisk in panel D) are attached to the femur. The GME does not have a definite tendon, but the muscle fibres are attached to the posterolateral aspect of the greater trochanter. The arrow in panel D indicates excess muscle fibres of the gluteus medius reaching the anterior side of the trochanter. The femoral head is directed posteromedially. Panels A–C (and D–F) were prepared at the same magnification. Scale bar: 3 mm in panel A; 1 mm in panel D. Abbreviations — see Figure 1.

413 Folia Morphol., 2019, Vol. 78, No. 2

Figure 4. A–G. A slightly wavy tendon of the iliopsoas muscle on the anteromedial side of the gluteus medius muscle (GME). Frontal sections of a specimen of crown-rump length 85 mm (gestational age 12 weeks). The upper side of each panel corresponds to the superior side of the pelvis. Panels A and F display topographical anatomy at lower magnification, while panels B and G are higher magnifications of the squares in panels A and F, respectively. Panels A and B (F and G) show the most posterior (anterior) sites in the figure. Panels C–E exhibit planes between panels A and F. Intervals between panels are 0.4 mm (B–C), 0.7 mm (C–D), 0.5 mm (D–E) and 0.7 mm (E–G). Fibres of the GME converge to a tendon wrapping around the greater trochanter (GT in panel B). The iliopsoas tendon exhibits a slightly wavy course (panel G) in association with muscle fibres attaching to the joint capsule (arrowheads in panel D). The femoral head is directed medially. Scale bar: 3 mm in panels A and F; 1 mm in the other panels. Abbreviations — see Figure 1.

morphology was similar to that of infraspinatus and mon to muscles surrounding a multiaxial spheroidal subscapularis muscle insertions, which wrap around joint. After GA 8 weeks, the femoral neck became the humerus head during early development [1]. elongated and the angle of inclination increased, These findings suggest that this configuration is com- resulting in an excess neck angle of almost 90 de-

414 P. Zhao et al., Foetal hip joint rotation

Figure 5. A–F. A lesser trochanter receiving a wavy tendon of the iliopsoas. Frontal sections of a specimen of crown-rump length 110 mm (gestational age 15 weeks). The right-hand side of each panel corresponds to the medial side of the pelvis. Panels A and F represent the most anterior and posterior sites in the figure, respectively. Intervals between panels are 0.2 mm (A–B, B–C), 0.6 mm (C–D), 0.8 mm (D–E) and 1.0 mm (E–F). Panel B contains an intramuscular tendon of the iliacus and another tendon (arrow) from the psoas; these two tendons of the iliopsoas join in panel C. These iliopsoas tendons are wavy (panels A–C), while the tendon of the gluteus medius muscle is straight (panel F). The greater trochanter is 2.6 mm posterior and 3–4 mm lateral to the lesser trochanter. The iliacus muscle fibres are attached to the femur near the lesser trochanter (panels A and B). All panels were prepared at the same magnification (scale bar in panel A: 1 mm). Abbreviations — see Figure 1.

grees (cf., 120 degrees in adults). This may facilitate torsion should occur at the femoral shaft distal to movements of the hip joint, even in utero. Torsion of the trochanters. the femur seemed to occur at GA 8–9 weeks (Fig. 7B) At GA 12–16 weeks, the IP tendon loosened, pro- and may subsequently increase in response to a physi- viding a wavy course, and the straight GME inser- ological external rotation at the hip joint to maintain tion (muscular and tendinous) wrapped around the a crossing of foetal legs with flexed knees. Because curved, lateral aspect of the greater trochanter. These the relative positions of the two trochanters changed, alterations may have been due to internal rotation

415 Folia Morphol., 2019, Vol. 78, No. 2

Figure 6. A–F. A cartilaginous lesser trochanter on the superior side of the gluteus medius tendon. Horizontal sections of a specimen of crown-rump length 130 mm (gestational age 16 weeks). The upper side of each panel corresponds to the posterior side of the pelvis. Panels A-C display topographical anatomy at lower magnification, while panels D–F are higher magnifications of squares in panels A–C, respectively. Panels A–C were prepared at the same magnification, as were panels D–F. Panels A and C represent the most superior and inferior sites in the figure, respectively. Intervals between panels are 1.5 mm (A–B) and 0.5 mm (B–C). A thick and wavy tendon of the psoas is inserted into the lesser trochanter (LT in panel D). Parts of iliacus muscle fibres converge to the tendon, while the other parts (asterisk in panel D) attach to the joint capsule. The gluteus medius muscle also carries a thick and straight tendon inserting into and wrapping around the lateral aspect of the greater trochanter (GT in panels E and F). The asterisks in panels A and D indicate excess fibres of the gluteus medius muscle reaching the supero-anterior side of the trochanter. The inferiormost part of the gluteus medius is located below. Scale bar: 3 mm in panel A; 1 mm in panel D. Abbreviations — see Figure 1.

416 P. Zhao et al., Foetal hip joint rotation

Figure 7. A–C. Summary of the study findings. Posterior views. Panel A shows the initial morphology of the proximal femur and its associated muscle insertions. The iliopsoas (IP) and gluteus medius (GME) muscles are inserted to the anterior and posterior aspects of the femur, respectively. Panel B exhibits an intermediate morphology. The greater trochanter (GT) develops first, followed by the development of the lesser trochanter (LT). The IP and GME insertions are shifted to the medial and lateral sides of the femur, respectively. The psoas tendon is wavy near the insertion, while the GME insertion is aponeurotic, wrapping around the trochanter (arrowheads). Panel C shows the almost final morphology, but the GME insertion provides an excess mass of muscles (arrow) at the supero-anterior side of the trochanter; F — femur.

caused by the developing torsion. Likewise, an initial 36 weeks [12], the present study found a drastic and tendon-like insertion of the GME changed at GA critical change in morphology before GA 12 weeks. 6–9 weeks to a thick muscular insertion, along with an The present study also showed that muscle fibres aponeurosis wrapping around the greater trochanter. of the iliacus ran along the psoas tendon and then The GME muscle fibres seemed to be rolled along the attached to the femur and were inserted into the trochanter by twisting of the femur, resulting not only hip joint capsule. Muscle fibres attaching to a joint in the attachment of muscle fibres to the trochanter capsule have been observed in many joints of human but also an excess mass or “slack” of muscles reach- adults [5]. However, in adults, IP muscle fibres are un- ing a site immediately below the trochanter. These likely to attach to the capsule because the IP tendon morphologies likely indicate the developing torsion reinforces the weak anterior aspect of the joint as of the femur. After GA 12 weeks, the IP insertion or a dynamic stabiliser, acting as a supporting structure the lesser trochanter was located on the anteromedial strengthened by muscle contractions, comparable side of the greater trochanter, a morphology that to the relationship between the rotator cuff and the persists into adulthood (Fig. 7C). glenohumeral joint [13]. The subscapularis muscle In adults, the angle of anteversion is 10–15 de- tendon, the strongest rotator cuff tendon, showed grees, equal to the range of twisting or torsion at GA a similar morphology [8], with this tendon in late 8–9 weeks. However, the greater (or lesser) trochanter stage foetuses composed of abundant fragments may twist 60–90 degrees from the posterior (or ante- of muscle fibres [2]. Moreover, flexor muscle fibres rior) side to the lateral (or anteromedial) side of the attached to the capsule in foetal elbow joints were femur shaft. This estimate of 60–90 degrees was ap- reported to degenerate after birth [5]. Foetal iliacus proximate, because topographical relationships may muscle fibres attaching to and along the capsule have altered in the pelvis, including the sacroiliac joint, seemed to degenerate from heavy mechanical stress as well as in the proximal femur. However, the discrep- when the tendon pushed the femoral head in the ancy between 10–15 and 60–90 degrees suggested posterior direction for stabilisation. that migration of muscle insertions along the bony “Proof of torsion” has been provided by a spi- surface may have led to an over-estimate of actual ral fibre arrangement in the ischiofemoral ligament twisting of the femur. Although a study of 140 foetuses covering the anterior aspect of the adult hip joint at GA 12–24 weeks, but decreases slightly after GA [6]. We found, however, that the joint capsule, in-

417 Folia Morphol., 2019, Vol. 78, No. 2

cluding the supporting ligaments, was too thin to tematic description of the branches of the human femoral identify the expected spiral configuration. Moreover, nerve. Okajimas Folia Anat Jpn. 1992; 69(1): 35–74, indexed in Pubmed: 1620529. even if present in foetuses, spiral collagenous bun- 4. Ishizawa A, Hayashi S, Nasu H, et al. An artery accompanying dles were not maintained until adulthood but were the sciatic nerve (arteria comitans nervi ischiadici) and the reconstructed depending on mechanical demands position of the hip joint: a comparative histological study after birth. Likewise, an oblique or longitudinal con- using chick, mouse, and human foetal specimens. Folia Mor- figuration of dermatomes in the lower extremities phol. 2013; 72(1): 41–50, indexed in Pubmed: 23749710. 5. Jin ZW, Jin Y, Yamamoto M, et al. Oblique cord (chorda obli- was unlikely to be connected with the development qua) of the forearm and muscle-associated fibrous tissues at of torsion because the courses of cutaneous nerve and around the elbow joint: a study of human foetal speci- were apparently established before GA 8–9 weeks. mens. Folia Morphol. 2016; 75(4): 493–502, doi: 10.5603/ Finally, the clinical relevance of prenatal torsion is dif- FM.a2016.0019, indexed in Pubmed: 27830875. 6. Kawashima T, Sasaki H. Reasonable classical concepts in ficult to determine, because little or no information human lower limb anatomy from the viewpoint of the primi- is available about pathologies of the proximal femur tive persistent sciatic artery and twisting human lower limb. in congenital dysplasia of the hip. This abnormality is Okajimas Folia Anat Jpn. 2010; 87(3): 141–149, erratum in usually attributed to a poorly developed acetabulum. 2011. 87 :195, indexed in Pubmed: 21174944. 7. Masłoń A, Sibiński M, Topol M, et al. Development of However, because the shape and depth of the socket human hip joint in the second and the third trimester are determined by the ball, a failure of foetal torsion of pregnancy; a cadaveric study. BMC Dev Biol. 2013; with an abnormal muscle insertion may be associated 13: 19, doi: 10.1186/1471-213X-13-19, indexed in Pub- with some types of dysplasia of the hip. med: 23651510. 8. Muraki T, Aoki M, Uchiyama E, et al. A cadaveric study of strain on the subscapularis muscle. Arch Phys Med Rehabil. Acknowledgements 2007; 88(7): 941–946, doi: 10.1016/j.apmr.2007.04.003, We are grateful to Dr. Daisuke Suzuki for his me- indexed in Pubmed: 17601478. ticulous work in taking photos in Madrid. This work 9. Ogoshi A, Kawashima T, Hanaoka Y, et al. Twisting of the supported by the National Natural Science Founda- human lower extremity changes the relationship of bones, ligaments, muscles, and nerves. J Tokyo Women’s Med Univ. tion of China (No. 31500968 by Peng Zhao and No. 2001; 71: 773–786. 81460471 by Zhe Wu Jin) and supported in part by 10. Shiraishi Y, Jin ZW, Mitomo K, et al. Foetal development a Grant-in-Aid for Scientific Research (JSPS KAKENHI of the human gluteus maximus muscle with special refer- No. 16K08435) from the Ministry of Education, Cul- ence to its fascial insertion. Folia Morphol. 2018; 77(1): 144–150, doi: 10.5603/FM.a2017.0060, indexed in Pub- ture, Sports, Science and Technology in Japan. med: 28653302. REFERENCES 11. Staheli LT, Corbett M, Wyss C, et al. Lower-extremity rota- tional problems in children. Normal values to guide manage- 1. Abe S, Nakamura T, Rodriguez-Vazquez JF, et al. Early fetal ment. J Bone Joint Surg Am. 1985; 67(1): 39–47, indexed in development of the rotator interval region of the shoulder Pubmed: 3968103. with special reference to topographical relationships among 12. Walker JM, Goldsmith CH. Morphometric study of the fetal related tendons and ligaments. Surg Radiol Anat. 2011; development of the human hip joint: significance for con- 33(7): 609–615, doi: 10.1007/s00276-011-0780-3, indexed genital hip disease. Yale J Biol Med. 1981; 54(6): 411–437, in Pubmed: 21249362. indexed in Pubmed: 7342490. 2. Abe SI, Aoki M, Nakao T, et al. Variation of the subscapularis 13. Yoshio M, Murakami G, Sato T, et al. The function of the tendon at the fetal glenohumeral joint. Okajimas Folia Anat psoas major muscle: passive kinetics and morphological Jpn. 2014; 90(4): 89–95, indexed in Pubmed: 24815107. studies using donated cadavers. J Orthop Sci. 2002; 7(2): 3. Aizawa Y. On the organization of the Plexus lumbalis. I. On 199–207, doi: 10.1007/s007760200034, indexed in Pub- the recognition of the three-layered divisions and the sys- med: 11956980.

Blind mole rat (Spalax leucodon) masseter muscle: structure, homology, diversification and nomenclature A. Yoldas1, M. Demir1, R. İlgun2, M.O. Dayan3

1Department of Anatomy, Faculty of Medicine, Kahramanmaras University, Kahramanmaras, Turkey 2Department of Anatomy, Faculty of Veterinary Medicine, Aksaray University, Aksaray, Turkey 3Department of Anatomy, Faculty of Veterinary Medicine, Selcuk University, Konya, Turkey

[Received: 10 July 2018; Accepted: 23 September 2018]

Background: It is well known that rodents are defined by a unique masticatory apparatus. The present study describes the design and structure of the masseter muscle of the blind mole rat (Spalax leucodon). The blind mole rat, which emerged 5.3–3.4 million years ago during the Late Pliocene period, is a subterranean, hypoxia-tolerant and cancer-resistant rodent. Yet, despite these impressive characteristics, no information exists on their masticatory musculature. Materials and methods: Fifteen adult blind mole rats were used in this study. Dissections were performed to investigate the anatomical characteristics of the masseter muscle. Results: The muscle was comprised of three different parts: the superficial masseter, the deep masseter and the zygomaticomandibularis muscle. The superficial masseter originated from the facial fossa at the ventral side of the infraorbital foramen. The deep masseter was separated into anterior and posterior parts. The anterior part of the zygomaticomandibularis muscle arose from the snout and passed through the infraorbital foramen to connect on the mandible. Conclusions: The construction of the deep masseter and zygomaticomandibularis muscles were of the Myomorpha type. Further studies are needed to reveal features such as muscle biomechanics, muscle types. (Folia Morphol 2019; 78, 2: 419–424)

Key words: anatomy, masseter muscle, Spalax leucodon

INTRODUCTION Animals from the Spalax leucodon spend the vast Spalax leucodon superspecies is a group of wild majority of their life in their underground burrows rodents belonging to the Spalacidae family. The general and tunnel systems. Their eyes are atrophied and the consensus is that they emerged 5.3–3.4 million years orbits are covered with skin. They are different from ago (Late Pliocene period) around Anatolia and spread other rodents, with their external ear pinnae and to Anatolia, the Balkans, the Russian steppes, the Mid- atrophied tails. The average weight of a mole rat is dle East and North Africa. Morphological investigations 100–570 g [9]. show that two species (Spalax leucodon and Spalax Spalax leucodon is an exceptional species that is ehrenbergi) of blind mole rats are found in Turkey [18]. able to survive in underground galleries at an oxygen The Spalacidae, or spalacids, are a family of rodents in level of 7%, while possessing certain morphological, the large and complex superfamily Muroidea [24, 25]. physiological and behavioural characteristics under

Address for correspondence: Dr. A. Yoldas, Department of Anatomy, Faculty of Medicine, University of Kahramanmaras, Bahçelievler, Kahramanmaras, Turkey, e-mail: [email protected]

419 Folia Morphol., 2019, Vol. 78, No. 2

routine underground conditions with high carbon and investigate its relationship in S. leucodon and dioxide levels. Therefore, they are used as an animal other rodents. model in hypoxic studies [21, 31]. Moreover, the species can live up to 21 years, and, interestingly, MATERIALS AND METHODS spontaneous tumours have never been observed [10]. Animals and experimental procedure They also have different structures of haemoglobin For this study, we collected 11 male and 4 female [19], myoglobin [42], respiratory system [17] and mole rats (Spalax leucodon) from the Aksaray Province myocardial performance [8]. of Turkey. All the animals were captured by villagers. It is well known that there is a significant re- Mole rats were taken to the Department of Anatomy, lationship between lifestyle, feeding strategy and Faculty of Veterinary Medicine, Aksaray University, in chewing muscles in mammalian species. Therefore, well-ventilated cages (dimensions: 1 × 0.7 × 1 m). the morphology of the masticatory muscles differs All experimental procedures were approved by the from species to species. Rodents, the largest group Animal Experiments Local Ethics Committee of the of mammals, are unique due to their extraordinary Afyonkarahisar Kocatepe University (approval no. masticatory apparatus. The masseter muscle com- 06.01.2012/1242). prises 60–80% of the total masticatory muscle mass At room temperature (20°C), the animals were in rodents [5, 33]. The morphological properties of anaesthetised with an intraperitoneal injection of the masseter muscle can be changed by the type of a combination of 10 mg/kg xylazine (Rompun enj; masticatory activity of the rodents. The structure Bayer Turk Chemistry Industry Ltd. Corp., Istanbul, Tur- of the masseter muscle has also been an important key) and 100 mg/kg ketamine HCl (Ketalar Eczacıbası; criterion in the classification of rodents [2]. In terms Istanbul, Turkey). Fifteen mole rats were killed by of masticatory muscle anatomy, rodents can be di- exsanguination of the left common carotid artery vided into four classes: Protrogomorpha (primitive under anaesthesia. Animals were then fixed in for- rodents), Sciuromorpha (squirrels), Myomorpha (rats mal saline (buffered 4% formaldehyde solution with and mice), and Hystricomorpha (porcupines and the 0.12 M NaCl) and stored in 70% ethanol. Mole rats South American caviomorph rodents) [23]. were dissected to further investigate and examine Kangaroos, wallabies, and rat-kangaroos (su- the anatomy of the masticatory muscles. For each perfamily Macropodoidea) are Australian dominant masticatory muscle, the positions of its origin and mammalian herbivores. They vary markedly in dietary insertion on the skull and the mandible were care- preferences, similar to the artiodactyl ungulates of fully inspected, as well as the direction of muscle other continents. Some species utilise many differ- fibres,in order to establish mycological maps. Finally, ent foods, whereas others are much more specific the findings were recorded and photographed using in their choice of food items. Different foods require a Canon Eos 600d. After being photographed, mac- processing in different ways. Masticatory morphol- eration was performed to describe the cranial bones ogy represents a trade-off between what is required and the attachment of the masseter muscle, as in to effectively break down food and phylogenetic previous studies. The bone structures related to the constraints. The correlation between feeding strategy masseter muscle were examined based on the work and morphology in kangaroos has been investigated done by Ozkan [27]. by several researchers, and the results have indicated differences in dental morphology and behaviour as- RESULTS sociated with the feeding types [35]. The masseter muscle consisted of three main Rodents have a very large population and consti- parts: superficial masseter, deep masseter and zygo- tute more than half of the known mammalian species. maticomandibularis muscle. The blind subterranean mole rat, Spalax leucodon superspecies, is a group of species of wild rodents. Superficial masseter Many studies [1, 4, 6, 12, 14, 20, 35, 36, 39–41] It is the outermost layer of the masseter muscle have been conducted on the masseter muscles of with a water-drop shape. The superficial masseter rodents living in different ecological conditions. How- was observed to cover two-thirds of the anterior ever, information about the masseter muscle of the deep masseter. However, the anteroventral portion Spalax genus is limited. This study aims to describe of the superficial masseter was clearly distinguishable the macro-anatomic structure of the masseter muscle from the deep masseter muscle with the naked eye

420 A. Yoldas et al., The masseter muscle in Blind Mole Rat

(Figs. 1–3). The aponeurosis of the origin of the superficial masseter formed a strong flattened tendinous (Figs. 2, 3). This aponeurosis originated from the oval shallow fossa and was located on the inferior portion of the zygomatic root, on the ventrolateral face of the maxilla. After its origin, the musculoaponeurotic fibres had an emphasised ventro-aboral obliquity, fanned out on the large part of the anterior deep masseter and the posteroventral portion of the posterior deep masseter, and continued with a massive muscular belly that could be seen on the angle and ventral edge of the angular process of the mandible. Figure 1. General view of skull and origins-insertions of the mas- The muscle fascicles were extended to the back of the seter muscle in S. leucodon (lateral view); 1 — incisor teeth; 2 — mandible and inserted between the posterolateral mental foremen; 3 — molar teeth; 4 — mandibular body; 5 — masseteric ridge; 6 — mandibular angle; 7 — masseteric fossa; surface of the angle of the mandible and the border 8 — angular process; 9 — condyloid process; 10 — external acous- of the masseteric fossa. Most of the muscle fascicles tic pore; 11 — occipital condyle; 12 — temporal process of the on the dorsal edge of this muscle were also integrated zygomatic arch; 13 — coronoid process; 14 — zygomatic arch; 15 — infraorbital foramen; 16 — the area of origin of the anterior into the aponeurosis of the deeper masseteric layer, deep masseter (the zygomatic plate); 17 — maxilla; 18 — the making the separation of these two parts of the mas- area of origin of the superficial masseter; SM (yellow lines) — the seter difficult (Figs. 2, 3). superficial masseter; PDM (red lines) — the posterior deep masseter; ADM (red lines) — the anterior deep masseter; PZM (purple lines) — the posterior zygomandibular masseter; AZM (purple Deep masseter lines) — the anterior zygomandibular masseter; IOZM (purple The deep masseter was divided into an anterior lines) — infraorbital part of zygomandibular masseter. and a posterior part. The anterior deep masseter originated from the zygomatic plate, which had a narrow pouch, and was horizontally located on the anteroventral face of the zygomatic process of the maxilla (Fig. 1). The point of origin of the anterior part of the deep masseter filled the pouch. After its origin, it passed horizontally, placed under the superficial masseter, and inserted on the masseteric ridge, placed rostral of the fossa mandible (Figs. 1–3). The posterior deep masseter originated along the caudoventral edge of the zygomatic arch and ran ventral-caudal to insert into the shallow masseteric fossa, located on the lateral face of the mandible Figure 2. Lateral view of masseter muscles; ADM — anterior deep (Figs. 1–3). masseter; IOZM — infraorbital part of the zygomaticomandibularis; PDM — posterior deep masseter; SM — superficial masseter; Zygomaticomandibularis muscle E — atrophied eye; ZA — zygomatic arch. It was the innermost layer of the masseter muscle and consisted of two parts: pars anterior and pars posterior. Then, it ran caudoventrally and inserted onto the The pars anterior originated from the medial face anterior root of the coronoid process (Figs. 1, 3, 4). of the temporal process of the zygomaticum bone. The pars posterior originated on the ventral and A part of the pars anterior, the infraorbital zygoma- medial surface along the posterior half of the zygo- ticomandibularis muscle also rose from the medial matic arch. The fibres of the posterior were weaker surface of the maxillary and the root of the zygo- than those of the anterior. The fibres of the pars matic arch. Its muscle fibres extended through the posterior passed straight rostroventrally to insert on infraorbital foramen, which was wide and ovoid, and the fosse, which was located on lateral surface of the combined with the main part of the pars anterior. coronoid process (Figs. 1, 4).

421 Folia Morphol., 2019, Vol. 78, No. 2

and the zygomaticomandibularis muscle. However, Hiiemae and Houston [14] reported that, in rodents, the masseter muscle was divided into two parts: the superficial masseter and the deep masseter. In our sample, the masseter muscle was observed with three main parts, as described in some rodents [5–7, 12]. It was reported that, in some rodents, the superficial masseter had two parts (superficial and lateral) [12, 19]. On the other hand, Miller and Gidley [22] and Howell [16] expressed that these two muscles were the same. In our samples, the superficial masseter was a single part and originated from the maxillary Figure 3. Lateral-ventral view of masseter muscles (the SM and bone via a flattened and strong tendon, as in Muroids, the part of the PDM have been removed); SM — superficial masseter (removed); OSM — the origin of SM; ADM — anterior deep Geomyida [15, 28], rats [14, 36], Laotian rock rats masseter; PDM — posterior deep masseter; AZM — anterior [12] and Marsupialia [35]. zygomaticomandibularis; ZA — zygomatic arch; E — atrophied Some researchers described the superficial mas- eye; M — mandibular body. seter in squirrels as originating from the masseteric tubercle, which is under the infraorbital foramen [41]. However, in our case, the masseteric tubercle was absent and the ovoid fossa had developed instead. This finding is supported by some studies conducted on squirrels [32], guinea pigs [37] and rats [11]. The deep masseter, namely the lateral masseter in some rodents [39], consists of two separate parts in rodents [5, 41]. This finding is consistent with our report. However, the deep masseter has been noted to be a single part in guinea pigs [5], squirrels [33] and mice [1]. Due to the anatomical variety of the zygomatic plate, the morphologic structure of the origin of the anterior deep masseter can be different in rodents. Thus, the structure of the zygomatic plate is critical for the classification of rodents. The anterior deep Figure 4. Left lateral view of masseter muscles (the SM, ADM, masseter originates from the zygomatic plate [5, 6]. PDM, the superior of ZA and the process frontal of the zygomatic However, unlike some Sciuromorpha and Hystrico- arch have been removed); ADM — anterior deep masseter (removed); AZM — anterior zygomaticomandibularis muscle; morpha, in our samples, the zygomatic plate was PZM — posterior zygomaticomandibularis muscle; IOZM — settled in the anteroventral face of the zygomatic inferior part zygomaticomandibularis muscle; ZA — zygomatic process of the maxilla. It did not expand onto the arch; E — atrophied eye; M — mandibular body; CP — coronoid rostrum, as in fossil rodents [34] and Aplodontia rufa, process; MF — masseteric fossamasseter. which is the only living protrogomorphous rodent [7]. In our samples, the posterior deep masseter originated from the part of the zygomatic arch that termi- DISCUSSION nated on the mandible fossa, with the same pattern Since rodents are the biggest mammalian group, as in squirrels [5, 33, 41], rats [5] guinea pigs [26], the classification of the masseter muscle of rodents Aplodontia rufa and Marmota monax [7]. is controversial. For many years, there have been ar- Some researchers [11] did not make any comments guments about the parts of the masseter muscle of about the presence of the zygomaticomandibularis mammals. Some researchers [1, 5, 35, 41] reported muscle in rats. Whereas Hiiemae and Houston [14] that the masseter muscle of rodents had three main could not distinguish the zygomaticomandibularis parts: the superficial masseter, the deep masseter muscle, and thus regarded it as a part of the deep

422 A. Yoldas et al., The masseter muscle in Blind Mole Rat

masseter, others [3, 4, 29] ignored this muscle. This 223(1): 46–60, doi: 10.1111/joa.12059, indexed in Pub- muscle was also termed the medial muscle of the med: 23692055. 2. Brandt JK. BeitragezurnahernKenntniss der Saugethiere- masseter muscle by some researchers [12]. We saw Russlands. Memoires de l’Acade´mieImpe´ riale des. that the zygomaticomandibularis muscle was the in- Sciences de St Petersbourg. 1855; 69: 1–375. nermost layer of the masseter, and was divided into 3. Bresin A, Bagge U, Kiliaridis S. Adaptation of normal and hypo- two parts (the pars anterior and pars posterior); this functional masseter muscle after bite-raising in growing rats. is the same as in squirrels, rats, guinea pigs [5], mice Eur J Oral Sci. 2000; 108(6): 493–503, doi: 10.1034/j.1600- 0722.2000.00915.x, indexed in Pubmed: 11153924. [1], Laotian rock rats [12], mountain beavers and 4. Byrd KE. Mandibular movement and muscle activity during Marmota monax [7]. mastication in the guinea pig (Cavia porcellus). J Morphol. In some living Myomorpha rodents, such as Spalax 1981; 170(2): 147–169, doi: 10.1002/jmor.1051700203, sp., Myospalax sp., and Oxycmterus sp., the zygomas- indexed in Pubmed: 7299825. 5. Cox PG, Jeffery N. Reviewing the morphology of the jaw- seteric structure has been shown to be an intermedi- closing musculature in squirrels, rats, and guinea pigs ate type between the hystricomorphous and myomor- with contrast-enhanced microCT. Anat Rec (Hoboken). phous types. The diameter of the infraorbital foramen 2011; 294(6): 915–928, doi: 10.1002/ar.21381, indexed is larger in Hystricomorpha than in Myomorpha [20]. in Pubmed: 21538924. 6. Druzinsky RE, Doherty AH, De Vree FL. Mammalian Furthermore, Hystricomorpha and Myomorpha, un- masticatory muscles: homology, nomenclature, and di- like Sciuromorpha have the infraorbital portion of the versification. Integr Comp Biol. 2011; 51(2): 224–234, pars anterior of the zygomaticomandibularis mus- doi: 10.1093/icb/icr067, indexed in Pubmed: 21742777. cle, which passes through the infraorbital foramen 7. Druzinsky RE. Functional anatomy of incisal biting in Aplo- [5, 33, 41]. This portion is enlarged and covers most dontia rufa and sciuromorph rodents - part 1: masticatory muscles, skull shape and digging. Cells Tissues Organs. of the lateral wall of the rostrum in Hystricomorpha. 2010; 191(6): 510–522, doi: 10.1159/000284931, indexed In our specimens, although this portion existed, it was in Pubmed: 20160428. not as wide as described in hystricomorphous [36]. 8. Edoute Y, Arieli R, Nevo E. Evidence for improved myo- In rats, squirrels, guinea pigs [5] and Laotian rock cardial oxygen delivery and function during hypoxia in the mole rat. J Comp Physiol B. 1988; 158(5): 575–582, rats [12], the anterior part of the zygomaticoman- indexed in Pubmed: 3249020. dibularis muscle is stronger than the posterior, as in 9. Eroglu F. Çorum city Spalax Leucodon Nordmann. 1840 the blind mole rat. Moreover, a large posterior part the caryiological and morphological analysis of Mammalia: of the zygomaticomandibularis muscle is also found Rodentia. Zonguldak Karaelmas University Science. 2006. in carnivores, in which it may function to resist forces 10. Gorbunova V, Hine C, Tian X, et al. Cancer resistance in the blind mole rat is mediated by concerted necrotic cell produced by struggling prey [30]. death mechanism. Proc Natl Acad Sci U S A. 2012; 109(47): In some Hystricomorpha rodents, an extra section 19392–19396, doi: 10.1073/pnas.1217211109, indexed of the posterior part of the zygomaticomandibularis in Pubmed: 23129611. muscle has been distinguished and named the mus- 11. Greene EC. Anatomy the Rat. Transactions of the Ameri- can PtvIosophicaI Society. Vol. XXVII, Hainer Publishing cles masseter lateralis profundus [40], the posterior Company, New York and London. 1963. masseter [7, 13], the pars posterior or the deep divi- 12. Hautier L, Saksiri S. Masticatory muscle architecture in sion [38]. We did not find any evidence of this part, the Laotian rock rat Laonastes aenigmamus (Mammalia, which was separated from the posterior part of the Rodentia): new insights into the evolution of hystricog- nathy. J Anat. 2009; 215(4): 401–410, doi: 10.1111/j.1469- zygomaticomandibularis muscle 7580.2009.01130.x, indexed in Pubmed: 19694873. 13. Herring SW. Masticatory muscles and the skull: a compara- Conclusions tive perspective. Arch Oral Biol. 2007; 52(4): 296–299, This study is a detailed description of the mor- doi: 10.1016/j.archoralbio.2006.09.010, indexed in Pub- phological features of the Spalax leucodon complex med: 17084804. 14. Hiiemae K, Houston W. The structure and function of the masseter muscle. The data were compared to those jaw muscles in the rat (Rattus norvegicus L.). Zoological for other rodents and may contribute to enhancing J Linnean Society. 1971; 50(1): 75–99, doi: 10.1111/j.1096- our morphological knowledge regarding the mastica- 3642.1971.tb00752.x. tory musculature in mammals. 15. Hill JE. Morphology of the pocket gopher mammalian genus thomomys. Univ Calif Publ Zool. 1937; 42: 1–171. REFERENCES 16. Howell A. The saltatorial rodent dipodomys: the functional and comparative anatomy of its muscular and osseous 1. Baverstock H, Jeffery NS, Cobb SN. The morphology systems. Proc Am Acad Arts Scien. 1932; 67(10): 377, of the mouse masticatory musculature. J Anat. 2013; doi: 10.2307/20022915.

423 Folia Morphol., 2019, Vol. 78, No. 2

17. İlgun R, Yoldas A, Kuru N, et al. Macroscopic anatomy 1–14, doi: 10.1002/(sici)1097-4687(199904)240:1<1::aid- of the lower respiratory system in mole rats (Spalax jmor1>3.0.co;2-d. leucodon). Anat Histol Embryol. 2014; 43(6): 474–481, 30. Scapino RP. Biomechanics of Feeding in Carnivora. Chi- doi: 10.1111/ahe.12098, indexed in Pubmed: 24450964. cago. Illinois: Dissertation, University of Illinois. 1968. 18. Kıvanc E. The geographic variations of Turkey Spalax. 31. Shams I, Aaron A, Eviatar N. Hypoxic stress tolerance Ankara (Notes in Turkish. 1988: 72–88. of the blind subterrean mole rat. Laboratory of Animal 19. Kleinschmidt T, Nevo E, Braunitzer G. The primary structure Molecular Evolution. Institute of Evolution, University of of the hemoglobin of the mole rat (Spalax ehrenbergi, Haifa, Mount Carmel, Haifa. 2004; 10(1): 9698–9703. rodentia, chromosome species 60). Hoppe Seylers Z Physiol 32. Thorington R, Darrow K. Jaw muscles of Old World squirrels. Chem. 1984; 365(5): 531–537, indexed in Pubmed: 6469215. J Morphol. 1996; 230(2): 145–165, doi: 10.1002/(sici)1097- 20. Klingener D. The Comparative Myology of Four Dipo- 4687(199611)230:2<145::aid-jmor3>3.0.co;2-g. doid Rodents: Genera Zapus, Napaeozapus, Sicista, and 33. Turnbull WD. Mammalian masticatory apparatus. Fieldiana Jaculus. Front Cover, Museum of Zoology. University of Geol. 1970; 18(2): 147–356. Michigan. 1964. 34. Vianey-Liaud M. Possible evolutionary relationships among 21. Malik A, Bicker A, Poetzsch G, et al. Hypoxia tolerance, Eocene and Lower Oligocene rodents of Asia, Europe, and longevity and cancer-resistance in the mole rat Spalax: North America. In: Luckett E.W.P., Hartenberger E.J.L.(eds): a liver transcriptomics approach. Scientific Reports. 2017; Evolutionary Relationships among Rodents: A Multidisci- 7(1), doi: 10.1038/s41598-017-13905-z. plinary Analysis. Plenum Press NATO ASI. New York 1985. 22. Miller G, Gidley J. Synopsis of the supergenerie groups 35. Warburton NM. Comparative jaw muscle anatomy in of rodents. J Washington Acad Scien. 1918; 8: 431–448, kangaroos, wallabies, and rat-kangaroos (marsupialia: doi: 10.5962/bhl.part.6490. macropodoidea). Anat Rec (Hoboken). 2009; 292(6): 23. Musser GG, Carleton MD. Suerfamily Muroidea. In: Wil- 875–884, doi: 10.1002/ar.20905, indexed in Pub- son DE, Reeder DM (eds.). Mammal Secies of the World: med: 19462457. a Taxonomic and Geograhic Reference. 3rd ed. Baltimore: 36. Weijs W. Morphology of the muscles of mastication in Johns Hokins University Press. 2005: 894–1531. the albino rat Rattusnorvegicus. Acta Morph Neerl-Scand. 24. Nevo E. utionary theory and processes of active speciation and 1973; 11: 312–340. adaptive radiation in subterranean mole rats. Spalaxehrenber- 37. William PW, Sherri JH. Biomechanical analysis of mastica- gisuperspecies in Israel Evol Biology. 1991; 25: 1–125. tion in the fossil rodent ischyromys and its bearing on 25. Norris RW, Zhou K, Zhou C, et al. The phylogenetic posi- the origin of sciuromorphs, Technical Report N.P. Service tion of the zokors (Myospalacinae) and comments on Paleontology Research. 1993; 2: 21. the families of muroids (Rodentia). Mol Phylogenet Evol. 38. Wood AE. The evolution of Old World and New World 2004; 31(3): 972–978, doi: 10.1016/j.ympev.2003.10.020, hystricomorphs. In: Rowlands I.W, Weir B.J. (eds): The indexed in Pubmed: 15120394. Biology of Hystricognath Rodents. Academic Press, Symp- 26. Odagiri N, Kubota K, Shibanai S. Density of muscle spindles ZoolSoc London. 1974; 34: 21–60. in the jaw muscles of the Japanese flying squirrel and the 39. Woods CA, Howland EB. Adaptive Radiation of Capromyid guinea pig. Ann Anat. 1993; 175(3): 263–270, indexed in Rodents: Anatomy of the Masticatory Apparatus. J Mam- Pubmed: 8338226. mal. 1979; 60(1): 95–116, doi: 10.2307/1379762. 27. Ozkan ZE. Macro-anatomical investigations on the 40. Woods CA. Comparative myology of jaw, hyoid, and skeletons of mole-rat (Spalax leucodon Nordmann). pectoral appendicular regions of new and old world III. Skeleton axiale. Veterinarski Arhiv. 2077; 7(3): 281–289. hystricomorph rodents. Bull Am Mus Nat Hist. 1972; 28. Rinker GC. The comparative myology of the mammalian 147: 115–198. genera Sigmodon, Oryzomys, Neotoma, and Peromyscus 41. Yalcın H, Arslan A, Tıpırdamaz S. Macro-anatomical inves- (Cricetinae), with remarks on their intergeneric relation- tigations on the masticatory muscles of tree squirrel. Vet. ships. MiscPublMusZoolUniv Michigan. 1954; 83: 1–124. Bilimleri, Dergisi 2003; 19. ; 3(4): 83–88. 29. Satoh K. Mechanical advantage of area of origin for the exter- 42. Zanjani ED, Poster J, Burlington H, et al. Liver as the prima- nal pterygoid muscle in two murid rodents,Apodemus spe- ry site of erythropoietin formation in the fetus. J Lab Clin ciosus andClethrionomys rufocanus. J Morphol. 1999; 240(1): Med. 1977; 89(3): 640–644, indexed in Pubmed: 839121.

Dermatoglyphics as a possible morphological biomarker in myopia: analysis of finger ridge counts and fluctuating asymmetry L. Sretić1, N. Labus1, T. Filipović2, M. Filipović3

1Department of Biology, Faculty of Sciences and Mathematics, University of Priština-Kosovska Mitrovica, Kosovska Mitrovica, Serbia 2Department of Anatomy, Medical Faculty, University of Priština-Kosovska Mitrovica, Kosovska Mitrovica, Serbia 3Department of Surgery, Medical Faculty, University of Priština-Kosovska Mitrovica, Kosovska Mitrovica, Serbia

[Received: 11 June 2018; Accepted: 11 July 2018]

Background: The aim of this study is to provide the first analysis of fingerridge counts and fluctuating asymmetry in myopia, in order to evaluate dermatoglyphic role as a morphological biomarker. Materials and methods: Study sample consisted of 102 participants recruited from freshman students’ population of the University of Priština-Kosovska Mitro- vica. Prints were taken by standard ink and paper method. Differences in mean ridge counts between examined groups were analysed by ANOVA analysis of variance. Fluctuating asymmetry assessment was performed by using correlation method (p < 0.05). Results: Analysis has identified myopic males as the group with the most prominent differences of examined dermatoglyphic parameters. Myopic males, compared to controls, have significantly higher ridge counts for left and right ring and little finger, as well as total ridge count. Also, this group has recorded significant difference in fluctuating asymmetry correlation score for middle finger, and borderline significance for thumb and ring finger. Conclusions: Overall findings of this study have indicated that dermatoglyphics might serve as a morphological biomarker, especially in myopic males, selecting them as the group with dermatoglyphic differences that might be suggestive of higher developmental instability. Although promising, the present results should be considered as preliminary until future investigations replicate them in a larger sample. (Folia Morphol 2019; 78, 2: 425–430)

Key words: dermatoglyphics, myopia, fluctuating asymmetry, developmental instability

INTRODUCTION are derived as an evolutionary adaptation that en- Dermatoglyphics is a term that originates from hances tactile sensation and friction, and are unique two Greek words, δέρμα — skin and γλύφή — carving, characteristic typically found in higher primates or, describing a study of a system of cutaneous ridges sporadically, in other mammals [31]. The pattern of and furrows flowing in distinctive paths or directions ridged skin is established from 6th to 17th week of on fingers, palmar and plantar surfaces. Dermal ridges gestation, when the basal layer of the volar epidermis

Address for correspondence: Dr. L. Sretić, Department of Biology, Faculty of Sciences and Mathematics, University of Priština-Kosovska Mitrovica, Ive Lole Ribara 29, 38220 Kosovska Mitrovica, Serbia, tel: +381638277826, e-mail: [email protected]

425 Folia Morphol., 2019, Vol. 78, No. 2

becomes folded forming primary ridges. This process is Bilaterally represented traits demonstrate three influenced by the volar pads, local temporary eminences types of asymmetry, differentiated by their causes of the subcutaneous tissue and the sites of ridge for- and biological significance: directional asymmetry, mation. Subsequent environmental insults, that follow antisymmetry and fluctuating asymmetry. The aetiol- full maturation of secondary ridges during 24th week, ogy of directional asymmetry and antisymmetry may remain basic dermatoglyphic structure unaffected, mak- be regarded as a part of the developmental plan, and ing them “a history of development” [3]. therefore likely to have an adaptive significance [5, Heredity of quantitative dermatoglyphic traits 13]. Fluctuating asymmetry signifies small, random conforms to polygenic system, with individual genes departures from perfect symmetry between the left contributing a small additive effect, but the prenatal and right side of a bilateral trait, where the right–left environment may also exert an important influence variation is normally distributed about a mean of [2]. Finger ridge count is one of the most heritable an- zero [32]. It is considered that fluctuating asymmetry thropometric features, and has been used as a model stems from inability of the organism to buffer nega- trait for the study of human quantitative genetics tive influences of disturbing developmental factors, [7, 29]. Since that genetic component affect ridge indirectly reflecting the level of stress experienced composition indirectly, through ontogenetic factors during development [6, 42]. Since that development such as embryonic pad topography, growth stress, of bilateral symmetrical traits is under control of iden- neurotrophic or skeletal factors, it is not the pattern tical genes the underlying assumption of fluctuating of friction ridge skin that is passed down through asymmetry analysis is that nondirectional differences heredity, but the shape and location of volar pads between two sides are of environmental origin, re- [3, 21]. As a result of their polygenic inheritance, flecting insults during developmental time [23, 37]. which makes them less susceptible and less vulner- The main objective of this study was to investigate able to stochastic processes, such as genetic drift, effects of myopic visual impairment on finger ridge dermatoglyphics are widely used in population stud- counts and levels of fluctuating asymmetry, in order ies [1]. Besides, due to their polygenic determination, to determine possible role of dermatoglyphics as genes underlying certain disorder may, by pleiotropy, a morphological marker. affect dermatoglyphic parameters [12]. This makes them, along with lifetime permanence and the fact MATERIALS AND METHODS that forces that channel ridge differentiation must be Study sample consisted of 102 participants re- operating prior the 19th week of gestation, a sensi- cruited from the freshman students’ population of tive indicator of intrauterine disturbances associated the University of Priština-Kosovska Mitrovica. Among with chromosomal/gene abnormalities, environmen- them, 51 were diagnosed with common myopia, tal stress, or a combination of these [28]. Altered from –1 to –5 dioptres, and 51 composed healthy dermatoglyphic configuration has been proven in control group (both groups included 26 females and numerous multifactorial or chromosomal disorders 25 males). Individuals were aligned by sex and vision [22, 28, 35]. as control males, control females, myopic males and One of the most common public health issues myopic females. Prints were taken by standard ink in the world is myopia [11]. Eye morphogenesis is and paper method described by Cummins and Midlo extremely precise, genetically determined process [8]. Scanned nail-to-nail rolled impressions of inked [15], involving ectodermal and neuroectodermal fingertips (CanoScanLiDE 25) were enlarged by image derivatives [39]. Uncoordinated contribution of ocu- editing software programme Adobe Photoshop CS3, lar components may lead to myopia, a multifacto- and scored. Comparisons were made between control rial eye disorder characterised by blurred vison of and myopic individuals of the same sex. distant objects [44]. Strong genetic background of Quantitative analysis encompassed finger ridge myopia onset has been proven in two recent large- counts (FRC) and total ridge count (TRC). Finger ridge scale genome-wide association studies that have count is defined as number of ridges intersected with identified significant association of several candidate the line drawn from the core, centre of a pattern, and genes, involved in neurotransmission, ion transport, triradius, point of three ridge systems confluence at retinoic acid metabolism, extracellular matrix remod- an angle of approximately 120o. Arch pattern has no elling and eye development, with this type of visual triradius, so the score for ridge count is zero, loops impairment [25]. have one and whorls two, yielding two counts, but 426 L. Sretić et al., Dermatoglyphics in myopia

only higher is used for finger ridge or total ridge DISCUSSION count. Total ridge count is a common composite trait Current paper is, to the best of our knowledge, calculated by summing ridge counts for all 10 fingers. the first attempt to investigate finger ridge counts Fingers of the left and right hand were designated and fluctuating asymmetry in myopia. as FRCL1–5 and FRCR1–5. Mean finger ridge counts Dermatoglyphic alterations obtained in this study, were analysed in relation to their variability between manifested by the degree of variations in measured examined groups using univariate analysis of variance parameters-ridge counts and fluctuating asymme- (ANOVA). try, identified myopic males as the group with the Fluctuating asymmetry assessment was per- most pronounced differences. The variables that have formed by using correlation method, since that it been highlighted for their significant scores between is not affected by directional asymmetry [36]. Pear- myopic and control males were higher ridge counts son’s product-moment correlation coefficient (r) was on left and right ring and little finger. Consequently, applied for the comparison of ridge counts between this group also recorded significant increase in TRC. 2 homologous fingers. r is a measure of their com- On the contrary, the only significant variation found 2 mon variance; 1-r is an estimate of error variance in females stemmed from the lower FRC on the left and thus a measure of fluctuating asymmetry [40]. little finger of myopic females in relation to control Statistical significance of differences in correlation females. Since that morphogenesis of some derma- coefficients between myopic and control males, toglyphic features is associated to specific stages of as well as females, was calculated using Fischer’s prenatal development, it has been suggested that z-transformation [10]. Analysis of correlation was TRC may be considered as an index of early foetal performed by SPSS7 (Statistica for Windows; Stat- growth rate [34], influenced by stimulating or- in Soft, Inc., Tulsa, OK, USA), while the Fischer’s z trans- hibiting factors [20]. According to Meier et al. [30] formation was calculated employing a test available males are late maturers and due to the later than on Internet [24]. The level of significance for all average ridge formation, along with a delay in volar reported differences was set at p < 0.05. pad regression, are found to have larger and more The protocol and informed consent procedure was complex dermatoglyphic patterns, such as whorls, approved by Institutional Ethical Committee. and increased digital ridge counts. Fluctuating asymmetry is a concept first described RESULTS by Ludwig [27] as a sign of ontogenic stability. Ever Means and standard deviations for left and right since it is one of the most common used tools in finger ridge counts in control and myopic males and measuring developmental stability, i.e. ability of an females are shown in Table 1. organism to moderate its development against ge- The results of univariate analysis of variance netic or environmental stresses. Increasing fluctuat- (ANOVA), presented in Table 2, point to significant ing asymmetry is in human populations linked to heterogeneity in ridge counts between myopic and some indicators of developmental stability such as control males for both left and right ring (p = 0.0198, morbidity and number of offspring [43] or length p = 0.0009. respectively) and little finger (p = 0.0026, of gestation [26], as well as to specific multifactorial p = 0.0005, respectively), and total ridge count disorders [9, 38]. (p = 0.0353), as a consequence of elevated values in Our analysis of fluctuating asymmetry between myopic males. The only significance in ridge counts homologous fingers showed that myopic individuals between myopic and control females was recorded of both sexes have higher degree of fluctuating asym- for left little finger (p = 0.027), being lower in my- metry in comparison to controls, but the only signifi- opic females. cance refers to middle finger in myopic males along Data presented in Table 3, displaying fluctuating with significance of borderline level for thumb and asymmetry indices (1-r2), Fisher’s z-transformation (z) ring finger. This might implicate somewhat greater and significance (p), reveal higher levels of fluctuating developmental instability in myopic males, i.e. greater asymmetry in myopic males and females compared to vulnerability to adverse environmental influences, and controls. The only significant variation was found in thus be in agreement with a hypothesis that males myopic males for middle finger (p = 0.017),but there may be less canalised in their growth and develop- was also a clear tendency to significance for thumb ment than females [41]. The earliest paper concerning and ring finger (p = 0.058, p = 0.054, respectively). environmental impact on males’ development appear 427 Folia Morphol., 2019, Vol. 78, No. 2

Table 1. Ridge count means and standard deviation (SD) Males Females Control (n = 25) Myopia (n = 25) Control (n = 26) Myopia (n = 26) Mean SD Mean SD Mean SD Mean SD FRCL1 17,480 6,838 16,480 5,058 14,692 5,562 14,884 5,361 FRCL2 12,280 7,242 13,800 6,626 11,076 6,157 8,692 8,073 FRCL3 12,320 5,843 14,680 2,511 14,307 5,576 11,846 7,708 FRCL4 14,520 4,601 17,360 3,684 16,769 4,411 15,961 5,257 FRCL5 12,760 4,719 16,200 2,692 15,038 4,753 12,153 4,369 FRCR1 19,360 4,414 20,440 2,814 20,115 4,537 18,153 4,522 FRCR2 12,080 7,222 15,120 4,918 11,115 7,016 11,653 7,699 FRCR3 11,840 5,997 12,520 6,028 13,384 4,079 11,461 7,100 FRCR4 14,520 5,205 18,920 3,463 17,346 4,621 15,307 5,409 FRCR5 11,680 5,406 16,320 3,091 13,961 4,677 12,307 5,626 TRC 139,160 45,697 160,800 20,194 147,423 36,001 132,423 45,007

FRC — finger ridge counts; TRC — total ridge count; L — left; R — right

Table 2. Differences in mean ridge counts between examined groups (ANOVA) Males Females F p F p FRCL1 0.3455 0.559412 0.0161 0.89951 FRCL2 0.5993 0.442636 1.4341 0.236748 FRCL3 3.442 0.069706 1.7404 0.193094 FRCL4 5.8026 0.019889 0.3601 0.551164 FRCL5 10.0205 0.002688 5.1889 0.02704 FRCR1 1.0637 0.307538 2.4375 0.124773 FRCR2 3.0259 0.088353 0.0695 0.793182 FRCR3 0.1599 0.691066 1.4337 0.2368 FRCR4 12.3827 0.000959 2.1344 0.15028 FRCR5 13.8769 0.000514 1.3283 0.254594 TRC 4.6903 0.035332 1.7611 0.190516

FRC — finger ridge counts; TRC — total ridge count; L — left; R — right

Table 3. Differences in fluctuating asymmetry correlation coefficients between examined groups Fingers Males Females Control Myopia Control Myopia 1-r2 1-r2 z p 1-r2 1-r2 z p Thumb 0.443 0.792 1.568 0.058 0.387 0.612 1.094 0.137 Index 0.429 0.552 0.584 0.28 0.430 0.758 1.514 0.065 Middle 0.517 0.955 2.127 0.017 0.478 0.686 0.948 0.172 Ring 0.365 0.707 1.608 0.054 0.414 0.674 1.222 0.111 Little 0.312 0.581 1.38 0.084 0.556 0.632 0.344 0.365

428 L. Sretić et al., Dermatoglyphics in myopia

to be Greulich’s et al. [14], who found that among 3. Babler WJ. Embryologic development of epidermal ridges children who survived the atomic bombing of Hiro- and their configurations. Birth Defects Orig Artic Ser. 1991; 27(2): 95–112, indexed in Pubmed: 1786361. shima and Nagasaki growth and maturation of males 4. Bailey SM, Gershoff SN, McGandy RB, et al. A longitudinal had been affected more adversely by environmental study of growth and maturation in rural Thailand. Hum stress than that of females. Biol. 1984; 56(3): 530–557, indexed in Pubmed: 6489996. Asymmetry between the dermatoglyphic char- 5. Boorman CJ, Shimeld SM. The evolution of left-right asym- acteristics, as postulated by Naugler and Ludman metry in chordates. Bioessays. 2002; 24(11): 1004–1011, doi: 10.1002/bies.10171, indexed in Pubmed: 12386931. [33], may signify relatively unstable genetic control 6. Bots J, ten Broek CMA, Belien JAM, et al. Higher limb asym- during embryogenesis, identifying males as more metry in deceased human fetuses and infants with ane- subjected to influences of intrauterine environment uploidy. Sci Rep. 2014; 4: 3703, doi: 10.1038/srep03703, [4, 19]. One of the theories that might shed more indexed in Pubmed: 24424506. 7. Bouchard TJ, Lykken DT, McGue M, et al. Sources of light on the hypothesis that males are more sensi- human psychological differences: the Minnesota Study tive to prenatal insults, which may be followed by of Twins Reared Apart. Science. 1990; 250(4978): atypical morphological traits and higher asymmetry, 223–228, doi: 10.1126/science.2218526, indexed in suggests the influence of prenatal circulating- hor Pubmed: 2218526. 8. Cummins H, Midlo C. Fingerprints, palms and soles. Dover mones, specifically prenatal testosterone. Prenatal Publications Inc., New York. 1976. testosterone might be related to slowed maturation, 9. de Bruin EI, Graham JH, Louwerse A, et al. Mild dermato- exposing dermatoglyphic morphogenesis longer to glyphic deviations in adolescents with autism spectrum adverse influences. Possible mechanism by which disorders and average intellectual abilities as compared testosterone modifies developmental rate might be to typically developing boys. Autism Res Treat. 2014; 2014: 968134, doi: 10.1155/2014/968134, indexed in through its stimulating effect on both epidermal and Pubmed: 25478224. nerve growth factor [16–18]. 10. Eid M, Gollwitzer M, Schmitt M. Statistik und Forschungs- Our results, although modest considering fluctu- methoden. Beltz Verlag, Weinheim, Basel. 2010. ating asymmetry, might be suggestive to an underly- 11. Foster PJ, Jiang Y. Epidemiology of myopia. Eye (Lond). 2014; 28(2): 202–208, doi: 10.1038/eye.2013.280, in- ing vulnerability that interacts with oculogenesis and dexed in Pubmed: 24406412. environmental factors, contributing to the later onset 12. Fuller IC. Dermatoglyphics: a diagnostic aid? J Med Genet. of myopia in males. The limitation of this study is 1973; 10(2): 165–169, doi: 10.1136/jmg.10.2.165, indexed relatively small sample, which may cause missing of in Pubmed: 4714584. some differences due to the lack of statistical power. 13. Gangestad SW, Thornhill R. Individual differences in developmental precision and fluctuating asymmetry: a model and its implications. J Evol Biol. 2001; 12(2): 402–416, CONCLUSIONS doi: 10.1046/j.1420-9101.1999.00039.x. Considering the overall findings of this study we 14. Greulich WW, Crimson CS, Turner ML. The physical growth may point to dermatoglyphics role as morphological and development of children who survived the atomic bombing of Hiroshima or Nagasaki. J Pediatr. 1953; 43(2): biomarker, especially in myopic males, selecting them 121–145, doi: 10.1016/S0022-3476(53)80001-6, indexed as the group with major alterations in finger ridge in Pubmed: 13070125. counts and fluctuating asymmetry score, which might 15. Heavner W, Pevny L. Eye development and retinogenesis. be suggestive of delayed maturation and higher devel- Cold Spring Harb Perspect Biol. 2012; 4(12), doi: 10.1101/ cshperspect.a008391, indexed in Pubmed: 23071378. opmental instability. Although promising, the present 16. Jamison CS. Dermatoglyphics and the Geschwindhy- results should be considered as preliminary until future pothesis. I. Theoretical background and palmar results of investigations replicate them in a larger sample. dyslexia. In: Durham NM, Plato CC (eds). Trends in Derma- toglyphic Research. Studies in Human Biology. Springer, REFERENCES Dordrecht. 1990: 99–113. 17. Jamison CS, Jamison PL, Meier RJ. Effect of prenatal testos- 1. Andreenko E, Baltova S. Sexual dimorphism in dermato- terone administration on palmar dermatoglyphic intercore glyphic traits and fluctuating asymmetry in Bulgarians ridge counts of rhesus monkeys (Macaca mulatta). Am from northeast Bulgaria. Homo. 2017; 68(4): 316–327, J Phys Anthropol. 1994; 94(3): 409–419, doi: 10.1002/ doi: 10.1016/j.jchb.2017.05.006, indexed in Pub- ajpa.1330940309, indexed in Pubmed: 7943194. med: 28867122. 18. Sorenson Jamison C, Meier RJ, Campbell BC. Dermato- 2. Arrieta MI, Criado B, Hauspie R, et al. Effects of genetic glyphic asymmetry and testosterone levels in normal males. and environmental factors on the a-b, b-c and c-d inter- Am J Phys Anthropol. 1993; 90(2): 185–198, doi: 10.1002/ digital ridge counts. Hereditas. 1992; 117(2): 189–194, ajpa.1330900205, indexed in Pubmed: 8430752. doi: 10.1111/j.1601-5223.1992.tb00173.x, indexed in 19. Jantz RL, Webb RS. Dermatoglyphic asymmetry as Pubmed: 1459858. a measure of canalization. Ann Hum Biol. 1980; 7(5):

429 Folia Morphol., 2019, Vol. 78, No. 2

489–493, doi: 10.1080/03014468000004611, indexed in 33. Naugler CT, Ludman MD. A case-control study of fluctuat- Pubmed: 7235627. ing dermatoglyphic asymmetry as a risk marker for de- 20. Kahn HS, Ravindranath R, Valdez R, et al. Fingerprint ridge- velopmental delay. Am J Med Genet. 1996; 66(1): 11–14, count difference between adjacent fingertips (dR45) predicts doi: 10.1002/(SICI)1096-8628(19961202)66:1<11::AID- upper-body tissue distribution: evidence for early gestational AJMG3>3.0.CO;2-Z, indexed in Pubmed: 8957503. programming. Am J Epidemiol. 2001; 153(4): 338–344, 34. Netley C, Rovet J. Relationships among brain organization, doi: 10.1093/aje/153.4.338, indexed in Pubmed: 11207151. maturation rate, and the development of verbal and non- 21. Kücken M, Newell AC. Fingerprint formation. J Theor Biol. verbal ability. Language Functions and Brain Organization. 2005; 235(1): 71–83, doi: 10.1016/j.jtbi.2004.12.020, Academic Press. 1983: 245–266, doi: 10.1016/b978-0-12- indexed in Pubmed: 15833314. 635640-3.50019-9. 22. Langsley N, Miller P, Byrne M, et al. Dermatoglyphics and 35. Nodari RJ, Sartoori G, Gin G, et al. Dermatoglyphic char- schizophrenia: findings from the Edinburgh high risk study. acteristics of hypertensives. Acta Medica Mediterr. 2016; Schizophr Res. 2005; 74(1): 122–124, doi: 10.1016/j. 32(4): 1015, doi: 0.19193/0393-6384_2016_4_125. schres.2004.02.006, indexed in Pubmed: 15694762. 36. Palmer AR, Strobeck C. Fluctuating Asymmetry: Measure- 23. Leamy L. Genetic analysis of fluctuating asymmetry for ment, Analysis, Patterns. Ann Rev Ecol Syst. 1986; 17(1): skeletal characters in mice. J Hered. 1997; 88(2): 85–92, 391–421, doi: 10.1146/annurev.es.17.110186.002135. doi: 10.1093/oxfordjournals.jhered.a023082. 37. Palmer A, Strobeck C, Chippindale A. Bilateral variation and 24. Lenhard W, Lenhard A. Hypothesis tests for comparing correla- the evolutionary origin of macroscopic asymmetries. Ge- tions. Psychometrica. 2014, doi: 10.13140/RG.2.1.2954.1367. netica. 1993; 89(1-3): 201–218, doi: 10.1007/bf02424514. 25. Li J, Zhang Q. Insight into the molecular genetics of 38. Russak OD, Ives L, Mittal VA, et al. Fluctuating derma- myopia. Mol Vis. 2017; 23: 1048–1080, indexed in Pub- toglyphic asymmetries in youth at ultrahigh-risk for med: 29386878. psychotic disorders. Schizophr Res. 2016; 170(2-3): 26. Livshits G, Davidi L, Kobyliansky E, et al. Decreased devel- 301–303, doi: 10.1016/j.schres.2015.12.013, indexed in opmental stability as assessed by fluctuating asymmetry of Pubmed: 26723845. morphometric traits in preterm infants. Am J Med Genet. 39. Sinn R, Wittbrodt J. An eye on eye development. 1988; 29(4): 793–805, doi: 10.1002/ajmg.1320290409, Mech Dev. 2013; 130(6-8): 347–358, doi: 10.1016/j. indexed in Pubmed: 3400724. mod.2013.05.001, indexed in Pubmed: 23684892. 27. Ludwig W. Rechts-Links-Problem im Tierreich und Beim 40. Sokal R. Rohlf F. Biometry. 4th ed. W.H. Freeman, New Menschen. Springer-Verlag, Berlin Heidelberg. 1932. York. 2011. 28. Martín B, Fañanás L, Gutiérrez B, et al. Dermatoglyphic 41. Stini WA. Reduced sexual dimorphism in upper arm muscle profile in 22q deletion syndrome. Am J Med Genet B Neu- circumference associated with protein-deficient diet in ropsychiatr Genet. 2004; 128B(1): 46–49, doi: 10.1002/ a South American population. Am J Phys Anthropol. 1972; ajmg.b.30034, indexed in Pubmed: 15211630. 36(3): 341–351, doi: 10.1002/ajpa.1330360304, indexed 29. Medland SE, Loesch DZ, Mdzewski B, et al. Linkage analy- in Pubmed: 5035060. sis of a model quantitative trait in humans: finger ridge 42. Van Dongen S, Ten Broek C, Bots J, et al. Changes of count shows significant multivariate linkage to 5q14.1. Fluctuating Asymmetry with Age in Human Fetuses and PLoS Genet. 2007; 3(9): 1736–1744, doi: 10.1371/journal. Young Infants. Symmetry. 2017; 9(3): 44, doi: 10.3390/ pgen.0030165, indexed in Pubmed: 17907812. sym9030044. 30. Meier RJ, Goodson CS, Roche EM. Dermatoglyphic devel- 43. Waynforth D. Fluctuating asymmetry and human male life- opment and timing of maturation. Hum Biol. 1987; 59(2): history traits in rural Belize. Proc Biol Sci. 1998; 265(1405): 357–373, indexed in Pubmed: 3596574. 1497–1501, doi: 10.1098/rspb.1998.0463, indexed in 31. Midlo C, Cummins H. Palmar and plantar dermatoglyph- Pubmed: 9744105. ics in primates. Wistar Institute of Anatomy and Biology, 44. Young TL, Metlapally R, Shay AE. Complex trait genet- Philadelphia. 1942. ics of refractive error. Arch Ophthalmol. 2007; 125(1): 32. Moler AP, Swaddle JP. Asymmetry, developmental stability 38–48, doi: 10.1001/archopht.125.1.38, indexed in Pub- and evolution. Oxford University Press, UK. 1997. med: 17210850.

Pancreatic theory of relativity… S. Hać

Department of General Endocrine and Transplant Surgery, Medical University of Gdansk, Poland

[Received: 8 July 2018; Accepted: 2 September 2018]

Pancreatic duct and parenchyma has different benchmarks in nomenclature. The author discusses the proposition to unify the description system of procedures and surgeries within pancreas according to the direction of pancreatic juice natural flow. (Folia Morphol 2019; 78, 2: 431–432)

Key words: pancreas

Pancreas is the organ of digestive tract playing and digestive system relations. An example might be: an important role in fat and protein digestion and “artery stenosis with proximal thrombus formation” glucose metabolism. The gland is composed of two or “bowel obstruction with proximal distension”. The embryo buds. Pancreas is conventionally divided, from matter is not so clear concerning the pancreatic gland the anatomical point of view, into the head, isthmus, and pancreatic duct. The gland has typical excretory corpus and tail, with no strict border between these duct. The description of pancreatic duct occlusion with parts. The pancreas has the typical configuration; glan- proximal distension means that the part of pancreas dular cells form glands with single small duct joining between occlusion and tail is involved. “Dilatation of together to form larger one and all are drained into the pancreatic duct proximal to the tumour…” means main pancreatic duct going along the whole pancreas. left to the tumour or “The abdominal computed to- Within the head of the pancreas, two ducts exist as mography scan revealed dilation of the proximal pan- a consequence of embryo development. Two ducts creatic duct with an irregular high density calcification have usually separate connections with the duodenum. shadow located at the head of the pancreas” [5]. On The lexical definitions of proximal and distal terms the other hand, the resection of pancreas is sometimes are descriptive and may have different meanings. Ac- called “distal pancreatic resection”, which means that cording to Oxford Medical Dictionary [7], ‘proximal’ left part of the gland is removed [1, 9]. Another de- means situated close to the origin or point of attach- scription of pancreatic relations is the following exam- ment or close to the median line of the body. ‘Distal’ is ple: “A 3-cm hypovascular mass accompanying a large situated away from the origin or point of attachment distal pseudocyst in the pancreatic tail” by Hamada or from the median line of the body. The term is applied et al. [3]. In that paper, the authors refer to the same to a part of the limb that is furthest from the body; situation as described by Liu et al. [5]. The part of to a blood vessel that is far from the heart; and to pancreatic duct after head resection is called distal a nerve fibre that is far from the central nervous system. pancreatic duct. The resection of pancreatic head is Proximal and distal — are universally used terms sometime called proximal pancreatic resection, like in describing the position or relation within human body. the paper by Lu et al. [6]: “For tumours in the head There are two main benchmarks; central part of the and uncinate process, the surgeon would palpate the body and direction of flow (blood or other fluids). The tumour with Kocher’s manoeuvre to evaluate the risk first one is used to express the positioning on extremi- and possibility of EU, followed by careful dissection of ties. The second one is used to describe the circulatory proximal pancreas”. Some authors call Santorini duct

Address for correspondence: S. Hać, MD, PhD, Department of General Endocrine and Transplant Surgery, Medical University of Gdansk, ul. Dębinki 7, 80–952 Gdańsk, Poland, e-email: [email protected]

431 Folia Morphol., 2019, Vol. 78, No. 2

as regressed proximal main duct of dorsal pancreatic 91(11): 1410–1427, doi: 10.1002/bjs.4794, indexed in bud [4]. Recent review papers of pancreatic anatomy Pubmed: 15499648. 2. Dimitriou I, Katsourakis A, Nikolaidou E, et al. The main an- also give unclear description: “The ventral pancreatic atomical variations of the pancreatic duct system: review duct and the distal part of the dorsal pancreatic duct of the literature and its importance in surgical practice. are anastomosed and form the MPD of Wirsung, while J Clin Med Res. 2018; 10(5): 370–375, doi: 10.14740/ the proximal part of the dorsal pancreatic duct be- jocmr3344w, indexed in Pubmed: 29581798. 3. Hamada T, Nanashima A, Hiyoshi M, et al. Curative distal comes the APD of Santorini” [2]. This few examples pancreatectomy in patients with acinar cell carcinoma of show dual benchmark related to pancreatic gland. pancreas diagnosed by endoscopic aspiration via esopha- One approach takes into account the pancreatic juice go-jejunostomy: A successful case report. Int J Surg Case flow, while the other one — the position in relation Rep. 2018; 42: 274–279, doi: 10.1016/j.ijscr.2017.12.040, indexed in Pubmed: 29331883. to the central axis of the body. Usually the pancreatic 4. Kamisawa T, Takuma K, Tabata T, et al. Clinical implications juice flow direction is refered by the endoscopists, of accessory pancreatic duct. World J Gastroenterol. 2010; relation to the duodenum is used by surgeons. How- 16(36): 4499–4503, indexed in Pubmed: 20857518. ever the surgical drainage procedures are performed 5. Liu Q, Wang Y, Zeng H, et al. Successful endoscopic removal of a rare, large impacted pancreatic duct stone proximally to the obstruction [8]. using grasping forceps: A case report with video. Med- The problem is not only restricted to nomenclature icine (Baltimore). 2018; 97(14): e0304, doi: 10.1097/ but it may also affect clinical practice. On the other MD.0000000000010304, indexed in Pubmed: 29620654. hand the sentence “Distal pancreatic resection was 6. Lu WJ, Cai HL, Ye MD, et al. Enucleation of non-invasive tumors in the proximal pancreas: indications and outcomes performed, proximal to obstruction site” or “Pancre- compared with standard resections. J Zhejiang Univ Sci B. atic head resection was performed and distal part 2017; 18(10): 906–916, doi: 10.1631/jzus.B1600597, of the gland was anastomosed with jejunal loop” indexed in Pubmed: 28990381. sounds strange. It might be confusing and leads to 7. Martin EA. Oxford Concise Medical Dictionary. 7th edition. Oxford University Press, cop. Market House Books Ltd, misunderstanding. Oxford New York 2004: 567. In my opinion it is reasonable to recommend that 8. Mihaljevic AL, Kleeff J, Friess H. Beger’s operation and the the term ‘proximal pancreas’ refers to the left part of Berne modification: origin and current results. J Hepato- the organ, and the term ‘distal pancreas’ — to its head, biliary Pancreat Sci. 2010; 17(6): 735–744, doi: 10.1007/ s00534-009-0179-2, indexed in Pubmed: 19798464. according to the pancreatic juice natural flow direction. 9. Toh BC, Rao J. Laparoscopic D2 total gastrectomy and en- References mass splenectomy and distal pancreatectomy for locally advanced proximal gastric cancer. Surg Endosc. 2018; 1. Alexakis N, Halloran C, Raraty M, et al. Current stand- 32(4): 2156, doi: 10.1007/s00464-017-5801-9, indexed ards of surgery for pancreatic cancer. Br J Surg. 2004; in Pubmed: 28842726.

Right circumcaval ureter and double right renal vein in the Brazilian shorthair cat (Felis catus): two case reports M. Abidu-Figueiredo1, A.V. Stocco1, C.A. Santos-Sousa1, 2, P. Souza Junior3, L.A.S. Pires4, M.A. Babinski4

1Universidade Federal Rural do Rio de Janeiro, Brazil 2Universidade Federal do Acre, Brazil 3Universidade Federal do Pampa, Brazil 4Universidade Federal Fluminense, Brazil, Brazil

[Received: 29 June 2018; Accepted: 19 August 2018]

Variations of the renal veins are well described in the literature, although variations concerning the ureter are considered a rare finding in cats. The circumcaval ureter is one of the rarest variations of the ureter and is characterised by a loop of the ureter posterior to the caudal vena cava. This variant is also known as preureteral vena cava and retrocaval ureter. It is thought to be caused by a deviation during embryonic development of the aforementioned vein. Due to its rarity, there are scarce reports of the circumcaval ureter in cats, and its association with two renal veins makes it less common as well. These variations should be preoperatively identified in order to avoid complications in kidney transplants, ureteral surgeries and cystoscopies, for instance. The present work aims to report two cases of a circumcaval ureter with two renal veins in two different Brazilian shorthair cats (Felis catus). (Folia Morphol 2019; 78, 2: 433–436)

Key words: anatomical variations, circumcaval ureter, preureteral vena cava, retrocaval ureter, cat, renal vein

INTRODUCTION pedicle and is essential to avoid confusion during The renal pedicle is composed by the ureter, one imaging exams [4, 8]. renal vein and one renal artery. The ureters are ret- On the other hand ureteral variations, such as the roperitoneal conduits that originate from the renal circumcaval ureter (also known as the retrocaval ure- pelvis. The right and left ureters run a lateral, cau- ter) are rarely reported in mammalian species other dal and ventral course in order to reach the urinary than humans. This variant has been associated with bladder [2, 9]. urinary tract infections, haematuria and hydronephro- In domestic mammalian species, both the right sis in humans. However, there is debate whether this and left renal veins drain their contents into the cau- variation can lead to symptoms in animals such as dal vena cava. However, the number, course, and site cats, due to its rarity. Understanding the congenital of drainage of the renal veins are largely variable. An- abnormalities of the ureter may prevent iatrogenic atomical variations of the renal veins are occasionally errors or confusion during surgery and exams such found in radiological studies and the dissecting room, as cystoscopy [2–5]. as knowledge of such variants is necessary prior to the The work presented herein aims to report the execution of vascular procedures involving the renal presence of doubled right renal veins associated with

Address for correspondence: M.A. Babinski, PhD, Universidade Federal Fluminense, Brazil, e-mail: [email protected]

433 Folia Morphol., 2019, Vol. 78, No. 2

Figure 1. Ventral view of the first cat. The right circumcaval ureter (RU) can be seen together with the right renal veins (RRV); RK — right kidney; LK — left kidney; CVC — caudal vena cava; LRV — left renal vein. a circumcaval ureter in two Brazilian shorthair cats After dissection of the retroperitoneal organs, it and discuss its clinical significance. was observed on the first cat two right renal veins and a right type 1 circumcaval ureter (Fig. 1). The CASE REPORT length of both veins were measured with the aid of The current research was approved by the Re- a digital calliper, and the most cranial and ventral vein search Ethics Committee of Rio de Janeiro Federal measured 1.45 cm, while the most caudal and dorsal Rural University (23083.005334/2009-49). vein measured 1.57 cm. The right kidney was 3.61 cm Two male Brazilian shorthair cats obtained from long, 2.32 cm wide and 1.97 cm thick. the Rio de Janeiro Federal Rural University Veterinary On the second cat, two right renal veins were also Hospital necropsy unit were previously identified and observed in association with a right type 1 circum- positioned in the right lateral recumbency. An incision caval ureter (Fig. 2). The length of the most cranial was performed on the thorax and their 6th and 7th ribs and ventral vein was 0.90 cm and the most caudal and were removed. Afterwards, their thoracic aorta was dorsal vein had a length of 2.40 cm. The right kidney minimally dissected, cannulated and further injected was 4.50 cm long, 2.63 cm wide and 2.58 cm thick. with saline solution, followed by a 10% formaldehyde No vascular variations regarding the left kidney solution. were found. Their vascular system was then filled with coloured stained Petrolatex S-65 solution (Refinaria Duque de DISCUSSION Caxias-REDUC-Petrobras, Duque de Caxias-RJ) mixed The most accepted theory regarding the circum- with red dye (Suvinilxadrez®). caval ureter is that this variant originates due to an Subsequently, the cadavers were immersed in embryonic anomaly of the caudal vena cava. Several a low-density polyethylene cardboard box with venous channels fuse and create several anastomoses a 500-litre capacity containing a 10% formaldehyde that eventually give rise to the caudal vena cava. solution to finalise the fixation and latex polymeri- Some of these vessels can be seen near the metane- sation process. Seven days after the latex injection, phros and the developing ureters during ontogenesis the cadavers were washed in running water and the [2, 5, 6]. abdominal cavity was dissected for teaching purpos- The right caudal cardinal veins should fuse and es. Measurements of the kidney and its vessels were disappear, although, when this channel gives origin performed with the aid of a digital calliper. to the infrarenal portion of the caudal vena cava

434 M. Abidu-Figueiredo et al., Right circumcaval ureter and double right renal vein in the Brazilian shorthair cat

Figure 2. Ventral view of the second cat. The right circumcaval ureter (RU) can be seen together with the right renal veins (RRV); RK — right kidney; LK — left kidney; CVC — caudal vena cava. instead of the right supracardinal vein, it causes the As previously stated, this variant in humans is ureter to form a loop on the caudal vena cava [5, 6]. seen more often and may be associated with hydro- In cats, the circumcaval ureter is known to be a nephrosis, haematuria and recurrent urinary tract rare variant with few case reports in the literature infections, although there has been no reports indi- [3, 5], although a study conducted by Bélanger et al. cating this association in cats, although some authors [2] observed the presence of a circumcaval ureter in observed ureteral strictures caused by this variant 35.2% of 301 cats, while the majority of these variants [7, 10]. Nonetheless, it has been related to these were right-sided (30.6%), while only 1.3% of the cats symptoms in dogs [3, 7]. had a left circumcaval ureter. The association of a right circumcaval ureter with In humans, this anatomical variation can be clas- a doubled renal vein is scarcely reported in the liter- sified into two distinct types: type 1, in which the ature. However, the variant ureter can be associated circumcaval ureter crosses posterior to the inferior with other venous anomalies, such as the double vena cava near the third lumbar vertebrae, thus pos- caudal vena cava [2, 3]. sessing an “S-shaped” aspect; and type 2, in which Duplication of the renal veins is an often variation. the aberrant ureter lies horizontally and immediately A study conducted by Rieck and Reis [8] observed that crosses posterior to the inferior vena cava, thus having 26.93% of 1000 cats had this variation, while only a “sickle shaped” aspect. It is known that the type 1 is 1.25% had three renal veins. This duplicity is found the most common between them [1, 2]. Both cats re- more commonly on the right side due to embryolog- ported herein presented the type 1 circumcaval ureter. ical causes [4, 8].

435 Folia Morphol., 2019, Vol. 78, No. 2

The presence of a circumcaval ureter may influ- J Vet Res. 2014; 75(1): 91–95, doi: 10.2460/ajvr.75.1.91, ence diagnostic procedures such as cystoscopy and indexed in Pubmed: 24370251. ureteral surgeries (e. g. ureteral bypass, ureteral stent, 3. Casteleyn C, Cornillie P, Van Cruchten S, et al. Left retrocaval ureter around the ipsilateral limb of a double caudal vena cava uretero-ureteral anastomosis, and others types of in a cat. J Comp Pathol. 2015; 152(4): 313–316, doi: 10.1016/j. anastomosis) [7], while clinically, it may cause lumen jcpa.2015.01.010, indexed in Pubmed: 25798957. obstruction [7, 10]. In addition, the presence of su- 4. Cáceres AV, Zwingenberger AL, Aronson LR, et al. Char- pernumerary renal veins can influence procedures acterization of normal feline renal vascular anatomy with such as transplants [4]. dual-phase CT angiography. Vet Radiol Ultrasound. 2008; 49(4): 350–356, indexed in Pubmed: 18720765. 5. Cornillie P, Baten T, Simoens P. Retrocaval ureter in a cat. Vet CONCLUSIONS Rec. 2006; 159(1): 24–25, indexed in Pubmed: 16816160. In conclusion, the retrocaval ureter is a rare varia- 6. Lesma A, Bocciardi A, Rigatti P. Circumcaval Ureter: Embry- tion with almost no recognition in domestic cats. De- ology. Eur Urol Suppl. 2006; 5(5): 444–448, doi: 10.1016/j. eursup.2006.02.008. spite the fact that it is not associated with symptoms 7. Pey P, Marcon O, Drigo M, et al. Multidetector-row computed in this mammal, the presence of a retrocaval ureter tomographic characteristics of presumed preureteral vena may lead to iatrogenic errors during surgery and con- cava in cats. Vet Radiol Ultrasound. 2015; 56(4): 359–366, fusion during exams such as cystoscopy; therefore, doi: 10.1111/vru.12251, indexed in Pubmed: 25786990. vascular variations and ureteral variations should be 8. Rieck AF, Reis RH. Variations in the pattern of renal vessels and their relation to the type of posterior vena cava in the cat Felis identified pre-operatively. domestica. Am J Anat. 1953; 93(3): 457–474, doi: 10.1002/ aja.1000930307, indexed in Pubmed: 13104340. REFERENCES 9. Testut L, Latarjet A. Tratado de Anatomía Humana. 8th 1. Bateson EM, Atkinson D. Circumcaval ureter: a new clas- ed. Barcelona: Salvat. 1958. sification. Clin Radiol. 1969; 20(2): 173–177, indexed in 10. Zaid MS, Berent AC, Weisse C, et al. Feline ureteral stric- Pubmed: 5771632. tures: 10 cases (2007-2009). J Vet Intern Med. 2011; 25(2): 2. Bélanger R, Shmon CL, Gilbert PJ, et al. Prevalence of cir- 222–229, doi: 10.1111/j.1939-1676.2011.0679.x, indexed cumcaval ureters and double caudal vena cava in cats. Am in Pubmed: 21314722.

Circumaortic left renal vein (circumaortic renal collar) associated with the presence of vascular anomalies: a case series and review of literature R. Haładaj1, M. Polguj2, G. Wysiadecki1, A. Żytkowski3, M. Topol1

1Department of Normal and Clinical Anatomy, Interfaculty Chair of Anatomy and Histology, Medical University of Lodz, Poland 2Department of Angiology, Interfaculty Chair of Anatomy and Histology, Medical University of Lodz, Poland 3Department of Biomechanics and Prosthetic-Orthopaedic Supply, Medical University of Lodz, Poland

[Received: 20 July 2018; Accepted: 27 July 2018]

Renal vessels exhibit a high degree of anatomical variations in terms of their number, level of origin, diameter and topographical relationships. In particular, it applies to the left renal vein which can take retroaortic or even circumaortic placement. Anatomical variations of the left renal vein may be of great clinical significance, particularly in the case of renal transplantation, retroperitoneal surgery as well as vascular or diagnostic procedures. Thus, the aim of this report was to present a complete anatomical description of two cases of the circumaortic left renal vein (CLRV; circumaortic renal collar) co-existing with the presence of various vascular anomalies. In the first case, the circumaortic renal collar was connected via a large anastomosis with the hemiazygos vein and was associated with the presence of the supernumerary left renal artery located below the main left renal artery. In the second case, the circumaortic renal collar was accompanied by the renal artery dividing close to its origin. Moreover, in the latter case, the fusiform aneurysm of the abdominal aorta was observed. In both cases, the CLRV began as a single and short trunk. On its further course, the initial segment of the CLRV was divided into two limbs — anterior (anterior left renal vein) and posterior (posterior left renal vein). Both anterior and posterior limb of the CLRV opened into the inferior vena cava. (Folia Morphol 2019; 78, 2: 437–443)

Key words: accessory renal vessels, anatomical variation, aneurysm, kidney, left renal vein, supernumerary renal arteries

INTRODUCTION importance to practitioners. Typically, a single renal Normal anatomical relationships between renal artery and vein is present on both the right and the vessels, abdominal aorta and inferior vena cava (IVC) left side. Each of the renal veins is located anterior to were illustrated for the first time in 1564 in the Eus- the appropriate renal artery and opens into the IVC tachi’s treatise entitled Opuscula anatomica. Anatom- [28]. However, there are numerous clinically relevant ical variants of the renal arteries and veins were also deviations from this arrangement [5]. shown for the first time in the illustrations contained Renal vessels demonstrate huge anatomical var- there [21]. Due to the development of advanced iability as regards to their number, level of origin, surgical and diagnostic procedures the issue of ana- diameter and topographical relationships [2, 8–10, tomical variations of renal vessels remains of utmost 17, 19, 25, 26, 30–36]. In particular, it applies to the

Address for correspondence: Dr. R. Haładaj, Department of Normal and Clinical Anatomy, Interfaculty Chair of Anatomy and Histology, Medical University of Lodz, ul. Narutowicza 60, 90–136 Łódź, Poland, tel: +48 42 630 49 49, e-mail: [email protected]

437 Folia Morphol., 2019, Vol. 78, No. 2

left renal vein which can present retroaortic [1, 10, performed in situ with a Digimatic digital calliper 12, 15] or even circumaortic placement [7, 11, 13, 20, (Mitutoyo Company, Kawasaki-shi, Kanagawa, Japan). 27]. Thus, the left renal vein may be a single vessel or Circumaortic left renal vein began as a single, double — with two limbs: one located anterior and the short (29 mm) trunk originating from a few tribu- second posterior to the abdominal aorta [5, 19, 33]. taries, 13 mm from the renal hilum (Fig. 1). This part Knowledge of the anatomical variations of renal of the left renal vein was in contact with the anterior vessels may be crucial during retroperitoneal surgery, surface of the psoas major muscle and occupied an renal transplants, as well as vascular or diagnostic antero-inferior position in relation to the main trunk procedures [3, 4, 7–11, 14, 16, 18, 19, 24, 30]. This of the left renal artery. On its further course, 42 mm applies in particular to the left kidney. Due to the fact from the renal hilum, the initial segment of CLRV was that the left renal vein is about three times longer divided into two limbs — anterior (anterior left renal than the right one, the left kidney is the preferred side vein [aLRV]; Fig. 1A, B) and posterior (posterior left for live donor nephrectomy [28]. Thus, different var- renal vein [pLRV]; Fig. 1B). Furthermore, the presence iations of the left renal vein, including the circumao- of the left accessory renal artery was observed below rtic renal collars, should be recorded and analysed, the main left renal artery (Fig. 1A). On its initial course especially in the context of their co-existence with the accessory artery ran posterior to the aLRV and at vascular variations. So, the aim of this case report the distance of 27 mm from the renal hilum arched was to present a complete anatomical description of over the left renal vein and then coursed along its two cases of the circumaortic left renal vein (CLRV; anterior aspect (Fig. 1A). The length (measured from circumaortic renal collar) associated with the presence the origin to the renal pelvis) of both renal arteries of the atypical arrangement of renal arteries. In the was respectively 57 mm for the main left renal artery first case, the circumaortic renal collar gave a large and 66 mm for the accessory left renal artery. The root to the hemiazygos vein and was accompanied by diameters of both arteries were 6.14 mm for the main the presence of the accessory (supernumerary) left re- trunk and 3.32 mm for the accessory one. nal artery running below the main left renal artery. In Both limbs of CLRV (namely aLRV and pLRV) ran the second case, the circumaortic renal collar was ac- obliquely crossing, respectively, the anterior (aLRV) or companied by renal artery dividing close to its origin. posterior (pLRV) aspect of the abdominal aorta. Anteri- Moreover, in the latter case the fusiform aneurysm or limb of CLRV (i.e. aLRV) crossed the anterior surface of the abdominal aorta was observed. Embryological of the abdominal aorta at the level of the inferior background and clinical significance of the observed border of LIII vertebra, draining into IVC at the level of variation have also been discussed. the intervertebral disc between LIII and LIV vertebrae. Posterior limb of CLRV (i.e. pLRV) crossed the LII ver- CASE REPORT tebra and opened into the IVC at the level of the su- Case no. 1. The circumaortic renal collar perior border of this vertebra. The length of aLRV was accompanied by the presence of the 51 mm. The diameter of aLRV was 10.45 mm before supernumerary left renal artery receiving the left suprarenal vein (of 4.32 mm diame- A 73-year-old female cadaver was subjected to ter), and 11.23 after receiving the left suprarenal vein dissection for teaching and research purposes. During (Fig. 1). Moreover, the left ovarian vein (of 2.41 mm the dissection of the posterior abdominal wall, the diameter) drained into the anterior surface of initial kidneys and renal vessels were exposed on both sides. part of aLRV (Fig. 1). The length of pLRV was 56 mm. Careful inspection of renal veins arrangements and The large anastomosis (of 7.71 mm diameter) with the tributaries revealed the presence of accessory (super- hemiazygos vein took origin from the upper surface numerary) left renal vein which displayed retroaortic of pLRV (14 mm from the origin of pLRV and 56 mm course. After a detailed inspection, the variation was from the renal pelvis; Figs. 1, 2). The diameters of classified as CLRV (circumaortic renal collar). The sub- pLRV measured proximal and distal to the anastomosis sequent stages of the dissection were subordinated with hemiazygos vein were 6.81 mm and 8.14 mm, to a thorough examination of the observed vascu- respectively. No venous valves were found in both the lar variations including measurements of the vessels CLRV and in the anastomosis between pLRV and the diameters and also examination for the presence hemiazygos vein. The renal vessels on the right side of possible venous valves. The measurements were were single and showed typical arrangement. The

438 R. Haładaj et al., CLRV (circumaortic renal collar) associated with the presence of vascular anomalies

Figure 1. Case no. 1. The circumaortic renal collar anastomosing with the hemiazygos vein and associated with the presence of the accessory left renal artery; A. Anterior view showing general topographical relationships within the dissected area. The left kidney has been exposed; B. Anterior view after partial removal of the abdominal aorta showing the location and course of the posterior limb (posterior left renal vein) of the circumaortic renal collar; aLRV — anterior left renal vein; aRA — accessory (supernumerary) left renal artery; AA — abdominal aorta; IMA — inferior mesenteric artery; IVC — inferior vena cava; OV — left ovarian vein; pLRV — posterior left renal vein; RA — left renal artery; SV — left suprarenal vein; U — ureter; *anastomosis between the posterior left renal vein and the hemiazygos vein.

diameter (measured at the midpoint of length) of the from a few tributaries, 9 mm from the renal hilum. right renal artery was 7.18 mm and the diameter of This part of the left renal vein was crossed (22 mm the right renal vein was 9.91 mm. from the renal hilum) by one of the anterior branches (4.06 mm of diameter) of the left renal artery (Fig. 3). Case no. 2. The circumaortic renal collar On its further course, 26 mm from the renal hilum, accompanied by the left renal artery the initial segment of CLRV was divided into the aLRV dividing close to its origin and pLRV. The length measured from the origin of The second case of the persistent circumaortic re- the left renal artery to the renal pelvis was 55 mm, nal collar was observed during dissection of the block while the place of division of the left renal artery into of abdominal organs harvested from the male body anterior and posterior branches was located 22 mm donor of unknown age. Also in this case the CLRV from its origin. The diameter of the main trunk of the began as a short, single trunk (17 mm) originating left renal artery was 7.62 mm.

439 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. Case no. 1. The circumaortic renal collar anastomosing with the hemiazygos vein. Anterior view after partial removal of the abdominal aorta showing the anastomosis between the posterior left renal vein and the hemiazygos vein (inferior vena cava has been cut and reflected together with the anterior left renal vein). Magnification of the anastomosis (region marked by dotted rectangle) has been shown in the right upper corner; AA — abdominal aorta; aLRV — anterior left renal vein; aRA — accessory left renal artery (cut and reflected); IVC — inferior vena cava; pLRV — posterior left renal vein; RA — left renal artery; RRV — right renal vein; U — left ureter; *anastomosis between the posterior left renal vein and the hemiazygos vein.

Both aLRV and pLRV ran obliquely crossing the renal artery occupied the most anterior position with- anterior or posterior aspect of the abdominal aorta, in the renal hilum). Additionally, the presence of the respectively (Fig. 3). The length of aLRV was 61 mm, fusiform aneurysm of the abdominal aorta located its diameter was 9.6 mm before receiving the left between the level of the pLRV and inferior mesenteric suprarenal vein (of 5.8 mm diameter), and 10.68 mm artery was revealed during the dissection. after receiving the left suprarenal vein. The length of pLRV was 63 mm. The small anastomosis (of 3.32 mm DISCUSSION diameter) with the hemiazygos vein took origin from The initial venous channels undergo several stages the upper surface of pLRV (19 mm from the origin of of remodelling early in embryonic development [23, pLRV and 45 mm from the renal pelvis; Fig. 3). The 28]. As the embryo increases in size, the postcardinal diameters of pLRV measured proximal and distal to veins, draining the body walls, are supplemented the anastomosis with hemiazygos vein were 7.13 mm by longitudinal vascular channels transforming into: and 7.66 mm, respectively. Moreover, the left gonadal subcardinal, supracardinal, azygos line, subcentral vein (of 4.55 mm diameter) drained into the anterior and precostal veins [28]. These venous channels anas- surface of initial part of pLRV (Fig. 3). No venous valves tomose with both posterior cardinal system and with were found both in the CLRV and in the anastomosis each other. The drainage of developing mesonephros between pLRV and the hemiazygos vein. The renal is provided mainly by the subcardinal veins. Subcardi- vessels on the right side were single. The diameter nal veins are connected by the pre-aortic anastomotic (measured at the midpoint of length) of the right re- plexus [28]. At the latter stages of development, the nal artery was 8.28 mm and the diameter of the right plexus is involved in the formation of the left renal renal vein was 12.59 mm. However, similarly to the vein, located anterior to the abdominal aorta. A series left side, the right renal vein was also crossed by the of complex changes leads to that end. Before the anterior branch of the right renal artery (of 4.34 mm renal vessels develop in their final form, a circumao- diameter). Thus, on both sides of the examined spec- rtic renal venous ring, known as ‘renal collar’, exists imen, the anomalous arrangement of hilar structures around the aorta. The renal collar is formed by: com- was observed (in which the anterior branch of the munications between subcardinal veins (anastomotic

440 R. Haładaj et al., CLRV (circumaortic renal collar) associated with the presence of vascular anomalies

Figure 3. Case no. 2. The circumaortic renal collar accompanied by the left renal artery dividing close to its origin. Anterior view of the isolated specimen. In this case, the circumaortic renal collar coexisted with the left renal artery dividing close to its origin. On both sides of the specimen, the anomalous arrangement of hilar structures was observed, in which the anterior branch of the renal artery occupied the most anterior position within the renal hilum; AA — abdominal aorta; abLRA — anterior branch of the left renal artery; abRRA — anterior branch of the right renal artery; aLRV — anterior left renal vein; CT — celiac trunk; GV — left gonadal vein; IMA — inferior mesenteric artery; IVC — inferior vena cava; pLRV — posterior left renal vein; SMA — superior mesenteric artery; SV — left suprarenal vein; U — ureter; *anastomosis between the posterior left renal vein and the hemiazygos vein. plexus mentioned above), communications between mated at 0.3% to 6.3% [25, 26, 29, 32]. According supracardinal veins, and supracardinal-subcardinal to Bergman’s Comprehensive Encyclopaedia of Hu- anastomoses [20, 28]. During normal formation of man Anatomic Variation, CLRV is observed in 1.02% the left renal vein, the dorsal part of the circumaortic of cases [33]. However, Bergman et al. [5] in their renal venous ring degenerates and the ventral vessel earlier work reported a wider range of frequency of becomes the renal vein. However, occasionally, the the circumaortic renal venous collar — from 1.5% post-aortic part of the renal collar may persist. Due to up to 8.7%. The circumaortic renal collar occurs with such complex developmental relationships and condi- a similar frequency in both sexes (48.1% in male tions, the number, orifices, and main tributaries of the and 51.9% female subjects with persistent renal renal veins may exhibit numerous variations. Also, the collar) [33]. Also, no significant correlation between arterial blood supply of both the kidneys and adrenal gender and variations of the left renal vein was ob- glands is extremely variable during the foetal period, served in the study of Boyaci et al. [6]. A study with which may give rise to several anatomical variations a multidetector computed tomography performed observed in the postnatal period and in adults [22]. in an adult population by Zhu et al. [36] confirmed, Anatomical variations of renal veins differ in terms that the frequency of right renal vein variations was of the prevalence depending on the side. Multiple re- significantly higher than the frequency of left renal nal veins are frequent on the right side (frequency to vein variations. Furthermore, in this study no sta- 28%), whereas they are less frequent on the left side tistically significant correlation was found between (frequency from 1% to 9%) [33]. The left renal vein variations of the renal vein and gender. Based on may be doubled. In this case, one limb of the left re- the morphology of the renal veins, Zhu et al. [36] nal veins runs anterior and the other runs posterior to classified variations of the left renal vein into five the abdominal aorta before opening into the IVC [5]. types. The CLRV was classified as the type I, and its This condition may be referred to as persistence of frequency was estimated at 2.1%, (31 out of 1452 the renal collar. The approximate prevalence of this patients who underwent multidetector computed variation differs among various authors and is esti- tomography angiography) [36].

441 Folia Morphol., 2019, Vol. 78, No. 2

Persistent renal collar may occur in numerous vari- different levels, which is similar to our observations. ations, especially when communications with neigh- According to Trigaux et al. [32] the mean distance be- bouring veins are taken into consideration [7, 11, 13, tween the openings of the retroaortic and preaortic left 20, 25-29, 32]. In the case described by Panagar et al. renal veins into the IVC was 39.0 ± 17.4 mm. The aLRV [20] the left preaortic renal vein receives left suprare- may also be absent. In such cases, the single retroaortic nal and left ovarian veins as tributaries, while the left left renal vein replaces the typical left renal vein [19, posterior renal vein ran obliquely downwards to open 29]. Anjamrooz et al. [1] described another case of the into IVC after being joined by lumbar veins. Thus, left renal vein bifurcating into two branches passing drainage of lumbar vein may be anomalous in the behind the aorta. Before its division, the left renal vein cases of persistent renal collar. In such cases the lum- received the left suprarenal and testicular veins. bar vein may open into retroaortic renal vein [20, 34]. Occasionally, the renal collar is also detected in Terayama et al. [30] described another case of double patients with different disorders during diagnostic pro- left renal vein, in which, in addition to the primary left cedures. Garg et al. [7], basing on computed tomogra- renal vein, the presence of a pLRV draining to the left phy angiography, described the presence of incidental ascending lumbar vein without communicating with finding of CLRV with “gross aneurysmal dilatation the IVC and other renal veins was observed. Our case of both pre- and retro-aortic part of the renal vein”. no. 1 differs from those previously described, as the These anomalies were found in a patient with delayed posterior renal vein was directly connected with the presentation of post-traumatic aortic pseudoaneurysm initial part of the hemiazygos vein via a strong anas- and its fistulous communication with the right renal tomosis (root). Thejodhar et al. [31], in turn, reported vein after a gunshot injury. Lee et al. [13] described the co-existence of circumaortic renal collar, accessory a case of double left renal veins in the form of a ve- renal artery and an anomalous arrangement of hilar nous collar in a 15-year-old girl who presented with structures on the left side in the same cadaver. The intermittent haematuria. Nishibe et al. [18] described atypical hilar arrangement presented in Thejodhar et the combination of abdominal aortic aneurysms and al.’s case [31] reminds our descriptions, except that in congenital anomalies of the IVC and its tributaries our case no. 2 renal arteries were single on both sides such as double IVC, left-sided IVC, circumaortic renal and were divided close to their origin, especially on collar and retroaortic renal vein. In our case no. 2, the left side. In addition, in our case no. 2, the unusual the fusiform aneurysm of the abdominal aorta was hilar arrangement occurred both on the right and found beginning at the level of the pLRV, which could the left side. Thus, the persistent renal collar may be potentially cause the difficulties in the operating field. accompanied by atypical topographical relationships In the group of patients with persistent renal collar between renal vessels, as well as by the possibility the risk of venous injury and subsequent bleeding of co-occurrence of arterial variations. According to increases, since typically the retroaortic component of Bergman et al. [5] the renal veins show less variation the left renal vein does not occur [7, 20, 24–26, 29]. than the renal arteries. The frequency of double renal Furthermore, the situation may be complicated by artery is estimated at 10% of cases [5]. Wherein, so- the presence of additional anastomoses reaching the called accessory (or supernumerary) renal arteries aris- pLRV. According to Bergman et al. [5] the root of the ing below the usual trunk (similar to those described hemiazygos vein originating from the left renal vein is in our case no. 1) are less frequent than those arising observed in 88% of the cases. However, the authors of above [5, 8]. For the kidney transplantation team, it is this report, making a detailed analysis of the literature, crucial to consider significant variations in the atypical did not find the description of the circumaortic renal hilar arrangement during the time of organ harvest collar giving a very strong root to the hemiazygos [3, 24]. This fact was reflected in the study of Bachul vein, like in our case no 1. In both cases described in et al. [3] who described “crossing anatomic barriers” our study the anastomosis with the hemiazygos vein during the transplantation of a kidney with five ar- was present. Thus, awareness of the variations in the teries, duplication of the pyelocalyceal system, and blood supply of kidneys is mandatory [2]. a double ureter. Sabouri et al. [24] in turn reported retro-aortic inverted left renal vein in a renal donor. CONCLUSIONS The openings of pre-aortic and retro-aortic left The persistent renal collar may be accompanied by renal veins in the renal collar are usually located at atypical topographical relationships, atypical connec-

442 R. Haładaj et al., CLRV (circumaortic renal collar) associated with the presence of vascular anomalies

tions to the adjacent veins as well as the possibility vena cava. J Anat. 2001; 199(Pt 5): 621–623, indexed in Pubmed: 11760894. of co-occurrence of arterial variations. Anatomical 16. Nam JK, Park SW, Lee SD, et al. The clinical significance of variations of renal vessels, due to their diversity and a retroaortic left renal vein. Korean J Urol. 2010; 51(4): 276–280, doi: 10.4111/kju.2010.51.4.276, indexed in Pubmed: 20428432. high prevalence, should be taken into account both 17. Nayak SB, Sirasanagandla SR, Shetty SD, et al. Multiple vascular variations at the vicinity of the left kidney. Anat Sci Int. 2013; in planning and conducting diagnostic and surgical 88(4): 230–233, doi: 10.1007/s12565-013-0177-y, indexed in medical procedures. Pubmed: 23649479. 18. Nishibe T, Sato M, Kondo Y, et al. Abdominal aortic aneurysm with left-sided inferior vena cava. Report of a case. Int Angiol. Acknowledgements 2004; 23(4): 400–402, indexed in Pubmed: 15767987. 19. Orczyk K, Wysiadecki G, Majos A, et al. What each clinical The authors wish to express their gratitude to all anatomist has to know about left renal vein entrapment syndrome (nutcracker syndrome): a review of the most those who donated their bodies to medical science. important findings. Biomed Res Int. 2017; 2017: 1746570, doi: 10.1155/2017/1746570, indexed in Pubmed: 29376066. References 20. Panagar AD, Subhash RL, Suresh BS, et al. Circumaortic left renal vein-a rare case report. J Clin Diagn Res. 2014; 8(3): 1. Anjamrooz SH, Azari H, Abedinzadeh M. Abnormal patterns of 111–112, doi: 10.7860/JCDR/2014/6634.4122, indexed in the renal veins. Anat Cell Biol. 2012; 45(1): 57–61, doi: 10.5115/ Pubmed: 24783096. acb.2012.45.1.57, indexed in Pubmed: 22536553. 21. Persaud TVN, Loukas M, Tubbs RSA. History of Human Anato- 2. Aristotle S, Felicia C. Anatomical study of variations in my. Second Edition. Charles C Thomas Publisher, Springfield. the blood supply of kidneys. J Clin Diagn Res. 2013; 7(8): 2014; 206. 1555–1557, doi: 10.7860/JCDR/2013/6230.3203, indexed in 22. Pityński K, Skawina A, Polakiewicz J, et al. Extraorganic vascular Pubmed: 24086837. system of adrenal glands in human fetuses. Ann Anat. 1998; 3. Bachul PJ, Osuch C, Chang ES, et al. Crossing anatomic barri- 180(4): 361–368, doi: 10.1016/S0940-9602(98)80045-9, ers-transplantation of a kidney with 5 arteries, duplication of indexed in Pubmed: 9728279. the pyelocalyceal system, and double ureter. Cell Transplant. 23. Polguj M, Szubert W, Topol M, et al. An unusual duplication 2017; 26(10): 1669–1672, doi: 10.1177/0963689717722169, of the inferior vena cava in a patient with endovascular repair indexed in Pubmed: 28933184. for abdominal aortic aneurysm. Rom J Morphol Embryol. 4. Bartle EJ, Pearce WH, Sun JH, et al. Infrarenal venous anomalies 2015; 56(2 Suppl): 875–878, indexed in Pubmed: 26429190. and aortic surgery: avoiding vascular injury. J Vasc Surg. 1987; 24. Sabouri S, Hosseini A, Shivaei SS. Retro-aortic inverted left 6(6): 590–593, indexed in Pubmed: 3694757. renal vein: a rare anomaly found in a renal donor. Iran J Radiol. 5. Bergman RA, Afifi AK, Miyauchi R (2015) Illustrated encyclo- 2015; 12(2): e11374, doi: 10.5812/iranjradiol.11374, indexed pedia of human anatomic variation: Opus II: cardiovascular in Pubmed: 25901255. system: Veins: Abdomen. http://www.anatomyatlases.org/ 25. Satyapal KS. Classification of the drainage patterns of the AnatomicVariants/Cardiovascular/Text/Veins/RenalSplenic renal veins. J Anat. 1995; 186 (Pt 2): 329–333, indexed in Testicular.shtml. Pubmed: 7649831. 6. Boyaci N, Karakas E, Dokumacı DS, et al. Evaluation of left 26. Satyapal KS, Kalideen JM, Haffejee AA, et al. Left renal vein renal vein and inferior vena cava variations through routine variations. Surg Radiol Anat. 1999; 21(1): 77–81, indexed in abdominal multi-slice computed tomography. Folia Morphol. Pubmed: 10370998. 2014; 73(2): 159–163, doi: 10.5603/FM.2014.0017, indexed 27. Senecail B, Bobeuf J, Forlodou P, et al. Two rare anomalies in Pubmed: 24902094. of the left renal vein. Surg Radiol Anat. 2003; 25(5-6): 7. Garg L, Jain N, Agrawal S, et al. Juxtarenal aortic pseudoan- 465–467, doi: 10.1007/s00276-003-0164-4, indexed in Pu- eurysm — right renal vein fistula with circumaortic renal bmed: 13680187. collar-delayed manifestation of a gunshot injury: an uncom- 28. Standring S. ed. Gray’s Anatomy: the Anatomical Basis of mon entity diagnosed with CT angiography. Pol J Radiol. Clinical Practice. 40th ed. Churchill Livingstone, Edinburgh, 2016; 81: 114–119, doi: 10.12659/PJR.896032, indexed in London. 2008: 207-208, 1233. Pubmed: 27057262. 29. Tatar I. Retroaortic and circumaortic left renal veins with their 8. Gulas E, Wysiadecki G, Cecot T, et al. Accessory (mul- CT findings and review of the literature. Anatomy. 2008; 2: tiple) renal arteries - Differences in frequency accord- 72–76, doi: 10.2399/ana.08.072. ing to population, visualizing techniques and stage 30. Terayama H, Yi SQ, Shoji S, et al. Anomalous connection of the of morphological development. Vascular. 2016; 24(5): left posterior renal vein with the left ascending lumbar vein 531–537, doi: 10.1177/1708538116631223, indexed in in a Japanese cadaver. Folia Morphol. 2015; 74(4): 544–547, Pubmed: 26945775. doi: 10.5603/FM.2015.0121, indexed in Pubmed: 26620520. 9. Gulas E, Wysiadecki G, Szymański J, et al. Morphological and 31. Thejodhar P, Kumar Po, Rao MS, et al. Co-existence of cir- clinical aspects of the occurrence of accessory (multiple) renal cumaortic renal collar, accessory renal artery and anomalous arteries. Arch Med Sci. 2018; 14(2): 442–453, doi: 10.5114/ arrangement of hilar structures in the same cadaver: A case aoms.2015.55203, indexed in Pubmed: 29593819. report. Eur J Anat. 2007; 11: 185–188. 10. Karkos CD, Bruce IA, Thomson GJ, et al. Retroaortic left renal 32. Trigaux JP, Vandroogenbroek S, De Wispelaere JF, et al. Con- vein and its implications in abdominal aortic surgery. Ann Vasc genital anomalies of the inferior vena cava and left renal vein: Surg. 2001; 15(6): 703–708, doi: 10.1007/s10016-001-0022-y, evaluation with spiral CT. J Vasc Interv Radiol. 1998; 9(2): indexed in Pubmed: 11769156. 339–345, indexed in Pubmed: 9540920. 11. Kottra JJ, Castellino RA. The circumaortic left renal vein. 33. Tubbs RS, Shoja MM, Loukas M. Bergman’s Compre- Angiographic appearance. Radiology. 1970; 95(1): 141–143, hensive Encyclopedia of Human Anatomic Variation. doi: 10.1148/95.1.141, indexed in Pubmed: 4906778. Wiley-Blackwell, Hoboken, New Jersey. 2016: 890–893, 12. Kyung DS, Lee JH, Shin DY, et al. The double retro-aortic left doi: 10.1002/9781118430309. renal vein. Anat Cell Biol. 2012; 45(4): 282–284, doi: 10.5115/ 34. Yao Y, Okada Y, Yamato M, et al. Communicating vein between acb.2012.45.4.282, indexed in Pubmed: 23301196. the left renal vein and left ascending lumber vein: incidence 13. Lee CM, Ng SH, Ko SF, et al. Circumaortic left renal vein: report and significance on abdominal CT. Radiat Med. 2003; 21(6): of a case. J Formos Med Assoc. 1992; 91(3): 356–358, indexed 252–257, indexed in Pubmed: 14743898. in Pubmed: 1354702. 35. Yi SQ, Ueno Y, Naito M, et al. The three most common varia- 14. Luo L, Liu YN, Zhang Y, et al. An easy and effective method tions of the left renal vein: a review and meta-analysis. Surg to locate renal vein during retroperitoneal laparoscopic rad- Radiol Anat. 2012; 34(9): 799–804, doi: 10.1007/s00276-012- ical nephrectomy: single-center experience. Med Sci Monit. 0968-1, indexed in Pubmed: 22535303. 2018; 24: 5147–5151, doi: 10.12659/MSM.911199, indexed 36. Zhu J, Zhang L, Yang Z, et al. Classification of the renal vein in Pubmed: 30040793. variations: a study with multidetector computed tomography. 15. Macchi V, Parenti A, De Caro R. Draining of a retroaortic Surg Radiol Anat. 2015; 37(6): 667–675, doi: 10.1007/s00276- left renal vein via the ‘subcentral veins’ into the inferior 014-1403-6, indexed in Pubmed: 25567101.

An atypical biceps brachii and coracobrachialis muscles associated with multiple neurovascular aberrations: a case report with clinical significance M. Piagkou1, T. Totlis2, N. Anastasopoulos2, N. Lazaridis2, K. Natsis2

1Department of Anatomy, School of Medicine, Faculty of Health Sciences, National and Kapodistrian University of Athens, Greece 2Department of Anatomy and Surgical Anatomy, Faculty of Health Sciences, Aristotle University of Thessaloniki, Greece

[Received: 8 July 2018; Accepted: 29 July 2018]

Neural and vascular variations in the axilla and upper limb area are usually paired, but coexistence of muscular aberration on top of this is uncommon. The current case report emphasizes on the unilateral coexistence of a three-headed (tricipital) biceps brachii muscle, a two-headed coracobrachialis with an accessory muscle bundle joining the superficial and deep heads of coracobrachialis muscle. On the ipsilateral side of the 72-year-old male cadaver, a connecting branch originated from the musculocutaneous nerve and joined the median nerve after surpassing the accessory muscle bundle. A large diameter subscapular trunk originated from the 2nd part of the axillary artery and after giving off the 1st lateral thoracic artery trifurcated into a common stem which gave off the 2nd and 3rd lateral thoracic arteries, the circumflex scapular artery and a common branch that gave off the 4th and 5th lateral thoracic arteries and the thoracodorsal artery, as the ultimate branch. All lateral thoracic arteries were accompanied by multiple intercostobrachial nerves. Documentation of such muscular and neurovascular variants and their embryologic origin increases awareness of their potential impact on diagnosis and treatment of upper limb pathology. To the best of our knowledge, the currently reported cadaveric observations seem to constitute a unique finding. (Folia Morphol 2019; 78, 2: 444–449)

Key words: biceps brachii, variation, coracobrachialis, lateral thoracic artery, intercostobrachial nerve, anastomosis, median nerve, musculocutaneous nerve

INTRODUCTION radial tuberosity. Furthermore, BB is quite variable The biceps brachii muscle (BB) is variable con- with regard to morphology, innervation and number cerning the origin and insertion of its typical heads. (3–7) of accessory heads among different populations In the majority of the reported cases, the long BB [3, 17, 20]. The existence of a 3rd BB head is usually head arises from the superior glenoid labrum and accompanied by axillary and brachial vessels with the supraglenoid tubercle [1], while the short head abnormal origin, branching pattern and course [15], arises from the apex of the coracoid process. The two variations in the musculocutaneous nerve (MCN) and heads join to form a single belly to be inserted to the coexistence of other multiple accessory muscles or

Address for correspondence: M. Piagkou, DDS, MD, MSc, PhD, Assistant Professor of Anatomy, School of Medicine, Faculty of Health Sciences, National and Kapodistrian University of Athens, 75 MikrasAsias Street, 11527 Athens, Greece, e-mail: [email protected]

444 M. Piagkou et al., Muscular and neurovascular aberrations

Figure 1. A. A three-headed biceps brachii muscle (BB) with the long, short and accessory heads (LHBB, SHBB and AHBB) combined with a two-headed coracobrachialis muscle (CB) with an accessory muscular bundle; C. The musculocutaneous nerve (MCN) connecting branch (***) initially coursed under the accessory muscular bundle, then advanced superficially to the brachial artery (BA) to ultimately reach and merge with the median nerve (MN); B. MCN innervated and BA supplied the atypical BB (**); UN — ulnar nerve. diversities of the coracobrachialis muscle (CB) [14]. bination of muscular, neural and vascular variations Knowledge of anatomical variants in neurovascular was detected in the right axilla and upper arm. The and muscular structures is fundamental in interpret- cadaver was donated to the Department of Anat- ing imaging studies or unusual clinical manifestations. omy and Surgical Anatomy of the Medical School The current case report emphasizes on the coex- of the Aristotle University of Thessaloniki, through istence of a right-sided tricipital (three-headed) BB, the “Anatomical Gift Programme” after a written a bicipital CB which also bears an interconnecting informed consent was obtained. A three-headed BB accessory muscle bundle (AMB) joining the super- was detected. The long head originated from the ficial with the deep head of the muscle. Neurovas- superior glenoid labrum and the supraglenoid tuber- cular aberrations coexisted on the ipsilateral side of cle similarly to Alashkham et al. study [1], the short a 72-year-old Greek Caucasian male formalin fixed head from the apex of the coracoid process and the cadaver. Documentation of such muscular and neuro- accessory head from the tendon of the short head of vascular variants along with their embryologic origin BB and CB. The accessory BB head after joining short explanation increases the awareness among clinicians and long heads inserted onto the posterior surface regarding their potential impact on diagnosis and of the radial tuberosity. Innervation and vascular sup- treatment of upper limb pathology. Complications ply originated from the MCN and the brachial artery, during surgical exploration of the axilla and upper arm respectively (Fig. 1). A bicipital (with superficial and could be avoided by keeping in mind these variants. deep heads) CB coexisted with a connecting branch To the best of our knowledge, the reported observa- joining the MCN with the median nerve (MN). This tions, combined on a cadaver, appear to constitute branch originated from the MCN. An AMB extended a unique finding. between superficial and deep CB head. The MCN connecting branch surpassed the muscle bundle, CASE REPORT on its way to the MN. The remaining MCN typically During the routine dissection of a 72-year-old coursed between superficial and deep head of CB, formalin fixed Greek male cadaver, an unusual com- supplying muscular branches to the anterior compart- 445 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. The tricipital biceps brachii muscle (BB) with the long, short and accessory heads (LHBB, SHBB and AHBB) combined with a bicipital coracobrachialis muscle (CB) (superficial and deep head SHCB and DHCB) with an accessory muscular bundle (white asterisks). Connect- ing branch (yellow asterisks) joining musculocutaneous nerve (MCN) with the median nerve (MN). MCN innervated and brachial artery (BA) supplied the atypical BB; UN — ulnar nerve; MBCN — medial brachial cutaneous nerve; LR — lateral root; LC — lateral cord.

ment of the arm (Fig. 2). The superior thoracic artery (7.1%), Chinese (8%), Africans (12%) and Turks (15%) was absent and a large diameter subscapular arterial and the highest in Colombians (37.5%), South Af- trunk (SST) originated just below the thoracoacromial ricans (21.5%) and Japanese (18%) [3, 6, 8, 9, 14, trunk’s origin (2nd part of the axillary artery). The SST, 21, 24]. Accessory heads, unilaterally or bilaterally just after the origin of the 1st lateral thoracic artery found [13], may originate from the coracoid process (LTA), trifurcated into a common stem giving off the and humeral shaft, the tendon or fascia of pectoralis 2nd and 3rd LTA, the circumflex scapular artery (CSA) minor or major, the deltoid muscle or CB insertion, and a long branch that gave off the 4th and 5th LTA, the intermuscular septum, the greater tuberosity or and the thoracodorsal artery (TDA) as the ultimate the shoulder joint capsule [14, 23] and the anterior branch. All LTAs were accompanied by multiple in- surface of brachialis muscle [18]. Although a male tercostobrachial nerves (ICBN). The 1st LTA after its predominance was referred concerning the occur- origin coursed anterior to the 1st ICBN and the 2nd, rence of the 3rd head [2], no gender dimorphism exists 3rd, 4th and 5th LTAs originated via a common trunk [8, 9, 13, 22]. Asymmetry has been reported by several with the CSA and the TDA (Fig. 3). The left axilla and authors [21] who predominantly reported a right- upper arm didn’t show any aberration. The present sided 3rd head, similarly to the current case, while study fully conforms to the provisions of the Helsinki others mentioned left side supremacy [2, 5]. Bilateral Declaration of 1995 (as revised in Edinburgh 2000) symmetrical accessory BB heads have an incidence and has been approved by the Ethics Committee ranging between 3.3% and 29.3% [3, 9, 13, 21]. All of the Medical School of the Aristotle University of reported supernumerary BB heads are innervated by Thessaloniki, Greece. the MCN, with the rare exceptions where accessory heads are innervated by the MN [8, 9, 11, 16]. DISCUSSION In humans, where long head of CB is absent, the Accessory heads of BB occur with an incidence third BB head, which arises in continuity with CB ranging from 0.5% to 37.5% [8, 12, 20–22]. Spe- insertion, may represent a remnant of the CB long cifically, the incidence of the three-headed BB var- head [22]. Muscles of the anterior arm compartment ies between 0.85% [14] and 20.5% [3, 7] with the develop from myogenic precursor cells that arise from lowest prevalence in Europeans (0.85–10%), Indians ventral dermomyotome of somites. Molecular chang-

446 M. Piagkou et al., Muscular and neurovascular aberrations

brachial plexus take an aberrant course, such as the connecting branch observed in the current study due to failure in differentiation [10]. Occasionally, the adjacent CB may have accessory muscle bundles inserting to the medial intermuscular septum, medial epicondyle, flexor carpi radialis, brachialis, brachioradialis or the pronator teres [13]. Course and direction of accessory heads or muscle bundles may occasionally entrap the MN and brachial artery [16, 19]. The formation of a musculoaponeurotic bundle built up by fibres from BB and CB, extending to the medial intermuscular septum and brachial fascia, may be a causative factor for forearm and hand neuro-vasculopathy [19]. MN entrapment and irritation may lead to tingling and neuropathic pain, followed or accompanied by reduced sensation or complete numbness. Compression of the brachial vessels may lead to ischaemia and oedema [19]. The accessory BB heads have been associated with MCN variations, such as an interconnecting branch with the MN [22], which was also found in our case. In such cases, the MCN course and branching pattern Figure 3. The subscapular arterial trunk (SST) coexisted with is usually affected and possible nerve entrapment may 5 lateral thoracic arteries (LTAs) and 3 intercostobrachial nerves ensue. Rare exceptions of MN innervation have been (ICBNs). The 1st LTA coursed anterior to the 1st ICBN and the 2nd, 3rd, 4th and 5th LTA originated via a common trunk with the circum- also reported [16]. The MCN-MN interconnections are flex scapular artery (CSA) and the thoracodorsal artery (TDA); highly important to anaesthesiologists performing MCN — musculocutaneous nerve; MN — median nerve; MBCN — selective motor nerve blocks, to surgeons dealing medial brachial cutaneous nerve. with upper limb trauma management, as well as to clinicians and neurologists involved in nerve impairment treatment. es in these cells induce muscle development. Further Presence of accessory BB heads and accompany- growth of muscle occurs by fusion of myoblasts and ing muscular aberrations may be combined with myotubes. Variation of muscle patterns may be the multiple neurovascular variations, as in the current product of altered signalling or stimulus between case where multiple LTAs and ICBNs were detected. mesenchymal cells. A first hypothesis supports that Five LTAs originated from the SST and the TDA. This the accessory BB heads may appear due to piercing of rare variation is significant in clinical and surgical BB by the MCN, thus resulting into a longitudinal split- practice and was also referred by Loukas et al. [12] ting of myotubules, which eventually will be covered in 3.09%. LTA variability was classified in types, in by a connective tissue layer producing a separate belly an effort to highlight the commonest (LTA from the [11]. Developmental studies described the variation thoracoacromial trunk, 67.62%) and rarest (from the of the 3rd BB head, as a portion of brachialis muscle axillary — 17.02%, the TDA — 5% and the subscapu- supplied by the MCN, in which its distal insertion lar artery — 3.93%) variations. LTA point of origin has been translocated from the ulna to the radius, is important, since the artery should remain intact according to the functional adaptation hypothesis. during reconstructive procedures involving head and Circulatory factors during brachial plexus formation neck malignancies, modified radical mastectomy, may affect the development of accessory heads [17]. mastopexy, breast reconstruction and reduction [12]. Aberrant muscular formations may be remnants of Concerning the ICBN variability, double, triple muscle primordia that failed to disappear or fuse to and multiple ICBNs with their possible anastomoses form a sole structure [19]. According to embryological have been reported [10]. Usually ICBN is called the studies, it is possible that some of the fibres of the lateral cutaneous branch of the 2nd intercostal nerve.

447 Folia Morphol., 2019, Vol. 78, No. 2

A 2nd ICBN often emanates from the anterior part of REFERENCES rd the 3 lateral cutaneous nerve supplying the axilla and 1. Alashkham A, Alraddadi A, Soames R. Origin of the long the medial side of the arm. The ICBN may be preserved head of biceps brachii from the supraglenoid tubercle and during radical mastectomy to prevent sensory loss or glenoid labrum. Eur J Anat. 2018; 22: 213–219. 2. Ashraf Nasr Y, Adel Hussein M. Morphology and clinical dysesthesia in the upper arm area [10]. Bratschi and implication of the extra-head of biceps brachii muscle. Haller [4] emphasized on the significance of the ICBN, Folia Morphol. 2013; 72(4): 349–356, doi: 10.5603/ since a sensory loss or regional dysesthesia appears in fm.2013.0058. approximately half of the performed breast cancer pro- 3. Asvat R, Candler P, Sarmiento EE. High incidence of the cedures and axillary lymphadenectomies (47.5%). Less third head of biceps brachii in South African populations. J Anat. 1993; 182 (Pt 1): 101–104, indexed in frequently, neural dysfunction may ensue when ap- Pubmed: 8509293. plying regional anaesthesia in upper limb procedures, 4. Bratschi HU, Haller U. [Significance of the intercostobra- with tourniquet compression, mechanical pressure chial nerve in axillary lymph node excision]. Geburtshilfe therapy devices, entrapment and/or nerve traction [10]. Frauenheilkd. 1990; 50(9): 689–693, doi: 10.1055/s-2008- 1026346, indexed in Pubmed: 2272434. From the clinical point of view, the supernumerary 5. Cheema P, Singla R. Lower incidence of the third head of heads of BB usually remain asymptomatic, incidentally biceps brachii in North Indian Population. J Clin Diagnostic found during imaging studies or surgery. Radiologists Res. 2011; 5(Suppl.2): 1323–1326. should take into consideration the possibility of an 6. Kervancioglu P, Orhan M. An anatomical study on the three-headed biceps brachii in human foetuses, and accessory head and the accompanied neurovascular clinical relevance. Folia Morphol. 2011; 70(2): 116–120, variations when assessing computed tomography and indexed in Pubmed: 21630233. magnetic resonance imaging scans, as these features 7. Kopuz C, Içten N, Yildirim M. A rare accessory coracobra- can be misinterpreted [24]. Clinicians should be also chialis muscle: a review of the literature. Surg Radiol Anat. alert when dealing with patients complaining of weak- 2003; 24(6): 406–410, doi: 10.1007/s00276-002-0079-5, indexed in Pubmed: 12652369. ness or pain in the anterior arm or when performing 8. Kopuz C, Sancak B, Ozbenli S. On the incidence of third selective motor nerve blocks to treat nerve impairment head of biceps brachii in Turkish neonates and adults. cases [20]. Neural and vascular supply of the acces- Kaibogaku Zasshi. 1999; 74(3): 301–305, indexed in sory heads is important to plastic and reconstructive Pubmed: 10429374. 9. Kosugi K, Shibata S, Yamashita H. Supernumerary head of surgeons [20]. Particularly, if the accessory head occurs biceps brachii and branching pattern of the musculocu- unilaterally [6], it might be misinterpreted as a soft tis- taneus nerve in Japanese. Surg Radiol Anat. 1992; 14(2): sue tumour due to the asymmetry on palpation [16]. 175–185, indexed in Pubmed: 1641744. Accessory heads may also provide an additional force 10. Kumar PA, Reddy DRK, Bapuji P. Multiple Intercostobra- for supination and flexion of the forearm or they may chial Nerves. J Evolution Med Dental Sciences. 2014; 57: 12978–12983. cause unusual displacement of the humeral fragments 11. Lokanadham S, Subhadra Devi V. Unusual presentation if a fracture occurs. The medical records of the body of supernumerary head of biceps brachii muscle in South donor don’t mention any symptoms related to the Indian population. World J Med Sciences. 2011; 6: 115. underlying variations. 12. Loukas M, du Plessis M, Owens DG, et al. The lateral thoracic artery revisited. Surg Radiol Anat. 2013; 36(6): 543–549, doi: 10.1007/s00276-013-1234-x, indexed in CONCLUSIONS Pubmed: 24281130. Accessory heads and muscular slips may alter 13. Nakatani T, Tanaka S, Mizukami S. Bilateral four-headed anatomy of the arm, since they may be combined with biceps brachii muscles: the median nerve and brachial artery passing through a tunnel formed by a muscle slip unexpected multiple neural and vascular variations from the accessory head. Clin Anat. 1998; 11(3): 209–212, predisposing to complications due to their potential doi: 10.1002/(SICI)1098-2353(1998)11:3<209::AID- compressive action on nerves and superficially cross- CA10>3.0.CO;2-N, indexed in Pubmed: 9579595. ing brachial vessels. Clinicians should consider such 14. Natsis K, Vlasis K, Tsakotos G, et al. case of a supernu- anatomical variations in the differential diagnosis of merary third head of the biceps brachii muscle - clinical significance. Aristotle University Med J. 2010; 37: 39–41. patients presenting with ischaemia, oedema or MN 15. Natsis K, Piagkou M, Panagiotopoulos NA, et al. An palsy symptoms. unusual high bifurcation and variable branching of the axillary artery in a Greek male cadaver. Springerplus. Acknowledgements 2014; 3: 640, doi: 10.1186/2193-1801-3-640, indexed in Pubmed: 25392808. The authors express their thanks to the body do- 16. Nayak SR, Krishnamurthy A, Kumar MN, et al. Four- nor and its family: without them this study would headed biceps and triceps brachii muscles, with neuro- not have been possible. vascular variation. Anat Sci Int. 2008; 83(2): 107–111, 448 M. Piagkou et al., Muscular and neurovascular aberrations

doi: 10.1111/j.1447-073X.2007.00171.x, indexed in 21. Rincón F, Rodríguez Z, Sánchez A, et al. The anatomic Pubmed: 18507620. characteristics of the third head of biceps branchii muscle 17. Nayak SR, Krishnamurthy A, Prabhu LV, et al. Multiple in a colombian population. Rev Chis Anat. 2002; 20(2): supernumerary muscles of the arm and its clinical signifi- 197–200, doi: 10.4067/s0716-98682002000200014. cance. Bratisl Lek Listy. 2008; 109(2): 74–76, indexed in 22. Rodríguez-Niedenführ M, Vázquez T, Choi D, et al. Su- Pubmed: 18457314. pernumerary humeral heads of the biceps brachii muscle 18. Ozan H, Atsev A. Sianau, A, Simsek, C, Basar R. An unusual revisited. Clin Anat. 2003; 16(3): 197–203, doi: 10.1002/ insertion of the accessory biceps brachii muscle. Am Anat ca.10060, indexed in Pubmed: 12673814. Nippon. 1997; 72: 515–519. 23. Sargon MF, Tuncali D, Celik HH. An unusual origin 19. Paraskevas G, Natsis K, Ioannidis O, et al. Accessory mus- for the accessory head of biceps brachii muscle. Clin cles in the lower part of the anterior compartment of the Anat. 1996; 9(3): 160–162, doi: 10.1002/(SICI)1098- arm that may entrap neurovascular elements. Clin Anat. 2353(1996)9:3<160::AID-CA4>3.0.CO;2-K, indexed in 2008; 21(3): 246–251, doi: 10.1002/ca.20608, indexed in Pubmed: 8740475. Pubmed: 18351653. 24. Vollala VR, Nagabhooshana S, Bhat SM, et al. Multiple 20. Rai R, Ranade AV, Prabhu LV, et al. Third head of biceps arterial, neural and muscular variations in upper limb of brachii in an Indian population. Singapore Med J. 2007; a single cadaver. Rom J Morphol Embryol. 2009; 50(1): 48(10): 929–931, indexed in Pubmed: 17909679. 129–135, indexed in Pubmed: 19221659.

Unusual arterial supply of the segment IV with interlobar bridge and right replaced hepatic artery: a case report M. Tiryakioğlu1, T. Koç2, N.S. İlgi1

1Department of Anatomy, Faculty of Medicine, Near East University, Nicosia, Cyprus 2Department of Anatomy, Faculty of Medicine, Mersin University, Mersin, Turkey

[Received: 21 May 2018; Accepted: 29 July 2018]

A replaced right hepatic artery (RHA) arising from the superior mesenteric artery and an interlobar parenchymal bridge over the sagittal fissure have been observed on a 64-year-old formalin-fixed male cadaver in the anatomy laboratory. As we had followed a detailed segmental anatomy, we encountered an arterial distribution of segment IV featuring a different pattern from the literature so far. According to our observations, the segment I is supplied by both left (LHA) and middle (MHA) hepatic arteries; the segments II and III are supplied by the LHA while the segment IV is supplied by both the MHA and replaced RHA. The segments V–VIII are supplied only by the replaced RHA. The case emphasizes the importance of arterial variations of liver in terms of the surgical procedures during the liver transplantation, hepatic resections, hepatic tumours, and etc. Our discussion focuses to the arterial supply of the segment IV and possible complications it may cause during/after the liver operations. (Folia Morphol 2019; 78, 2: 450–454)

Key words: replaced right hepatic artery, liver, anatomy, cadaver

INTRODUCTION artery (RHA) supplies the four segments: V, VI, VII, and During the 8th and 10th gestational weeks, hepatic VIII; the left hepatic artery (LHA) supplies the lateral arteries start to be visible in the hepatic hilum and superior and lateral inferior segments: II and III, and take their last form in the liver parenchyma in the end the middle hepatic artery (MHA) supplies the medial of 10th gestational week [7]. According to Couinaud superior and medial inferior segments: segment IV. classification, three main lobes are described during The caudate lobe (segment I) is supplied by both RHA the embryological stage as the lateral (segment II), and LHA [5]. medial-anterior (segments III, IV, V, and VIII), and the In the literature, hepatic arteries have many varia- posterior lobe (segments VI and VII). All of the lobes tions in terms of their origins. Two main classifications supplied by an embryonic artery of its own during the have been described so far: replaced [2, 4, 6, 12, development of hepatic formations which are named 16] and accessory hepatic arteries [11, 12, 16]. The as the embryonic left, medial, and right hepatic ar- accessory hepatic artery is present in addition to the teries [9]. Development of the intrahepatic arterial main artery of the right and/or left lobes [2, 16] while formation is parallel to the order of the portal and the replaced hepatic artery originates from another biliary systems [7]. In the adult liver, the right hepatic artery instead of the proper hepatic artery (usually

Address for correspondence: Dr. M. Tiryakioğlu, Near East University, Faculty of Medicine, Department of Anatomy, Nicosia, Cyprus, e-mail: [email protected]

450 M. Tiryakioğlu et al., Right replaced hepatic artery

from the superior mesenteric artery or the coeliac trunk) [2, 12, 16]. The vascular segmental distribution of liver has critical importance in liver transplantation [10, 14]; particularly the segment IV of liver is at greater risk for development of ischaemia, necrosis, and hepatic artery thrombosis as a post-transplant/ post-operational complication [1, 14]. In this particular report, we aimed to emphasize the arterial supply of segment IV and replaced RHA (rRHA) with their clinical and morphometric features besides drawing attention to an interlobar parenchymal bridge running between the right and left lobes of liver on its visceral surface. Figure 1. Schematic drawing showing the segmental distribution CASE REPORT of hepatic arteries; I–VIII — segments of liver; CHA — common hepatic artery; CT — coeliac trunk; GA — gastroduodenal artery; A rRHA (Figs. 1, 2) and an interlobar parenchymal LHA — left hepatic artery; MHA — middle hepatic artery; rRHA — bridge over the sagittal fissure (Fig. 3) have been en- replaced right hepatic artery; SMA — superior mesenteric artery; countered during a routine abdominal dissection of Triangle — Calot’s triangle. 64-year-old male cadaver in the Anatomy Department of the Near East University, Faculty of Medicine. The cadaver identification and medical history did not show artery separated into two branches on its course; one any of the diseases, abdominal surgeries, or cirrhosis. ran superiorly and supplied the anterior surface of the In our case the common hepatic artery (CHA) gallbladder and it ended in the right lobe while the originated from the coeliac trunk (CT) and passed other branch supplied the cystic duct and reached the anterior to the portal vein (PV) before it gave off the gallbladder, inferiorly (Fig. 1). In our case the arterial LHA and the gastroduodenal artery (GA) (Figs. 1, 2). supply to the segment IV was showing a significant Its length and diameter were measured as 22 mm difference as it is supplied by the branches from both and 4 mm, respectively. MHA and the rRHA that was originated from the The LHA arose from the CHA with a diameter superior mesenteric artery. of 3 mm and run anterior to the PV. It gave off the In addition to the deviation on the arterial supply MHA 4 mm distal to its origin and ramified into three of the segment IV, there was an interlobar parenchy- terminal branches. The terminal branches of the LHA mal bridge running between the right and left hepatic entered into the segments II (dia = 2 mm) and III lobes on its visceral surface right behind the sagittal (dia = 2 mm) of the left lobe. The LHA also gave off fissure. It was 2 cm long and 2 cm inferior to porta small branches (dia = 0.8 mm) to the caudate lobe hepatis as appeared on Figure 3. The bridge can also before it gave off the terminal branches. The length be described like it was where the sagittal fissure and diameter of the MHA were 41 mm and 2 mm, missing or became a tunnel with an uninterrupted respectively. The MHA gave five branches off: one of parenchyma in between the right and left lobes. them ran through the umbilical fissure; three of them We could not trace the arterial structure within the entered to the segment IV, and one branch entered bridge, but when we split the tissue parallel to the to the caudate lobe (segment I) (Figs. 1, 2). sagittal fissure, we observed that the round ligament The rRHA occurred as arising 20 mm distal to the of liver was located right in front of it. origin of the SMA, posterior to the PV, and coursed lateral to the PV. The length and diameter of the rRHA DISCUSSION were measured as 77 mm and 3.5 mm, respectively. In the literature, 10 variant subtypes of the hepatic It divided into two thick branches in Calot’s triangle; arterial system were reported by Michels [16] and one of these branches (with dia = 2 mm) entered into 6 subtypes were described by Hiatt et al. [12]. Our the liver segments V–VIII while the other one gave case is in accordance with both Michels’s and Hiatt’s off the cystic artery (dia = 1.5 mm) and entered into classification as type III (as the RHA is being originated the segment IV (dia = 2 mm) (Figs. 1, 2). The cystic from the superior mesenteric artery).

451 Folia Morphol., 2019, Vol. 78, No. 2

Figure 2. Visceral surface of liver showing the hepatic arteries; AA — abdominal aorta; CA — cystic artery; CHA — common hepatic artery; CL — caudate lobe; CT — coeliac trunk; GA — gastroduodenal artery; GB — gallbladder; IMV — inferior mesenteric vein; IVC — inferior vena cava; LHA — left hepatic artery; LL — left lobe; MHA — middle hepatic artery; PV — portal vein; QL — quadrate lobe; RL — right lobe; rRHA — replaced right hepatic artery; SA — splenic artery; SMA — superior mesenteric artery; SMV — superior mesenteric vein; SV — splenic vein.

Figure 3. Photograph showing the interlobar bridge over the sagittal fissure on the visceral surface of liver; AA — abdominal aorta; CA — cystic artery; CHA — common hepatic artery; CL — caudate lobe; CT — coeliac trunk; FL — falciform ligament; GA — gastroduodenal artery; GB — gallbladder; IVC — inferior vena cava; LHA — left hepatic artery; LL — left lobe; MHA — middle hepatic artery; RL — right lobe; rRHA — replaced right hepatic artery; SA — splenic artery; SMA — superior mesenteric artery; red circular line showing the interlobar bridge.

452 M. Tiryakioğlu et al., Right replaced hepatic artery

The vascular segmental distribution of liver has for surgical/invasive approaches to the liver both in liv- critical importance during the liver transplantations ing donor and split liver transplantations. The present [1, 5, 10, 21]; particularly the segment IV is at greater case with the segment IV supplied by both the rRHA risk as it may cause a hepatic artery thrombosis as and MHA is of great importance due to their possible a post-transplant complication [10, 13, 14]. The MHA implications in liver transplantations/surgeries. There generally supplies the segment IV [10]. In the litera- is also a possibility to interrupt the blood supply to ture, researchers reported that the MHA originated the liver when there is pancreaticoduodenectomy in about equal proportions from the RHA and the attempted to the patients with rRHA [15, 21, 22]. It LHA [12, 21] whereas some studies showed that the should always be kept in mind that the origin of the MHA originates from the LHA more often than from hepatic artery and/or its branches may show many the RHA [17, 20]. Wang et al. [21] reported that MHA variations and should be under consideration before- could originate from the RHA in the presence of a re- hand the planned surgery. Also, using a multidetector placed LHA as well as from the LHA in the presence of computed tomography angiography can detect the rRHA as seen in our case. Recent studies show that the rare hepatic arterial vascularisation [18]. MHA diverged from the LHA in 54–61.5% of the cases The parenchymal variations like the one we have while originated from the RHA in 27.5–34% of the reported in this case should be kept in mind in terms cases [17, 20], also the MHA originated from both the of possible vascular variations associated their de- RHA and LHA in 5.5% of the cases [17]. The origin of velopment. the MHA may also be very important for some of the clinical practices such as a right lobe resection and/or Acknowledgements pancreaticoduodenectomy operations. Interestingly, We would like to commemorate the cadaver donor MHA originated from the LHA could also originate for this study. Also, we are grateful to our students, from the RHA, especially in the presence of a replaced Dilara Unalan and Ali Eren Ozturk for their endless LHA according to Azoulay et al. [3]. helps for taking the pictures and recording the video In our case, both rRHA and MHA occurred as sup- during our cadaver dissection. plying the segment IV indicating a possible increased risk of intraoperative injury to the MHA during right/ References left lobe living donor liver transplantation. Further- 1. Alghamdi T, Viebahn C, Justinger C, et al. Arterial blood more, complications of surgery such as bleeding supply of liver segment IV and its possible surgical consequences. Am J Transplant. 2017; 17(4): 1064–1070, would be eliminated by acknowledging the relation- doi: 10.1111/ajt.14089, indexed in Pubmed: 27775870. ship of rRHA with the PV [19]. In our case, rRHA 2. Allendorf JD, Bellemare S. Reconstruction of the replaced coursed posterior to the PV first and then it coursed right hepatic artery at the time of pancreaticoduo- laterally and superiorly as it goes farther from it. denectomy. J Gastrointest Surg. 2009; 13(3): 555–557, doi: 10.1007/s11605-008-0578-8, indexed in Pub- As for the parenchymal interlobar bridge, it med: 18642051. seemed to be a continuous parenchymal tissue jump- 3. Azoulay D, Castaing D, Adam R, et al. Transplantation of ing over the sagittal fissure in between the right and three adult patients with one cadaveric graft: wait or in- left lobes creating a gross anatomical variation, yet novate. Liver Transpl. 2000; 6(2): 239–240, doi: 10.1002/ we could not trace its vascular structure. The sagittal lt.500060208, indexed in Pubmed: 10719027. 4. Beger O, Elvan Ö, Kurtoğlu Z. Replaced right hepatic ar- fissure partially has become a 2 cm long tunnel right tery and its segmental distribution. Anatomy. 2016; 9(3): in front of the parenchymal bridge, 2 cm inferior to 185–188, doi: 10.2399/ana.15.019. the porta hepatis. We have come across a similar 5. Bismuth H. Surgical anatomy and anatomical surgery case in a report of Ebby et al. [9]. They reported the of the liver. World J Surg. 1982; 6(1): 3–9, doi: 10.1007/ bf01656368. sagittal fissure have become a complete tunnel hav- 6. Caliskan E, Acar T, Ozturk M, et al. Coeliac trunk and com- ing the round ligament in it with a missing quadrate mon hepatic artery variations in children: an analysis with lobe [8]. Our case shows a normal quadrate lobe as computed tomography angiography. Folia Morphol. 2018; seen in Figure 3. 77(4): 670–676, doi: 10.5603/FM.a2018.0037, indexed in Pubmed: 29651794. 7. Collardeau-Frachon S, Scoazec JY. Vascular development CONCLUSIONS and differentiation during human liver organogenesis. The origin, diameter, length, course and segmen- Anat Rec (Hoboken). 2008; 291(6): 614–627, doi: 10.1002/ tal distribution of the hepatic arteries are important ar.20679, indexed in Pubmed: 18484606.

453 Folia Morphol., 2019, Vol. 78, No. 2

8. Couinaud C. Le foie. Etudes anatomiques et chirurgicales. 16. Michels NA. Newer anatomy of the liver and its variant Masson, Paris 1957. blood supply and collateral circulation. Am J Surg. 1966; 9. Ebby S, Ambike MV. Anatomical variation of ligamentum 112(3): 337–347, indexed in Pubmed: 5917302. teres of liver: a case report. WebmedCentral ANATOMY. 17. Onishi H, Kawarada Y, Das BC, et al. Surgical anatomy of 2012; 3(5): WMC003389. the medial segment (S4) of the liver with special reference 10. Ghosh SK. Variations in the origin of middle hepatic to bile ducts and vessels. Hepatogastroenterology. 2000; artery: a cadaveric study and implications for living 47(31): 143–150, indexed in Pubmed: 10690598. donor liver transplantation. Anat Cell Biol. 2014; 47(3): 18. Skórzewska A, Stajgis P, Grzymisławska M, et al. Rare varia- 188–195, doi: 10.5115/acb.2014.47.3.188, indexed in tions of hepatic arteries in association with variable origin of Pubmed: 25276478. gastroduodenal artery found in multidetector computed to- 11. Gurgacz AM, Horbaczewska A, Klimek-Piotrowska W, et mography angiography. Folia Morphol. 2014; 73(4): 531–535, al. Variations in hepatic vascularisation: lack of a proper doi: 10.5603/FM.2014.0082, indexed in Pubmed: 25448917. hepatic artery. Two case reports. Folia Morphol. 2011; 19. Sureka B, Mittal MK, Mittal A, et al. Variations of ce- 70(2): 130–134, indexed in Pubmed: 21630235. liac axis, common hepatic artery and its branches in 12. Hiatt JR, Gabbay J, Busuttil RW. Surgical anatomy of the 600 patients. Indian J Radiol Imaging. 2013; 23(3): hepatic arteries in 1000 cases. Ann Surg. 1994; 220(1): 223–233, doi: 10.4103/0971-3026.120273, indexed in 50–52, indexed in Pubmed: 8024358. Pubmed: 24347852. 13. Holbert BL, Baron RL, Dodd GD. Hepatic infarction caused by 20. Suzuki H. [Correlation and anomalies of the vascular struc- arterial insufficiency: spectrum and evolution of CT findings. ture in Glisson’s area around the hepatic hilum, from the AJR Am J Roentgenol. 1996; 166(4): 815–820, doi: 10.2214/ standpoint of hepatobiliary surgery]. Nihon Geka Hokan. ajr.166.4.8610556, indexed in Pubmed: 8610556. 1982; 51(5): 713–731, indexed in Pubmed: 7159135. 14. Jin GYu, Yu HC, Lim HS, et al. Anatomical variations of 21. Wang S, He X, Li Z, et al. Characterization of the mid- the origin of the segment 4 hepatic artery and their clini- dle hepatic artery and its relevance to living donor liver cal implications. Liver Transpl. 2008; 14(8): 1180–1184, transplantation. Liver Transpl. 2010; 16(6): 736–741, doi: 10.1002/lt.21494, indexed in Pubmed: 18668651. doi: 10.1002/lt.22082, indexed in Pubmed: 20517907. 15. Kim JiH, Gonzalez-Heredia R, Daskalaki D, et al. Totally replaced 22. Yang SH, Yin YHu, Jang JY, et al. Assessment of hepatic arte- right hepatic artery in pancreaticoduodenectomy: is this ana- rial anatomy in keeping with preservation of the vasculature tomical condition a contraindication to minimally invasive while performing pancreatoduodenectomy: an opinion. surgery? HPB (Oxford). 2016; 18(7): 580–585, doi: 10.1016/j. World J Surg. 2007; 31(12): 2384–2391, doi: 10.1007/ hpb.2016.04.009, indexed in Pubmed: 27346138. s00268-007-9246-5, indexed in Pubmed: 17922256.

454 An international multidisciplinary journal devoted to fundamental research in the morphological sciences Official Journal of the Polish Anatomical Society Investigation of the styloid process length in a Greek population...... 378 N. Zokaris, I. Siska, K. Natsis, M. Piagkou, N. Lazaridis, A. Skolka, E. Piehslinger (a Constituent Member of European Federation for Experimental Morphology — EFEM) Studies on clinical anatomy of the maxillofacial and mandibular regions of the Madras EDITOR-IN-CHIEF Red sheep (Ovis aries) in India...... 389 Janusz Moryś V. Sundaram, P. Dharani, R. Gnanadevi, R. Kavya Department of Anatomy and Neurobiology Accessory head of the flexor pollicis longus muscle: anatomical study and clinical significance.....394 Medical University of Gdańsk D.R. Ballesteros, P.L. Forero, L.E. Ballesteros http://journals.viamedica.pl/folia_morphologica Age-related structural-mechanical property changes in human peroneus longus muscle...... 401 A. Niyomchan, B. Panichareon, A. Siriphorn, T. Wongtawatchai See our website for information on manuscript status, Differences in foetal topographical anatomy between insertion sites of the iliopsoas aims and scope, instructions for authors and gluteus medius muscles into the proximal femur: a consideration of femoral torsion...... 408 as well as editorial board. P. Zhao, Z.W. Jin, J.H. Kim, H. Abe, G. Murakami, J.F. Rodríguez-Vázquez Blind mole rat (Spalax leucodon) masseter muscle: structure, homology, diversification Folia Morphologica and nomenclature ...... 419 A. Yoldas, M. Demir, R. İlgun, M.O. Dayan Publishing, Subscription and Advertising Office: VM Media sp. z o.o. VM Group sp.k., Grupa Via Medica Dermatoglyphics as a possible morphological biomarker in myopia: ul. Świętokrzyska 73, 80–180 Gdańsk, Poland analysis of finger ridge counts and fluctuating asymmetry...... 425 L. Sretić, N. Labus, T. Filipović, M. Filipović tel. (+48 58) 320 94 94, fax (+48 58) 320 94 60 Managing editor SHORT COMMUNICATION Joanna Niezgoda e-mail: [email protected] Pancreatic theory of relativity… ...... 431 Cover designer S. Hać Sylwia Scisłowska CASE REPORTS The journal is published at: www.fm.viamedica.pl in one volume per year consisting of four numbers. Right circumcaval ureter and double right renal vein in the Brazilian shorthair cat (Felis catus): The annual price for access to electronic version of the journal is 120 PLN for individual subscribers and 240 PLN two case reports...... 433 for institutions from Poland and 40 EURO for foreign individual subscribers and 80 EURO for foreign institu- M. Abidu-Figueiredo, A.V. Stocco, C.A. Santos-Sousa, P. Souza Junior, L.A.S. Pires, M.A. Babinski tions. Payment should be made to: VM Media sp. z o.o. VM Group sp.k., Grupa Via Medica, Bank BGŻ Paribas SA account number: 15 1600 1303 0004 1007 1035 9021; SWIFT: PPABPLPK. Electronic orders option available at: Circumaortic left renal vein (circumaortic renal collar) associated with the presence www.fm.viamedica.pl of vascular anomalies: a case series and review of literature...... 437 The publisher must be notified of a cancellation of access to electronic version not later then two months before R. Haładaj, M. Polguj, G. Wysiadecki, A. Żytkowski, M. Topol the end of a calendar year. After that date electronic access will be automatically prolonged for another year. An atypical biceps brachii and coracobrachialis muscles associated with Advertising. For details on media opportunities within this electronic version of journal please contact multiple neurovascular aberrations: a case report with clinical significance...... 444 the advertising sales department, ul. Świętokrzyska 73, 80–180 Gdańsk, Poland, tel: (+48 58) 320 94 57, M. Piagkou, T. Totlis, N. Anastasopoulos, N. Lazaridis, K. Natsis e-mail: [email protected] Unusual arterial supply of the segment IV with interlobar bridge and right replaced The editors accept no responsibility for advertisement contents. hepatic artery: a case report...... 450 Folia Morphologica is the official journal of the Polish Anatomical Society. For information about the Society, M. Tiryakioğlu, T. Koç, N.S. İlgi please contact: Prof. Marek Grzybiak, Department of Clinical Anatomy, Medical University of Gdansk, ul. Dębinki 1, 80–211 Gdańsk, Poland, tel: +48 58 349 14 22, e-mail: [email protected] All rights reserved, including translation into foreign languages. No part of this periodical, either text or illustration, may be used in any form whatsoever. It is particularly forbidden for any part of this material to be copied or translated into a mechanical or electronic language and also to be recorded in whatever form, stored in any kind of retrieval system or transmitted, whether in an electronic or mechanical form or with the aid of photocopying, microfilm, recording, scanning or in any other form, without the prior written permission of the publisher. The rights of the publisher are protected by national copyright laws and by international conventions, and their violation will be punishable by penal sanctions. Editorial policies and author guidelines are published on journal website: http://journals.viamedica.pl/folia_morphologica Legal note: http://journals.viamedica.pl/FM/about/legalNote Folia Morphologica is indexed by: BIOSIS Previews, CAS, CINAHL, CrossRef, Dental Abstracts, EBSCO, Elsevier BIOBASE, EMBIOLOGY, FMJ, Google Scholar, Index Copernicus (143.41), Index Medicus/MEDLINE, Index Scholar, Ministry of Science and Higher Education (15), NCBI/National Center for Biotechnology Information, Polish Medical Bibliography, Scopus, SJR, Thomson Reuters, Thomson Scientific Products — Biological Abstracts, Ulrich’s Periodicals Directory, Veterinary Bulletin, WorldCat and Zoological Record. Position in Index Copernicus ranking systems is available at: www.indexcopernicus.com. Current Impact Factor of Folia Morphologica (2017) is 0.497.

ISSN 0015–5659 ISSN 0015–5659 Impact Factor: 0.497

Folia Morphol. Vol. 78, No. 2, May 2019 2018 Vol. 77 No. 2, pp. 171–408 2019, Vol. 78, No. 2, pp. 221–454 CONTENTS