Royal College of Surgeons of Thailand Secretariat Office : Royal Golden Jubilee Building, 2 Soi Soonvijai, New Petchburi Road, Bangkok 10310, Thailand
OFFICERS 2015 - 2017
President : Parinya Thavichaigarn President-Elect : Tanaphon Maipang Vice President : Pramook Mutirangura Secretary General : Rattaplee Pak-Art Assistant Scretary : Sutdhachit Linananda Treasurer : Preecha Siritongtaworn Past Presidents : Udom Poshakrisna Sem Pring-Puang-Geo Kasarn Chartikavanij Charas Suwanwela Thira Limsila Kijja Sindhvananda Arun Pausawasdi Kitti Yensudchai Thongueb Uttaravichien Chomchark Chuntrasakul Prinya Sakiyalak Vithya Vathanophas Naronk Rodwarna Nopadol Wora-Urai Soottiporn Chittmittrapap Vajarabhongsa Bhudhisawasdi
BOARD OF DIRECTORS
Songchai Simaroj Representative of General Surgeons Sani Molkul Representative of Pediatric Surgeons Sirachai Jindarak Representative of Plastic & Reconstructive Surgeons Taweesak Chotivatanapong Representative of Thoracic Surgeons Siraruj Sakoolnamarka Representative of Neurosurgeons Paisit Siriwittayakorn Representative of Colorectal Surgeons Choosak Pripatnanont Representative of Urological Surgeons Manoj Chantarasorn Representative of Orthopedic Surgeons Darin Lohsiriwat Member Pornchai O-Charoenrat Member Vitoon Chinswangwatanakul Member The THAI Journal of SURGERY Official Publication of Royal College of Surgeons of Thailand
ISSN 0125-6068 Editorial Board First Editor : Thongdee Shaipanich International Editorial Board : Lohfa Chirdan (Nigeria) Emeritus Editor : Soottiporn Chittmittrapap Suzanne A. Eccles (United Kingdom) Nopadol Wora-Urai Young-Chul Jang (Korea) Clifford P. Shearman (United Kingdom) Editor-in-Chief : Pornchai O-charoenrat Shuji Shimizu (Japan) Anusak Yiengpruksawan (USA) Editorial Team : Akkaraporn Deeprasertvit Apirag Chuangsuwanich Doonyapat Sa-Nguanraksa Kittipan Rerkasem Kriengchai Prasongsukarn Monthira Tanthanuch Nakornchai Phuenpathom Paisarn Vejchapipat Panuwat Lertsithichai Pornprom Muangman Potchavit Aphinives Somkiat Sunpaweravong Sukchai Satthaporn Thawatchai Akaraviputh Advisory Board : Apichat Kongkanand Arun Pausawasdi Auchai Kanjanapitak Banterng Rajatapiti Chalit Chienvichai Chana Sathornkich Chanvit Thanphiphat Charan Mahatumarat Charas Suwanwela Charoen Chotigavanich Chawalit Ongcharit Chesada Sangsubhan Chomchark Chuntrasakul Chumsak Pruksapong Chusak Kuptarnond Darin Lohsiriwat Kamphol Laohapensang Kijja Sindhvananda Kris Bhothisuwan Krit Chatamra Maiyadhaj Samsen Manus Satienchoak Montri Kitmanee Nakornchai Phuenpathom Naris Janviriya Naronk Rodwarna Ouichai Pleangprasit Paisal Pongchairerks Phaibul Jitpraphai Pichit Anuwutnavin Piyasakol Sakolsatayadorn Polpatt Talalak Pongsak Vathana Pradischai Chaiseri Prapan Kitisin Prasert Sarnvivad Prasert Trairatvorakul Prasert Vasinanukorn Prasith Gonggetyai Preeyaphas Nilubol Prinya Sakiyalak Prompong Peerabool Ravi Phimolsarnti Sanchai Kanchanalarp Sathit Ruangdilokrat Somboon Boonkasem Somprasong Songcharoen Sonchai Sirivunnabood Suchai Soontrapa Suchart Chaimuangraj Suchat Areemit Sukawat Watanatittan Supot Wudhikarn Suprecha Mokkhavesa Surachai Khaoroptham Surapong Supaporn Suthorn Bavonratanavech Thongueb Uttaravichien Trichak Sundhu Vajarabhongsa Bhudhisawasdi Vatana Supromajakr Vibul Sachakul Vibul Trakulhoon Vinai Parkpian Visist Dhitavat Vivat Visuthikosol Voravut Chanyavanich Yongyudh Vajaradul Editorial Board: Adisorn Patradul Akkrapol Mungnirandr Anan Manomaipiboon Anupan Tantiwong Apichart Ploysangwal Apirak Santingamkun Arkom Chaiwerawattana Arthi Kruavit Arun Rojanasakul Asada Methasate Banchong Mahaisavariya Boonchoo Sirichindakul Burapat Sangthong Chaichoompol Suwantemee Chairat Pokawattana Chaiwat Palawatvichai Chaiya Chansai Chaiyut Thanapaisal Chakrapan Euanorrasetr Chanchai Nimitrvanich Chanvate Sattaputh Chareonkiat Rergkliang Choosak Kasemsarn Choosak Pripatnanont Chopeow Taecholarn Chucheep Sahakitrungruang Chumpon Wilasrusmee Danaiphand Akarasakul Dusit Viravaidhya Kriengsak Limpastan Kriangsak Prasopsanti Kitti Kanpirom Kittichai Luengtaviboon Krisada Ratana-olarn Mongkol Laohapensang Narain Chotirosniramit Narong Kulvatanyou Narong Lertakyamanee Narong Punyahotra Nopporn Choeypunt O-Tur Saeseow Paiboon Jivapaisarnpong Paisit Siriwittayakorn Panupan Songcharoen Parinya Thavichaigarn Patpong Navicharern Peenutchanee Chartiburus Pichai Boonyongsunchai Phichaya Sujijantararat Piya Teawprasert Pongnares Purasiri Pornthep Pramyothin Pramook Mutirangura Prasit Watanapa Prayuth Chokrungvaranont Prayuth Sirivongs Preecha Siritongtaworn Puttisak Puttawibul Rangsan Niramis Sahachart Pipithkul Saranatra Waikakul Sathien Tumtavitikul Sathit Karanes Sindhujai Thansathit Siraruj Sakoolnamarka Smarn Tangaroonsin Somkiat Wattanasirichaikul Sopon Jirasiritham Sorawuth Chu-ongsakul Srichai Chaiyaphruk Sriprasit Boonvisut Sukasom Attanavanich Sukij Panpimanmas Sukij Mekraksavanich Sumit Wongkietkachorn Supachoke Chitvanith Supanit Nivatvong Supoj Ratchanon Supol Chindasub Surithorn Soontornpun Suriya Chakkaphak Suvit Sriussadaporn Tanaphon Maipang Tara Vongviriyangkoon Tavatchai Kanchanarin Taveesin Tanprayoon Tawee-sak Trivej Taweewong Chulakamontri Thamrongroj Temudom Thanyadej Nimmanwudipong Thirawud Khuhaprema Tongkao Rattanasuwan Vicha Charoonratana Vichai Benjachonlamas Vichai Srimuninnimit Vichien Laohacharoensombut Vinaisak Kattipattanapong Viroon Boonnuch Vitoon Chinsawangwatanakul Wachira Kachakarn Wanchai Krailadsiri Wichai Pansrimangkorn Wichai Vassanasiri Witaya Chadbunchachai Woramin Riansuwan Wuttichai Thanapongsatorn Yodruk Prasert Youwanush Kongdan
Published quarterly by : Royal College of Surgeons of Thailand The THAI Journal of SURGERY Official Publication of the Royal College of Surgeons of Thailand
INFORMATION FOR AUTHORS The Thai Journal of Surgery is the official publication of the Royal order in which they are first mentioned in the text. Identify references in text, College of Surgeons of Thailand issued quarterly. tables, and legends by arabic numerals (in superscript). The Thai Journal of Surgery invites concise original articles in the field The references must be verified by the author(s) against the original of clinical and experimental surgery as well as surgical education. Articles must documents. be contributed solely to The Thai Journal of Surgery and when published Example forms of references are given below. become the property of the Royal College of Surgeons of Thailand. The Royal Journal College of Surgeons of Thailand reserves copyright on all published materials 1. Standard Journal Article : and such materials may not be reproduced in any form without the written List all authors when six or less; when seven or more, list only first permission. three and add et al. The Royal College of Surgeons of Thailand accepts no responsibility for Soter NA, Wasserman SI, Austen KF. Cold urticaria: release into the opinions expressed by contributors. Manuscripts submitted for publication circulation of histamine and eosinophil chemotactic factor of ana- should be sent to : phylaxis during cold challenge. N Engl J Med 1976; 294:687-90. 2. Corporate Author : Pornchai O-charoenrat, M.D. The Committee on Enzymes of the Scandinavian Society for Clinical The Thai Journal of Surgery Chemistry and Clinical Physiology. Recommended method for the Royal College of Surgeons of Thailand determination of gamma glutamyltransferase in blood. Scand J Clin Royal Golden Jubilee Building Lab Invest 1976; 36:119-25. 2 Soi Soonvijai, New Petchburi Road Books and Other Monographs Huay Kwang, Bangkok 10310, Thailand 3. Personal Author (s) : Tel. : +66 2716 6141-3 Fax : +66 2716 6144 Osler AG. Complement: mechanisms and functions. Englewood E-mail : [email protected] Cliffs: Prentice - Hall, 1976. 4. Corporate Author : Manuscripts American medical Association Department of Drugs. AMA drug Manuscripts should be typewritten on one side of A4 paper only with evaluations. 3rd ed. Littleton: Publishing Sciences Group, 1977. double spacing and at least one inch margins. Separate pages: title page, text, 5. Editor, Compiler, Chairman as Author : acknowledgments, references, individual tables, and legends (as the format in Rhooder AJ, Van Rooyen CE, comps. Textbook of virology: for this journal). Number pages consecutively, beginning with the title page. Type students and practitioners of medicine and the other health sciences. the page number in the upper right-hand corner of each page. An orginal and 5th ed. Baltimore: Williams & Wilkins, 1968. one additional copy of all materials including figures and tables, using the 6. Chaper in Book : Microsoft Words Program, must be submitted for editorial review. Metric Weinstein L, Swartz MN. Pathogenic properties of invading micro- measurements should be used. Generic names for drugs should be used and if organisms. In: Sodeman WA Jr. Sodeman WA, eds. Pathologic trade name is mentioned, it should be put in parenthesis. physiology: mechanisme of disease. Philadelphia: WB Saunders, The title page of the manuscript should be typed on a separate sheet and 1974:457-72. contain the following informations : 7. Agency Publication : 1. Title of the paper National Center for Health Stastistics. Acute conditions: incidence 2. Name of Author(s), including first name(s) with academic degree(s) and associated disability, United States, July 1968-June1969. Rockville. 3. Name of department and institution in which the work was done Md.: National Center for Health statistics, 1972. Vital and health 4. Index word(s) statistics. Series 10: Data from the National health Survey, No. 69 : 5. Short running title (DHEW publication no. (HSM) 72-1036). Other Articles Illustrations and Tables 8. Newspaper Article : Glossy, black and white photographs of illustration are requested. Shaffer RA. Advances in chemistry are starting to unlock mysteries Original drawing should be professionally drawn in black India ink. Typewritten of the brain: discoveries could help cure alcoholism and inso mnia, or freehand lettering is not acceptable. Illustration should be numbered, the explain mental illness. How trhe messengers work. Wall Street tops indicated and the title of the article and the authors name in brief typed Journal 1977 Aug 12:(col. 1), 10(col.1). on a label pasted on its back. Do not mount the illustrations in the manuscripts. 9. Magazine Article : Each table should be prepared on a separate sheet and should be numbered Roueche B. Annals of medicine: the Santa Claus culture. The New consecutively as mentioned in the text and each must have a heading. Illustra- Yorker 1971 Sep 4:66-81. tions are accepted with color prints or color slide (not the negatives). Abbreviations References Use only standard abbreviations of commonly used approved abbrevia- References must be listed on a separate sheet in numeric order as tions. Avoid abbreviations in the title. The full term for which an abbreviation referred to in the article, not alphabetically. Only references mentioned in the stands should precede its first use in the text unless it is a standard unit of text should be listed and should be selective with not more than 30 references measurement. except under unusual circumstances. Number references consecutively in the
SUBSCRIPTION INFORMATION Subscription rate : B 200.00. single issue B 60.00 International : Individual US$ 35.00, Institutional US$ 45.00 Postage and handling - Surface mail included. Air mail added US$ 15.00 for each or US$ 40.00 for bulk of 4 copies. Combined subscriptions > 4 copies available for institutional only at US$ 35.00 each subscription. Medical Students, Intern and Residents : B 100.00 Business Address : Subscriptions, advertising, and all business communications should be addressed to: THE THAI JOURNAL OF SURGERY Royal Golden Jubilee Building 2 Soi Soonvijai, New Petchburi Road Bangkok 10310, Thailand Change of Address : Please allow 30 days before date of issue for change of address. Notification should include the old as well as the new address, with indication of whether the change is temporary or permanent. The THAI Journal of SURGERY Official Publication of the Royal College of Surgeons of Thailand
Vol. 36 April - June 2015 No. 2
Original Article
Management and Outcome of Ruptured Hepatocellular Carcinoma
Surawut Charoenkajonchai, MD General Surgery Unit, Department of Surgery, Chonburi Hospital, Chonburi 20000, Thailand
Abstract This study reviewed 84 patients with ruptured hepatocellular carcinoma (HCC) who were treated at Chonburi Hospital during 2006-2012. There were 72 males and 12 females. In acute management, 52 were treated conservatively, 23 were treated by transarterial embolization (TAE), 5 were treated by surgery followed by TAE and 4 were treated by surgery alone. Overall 30-day mortality was 61.9%. Factors related to early mortality are Child’s classification, MELD score and age. Four patients underwent delayed hepatectomy as a definitive treatment. The median survival after hepatectomy was more than 34 months. The longest follow up was a patient who underwent delayed hepatectomy as a definitive treatment and remained disease-free after 60 months. Hepatectomy should be considered as a definitive treatment for ruptured HCC in delayed phase. A suitable candidate for hepatectomy is a patient who has good performance status and normal liver function.
Keywords: Ruptured, hepatocellular carcinoma, management, outcome
INTRODUCTION hypovolemic shock, and eventually death. The rate of Hepatocellular carcinoma (HCC) usually ruptured HCC varied between 3-26%1-8 . These patients develops in cirrhosis. Most patients have no symptom have very high mortality rate of 20-71%1-3,6,7,9-12. The of tumor but cirrhosis. A few patients have symptoms initial treatments aim to control bleeding of the rupture related to tumor including mass effect and rupture of site. The most common method to control bleeding is tumor. Ruptured HCC can lead to massive blood loss, trans-arterial embolization (TAE) with success rate of
Correspondence address: Surawut Charoenkajonchai, MD, Department of Surgery, Chonburi Hospital, Amphur Muang, Chonburi 20000, Thailand; Telephone: +66 3893 1203; Fax: +66 3893 1202; Email : [email protected]
47 48 Charoenkajonchai S Thai J Surg Apr. - Jun. 2015
bleeding control around 30-100%8,13,14 whereas The patients who had been confirmed to have ruptured surgery has a role in control bleeding in selected hepatocellular carcinoma were included in this study. situation, with success rate lower than TAE. Surgery in The diagnosis of ruptured HCC was confirmed by acute management includes packing, hepatic artery imaging studies plus one of the following evidences: ligation and wedge resection or anatomical hepa- 1) abdominal tapping showed unclotted bloody fluid, tectomy. Definite hepatectomy is not recommended 2) imaging demonstrated disruption of the peritumoral in acute situation15. Patients who survive after acute liver capsule with enhanced fluid collection in the bleeding should be considered for a definitive perihepatic area adjacent to the tumor. Eleven patients treatment same as patients who never have rupture. were excluded because they were not confirmed for The best choice for definitive treatment is hepatectomy. ruptured HCC. Eighty four cases with ruptured HCC The alternative treatments are trans-arterial chemo- were included in the analysis. Details of patient embolization (TACE) and radiofrequency ablation characteristics including cirrhosis classifications, tumor (RFA). Liver transplantation which is the best definitive characteristics, early and late treatment, early and late treatment for non-ruptured HCC has no role for survival and results of hepatectomy were recorded. ruptured HCC at the present time. The longest follow up was 60 months. This study was The present study reviewed series of patients with approved by the hospital research committee. ruptured HCC at the Chonburi Hospital during 2006 to 2012. Outcomes of different treatment strategies RESULTS were reviewed and analyzed. There were 72 males and 12 females. The mean age was 56.1 year (range 34-87). Eighty three patients METHODS (98.8%) had hemodynamic or were clinically unstable. Medical records of patients with HCC during The mean systolic blood pressure was 107 mmHg 2006-2012 at Chonburi Hospital were reviewed. There (range 0-180). The mean hematocrit level was 25.6% were 748 patients who admitted for treatment of (range 7.4-40.1). The mean MELD score was 18.3 hepatocellular carcinoma. Ninety five charts with (range 6-47). 30-day mortality was 61.9% (Figure 1). additional record of hemoperitoneum were reviewed. Trans-arterial embolization (TAE) was performed in
Figure 1 Flow chart depicting acute management of patients with ruptured HCC. (TAE, transcatheter arterial embolization ; numbers in parenthesis are numbers of patients) Vol. 36 No. 2 Management and Outcome of Ruptured Hepatocellular Carcinoma 49
28 patients with success rate of 85.7% (24 in 28). Other Table 1 Characteristics of survivors and non survivors in 30 treatments were perihepatic packing (N=5), surgical days of ruptured HCC hemostasis (N=3) and wedge hepatectomy (N=1). Survivor Non survivor P value Success rate of overall surgical control of bleeding was (n=32) (n=52) 66.7% (6 in 9). Age (range) 59.3 ± 11.8 (34-87) 54.1 ± 11.2 (36-87) 0.04 Most patients had hemodynamic unstable and Gender male (%) 25 (78%) 47 (90%) 0.2 57.8% were not candidates for any treatment because Child-Pugh they were moribund or had very poor liver status. classification (%) Among those who had treatment, TAE was the most A 12 (37.5) 0 frequently used modality. Surgery was performed to B 18 (56.3) 23 (44.2) <0.0001 control bleeding in acute phase for 10.7%. In later C 2 (6.2) 29 (55.8) HBsAg +(%) 48.1 59.5 0.52 management, BCLC classification was applied in order AntiHCV+ (%) 27.3 26.5 0.8 to determine the opportunity of definitive treatment. Chronic Alcoholic (%) 65.6 76.5 0.41 Four hepatectomies were performed for definitive SBP (mmHg) 109 ± 26 106 ± 32 0.67 treatment. Hct (%) 27.3 ± 6.1 24.6 ± 7.5 0.08 According to analysis of factors effecting early MELD score 13.0 ± 7.1 21.5 ± 9.4 <0.0001 ± ± survival, Child Classification, MELD score and younger Tumor size (cm.) 8.6 4.2 10.4 3.9 0.05 > 3 tumors (%) 31.3% 52.9% 0.09 age were significantly related with survival (Table 1). Tumor location Other potential factor was tumor size but the level of Right 17 20 significance was not reached. Hematocrit and blood Left 6 6 0.11 pressure were not related with survival. Treatment Both 8 25 methods were significantly related to survival because Treatment potential survivors were usually selected for TAE or TAE 15 8 Surgery only 3 1 surgery whereas moribund cases were usually chosen <0.0001 for conservative treatment. There were 32 patients Surgery plus TAE 5 0 Conservative 9 43 who survived until subsequent evaluation for definite Median Survival (Days) 109 3.5 NA treatment. BCLC classification was used as a strategy
Figure 2 Flow chart of late management of patients with ruptured HCC according to BCLC classification. (Numbers in parenthesis are numbers of patients.) 50 Charoenkajonchai S Thai J Surg Apr. - Jun. 2015
for determination of treatment (Figure 2). Four of Table 2 Late management of ruptured HCC patient them had hepatectomy as definitive treatment. The TACE Hepatectomy Palliative others were offered transarterial chemoembolization (n=11) (n=4) (N=17) (TACE) if deemed suitable or best supportive care Age 62.3 ± 14.3 50.5 ± 14.8 59.5 ± 8.8 for palliative treatment. Fourteen patients underwent Gender male (%) 10 (91%) 3 (75%) 12 (71%) TACE as palliative treatment. The result of TACE were Child-Pugh Classification (%) partial response (3 in 14), stable disease (6 in 14) and A 6 (54.5) 4 (100) 2 (11.8) no response (5 in 14). None of them had complete B 5 (45.5) 0 13(76.4) response. Median survival of TACE group was 15.5 C 0 0 2(11.8) Tumor size (cm.) 9.2 ± 4.5 6.7 ± 3.7 8.6 ± 4.2 months. Table 2 showed characteristic of patients Tumor Location classified by treatment. Three patients underwent Right 5 3 9 hepatectomy after TACE (Table 3, case #2, 3 and 5). Left 2 1 3 The detail and result of five patients who underwent Both 4 0 5 hepatectomy was shown in Table 3. The other 17 Extrahepatic disease 3 (2 lung 1 (peritoneal 4 (3 lung patients who did not undergo any definite treatment metastasis, seeding) metastasis, 1 LN 1 lung & bone received palliative treatment because of poor metastasis) metastasis) performance status, advanced cirrhosis, bilateral tumor Previous treatment or metastatic disease. These patients had very poor TAE / TACE 9 3 8 median survival of 2.5 months. Survival of each Emer. Surgery 1 2 3 treatment was shown in Figure 3. None 1 - 7 Median survival (months) 15.5 not reach 2.5 (>34)
Table 3 Characteristics of five patients who underwent hepatectomy
Case # Characteristic Tumor Previous Operation Survival Treatment (months)
1 M, 59 yrs, 5 cm. Segment 4 none, emergency surgery wedge hepatectomy 33 months, alcoholic cirrhosis dead from massive variceal bleeding 2 M, 43 yrs, HBVcirrhosis 7 cm. segment 5,8 Acute phase : extended right 60 months, alive, liver packing follow by TAE hepatectomy disease free, **pathology result showed regularly follow up Secondary Phase : complete tumor necrosis TACE × 1 by effect of TACE 316 M,34 yrs, HBVcirrhosis 10 cm. Rt.lobe + 30 cm Acute phase : TAE 1. Right hepatectomy + 34 months, seeding tumor at Enbloc resection of dead from tumor Lt.lower abdomen Secondary Phase : tumor seeding recurrence at TACE × 4 2. Re-resection of peritoneum seeding recurrence seeding 3. Re-resection of recurrence seeding 4 M,67 yrs, HBVcirrhosis 8 cm. Rt.lobe Acute phase : conservative Right hepatectomy 32 months, alive, (stable VS., bleeding stop disease free, spontaneously) regularly follow up 5 F,58 yrs 1.5 cm. segment 2 Acute phase : TAE Left hepatectomy 40 months, Secondary Phase: TACE × 1 alive, disease free, regularly follow up Vol. 36 No. 2 Management and Outcome of Ruptured Hepatocellular Carcinoma 51
hepatocellular carcinoma]. Korean J Hepatol 2009;15:148- 58. [Abstract] 4. Pawarode A, Voravud N. Ruptured primary hepatocellular carcinoma at Chulalongkorn University Hospital: a retrospective study of 32 cases. J Med Assoc Thai 1997;80: 706-14. 5. Castells L, Moreiras M, Quiroga S, et al. Hemoperitoneum as a first manifestation of hepatocellular carcinoma in western patients with liver cirrhosis: effectiveness of emergency treatment with transcatheter arterial embolization. Dig Dis Sci 2001;46:555-62. 6. Chearanai O, Plengvanit U, Asavanich C, Damrongsak D, Sindhvananda K, Boonyapisit S. Spontaneous rupture of primary hepatoma: report of 63 cases with particular reference to the pathogenesis and rationale treatment by Figure 3 Kaplan-Meier survival curve of late management of hepatic artery ligation. Cancer 1983;51:1532-6. ruptured HCC patients 7. Bassi N, Caratozzolo E, Bonariol L, et al. Management of ruptured hepatocellular carcinoma: implications for therapy. World J Gastroenterol 2010;16:1221-5. CONCLUSION 8. Lai ECH, Lau WY. Spontaneous rupture of hepatocellular carcinoma: a systematic review. Arch Surg 2006;141:191-8. Ruptured hepatocellular carcinoma is a lethal 9. Tan FL-S, Tan Y-M, Chung AY-F, Cheow PC, Chow PK-H, Ooi presentation of hepatocellular carcinoma. The LL. Factors affecting early mortality in spontaneous rupture mortality is likely related to liver status more than the of hepatocellular carcinoma. ANZ J Surg 2006;76:448-52. tumor status (Child-Pugh classification and MELD 10. Miyamoto M, Sudo T, Kuyama T. Spontaneous rupture of score rather than tumor size). Secondary treatment hepatocellular carcinoma: a review of 172 Japanese cases. Am J Gastroenterol 1991;86:67-71. was selected according to patient’s condition. Patients 11. Lai EC, Wu KM, Choi TK, Fan ST, Wong J. Spontaneous who have tolerable liver status deserve better chance of ruptured hepatocellular carcinoma. An appraisal of surgical survival and opportunity for definitive treatment, even treatment. Ann Surg 1989;210:24-8. though survival after hepatectomy in ruptured HCC is 12. Yang H, Chen K, Wei Y, et al. Treatment of spontaneous not as good as in non-ruptured HCC2,17. Hepatectomy ruptured hepatocellular carcinoma: A single-center study. is still the most appropriate definitive treatment Pak J Med Sci 2014;30:472-6. 13. Leung CS, Tang CN, Fung KH, Li MKW. A retrospective review 8,10,18,19 available so far . BCLC classification which of transcatheter hepatic arterial embolisation for ruptured includes liver status, tumor status and patient perfor- hepatocellular carcinoma. J R Coll Surg Edinb 2002;47:685-8. mance status for grading is a good prognosis predic- 14. Nouchi T, Nishimura M, Maeda M, Funatsu T, Hasumura Y, tor in long term. BCLC classification is also a good Takeuchi J. Transcatheter arterial embolization of ruptured strategy to determine treatment. Hepatectomy should hepatocellular carcinoma associated with liver cirrhosis. be offered for BCLC grade 0-B in order to meet a long Dig Dis Sci 1984;29:1137-41. 15. Chiappa A, Zbar A, Audisio RA, Paties C, Bertani E, Staudacher term survival in these patients. Patient with peritoneal C. Emergency liver resection for ruptured hepatocellular seeding (BCLC grade C) could meet a long term carcinoma complicating cirrhosis. Hepatogastroenterology. survival if resection could be performed. 1999;46:1145-50. 16. Charoenkajonchai S. Synchronous resection of large peritoneal implantation with right hemihepatectomy after REFERENCES ruptured hepatocellular carcinoma: a case report. Chonburi 1. Liu CL, Fan ST, Lo CM, et al. Management of spontaneous Hospital Journal 2011;36:93-8. rupture of hepatocellular carcinoma: single-center 17. Yeh C-N, Lee W-C, Jeng L-B, Chen M-F, Yu M-C. Spontaneous experience. J Clin Oncol 2001;19:3725-32. tumour rupture and prognosis in patients with hepatocellular 2. Chen CY, Lin XZ, Shin JS, et al. Spontaneous rupture of carcinoma. Br J Surg 2002;89:1125-9. hepatocellular carcinoma. A review of 141 Taiwanese cases 18. Shuto T, Hirohashi K, Kubo S, et al. Delayed hepatic resection and comparison with nonrupture cases. J Clin Gastroenterol for ruptured hepatocellular carcinoma. Surgery 1998;124:33-7. 1995;21:238-42. 19. Vergara V, Muratore A, Bouzari H, et al. Spontaneous rupture 3. Kim Y-I, Ki H-S, Kim M-H, et al. [Analysis of the clinical of hepatocelluar carcinoma: surgical resection and long- characteristics and prognostic factors of ruptured term survival. Eur J Surg Oncol 2000;26:770-2. The THAI Journal of SURGERY 2015;36:52-62. Official Publication of the Royal College of Surgeons of Thailand Original Article
En Bloc Sacrectomy for Retrorectal Tumor: Factors Influencing Immediate Outcome
Teerachai Yongchaitrakul, MD* Piya Kiatisevi, MD** *Department of Surgery, Lerdsin Hospital, Ministry of Public Health, Bangkok, Thailand **Department of Orthopedics, Lerdsin Hospital, Ministry of Public Health, Bangkok, Thailand
Abstract Background: En bloc sacrectomy is the only chance for cure of retrorectal tumor. Total sacrectomy is a complicated, time-consuming and collaborative operation with multiple potential postoperative morbidities and mortality. Well-planned surgery, perioperative care and rehabilitation would be beneficial in surgical outcomes and decrease postoperative complications. Objective: To define the factors affecting the immediate outcomes of en bloc sacrectomy in patients with retrorectal tumor. Study design: A retrospective cohort at a single academic institute. Methods: Data were collected retrospectively in all patients who underwent en bloc sacrectomy at the Lerdsin Hospital between June 2008 and August 2013. We reviewed demographic data, clinical findings, past medical history, tumor characteristics, tumor invasion, preoperative preparation: iliac artery embolization, ureteric stenting, surgical procedures, intraoperative complication, estimated blood loss, units of blood transfusion, histopathologic findings, postoperative morbidities and mortality, functional outcomes: constipation, urinary retention and gait disturbance. Immediate surgical outcome was defined according to the postoperative complications, hospital length of stay, functional outcomes and gait disturbance. All parameters were analyzed and determined for their significance. Result: Fifteen patients (6 males and 9 females) who underwent sacrectomy (6 subtotal sacrectomy and 9 total sacrectomy) were studied. The mean age was 53.8 years (range 26-83 years). Of these, 13 had preoperative iliac artery embolization, 6 had ureteric stenting, 8 had anterior approach with bilateral internal iliac vessels ligation, 2 had protective colostomy and 3 had lumbosacral reconstruction. The average operative time was 8.4 hours (4-14 hours). The mean estimated blood loss was 5,020 mL (1,600-10,000 mL) and the average blood transfusion was 17.6 units (range 1-40 units). Pathological diagnosis were chordoma (n=7), malignant peripheral nerve sheath tumor (n=3), chondrosarcoma (n=2), and giant cell tumor, gastrointestinal stromal tumor (GIST) and leiomyosarcoma (1 each). Based on the Clavien-Dindo Grading Systems for postoperative complications, 1 patient was categorized into Class I, 2 in Class II, 4 in Class IIIa, 7 in Class IIIb and 1 in Class V. Eight patients (53%) had major complications and required re-operation and 1 patient died from severe sepsis. Regarding functional outcomes, 14 of 15 patients had postoperative urinary retention and 10 of them (67%) had constipation. The gait disturbance was as follows: 5 patients walked without assistance, 6 patients walked with assistance and 3 patients needed wheelchairs. None had sacral hernia. The mean duration of hospital stay was 77 days (range 24-186 days). Conclusion: Subtotal sacrectomy was safe and had less complications whereas preoperative iliac artery embolization did not influence intraoperative blood loss. The anterior approach with internal iliac vessel ligation decreased blood loss especially in total sacrectomy and the posterior approach is a safe option for subtotal sacrectomy. The protective colostomy did not affect the incidence of surgical site infection but helped in decreasing its severity. Despite preoperative ureteric stenting, ureteric injury could not be prevented. The closure of sacral defect could be adequately undertaken without prosthesis sheath by gluteal advancement flap or gluteus maximus approximation. There was no difference between total and subtotal sacrectomy in terms of urinary retention and constipation. Gait disturbance is less in subtotal sacrectomy compared to total sacrectomy. Without lumbopelvic reconstruction, total sacrectomy could give favorable intermediate and long term outcome. Keywords: pelvic surgery, presacral tumor, retrorectal tumor, sacrectomy
Correspondence address: Teerachai Yongchaitrakul, MD, Department of Surgery, Lerdsin Hospital, 190 Silom Road, Bangrak, Bangkok 10500, Thailand: Telephone: +66 2353 9800; E-mail: [email protected]
52 Vol. 36 No. 2 En bloc Sacrectomy for Retrorectal Tumor 53
Statistical analysis INTRODUCTION The studied data would be presented in mean or Retrorectal tumors are rare neoplasms locating median and range if they were continuous variables. in the retrorectal (presacral) space. These neoplasms These variables were analyzed by student t test if they originate from various embryonic remnants and have had normal distribution and by Chi-squared test or a large variety in either gross or histological findings. Fisher exact test if they were categorical variables. True incidence of retrorectal tumors remains unclear Mann-Whitney U test was used to test whether two with approximately 1 in 40,000 hospital admissions. independent samples of observations were drawn from Although the majority of retrorectal tumors are benign, the same or identical distributions, such as tumor size, malignant neoplasms are observed especially in estimated blood loss and subgroup (subtotal/total pediatric population and solid tumor type1-6. Because sacrectomy) analysis. The overall survival and overall pelvic cavity is a closed narrow space which contains a disease-free survival rates were traced using the Kaplan- variety of vital organs, pelvic surgery requires a Meier method. P value less than 0.05 was considered comprehensive knowledge of pelvic anatomy, expertise statistically significant. Statistical analysis was done of involved surgeons, and a well-organized teamwork using SPSS software. to achieve a safe and successful operation. The primary curative modality for retrorectal Preoperative assessment and evaluation tumor is en bloc resection which requires multi- Patients were carefully evaluated following the disciplinary specialists. Operation in this surgical American Society of Anesthesiologists (ASA) standards. field, however, remains vulnerable to complication Patients in ASA class I-III were eligible for surgery. and the operative environment is often hostile. In this Clinical data and radiologic imaging (CT/MRI) of study, we collected case series and determined the patients were discussed in the manner of team approach possible factors that affect immediate surgical outcomes involving general surgeon, orthopedist, urologist, and complications of the en bloc sacrectomy for plastic surgeon, gynecologist and radiologist. Operative retrorectal tumor. The data acquired from this study planning and sequence of procedures were summarized would facilitate perioperative care and improve in the multidisciplinary preoperative meeting. In the morbidities, mortality and quality of life of the patients. case that surgeon considered preoperative arterial Moreover, we may utilize these data to develop standard embolization, the procedure would be performed by guideline for the treatment of retrorectal tumors in interventional radiologist. The embolization of the the future. tumor-feeding internal iliac artery was done either unilaterally or bilaterally depending on the angio- PATIENTS AND METHODS graphic findings 2-3 days prior to surgery.
We retrospectively studied 15 patients who Surgical technique underwent surgery for retrorectal tumor at the Lerdsin En bloc sacrectomy consisted of two parts: anterior Hospital, Department of Medical Services, Ministry of approach and posterior approach. In some circum- Public Health of Thailand between September 2008 stances such as large tumor, obstructive uropathy and and August 2013. We reviewed demographic data, unclear ureteral identification from preoperative clinical findings, past medical history, tumor imaging, the surgeons might consider ureteric stent characteristics, tumor invasion, preoperative placement which would be done by urologist. In preparation: iliac artery embolization, ureteric stenting, anterior approach stage, the patient was in a supine surgical procedures, intraoperative complication, position. After a standard midline laparotomy incision, estimated blood loss, units of blood transfusion, sigmoid colon and rectum were mobilized from tumor. histopathological findings, postoperative morbidities The rectum must be dissected as distal as possible. In and mortality, functional outcomes: constipation, some patients whose rectum tightly adhered to tumor urinary retention and gait disturbance. We attained or suspected tumor invasion, the rectum was transected patient data only in the admission period by browsing proximal to the adhesion with linear staple, then the hospital archives. proximal end was brought out through abdominal 54 Yongchaitrakul T, Kiatisevi P Thai J Surg Apr. - Jun. 2015
wall and matured later. The distal end of rectum was (inverted Y incision). The incision was carried through left in place and removed altogether with the tumor subcutaneous tissue and lumbosacral fascia. After the in posterior approach stage. Both ureters would be fascia was incised, it was elevated off the paraspinous freed away from the tumor and also dissected as distal muscles laterally. There was an avascular plane that as possible. Both sides of common iliac, external and could be followed to the iliac crest bilaterally. The internal iliac vessels were identified. The internal iliac fascia was elevated off of the iliac crest. Using a arteries and veins were ligated and transected at their subperiosteal technique, the paraspinous muscles were origin bilaterally, only in the patients who could be elevated off of the spinous process, the lamina, the safely done. Other soft tissues and structures around facets, and the transverse process of the spine. If a total the tumor and sacroiliac joints were dissected and sacrectomy was planned, this exposure must encompass cleared out. The abdominal wall was closed as usual. L3 through L5. If a subtotal sacrectomy was to be In posterior approach stage, the patient was in a performed, then exposure of L5 was sufficient. The prone position. A midline incision was made beginning paraspinous muscles were then cut transversely at the at the level of L3 down to the coccygeal tip 5-10 cm same level at which the fascia was incised. The gluteal above the anus. If there was a preoperative core needle muscles inserting on the lateral aspect of the ilium biopsy tract, the tract and adjacent tissue were removed were elevated, allowing visualization of the sciatic notch. with the specimen using an elliptical skin incision Laminectomy was performed at the level of the most around the biopsy tract. The incision was extended caudal nerve root to be preserved. For the subtotal laterally toward the greater trochanter on both sides sacrectomy, S1 laminectomy was made and the S1
Figure 1 The summary illustration of total sacrectomy from posterior aspect. (Printed with permission. ©2008 The Johns Hopkins University Neurosurgery-Ian Suk) (From Ian suk. Dissecting a complex neurosurgical illustration: step-by-step development. World Neurosurg. 76, 6:497-507, 2011.) Vol. 36 No. 2 En bloc Sacrectomy for Retrorectal Tumor 55
roots were carefully identified and preserved. Dural Fifteen patients with retrorectal tumors underwent en sac was divided and ligated at the S1-2 level, and then bloc sacrectomy from September 2008 to August 2013. the sacral body was exposed for a safe osteotomy at the They were 6 male (40%) and 9 female (60%) with a S1-2 junction. Posterior osteotomy was continued on mean age at presentation of 53.8 years (range 26-83 the ala of sacrum and the sacroiliac joints, midline years). All of them were ASA functional classification osteotomy was performed, connecting the lateral I-III. Pathological diagnosis included chordoma (n=7), osteotomies through the vertebral body, and malignant peripheral nerve sheath tumor (n=3), subsequently the sacrum could be mobilized. S2-4 chondrosarcoma (n=2) and 1 each of giant cell tumor, nerves were separated from sciatic nerves, piriformis GIST and leiomyosarcoma. The mean tumor volume muscles and ligamentous structures (sacrotuberous, was 1,755 mL (average 12 × 12 × 12 cm) and 5 of them sacrospinous ligament) were divided. The internal (30%) had sacral bone invasion. iliac arteries and veins were ligated and transected if Of 15 patients, 6 underwent subtotal sacrectomy they were not previously performed in anterior and 9 underwent total sacrectomy while 13 of them approach stage. In the same way, these processes were had preoperative iliac artery embolization. Ureteric done with other branches of the vessels such as stent placements were also done in six patients. For iliolumbar vessels and lateral sacral vessels. The anterior approach: eight patients underwent bilateral mesorectum was carefully dissected off the capsule internal iliac vessels ligation and two patients underwent of tumor, in cases where rectoproctectomy was protective colostomy. Among the total sacrectomy performed, distal rectum and anus were removed group, lumbosacral reconstruction was done in 3 together with the tumor, then en bloc tumor resection patients and hemipelvectomy was done in 1 patient was accomplished. For total sacrectomy, an L5-S1 due to tumor extension who later had severe sepsis lamnicectomy was performed and the dural sac was and died after 71 days of admission. The mean operative ligated at L5-S1 with preservation of the L5 roots. An time, estimated blood loss and units of blood transfusion L5-S1 discectomy was performed followed by a posterior were 8.4 hours (range 4-14 hours), 5,020 mL (range osteotomy at the sacroiliac joints for total sacrectomy. 1,600-10,000 mL), and 17.6 units (range 1-40 units), The S1-4 nerves were then separated from the sciatic respectively. The details of studied data are listed in nerves, the following steps were the same as previously Table 1. described in subtotal sacrectomy. Figure 1 demonstrates According to Clavien-Dindo Grading System for summary of the total sacrectomy operation. the Classification of Surgical Complications (Table Lumbopelvic reconstruction was done in three 2)7, patients were classified as 1 in class I, 1 in class II, patients by Galveston reconstruction system, however, 4 in class IIIa, 7 in class IIIb, 0 in class IV and 1 in class two patients had postoperative pelvic infection that V. As mentioned above, there were 8 patients (53%) required instrument removal. Thereafter, the surgeons with severe complications (Class IIIb, IV, V) who needed did not consider lumbopelvic reconstruction in the re-operations or died. Common complications rest of total sacrectomy cases. Sacral defect closure was included 11 wound infections (73%) 7 of which done by approximation of the gluteus maximus. In required re-operations for wound debridement, 1 intra- the cases with large sacral defect with inadequate abdominal collection (6.7%) which required re- remained viable muscles, reconstruction has been exploration and drainage, 2 ureteric injuries (13.4%) done with gluteal advancement flap. Subsequently, which required percutaneous nephrostomy, 2 rectal the skin flap was closed, self-suctioning drains were injuries (13.4%) which required re-operation for placed. In the patients who underwent rectoproc- Hartmann’s procedure. There were three patients tectomy or Hartmann’s procedure concurrent with en who underwent a combined total sacrectomy and bloc sacrectomy, colostomy would be matured after the lumbopelvic reconstruction, two of them experienced patients were repositioned back to supine position. postoperative wound infection with instrument failure. Both of them needed re-operation for instrument removal. None of patients in this study had sacral RESULTS hernia during admission or after the 2-4 year follow- Summary of patients’ data was shown in Table 1. up.
56 Yongchaitrakul T, Kiatisevi P Thai J Surg Apr. - Jun. 2015 LOS (days) LOS
Others Gait disturbance Gait
with
assistance
assistance
assistance
Urine retention Urine Constipation
Functional outcomes
Clavien-Dindo score Clavien-Dindo utcomes and length of stay re-operation of No.
1
wound assistance
removal)
collection chair
instrument removal, 7
debridement
debridement) Others
DVT, removal, 3
injury, Hartmann’s with
injury, instrument chair
tal failure procedure,
tal failure 1 PCN)
instrument debridement,
Complications
instrumen- assistance
Sacral hernia Sacral
Collection
Wound infection Wound
TBT (units) TBT EBL (cc) EBL
Others Operative time (hrs) time Operative
8 15,000 30 - + - - 1 (1 IIIb - + wheel 110
5 300 1 - - - - - I + + walk, no 24
6 4,000 14 - - - - - II - + walk, 35
4 3,800 12 + - - - - IIIa + + walk, no 33
12 6,000 40 + + - rectal 9 (1 IIIb - + walk, 107
10 1,600 11 + - - (8 8 IIIb + + walk, no 90
14 8,000 28 + + - ureter 5 (1 IIIb + + wheel 186
Operative Factors:
Cystostomy
Colostomy
Internal iliac vessels ligation vessels iliac Internal Operation type Operation
+LPR
+ LPR
+ LPR
Operative Factors: Technique
sacrectomy
sacrectomy
Ureteric stening Ureteric Embolization
Factors
Preoperative
Bone invasion Bone Tumor size (cm) size Tumor
14 sacrectomy
16 sacrectomy Tumor Factors pathology Tumor
tumor 11 sacrectomy
Summary data of the Patients’ demographic, tumor factors, preoperative factors, operative factors, complications, functional o
sarcoma 24 acrectomy
sarcoma 0 sacrectomy ASA
Table 1 Gender
data Age
Demographic Patient
1 59 M 2 Chondro- 13x20x + + + subtotal - - -
2 32 F 1 Chondro- 7x11x1 + + + total + - -
3 54 F 1 Giant c ell 12x25x - + - total - + -
4 56 F 1 Chordoma 7x6x7 - + - subtotal - - -
5 58 F 2 MPNST 12x13x + + + total + - -
6 60 M 1 Chordoma 13x12x - + + total + - -
7 26 F 1 GIST 4x3x2 - + - subtotal + - -
Vol. 36 No. 2 En bloc Sacrectomy for Retrorectal Tumor 57 LOS (days) LOS
Others Gait disturbance Gait
with
with
assistance
assistance
assistance
assistance
Urine retention Urine Constipation
Functional outcomes
Clavien-Dindo score Clavien-Dindo utcomes and length of stay re-operation of No.
wound chair
debridement
debridement asistance
debridement Others
dead wound
injury wound with
injury wound with
foot drop debridement asistance
Complications
Sacral hernia Sacral
TBT = Total Blood Transfusion; LOS = Length of Stay
Collection
Wound infection Wound
TBT (units) TBT EBL (cc) EBL
Others Operative time (hrs) time Operative
8 4,000 12 + - - - - IIIa + + walk, 74
4 1,500 6 + - - - - IIIa + - walk, no 34
4 2,600 7 - - - - - I + + walk, no 37
9 15,500 29 + - - - - IIIa - + walk, 80
6 10,000 25 + - - (2 2 IIIb + + whell 99
12 5,500 13 + - - sepsis, 4 (4) V 71
12 11,000 21 + + - ureter 3 (2 IIIb + + walk, 82
12 4,000 15 + - - rectal 8 (8 IIIb - + walk, 97
Operative Factors:
Cystostomy
Colostomy
Internal iliac vessels ligation vessels iliac Internal Operation type Operation
+ It.
tomy
Operative Factors: Technique
sacrectomy
sacrectomy
sacrectomy
hemipelvec-
Ureteric stening Ureteric Embolization
Factors
Preoperative
Bone invasion Bone Tumor size (cm) size Tumor
0 sacrectomy
11 sacrectomy
10 sacrectomy
14 sacrectomy
10 sacrectomy Tumor Factors pathology Tumor
Summary data of the Patients’ demographic, tumor factors, preoperative factors, operative factors, complications, functional o ASA
Table 1 Gender
data Age
Demographic Patient
8 76 M 3 Chordoma 17x15x - - + subtotal + - -
9 62 F 2 Chordoma 12x10x - - - subtotal - - -
10 36 M 1 MPNST 12x7x6 + + - subtotal - - -
11 42 F 2 MPNSt 6x6x10 + + + Total + - -
12 60 F 2 Sarcoma 12x7x1 - + - total - + -
13 31 M 1 Chordoma 17x15x - + - total - - -
14 83 F 3 Chordoma 8x6x10 - + - total + - -
15 72 M 3 Chordoma 12x13x - + - total + - -
ASA = American Society of Anesthesiologists Physical Status Classification System; EBL = Estimated Blood Loss (intraoperation); 58 Yongchaitrakul T, Kiatisevi P Thai J Surg Apr. - Jun. 2015
Table 2 Clavien-Dindo Grading System for the Classification of Surgical Complications
Grades Definition
Grade I: Any deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic and radiological interventions. Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgetics, diuretics and electrolytes and physiotherapy. This grade also includes wound infections opened at the bedside. Grade II: Requiring pharmacological treatment with drugs other than such allowed for grade I complications. Blood transfusions and total parenteral nutrition are also included. Grade III: Requiring surgical, endoscopic or radiological intervention Grade III-a: Intervention not under general anesthesia Grade III-b: Intervention under general anesthesia Grade IV: Life-threatening complication (including CNS complications: brain haemorrhage, ischaemic stroke, subarachnoid bleeding, but excluding transient ischaemic attacks) requiring IC/ICU management. Grade IV-a: Single organ dysfunction (including dialysis) Grade IV-b: Multi-organ dysfunction Grade V: Death of a patient Suffix ‘d’: If the patients suffers from a complication at the time of discharge, the suffix ‘d’ (for ‘disability’) is added to the respective grade of complication. This label indicates the need for a follow-up to fully evaluate the complication.
Daniel D, Nicolas D, Pierre-Alain C: Classification of Surgical Complications: A New Proposal With Evaluation in a Cohort of 6336 Patients and Results of a Survey. Ann Surg. Aug 2004; 240(2): 205-213.
Table 3 Patient status in October 2014
Patient Follow-up Type of Functional outcomes Tumor Sacral Remarks duration operation (years) Defecation Urination Gait recurrence hernia 1 loss F/U subtotal moderate difficulty/ straining walk, no assistance Yes (at 2009) No re-operation since 2011 sacrectomy laxative use (wide excision (3-yr. F/U) at 2009) 2 4 total sacrectomy required enema/ on urinary walk, No No + LPR evacuation catheter no assistance 3 4 total sacrectomy on colostomy on urinary wheel chair, No No catheter standable 1-2 min. 4 4 subtotal sacrectomy moderate difficulty/ straining walk, No No laxative use no assistance 5 4 total sacrectomy required enema/ on PCN walk, No No + LPR evacuation with assistance 6 4 total sacrectomy moderate difficulty/ self walk, No No + LPR laxative use catheterization no assistance 7 2 subtotal required enema/ straining walk, No No sacrectomy evacuation no assistance 8 3 subtotal required enema/ on urinary walk, No No sacrectomy evacuation catheter with assistance 9 3 subtotal normal normal walk, No No sacrectomy defecation urination no assistance 10 3 subtotal required enema/ self walk, No No sacrectomy evacuation catheterization no assistance 11 - total sacrectomy + lt. - - - - - Death hemipelvectomy 12 2 total sacrectomy on colostomy self walk, No No catheterization with assistance 13 2 total sacrectomy required enema/ self walk, No No evacuation catheterization with assistance 14 2 total sacrectomy required enema/ self wheel chair No No evacuation catheterization 15 1 total sacrectomy on colostomy self walk, No No catheterization with assistance LPR = lumbopelvic reconstruction Vol. 36 No. 2 En bloc Sacrectomy for Retrorectal Tumor 59
Postoperative functional outcomes were also Table 4 Factors influencing Clavien-Dindo Grading System for reviewed. Ten patients experienced constipation (67% the Classification of Surgical Complications of all patients, 90% of patients without colostomy) and Clavien-Dindo Clavien-Dindo 14 of 15 patients had urinary retention. Gait Score Score P-Value ≤ ≥ disturbances were noted as follows: 5 patients (33%) ( IIIa) ( IIIb) could walk without assistance, 6 patients (40%) could Demographic data walk with assistance, 3 patients (20%) needed Age (year) : N (x, SD.) 7 (53.71, 16.987) 8 (53.88, 18.373) 0.986 wheelchairs and 1 died (7%). The average hospital Gender : N (%) 1.000 Male 3 (42.9) 3 (37.5) stay was 77 days (range 24-186 days). Female 4 (57.1) 5 (62.5) Although it is beyond the scope of this study, in ASA : N (%) 1.000 October 2014, 13 patients (87%) had the follow-up I 4 (57.1) 3 (37.5) period lasting 2-4 years, 1 was lost to follow-up and 1 II 2 (28.6) 3 (37.5) died after the operation. After analyzing the data with III 1 (14.3) 2 (25.0) Tumor Factors Kaplan-Meier method, the overall survival rate was Tumor Patholog : N (%) 0.576 92.9% and overall disease free survival rate was 93.3%. Chondrosarcoma - 2 (25.0) A local recurrence was discovered in a single patient Giant cell tumor - 1 (12.5) one year after surgery. The functional outcomes by the Chordoma 4 (57.1) 3 (37.5) time of follow-up were listed in Table 3. Three patients MPNST 1 (14.3) 2 (25.0) GIST 1 (14.3) 0 (0.0) (21%) had colostomy, 1 patient had normal defecation, Leiomyosarcoma 1 (14.3) 0 (0.0) 3 patients (21%) had moderate difficult defecation Tumor Size (cm3) : N (Median) 7 (840) 8 (1,873) 0.247 and required laxative, 6 patients (42%) required enema Sacral bone invasion : N (%) 0.282 or evacuation. Urinary dysfunctions were also noted, No 6 (85.7) 4 (50.0) 4 patients (28%) retained urinary catheter, 1 patient Yes 1 (14.3) 4 (50.0) (7%) had normal urination, 3 patients (21%) had Preoperative Factors Embolization 0.200 straining, and 6 patients (42%) performed self-urinary No 2 (28.6) 0 (0.0) catheterization. Patients’ gait disturbances were Yes 5 (71.4) 8 (100.0) recorded, 2 patients (14%) used wheelchairs, 7 patients Ureteric stenting 0.608 (49%) could ambulate without assistance and 5 patients No 5 (71.4) 4 (50.0) (35%) could ambulate with assistance. Yes 2 (28.6) 4 (50.0) Operation Factors : Technique Reviewing record of the patient who lost follow- Operation : N (%) 0.041* up, he regularly came in the first 3 years (2008-2010), Subtotal sacrectomy 5 (71.4) 1 (12.5) however tumor recurrence was detected in one year Total sacrectomy 2 (28.6) 7 (87.5) after the operation. He underwent the re-operation of IIV ligation : N (%) 0.619 wide excision to remove the recurrent tumor. After No 4 (57.1) 3 (37.5) Yes 3 (42.9) 5 (62.5) the second operation, he continued another 2 years of Colostomy : N (%) follow-up until he lost contact in 2011. However, No 6 (85.7) 7 (87.5) 1.000 during his follow-up, he experienced moderate difficult Yes 1 (14.3) 1 (12.5) defecation and required laxative , straining of urination Cystostomy : N (%) - and ambulation without assistance. No 7 (100.0) 8 (100.0) Yes - -
DISCUSSION could resume their activities of daily life. Various studies reported that en bloc sacrectomy is Focusing on immediate outcomes, significant the only chance for cure of retrorectal tumors8-10. Our factors for a successful en bloc sacrectomy in retrorectal study showed that the intermediate and long term tumors were: preoperative preparation, a compre- outcomes are favorable after 2-4 year follow-up (Table hensive knowledge of sacral anatomy and biomechanics 3) with the overall survival rate of 92.9% and overall of lumbosacropelvic region, surgical technique and disease-free survival rate of 93.3%. Most of patients prophylactic procedures. As mentioned earlier, in 60 Yongchaitrakul T, Kiatisevi P Thai J Surg Apr. - Jun. 2015
Table 5 Factors influencing estimated blood loss (EBL) Among demographic data, tumor factors,
EBL EBL P-Value preoperative factors and operative technique factors, (≤ 2,000 cc.) (≥ 2,000 cc.) the statistical analysis showed that the only single
Tumor Pathology : N (%) 0.862 factor associated with severe complications leading to Chondrosarcoma 1 (33.3) 1 (8.3) re-operation and death (Clavien-Dindo classification Giant cell tumor - 1(8.3) IIIb, IV, V) was operation types (subtotal/total Chordoma 2 (66.7) 5 (41.7) sacrectomy, p=0.041), as listed in Table 4. Of 6 patients MPNST - 3 (25.0) with subtotal sacrectomy, only 1 patient (16.7%) had GIST - 1 (8.3) severe complication while 7 of 9 patients (77.8%) who Leiomyosarcoma - 1 (8.3) Tumor Size (cm3) : N (Median) 3 (1,200) 12 (1,200) 0.885 underwent total sacrectomy had severe complica- Sacral bone invasion : N (%) 1.000 tions. We concluded that subtotal sacrectomy was No 2 (66.7) 8 (66.7) significantly safe and caused less severe complication Yes 1 (33.3) 4 (33.3) comparing to total sacrectomy. Preoperative embolization : N (%) 0.371 We evaluated the effects of tumor factors and No 1 (33.3) 1 (8.3) Yes 2 (66.7) 11 (91.7) preoperative embolization on intraoperative blood loss. Massive blood loss was defined as 50% of blood volume loss within 3 hours or a rate of blood loss more Table 6 Number of patients and estimated blood loss in subtotal/ than 150 cc/min11. We found that there was no factor total sacrectomy group, separate to non ligation and significantly affecting intraoperative bleeding as ligation of Internal Iliac Vessels group shown in Table 5. However, when subgroup analysis Non Ligate Ligate P-Value was done affecting among operation types, the anterior approach with internal iliac vessels ligation in subtotal Subtotal sacrectomy: 4 (1,550) 2 (3,900) 0.133 sacrectomy did not reduce intraoperative bleeding N (Median EBL cc.) Total sacrectomy: 3 (15,000) 6 (5,500) 0.024* (p=0.133). On the other hand, for total sacrectomy, N (Median EBL cc.) anterior approach with internal iliac vessels ligation significantly reduced the intraoperative bleeding (p=0.024) as shown in Table 6. Therefore, subtotal order to provide a comprehensive surgical care in en sacrectomy may not require anterior approach for bloc sacrectomy, a multidisciplinary team composing of internal iliac vessels ligation which is consistent with general surgeon, orthopedist, urologist, plastic surgeon, the report of various authors that subtotal sacrectomy nursing staff, physiotherapist, and psychologist should can be safely performed with posterior approach be involved to achieve an effective outcome. alone12,13 while anterior approach with internal iliac Sacral resection is a challenging procedure and at vessels ligation should be conducted in all total the same time is also a worrisome operation for the sacrectomy to reduce intraoperative blood loss14-16. involved surgeons because of 1) the complexity of When determining the correlation of protective anatomy and surgical technique, which is compulsory colostomy and wound infection, wound infection could to perform both anterior and posterior approach; not be prevented by protective colostomy (p=0.506) 2) the multiple potential complications including listed in Table 7. However, two patients with protective massive bleeding, associated organ injuries and sacral colostomy had less severe postoperative wound hernia; 3) degree of functional impairment; more or infection that required no further re-operation for less, patients will experience anorectal and urogenital surgical debridement. We proposed that the small dysfunction which might need permanent colostomy sample size may not allow us to see the significant or cystostomy; and 4) gait disturbance and remained statistical difference. lumbosacropelvic stability. When surgeons are aware The preoperative ureteric stenting for prevention of factors affecting surgical outcomes, then the possible of ureteric injury was also reviewed. Preoperative complications could be prevented. It would be ureteric stenting did not significantly decrease the rate enormously beneficial for surgical performance, of ureter injury, and could not prevent ureter injury, effectiveness and patient safety. listed in Table 8. We suggest that because of the unfa- Vol. 36 No. 2 En bloc Sacrectomy for Retrorectal Tumor 61
Table 7 Effect of colostomy on severity of wound infection Table 9 Effect of operation type on functional outcomes
Not severe Severe P-Value Subtotal Total P-Value sacrectomy Sacrectomy Colostomy : N (%) 0.506 Not done 7 (77.8) 5 (100.0) Constipation : N (%) 0.182 Done 2 (22.2) - No - 2 (40.0) Yes 6 (100.0) 4 (60.0) Not severe = no need re-operation (Clavien-Dindo classification I, II, IIIa); Urine retention : N (%) 0.429 Severe = need re-operation (Clavien-Dindo classification IIIb, IV, V) No 1 (16.7) - Yes 5 (83.3) 8 (100.0) Table 8 Effect of ureteric stenting on incidence of ureter injury Gait disturbance : N (%) 0.003* Walk, no assistance 5 (83.3) - Ureter injury Ureter injury P-Value Walk, with assistance 1 (16.7) 5 (62.5) (No) (Yes) Wheel chair - 3 (37.5) Ureteric stenting 1.000 Not done 8 (61.5) 1 (50.0) Done 5 (38.5) 1 (50.0) performing or not performing. The explanation came from the evidence that after completely healed, ileolumbar muscle and ligament would form biologic miliar posterior approach position, surgeons should sling and stabilize the spine which adequately function pay more attention in order to clearly identify ureters as weight-bearing joint for individual gait. There was and conduct the operation cautiously. one patient in our study who underwent total In the subgroup analysis of operation types and sacrectomy with successful lumbopelvic reconstruction, functional outcomes, listed in Table 9, 10 patients and had an excellent gait function outcome. The experienced constipation, 6 were in subtotal sacrectomy patient could walk without any assistance. group, 4 were in total sacrectomy group, and there was There was no report of sacral hernia among the no significant difference between the groups (p=0.182). studied group either during admission or after 2-4 Nearly all of patients (14 from 15) had postoperative years of follow-up. Regardless of postoperative wound urinary retention. We concluded that urinary retention infection and re-operative debridement (7 out of 11 or constipation could occur regardless of operation patients), we proposed that sacral defect closure types. In contrast, gait disturbance was associated with following sacrectomy could be adequately done by operative procedure. Patients undergoing subtotal gluteus maximus approximation or gluteal advance- sacrectomy tended to have less gait disturbance; 5 ment flap. Prosthesis sheath placement which might patients could walk without assistance and 1 patient increase the rate of wound infection was not routinely could walk with assistance while all patients with total necessary for all patients20. sacrectomy need external support (5 of 6 patients) or wheelchairs (3 of 3 patients) (p=0.003). Our findings CONCLUSION supported the theory of subtotal sacrectomy in which patients would not lose the lumbopelvic ring stability, En bloc sacrectomy is a complex procedure which thus lumbopelvic reconstruction was not necessary. is the only chance of cure for retrorectal tumor. The However, after the 2-4 year follow-up, our data showed operation requires multidisciplinary team of surgeons that 75% of patients (6 of 8 patients) undergoing total and other medical personnel. Even though it is an sacrectomy whether received lumbopelvic recon- operation with various possible complications, the struction or not, was able to ambulate either with or intermediate and long term outcomes are favorable, without assistance. Our findings were consistent with most of patients could resume their activities of daily the previous studies that total sacrectomy without life. The operation of choice depends on tumor size lumbopelvic reconstruction could give favorable and location, when free margin resection is obtainable, outcomes17-19. Wuisman et al.17 concluded that there subtotal sacrectomy is preferred. And when subtotal was no significant difference in functional outcomes sacrectomy is a procedure of choice, posterior approach among two groups of lumbopelvic reconstruction alone can be safely performed. On the other hand, for 62 Yongchaitrakul T, Kiatisevi P Thai J Surg Apr. - Jun. 2015
all total sacrectomy, anterior approach should be 6. Jao SW, Beart RW Jr, Spencer RJ, et al. Retrorectal tumors. performed for intraoperative internal iliac vessels Mayo Clinic experience, 1960-1979. Dis Colon Rectum 1985; 28(9):644-52. ligation to decrease intraoperative blood loss. Although 7. Daniel D, Nicolas D, Pierre-Alain C: Classification of Surgical protective colostomy does not decrease the incidence Complications: A New Proposal With Evaluation in a Cohort of of surgical wound infection but the severity tends to 6336 Patients and Results of a Survey. Ann Surg 2004; 240(2): have less virulent. The evidence showed that pre- 205-213. operative ureteric stenting could not prevent ureteric 8. Fourney DR, Rhines LD, Hentschel SJ, et al. En bloc resection injury. In addition, there were no difference in of primary sacral tumors: classification of surgical approaches and outcome. J Neurosurg Spine 2005;3(2):111-22. postoperative urinary-bowel function between patients 9. Randall RL, Bruckner J, Lloyd C, Pohlman TH, Conrad EU 3rd. with subtotal sacrectomy and total sacrectomy while Sacral resection and reconstruction for tumors and tumor-like patients underwent subtotal sacrectomy had less gait conditions. Orthopedics 2005;28(3):307-13. disturbance compared to total sacrectomy. We also 10. Wuisman P, Lieshout O, Sugihara S, van Dijk M. Total sacrectomy suggest that in total sacrectomy, instrument placement and reconstruction: oncologic and functional outcome. Clin for lumbopelvic reconstruction is not always necessary, Orthop Relat Res 2000;(381):192-203. 11. Fakhry SM, Sheldon GF. Massive transfusion in the surgical because most of patients without lumbopelvic patient. In: Jeffries LC, Brecher ME, editors. Massive Transfusion reconstruction had favorable outcomes after the American Association of Blood Banks, Bethesda; 1994. intermediate to long term follow-up. Lastly, sacral 12. McLoughlin GS, Sciubba DM, Suk I, et al. En bloc total defect after sacrectomy could be closed with gluteus sacrectomy performed in a single stage through a posterior maximus approximation or gluteal advance flap without approach. Neurosurgery 2008;63(1 suppl 1). prosthesis sheath placing. 13. Clarke MJ, Dasenbrock H, Bydon A, Sciubba DM, McGirt MJ, Hsieh PC, Yassari R, Gokaslan ZL, Wolinsky JP. Posterior-only approach for en bloc sacrectomy: clinical outcomes in 36 ACKNOWLEDGEMENT consecutive patients. Neurosurgery 2012;71(2):357-64 14. Localio SA, Eng K, Ranson JH: Abdominosacral approach for The authors gratefully thank Dr. Alisa Yanasan retrorectal tumors. Ann Surg 1980;191:555-60. for her support and valuable assistance and Miss 15. Wanebo HJ, Marcove RC: Abdominal sacral resection of Chatravee Jindapol for helping with statistical analysis locally recurrent rectal cancer. Ann Surg 1981;194:458-71. 16. McCormick PC, Post KD: Surgical approaches to the sacrum, method. in Doty JR, Rengachary SS (eds): Surgical Disorders of Sacrum. New York: Thieme Medical Publishers; 1994. p. 257-65. 17. Wuisman P, Lieshout O, Sugihara S, van Dijk M. Total sacrectomy and reconstruction: oncologic and functional outcome. Clin REFERENCES Orthop Relat Res 2000:192Y203. 1. Hobson KG, Ghaemmaghami V, Roe JP, et al. Tumors of the 18. Ohata N, Ozaki T, Kunisada T, Morimoto Y, Tanaka M, Inoue H. retrorectal space. Dis Colon Rectum 2005;48(10):1964-74. Extended total sacrectomy and reconstruction for sacral 2. Uhlig BE, Johnson RL. Presacral tumors and cysts in adults. Dis tumor. Spine 2004;29:123Y6. Colon Rectum 1975;18(7):581-9. 19. Santi MD, Mitsunaga MM, Lockett JL. Total sacrectomy for a 3. Johnson WR. Postrectal neoplasms and cysts. Aust N Z J Surg giant sacral schwannoma. A case report. Clin Orthop Relat 1980;50(2):163-6. Res 1993:285Y9. 4. Cody HS 3rd, Marcove RC, Quan SH. Malignant retrorectal 20. Miles WK, Chang DW, Kroll SS, Miller MJ, Langstein H, Reece tumors: 28 years’ experience at Memorial Sloan-Kettering GP, et al. Reconstruction of large sacral defects following Cancer Center. Dis Colon Rectum 1981;24(7):501-6. total sacrectomy. Plast Reconstr Surg 105:2387-2394, 2000 5. Freier DT, Stanley JC, Thompson NW. Retrorectal tumors in 21. Ian suk. Dissecting a complex neurosurgical illustration: step- adults. Surg Gynecol Obstet 1971;132(4):681-6. by-step development. World Neurosurg 2011;76:497-507. The THAI Journal of SURGERY 2015;36:63-66. Official Publication of the Royal College of Surgeons of Thailand Original Article
The Effectiveness of Meticulous Surgery and Postoperative Steroid Injection for Treatment of Auricular Keloids
Anan Watcharhachittitam, MD Plastic and Reconstructive Surgery Unit, Nakornping Hospital, Chiang Mai, Thailand
Abstract Background: The treatment of auricular keloids is known to be difficult because of high recurrence rate after surgery. Many adjuvant therapies had been used including intralesional steroid injection. Objective: To evaluate the effectiveness of meticulous surgery and postoperative steroid injection for the treatment of auricular keloids. Methods: From October 2011 to September 2014, 38 patients with ear keloids (total 62 ear keloids) were treated with intralesional excision and closure of the defect using keloid fillet flap. Postoperative intralesional steroid injection was administered for the first time at 2- week and then every 4 weeks for 2 times (total 3 times). Recurrence rate, aesthetic results, symptomatic improvement and the adverse effects were evaluated. Results: The follow-up period ranged from 6-32 months (mean 18.09 months). Four from 38 patients recurred (recurrent rate 10.52%). All patients were satisfied with the aesthetic results and had symptomatic improvement without any complications. Conclusion: Meticulous surgery (excision and keloid fillet flap) and 3-time postoperative intralesional steroid injection can be effective for treatment of the auricular keloids.
Keywords: Auricular keloid, excision and keloid fillet flap, intralesional steroid injection
INTRODUCTION ear piercing, trauma or burn. Keloids are benign, hard and persistent fibrous Surgery is the gold standard to treat any size of proliferations that developed in predisposed persons keloids, as it corrects the obvious distortion of the ear at site of the cutaneous injury1. They are characterized and reduces the mass. Various surgical methods such by extension beyond the borders of the original wound as excision followed by primary suture2, healing by and growth in a pseudotumor fashion with tissue secondary intention, skin graft3 or local flap4,5 have distortion and high recurrence rates after excision. been used. But the reported recurrence rates after They could be asymptomatic, elicit burning, pain or surgery alone were 50-100%6. itching. The common affected areas are anterior Several methods to prevent post-surgical chest, shoulders and ears. Auricular keloids usually recurrence of keloids include intralesional steroid appear as shiny, smooth and globular growths on one injections7-10, radiation therapy11-14, mechanical or both sides of the ear. They are usually preceded by compression15-17, silicone occlusive dressing18,
Correspondence address: Anan Watcharhachittitam, MD, Plastic and Reconstructive Surgery Unit, Nakornping Hospital, 159 Moo 10, Chotana Road, Maerim District, Chiang Mai 50180, Thailand; Email: [email protected]
63 64 Watcharhachittitam A Thai J Surg Apr. - Jun. 2015
Figure 1 Patient with bilateral ear lobule keloids, before (left panel) and after surgery (right panel). cryosurgery19, 5-Fluorouracil, imiquimod, mitomycin- C 20 and interferon-alpha, -beta and -gamma. The objective of this study was to determine the effectiveness of meticulous surgery by local flap technique combined with postoperative intralesional steroid injection for treatment of the auricular keloids. The primary goals were low recurrence rate, significant aesthetic results, symptomatic improvement and minimal adverse effects.
MATERIALS AND METHODS
This study was carried out in outpatients attending Figure 2 Patient with left ear keloid (helical rim), before (left the Plastic Surgery Unit of Nakornping Hospital, panel) and after surgery (right panel). Chiangmai, Thailand from October 2011 to September 2014. Sixty-two ear keloids in 38 patients, were treated growing scar was defined as thickening of scar, with intralesional excision of the keloids and closure of distortion of surrounding tissue or stiff raised scar. the defect using keloid fillet flap. The surgery was performed under local anesthesia using 1% lidocaine RESULTS and 1:100,000 epinephrine and closed with one layer of 6-0 Nylon. Terramycin eye ointment and pressure From October 2011 to September 2014, 62 ear dressing with gauze were applied after surgery for 24 keloids in 38 patients (28 female and 10 male) were hours. Two weeks after surgery, sutures were removed treated with this protocol. The patient age ranged and intralesional Triamcinolone acetonide (TA) from 10 to 73 years (mean age 21.26 years). All keloids injection (40 mg/ml) 0.1-0.5 ml (depending on the occurred after ear piercing. At the time of treatment, size of the lesion) was administered for the first time the keloids have been presented from 6 months to 3 and then every 4 weeks for 2 times (total 3 times). The years (mean 1.74 years). Twenty two patients had patients came back for checking every 2-6 months or keloids on the left ear, 8 patients on the right ear and when clinically recurrence occurred. 8 patients had bilateral lesions. Twenty two patients The final results were recurrence rate. Recurrence had keloids at ear lobules, 15 patients at the helical rim was defined as postoperative emergence of growing or and only one patient at both areas. persistently symptomatic scar tissue. Symptoms were Twenty three patients had only one ear keloid, 8 defined as pain, itching or pigmentation whereas patients had 2 keloids, 5 patients had 3 keloids and 2 Vol. 36 No. 2 The Effectiveness of Meticulous Surgery and Postoperative Steroid Injection 65
patients had 4 keloids. The size of keloids ranged from 23%, 16.6% and 9.5% in patients treated with excision 0.8-2.5 cm in diameter (mean 1.55 cm). Among the 38 and keloid fillet flap combined with intraoperative and patients, 7 patients had received prior surgery but no postoperative steroid injections. In my experience, patient had received intralesional steroid injection or three patients treated with intraoperative steroid other treatment. injection had wound dehiscence at two weeks after All 62 ear keloids had no surgical site infection or stitch removal. Then, I only administered postoperative flap necrosis. All 38 patients were satisfied with the steroid injection and the results were better (recurrent aesthetic results and had symptomatic improvement. rate 10.52%) Telangiectasia and skin hypopigmentation were noted 2. Surgery and radiation therapy in five patients, but resolved later. Ragoowansi et al.12 reported 8.8% recurrence The follow-up period ranged from 6-32 months after 12-month follow-up in 34 patients who were (mean 18.09 months). Four from 38 patients recurred treated with excision and immediate postoperative (recurrence rate 10.52%). Two patients recurred six adjuvant radiotherapy of 10 Gy. Ogawa et al.14 reported months after treatment and the other two patients overall recurrence rate only 4% after 18-month follow- recurred four months after treatment. All recurrent up in the patients who were treated by post-surgical cases were treated with intralesional steroid injection radiotherapy with 10 Gy in two fractions over 2 years. until the keloids were resolved. Postoperative radiation therapy tends to have better result but it is limited only in the university hospital or cancer center hospital, not in the most general hospital DISCUSSION in Thailand. Postoperative steroid injection is easier Keloids are benign fibroproliferative lesions that than radiation therapy to treat the auricular keloids. the etiopathogenesis is not clear. Several mechanisms 3. Surgery and pressure therapy have been implicated in formation of keloids, including Many pressure devices were used for treatment growth factor abnormalities, defective collagen of keloids after surgery such as Zimmer splintage15, turnover, changes in the orientation of collagen fiber custom-made methylmethacrylate stent, and new due to tension, immune system dysfunction or pressure device by Gregor et al.17. These studies hypersensitivity to sebum. reported that this treatment was safe and effective. The Multiple therapeutic options are available for the author suggests that the triple therapy (surgery + treatment of auricular keloids, but in the present day, postoperative steroid injection + pressure therapy) combination therapy (with surgery) appears to be the can be the most appropriate treatment for general most effective. hospitals. 1. Surgery and intralesional steroid injection 4. Surgery and other methods The most literatures reported the use of steroid Several methods to prevent post-surgical injection as an adjunctive procedure after keloid recurrence, for examples cryosurgery19, colchicine, 5- excision but with various doses, schedules and concen- fluorouracil injection, application of 5% imiquimod trations. cream, mitomycin-C20, or interferon-alpha 2b, had Brown and Ortega3 reported no recurrence after been studied but the results were still not satisfactory. 11-month follow-up in 10 patients with auricular keloids treated with total excision in combination with coverage of the defect with a full-thickness skin graft and REFERENCES intradermal injection of steroid. Limitations of this 1. Kelly AP. Keloids and hypertrophic scars. In: Parish LC, Lask GP, research were small number of patients and more editors. Aesthetic dermatology. New York: McGraw-Hill, donor site scars. 1991:58-64. 2. Lee Y, Minn KW, Baek RM, Hong JJ. A new surgical treatment In my opinion, keloid excision and closure of the of keloid: keloid core excision. Ann Plast Surg 2001;46:135-40. defect using keloid fillet flap is the better surgical 3. Brown NA, Ortega FR. The role of full-thickness skin grafting technique because of no donor site scar as reported in and steroid injection in the treatment of auricular keloids. Ann many literatures. Plast Surg 2010;64:637-8. Many authors5,8,9 reported recurrence rates of 4. Kim DY, Kim ES, Eo SR, Kim KS, Lee SY, Cho BH. A surgical 66 Watcharhachittitam A Thai J Surg Apr. - Jun. 2015
approach for earlobe keloid: keloid fillet flap. Plast Reconstr Plast Surg 2001;54:504-8. Surg 2004;113:1668-74. 13. Stahl S, Barnea Y, Weiss J, et al. Treatment of earlobe keloids 5. Al Aradi IK, Alawadhi SA, Alkhawaja FA, Alaradi I. Earlobe by extralesional excision combined with preoperative and keloids: a pilot study of the efficacy of keloidectomy with core postoperative çsandwiché radiotherapy. Plast Reconstr Surg fillet flap and adjuvant intralesional corticosteroids. Dermatol 2010;125:135-41. Surg 2013;39:1514-9. 14. Ogawa R, Huang C, Akaishi S, et al. Analysis of surgical 6. Mafong EA, Ashinoff R. Treatment of hypertrophic scars and treatments for earlobe keloids: analysis of 174 lesions in 145 keloids: A review. Aesth Surg 2000;3:114-21. patients. Plast Reconstr Surg 2013;132:818-25. 7. Shons AR, Press BH. The treatment of earlobe keloids by 15. Russell R, Horlock N, Gault D. Zimmer splintage: a simple surgical excision and postoperative triamcinolone injection. effective treatment for keloids following ear-piercing. Br J Ann Plast Surg 1983;10:480-2. Plast Surg 2001;54:509-10. 8. Rosen DJ, Patel MK, Freeman K, Weiss PR. A primary protocol 16. Park TH, Seo SW, Kim JK, Chang CH. Outcomes of surgical for the management of ear keloids: results of excision excision with pressure therapy using magnets and identification combined with intraoperative and postoperative steroid of risk factors for recurrent keloids. Plast Reconstr Surg injections. Plast Reconstr Surg 2007;120:1395-400. 2011;128:431-9. 9. Jin YJ, Mi RR, Yeon SK, Kee YC. Surgery and perioperative 17. Gregor MB, Jorn B, Karl H, Boris AS. Auricular keloids: combined intralesional corticosteroid injection for treating earlobe keloids therapy with a new pressure device. Arch Facial Plast Surg : A Korean experience. Ann Dermatol 2009; 21: 221-5. 2012;14:20-6. 10. Reuben FS, Bijitesh C, Alok S, M Alam P, Anil M. Efficacy of triple 18. Yigit B, Yazar M, Alyanak A, Guven E. A custom-made silicon therapy in auricular keloids. Cutan Aesth Surg 2014;7:98-102. mold for pressure therapy to ear keloids. Aesth Plast Surg 11. Sclafani AP, Gordon L, Chadha M, Romo T. Prevention of 2009;33:849-51. earlobe keloid recurrence with postoperative corticosteroid 19. Careta MF, Fortes AC, Messina MC, Maruta CW. Combined injections versus radiation therapy: a randomized, prospective treatment of earlobe keloids with shaving, cryosurgery and study and review of the literature. Dermatol Surg 1996;22:569- intralesional steroid injection: a 1-year follow-up. Dermatol 74. Surg 2013;39:734-8. 12. Ragoowansi R, Cornes PGS, Glees JP, Powell BW, Moss ALH. 20. Chi SG, Kim JY, Lee WJ, et al. Ear keloids as a primary candidate Earlobe keloids: treatment by a protocol of surgical excision for the application of mitomycin C after shave excision: in vivo and immediate postoperative adjuvant radiotherapy. Br J and in vitro study. Dermatol Surg 2011; 37: 168-75. The THAI Journal of SURGERY 2015;36:67-72. Official Publication of the Royal College of Surgeons of Thailand Original Article
Low Adherence to Guideline for the Control of Major Cardiovascular Risk Factors in Diabetic Patients with Peripheral Arterial Disease: A Study in Northern Part of Thailand
Supapong Arworn, MD* Saranat Orrapin, MD* Natapong Kosachunhanun, MD† Kiran Sony, MD‡ Nimit Inpankaew, MD§ Kittipan Rerkasem, MD, PhD*,¶ *Department of Surgery, Faculty of Medicine, Chiang Mai University, †Department of Internal Medicine, Faculty of Medicine, Chiang Mai University, ‡Department of Internal Medicine, Chiangrai Prachanukroh Hospital, §Department of Internal Medicine, Lamphun Hospital ¶Research Institute of Health Science, Chiang Mai University
Abstract Objective: The control of major cardiovascular risk factors is an important issue in the treatment of patients with peripheral arterial diseases (PAD), and several large trials and many guidelines have stressed the importance of detection and treatment of these risk factors. This is even critical in diabetic patients with PAD because such patients carry higher risk of cardiovascular problem than non diabetic patients. However, it is not known to what extent this has been applied to such patients. The objective of this study was to investigate the adequacy of risk factor management in diabetic patients with PAD. Methods: A cross-sectional study was conducted in 286 consecutive diabetic patients with PAD. A medical history, medication history, and fasting venous blood sample were taken for each patient. The plasma concentrations of cholesterol were determined. PAD was diagnosed when patients had low ankle brachial index (≤ 0.9). Then we classified whether patients had good or poor control in each criteria according to the recommended guidelines between PAD and non PAD. Results: There were 2,247 diabetic patients, 874 (38.9) males and 1373 (61.1) females. Among them were 286 PAD patients. Diabetic patients with PAD had histories of cardiovascular events (angina pectoris, myocardial infarction, unstable angina, stroke and transient ischemic attack) more than those without PAD. Apart from smoking, the percentage of appropriate control of risk factors was around only 15-30% in PAD patients. When authors focused on further details between non PAD, mild PAD and severe PAD, severe PAD had lower BMI and lower waist circumference than those in other groups. Eighteen out of 286 PAD patients were still smoking. Conclusion: Although the risk factors in vascular disease management are well understood and many guidelines have been published, they do not seem to be well managed in diabetic patients with PAD.
Keywords: blood pressure, diabetes, dyslipidemia, peripheral arterial disease, smoking
Correspondence address: Kittipan Rerkasem, MD, Division of Vascular Surgery, Department of Surgery, Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand; Tel +66 5394 5532; Fax +66 5394 6139; Email: [email protected]
67 68 Arworn S, et al. Thai J Surg Apr. - Jun. 2015
INTRODUCTION is the ABI ≤ 0.9 or a history of leg ischemia with a Hyperlipidaemia, hypertension, smoking, obesity previous intervention (angioplasty, bypass graft or and hyperglycemia are the most important risk factors other vascular intervention including amputation)2. in atherosclerosis, which is the major cause of arterial Then authors further classified PAD group into mild diseases1. Aggressive risk reduction to prevent PAD (ABI 0.9-0.5) and severe PAD (ABI < 0.5 or history cardiovascular events in patients with arterial disease is of a previous intervention). well known2. Many recommendations have been made specifically for PAD patients2-5. Some even more focus Statistical methodology on PAD in diabetic patients, because these patients are The data was showed in descriptive study in remarkably at high risk for cardiovascular morbidity percentage. We categorized each risk factor as and mortality. The guidelines of the American Heart “controlled” if it was at the target level of the American Association (AHA) and The Royal College of Physicians Heart Association recommendation and The Royal of Thailand indicated risk factor control in diabetic College of Physician of Thailand as Table 1. Chi- patients with PAD as in Table 13-5. squared test was used to analyse the difference between Despite much evidence about the danger of PAD and non PAD group. The protocol was approved inappropriate risk factor control in diabetic patients by our local ethic committee in each hospital and with PAD, it is not known to what extent this has been signed, informed consent was obtained for all patients. applied to Thai patients. The objective of this study Analysis was performed by STATA for Windows version was to study the extent to which participants’ risk 13.0. factors this study had been managed with regard to such guidelines. RESULTS
There were 2,247 diabetic patients 874 (38.9) MATERIALS AND METHODS males and 1373 (61.1) females (Table1). There were This study recruited outpatients aged 45 years 286 PAD patients. The number of severe and mild PAD with diabetic mellitus between May 2014 and August was 35 and 251 cases respectively. Diabetic patients 2014. This study was conducted in three hospitals in with PAD had histories of cardiovascular events (angina Northern part of Thailand namely Maharaj Nakorn pectoralis, myocardial infarction, unstable angina, Chiang Mai Hospital, Chiangrai Prachanukroh stroke and transient ischemic attack) more than those Hospital and Lamphun Hospital. Medical records were without PAD (Table 2). Apart from smoking, the reviewed. Patients’ records and the drug charts in the percentage of appropriate control of risk factors was patient’s notes were checked for record of laboratory around only 15-30% in PAD patients (Table 3). Non within three months during enrolment including low PAD patients had more BMI and waist circumference density lipoprotein (LDL) and HbA1C. The test than those in PAD patients. When authors focused on consisted of the participants sitting quietly in a room at the clinic for at least 20 minutes. Twice, at intervals of 5-10 minutes, their blood pressure was measured on Table 1 Criteria of good control of risk factors according to AHA left arm at heart level. The mean of these two guideline and the guideline of The Royal College of measurements was used for the analysis. Anthro- Physicians of Thailand pometric measurements were performed. While Item “Good control” definition wearing indoor clothes, the participant’s height, weight, Blood pressure (BP) systolic BP < 130 mmHg and diastolic BP and waist circumference measurement were taken. < 80 mmHg Each participant filled out questionnaires on various LDL < 70 mg/dL cardiovascular risk factors, such as family history of HbA1C < 7% ≤ smoking experience, cardiovascular disease, current Body mass index 23 drug habits, and past medical history. Ankle brachial Waist circumference < 90 cm (male) or < 80 cm (female) Smoking Stop index (ABI) was determined in all patients using autonomic machine. The criteria for diagnosis of PAD LDL low density lipoprotein Vol. 36 No. 2 PAD in Diabetic Patients 69
Table 2 Baseline clinical data
Characteristics Total PAD Non-PAD (2247 patients) (286 patients) (1961patients) N (%) N (%) N (%)
Age (years) 45 - 49 151 (6.7) 11 (3.9) 140 (7.1) 50 - 59 759(33.8) 63 (22.0) 696 (35.5) 60 - 69 813(36.2) 97 (33.9) 716 (36.5) 70 - 79 400(17.8) 81 (28.3) 319 (16.3) ≥ 80 124 (5.5) 34 (11.9) 90 (4.6) Gender Male 874(38.9) 112 (39.2) 762 (38.9) Female 1373(61.1) 174 (60.8) 1199 (61.1) Previous history of cardiovascular events 1. Chronic stable angina 8 (0.4) 3 (1.1) 5 (0.3) 2. Myocardial infarction 73 (3.3) 19 (6.6) 54 (2.8) 3. Unstable angina 35 (1.6) 9 (3.2) 26 (1.3) 4. Stroke (ischemic) 55 (2.5) 15 (5.2) 40 (2.0) 5. Transient ischemic attack 3 (0.13) 1 (0.35) 2 (0.10)
PAD = peripheral arterial disease
Table 3 Distribution of patients according to recommended criteria between PAD and non PAD
Item Non PAD PAD P value
Blood pressure (BP) Systolic BP < 130 mmHg and Diastolic BP < 80 mmHg) 29.8 28.7 0.70 Systolic BP _ 130 mmHg or BP Diastolic _ 80 mmHg) 70.2 71.3 LDL < 70 mg/dL 18.7 18.8 0.95 _ 70 mg/dL 81.4 81.3 HbA1C (%) < 7 25.8 30.1 0.18 ≥ 7 74.2 69.9 BMI (kg/m2) ≤ 23 31.1 37.1 0.04 >23 68.9 62.9 Waist circumference (cm.) < 90 (Male) or < 80 (Female) 29.9 35.6 0.05 ≥ 90 (Male) or ≥ 80 (Female) 70.1 64.4 Smoking Never or Former 96.8 93.7 0.008 Current 3.2 6.3
PAD = peripheral arterial disease, BP = blood pressure, LDL = low density lipoprotein, BMI=body mass index further details between non PAD, mild PAD and severe the proper treatment of risk factor control in patients PAD, severe PAD had lower BMI and waist with PAD, it does not seem to be well managed in circumference than those in other groups (Table 4). clinical practice6. Indeed mismanagement is evident in patients in this study, only 15-30% of diabetic patients with PAD had proper control in each risk factor. It DISCUSSION seems that some patients are not aware of this problem, Although a huge amount of evidence supports which may imply that clinicians fail to stress the 70 Arworn S, et al. Thai J Surg Apr. - Jun. 2015
Table 4 Distribution of patients according to recommended criteria between non PAD, Mild PAD (ABI 0.5-0.9), severe PAD (ABI < 0.5)
Item Non PAD Mild PAD Severe PAD P Value
Blood pressure (BP) Systolic BP < 130 mmHg and Diastolic BP < 80 mmHg) 29.8 27.9 34.3 0.69 Systolic BP ≥ 130 mmHg or BP Diastolic ≥ 80 mmHg) 70.2 72.1 65.7 LDL < 70 mg/dL 18.7 16.6 35.0 0.14 ≥ 70 mg/dL 81.4 83.5 65.0 HbA1C (%) < 7 25.8 30.9 22.7 0.29 ≥ 7 74.2 69.1 77.3 BMI (kg/m2) ≤ 23 31.1 32.7 68.6 <0.001 >23 68.9 67.3 31.4 Waist circumference (cm.) < 90 (Male) or < 80 (Female) 29.9 33.3 51.4 0.01 ≥ 90 (Male) or ≥ 80 (Female) 70.1 66.7 48.6 Smoking Never or Former 96.8 93.6 94.3 0.03 Current 3.2 6.4 5.7
PAD = peripheral arterial disease, BP = blood pressure, LDL = low density lipoprotein, BMI=body mass index importance of the risk factor control to the patients do not understand the importance of risk factor with PAD. Only 20-35 % of severe PAD group had good management in these patients. There had been a lack control in blood pressure, LDL and HbA1C. of direct evidence about the benefits of risk factor Since severe PAD patients mainly came to see modifications such as lipid-lowering therapy for diabetic physicians due to rest pain, gangrene or ulcer, they are patients with PAD. Some clinicians might be concerned waiting for surgery. We hypothesised that some about the cost-effectiveness of lipid-lowering treatment, physicians believe that once a patient has PAD, risk more specifically about its lack of cost-effectiveness for factor control is no longer beneficial. In other words, patients with PAD. This might come from confusion it is already too late to treat a severely stenotic lesion. about the correct values for good control of plasma While it is true that a severely stenotic lesion is highly cholesterol. There are several recommendations in unlikely to regress in a short period, it can reduce the clinical practice. For example, in the USA, the National cardiovascular event. For example, in many trials Cholesterol Education Program used 100 mg/dl as the statins have reduced the incidence of myocardial desirable blood cholesterol level8. However, for PAD infarction, which is the main cause of death peri- patients who have a high risk of developing associated operatively and long-term after surgery7. To achieve a ischaemic heart disease the recommendation is 70 good outcome for these patients, adequate risk factor mg/dl rather than 100 mg/dl (AHA guideline)3,4. management would be beneficial. Therefore, it is important to educate all clinicians who This study found that hyperlipidemia was not well take care of vascular patients about proper management controlled in clinical practice in Thailand. The main of risk factor control. reason is probably no one takes responsibility for the In addition, serum cholesterol is an independent management of hyperlipidaemia. Some general predictor of atherosclerosis9,10 and the level of practitioners treat hyperlipidemia intensively in patients hyperlipidaemia is a prognostic factor for recurrence with PAD and some do not. Some vascular surgeons do of peripheral artery stenosis following carotid not treat it because of the expense and ask general endarterectomy (CEA)11,12. If hyperlipidaemia can be practitioners to handle it instead, which makes this adequately treated, the patency of the artery may last matter even more complicated if general practitioners longer. Ultrasonic evidence has shown that lipid- Vol. 36 No. 2 PAD in Diabetic Patients 71
lowering treatments result in regression of carotid clinical practice. This problem persists in spite of plaques and Furberg et al. found that the average much evidence of the benefits of risk factor control. intimal medial thickness was significantly lower in Management of risk factor should be strictly performed. patients taking lovastatin than in those of control The findings in current study support the need for group13. This might be particularly beneficial for greater education of clinicians who involved in the patients with bilateral PAD who have a unilateral leg management of risk factor of PAD patients. To control surgery and conservative treatment for the other lesion. the majority of these risk factors effectively, much Thus proper management of hyperlipidaemia after expertise is needed and many services are involved, surgery for PAD on one side may prevent progression including those that lead to a cessation of smoking. A of the disease on the non-operative side. risk factor clinic with a multidisciplinary approach It is interesting to note that patients in severe might therefore be beneficial. PAD group had significantly lower BMI and waist circumference than those in other groups. This can ACKNOWLEDGEMENT partially be explained by BMI in PAD patients which was lower than those in non PAD patients. Although This study was supported by the Health Systems originally we believe obesity is an important risk factor Research Institute and National Research Council of of PAD, this seems contrast to our original concept. Thailand. We also would like to thank personnel in We hypothesized an explanation that these patients OPD 21, OPD 9, Pain Clinic and OPD 101 at Maharaj were in starvation stage due to excruciating foot pain Nakorn Chiang Mai Hospital and outpatient in severe PAD. Further nutritional analysis is needed to department at Chiangrai Prachanukroh Hospital and confirm this hypothesis. Lamphun Hospital for assisting our recruitment. We Another risk factor that can be managed is also would like to thank Dr. Ampica Mangklabruks, Dr. smoking. Eighteen patients (6.3%) in our study were Piyamitr Sritara, Dr. Arintaya Phrommintikul, Ms. still smoking (Table 3). It is well known that smoking Chonlisa Chariyalertsak, Ms. Antika Wongthanee and is strongly associated with atherogenesis, and it has Ms.Orapin Pongtam for designing this study. been reported in many papers that smoking is associated with recurrent PAD after surgery14. Cessation of smoking rapidly decreases the risk of ischaemic stroke REFERENCES and myocardial infarction15,16. This is another way in 1. Castelli WP. Epidemiology of coronary heart disease: the which long-term results for these patients may be Framingham study. Am J Med 1984;76:4-12. improved. 2. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, Fowkes FG. Inter-society consensus for the management of Limitations peripheral arterial disease (TASC II). J Vas Surg 2007;45 Suppl S:S5-67. There are some limitations of this study. Firstly, 3. Hirsch AT, Haskal ZJ, Hertzer NR, Bakal CW, Creager MA, all patients in this study came from one region in Halperin JL, et al. ACC/AHA 2005 guidelines for the Northern part of Thailand, which might not represent management of patients with peripheral arterial disease all PAD patients in this country, including those in (lower extremity, renal, mesenteric, and abdominal aortic): other settings. Secondly, only the LDL was used as a executive summary a collaborative report from the American Association for Vascular Surgery/Society for Vascular Surgery, measurement of lipid abnormality, but low serum Society for Cardiovascular Angiography and Interventions, HDL and high triglyceride levels may be better measures Society for Vascular Medicine and Biology, Society of of lipid abnormality in some subgroup. Interventional Radiology, and the ACC/AHA Task Force on Practice Guidelines. J Am Coll Cardiol 2006;47:1239-312. 4. Rooke TW, Hirsch AT, Misra S, Sidawy AN, Beckman JA, Findeiss CONCLUSION LK, et al. 2011 ACCF/AHA focused update of the guideline for the management of patients with peripheral artery disease In this study it was found that one-third of PAD (updating the 2005 guideline): a report of the American patients had proper control for each risk factor College of Cardiology Foundation/American Heart according to the recommended value. Awareness of Association Task Force on Practice Guidelines. J Am Coll the significance of risk factor control is inadequate in Cardiol 2011;58:2020-45. 72 Arworn S, et al. Thai J Surg Apr. - Jun. 2015
5. Thai Medical Council TRCoPoT. Guideline for tertiary care in 1985;202:28-35. arterial disease prevention Bangkok: 2011. 12. Rapp JH, Qvarfordt P, Krupski WC, Ehrenfeld WK, Stoney RJ. 6. Aspry KE, Holcroft JW, Amsterdam EA. Physician recognition of Hypercholesterolemia and early restenosis after carotid hypercholesterolemia in patients undergoing peripheral and endarterectomy. Surgery 1987;101:277-82. carotid artery revascularization. Am J Prev Med 1995;11:336- 13. Furberg CD, Adams HP Jr, Applegate WB, Byington RP, 41. Espeland MA, Hartwell T, et al. Effect of lovastatin on early 7. Musser DJ, Nicholas GG, Reed JF 3rd. Death and adverse carotid atherosclerosis and cardiovascular events. cardiac events after carotid endarterectomy. J Vas Surg Asymptomatic Carotid Artery Progression Study (ACAPS) 1994;19:615-22. Research Group. Circulation 1994;90:1679-87. 8. Talbert RL. New therapeutic options in the National Cholesterol 14. Cuming R, Worrell P, Woolcock NE, Franks PJ, Greenhalgh RM, Education Program Adult Treatment Panel III. Am J Managed Powell JT. The influence of smoking and lipids on restenosis Care 2002;8:S301-7. after carotid endarterectomy. Eur J Vas Surg 1993;7:572-6. 9. Joensuu T, Salonen R, Winblad I, Korpela H, Salonen JT. 15. Ockene IS, Miller NH. Cigarette smoking, cardiovascular Determinants of femoral and carotid artery atherosclerosis. J disease, and stroke: a statement for healthcare professionals Int Med 1994;236:79-84. from the American Heart Association. American Heart 10. Rittoo D, Cramb R, Odogwu S, Khaira H, Duddy M, Smith S, et Association Task Force on Risk Reduction. Circulation 1997; al. Worsening lipid profile is associated with progression of 96:3243-7. carotid artery stenosis. Int Angiol 2001;20:47-50. 16. Wolf PA, D’Agostino RB, Kannel WB, Bonita R, Belanger AJ. 11. Das MB, Hertzer NR, Ratliff NB, O’Hara PJ, Beven EG. Recurrent Cigarette smoking as a risk factor for stroke. The Framingham carotid stenosis. A five-year series of 65 reoperations. Ann Surg Study. JAMA 1988;259:1025-9. The THAI Journal of SURGERY 2015;36:73-80. Official Publication of the Royal College of Surgeons of Thailand Original Article
Safety and Feasibility of Pancreaticoduodenectomy with Venous Reconstruction: Maharaj Nakorn Chiang Mai Hospital’s Experience
Sunhawit Junrungsee MD* Wasana Ko-iam RN† Anon Chotirosniramit, MD* Trichak Sandhu, MD* *Division of Hepatobiliary-pancreatic Surgery, Department of Surgery, Faculty of Medicine, Chiang Mai University, Thailand, †Research Unit, Department of Surgery, Faculty of Medicine, Chiang Mai University, Thailand
Abstract Background: Pancreaticoduodenectomy (PD) with venous reconstruction is becoming a more common practice in locally advanced pancreatic adenocarcinoma in which tumors adhere or invade portal vein (PV) or superior mesenteric vein (SMV). Our objective was to compare short-term outcomes of PD with venous reconstruction (VR group) to a standard PD (non-VR group) in our center. Materials and Methods: This retrospective study recruited 43 patients who underwent PD by a single surgeon from July 2010- June 2013. Seventeen patients underwent PD with VR and 26 were in non-VR group. All charts were reviewed, including demographic data, operative details, and complications. Disease-free survival (DFS) and overall survival were evaluated. Results: Of 43 patients, there were significant differences in operative time and amount of blood loss between both groups. However, length of stay of both groups had no difference. Pathological PV/SMV invasion in VR group was 41.7%. Mortality rate was not different in both groups. Overall survival and DFS in the VR group were 9 months and 6.5 months respectively. Conclusion: PD with venous reconstruction is safe and effective in selected cases. Short-term outcomes are similar to those in non-vascular reconstruction group. Mortality rate in both groups are high. PD with VR should remain the standard of care for locally advanced peri-ampullary cancer.
Keywords: Pancreaticoduodenectomy, whipple operation, venous reconstruction, pancreatic cancer
INTRODUCTION reported globally, with 213,000 resultant deaths2. Pancreatic adenocarcinoma is the fourth leading Pancreatic resection is known as one of the most cause of cancer death in men, after lung, prostate, and complicated and technically challenging surgical colorectal cancer, and the fifth leading cause of cancer procedures known to general and hepatobilia- death in women, following lung, breast, colorectal, rypancreatic (HBP) surgeons. It is not only technically and ovarian cancer1. In 2000, 217,000 new cases were challenge, but also substantial logistical strain on health
Correspondence address: Sunhawit Junrungsee, MD, Division of Hepatobiliary-pancreatic Surgery, Department of Surgery, Faculty of Medicine, Chiang Mai University, 110 Intravarorod Road, Sripoom, Muang, Chiang Mai 50200, Thailand; Telephone: +66 5394 5533; Email: [email protected]
73 74 Junrungsee S, et al. Thai J Surg Apr. - Jun. 2015
care resources. For tumors that are located in the tests and good nutritional status before their operation peripancreatic head area, pancreaticodudenectomy was performed. Operative details, including patient (PD) is recognized as the optimal definite treatment. factors (gender, age, weight, type of tumor and tumor In 1973, Fortner3 showed that a more radical resection size), operative time, blood loss, complications and should improve survival by better tumor clearance. one year survival were recorded. The neoadjuvant When tumor adhered to portal vein (PV) or superior treatment protocol was not established during time of mesenteric vein (SMV), often regarded as unresec- study period. All works were performed after the tability, en bloc resection of the involved vessels should approval of Institutional Review Board. be attempted. As surgical expertise has improved over the past few decades, and the quality of high-volume Operative details centers especially for technically demanding surgeries, Majority of cases underwent standard pylorus- such as pancreatic resection has been proven, thus, it preserving PD (PPPD). Classical PD was performed in is not surprising that perioperative morbidity and cases of inadequate surgical margins. Pancreatic-biliary- mortality rates reported for pancreatic resections with enteric continuity was performed by using a retrocolic or without PV or SMV resection are identical4,5,6,7,8. jejunal limb with end-to-side pancreaticojejunostomy Moreover, the long-term survival of patients who have (duct to mucosa), end-to-side hepaticojejunostomy, had pancreatic resection in combination with PV and end-to-side duodenojejunostomy in PPPD. The resection is far superior compared with those in whom author preferred a small feeding tube number 5 or 8 only palliative surgical management was performed. French for stenting pancreaticojejunostomy anasto- Based on these facts, we would like to report the results mosis when encountered with small pancreatic duct. from 46 patients that underwent PD and PD with The external pancreatic stent was never used. When venous reconstruction (VR) to compare perioperative evidence of direct extension to the PV or SMV was outcome. One-year survival in VR group was also encountered, en bloc vascular resection was performed examined. with vascular reconstruction accomplished by either primary end-to-end anastomosis or by placing an interposition graft. We ligated the splenic vein and PATIENT SELECTION AND METHODS performed proximal and distal control at PV-SMV. Data of patients who underwent PD by the first The vein was divided and the remaining tissue was author (S.J.) at the Division of Hepatobiliary-pancreatic dissected at the retroperitoneal margin. In our early Surgery, Department of Surgery, Maharaj Nakorn experience, we routinely used a shunt between the PV Chiang Mai Hospital, Faculty of Medicine, Chiang Mai and SMV to reduce visceral congestion and liver University between July 2010 and June 2013 were ischemia. More recently, a shunt is used just in complex reviewed. Tri-phase computed tomography (CT) scan case as an interposition graft. For reducing the gap of the abdomen and pelvis or magnetic resonance between vascular anastomosis, falciform and triangular imaging (MRI) of the abdomen and pelvis were used ligaments were divided to mobilize right and left lobes to assess resectability. Endoscopy and biopsy were of liver. In some cases, we clamped the superior performed in patients who were suspected of carcinoma mesenteric artery (SMA) to reduce congestion of small of the ampulla of Vater and carcinoma of the bowel. duodenum. Patients who had clinical cholangitis, Clamping time, operative time and blood loss malnutrition or needed to wait for a long time before were recorded in operative note. Liver function test surgery received placement of preoperative biliary was examined when patient returned to ward. Patency drainage by Percutaneous Transhepatic Biliary of vascular anastomosis was checked by Doppler Drainage (PTBD) or Endoscopic Retrograde ultrasound on the first post-operative day. Amylase Cholangio-Pancreatography (ERCP) with plastic stent. levels in serum and drain fluid were examined on the The decision to perform endoscopic biliary drainage third post-operative day. When patients had a pancreatic depended on clinical situation and opinion of the fistula according to the International Study Group of gastroenterologists or surgeons. We waited until this Pancreatic Fistula (ISGPF) criteria9, the drain was group of patients regained near normal liver function retained until volume and level of amylase declined to Vol. 36 No. 2 Safety of Pancreaticoduodenectomy with Venous Reconstruction 75
normal. If all laboratory values were in normal range Table 1 Demographics, tumor histology, operative detail and and patients resumed oral food, they were discharged. post-operative data of all 43 patients that underwent PD from July 2010 to June 2013. All patients were scheduled to meet the attending staff in the Division of HBP Surgery at two weeks, one Variable month and then every three months until one year. Age; year 59.2 ± 10.8 ± Pathology reports including type and grading of tumor, Weight; kg 55.9 11.0 Pre-op albumin 3.01 ± 0.8 lymph node status, margin, venous invasion status Carbohydrate antigen 19-9 (range) 566.6 (0.7- 2182) were reported by a specialized HBP pathologist. The Primary (%) use of adjuvant chemotherapy in each patient was Pancreatic adenocarcinoma 13 (30.2) determined by clinical performance of patient and Cholangiocarcinoma 7 (16.3) agreement between oncologist and surgeon. Ampulla of vater carcinoma 17 (39.5) Duodenal carcinoma 1 (2.3) Statistical analysis Pancreatic neuroendocrine tumor 2 (4.6) Benign disease 3 (7) Data were analyzed by STATA version 11.0. Operation (%) Continuous data from both groups were compared by Non Vascular reconstruction 26 (60.5) student’s t test to report p-value, if data had a normal Vascular reconstruction 17 (39.5) distribution. When data had distribution that was not Mean operative time; min (range) 487.4 ± 126.2 (215- 840) ± normal, the p-value was reported by a rank sum test Mean Blood loss; ml (range) 774.1 686.6 (50 - 3000) Mean Hospital stay; day (range) 19.03 ± 23.6 (7- 125) (Mann-Whitney U test). Categorical data were 30 day mortality 4 (9.3%) compared by Chi-square test/Fisher’s exact test. Median Morbidity (%) 8 (18.6) survival time and one year survival were analyzed. Sepsis 3 (7) Pancreatic fistula 5 (11.6) Intraabdominal fluid collection 1 (7) ESULTS R Acute renal failure 1 (2.3) Bleeding 2 (4.6) During the study period, records from 43 patients Re-operation 2 (4.6) who underwent PD were evaluated. The baseline characteristics of the patients are shown in Tables 1 and 2. There were 21 female and 22 male patients. The performed by end to end anastomosis in nine patients primary diseases treated were pancreatic adenocar- (Figure 3). In seven patients an autogenous vein graft cinoma (n=13), cholangiocarcinoma (n=7), carcinoma was used for the left renal vein in four cases (Figure 4), of ampulla of Vater (n=17), benign disease (n=3), and for the inferior mesenteric vein in three cases pancreatic neuroendocrine tumor (n=2) and duodenal (Figure 5). Polytetrafluoroethylene (PTFE) graft was carcinoma (n=1). Overall morbidity and 30-day used in one case due to lack of autogenous vein graft. mortality rates were 18.6% and 9.3 % respectively. The The median operative blood loss in this group was pancreatic fistula rate in overall patients was 11.6 %. 1,217.6 mL. (300 - 3,000 mL) and the mean vascular In the VR group, the mean patient age was 56.7 gap was around 3 cm. (1.8- 5 cm). The vascular years. Patients receiving treatment included pancreatic anastomosis was evaluated for all patients by adenocarcinoma (n=10), cholangiocarcinoma (n=4), intraoperative and postoperative Doppler ultrasound. ampulla of Vater carcinoma (n=1) and nonfunctioning None of these patients had occlusion of their pancreatic neuroendocrine tumor (n=2) as shown in reconstructed vessel. There was a positive margin in 4 Figure 1. Two patients underwent concomitant arterial patients (23.5%) and lymph node metastasis was found reconstruction because their replaced right hepatic in 11 patients (64.7 %). In this series, pathology of the artery came from the superior mesenteric artery and excised pancreas was confirmed with PV or SMV passed through the tumor. A colectomy was performed invasion in 41.7% of cases. due to invasion of the colic vessel and colonic serosa. Two patients (11.7%) died in the postoperative Proportion of vascular reconstruction technique was period from septic shock at 8 and 14 days after surgery. shown in Figure 2. Venous reconstruction was The median hospital stay was 16 days (9 - 60). One 76 Junrungsee S, et al. Thai J Surg Apr. - Jun. 2015
Table 2 Comparative demographic and operative data between non venous reconstruction (non VR) and venous reconstruction (VR) group
Variable non VR VR p-value
Number (%) 26 (60.47) 17 (39.53) - Age in years 61 ± 12.3 56.7 _ 10.4 0.216 Weight; kg 55.9 ± 10.8 55.9 _ 11.7 0.990 Mean blood loss; ml (range) 373.1 ± 276.1 (50-1200) 1217.6 ± 741.8 (300-3000) <0.001 Mean operative time; min (range) 390.4 ± 79.39 (300-530) 564.1 ± 119.7 (440-840) <0.001 Mean hospital stay; day (range) 18.4 ± 24.16 (7-125) 16.17 ± 12.62 (9-60) 0.726 Tumor size; cm 3.3 ± 5.02 3.9 ± 1.57 0.639 Positive margin (%) 2/26 (7.7) 4/17 (23.5) 0.539 Node positive (%) 7/26 (26.9) 11/17 (64.7) 0.026 Pancreatic fistula (%) 3/26 (11.5) 2/17 (11.8) 1.000 30 days mortality (%) 2/26 (11.5) 2/17 (11.8) 1.000 Re-operation 1/26 (3.8) 1/17 (5.9) 1.000
Table 3 Operative characteristics and perioperative complications in 17 patients in venous reconstruction group 25 Pancreatic adenocarcinoma
1234 Variables 1234 1234 Distal cholangiocarcinoma
123
123 63 6 123 Ampulla of vater carcinoma Primary disease 6 Distal cholangiocarcinoma Pancreatic adenocarcinoma 10 (58.8) Cholangiocarcinoma 4 (23.5) Ampulla of vater carcinoma 1 (5.9) Pancreatic neuroendocrine tumor 2 (11.8) ± Figure 1 Proportion of primary tumors in VR group. Mean mesoportal clamping time; min (range) 31.8 6.2 (23-42) Mean vascular gap; cm(range) 3 ± 1.2 (1.8-5) Mean post-operative ALT 182.7 ± 395 30 days mortality (%) 2 (11.7) 6 Pathological SMV invasion 7 (41.7) 18 Major perioperative complication* End to end anastomosis Reoperation 1 (5.9) 1234
1234 1234 Lt. renal vein graft Bleeding 1 (5.9)
123
123
23 123 Inferior mesenteric vein graft Pancreatic fistula (%) 2 (11.7) 53 PTFE graft Intra-abdominal fluid collection (%) 1 (5.9) Pulmonary complications 1 (5.9) Sepsis syndrome (%) 2 (11.7) Acute renal failure (%) 3 (17.6)
Figure 2 Proportion of vascular reconstruction technique in VR group Subsequently the patient developed sepsis and died. In the non VR group the majority of primary patient in this group had pancreatic fistula and multiple tumors were ampulla of Vater adenocarcinoma. The intra-abdominal fluid collections that required a 60- mean operative time was 390 minutes and mean blood day hospital stay. Another patient sustained acute loss was 373 mL. Two patients (11.5%) in this group renal failure from the sacrificed left renal vein for died from septic shock and hospital-acquired venous graft. However, after multiple hemodialysis pneumonia. A pancreatic fistula was observed in 3 sessions, his renal function recovered to his normal patients (11.5 %) that were treated conservatively. All level. Reoperation was performed in one patient due operative data is shown in Figures 1, 2 and Tables 1, 2, to bleeding from inferior pancreaticoduodenal artery. 3. Vol. 36 No. 2 Safety of Pancreaticoduodenectomy with Venous Reconstruction 77
Table 4 Comparative survival data between non venous reconstruction (non VR) and venous reconstruction (VR) group
non VR VR
Median survival (month) - 9 1 year survival (%) 83.3 27.3 Disease free survival (month) 12 6.5 _ 3.6 Metastatic site Liver (%) 4 (36) Peritoneum (%) 2 (18) Lung (%) 1 (9) Node (%) 1 (9)
Figure 3 End to End anastomosis
Figure 6 One-year survival graph of non VR and VR group. Figure 4 Left renal vein graft Group 1: non VR group, Group 2: VR group
survival was 9 months. Median disease-free time is 6.5 months. None of the patients in the non VR group died in one year except for three patients who died in the early post-operative period. All of those who survived are still free of disease. The estimated one- year survival in the non VR group was 83.3%.
DISCUSSION
SMV resection was first reported by Moore in 19519. In 1956, Kikuchi described the use of a Figure 5 IMV graft homologous vein graft and polyethylene splint for reconstruction of the portal vein. In 1963, Asada et al.10 Survival analysis reported two cases of reconstruction with homologous Survival data and survival curves are shown in vein grafts. In 1965, Siegel et al.11 described autologous Table 4 and Figure 6. There were significant differences vein graft reconstruction. In 1966, Longmire12 used a in survival between both groups. In the VR group, the prosthetic graft for reconstruction. In 1973, Fortner estimated one-year survival was 27.3% and median proposed a regional pancreatectomy for removal of 78 Junrungsee S, et al. Thai J Surg Apr. - Jun. 2015
peripancreatic soft tissue in conjunction with major insufficiency. We suppose that because this patient vascular resection, but the frequency of complications had massive blood loss during the operation and this was high, and there was no improvement in survival13. might be the cause that aggravated acute tubular VR was initially performed with the objective to necrosis. Finally, his normal renal function resumed maximize soft tissue and lymphatic excision. However, after receiving hemodialysis for a couple of weeks. The from subsequent work done by Yeo et al.14, it became left renal vein provides a graft with good length, evident that widening the surgical margins to include optimal caliber, and is easily accessible. The left renal more lymphatic tissues had little impact on survival. vein typically provides a graft of 3-4 cm in length when One large series demonstrated that with proper patient harvested from the junction of the left gonadal and left selection and surgeon experience, VR can be adrenal vein proximally and the inferior vena cava performed safely with complication rates similar to distally, although some reports have indicated lengths standard pancreatic resections15. The low complication up to 6 cm19. The caliber of the graft allows for excellent rate reported with vascular reconstruction and the flow. The ease of harvesting the graft is also an important improving operative morbidity and mortality after PD consi-deration. Exposure of the left renal vein can be makes it reasonable to consider vascular resection to accomplished through a standard PD incision, without achieve an R0 resection16. requiring any further prepping, an additional incision, Siriwardana et al. have reviewed the outcome of or the need for an additional operating team. PV reconstruction during pancreatic resection for Furthermore, use of the left renal vein leaves the cancer17. Fifty two non-duplicated papers were patient with all possible routes of central venous access. reviewed. Pathological evidence of PV invasion was Inferior mesenteric vein (IMV) is another option for a detected in 668 (63.4%) of 1,054 portal vein resection venous graft for the distal SMV segment because the specimens. The rates of invasion ranged from 3% to diameter of the IMV is usually smaller than the PV/ 86% in 30 studies. PV margins were positive in 346 SMV segment. The IMV can be harvested from a (39.8%) of 870 patients with PV resection in 23 studies, standard incision at the junction between the splenic with a range of 0-85%. Postoperative morbidity ranged vein and IMV. The length of an IMV graft can be 5-6 from 9% to 78%. There were 73 (5.9%) reported cm. However, the disadvantage of an IMV graft is deaths among 1,235 patients in 39 studies that reported interrupted collateral visceral venous drainage during mortality after PV resection. The reported mortality the vascular reconstruction phase that may cause bowel rates in these studies ranged from 0 to 26%. However, congestion. the mortality rate has decreased to <5% in recent years. The major reconstruction technique in the VR In this present study, mortality was found in 2 patients group was primary end-to-end anastomosis. By complete (11%) from the VR group which was more than many mobilization of liver and colon, the gap between the recent studies such as Nakao et al. (3.6% in VR group). anastomosis could be reduced approximately 2 to 3 These 2 patients died from gram negative sepsis on cm. The Ligamentum Teres and Falciform ligament day 8 and 14 after surgery and we could not conclude should be fixed in a proper position after all anasto- that his death was correlated with the operation. moses are completed to reduce tension between Morbidity in the VR group was found in 2 patients vascular anastomoses. The intra-operative ultrasound (11%) that included pancreatic fistula, intra-abdominal was utilized to demonstrate patency of the anastomosis. fluid collection and acute renal failure. The interes- We did not find any vascular complication in all patients. ting point from one patient in this group is that he More invasive operations (concomitant arterial developed acute renal failure after his left renal vein reconstruction and combine colectomy) have not was sacrificed to be a substitution for SMV. Previous shown a higher mortality. In case in which the RHA work demonstrated that good collateral flow and originated from the SMA and passed through the functional capacity of the left kidney was preserved specimen, the RHA was divided and reconstructed by despite ligation of the left renal vein. McCullough and anastomosis to the gastroduodenal artery (GDA) or colleagues18 reported that after a right nephrectomy common hepatic artery proper (CHA). Indications for and ligation of the left renal vein for malignancy, only colectomy in PD are for achieved R0 resection and one of three patients experienced transient renal ischemia of colon. We preferred to perform colonic Vol. 36 No. 2 Safety of Pancreaticoduodenectomy with Venous Reconstruction 79
anastomosis immediately after the colon was resected operations by a single primary surgeon indicating a to minimize contamination in the operative field. strong point in this study. Histological evidence of true PV/SMV invasion In summary, PD with VR is safe to perform. was detected in 41.7% which was lower than many Morbidity and mortality is not different from standard recent studies. However, the resection margin in the PD in selected patients. However, survival in the VR VR group was positive in 4 patients (23.5 %) which was group is small and significantly different from the non better than nearly 40% (0 -85 %) from the Siriwardana VR group that may be influenced by the type of study17. Estimated blood loss and operative time in the primary tumor. Future studies may be needed to VR group was statistically different from the non VR increase the number of patients and compare survival group but not far from Siriwardana study (median in a palliative group. Until then, we suggest that this 1,750 ml). However, length of hospital stay in both type of procedure be performed by a hepatobiliary- groups showed no difference. These results showed pancreatic surgeon in high-volume centers. that a PD with VR is safe and the mortality is not higher than in a non VR group. Our mortality rate in VR group was 11% which was comparable to the mortality REFERENCES rate in the current era from the Siriwardana study17. 1. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer Two patients in non VR group who died in the early statistics, 2009. CACancer J Clin 2009;59(4):225-49. post-operative period were elderly (aged 70 and 80) 2. Hariharan D, Saied A, Kocher H.M. Analysis of mortality rates which might be the risk factor for morbidity. Another for pancreatic Cancer across the world. HBP 2008;10(1):58-62. 3. Fortner JG. Regional resection of cancer of the pancreas: a patient in the non VR group had bleeding from GDA new surgical approach. Surgery. 1973;73:307-20. stump that was detected on post-operative day zero. 4. Bachellier P, Nakano H, Oussoultzoglou PD, et al. Is We brought her back to the operative room to stop pancreaticoduodenectomy with mesentericoportal venous bleeding. However, massive blood loss resulted in resection safe and worthwhile? Am J Surg. 2001;182:120-29. acute renal failure and pancreatic anastomosis failure 5. Carrére N, Sauvanet A, Goere D, et al. Pancreatico- later. She had re-bleeding three times despite multiple duodenectomy with mesentericoportal vein resection for adenocarcinoma of the pancreatic head. World J Surg attempts to embolize the vessel. Eventually, the patient 2006;30:1526-35. died 120 days after surgery from hospital-acquired 6. Leach SD, Lee JE, Charnsangavej C, et al. Survival following pneumonia. We realized that this is not a good result pancreaticoduodenectomy with resection of the superior for PD in modern era which we should expect zero mesenteric-portal vein confluence for adenocarcinoma of mortality in any kind of procedure. the pancreatic head. Br J Surg 1998;85:611-17. Survival analysis shows that survival in the VR 7. Nakao A, Takeda S, Inoue S, et al. Indications and techniques of extended resection for pancreatic cancer. World J Surg group was definitely worse than the non VR group. 2006;30:976-82. There was a 27.3% survival beyond one year. All patients 8. Riediger H, Makowiec F, Fischer E, Adam U, Hopt UT. in the non VR group except three patients who died in Postoperative morbidity and long-term survival after the early period achieved one year survival. However, pancreaticoduodenectomy with superior mesenterico-portal 55% in the non VR group had ampulla carcinoma and vein resection. J Gastrointest Surg 2006;10:1106-15. we know that this type of tumor has a better prognosis 9. Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery than other periampullary carcinoma, particularly 2005;138(1):8-13. pancreatic cancer. We cannot conclude that the factors 10. Asada S, Itaya H, Nakamura K, Isohashi T, Masuoka S. Radical that impact survival is the type of operation because of pancreatectomy and portal vein resection: report of two the heterogeneity of the population in this study. successful cases with transplantation of portal vein. Arch Surg However, we observed that aggressive tumors that 1963;87:609-13. need vascular reconstruction tend to have grave survival. 11. Sigel B, Bassett JG, Cooper D, Dunn MR. Resection of the superior mesenteric vein and replacement with a venous The major limitation of this study is the retrospective autograft during pancreaticoduodenectomy: case report. nature that could result in variations in pre-operative Ann Surg 1985;162:941-45. selection criteria and mixed patient populations. The 12. Longmire WP Jr. The technique of pancreaticoduodenal number of patients in this study is small. However, our resection. Surgery 1966;59:344-52. study showed a one year result from a single center and 13. Amano H, Miura F, Takada T, Sano K. Portal vein resection in 80 Junrungsee S, et al. Thai J Surg Apr. - Jun. 2015
pancreaticoduodenectomy (with video). J Hepatobiliary 2007;11:1168-74. Pancreat Sci. 2012;19:109-15. 17. Siriwardana HP, Siriwardena AK. Systematic Review of 14. Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreatico- Outcome of Synchronous Portal- superior Mesenteric Vein duodenectomy with or without distal gastrectomy and Resection During Pancreatectomy for Cancer. Br J Surg extended retroperitoneal lymphadenectomy for periam- 2006;93: 662-73. pullary adenocarcinoma, part 2: randomized controlled trial 18. McCullough DL, Gittes RF. Ligation of the renal vein in the evaluating survival, morbidity, and mortality. Ann Surg. solitary kidney: Effects on renal function. J Urol 1975;113 (3):295- 2002;236:355-66. 98. 15. Tseng JF, Raut CP, Lee JE, et al. Pancreaticoduodenectomy 19. Miyazaki M, Itoh H, Kaiho T, et al. Portal vein reconstruction at with vascular resection: margin status and survival duration. J the hepatic hilus using a left renal vein graft. J Am Coll Surg Gastrointest Surg 2004;8:935-49. 1995;180(4):497-98. 16. Al-Haddad M, Martin JK, Nguyen J, et al. J Gastrointest Surg The THAI Journal of SURGERY 2015;36:81-87. Official Publication of the Royal College of Surgeons of Thailand Original Article
Comparison of Locking Plate and Percutaneous K-wire in Treatment of Proximal Humerus Fractures
Somboon Wutphiriya-angkul, MD Sawangdandin Crown Prince Hospital, Sakon Nakhon, Thailand
Abstract Background and Objectives: The most effective method for the surgical treatment of proximal humerus fractures has not been established. Two commonly used techniques are locking plate and percutaneous K-wire. We performed a retrospective study to compare these two treatment strategies. Materials and Methods: A total of 62 patients were selected for review. Each patient was treated by one of the two methods. Data including operative time, blood loss, pain scale, neck-shaft angle, shoulder score, and operative complications were collected. Results: Thirty patients were treated with locking plate fixation, and the rest were treated with percutaneous K- wire fixation. Percutaneous K-wire group was associated with significantly shorter operative time and length of hospital stay, lesser blood loss and lower postoperative pain (P < 0.05), but had more complications (P = 0.046). There were no significant differences in shoulder score and neck-shaft angle between the two groups. Conclusion: Percutaneous K-wire technique is similar in effectiveness to the locking plate in the treatment of proximal humerus fractures with a shorter operative time and hospital stay, lesser blood loss and lower postoperative pain, but had more complications
Keywords: K-wire fixation, plate fixation, proximal humerus fractures
INTRODUCTION morbidities: conservative treatment4, open reduction Proximal humerus fractures are on the rise due to and internal fixation (ORIF)5, joint replacement6,7 road traffic accidents and increase in the incidence of and percutaneous fixation7,8. Good clinical results osteoporosis. They constitute about 4-5% of all were obtained in 92% of cases treated with ORIF, 87% fractures1. Minimally displaced fractures, regardless for cases treated with conservative treatment and 87.5% of the number of fracture lines, can be treated with of cases treated with shoulder arthroplasty6,7. Open closed reduction but displaced fractures require reduction and internal fixation (ORIF) has the anatomical reduction with internal fixation2,3. advantages of anatomical reduction and early Several treatment modalities have been proposed, mobilization. It may however be associated with higher depending upon the fracture pattern, patients’ age rates of infection, damage to arteries and nerves, and level of activity, and associated medical co- reduction loss, implant failure, nonunion or malunion,
Correspondence address: Somboon Wutphiraya-angkul, MD, Sawangdandin Crown Prince Hospital, Sakon Nakhon 4710, Thailand; Email: [email protected]
81 82 Wutphiriya-angkul S Thai J Surg Apr. - Jun. 2015
impingement, and osteonecrosis of the humeral was exposed through a delto-pectoral approach and head9,10. If adequate reduction is not achieved or fracture fragments were reduced. The reduced medial buttress is not sufficient, reduction loss and fragments were held in position with K-wires under subsequent fixation failure are possible, especially in guidance of image intensifier. Definitive fixation with osteoporotic elderly patients11. locking plate was done with the plate positioned lateral Conservative treatment with closed reduction and to the bicipital groove, sparing the tendon of long percutaneous pinning has limited indications, less head of biceps. The plate was placed at least 1 cm distal blood loss, lower risk of neurovascular complications to the upper end of greater tubercle. The required and less interference with glenohumeral joint motion. lengths of the locking screws were determined with a This technique, however may not ensure anatomical direct measuring device over the K-wire, and at least 6 reduction and early mobilization. It is also associated locking screws were inserted in the humeral head. with pin tract infection and a long recovery period12. Lesser tuberosity was fixed with separate screws or The purpose of this study was to retrospectively compare wires if found to be avulsed. If the fracture site had a between the two techniques in terms of clinical outcomes. Also, the functional recovery and operative complication were examined in details.
MATERIALS AND METHODS
A retrospectively review was conducted between 2007 and 2014, 62 patients with proximal humerus fractures were surgically treated at the Sawangdandin Crown Prince Hospital. Inclusion criteria for this study were: (a) patients with 2, 3 or 4 parts of fractures, (b) acute and unilateral fracture, (c) internal fixation with either a locking plate or percutaneous K-wire fixation, and (d) normal shoulder function before injury. Exclusion criteria included: (a) pathological fractures, (b) primary or metastatic bone tumors, (c) major neurological deficits, (d) open fractures, (e) other Figure 1 Radiograph of a proximal humerus fracture of a 52- injuries to the same upper limb requiring surgery, (f) year-old woman fracture dislocation, and (g) any medical condition that excludes surgical treatment. The locking plate used in this study is Proximal Humeral Internal Locking System (titanium; thickness: 4.2 mm; width: 12 mm; length: 105-231 mm; Synthes) which was anatomically pre-contoured with 3 to 10 holes on the plate shaft and 9 holes for head screws. The proximal suture holes were applied to secure the tuberosity fragments and the plate. Operative techniques for each group were as follows: Locking plate group patients with proximal humerus fractures were treated with open reduction and internal fixation (ORIF) with locking plate. Surgery was performed under general anesthesia, patient in supine position with a small sand bag under the shoulder. All patients received prophylactic dose of intravenous antibiotic preoperatively. The fracture Figure 2 Radiographs 2 weeks after the surgery Vol. 36 No. 2 Comparison of Locking Plate and Percutaneous K-wrie in Treatment of Proximal Humerus Fractures 83
bone gap, an iliac bone graft was inserted. Range of radiographs were obtained and evaluated for fracture motion of shoulder and impingement were checked healing, non-union, malunion, loosening of implant, on the table. Wound was closed in layers with suction loss of reduction and avascular necrosis of head of drain (Figure 1, 2). Active range of motion (ROM) humerus. The criteria for radiographic healing was exercises were initiated on the first postoperative day. when all fragments showed substantial cortical Sutures were removed after 10 days. Follow up was at continuity. The neck-shaft angle was formed by the one week, then every month for 6 months, and then at first line drawn from the superior border to inferior 12 months for final evaluation. border of the articular surface. The second line was In percutaneous K-wire fixation group, surgery then drawn perpendicular to the first and through the was performed under general anesthesia with the center of the humeral head. The angle created by this patient in supine position. Near anatomical reduction line and the line bisecting the humeral shaft were was achieved by manual traction and arm mobilization. measured as the neck-shaft angle. K-wires under image intensifier were inserted The data recorded for all patients included depending on the number of fracture fragments. For operative time, blood loss, visual analogue pain scale difficult reduction, one K-wire was used as a joystick. (0, none to 10, severe) on the first post-operative day Care was taken on the orientation and pin placement and operative complications. At the follow-up in 6 to avoid injury to the axillary nerve, the radial nerve months, shoulder scoring system of Constant and and the anterior circumflex humeral vessels lying Murley13 was applied. In this system, both subjective medially. K-wires were left out of skin and bent at the and objective clinical data were included, with a extremity to control migration. If the fracture site had maximum score of 100 points. Pain (15 points), activities a bone gap, an iliac bone graft was inserted. (Figure 3, of daily living (20 points), range of motion of the 4). Patients were encouraged to start active mobilization shoulders (40 points), and muscle power (25 points) of wrist and elbow on the first postoperative day. were evaluated. The Student’s t-test, chi-square test, Dressing of the pin tracts were done on alternate days. and Fisher’s exact test were used in the comparison of Postoperative care was as for open reduction and outcomes between the two groups. Two-sided p-values internal fixation (ORIF) with locking plate technique. less than 0.05 were considered statistically significant. Standard antero-posterior and axillary view
Figure 3 Radiograph of a proximal humerus fracture of a-46- Figure 4 Radiograph 2 weeks after the surgery year old women 84 Wutphiriya-angkul S Thai J Surg Apr. - Jun. 2015
RESULTS summarized characteristic of fracture. There were 62 patients in the present study, with Mean operative time was 112 minutes (range, 89- an average age of 46.7 years (range 16- 68 years). All 141 minutes) for the locking plate group and 52 were followed up for more than six months after minutes (range, 41- 68 minutes) for the percutaneous discharge from the hospital. The average follow-up was K-wire group; this difference was significant. Average 12.7 months (range 6- 24 months). The patients were blood loss during surgery was 212 ml (range, 100-300 divided into two groups, based on the method of ml) for the locking plate group and 34 ml (range, 10- treatment. The locking plate group included 30 50 ml) for the percutaneous K-wire group; this patients and the percutaneous K-wire group included difference was significant. Average pain score on the 32 patients. Length of hospital stay was 5.6 days (range first post-operative day was 4.8 for the locking plate 4-7 days) for the locking plate group and 3.6 days group and 2.3 for the percutaneous K-wire group; this (range 3-4 days) for the percutaneous K-wire ; this difference was significant. Details of outcomes were difference was significant (p<0.01). The mechanisms given in Table 3. of injury and demographics data related to each group At 6 months follow-up of the locking plate group, were shown in Table 1. mean score for the affected shoulder using the scoring Both groups were similar in fracture patterns on system of Constant and Murley was 86.3 points, and Neer classification. The difference between in the two mean score for the contralateral shoulder was 94.2 groups was not significant (p=0.590). Table 2 points. In the percutaneous K-wire group, mean score
Table 1 The injury mechanism, length of hospital stay, preoperative demographics for both treatment groups
Locking plate Percutaneous Characteristics SD SD p-value (n=30) K-wire (n=32)
Gender (M/F) 16/14 18/14 0.818 Age (years): mean(SD) 45.6 4.3 48.3 5.1 0.792 Follow-up (months) : mean(SD) 12.7 5.6 12.8 6.2 0.922 Vehicular trauma: number (%) 16(53.3%) 19(59.3%) 0.632 Injury time (days) :mean(SD) 1.4 0.3 1.6 0.4 0.685 Length of hospital stay: mean(SD) 5.6 0.5 3.6 0.4 <0.01
Table 2 Characteristic of fractures
Fracture type Locking plate Percutaneous K-wire
2-part 14 17 3-part 12 9 4-part 4 6 Total 30 32
Table 3 Comparison of outcome between the two treatment groups
Outcome Locking plate Percutaneous SD SD p-value (n=30) K-wire (n=32)
Operative time (min) 112 (range, 89-141 ) 13 52 (range, 41-68 ) 7 <0.01 Blood loss (ml) 212(range, 100-300 ml) 48 34 ml (range, 10-50ml). 9 <0.01 Pain scale 4.8 (range, 2-8 ) 1.4 2.3 (range, 1-4 ) 0.6 <0.01 Vol. 36 No. 2 Comparison of Locking Plate and Percutaneous K-wrie in Treatment of Proximal Humerus Fractures 85
Table 4 Comparison of shoulder scores and neck-shaft angle between the two groups
Locking plate SD Percutaneous K-wire SD p-value
Affected shoulder 86.3 points 7.8 82.3 points (68-100) 8.2 0.312 (71-100) Contralateral shoulder 94.2 points 3.2 93.6 points (84-100) 3.8 0.532 (86-100) Neck-shaft angle (degrees) 126 3 128 (120-136) 4 0.914 (120-132)
for the affected shoulder was 82.3 points, and for the Table 5 Comparison of complication in both groups contralateral shoulder was 93.6 points. There was no Locking plate Percutaneous significant differences in the scores between the two K-wire groups. The scores varied depending upon the frac- Non union 1 1 ture type with the worst in four-part fractures. Infection 2 3 The neck-shaft angle measured on radiographs at Malunion 2 3 healing was 126˚ in the locking plate group and 128˚ in Avascular necrosis of humeral head 1 0 the percutaneous k-wire group. The difference between Implant loosening 0 3 neck-shaft angles in the two groups was not significant Loss of reduction 0 4 (p=0.914) (Table 4). Total 6 14 Post-operative complications were noted in 6 patients in the locking plate group and in 14 patients in the percutaneous K-wire group. In the locking plate This difference was significant (p=0.046) (Table 5). group, one patient had non-union (four-part fracture), two patients had infection (one patient with two-part DISCUSSION fracture and one patient with three-part fracture) and one had avascular necrosis of the humeral head (four- Proximal humerus fracture is the most common part fracture). For the patients with nonunion, bone fracture of the shoulder. It is the second most common grafting with removal of the previous implant was site of fracture in the upper limb after distal radius. performed. The patients with infection were treated These fractures have been treated with a wide range of with intravenous antibiotics after obtaining culture options, namely, non-operative, ORIF, percutaneous sensitivity reports. One patient with avascular necrosis screw/pin fixation and external fixation. Fractures of of the head of humerus refused arthroplasty. In the this region are common with both high-energy injuries percutaneous K-wire group, three patients had pin in people of all ages, as well as simple falls in older tract infection (one patient with two-part fracture and people with osteoporosis. In elderly patients, fragility two patients with three-part fracture), one patient had of the bone complicates the pattern of fracture. These non-union (three-part fracture), three patients had patients also have comorbidities which makes treating malunion (one patients with two-part fracture and two them even more challenging. Zyto and colleagues patients with three-part fracture), three patients had reported mean constant score of 65 points and no K-wire loosening (one patient with three-part fracture complications with conservative treatment compared and two patients with four-part fracture), and four with surgical approach, resulting in mean value of 60 patients had loss of reduction (two patients with three- points and with complications (avascular necrosis, part fracture and two patients with four-part fracture), infection)4. Magovern, Kenner, and Nho found good The patients with pin tract infection were treated with constant scores with surgery and relatively few daily dressings and antibiotics.Those with non-union complications, with better functional scores for were treated with ORIF and bone grafting. The patients percutaneous fixation8,14,15. Percutaneous fixation has with K-wire loosening underwent K-wire replacement. limitations of poor reduction of fracture fragments, 86 Wutphiriya-angkul S Thai J Surg Apr. - Jun. 2015
pin tract infection and long period of recovery8,15. But humeral head is the most common technical error in it has the advantages of less soft tissue stripping with plate fixation. Patients with a poor reduction had less exposure, less blood loss and minimal invasiveness. greater than 20˚ of varus malreduction, which resulted In cases where there is loss of reduction due to pin in the mechanical loss of fixation and poor subsequent loosening, ORIF can be performed15. ORIF with clinical outcomes. PHILOS plate for treatment of proximal humerus Care should be taken to prevent axillary nerve fractures has the advantages of accurate reduction, injury21. Gardner et al.22, through a cadaveric study, early mobilization, better fixation in osteoporotic bones reported that the axillary nerve was located 6.3 ± 0.5 and ease of reconstruction of comminuted irreducible cm below acromion. Smith et al.23 reported that the fractures. On the other hand, it has the disadvantages safe zone of the PHILOS plate for proximal screw of excessive soft tissue dissection and blood loss, risks insertion consisted of the six most proximal holes only. of injury to the neurovascular structures and increased They emphasized the necessity of gentle palpation and risk of avascular necrosis of humeral head16,17. However, digital protection of the nerve during proximal plate recent studies have shown good long term results of and screw insertion. Nonetheless, routine identification proximal humerus fractures managed by the PHILOS of the axillary nerve to avoid traction injury is not plate18,19. recommended. Six patients were suspected of having We believe that age and osteoporosis are the main an axillary nerve injury as they had difficulty elevating reasons for the failure of the pins to hold the reduction. in forward flexion and lateral elevation. One patient Other authors have described a combination of pins was diagnosed with axonotmesis. with other forms of internal fixation, such as screws, to The common complication found in locking plate overcome this problem20. Recently we have added two fixation was primary screw perforation, resulting from additional antegrade pins from the greater tuberosity incorrect surgical technique and fracture com- into the medial cortex in order to improve stability. pression24-25. Skill and comfort of the surgeon play a These pins are left outside the skin, and could increase significant factor in the decision to use one method of the rate of infection. fixation over the other. In a study conducted by Fazal Varus mal-reduction is a complication reported et al. it was seen that PHILOS plate fixation provided in both types of fixation. Agudelo et al.12 found that stable fixation with minimal implant problems and among 73 patients treated with a locking plate, 30.4 % enabled early range of motion exercises to achieve of those with a neck-shaft angle less than 120˚ developed acceptable functional results26. loss of fixation as compared to only 11 % in patients This study had a few limitations: (a) it was a with a postoperative neck-shaft angle greater than retrospective study and not randomized, so there was 120˚. This suggests that varus malreduction is a risk selection bias; and (b) the size of the study was relatively factor for loss of fixation and should be carefully small, thus a few comparisons lacked statistical power. assessed during surgery. Our study demonstrated that Finally, the follow-up period of one year may be too both locking plate and percutaneous K-wire fixation short to draw final conclusions on long-term outcome methods yielded no significant difference in and complications such as avascular necrosis of the postoperative neck-shaft angle with 126˚ and 128˚ in humeral head. the locking plate and percutaneous K-wire group, respectively (p=0.914). Both treatments resulted in CONCLUSION neck-shaft angles that were at or greater than 120˚, which proved that varus malreduction can be avoided Both locking plate and percutaneous K-wire for and is not attributed to the type of implant used. treatment of proximal humerus fractures could achieve The maintenance of neck-shaft angle and good results. However, percutaneous K-wire fixation restoration of medial support are important in had more advantages, such as shorter operative time achieving an adequate reduction. The degree of and hospital stay, lesser blood loss and lower humeral head angulation has a substantial effect on postoperative pain, but had more complications than the final clinical outcomes. Malreduction of the locking plate fixation. Vol. 36 No. 2 Comparison of Locking Plate and Percutaneous K-wrie in Treatment of Proximal Humerus Fractures 87
REFERENCES literature. J Bone Joint Surg[Am 2007;89:44-58. 1. Handoll HH, Gibson JN, Madhok R. Interventions for treating 15. Kenner JD, Parsons BO, Flatow EL, Rogers K, Williams GR, proximal humeral fractures in adults. Cochrane Database Galatz LM. Outcomes after percutaneous reduction and Syst Rev 2003;4:434-43. fixation of proximal humerus fractures. J Shoulder Elbow Surg 2. Clifford PC. Fractures of the neck of humerus: A review of the 2007;16:330-8. late results. Injury 1980;12:91-5. 16. Siegel J, Dines D. Proximal humerus malunions. Orthop Clin 3. Stableforth PG. Four part fractures of the neck of humerus. J North Am 2000;31:35-49. Bone Joint Surg Br 1984;66:104-8. 17. Wijgman AJ, Roolker W, Pall TW, Raaymakers EL, Marti RK. 4. Zyto K, Ahrengart L, Sperbert A, Torknvist H. Treatment of Open reduction and internal fixation of three- and four-part displaced proximal humeral fractures in elderly patients. J fractures of the proximal part of the humerus. J Bone Joint Surg Bone Joint Surg Br 1997;79:512-7. 1970;52:1077-89. 5. Ring D. Current concepts in plate and screw fixation of 18. Hirschmann MT, Falleger B, Amsler F, Regazzoni P, Gross T. osteoporotic proximal humerus fractures. Injury 2007;3:559-68. Clinical longer term results after internal fixation of proximal 6. Reitman RD, Kerzhner E. Reverse shoulder arthroplasty as humerus fractures with a locking compression plate (PHILOS). treatment for comminuted proximal humeral fractures in J Orthop Trauma 2011;25:286-93. elderly patients. Am J Orthop 2011;40:458-61. 19. Olerud P, Ahrangart L, Ponzer S, Saving J, Tidermark J. Internal 7. Esen E, Dogramci Y, Gultken S, Deveci MA. Factors affecting fixation versus non operative treatment of displaced 3 part results of patients with humeral proximal end fractures proximal humerus fractures in elderly patients: A randomised undergoing primary hemiarthroplasty: A retrospective study controlled trial. J Shoulder Elbow Surg 2011;20:747-55. in 42 patients. Injury 2009;40:1336-41. 20. Resch H. Percutaneous fixation of three and four parts fractures 8. Magovern B, Ramsay ML. Percutaneous fixation of proximal of the proximal humerus. J Bone Joint Surg Br 1997;79:295-300. humerus fractures. Orthop Clin North Am 2008;39:405-16. 21. Liu KY, Chen TH, Shyu JF, Wang ST, Liu JY, Chou PH. Anatomic 9. Bogner R, Hubner C, Matis N, Auffarth A, Lederer S, Resch H. study of the axillary nerve in a Chinese cadaveric population: Minimally-invasive treatment of three- and four-part fractures correlation of the course of the nerve with proximal humeral of the proximal humerus in elderly patients. J Bone Joint Surg fixation with intramedullary nail or external skeletal fixation. Br 2008;90:1602-7. Arch Orthop Trauma Surg 2011;131:669-74. 10. Krappinger D, Bizzotto N, Riedmann S, Kammerlander C, 22. Gardner MJ, Griffith MH, Lorich DG. Helical plating of the Hengg C, Kralinger FS. Predicting failure after surgical fixation proximal humerus. Injury 2005;36:1197-200. of proximal humerus fractures. Injury 2011;42:1283-8. 23. Smith J, Berry G, Laflamme Y, Blain-Pare E, Reindl R, Harvey E. 11. Lau TW, Leung F, Chan CF, Chow SP. Minimally invasive plate Percutaneous insertion of a proximal humeral locking plate: osteosynthesis in the treatment of proximal humeral fracture. an anatomic study. Injury 2007;38:206-11. Int Orthop 2007;31:657-64. 24. Neer CS. Displaced proximal humeral fractures. I. Classifica- 12. Agudelo J, Schurmann M, Stahel P, et al. Analysis of efficacy tion and evaluation. J Bone Joint Surg Am 1970;52:1077-89. and failure in proximal humerus fractures treated with locking 25. Tanner MW, Cofield RH. Prosthetic arthroplasty for fractures plates. J Orthop Trauma 2007;21:676-81 and fracture dislocations of the proximal humerus. Clin 13. Constant CR, Murley AHG. A clinical method of functional Orthop Relat Res 1983;179:116-28. assessment of the shoulder. Clin Orthop 1987;214:160-4. 26. Fazal. PHILOS plate fixation for displaced proximal humeral 14. Nho SJ, Brophy RH, Baker JU, Cornell CN, Mac Gillivray JD. fractures. J Orthop Surg 2009;17:15-8. Management of proximal humerus fractures based on current