UvA-DARE (Digital Academic Repository)

Management of early neoplasms and surgical complications of the rectum

Musters, G.D.

Publication date 2016 Document Version Final published version

Link to publication

Citation for published version (APA): Musters, G. D. (2016). Management of early neoplasms and surgical complications of the rectum.

General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl) Download date:11 Oct 2021 MANAGEMENT OF EARLY NEOPLASMS AND SURGICAL COMPLICATIONS OF THE RECTUM

GIJSBERT MUSTERS

Management of early neoplasms and surgical complications of the rectum

Gijsbert D. Musters

ISBN: 978-94-028-0045-6

Copyright © Gijsbert D. Musters, 2016. No parts of this thesis may be produced, stored or transmitted in any form by any means, without prior permission of the author.

Financial support for the printing of this thesis was kindly provided by: Rembrandt Medical B.V., Chipsof B.V., Olympus Nederland B.V, B. Braun Medical B.V., Acelity company, Nederlandse Vereniging voor Gasteroenterologie, Tergooi ziekenhuizen,Wetenschappelijk Fonds Chirurgie AMC. Management of early neoplasms and surgical complications of the rectum

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnifcus prof. dr. D.C. van den Boom

ten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op woensdag 6 april 2016, te 14:00 uur

door Gijsbert Daniël Musters

geboren te Huizen Promotiecommissie:

Promotor: Prof. dr. W.A. Bemelman Universiteit van Amsterdam

Copromotores: Dr. P.J. Tanis Universiteit van Amsterdam Prof. dr. E. Dekker Universiteit van Amsterdam

Overige leden: Prof. dr. O.R.C. Busch Universiteit van Amsterdam Prof. dr. M.A. Boermeester Universiteit van Amsterdam Prof. dr. P. Fockens Universiteit van Amsterdam Prof. dr. L.P.S. Stassen Universiteit van Maastricht Dr. E.J.R. de Graaf IJsselland ziekenhuis

Faculteit der Geneeskunde TABLE OF CONTENT

General Introduction and outline of the thesis 7

CHAPTER 1 Endoscopic mucosal resection of large rectal adenomas in the era of 19 centralization: results of a multicenter collaboration. United European Gastroenterol J 2014; 2(6): 497-504

CHAPTER 2 Perineal wound healing afer abdominoperineal resection for rectal 33 cancer; a systematic review and meta-analysis. Dis Colon Rectum 2014; 57(9): 1129-1139

CHAPTER 3 Perineal wound problems afer abdominoperineal resection for 53 rectal cancer; a two-institutional experience in the era of intensifed oncological treatment. Int J Colorectal Dis 2014; 29(9):1151-1157

CHAPTER 4 Biological mesh closure of the pelvic foor afer extralevator 67 abdominoperineal resection for rectal cancer (BIOPEX-study). BMC Surg 2014; 27: 58 CHAPTER 5 Biological mesh closure of the pelvic foor afer extralevator 79 abdominoperineal resection for rectal cancer: a single blinded, multicentre randomised controlled trial (the BIOPEX-study). Submitted CHAPTER 6 Local application of gentamicin in the prophylaxis of perineal wound 109 infection afer abdominoperineal resection, a systematic review. World J Surg 2015; 39(11): 2786-2794

CHAPTER 7 Is there a place for a biological mesh in perineal hernia repair? 125 Submitted

CHAPTER 8 Surgery for complex perineal fstula following rectal cancer treatment 139 using biological mesh combined with gluteal perforator fap. Tech Coloproctol 2014; 18(10): 955-959

CHAPTER 9 Early reconstruction of the leaking ileal pouch-anal anastomosis: a novel 149 solution to an old problem. Colorectal Dis 2015; 17(5): 426-432

CHAPTER 10 Major morbidity afer ileostomy closure: results from an institutional 161 change in practice and awareness. Int J Colorectal Dis 2016

CHAPTER 11 Intersphincteric completion proctectomy with omentoplasty for chronic 173 presacral sinus afer low anterior resection for rectal cancer. Colorectal Dis 2015 Appendices Summary and future perspectives 187 Samenvatting en toekomstige perspectieven 195 Ph.D. portfolio 203 List of publications 207 Dankwoord 213 About the author 219

General introduction and outline of the thesis

General introduction and outline of the thesis

8 General introduction and outline of the thesis

GENERAL INTRODUCTION

Colorectal cancer (CRC) is one of the most common cancers in the world and is expected to rise from 1.4 million annually diagnosed patients worldwide in 2012 to 2.4 million patients by 2035.1 In the Netherlands, currently 13.000 new patients are diagnosed with CRC each year, of which approximately one third have rectal cancer.2

Te prognosis of patients with rectal cancer has improved signifcantly during the last two decades in the Netherlands. Te cornerstone of locoregional treatment is still surgery, and improved quality of resection as well as locally advanced rectal cancer by specialised surgical teams contributed substantially to these improvements. Tis was also facilitated by optimized pre-operative staging using MRI and the introduction of down-sizing neoadjuvant treatment. Furthermore, improvement has been achieved in the treatment of distant metastases, both by more efective systemic treatment and by increased application of local treatment modalities of liver and/or lung metastases. Further improvement of the prognosis can be expected from the recently started screening programme in the Netherlands, enabling treatment of (advanced) adenomas and early detection of rectal cancer.

Rectal cancer surgery is complex and requires specifc expertise. But despite specialisation and centralisation, there is still a relatively high risk of postoperative complications. In contrast to literature on treatment of rectal cancer, prevention and treatment of short and long-term surgical complications has less extensively been explored. At the time prognosis improves, quality of life and management of treatment related morbidity becomes more and more important. Treatment of rectal cancer requires a multidisciplinary evidenced based approach, in which the treatment related morbidity has to be weighed against the expected health beneft.

Treatment of rectal adenomas With the start of screening programs, adenomas are more frequently encountered.3 Early detection and removal of colorectal adenomas reduces the incidence of colorectal cancer and its associated morbidity.4 Removal of these rectal adenomas face their own therapeutic challenges. Resection of adenomas can be performed endoscopically or with microsurgery. For small adenomas, endoscopic resection can be performed by simple loop polypectomy, piecemeal endoscopic mucosal resection (EMR), or single-piece endoscopic submucosal dissection (ESD). For larger non-pedunculated adenomas, EMR is most ofen being performed. With EMR, the large rectal adenoma is submucosal lifed and snare piecemeal resection is performed. For EMR, light sedation is most ofen sufcient and no specialized equipment or overnight hospital stay is required. With transanal endoscopic microsurgery (TEM), the adenoma can be dissected submucosally or by full-thickness en bloc resection. TEM requires general anaesthesia, hospital admission and specialized instrumentation. TEM can be replaced by transanal minimally invasive surgery (TAMIS) platforms using standard laparoscopic equipment.5 Te results of TEM have been extensively reported in literature whereas data on endoscopic resection is limited for large rectal adenomas.6

9 General introduction and outline of the thesis

Surgical treatment principles in rectal cancer Afer low anterior resection (LAR) following total mesorectal excision (TME) and beyond TME principles, continuity can be restored by a colo-anal anastomosis in diferent confgurations and using both stapling and hand-sewn techniques. Te decision to restore continuity depends on the sphincter function of the patient preoperatively, the distance of the tumour to the anal verge, the resection margins that can be achieved in relation to the sphincter complex and levator muscle, age and comorbidity, lower border of the radiation feld, and patient preference. When an anastomosis is not made for any reason, there are two alternatives. If the sphincter complex and pelvic foor can be preserved, the rectal stump is closed and an end colostomy is created, also referred to as a low Hartmann’s procedure. When the sphincter complex and/or levator muscle cannot be spared anymore, an abdominoperineal resection (APR) is performed.

Low anterior resection with colo-anal anastomosis One of the most feared complications afer reconstructive rectal surgery is an anastomotic leakage, occurring between 6 to 20% of the patients afer LAR.7 An anastomotic leakage is a defect of the intestinal wall integrity at the anastomotic site, leading to a communication between the intra- and extraluminal compartments.8 Te mainstay in the treatment of an anastomotic leakage in patients in which taking down the anastomosis is not indicated, is the creation of a defunctioning stoma if not done so during primary surgery. Te cavity that occurs in an anastomotic leakage will fll with debris and pus, even when the anastomosis is defunctioned. Te resulting presacral abscess is conventionally treated with transanal or percutaneous drainage. However, treatment fails between 17 and 25% of the patients.9 As a result, a non-healing chronic presacral sinus lasting longer than one year may develop. It is generally thought to be a rare clinical problem, but recent population based data in the Netherlands suggest an 8 percent incidence (unpublished data). A chronic presacral sinus is of signifcant clinical impact because of the severe associated morbidity.10 Te increased use of neoadjuvant (chemo)radiotherapy is likely to afect the healing of the anastomotic leakage and may contribute to an increase in chronic presacral sinuses.11

Patients with a chronic presacral sinus are ofen subjected to many diferent operations and face signifcant morbidity without signifcant clinical improvement.12 Tis is probably the result of the limited available literature about the treatment options and the outcome of a chronic presacral sinus afer low anterior resection for rectal cancer. Minimally invasive treatment approaches such as fbrin glue application or stapler marsupialisation have been reported with short term success. However, long term follow-up is required because the presacral abscess may reoccur and progress with fstula formation in the years thereafer. Terefore, an intensive surgical treatment strategy has been developed at the Academic Medical Centre to treat this difcult problem. Tis entails resection of the old anastomosis with complete debridement of the sinus, followed by either a colo-anal anastomotic reconstruction or completion proctectomy with end colostomy and complete flling of the presacral cavity by an omentoplasty.12

A colo-anal anastomosis is ofen diverted using a double loop ileostomy, or some surgeons prefer a diverting colostomy for this purpose. Currently, up to 20% of the temporary stomas afer LAR are never being reversed.11 A defunctioning stoma is primarily performed to protect the anastomosis.

10 General introduction and outline of the thesis

It probably does not reduce the incidence of anastomotic leakage, but rather limits the consequences.13 Although a defunctioning stoma may reduce leakage induced pelvic sepsis, stoma related morbidity should be taken into account as well. Besides the complications of having a stoma (i.e. dehydration, stoma prolapse, parastomal herniation), reversal of a defunctioning stoma is associated with an additional morbidity risk of up to 17%, and even a few percent risk of mortality.14 Ileostomy reveral is ofen considered a low complex procedure. However, little is known about the efect of surgical expertise on ileostomy reversal and its associated morbidity.

Abdominoperineal resection When the resection margins are too small to spare the sphincter complex or levator muscle, an APR will be performed. With the conventional abdominoperineal resection (cAPR), the TME plane is followed all the way down to the pelvic foor, resulting in a typical coning of the specimen. Te mesorectum becomes smaller at the level of the levator muscles and disappears in the sphincter area. Tumours at this level, even early T2 stage, are at increased risk of incomplete resection by cAPR. A pooled analysis of fve European randomised controlled trials revealed that cAPR in comparison to LAR is an independent risk factor for local disease recurrence (19.7% vs 11.4%) and overall survival (59% vs 70%).15

Due to the high local recurrences afer cAPR, wider excisions have been advocated. Using an extralevator abdominoperineal resection (eAPR) the TME plane is not followed all the way down to the pelvic foor, but an en bloc resection of the distal rectum, sphincter complex and levator muscles is performed. Te eAPR resulted in a cylindrical specimen with a reduction in positive resection margins and intraoperative perforations.16 A downside of the neoadjuvant (chemo)radiotherapy and wider surgical resection is the increased risk of perineal wound problems such as perineal wound infection and incisional perineal herniation. A perineal wound infection is described in up to 57% of the patients afer eAPR.17 Perineal wound infections are conventionally treated with draining of the abscesses and rinsing of the perineal wound, which may even take up to one year before the perineal wound is healed.18 In an attempt to overcome this problem, biological mesh assisted perineal wound closure, local application of gentamicin and musculocutaneous faps are being used, but not supported by evidence of a sufcient level.19-21 A less frequent complication afer APR is an incisional perineal hernia. A perineal hernia can cause discomfort and pain, urinary dysfunction, and in rare occasions bowel obstruction. Treatment of perineal hernia is most ofen with supportive undergarments. When surgical intervention is indicated, mesh repair is ofen being performed. With the success achieved by the use of a biological mesh in (contaminated) abdominal hernia repair, biological meshes have also been used for perineal hernia correction, but there is almost no literature on its efectiveness.

Ileal pouch-anal anastomosis In severe cases of ulcerative colitis (UC) or selected patients with Crohn’s disease refractory to medication, or in patients with familial adenomatous polyposis (FAP), a restorative proctocolectomy with an ileal pouch-anal anastomosis (IPAA) is the treatment of choice. As with low anterior resection, the Achilles heel of this operation is an anastomotic leakage. When an anastomotic leakage develops, a defunctioning ileostomy is constructed, if not done so primarily, and the presacral

11 General introduction and outline of the thesis

abscess is transanally or percutaneously drained. A relative new technique to drain the presacral sinus afer IPAA is by applying the Endo-sponge® technique as described by Weidenhagen et al..22 In this technique a sponge is placed with sigmoidoscopy through the anastomotic defect in the presacral space and connected to a low vacuum suction bottle. By changing the Endo-sponge® every 3-4 days and reducing the size of the sponge, the abscess cavity gradually collapses. However, this may take many weeks and many Endo-sponge® changes before the abscess cavity is small enough to close on itself. A modifcation to this technique that is currently being used is to clean the presacral cavity with two or three Endo-sponge® changes and then surgically close the anastomotic defect. It is expected that this modifcation has a higher success rate and reduces the treatment duration.

12 General introduction and outline of the thesis

AIM AND OUTLINE OF THE THESIS

Tis thesis starts with a chapter on endoscopic treatment of the precursor lesions of rectal cancer. In the subsequent chapters, several complications afer rectal surgery are discussed, focussing on predictive factors, prevention, and evaluation of treatment modalities.

Rectal adenomas are considered to precede rectal cancer, and complete excision of these precursor lesions may therefore prevent cancer development. Removal of rectal adenomas can be challenging, especially the large lesions. One of the techniques to excise large rectal adenomas is endoscopic mucosal resection (EMR). In chapter 1 we assessed the safety and efectiveness of EMR for large rectal adenomas.

Since the introduction of neoadjuvant (chemo)radiotherapy and the extralevator abdominoperineal resection (eAPR), oncological outcome afer distal rectal cancer treatment has been improved, but at the expense of more perineal wound infections and incisional perineal herniation. Te primary aim of chapter 2 was to determine the impact of (chemo)radiotherapy and an extralevator approach on perineal wound healing afer abdominoperineal resection for rectal cancer by performing a systematic review of the literature with meta-analysis. Besides the eAPR and neoadjuvant (chemo) radiotherapy, other clinical changes have been made over time. Little is known about the impact of these changes on perineal wound complications. Te aim of chapter 3 was to analyse perineal wound healing in relation to changes in treatment approaches for patients undergoing APR for distal rectal cancer in a retrospective cohort study of patients who were treated at an academic and a regional hospital.

Due to the increase in perineal wound healing problems over time, the standard primary perineal wound closure technique afer eAPR is being replaced by other closure techniques. One of those new techniques is the use of biological mesh assisted perineal wound closure. Te biological meshes are increasingly used for perineal wound closure, thereby increasing costs, despite sufcient evidence. In chapter 4 we describe a study protocol of a multicentre randomised controlled trial in which we wanted to determine the efectiveness of pelvic foor reconstruction using a biological mesh afer eAPR with neoadjuvant (chemo)radiotherapy compared to primary perineal wound closure. Te outcome of this multicentre randomised controlled trail is being described in chapter 5. Other prophylactic measures to prevent perineal wound infections afer APR, might be prophylactic antibiotics. However, for the antibiotics to be efective, a high antibiotic concentration at the designated site is warranted. Systemic or oral prophylactic antibiotics seem not to efect perineal wound infections. Local application of antibiotics can obtain higher local concentrations without systemic adverse reactions. In chapter 6 we describe in a systematic review the efect of local application of gentamicin on perineal wound healing afer APR.

It is expected that increased use of eAPR is associated with an increased incidence of perineal hernia. Surgical treatment of an incisional perineal hernia is performed with several types of meshes. One of these meshes currently being used is a biological mesh. However, current literature on perineal hernia repair with a biological mesh mainly consists of case reports.Terefore, the aim of chapter 7 is to describe the postoperative outcome and follow-up of a consecutive cohort of perineal hernia repair using a biological mesh.

13 General introduction and outline of the thesis

For patients in which extensive pelvic surgery is required in the treatment of pelvic malignancy, reconstruction of the pelvic foor can be challenging due to the limited options available. In chapter 8 we describe three patients with complex perineal fstula afer extensive pelvic surgery and radiotherapy whom underwent surgical treatment by combining a biological mesh for pelvic foor reconstruction and a unilateral superior gluteal artery perforator (SGAP) fap for flling of the perineal defect.

Anastomotic leakage is one of the most feared complications afer low pelvic anastomosis. To increase the success rate of the treatment of anastomotic leakage afer restorative rectal resection or proctocolectomy, Endo-sponge® treatment is currently being used. A modifcation to the conventional Endo-sponge® treatment is early surgical closure of the anastomotic defect afer a short course of Endo-sponge® therapy. In chapter 9 we determined the efectiveness and direct medical costs of this modifcation, compared to conventional treatment in patients with anastomotic leakage afer ileal pouch-anal anastomosis for ulcerative colitis or familial adenomatous polyposis.

Currently, diverting ileostomies are ofen being performed during primary rectal surgery to protect the anastomosis. A previous study from the Academic Medical Centre revealed substantial morbidity afer ileostomy reversal, which is one of the issues in the ongoing debate about the value of routine diverting ileostomy in low pelvic anastomoses. Tis study resulted in an institutional change in practise. In chapter 10 we aimed to determine if these changes in clinical practice resulted in reduced complication rates.

If a leaking colo-anal anastomosis does not heal by conventional treatment using fecal diversion and drainage, a chronic presacral sinus may develop with signifcant morbidity and impact on quality of life. Little is known about natural course of this long-term surgical complication, possible treatment options and long-term outcome. In chapter 11 we aimed to determine the clinical outcome of salvage surgery without restoring continuity for symptomatic chronic presacral sinus afer low anterior resection for rectal cancer.

14 General introduction and outline of the thesis

REFERENCE LIST

1. Stooker T, van Acht JW, van Barneveld EM, et al. Costs in the last year of life in Te Netherlands. Inquiry 2001; 38(1): 73-80.

2. van Hout BA, Simoons ML. Cost-efectiveness of HMG coenzyme reductase inhibitors; whom to treat? European heart journal 2001; 22(9): 751-61.

3. Robertson DJ, Kaminski MF, Bretthauer M. Efectiveness, training and quality assurance of colonoscopy screening for colorectal cancer. Gut 2015; 64(6): 982-90.

4. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. Te National Polyp Study Workgroup. Te New England journal of medicine 1993; 329(27): 1977-81.

5. Nieuwenhuis DH, Draaisma WA, Verberne GH, van Overbeeke AJ, Consten EC. Transanal endoscopic operation for rectal lesions using two-dimensional visualization and standard endoscopic instruments: a prospective cohort study and comparison with the literature. Surgical endoscopy 2009; 23(1): 80-6.

6. Barendse RM, van den Broek FJ, Dekker E, et al. Systematic review of endoscopic mucosal resection versus transanal endoscopic microsurgery for large rectal adenomas. Endoscopy 2011; 43(11): 941-9.

7. Wu SW, Ma CC, Yang Y. Role of protective stoma in low anterior resection for rectal cancer: a meta- analysis. World journal of gastroenterology : WJG 2014; 20(47): 18031-7.

8. Rahbari NN, Weitz J, Hohenberger W, et al. Defnition and grading of anastomotic leakage following anterior resection of the rectum: a proposal by the International Study Group of Rectal Cancer. Surgery 2010; 147(3): 339-51.

9. Kirat HT, Remzi FH, Shen B, Kiran RP. Pelvic abscess associated with anastomotic leak in patients with ileal pouch-anal anastomosis (IPAA): transanastomotic or CT-guided drainage? International journal of colorectal disease 2011; 26(11): 1469-74.

10. van Koperen PJ, van der Zaag ES, Omloo JM, Slors JF, Bemelman WA. Te persisting presacral sinus afer anastomotic leakage following anterior resection or restorative proctocolectomy. Colorectal disease 011; 13(1): 26-9.

11. den Dulk M, Smit M, Peeters KC, et al. A multivariate analysis of limiting factors for stoma reversal in patients with rectal cancer entered into the total mesorectal excision (TME) trial: a retrospective study. Te Lancet Oncology 2007; 8(4): 297-303.

12. Sloothaak DA, Buskens CJ, Bemelman WA, Tanis PJ. Treatment of chronic presacral sinus afer low anterior resection. Colorectal Dis 2013; 15(6): 727-32.

13. Huser N, Michalski CW, Erkan M, et al. Systematic review and meta-analysis of the role of defunctioning stoma in low rectal cancer surgery. Annals of surgery 2008; 248(1): 52-60.

14. Mennigen R, Sewald W, Senninger N, Rijcken E. Morbidity of loop ileostomy closure afer restorative proctocolectomy for ulcerative colitis and familial adenomatous polyposis: a systematic review. Journal of gastrointestinal surgery 2014; 18(12): 2192-200.

15. den Dulk M, Putter H, Collette L, et al. Te abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of fve European randomised clinical trials on rectal cancer. European journal of cancer 2009; 45(7): 1175-83.

16. West NP, Finan PJ, Anderin C, Lindholm J, Holm T, Quirke P. Evidence of the oncologic superiority of cylindrical abdominoperineal excision for low rectal cancer. Journal of clinical oncology 2008; 26(21): 3517-22.

15 General introduction and outline of the thesis

17. de Bruin AF, Gosselink MP, Wijfels NA, Coene PP, van der Harst E. Local gentamicin reduces perineal wound infection afer radiotherapy and abdominoperineal resection. Techniques in coloproctology 2008; 12(4): 303-7.

18. Collin A, Gustafsson UM, Smedh K, Pahlman L, Graf W, Folkesson J. Efect of local gentamicin-collagen on perineal wound complications and cancer recurrence afer abdominoperineal resection: a multicentre randomized controlled trial. Colorectal Dis 2013; 15(3): 341-6.

19. Gruessner U, Clemens M, Pahlplatz PV, et al. Improvement of perineal wound healing by local administration of gentamicin-impregnated collagen feeces afer abdominoperineal excision of rectal cancer. American journal of surgery 2001; 182(5): 502-9.

20. Han JG, Wang ZJ, Wei GH, Gao ZG, Yang Y, Zhao BC. Randomized clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. American journal of surgery 2012; 204(3): 274-82.

21. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus fap reconstruction of the pelvic foor for rectal cancer. Te British journal of surgery 2007; 94(2): 232-8.

22. Weidenhagen R, Gruetzner KU, Wiecken T, Spelsberg F, Jauch KW. Endoscopic vacuum-assisted closure of anastomotic leakage following anterior resection of the rectum: a new method. Surgical endoscopy 2008; 22(8): 1818-25.

16 General introduction and outline of the thesis

17

CHAPTER 1

Endoscopic mucosal resection of large rectal adenomas in the era of centralization: results of a multicenter collaboration

R.M. Barendse, G.D. Musters, P. Fockens, W.A. Bemelman, E.J. de Graaf, F.J. van den Broek, K. van der Linde, M.P. Schwartz, M.H. Houben, A.W. van Milligen de Wit, B.J. Witteman, R. Winograd, and E. Dekker on behalf of the TREND study group

United European Gastroenterol J 2014; 2(6): 497-504 Chapter 1

ABSTRACT

Aim: Endoscopic mucosal resection (EMR) of large rectal adenomas is largely being centralized. We assessed the safety and efectiveness of EMR in the rectum in a collaboration of 15 Dutch hospitals.

Methods: Prospective, observational study of patients with rectal adenomas >3 cm, resected by piecemeal EMR. Endoscopic treatment of adenoma remnants at 3 months was considered part of the intervention strategy. Outcomes included recurrence afer 6, 12 and 24 months and morbidity.

Results: Sixty-four patients (50% male, age 69 ± 11, 96% ASA 1/2) presented with 65 adenomas (diameter 46 ± 17 mm, distance ab ano 4.5 cm (IQR 1–8), 6% recurrent lesion). Sixty-two procedures (97%) were technically successful. Histopathology revealed invasive carcinoma in three patients (5%), who were excluded from efectiveness analyses. At 3 months’ follow-up, 10 patients showed adenoma remnants. Recurrence was diagnosed in 16 patients during follow-up (recurrence rate 25%). Fifeen of 64 patients (23%) experienced 17 postprocedural complications.

Conclusion: In a multicentre collaboration, EMR was feasible in 97% of patients. Recurrence and post-procedural morbidity rates were 25% and 23%. Our results demonstrate the outcomes of EMR in the absence of tertiary referral centres.

20 Endoscopic mucosal resection of large rectal adenomas

INTRODUCTION

Early endoscopic detection and removal of colorectal adenomas reduces the incidence of colorectal cancer.1 It has recently been shown that colonoscopic polypectomy also significantly reduces colorectal cancer associated mortality as compared to the general population.2 Colorectal cancer screening programs, which are currently broadly implemented, will facilitate the detection of cancers in a curable stage, but will also diagnose more adenomas. Tis increased adenoma detection 1 demands expansion of treatment capacities and optimization of treatment strategies to facilitate cost-efective care.

Endoscopic treatment of colonic adenomas may prevent surgical interventions such as segmental resection or hemicolectomy and their associated complications. However, large colonic adenomas demand treatments beyond simple loop polypectomy or single-piece endoscopic mucosal resection. Te most common endoscopic resection technique for large adenomas is piecemeal endoscopic mucosal resection (EMR). In a systematic review of 20 studies on EMR of large (>2 cm) colorectal adenomas, we have previously described an early recurrence rate (afer a single attempt) of 11.2%; in fact, these lesions are residual adenomas rather than recurrences. When permitting re-treatment of these adenoma remnants within 6 months, the late recurrence rate was 1.5% and complication rate 3.8%.3 Te majority of the studies included in this review consisted of retrospective case series from specialized centres and might not reflect daily practice.

Piecemeal adenoma resection is associated with high rates of residual adenoma at follow-up.4 Two recently published prospective studies on piecemeal EMR of colorectal adenomas, published by endoscopists with extensive experience and high case-volumes, report residual adenoma rates of 20–38%. However, late recurrence was eventually diagnosed in less than 5% of patients.5,6 In many countries, EMR is considered a complex endoscopic procedure and is mostly being centralized to guard its efectiveness and safety.7 In the Netherlands, tertiary referral centres are developing but most patients are still treated by general endoscopists with low case-volumes of these large lesions.

With regards to large rectal adenomas, it is still debated whether endoscopic or surgical resection is the preferable treatment method. An ongoing prospective randomised trial by a multicentre collaboration of dedicated endoscopists and surgeons (TREND-study) compares the cost- efectiveness of EMR and transanal endoscopic microsurgery (TEM) for the resection of large (>3 cm) rectal adenomas.8 Whereas the results of TEM have been extensively reported in previous literature, data on EMR in the rectum alone seem absent, with the exception of one single-centre prospective study of 62 patients, reporting a local recurrence rate of 8% within a median follow-up of 14 months and a post procedural bleeding rate of 8%.9 Additional prospective data on EMR in the rectum are important to assess the outcomes of EMR in the absence of expert centres with high case-volumes, especially while the results of the TREND study are awaited until the fall of 2015. Terefore, we aimed to assess the outcomes, including safety and efectiveness, of EMR of large adenomas located in the rectum only, performed in a dedicated multicentre setting.

21 Chapter 1

METHODS

Patients Patients who were eligible for inclusion in the TREND study, but refused randomisation because of a distinct preference for endoscopic treatment between January 2009 and August 2011 were included. Patients were eligible if they presented with non-pedunculated rectal adenomas larger than 3 cm, of which the distal border was situated within 15 cm from the dentate line. Adenoma size was estimated endoscopically using an open biopsy forceps or resection snare. Patients were excluded if there was any suspicion of submucosal invasion during diagnostic workup, if patients were medically unfit for surgery or if either EMR or TEM was considered non-feasible by the gastroenterologist or surgeon. Inclusion and exclusion criteria for the TREND-study are extensively described elsewhere.8 If patients fulfilled criteria but refused randomization in the TREND-study and requested endoscopic treatment, informed consent was asked for prospective follow-up in the current study. Te Institutional Review Boards (IRB) of all participating hospitals reviewed and approved the TREND-study protocol as well as the current study protocol.

Participating centres and endoscopists Patients could be enrolled in one of 18 Dutch and Belgian hospitals also involved in the TREND- study. All participating hospitals were referral centres for interventional endoscopy in their geographical region. To strive for a centralization efect, we coopted a single gastroenterologist per hospital to perform all study procedures. As the extent of EMR experience varied among participating endoscopists, an expert panel reviewed video-recorded cases for quality assurance purposes prior to study involvement. A consensus meeting was held prior to the start of the study to agree on EMR equipment and technique.

Interventions Split-dose bowel preparation was used, but not otherwise standardized. Anticoagulant therapy was typically discontinued three days prior to the procedure; single agent antiplatelet therapy such as acetylsalicylic acid was continued. EMR was performed as described by Karita and Hurlstone.9,10 At the discretion of the endoscopist, conscious or deep sedation could be administered with midazolam, fentanyl or propofol. Diagnostic gastroscopes (such as GIF-H180 or GIF- HQ180, Olympus, Tokyo, Japan) were preferably used for the procedure because of their flexibility. Afer proper endoscopic visualization of the adenoma, it was lifed submucosally using a solution of saline 0.9%, 1mL methylene blue, and 1:10,000 units adrenaline. Adenoma margins were not routinely marked with mucosal cautery. Te lesions were removed by sequential inject and resect EMR technique.

Te adenoma was resected in a piecemeal fashion by alternated use of standard large (30 mm, Boston Scientific, Natick, Massachusetts, USA), barbed (20 mm, Olympus, Tokyo, Japan) and standard small (13 mm, Boston Scientific, Natick, Massachusetts, USA) snares and electro-coagulation (endocut, ERBE Elektromedizin GmbH, Vienna, Austria). Snare resection was continued until the lesion was macroscopically entirely removed and the blue colored submucosa was visualized. Tereafer, the borders of the mucosal defect and potential remnants within the resection crater were treated with argon plasma coagulation (APC) to enhance adenoma clearance. Visible vessels in the EMR

22 Endoscopic mucosal resection of large rectal adenomas

defect that were not bleeding were not treated prophylactically. Polypectomy sites were not closed with clips. Tattoos were not routinely placed at the polypectomy site in the rectum. Patients were discharged afer the procedure in accordance with current practice.

Histopathological evaluation All pieces that were resected by EMR were processed in a standard fashion by the pathologist. As evaluation of the resection margins is impossible in piecemeal resection specimens, only the basal 1 margins, lesion type, and degree of dysplasia were assessed. When submucosally invasive carcinoma was diagnosed in the resection specimen, the necessity of additional treatment was discussed in a multidisciplinary team, and the patient was excluded from further analyses in the current study.

Follow-up and outcome parameters All patients underwent repeat endoscopy at 3 months. Endoscopic treatment of remnants at 3 months was considered part of the initial intervention strategy. Surveillance endoscopies were scheduled at 6, 12, and 24 months afer EMR. In all endoscopies following EMR, biopsies of presumed adenomatous tissue were taken to confirm the presence of neoplasia by histology. All remnants and recurrences were treated endoscopically if feasible. If the scar appeared normal without the presence of adenomatous tissue, random biopsies were taken from the centre of the scar and its edges. Te primary outcome was recurrence of neoplasia, defined as the presence of histologically proven neoplastic tissue in either visible recurrent lesions or in random biopsies at 6, 12, or 24 months’ follow-up.

Secondary outcomes were post-procedural complication rates, including complications associated with additional endoscopic treatment of adenoma remnants at 3 months’ follow-up. All patients who had consented to follow-up by telephone, in addition to solely endoscopic follow-up, were contacted by a research nurse on the first post-procedural day as well as 2 weeks and 3, 6, 12, and 24 months afer the procedure for inquiry into possible adverse events other than those already reported in patients’ medical correspondence during follow-up. Te correlation of each event to the intervention strategy was classified according to likelihood (unlikely, possibly, probably or definitely related, respectively).

Statistical analysis Results for continuous variables were summarized using mean (standard deviation (SD)) or median (interquartile range (IQR)) for skewed data. Frequencies (%) were used to summarize categorical variables. By univariable logistic regression analyses, possible predictors of recurrence and complications were identified. All predictors identified by a P-value of <0.1 were candidate variables for inclusion in a multivariable model. Multiple regression with backward stepwise Wald variable selection, in which a P-value of <0.05 was considered statistically significant was then used to identify independent predictors of the outcomes. Calculations were made in SPSS statistics package for Windows (International Business Machines Corp., Armonk. NY, USA), version 19.0.

23 Chapter 1

RESULTS

Patients and lesions A total of 64 patients who were eligible for the TREND study between January 2009 and August 2011 denied randomization because of a distinct preference for endoscopic treatment, but all agreed on being followed in the current study protocol. Tey underwent treatment and surveillance in one of 15 Dutch hospitals (three academic, 12 general). Participating endoscopists had a mean annual case volume of 2.6 ± 2.1 (range 0.3–7.7) rectal adenomas ≥3 cm, assuming that all cases were registered in the TREND study or current study. Eleven patients (17%) were treated in academic hospitals; the remaining 53 were treated in general hospitals. In these 64 patients, 65 large rectal adenomas (mean diameter 46 ± 17 mm) were diagnosed; one patient had two adenomas. Patient and adenoma characteristics are summarised in Table 1. Te mucosal aspect (Kudo pit pattern) of the resected adenoma was only described in 12% of lesions.

Table 1 Patient and lesion characteristics

Patient characteristics (n =64) Male sex (%) 32(50) Mean age ± SD 69 ± 11 ASA classifcation (missing 2) Grade 1 (n,%) 26 (41) Grade 2 (n,%) 31 (48) Grade 3 (n,%) 5 (8) Lesion characteristics (n = 65) Mean maximum diameter (mm ± S D ) 46 ±17 Median distance ab ano (cm, IQR) 4.5 (1–8) Mean rectal circumference covered (%, ± SD) 39 ± 13 Paris classifcation Grade Is (n,%) 37 (57) Grade IIa (n,%) 15 (23) Grade Is + IIa (n,%) 12 (18) Unknown (n,%) 1 (2) Recurrent or remnant lesion (n, %) 4 (6) Prior treatment Endoscopic mucosal resection (n,%) 4 (6) Transanal endoscopic microsurgery (n,%) 0 Conventional transanal excision (n,%) 0 Duration Since prior treatment (months, IQR) 3.5 (2.4–161.1)

EMR procedures Te majority of EMRs were performed under conscious sedation, mostly consisting of midazolam and/or fentanyl. Two patients were treated under deep sedation administered with propofol and three patients were not sedated. Te patient with two adenomas had both lesions treated in a single session.

24 Endoscopic mucosal resection of large rectal adenomas

EMR was technically successful in 62 of 64 patients (97%), although a perforation demanded a staged procedure in one patient. Te two remaining patients were referred for TEM because of procedure intolerance and insufcient lifing of the submucosal plane respectively. Tese patients were excluded from follow-up analysis as EMR had failed. APC was administered afer snare resection in only 55 patients (89%) even though the protocol prescribed it to be applied in all patients. Afer treatment, 54 resections (87%) were considered endoscopically complete. Following EMR, patients were discharged from the hospital afer a median of 0 days (IQR 0–0.25, range 0–16). One patient 1 was hospitalised for 16 days. She required peri-procedural heparin because of a mechanic aortic valve implant. Te procedure was complicated by two haemorrhagic events, which required re- endoscopy without transfusion.

Intra-procedural complications Other than the abovementioned perforation, no other intra-procedural complications occurred (complication rate 2%). Te perforation was closed using endoclips and EMR could be completed uneventfully four weeks later.

Histopathology In both patients who underwent TEM afer EMR failure, the resection specimen contained villous adenoma with high-grade dysplasia. Histopathological evaluation of the 63 adenomas (in 62 patients) resected by EMR revealed adenoma in 60 cases (95%); dysplasia was low-grade in 34 cases, high-grade in 25 patients, and was not specified in one patient. In three patients (5%), the lesion appeared to be a submucosally invasive adenocarcinoma. Two patients underwent subsequent total mesorectal excision (TME) afer neoadjuvant therapy. Both TME specimens showed no tumour remnant. Te third patient chose not to undergo additional therapy because of her old age and significant cardiovascular comorbidity. Afer 24 months’ follow-up, all three patients were alive and did not show signs of local recurrence or distant metastasis. Tese patients were excluded from further analysis in the current study.

Follow-up Total median follow-up of the 59 patients who had undergone EMR of a histopathologically confirmed rectal adenoma was 24 months (IQR 18–25). Fify-three patients (90%) underwent a per protocol repeat sigmoidoscopy afer 3 months. Remnant adenomatous tissue was found in 10/53 patients (19%). Two remnants were diagnosed by random biopsies of normally appearing scars. One patient was referred for TEM because biopsies of the remnant lesion had shown intramucosal carcinoma with possible submucosal invasion. Eventually, the TEM specimen only contained high-grade dysplasia. EMR was considered to have failed in this patient, who was excluded from further efectiveness analyses. Recurrence was diagnosed in five patients at 6 months, nine patients at 12 months, and two patients at 24 months’ follow-up (overall recurrence rate 25%). Again, two recurrences were diagnosed by random biopsies. Table 2 summarizes the follow-up of all patients with remnants and recurrences. Univariable regression analyses did not reveal any associations between recurrence and tumour size, distance from the anus, prior attempt of resection, use of APC, macroscopically complete character of the resection as judged by the endoscopist afer the procedure, or the presence of remnant adenomatous tissue at 3 months’ follow-up (Table 3).

25 Chapter 1

Table 3 Univariable analysis of possible predictors of recurrence, with odds ratios (OR), 95% confdence intervals (95% CI) and p-values

OR 95% CI p-value

Tumour size (per cm) 1.35 0.83–2.98 0.230 Distance of distal adenoma margin to anus (per cm) 0.99 0.88–1.14 0.982 Prior attempt of adenoma resection 3.23 0.41–25.26 0.264 APC used during EMR procedure 2.21 0.24–20.03 0.481 Resection judged macroscopically incomplete by endoscopist 3.64 0.78–16.93 0.100 Remnant adenoma at 3 months 1.25 0.27–5.68 0.230

Post-procedural complications Fifeen of 64 patients in whom EMR was attempted (23%) had at least one post-procedural complication that was probably or definitely related to the EMR procedure. Te total number of complications was 17; 14 patients encountered one event and one patient encountered three events. Te most common complication was haemorrhage (n=13). Seven patients in whom haemorrhage occurred were on anticoagulant or antiplatelet therapy. Less frequent complications included post- procedural abdominal pain (n= 2), rectal stenosis (n=1), and atrial fibrillation that occurred during bowel preparation (n=1). Eight patients underwent endoscopic inspection or treatment for post- procedural haemorrhage. One case of haemorrhage was surgically sutured afer three attempts of endoscopic therapy using adrenaline, gold probe coagulation and APC. Te patient with rectal stenosis required surgical stenoplasty.

Multivariable regression analysis identified ASA classification as the only independent predictor of post-procedural complications. No significant associations were found between complications and tumour size, distance from the anus, or prior attempt of resection. Te use of anticoagulant medication was significantly associated with complications in univariable analyses, but could not be identified as an independent predictor of complications in multivariable analyses (Table 4).

Table 4 Multivariable analysis of possible predictors of complications, with odds ratios (OR), 95% confdence intervals (95% CI)

Univariable analysis Multivariable analysis OR 95% CI p-value OR 95% CI p-value Tumor size (per cm) 1.19 0.74–1.93 0.469 - - - Distance of distal adenoma margin 0.93 0.81–1.08 0.377 - - - to anus (per cm) ASA Classifcation I - - 0.019 - - 0.019 II 6.00 1.15–31.23 - 6.00 1.15–31.23 - III 36.00 2.46–527.06 - 36.00 2.46– - 527.06 Use of anticoagulant medication 7.24 1.91–27.38 0.004 2.50 0.49–12.89 0.273 Prior attempt of adenoma resection 1.08 0.10–11.26 0.951 - - -

ASA: American Society of Anaesthesiologists.

26 Endoscopic mucosal resection of large rectal adenomas

DISCUSSION

Tis multicentre prospective study provides a faithful reflection of current daily practice in the Netherlands, although the potential bias of including only patients who refused participation in a randomised trial should be considered. EMR of large rectal adenomas was technically successful in 97% of patients. Subsequent endoscopic follow-up and treatment of adenoma remnants was feasible in 98% of patients. Te recurrence and complication rates in our study compare unfavourably with 1 prior published data on large colonic EMRs. In a recent, prospective study, colleagues Moss and Bourke have demonstrated the feasibility of piecemeal EMR in colorectal adenomas, reporting very high case-volumes and demonstrating elaborate expertise in a multicentre collaboration. Teir results (early recurrence rate 20.4%; morbidity 7.7%) have established a sound standard for future endoscopists pursuing similar proficiency.6 In the Netherlands, EMR was slowly implemented from the early 2000s onwards. When our current study started in 2009, the participating endoscopists from 15, largely non-academic hospitals were gradually acquiring a tertiary referral setting in their geographical region. However, the low number of included patients in our study reflects their low case volumes compared to those of the few expert centres dominating the current EMR literature. Although our endoscopists were considered as experienced, some of them may have still been early in their learning curve when the study was commenced. Due to the small numbers in our study, a comparison of patients treated in academic and non-academic hospitals could not be made.

According to a recent study by Buchner et al., who reported the results of colorectal EMR in a single endoscopist tertiary referral setting, independent predictors of recurrence and complications were larger polyp size and piecemeal character of the resection.5 In our study, tumour size did not independently predict recurrence or complications. Tis may be explained by our inclusion criteria; the relatively large size of all included adenomas (mean diameter 46 mm) demanded all resections to be performed in a piecemeal fashion. Whether non-adherence to the study protocol with regards to the use of APC has influenced recurrence rates, remains questionable; previous studies have shown contradictory results with regards to the efect of APC on adenoma recurrence.6,13

We have shown that remnant or recurrent adenomatous tissue may present as dysplasia in random biopsies of an apparently healed mucosectomy scar. Tis may be explained by the fact that chromoendoscopy or virtual chromoendoscopy-like narrow band imaging were not routinely used for the detection of diminutive dysplastic lesions located at the scar. Moreover, we have detected recurrences in patients at 6, 12, and 24 months afer EMR, following a prior negative surveillance endoscopy. In these cases, the adenomatous tissue might well have been missed during the prior endoscopy due to insufcient imaging, poor bowel preparation or localization behind a fold. Strict surveillance with high quality endoscopes is essential in the follow-up of these patients, and random biopsies should be considered in the absence of visible recurrent lesions.

Nevertheless, the rather high recurrence rate may be clinically overcome by the overall success rate of the total intervention strategy including endoscopic surveillance and concurrent treatment of recurrence. Fourteen of our patients, in whom a remnant or recurrent adenoma was treated endoscopically, were free of recurrence at subsequent surveillance endoscopies. All recurrences were benign and no invasive cancers were detected during follow-up endoscopies.

27 Chapter 1

Te diagnostic workup of large rectal adenomas remains challenging. In our study cohort, three adenomas appeared to be submucosally invasive carcinomas. Polyp morphology and mucosal aspect may indicate a risk of submucosal invasion. Moss and Bourke identified the Paris 0-IIa+c classification as a risk factor, especially in combination with nongranular surface morphology.4 Although none of the included lesions in our study was classified as IIa+c, the mucosal aspect (Kudo pit pattern) of the resected adenoma was not routinely assessed and not described in 88% of lesions. Tis may reflect the unfamiliarity of all-round endoscopists with dedicated classification systems such as the Kudo pit pattern. Standard biopsies of large lesions will not be able to rule out invasive growth and may even cause submucosal fibrosis, thereby compromising submucosal lifing during the following therapeutic procedure. Endoscopic pattern recognition should be trained among colonoscopists, especially those who resect advanced adenomas.

With our study, we gained a clearer understanding of what is achievable when proven expert centres are scarce and dedicated gastroenterologists with lower case volumes treat these patients. By involving a single endoscopist per hospital and organizing consensus meetings, we aimed to approach centralization and uniform treatment protocols. However, our recurrence and complication rates emphasize the importance of training and centralization. Tis was also suggested by suboptimal diagnostic workup and possibly missed remnants and recurrences during follow-up. Perhaps, analogous to many surgical procedures, highly complex endoscopic procedures should become subject to formal training programs, strict volume standards and quality monitoring.

28 Endoscopic mucosal resection of large rectal adenomas ------APC Biopsy Biopsy Biopsy Treatment Biopsy + APC Biopsy ------Visible Visible Visible Visible Visible Visible Visible lesion / lesion random Bx random Lost to FU Lost to FU Lost to 1 FU24m (49 patients) Lost to FU (EMR failure) Lost to Outcome Recurrence Recurrence Recurrence Recurrence Recurrence No endoscopy No No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No ------APC APC Biopsy Biopsy Biopsy Treatment EMR + APC EMR + APC Biopsy + APC Biopsy ------visible Visible Visible Visible Visible Visible Visible Visible Visible Visible lesion / lesion random Bx random Lost to FU Lost to FU Lost to Random Bx Random Bx Random FU12m (49 patients) Lost to FU (EMR failure) Lost to Outcome Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence no recurrence no recurrence no recurrence no recurrence no recurrence no No recurrence No recurrence No recurrence No recurrence No recurrence No ------Biopsy Treatment EMR + APC EMR + APC Biopsy + APC Biopsy + APC Biopsy ------Visible Visible Visible Visible Visible Visible Visible lesion / lesion random Bx random FU 6m (43 patients) Lost to FU (EMR failure) Lost to Outcome Recurrence Recurrence Recurrence Recurrence Recurrence no endoscopy no no recurrence no recurrence no recurrence no recurrence no recurrence no recurrence no recurrence no recurrence no No endoscopy No endoscopy No endoscopy No No recurrence No recurrence No recurrence No recurrence No recurrence No ------APC APC TEM None Biopsy Biopsy Treatment Biopsy + APC Biopsy + APC Biopsy ------Visible Visible Visible Visible Visible Visible Visible Visible Random Bx Random Bx Random / random Bx / random Visible lesion lesion Visible FU 3m (53 patients) Outcome Remnant Remnant Remnant Remnant Remnant Remnant Remnant Remnant Remnant Remnant No remnant No remnant No remnant No remnant No remnant No remnant No remnant No remnant No remnant No remnant No remnant No remnant No No endoscopy No Summary of the follow-up of all patients with remnants and/or recurrences and/or remnants with all patients of the follow-up Summary of Table 2 Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

29 Chapter 1

REFERENCE LIST

1. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. Te National Polyp Study Workgroup. N Engl J Med 1993; 329: 1977–1981.

2. Zauber AG, Winawer SJ, O’Brien MJ, et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012; 366: 687–696.

3. Barendse RM, van den Broek FJ, Dekker E, et al. Systematic review of endoscopic mucosal resection versus transanal endoscopic microsurgery for large rectal adenomas. Endoscopy 2011; 43: 941–949.

4. Woodward TA, Heckman MG, Cleveland P, et al. Predictors of complete endoscopic mucosal resection of fat and depressed gastrointestinal neoplasia of the colon. Am J Gastroenterol 2012; 107: 650–654.

5. Buchner AM, Guarner-Argente C, Ginsberg GG. Outcomes of EMR of defant colorectal lesions directed to an endoscopy referral center. Gastrointest Endosc 2012; 76: 255–263.

6. Moss A, Bourke MJ, Williams SJ, et al. Endoscopic mucosal resection outcomes and prediction of submucosal cancer from advanced colonic mucosal neoplasia. Gastroenterology 2011; 140: 1909–1918.

7. Cotton PB, Eisen G, Romagnuolo J, et al. Grading the complexity of endoscopic procedures: results of an ASGE working party. Gastrointest Endosc 2011; 73: 868–874.

8. van den Broek FJ, de Graaf EJ, Dijkgraaf MG, et al. Transanal endoscopic microsurgery versus endoscopic mucosal resection for large rectal adenomas (TREND- study). BMC Surg 2009; 9: 4.

9. Hurlstone DP, Sanders DS, Cross SS, et al. A prospective analysis of extended endoscopic mucosal resection for large rectal villous adenomas: An alternative technique to transanal endoscopic microsurgery. Colorectal Dis 2005; 7: 339–344.

10. Karita M, Tada M, Okita K, et al. Endoscopic therapy for early colon cancer: Te strip biopsy resection technique. Gastrointest Endosc 1991; 37: 128–132.

11. Dindo D, Demartines N, Clavien PA. Classifcation of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004; 240: 205–213.

12. Cotton PB, Eisen G, Romagnuolo J, et al. A lexicon for endoscopic adverse events: report of an ASGE workshop. Gastrointest Endosc 2010; 71: 446–454.

13. Brooker JC, Saunders BP, Shah SG, et al. Treatment with argon plasma coagulation reduces recurrence afer piecemeal resection of large sessile colonic polyps: A randomised trial and recommendations. Gastrointest Endosc 2002; 55: 371–375.

30 Endoscopic mucosal resection of large rectal adenomas

1

31

CHAPTER 2

Perineal wound healing afer abdominoperineal resection for rectal cancer; a systematic review and meta-analysis

G.D Musters, C.J. Buskens, W.A. Bemelman, and P.J. Tanis

Dis Colon Rectum 2014;57(9):1129-1139. Chapter 2

ABSTRACT

Aim: Te primary aim of this systematic review with meta-analysis was to determine the impact of neoadjuvant (chemo)radiotherapy and an extralevator approach on perineal wound healing afer abdominoperineal resection for rectal cancer.

Methods: In March 2014, electronic databases were searched. Studies describing any outcome measure on perineal wound healing afer abdominoperineal resection for rectal cancer were included. Te primary endpoint was overall perineal wound problems within 30 days afer conventional or extralevator abdominoperineal resection with or without neoadjuvant radiotherapy. Secondary endpoints were primary wound healing, perineal hernia rate, and the efect of biological mesh closure on perineal wound problems.

Results: A total of 32 studies were included. Te pooled percentage of perineal wound problems afer primary perineal wound closure in patients who did not undergo neoadjuvant radiotherapy was 15.3% (95% Ci, 12.1–19.2) afer conventional abdominoperineal resection and 14.8% (95% Ci, 9.5–22.4) afer extralevator abdominoperineal resection. afer neoadjuvant radiotherapy, perineal wound problems occurred in 30.2% (95% Ci, 19.2–44.0) afer conventional abdominoperineal resection and in 37.6% (95% Ci, 18.6–61.4) afer extralevator abdominoperineal resection. Radiotherapy signifcantly increased perineal wound problems afer abdominoperineal resection (OR 2.22; 95% Ci, 1.45–3.40; p < 0.001). Afer biological mesh closure of the pelvic foor following extralevator abdominoperineal resection with neoadjuvant radiotherapy, the percentage of perineal wound problems was 7.3% (95% Ci, 1.5–29.3).

Conclusion: Neoadjuvant radiotherapy signifcantly increases perineal wound problems afer abdominoperineal resection for rectal cancer.

34 Perineal wound healing afer abdominoperineal resection for rectal cancer

INTRODUCTION

Perineal wound problems afer abdominoperineal resection (APR) for rectal cancer is reported in up to 57% of patients.1 If the perineum does not heal primarily, secondary wound healing may prolong hospital stay, may necessitate surgical reintervention, and ofen requires intensive wound care for several months, with the risk of developing a persistent sinus afer 1 year.2–4 In the past decade, perineal wound healing afer APR has gained more attention because of the intensifed treatment of distal rectal cancer. Afer the widespread adoption of the technique of total mesorectal excision and the increasing use of neoadjuvant radiotherapy, locoregional disease control and survival of rectal cancer have improved.5,6 Te oncological outcome remained poor 2 due to positive circumferential resection margins and tumour perforations in patients with low rectal cancer undergoing APR.7 When the total mesorectal excision plane is followed all the way down to the pelvic foor, as performed in conventional APR (cAPR), a typical coning of the specimen occurs by which the resection margins are compromised. To overcome this problem, wider excisions with en bloc resection of the distal rectum, sphincter complex, and levator muscles have been introduced, resulting in a cylindrical specimen. Tis extralevator APR (eAPR) reduces the rate of positive resection margins and tumour perforation in distal rectal cancer, and improves oncological outcome.8,9

As a downside of improved oncological outcome, both radiotherapy and extended resections have been suggested to increase perineal wound healing problems afer APR.10 Terefore, the aim of this systematic review with meta-analysis was to determine the impact of these changes in rectal cancer management on perineal wound healing afer APR. Secondarily, the recent introduction of biological mesh reconstruction of the pelvic foor, aiming at improved perineal wound healing and the prevention of perineal hernia, was evaluated.

METHODS

Search strategy All studies reporting on perineal wound healing afer either cAPR or eAPR for rectal cancer were considered eligible for review. Te electronic databases of the national institutes of health Pubmed (1952–2014), EMBASE (1984–2014), and Cochrane library (2008–2013) were systematically searched in March 2014. Te following medical subject heading (mesh) terms were used; rectal neoplasms, rectal, neoplasms, abdomen, perineum, colorectal surgery, surgical procedures, operative, general surgery, wound healing, reconstructive surgical procedures, wounds, and injuries. Other search terms were abdominal perineal resection, abdominoperineal resection, abdominoperineal excision, abdominal perineal excision, perineum surgery, perineal surgery, primary healing, repair, wound, and healing. No restrictions considering patient age or technique of APR were applied. Anal cancer, benign disorders, gynaecological cancers, exenterative procedures, sacral resections, use of myocutaneous faps for perineal closure, delayed perineal wound closure, animal studies, non-English articles, congress abstracts, studies before 1990, previously published data, and studies with fewer than 5 patients were excluded. Of the initially identifed publications, titles and abstracts

35 Chapter 2

were screened to exclude nonrelated publications. Ten, the full text of the remaining articles was read to determine whether they were eligible for inclusion. Of the included articles, the references were cross-checked to identify any additional studies of interest. Studies were included by two independent authors (G.M., P.T.). Any disagreements between both reviewers were resolved by consultation with an expert specialist (W.B.).

Outcome parameters Te primary end point was the percentage of overall perineal wound problems within 30 days afer cAPR or eAPR with or without neoadjuvant radiotherapy. Besides this combined end point, a distinction between superfcial (wound infection and dehiscence) and deep perineal wound problems (perineal abscess and presacral abscess) was made if relevant data were provided by the selected articles. Secondary outcome measures were superfcial perineal wound problems, deep perineal wound problems, primary wound healing rate, perineal hernia rate at the end of follow- up, and the efect of biological mesh assisted closure on perineal wound problems and perineal hernia. Primary wound healing was included as a secondary end point, because this is not always the complement of perineal wound problems. For example, a seroma or hematoma may be present, while the perineal wound has been healed primarily. Furthermore, some studies only report on perineal wound healing rate without mentioning the perineal wound complications.

Validity and eligibility assessment All included studies were assessed for methodological quality and risk of bias. For cohort studies, the Newcastle Ottawa quality assessment scale for cohort studies was used to assess risk of bias.11 Te quality items scored were as follows: representativeness of the exposed cohort, selection of the nonexposed cohort, ascertainment of exposure, the absence of outcome of interest at the start of the study, comparability of the design or analysis, assessment of outcome, duration of follow-up and lost to follow-up. For randomised controlled trials (RCTs), the Jadad scoring system was used to assess the risk of bias.12 When the randomised groups of the RCTs were not used as described in the RCT, the Newcastle Ottawa quality assessment scale for cohort studies was also used to assess risk of bias.11

Data extraction Two authors independently extracted data from the included studies (G.M., P.T.). Disagreements were resolved by discussion between the two reviewers. If no consensus could be reached, a third specialist author was consulted and decided (W.B.). In the case of RCTs on oncological interventions in which perineal wound healing was a secondary outcome parameter, data of both arms of the trial were combined or data from the appropriate arm was used. Data on outcome measures were extracted for specifc subgroups of patients from the included articles depending on the availability of separate data with regard to the extent of resection (eAPR or cAPR) and the use of neoadjuvant radiotherapy.

36 Perineal wound healing afer abdominoperineal resection for rectal cancer

Statistical analysis From each study, a weighted average of the logit proportions was determined by the use of the generic inverse variance method. Te logit proportions were back transformed to the summary estimate and 95% CIs were obtained in a summary proportion. Heterogeneity was assessed by using the I2 and χ2 statistics, and the data were considered signifcant if the p value (χ2) was <0.1 and I2 was >75%. Analyses were performed with the use of Review manager (Revman, Version 5.2. Copenhagen: the nordic Cochrane Centre, the Cochrane Collaboration, 2012). Analyses were performed with the inverse variance analysis by using a random-efects model. Te estimated pooled OR and the corresponding 95% CIs were illustrated in forest plots. Te pooled data were considered statistically signifcant if the p-value was <0.05. Publication bias was assessed by using a funnel plot. 2

RESULTS

Te systematic search resulted in a total of 2318 articles. Afer the removal of duplicates, 2039 articles remained. Afer screening of publication titles and abstracts, 145 publications were retrieved for full text review. Of these 145 articles, 118 were excluded for reasons as summarized in fgure 1. Of the included articles, the reference lists were checked, which resulted in 5 additional eligible studies. In total, 32 publications were included with a total number of 7247patients. Seven studies were RCTs,13–19 and 25 were cohort studies. Of the cohort studies, 17 studies were consecutive series.1,2,4,9,10,20–31 Te following groups of the RCTs were used: in 1 RCT, only the control arm of the study was included because local gentamicin was used in the experimental arm.15 In 1 RCT, only 1 study arm was included, because in the other study arm the perineum was packed and not primarily closed.13 In 1 RCT, only the patients undergoing a cAPR were included.17 In 1 RCT, only the patients receiving preoperative radiotherapy and no postoperative radiotherapy were included.19 In 2 RCTs, both study arms were merged and described as one,14,18 and each individual study arm was included of the remaining RCTs.16 Tis resulted in a fnal number of 3941 patients to be included for data analysis. Te risk of bias of all included studies is displayed in appendix 1.

Study characteristics APR with primary perineal closure Of the 32 included studies, primary perineal wound closure afer cAPR was described in 21 studies, with a total number of 2989 patients (table 1).1,2,9,10,13–20,22,25,29,30,32–36 Five studies described the use of an extralevator approach (eAPR) in 728 patients.4,21,32,37,38 All patients underwent APR for primary rectal cancer, with the exception of 2 studies in which a minority of patients had recurrent rectal cancer (n=41;1%).25,39 Only 3% (n=115) of the patients underwent minimally invasive surgery as described in 3 studies.9,25,32 Te use of an omentoplasty was described in 3 studies and ranged between 20% and 88%.14,35,37 Te use of a presacral drain was described in 14 studies (56%–100%)1,13– 15,18,21,29,32–38 and a perineal wound drain in 3 studies.20,36,38 Radiotherapy was applied in 16 studies, and the proportion of patients receiving radiotherapy ranged from 28% to 100%.1,2,4,9,10,16–21,32,33,36–38

37 Chapter 2

Figure 1 Flow-diagram

Perineal wound problems afer APR with primary perineal wound closure Te percentages of perineal wound problems afer primary perineal wound closure for each individual study are displayed in Table 1. Afer cAPR without neoadjuvant radiotherapy (6 studies; 1000 patients; Table 2), overall perineal wound problems occurred in a pooled percentage of 15.3% (95% CI 12.1-19.2; I2=50%, χ2=0.07).2,10,14,15,17,36 Assuming that radiotherapy was not applied in an additional number of four studies published between 1991 and 1995, although this was not explicitly described, the pooled percentage of all perineal wound problems was 18.7% (95% CI 13.9-24.9; I2=78%, χ2<0.01; 10 studies, n=1218).2,10,13-15,17,29,34-36 Superfcial perineal wound problems occurred with a pooled percentage of 15.0% (95% CI 10.3-21.6; I2=78%, χ2<0.01; 8 studies, n=1032)2,10,15,17,29,34-36 and deep perineal wound problems with a pooled percentage of 8.8% (95% CI 2.7- 25.0, I2=92%, χ2<0.01; 5 studies, n=744).2,13-15,36

Afer cAPR with neoadjuvant radiotherapy in all patients, the pooled percentage of perineal wound problems was 30.2% (95% CI 19.2-44.0, I2=90%, χ2<0.01;5 studies, n= 730).1,2,10,17,36 Superfcial perineal wound problems had a pooled percentage of 19.4% (95% CI 14.5-25.4; I2=64%, χ2=0.026; 38 Perineal wound healing afer abdominoperineal resection for rectal cancer

5 studies, n=730)1,2,10,17,36 and deep perineal wound problems had a pooled percentage of 7.1% (95% CI 1.3-31.1; I2=87%, χ2<0.01; 3 studies, n=362).1,2,36 Radiotherapy showed a signifcant increase in overall perineal wound problems afer cAPR (OR 1.74; 95% CI 1.29-2.34; p=0.0003; I2=13%, χ2=0.33; 4 studies, n=1496). 2,10,17,36

Te pooled percentage of overall perineal wound problems afer eAPR without neoadjuvant radiotherapy was 14.8% (95% CI 9.5-22.4; I2=30%, χ2=0.233; 2 studies, n=217).4,37 In patients who underwent neoadjuvant radiotherapy, the pooled percentage of perineal wound problems afer eAPR was 37.6% (95% CI 18.6-61.4; I2=85%, χ2=0.01; 2 studies, n=169).4,37 Neoadjuvant radiotherapy signifcantly increased the risk of perineal wound problems afer both cAPR and eAPR (OR 2.22; 2 95% CI 1.45-3.40; p=0.0003; Figure 2). Te pooled percentages with 95% confdence intervals of the diferent subgroups are summarized in Figure 3.

Figure 2 Forrest plot of perineal wound problems afer cAPR and eAPR in patients who underwent neoadjuvant radiotherapy compared to those who did not undergo radiotherapy

Figure 4 Forrest plot of perineal wound healing afer cAPR and eAPR in patients who underwent neoadjuvant radiotherapy compared to those who did not undergo radiotherapy

Perineal wound healing and perineal hernia afer APR with primary perineal wound closure Afer cAPR without radiotherapy, primary perineal wound healing had a pooled weighted estimate of 81.8% (95% CI 64.7-91.7; I2=93%, χ2<0.01; 5 studies, n=736).2,10,14,15,33 Afer neoadjuvant radiotherapy, primary perineal wound healing had a pooled estimate of 63.9% (95% CI 45.7- 39 Chapter 2

78.8; I2=93%, χ2<0.01; 5 studies, n=650).1,2,10,19,33 Tere was only one study reporting the primary perineal wound healing rate afer eAPR with all patients undergoing radiotherapy, which was 43% (n=53/123).4 Radiotherapy showed a signifcant reduction in primary perineal wound healing afer both cAPR and eAPR (OR 0.27; 95% CI 0.16-0.45;p<0.001; Figure 4).2,4,10,33

Perineal hernias afer cAPR occurred in a pooled percentage of 1.8% (95% CI 0.4-8.3; I2=84%, χ2<0.01; 4 studies, n=1012)2,9,10,16, and in 2.0% afer eAPR (95% CI 0.5-7.0; I2=59%, χ2=0.09; 3 studies, n=367).21,37,38

Perineal wound problems and perineal hernia afer biological mesh assisted closure Biological mesh assisted perineal wound closure was performed in 224 patients from 8 studies (Table 3). 16,23,24,26-28,31,39 All 224 patients underwent eAPR. Te use of a presacral drain was reported in two studies26,39 and a perineal wound drain was reported in four studies.16,24,27,39 Te proportion of patients receiving radiotherapy ranged from 43% to 100% and an omentoplasty was used in 28% (n=36) and described in two studies.23,31

Te biological meshes consisted in four studies of cross linked porcine meshes23,26,27,39, and in one studies of non-cross linked porcine meshes.28 In one study, both cross linked and non-cross linked meshes were used31, and in two studies human dermal meshes were used.16,24 Perineal wound problems afer eAPR and at least 75% of the patients treated with radiotherapy occurred in a pooled percentage of 24.3% (95% CI 14.1-38.6; I2=21%, χ2=0.28; 4 studies, n=78).24,28,31,39 If only two studies were included in which all patients underwent neoadjuvant radiotherapy, perineal wound problems afer eAPR occurred in a pooled percentage of 7.3% (95% CI 1.5-29.3; I2=0%, χ2=0.81; n=20).24,31 Perineal hernias afer biological mesh closure following eAPR were reported in a pooled percentage of 8.2% (95% CI 4.8-13.6; I2=0%, χ2=0.45; 6 studies, n=188).16,23,24,26,28,39

Figure 3 Te pooled percentages with 95% confdence intervals

40 Perineal wound healing afer abdominoperineal resection for rectal cancer

DISCUSSION

Neoadjuvant radiotherapy signifcantly increased perineal wound problems afer APR for rectal cancer with an Odds Ratio of 2.22. Although an extralevator approach has also been associated with higher wound complication rates, similarly weighted pooled percentages of perineal wound problems afer cAPR and eAPR were found in homogeneous subgroups depending on whether or not radiotherapy was applied. Without neoadjuvant radiotherapy, perineal wound problems occurred in 15.3% afer cAPR and 14.8% afer eAPR; corresponding wound problem rates following radiotherapy were 30.2% and 37.6%. Tis suggests that, not the extent of the resection, but mainly neoadjuvant radiotherapy is the most important therapy-related risk factor for perineal wound 2 complications.

Te majority of the available studies have signifcant methodological shortcomings, and these data should be interpreted with caution, especially the studies with high statistical heterogeneity. Te relatively high statistical heterogeneity could be the result of the non-standardized prospective evaluation and registration of the perineal wounds and the non-uniform defnitions used for classifying perineal wound complications. Both the use of neoadjuvant radiotherapy and an extralevator approach for the APR procedure have increased over time, with more focus on perineal wound complications in more recently published articles. Terefore, a restriction of the studies before 1990 was applied, and subgroup analyses for the amount of radiotherapy and the extent of resection have been performed in this systematic review. Also, a detailed description of the diferent types of perineal wound problems is lacking in a substantial number of studies, such as superfcial wound infection, dehiscence without infection, and deep pelvic abscess. Tis was the reason for choosing a combined end point of all perineal wound problems as primary outcome measure for the present analysis. Although all these limitations may have infuenced the pooled analyses, the presently reported data are the best available evidence at this time. Neoadjuvant radiotherapy might increase perineal wound problems and reduce perineal wound healing because it damages DNA and proteins, which could lead to apoptotic cell death.40Apoptosis in endothelial cells leads to increased vascular permeability. Tis permeability results in thrombosis and intima proliferation.40 Furthermore, radiotherapy afects pro-infammatory cytokines in the early phase.40 Terefore, infammation may not resolve adequately, leading to uncontrolled matrix accumulation and fbrosis. Also decreased levels of nitric oxide and matrix metalloproteinase may contribute to the inadequate sof tissue reconstruction.41

Primary perineal wound closure afer APR has been the standard technique for a long period of time. Tis entails a layered closure of the remaining structures, namely, the remnants of the pelvic foor, ischioanal and subcutaneous fat, and perineal skin. Afer an extralevator approach, only the fat and perineal skin can be closed because the pelvic foor has been resected. Despite the much larger pelvic defect, satisfying outcome on perineal wound healing has been published for primary perineal closure afer eAPR.4,21,37 But this is not uniformly reported, and others have used additional surgical procedures to improve perineal wound healing afer eAPR. Autologous tissue fap such as the vertical rectus abdominis myocutaneous fap and gluteal fap, have been used for closure of large perineal defects afer extended resections for primary or recurrent rectal cancer or salvage surgery for anal cancer.42–45 Because of the donor site morbidity, increased operative time, and higher

41 Chapter 2

costs, it is questioned whether autologous tissue fap should be performed as a routine in patients undergoing APR for rectal cancer. Pelvic foor reconstruction with the use of a biological mesh has been introduced recently and appears to be an attractive alternative for tissue fap based on a recent systematic review.46 However, a comparison between primary perineal wound closure and biological mesh-assisted closure has not yet been made. Literature on biological mesh-assisted perineal closure afer APR is scarce (table 3). In contrast, most of the APR studies have a primary oncological focus, whereas perineal wound healing and perineal hernia formation are ofen primary outcome measures in biological mesh studies. Tis does not allow for a formal comparison between primary perineal wound closure and biological mesh assisted closure based on literature data. Although the pooled analysis suggests a reduced percentage of perineal wound problems afer biological mesh assisted closure, RCTs are needed to draw defnitive conclusions. To date, 2 multicentre trials are being conducted in which pelvic foor reconstruction using a biological mesh is compared with primary perineal wound closure (BIOPEX-study; Clinicaltrials.gov identifer NCT01927497) or gluteus maximus myocutaneous fap closure (NEAPE; Clinicaltrials.gov identifer NCT01347697).

Perineal wound problems are also likely to be related to the dead space in the pelvis with bacterial contamination. An omentoplasty may solve this problem by flling of the pelvic cavity, improving the local immune response and promoting angiogenesis.47,48 Terefore, the application of an omentoplasty has been suggested to improve primary perineal wound healing.49,50 Only fve selected studies described the use of an omentoplasty without separate data on perineal wound healing. Te efect of this additional surgical procedure could therefore not be analysed in the present meta-analysis.

In our systematic review, we found a rather high weighted pooled percentage of perineal hernia in patients undergoing pelvic foor reconstruction using a biological mesh. Tis contradictory fnding is most likely related to a more detailed description of perineal complications with more adequate follow-up in biological mesh studies compared to oncological APR studies using primary perineal wound closure. In addition, the extent of the resection (only eAPR in biological mesh studies) and a learning curve efect of a new surgical technique may explain the relatively high perineal hernia rate.

In conclusion, our systematic review demonstrates that neoadjuvant radiotherapy signifcantly increases perineal wound problems afer APR, while the extralevator approach seems to be of less infuence. Biological mesh assisted perineal wound closure is a promising technique to improve perineal wound healing, but we have to await results of ongoing randomised trials.

42 Perineal wound healing afer abdominoperineal resection for rectal cancer + ~ + ~ + + ^ ^ ^ ^ + ~ ^ ^ - - - - - 2 6 60 60 12 12 24 25 (range) Months Months Follow-up 45 (1-89) 29 (12-48) 34 (25-57) 45 (1-130) 38 (16-58) 40 (25-55) ------7 (1) 1 (1) 1 (1) n(%) 4 (13) hernia Incisional Incisional ------n(%) PWH 9 (43) 6 (29) 26 (33) 61 (81) 25 (86) 57 (61) 36 (75) 74 (89) 633 (91) 132 (65) 122 (76) 111 (74) 112 (68) - - - 7 (6) Perineal wound wound Perineal n(%)* 6 (19) 3 (17) cPWP 65 (27) 22 (28) 15 (20) 23 (15) 12 (57) 44 (47) 65 (41) 10 (21) 18 (11) 10 (48) 19 (23) 106 (15) 107 (22) 2 ------0 0 1 (1) n(%)* 26 (4) 6 (29) 5 (10) dPWP 18 (11) 10 (48) - - - - - 7 (6) n(%)* 6 (19) 6 (29) 3 (17) 5 (10) sPWP 65 (27) 22 (28) 14 (19) 23 (15) 80 (11) 44 (47) 65 (41) 19 (23) 107 (22) ------0 5 (6) n(%) 9 (28) 25 (33) 40 (26) 48 (39) 273 (39) 107 (67) Concomitant chemotherapy ------0 0 n(%) 9 (28) 68(44) 66 (84) 25 (33) 90 (73) 39 (41) 273 (39) 21 (100) 117 (73) 251 (52) Neoadjuvant treatment Neoadjuvant 202 (100) Radiotherapy ------Yes Yes Yes Yes Yes Yes Yes (Yes/no) antibiotics Preoperative Preoperative # cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR eAPR cAPR/ Operation ------No No No Yes Yes Yes Yes Yes Yes Yes Yes series Unclear Consecutive Consecutive RCT RCT RCT RCT RCT RCT RCT Study Study design Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Study characteristics Study n 75 94 155 696 160 165 32 of 67 32 of 40 21 of 51 18 of 38 29 of 97 48 of 45 21 of 79 of 158 79 of 103 83 of 239 of 249 239 of 176 124 of 231 149 of 202 of 1350 202 of 1805 485 of year 2013 2012 2011 2011 2010 2009 2009 2009 2008 2008 2006 2006 2005 2001 2001 2001 1995 1994 1994 9 1 8 .1

33

20 5 13 1 15 30 2

2

14 10 . 17 32 36

.

9

35 . 16 22 er APR with primary perineal closure for rectal cancer primary with er APR for perinealf a closure problems wound Perineal Table 1 Table Included studies author First et al. AT Hawkins et al. JG Han D et al. Asplund f et alakis MG Pramate L et al Zorcolo El-Gazzaz G et al. D et al. f ore Sebag-Monte NP et al. West De Bruin AF et.al. S et al. Chan et al. Artioukh DY MA et al. Chadwick KM et al. Bullard U et al. Gruessner E et al. Kapiteijn H et al. Scheibach A et al Fingerhut JP et al. Delalande JY et al Wang

43 Chapter 2 + + ^ ^ ~ - - 60 24 120 26 (1-88) 30 (1-91) - - - - 3 (6) 2 (1) 1 (1) - - 7 (15) 47 (69) 19 (24) 83 (80) 182 (65) 2 (4) 6 (11) 21 (31) 36 (46) 37 (18) 18 (17) 87 (31) - - - - 3 (1) 6 (6) 18 (6) 2 (4) 6 (11) 21 (31) 36 (46) 34 (16) 12 (12) 69 (24) - - - 19 (9) 46 (87) 15 (19) 16 (15) - - 46 (87) 60 (76) 65 (31) 62 (60) 123(44) - - - Yes Yes Yes Yes cAPR cAPR eAPR eAPR eAPR eAPR eAPR No No No No Yes Yes Yes Cohort Cohort Cohort Cohort Cohort Cohort Cohort 53 210 104 282 68 of 102 68 of 288 46 of 158 79 of 1992 1991 2013 2011 2009 2005 2001

29

1 38 37 34 2 32

4 Continuation of Table 1 Table of Continuation Robles Campos R et al. Campos Robles WP et al. Mazier S et al. Toshniwal D et al. Asplund Bbenek M et al. ę E et al. De Broux A et al. Nissan += median follow-up. ^= protocolised follow-up. ~= mean follow-up. RCT= randomised controlled trial. resection. cAPR= abdominoperineal cial f sPWP=super perineal conventional resection. abdominoperineal wound problems (infections and eAPR= dehiscence) within 30 days. extralevator dPWP=deep perineal wound problems (perineal and presacral 30 was counted within *=a days. only patient abscesses) and presacral abscesses perineal dehiscence, (infections, problems wound perineal 30 within cPWP=combined days. abscesses) healing. wound perineal PWH=primary present. was complication one than more if once

44 Perineal wound healing afer abdominoperineal resection for rectal cancer

2 ------1 (1) 6 (1) 6 (5) n(%) 1 (0.4) 5 of 94 (5) 5 of Incisional hernia hernia Incisional ------n(%) PWH 86 (74) 36 (84) 13 (35) 44 (77) 53 (43) 234 (86) 399 (94) 129 (81) Perineal wound Perineal - - n(%)* 6 (10) cPWP 55 (47) 10 (23) 55 (20) 51 (12) 65 (26) 42 (18) 10 (15) 13 (15) 12 (26) 60 (49) 27 (17) ------0 0 6 (4) n(%) 14 (5) 12 (3) dPWP 12 (10) - - - - 0 n(%) sPWP 16 (14) 49 (18) 46 (11) 65 (26) 42 (18) 10 (15) 13 (15) 48 (39) 21 (13) No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes/no Radiotherapy N 43 37 57 68 87 46 58 117 273 423 251 234 123 159 # er APR with primary perineal closure in patients being treated with or without neoadjuvant radiotherapy. neoadjuvant without or with being treated primary with er APR in patients perinealf a closure

Operation cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR cAPR eAPR eAPR eAPR eAPR cAPR/eAPR 3 3 0 ^ 2 7 1 3 17 6 3 4 Perineal wound problems wound Perineal Included studies author First KM et al. Bullard MA et al. Chadwick El-Gazzaz G et al. E et al. Kapiteijn L et al. Zorcolo E et al. De Broux A et al. Nissan Table 2 Table once if *=a was than more resection. counted only patient abdominoperineal eAPR= resection. extralevator abdominoperineal cAPR= conventional wound perineal dPWP= deep 30 days. within dehiscence) and (infections problems wound perineal cial f sPWP=super present. was complication one within dehiscence) and abscesses (infection, problems wound perineal cPWP= 30 combined days. within abscesses) presacral and (perineal problems trial. controlled is a randomised study ^= this healing. wound perineal PWH= primary 30 days.

45 Chapter 2 + + + + + + + + (range) Months 9 (0-21) 8 (2-16) 36 (1-67) 21 (2-54) 20 (5-26) 20 (0-45) Follow-up 22 (16-46) 38 (23-66) - - 0 0 3 (6) 1 (3) n(%) 5 (14) 3 (10) hernia Incisional Incisional - - - - - n(%) PWH 20 (59) 8 (100) 11 (92) 0 1 (8) n(%)* 3 (11) 4 (11) 6 (25) cPWP 16 (30) 11 (32) 15 (50) Perineal wound Perineal - - - - 0 0 4 (8) 3 (9) n(%)* dPWP 0 1 (8) n(%)* 3 (11) 8 (24) 4 (11) 6 (25) sPWP 12 (23) 15 (50) - n(%) 23 (43) 17 (61) 26 (76) 10 (29) 20 (83) 14 (47) 12 (100) Concomitant chemotherapy n(%) 23 (43) 19 (67) 28 (82) 10 (29) 20 (83) 14 (47) 8 (100) 12 (100) Neoadjuvant treatment Neoadjuvant Radiotherapy ------Yes (Yes/no) antibiotics Preoperative Preoperative # eAPR eAPR eAPR eAPR eAPR eAPR eAPR eAPR Operation cAPR/eAPR - No Yes Yes Yes Yes Yes Yes series Consecutive Consecutive RCT Study Study design Cohort Cohort Cohort Cohort Cohort Cohort Cohort Study characteristics Study n 53 28 34 12 8 of 13 8 of 35 of 67 35 of 57 24 of 31 30 of year 2013 2013 2013 2012 2011 2011 2011 2010 1 3 9 3 8 7 6 23 2 2 2 6 4 1 2 er APR for rectal cancer for er APR the perineumf of a closure assisted mesh Biological Table 3 Table Included studies author First KK et al. Jensen et al. SL Kipling O et al. Peacock et al. JG Han HK et al. Christensen et al. RSJ Dalton PG et al. Vaughan-Shaw Han JG et al. JG Han problems += wound perineal median deep dPWP= follow-up. days. 30 RCT= within randomised trial. controlled dehiscence) cAPR= resection. and abdominoperineal eAPR= conventional resection. (infections abdominoperineal extralevator *= problems a wound perineal patient was cial f super sPWP= only present. was complication one than more if once counted PWH= days. 30 within abscesses) presacral and abscesses perineal dehiscence, (infections, problems wound perineal combined cPWP= days. 30 within abscesses) presacral and (perineal healing. wound perineal primary

46 Perineal wound healing afer abdominoperineal resection for rectal cancer

Appendix 1

Newcatle Ottawa quality assessment scale1 cohort studies Included Studies Jadad et al.2 Selection Comparability Outcome Total (0-5) (0-4) (0-2) (0-3) (0-9) Artioukh DY 200620 - * - ** 3 Asplund D 201132 - *** - ** 5 Bębenek M 200921 - * - ** 3 Bullard KM 200510 - *** ** * 6 Chadwick MA 200633 - * - ** 3 2 Chan S 200822 - ** - ** 4 Christensen HK 201139 - *** - * 4 Dalton RSJ 201123 - ** - ** 4 De Broux E 200537 - *** - ** 5 De Bruin AF 20081 - ** - * 3 Delalande JP 199413 3 ** - ** 4 El-Gazzaz G 20092 - ** - ** 4 Fingerhut A 199514 3 ** - ** 4 Gruessner U 200115 1 ** - ** 4 Han JG 201216 1 ** - ** 4 Han JG 201024 - - - ** 2 Hawkin AT 201325 - ** * ** 5 Jensen KK 201326 - * - * 2 Kapiteijn E 200117 3 ** - ** 4 Kipling SL 201327 - * - * 2 Mazier WP 199134 - *** - * 4 Nissan A 20014 - *** - ** 5 Peacock O 201328 - ** - ** 4 Pramatefakis MG 201118 3 ** - ** 4 Robles Campos R 199229 - ** - ** 4 Sebag-Montefore D 200919 3 ** - * 3 Scheidbach H 200130 - ** - ** 4 Toshniwal S 201338 - ** - * 3 Vaughan-Shaw PG 201131 - * - ** 3 Wang JY 199435 - ** - * 3 West NP 20099 - *** ** ** 7 Zorcolo L 201036 - *** - ** 5

47 Chapter 2

REFERENCE LIST

1. de Bruin AF, Gosselink MP, Wijfels NA, Coene PP, van der Harst E. Local gentamicin reduces perineal wound infection afer radiotherapy and abdominoperineal resection. Tech Coloproctol 2008; 12(4): 303-7.

2. El-Gazzaz G, Kiran RP, Lavery I. Wound complications in rectal cancer patients undergoing primary closure of the perineal wound afer abdominoperineal resection. Dis Colon Rectum 2009; 52(12): 1962-6.

3. Marks CG, Leighton M, Ritchie JK, Hawley PR. Primary suture of the perineal wound following rectal excision for adenocarcinoma. Br J Surg 1976; 63(4): 322-6.

4. Nissan A, Guillem JG, Paty PB, et al. Abdominoperineal resection for rectal cancer at a specialty center. Dis Colon Rectum 2001; 44(1): 27-35.

5. Fleming FJ, Pahlman L, Monson JR. Neoadjuvant therapy in rectal cancer. DisColon Rectum 2011; 54(7): 901-12.

6. Heald RJ, Moran BJ, Ryall RD, Sexton R, MacFarlane JK. Rectal cancer: the Basingstoke experience of total mesorectal excision, 1978-1997. ArchSurg 1998; 133(8): 894-9.

7. den Dulk M, Putter H, Collette L, et al. Te abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of fve European randomised clinical trials on rectal cancer. EurJCancer 2009; 45(7): 1175-83.

8. Stelzner S, Koehler C, Stelzer J, Sims A, Witzigmann H. Extended abdominoperineal excision vs. standard abdominoperineal excision in rectal cancer--a systematic overview. Int J Colorectal Dis 2011; 26(10): 1227-40.

9. West NP, Anderin C, Smith KJ, Holm T, Quirke P. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. BrJSurg 2010; 97(4): 588-99.

10. Bullard KM, Trudel JL, Baxter NN, Rothenberger DA. Primary perineal wound closure afer preoperative radiotherapy and abdominoperineal resection has a high incidence of wound failure. Dis Colon Rectum 2005; 48(3): 438-43.

11. Te Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Ottawa, Ontario: 2008. Available at: http://www.ohri.ca/programs/clinical_epidemiology/ oxford.htm Accessed September 20, 2013.

12. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomised clinical trials: is blinding necessary? Control Clin Trials 1996; 17(1): 1-12.

13. Delalande JP, Hay JM, Fingerhut A, Kohlmann G, Paquet JC. Perineal wound management afer abdominoperineal rectal excision for carcinoma with unsatisfactory hemostasis or gross septic contamination: primary closure vs. packing. A multicenter, controlled trial. French Association for Surgical Research. Dis Colon Rectum 1994; 37(9): 890-6.

14. Fingerhut A, Hay JM, Delalande JP, Paquet JC. Passive vs. closed suction drainage afer perineal wound closure following abdominoperineal rectal excision for carcinoma. A multicenter, controlled trial. Te French Association for Surgical Research. Dis Colon Rectum 1995; 38(9): 926-32.

15. Gruessner U, Clemens M, Pahlplatz PV, Sperling P, Witte J, Rosen HR. Improvement of perineal wound healing by local administration of gentamicin-impregnated collagen feeces afer abdominoperineal excision of rectal cancer. American Journal of Surgery 2001; 182(5): 2001.

16. Han JG, Wang ZJ, Wei GH, Gao ZG, Yang Y, Zhao BC. Randomised clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. Am J Surg 2012; 204(3): 274-82.

48 Perineal wound healing afer abdominoperineal resection for rectal cancer

17. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001; 345(9): 638-46.

18. Pramatefakis MG, Raptis D, Kanellos D, et al. Abdominoperineal resection for rectal cancer: is the pelvic drain externalization site an independent risk factor for perineal wound healing? Int J Surg Oncol 2012; 2012: 156935.

19. Sebag-Montefore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus selective postoperative chemoradiotherapy in patients with rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomised trial. Lancet 2009; 373(9666): 811-20.

20. Artioukh DY, Smith RA, Gokul K. Risk factors for impaired healing of the perineal wound afer abdominoperineal resection of rectum for carcinoma. Colorectal Dis 2007; 9(4): 362-7. 2 21. Bebenek M. Abdominosacral amputation of the rectum for low rectal cancers: ten years of experience. Ann Surg Oncol 2009; 16(8): 2211-7.

22. Chan S, Miller M, Ng R, et al. Use of myocutaneous faps for perineal closure following abdominoperineal excision of the rectum for adenocarcinoma. Colorectal Dis 2010; 12(6): 555-60.

23. Dalton RS, Smart NJ, Edwards TJ, Chandler I, Daniels IR. Short-term outcomes of the prone perineal approach for extra-levator abdomino-perineal excision (elAPE). Surgeon 2012; 10(6): 342-6.

24. Han JG, Wang ZJ, Gao ZG, Xu HM, Yang ZH, Jin ML. Pelvic foor reconstruction using human acellular dermal matrix afer cylindrical abdominoperineal resection. Dis Colon Rectum 2010; 53(2): 219-23.

25. Hawkins AT, Berger DL, Shellito PC, Sylla P, Bordeianou L. Wound dehiscence afer abdominoperineal resection for low rectal cancer is associated with decreased survival. Dis Colon Rectum 2014; 57(2): 143-50.

26. Jensen KK, Rashid L, Pilsgaard B, Moller P, Wille-Jorgensen P. Pelvic foor reconstruction with a biological mesh afer extralevator abdominoperineal excision leads to few perineal hernias and acceptable wound complication rates with minor movement limitations: single-centre experience including clinical examination and interview. Colorectal Dis 2014; 16(3): 192-7.

27. Kipling SL, Young K, Foster JD, et al. Laparoscopic extralevator abdominoperineal excision of the rectum: short-term outcomes of a prospective case series. Tech Coloproctol 2013.

28. Peacock O, Simpson JA, Tou SI, et al. Outcomes afer biological mesh reconstruction of the pelvic foor following extra-levator abdominoperineal excision of rectum (APER). Tech Coloproctol 2014.

29. Roboles campos R, Garcia AJ, Parrilla PP, et al. Management of the perineal wound following adominoperineal resection: Prospective study of three methods. British Journal of Surgery 1992; 79(1): 1992.

30. Scheidbach H, Schneider C, Konradt J, et al. Laparoscopic abdominoperineal resection and anterior resection with curative intent for carcinoma of the rectum. Surg Endosc 2002; 16(1): 7-13.

31. Vaughan-Shaw PG, King AT, Cheung T, et al. Early experience with laparoscopic extralevator abdominoperineal excision within an enhanced recovery setting: analysis of short-term outcomes and quality of life. Ann R Coll Surg Engl 2011; 93(6): 451-9.

32. Asplund D, Haglind E, Angenete E. Outcome of extralevator abdominoperineal excision compared with standard surgery: results from a single centre. Colorectal Dis 2012; 14(10): 1191-6.

33. Chadwick MA, Vieten D, Pettitt E, Dixon AR, Roe AM. Short course preoperative radiotherapy is the single most important risk factor for perineal wound complications afer abdominoperineal excision of the rectum. Colorectal Dis 2006; 8(9): 756-61.

49 Chapter 2

34. Mazier WP, Surrell JA, Senagore AJ. Te bottom end. Handling of the perineal wound afer abdominoperineal resection. Am Surg 1991; 57(7): 454-8.

35. Wang JY, Huang CJ, Hsieh JS, Huang YS, Juang YF, Huang TJ. Management of the perineal wounds following excision of the rectum for malignancy. Gaoxiong Yi Xue Ke Xue Za Zhi 1994; 10(4): 177-81.

36. Zorcolo L, Restivo A, Capra F, Fantola G, Marongiu L, Casula G. Does long-course radiotherapy infuence postoperative perineal morbidity afer abdominoperineal resection of the rectum for cancer? Colorectal Dis 2011; 13(12): 1407-12.

37. de Broux E, Parc Y, Rondelli F, Dehni N, Tiret E, Parc R. Sutured perineal omentoplasty afer abdominoperineal resection for adenocarcinoma of the lower rectum. Dis Colon Rectum 2005; 48(3): 476-81.

38. Toshniwal S, Perera M, Lloyd D, Nguyen H. A 12-year experience of the Trendelenburg perineal approach for abdominoperineal resection. ANZ JSurg 2013.

39. Christensen HK, Nerstrom P, Tei T, Laurberg S. Perineal repair afer extralevator abdominoperineal excision for low rectal cancer. Dis Colon Rectum 2011; 54(6): 711-7.

40. Dormand EL, Banwell PE, Goodacre TE. Radiotherapy and wound healing. Int Wound J 2005; 2(2): 112-27.

41. Haubner F, Ohmann E, Pohl F, Strutz J, Gassner HG. Wound healing afer radiation therapy: review of the literature. Radiat Oncol 2012; 7: 162.

42. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus fap reconstruction of the pelvic foor for rectal cancer. Br J Surg 2007; 94(2): 232-8.

43. Houvenaeghel G, Ghouti L, Moutardier V, Buttarelli M, Lelong B, Delpero JR. Rectus abdominis myocutaneous fap in radical oncopelvic surgery: a safe and useful procedure. Eur J Surg Oncol 2005; 31(10): 1185-90.

44. Loessin SJ, Meland NB, Devine RM, Wolf BG, Nelson H, Zincke H. Management of sacral and perineal defects following abdominoperineal resection and radiation with transpelvic muscle faps. Dis Colon Rectum 1995; 38(9): 940-5.

45. Petrie N, Branagan G, McGuiness C, McGee S, Fuller C, Chave H. Reconstruction of the perineum following anorectal cancer excision. Int J Colorectal Dis 2009; 24(1): 97-104.

46. Foster JD, Pathak S, Smart NJ, et al. Reconstruction of the perineum following extralevator abdominoperineal excision for carcinoma of the lower rectum: a systematic review. Colorectal Dis 2012; 14(9): 1052-9.

47. Goldsmith HS, Grifth AL, Kupferman A, Catsimpoolas N. Lipid angiogenic factor from omentum. JAMA 1984; 252(15): 2034-6.

48. Walker F, Rogers A. Te greater omentum as a site of antibody synthesis. Br J Exp Pathol 1961; 42: 222-31.

49. Irvin TT, Goligher JC. A controlled clinical trial of three diferent methods of perineal wound management following excision of the rectum. Br J Surg 1975; 62(4): 287-91.

50. Terranova O, Sandei F, Rebufat C, Maruotti R, Pezzuoli G. Management of the perineal wound afer rectal excision for neoplastic disease: a controlled clinical trial. Dis Colon Rectum 1979; 22(4): 228-33.

50 Perineal wound healing afer abdominoperineal resection for rectal cancer

2

51

CHAPTER 3

Perineal wound healing afer abdominoperineal resection for rectal cancer; a two centre experience in the era of intensifed oncological treatment

G.D. Musters, D.A Sloothaak, S. Roodbeen, A.A.W. van Geloven, W.A. Bemelman, and P.J. Tanis

Int J Colorectal Dis 2014; 29(9): 1151-7 Chapter 3

ABSTRACT

Aim: Intensifed treatment for distal rectal cancer has improved oncological outcome, but at the expense of more perineal wound complications in patients undergoing an abdominoperineal resection (APR). Te aim of this study was to analyse perineal wound healing afer APR with primary perineal wound closure over time.

Methods: All patients undergoing APR for primary rectal cancer with primary wound closure between 2000 and 2013 were included and analysed in three consecutive time periods. Both early (<30 days postoperatively) and late perineal wound complications were determined. Independent risk factors of early perineal wound complications were identifed using multivariable analysis.

Results: In total 136 patients were identifed, of whom 129 patients underwent primary perineal wound closure. Te use of neoadjuvant (chemo)radiotherapy increased from 70% to 91% and the use of an extralevator approach increased from 9% to 19%. Te rate of early perineal wound complications increased from 18% to 31% and was independently associated with an extralevator approach (OR 3.17; 95% CI 1.16-8.66) and intra-operative perforation (OR 3.35; 95% CI 1.06-10.57). Perineal wound complications had no impact on local recurrence or 3-year overall survival rate. During a median follow-up of 28 months (IQR 14-56), a persistent presacral sinus was diagnosed in 10%, and a perineal hernia occurred in 8% of the patients.

Conclusion: Te increased use of an extralevator APR for rectal cancer signifcantly increased the risk of perineal wound complications over time. Intra-operative perforation was also an independently associated with impaired perineal wound healing.

54 Perineal wound healing afer abdominoperineal resection for rectal cancer

INTRODUCTION

Conventional abdominoperineal resection (cAPR) for rectal cancer has been associated with a relatively high rate of positive resection margins and intraoperative perforations.1,2 Tis is owing to a non-cylindrical resection without mesorectal covering of the muscular wall of the rectum at the level of the pelvic foor. Te suboptimal cAPR translated into a lower overall survival and higher local recurrence rate if compared to low anterior resection.3 As a consequence, the extralevator APR (eAPR) is increasingly used, which entails en bloc resection of the levator muscles covering the distal mesorectum. Systematic review of non-randomised studies showed a reduced rate of positive resection margins and local disease recurrence by performing an extralevator approach compared to cAPR.4,5 In addition, the increased use of neoadjuvant radiotherapy over the past three decades contributed to improved locoregional control rates.6

Although neoadjuvant therapy and wider excisions have improved the oncological outcome afer 3 APR, it has been suggested that this is at the expense of more perineal wound complications.7,8 However, literature on perineal wound healing afer APR is difcult to interpret. Mostly perineal wound healing was a secondary endpoint in primary oncological studies and several aspects of rectal cancer management have changed over time such as neoadjuvant therapy, perioperative care, and surgical techniques. To date, little is known about the impact of several changes in rectal cancer management on perineal wound complications. Terefore, the aim of the present study was to analyse perineal wound healing in relation to changes in treatment approaches over time for patients undergoing APR for distal rectal cancer in two diferent hospital settings.

PATIENTS AND METHODS

Patients All consecutive patients from one university hospital and one regional hospital who underwent APR for primary rectal cancer between March 2000 and April 2013 were retrospectively identifed. Patients were included in the present analysis when the perineal wound was primarily closed. A distinction was made between a conventional (cAPR) and extralevator (eAPR) approach. All APR proceducres were performed or supervised by colorectal surgeons. During the whole study period, standard treatment at the regional hospital was cAPR in supine position with selective use of an omentoplasty. In the university hospital, the extralevator approach was increasingly applied during the study period and performed as a routine in prone position since 2010. An intersphincteric APR was used at the university hospital since 2010 if MRI revealed at least 1 mm distance between the tumour and the levator muscle / external sphincter. An omentoplasty was performed as a routine if technically possible. Te indication for neoadjuvant radiotherapy was discussed in a pre-operative multidisciplinary team meeting and was based on the national rectal cancer guideline.

Data extraction Patient and treatment characteristics were retrospectively collected from patient records. Operative reports were searched for the description of the extent of resection (cAPR, eAPR or intersphincteric APR), the operative approach (open or laparoscopic), the use of an omentoplasty,

55 Chapter 3

and intraoperative complications (i.e. perforation). In addition, data were extracted on early and late perineal complications, other postoperative complications, hospital stay, surgical and non-surgical reintervention for perineal complications, local recurrence and survival. Early perineal wound complications were recorded as a presacral abscess, perineal wound infection or perineal wound dehiscence without signs of infection if these occurred within 30 days afer the surgical procedure. Late perineal complications such as perineal hernia, and a persistent presacral sinus were recorded until end of follow-up.

Statistical analysis Descriptive data were reported according to distribution as median with interquartile range (IQR) or mean ± standard deviation (SD). Categorical data were analysed with the Chi-square-test or Fisher’s exact test and continuous variables were analysed using the Mann-Whitney-Wilcoxon test. Te inclusion period was categorized in 3 groups (2000-2005, 2006-2009, and 2010-2013). Te relationship between tumour stage, neoadjuvant therapy and the date of operation was studied by Chi-square-test for trend. For this purpose, time of inclusion was categorized as ordinal data. All predictors identifed in the univariable analysis with a P-value of <0.2 were candidate variables for inclusion in a multivariable model. Multiple regression analysis was used to identify independent predictors of early perineal wound complications. Tree-year survival rates were determined using Kaplan Meier analysis and its association with perineal wound complications was analysed using a log rank test. Te signifcance level was set at P < 0.05. All analyses were performed with IBM SPSS statistics, version 20.0.0 (IBM Corp., Armonk, NY, United States).

RESULTS

Patient and surgical characteristics In the period 2000-2013, a total of 136 patients with primary rectal cancer underwent APR, of whom 51 patients were operated upon at the regional hospital and 85 patients at the university hospital. Seven of these 136 patients did not undergo primary perineal wound closure: musculocutaneous fap in one patient, biological mesh in two patients, delayed primary closure (2 days) in one patient, and in three patients the closure of the perineum was not specifed. Patient, treatment and tumour characteristics of 129 included patients for each time period are provided in Table 1. A signifcant increase in the number of APR procedures performed at the regional hospital was observed, with a signifcant decrease over time for the university hospital (p<0.01). A signifcant increase in age (p=0.018) and a non-signifcant trend towards an increase in BMI (p=0.381) were observed over time. Tere was an increased use of neoadjuvant (chemo)radiotherapy (p=0.038) with overall down staging over time, as shown in Figure 1. Laparoscopic surgery was performed in 54% (70/129) and signifcantly increased over time (p<0.001; Fig. 2). An omentoplasty was placed in the pelvic cavity in 37% (48/129) of the patients. An eAPR was performed in 9% (3/33) of the patients in the period between 2000 and 2005, in 28% (12/42) between 2006 and 2009, and in 19% (10/54) between 2010 and 2013. Resection margins were positive in 11% (14/129) of the patients and incompleteness of resection showed a non-signifcant decrease over time (p=0.161, Fig. 2).

56 Perineal wound healing afer abdominoperineal resection for rectal cancer

Table 1 Patient, treatment and tumour characteristics

Year 2000-2005 2006-2009 2010-2013 (n=33) (n=42) (n=54) Hospital Regional (n, %) 3 (9) 13 (31) 34 (63) University hospital (n, %) 30 (91) 29 (69) 20 (37) Sex Male (n, %) 21 (64) 28 (67) 37 (69) Age Years ± SD 59 ± 14 63 ± 16 67 ± 10 BMI BMI (IQR) 25 (21-30) 25 (24-29) 27 (24-30) ASA-classifcation ASA 1 (n, %) 11 (33) 11 (26) 10 (19) ASA 2 (n, %) 14 (42) 16 (38) 34 (63) ASA 3 (n, %) 7 (21) 4 (10) 6 (11) 3 Unknown (n, %) 1 (3) 11 (26) 4 (7) Comorbidity Diabetes (n, %) 2 (6) 6 (14) 5 (9) Respiratory (n, %) 2 (6) 2 (5) 1 (2) Cardiac (n, %) 5 (15) 8 (19) 15 (28) Radiotherapy for other malignancy (n, %) 2 (6) 2 (5) 4 (7) Preoperative imaging MRI (n, %) 19 (58) 39 (93) 49 (91) Radiotherapy Short course 5x5 Gy (n, %) 13 (39) 18 (43) 18 (33) Chemoradiotherapy (n, %) 3 (9) 17 (40) 31 (57) Long course without chemotherapy (n, %) 6 (18) 1 (2) 0 (0) Not specifed (n, %) 1 (3) 2 (5) 0 (0) Surgery Laparoscopic surgery (n, %) 4 (12) 23 (55) 43 (80) Prone position (n, %) 0 (0) 0 (0) 16 (30) Omentoplasty (n, %) 13 (39) 19 (45) 16 (30) Pelvic drain (n, %) 19 (58) 28 (67) 41 (76) Type of APR cAPR (n, %) 29 (88) 30 (71) 35 (65) eAPR (n, %) 3 (9) 12 (29) 10 (19) Intersphincteric APR (n, %) 1 (3) 0 (0) 9 (17) Hospital stay Days (median, IQR) 13 (9-19) 11 (8-14) 8 (6-16)

pTNM-stage T0-2 (n, %) 11 (33) 19 (45) 33 (61)

T3-4 (n, %) 22 (67) 23 (55) 21 (39)

N0 (n, %) 23 (70) 25 (60) 39 (72)

N+ (n, %) 10 (30) 16 (38) 15 (28)

Nx (n, %) 0 (0) 1 (2) 0 (0)

M0 (n, %) 24 (73) 40 (95) 52 (96)

M+ (n, %) 4 (12) 1 (2) 2 (4)

Mx (n, %) 5 (15) 1 (2) 0 (0) cAPR=conventional abdominoperineal resection eAPR= extralevator abdominoperineal resection

57 Chapter 3

Overall intra-operative perforation rate was 14% (18/129); 16% (15/94) afer cAPR, 8% (2/25) afer eAPR, and 10% (1/10) afer intersphincteric APR. Over time, intra-operative perforation occurred in 21% (7/33) of the patients between 2000 and 2005, in 10% (4/42) between 2006 and 2009, and in 13% (7/54) of the patients between 2010 and 2013 (p=0.336, Fig. 2). Perforation rate for prone and supine positioning was 6% (1/16) and 15% (17/113), respectively (p=0.468).

Figure 1 An overview of tumour stage and neoadjuvant therapy over time

Early postoperative perineal wound complications Te overall 30-day perineal wound complication rate was 29% (37/129); 18% (6/33) between 2000 and 2005, 33% (14/42) between 2006 and 2009, and 31% (17/54) between 2010 and 2013 (p=0.297; Fig. 2). Tese perineal wound complications consisted in 18% (23/129) of a perineal wound infection, in 6% (8/129) of a presacral abscess, and in 5% (6/129) of a perineal wound dehiscence without infection (Table 2). Perineal wound complications occurred in 48% (n=12/25) afer eAPR, 23% (22/94) afer cAPR, and 30% (3/10) afer intersphincteric APR. Signifcantly more perineal wound complications were observed afer eAPR compared to cAPR as shown in Table 3 (p=0.02). ASA- score, chemoradiotherapy, patient positioning and intra-operative perforation were also potential predictors for perineal wound complications in univariable analysis, while age, comorbidity, any kind of radiotherapy and the use of an omentoplasty did not reach a P-value below 0.2. Multivariable analysis revealed that intra-operative perforation and eAPR were both independently associated with an increased risk of perineal wound complications (Table 4). Patients with a perineal wound complication had a signifcantly longer median postoperative stay compared to patients with uncomplicated wound healing: 14 days (IQR 11-36) versus nine days (IQR 6-14) (p<0.0001).

58 Perineal wound healing afer abdominoperineal resection for rectal cancer

Table 2 Impaired perineal wound healing

Year 2000-2005 2006-2009 2010-2013 (n=33) (n=42) (n=54) Early (<30 days) Perineal wound infection (n, %) 5 (15) 7 (17) 11 (20) Perineal or presacral abscess (n, %) 1 (3) 2 (5) 5 (9) Perineal wound dehiscence (n, %) 0 (0) 4 (10) 2 (4) Late Persistent presacral sinus (n, %) 3 (9) 4 (10) 6 (11) Perineal hernia (n, %) 2 (6) 6 (14) 3 (6) Re-intervention for perineal Percutaneous (n, %) 1 (3) 3 (7) 4 (7) complication Surgical (n, %) 2 (6) 1 (2) 3 (6) 3 Table 3 Univariable analysis of possible predictors of perineal wound complications

OR 95% CI P-value Age Per 10 years 1.14 0.85-1.53 0.389 ASA-classifcation (vs.1) Score 2 2.46 0.83-7.31 0.107 Score 3 3.78 0.97-14.70 0.055 Comorbidity Diabetes 1.62 0.49-5.33 0.426 Pulmonary 0.60 0.07-5.59 0.657 Cardiac 0.98 0.39-2.47 0.965 BMI >30 1.00 0.30-3.34 1.000 Neoadjuvant therapy Any type of radiotherapy 1.05 0.35-3.20 0.927 APR (vs. cAPR) eAPR 3.02 1.21-7.57 0.018 Intersphincteric APR 1.40 0.33-5.89 0.644 Surgery Laparoscopic 0.63 0.29-1.35 0.231 Perineal phase frst 1.79 0.59-543 0.307 Pelvic drain 1.00 0.99-1.01 0.463 Perforation 2.96 1.07-8.21 0.037 Omentoplasty 1.22 0.56-2.67 0.620

OR=odds ratio, eAPR=extralevator abdominoperineal resection.

Table 4 Multivariable logistic regression analysis of 30-day perineal wound complications afer APR with primary perineal wound closure

OR 95% CI P-value ASA-classifcation (vs.1) Score 2 2.14 0.70-6.59 0.184 Score 3 2.77 0.67-11.54 0.160 Type of APR procedure eAPR 3.17 1.16-8.66 0.024 Intra-operative perforation 3.35 1.06-10.57 0.039 eAPR=extralevator abdominoperineal resection. OR= odds Ratio.

59 Chapter 3

Follow-up Te median follow-up was 28 months (IQR 14-56). During follow-up, 28 patients have died, of which three patients died postoperatively because of myocardial infarction (n=1) and respiratory insufciency (n=2). Of the 126 evaluable patients, in 10% (13/126) a persistent presacral sinus was diagnosed. A perineal hernia occurred in 8% (11/126) of the patients. Te median time between the operation and the development of a perineal hernia was 17 months (IQR 5-42). A clinically diagnosed perineal hernia occurred in 8% (7/92) afer cAPR, in 13% (3/24) afer eAPR, and in 10% (1/10) afer intersphincteric APR (p=0.74). All clinical perineal hernias occurred afer neoadjuvant (chemo)radiotherapy and occurred in 6% (2/32) between 2000 and 2005, in 14% (6/42) between 2006 and 2009, and in 6% (3/52) between 2010 and 2013 (p=0.294).

A local recurrence occurred in 12 patients afer a median of 19 months (IQR 5-25). Local recurrence rate was similar (p=0.773) for patients with and without perineal wound complications: 8% (3/36) versus 10% (9/90), respectively. Te 3-year overall survival rate was 75% and was not signifcantly associated with perineal wound complications (3-year OS 73% vs. 82%; p=0.286).

Figure 2 An overview of diferent operation techniques and perineal wound complications over time

60 Perineal wound healing afer abdominoperineal resection for rectal cancer

DISCUSSION

In this two-institutional experience with APR for rectal cancer, the rate of perineal wound complications increased from 18% to 31% over time, and an extralevator approach and intra- operative perforation appeared to be independent risk factors. Tese wound complications may also result in a persistent presacral sinus in a few percent of the patients. Another late perineal complication afer APR is perineal hernia, which occurred in 8% in this cohort of patients.

To our knowledge, the association between intra-operative perforation and complicated perineal wound healing has not been described in literature before. Despite this theoretically obvious association because of spill of bowel contents, the relatively low incidence of intra-operative perforation may explain the difculty to use it as a risk factor in multivariable analysis. Despite the increased use of eAPR in the university hospital, the perforation rate remained relatively high because of the increasing numbers of cAPR performed at the regional hospital over time. Te 3 diferences in operative approaches between the two institutes enabled us to analyse the association between perineal wound complications and technical aspects of the APR procedure. Te extralevator approach in prone position has been recently implemented in the regional hospital, proctored by the university hospital. Although this is likely to reduce the risk of intra-operative perforation, perineal wound complications are expected to remain at a similar level because of the extralevator approach.

Radiotherapy was not identifed as a risk factor for perineal wound complications in the present study, in contrast to several other studies.9 Te proportion of patients undergoing neoadjuvant radiotherapy was relatively high throughout the study period, which is in line with Dutch practice, as found in a population based study.10 Although radiotherapy was likely to have contributed to the observed perineal wound complications, it’s almost routine use did not allow for a meaningful comparison between patients with and without neoadjuvant radiotherapy. Neoadjuvant (chemo) radiotherapy may result in fbrosis of the mesorectal fat and rectal wall stifness.11,12 Tis might complicate the resection and increase the risk of an intra-operative perforation. Besides a suboptimal surgical technique, frequent use of radiotherapy might have resulted in the relatively high intra- operative perforation rate.

Te higher perineal wound complication rate using an extralevator approach might be a result of the increased closure under tension of the remaining perineal structures following wider excisions, and the increased pressure of the abdominal content on the perineal wound while standing in the absence of a pelvic foor. Perineal wound complications are a signifcant clinical problem and increased hospital stay with a median of fve days. Furthermore, these patients ofen need intensive wound care afer hospital discharge or need to be admitted to a nursing home for a certain period of time. It may even result in a chronic perineal fstula, compromising quality of life. Tis stresses the importance of fnding efective measures to prevent perineal wound complications afer APR. One of these measures might be the use of vessel sealing equipment, which might reduce perineal wound complication because it results in a better haemostasis and therefore might reduce wound complications.13 Another potentially efective measure, although not supported by our results, is the placement of an omentoplasty in the presacral cavity which may positively infuence primary perineal wound healing. An omentoplasty has an efect on the local immune response, promotes

61 Chapter 3

angiogenesis and may prevent abscess formation in the pre-sacral dead space.14,15 Besides flling of the pelvic cavity with an omentoplasty, pelvic foor reconstruction using a myocutaneous or fasciocutaneous fap may add to better perineal wound healing.16,17 However, the mucotaneous faps are associated with a risk of fap necrosis and donor site morbidity, and therefore questions their routine applicability.18,19 Te use of a biological mesh for pelvic foor reconstruction may also serve as a valid alternative and seems to be equally efective as myocutaneous faps.20 However, Jensen et al. recently reported an overall 55% rate of infectious and non-infectious perineal wound complications afer biological mesh assisted closure.21 A higher level of evidence is needed to justify the routine use of a biological mesh for perineal closure afer eAPR.

With the resection of the pelvic foor and only closure of the subcutaneous fatty tissue and skin, a perineal hernia is more likely to occur. Nevertheless, we observed only a 13% perineal hernia rate for eAPR with primary perineal wound closure. Reported incidences in literature are even lower, ranging from 1% to 13% afer cAPR7,8,22,23 and from 1% to 6% afer eAPR.14,24,25 But these studies date back to 2005 at which time fewer patients were treated with neoadjuvant radiotherapy. Radiotherapy afects in the early phase pro-infammatory cytokines which may result in an infammation that does not resolve adequately, leading to inadequate fbrotic matrix accumulation.26 In addition, metalloproteinase and nitric oxide are reduced which may contribute to inadequate sof tissue reconstruction.27 Other explanations for the relatively low perineal hernia rate in literature may be underreporting by patients of asymptomatic hernia, retrospective study design and the primary focus on oncological outcome in most rectal cancer studies.

Limiting factors of this study are its retrospective design and the unstructured operation reports, which may have resulted in the omission of data. In addition, a non-standardised defnition of perineal wound infection was used between the surgeons and hospitals, which might have infuenced the results. Despite these limitations, the present study clearly illustrates the increasing clinical problem of early and late perineal wound complications afer APR for rectal cancer over time, with an extralevator approach and intraoperative perforation as independent risk factors for complicated wound healing.

62 Perineal wound healing afer abdominoperineal resection for rectal cancer

REFERENCE LIST

1. Eriksen MT, Wibe A, Syse A, Hafner J, Wiig JN. Inadvertent perforation during rectal cancer resection in Norway. Br J Surg 2004; 91(2): 210-216.

2. Nagtegaal ID, van de Velde CJ, Marijnen CA, van Krieken JH, Quirke P. Low rectal cancer: a call for a change of approach in abdominoperineal resection. J Clin Oncol 2005; 23(36): 9257-9264.

3. den Dulk M, Putter H, Collette L et al. Te abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of fve European randomised clinical trials on rectal cancer. Eur J Cancer 2009; 45(7): 1175-1183.

4. Stelzner S, Koehler C, Stelzer J, Sims A, Witzigmann H. Extended abdominoperineal excision vs. standard abdominoperineal excision in rectal cancer--a systematic overview. Int J Colorectal Dis 2011; 26(10): 1227-1240. 5. West NP, Finan PJ, Anderin C et al. Evidence of the oncologic superiority of cylindrical abdominoperineal 3 excision for low rectal cancer. J Clin Oncol 2008; 26(21): 3517-3522.

6. Fleming FJ, Pahlman L, Monson JR. Neoadjuvant therapy in rectal cancer. Dis Colon Rectum 2011; 54(7): 901-912.

7. Bullard KM, Trudel JL, Baxter NN, Rothenberger DA. Primary perineal wound closure afer preoperative radiotherapy and abdominoperineal resection has a high incidence of wound failure. Dis Colon Rectum 2005; 48(3): 438-443.

8. West NP, Anderin C, Smith KJ, Holm T, Quirke P. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010; 97(4): 588-599.

9. Musters GD, Buskens CJ, Bemelman WA, Tanis PJ. Perineal wound healing afer abdominoperineal resection for rectal cancer; a systematic review and meta-analysis. Dis Colon Rectum 2014; 57(9):1129-39.

10. van Leersum NJ, Snijders HS, Wouters MW et al. Evaluating national practice of preoperative radiotherapy for rectal cancer based on clinical auditing. Eur J Surg Oncol 2013; 39(9): 1000-1006.

11. Krol R, Hopman WP, Smeenk RJ, van Lin EN. Increased rectal wall stifness afer prostate radiotherapy: relation with fecal urgency. Neurogastroenterol Motil 2012; 24(4): 339-e166.

12. O’Brien PC. Radiation injury of the rectum. Radiother Oncol 2001; 60(1): 1-14.

13. Nanashima A, Arai J, Oyama S et al. Associated factors with surgical site infections afer hepatectomy: predictions and countermeasures by a retrospective cohort study. Int J Surg 2014; 12(4): 310-314.

14. De Broux E, Parc Y, Rondelli F et al. Sutured perineal omentoplasty afer abdominoperineal resection for adenocarcinoma of the lower rectum. Dis Colon Rectum 2005; 48(3): 476-481.

15. Hultman CS, Carlson GW, Losken A et al. Utility of the omentum in the reconstruction of complex extraperitoneal wounds and defects: donor-site complications in 135 patients from 1975 to 2000. Ann Surg 2002; 235(6): 782-795.

16. Howell AM, Jarral OA, Faiz O et al. How should perineal wounds be closed following abdominoperineal resection in patients post radiotherapy--primary closure or fap repair? Best evidence topic (BET). Int J Surg 2013; 11(7): 514-517.

17. Petrie N, Branagan G, McGuiness C et al. Reconstruction of the perineum following anorectal cancer excision. Int J Colorectal Dis 2009; 24(1): 97-104.

18. Bakx R, van Lanschot JJ, Zoetmulder FA. Inferiorly based rectus abdominis myocutaneous faps in surgical oncology: Indications, technique, and experience in 37 patients. J Surg Oncol 2004; 85(2): 93-97.

63 Chapter 3

19. Loessin SJ, Meland NB, Devine RM et al. Management of sacral and perineal defects following abdominoperineal resection and radiation with transpelvic muscle faps. Dis Colon Rectum 1995; 38(9): 940-945.

20. Foster JD, Pathak S, Smart NJ et al. Reconstruction of the perineum following extralevator abdominoperineal excision for carcinoma of the lower rectum: a systematic review. Colorectal Dis 2012; 14(9): 1052-1059.

21. Jensen KK, Rashid L, Pilsgaard B, Molle P, Wille-Jorgensen P. Pelvic foor reconstruction with a biological mesh afer extralevator abdominoperineal excision leads to low wound complication and perineal hernia rates with minor movement limitations. Colorectal Dis 2013; 16(3): 192-7.

22. El-Gazzaz G, Kiran RP, Lavery I. Wound complications in rectal cancer patients undergoing primary closure of the perineal wound afer abdominoperineal resection. Dis Colon Rectum 2009; 52(12): 1962-1966.

23. Han JG, Wang ZJ, Wei GH et al. Randomised clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. Am J Surg 2012; 204(3):274-282.

24. Bebenek M. Abdominosacral resection is not related to the risk of neurological complications in patients with low-rectal cancer. Colorectal Dis 2009; 11(4): 373-376.

25. Toshniwal S, Perera M, Lloyd D, Nguyen H. A 12-year experience of the Trendelenburg perineal approach for abdominoperineal resection. ANZ J Surg 2013; 83(11): 853-8.

26. Dormand EL, Banwell PE, Goodacre TE. Radiotherapy and wound healing. Int Wound J 2005; 2(2): 112-127.

27. Haubner F, Ohmann E, Pohl F, Strutz J, Gassner HG. Wound healing afer radiation therapy: review of the literature. Radiat Oncol 2012; 7:162.

64 Perineal wound healing afer abdominoperineal resection for rectal cancer

3

65

CHAPTER 4

Randomised controlled multicentre study comparing biological mesh closure of the pelvic foor with primary perineal closure afer extralevator abdominoperineal resection for rectal cancer (BIOPEX-study)

G.D. Musters, C.E.L. Klaver, R.J.I. Bosker, J.W.A. Burger, P. van Duijvendijk, B. van Etten, A.A.W. van Geloven, E.J.R. de Graaf, C. Hof, J.W.A. Leijtens, H.J.T. Rutten, B. Singh, R.J.C.L.M. Vuylsteke, J.H.W. de Wilt, M.G.W. Dijkgraaf, W.A. Bemelman, and P.J. Tanis

BMC Surg 2014; 27: 58 Chapter 4

ABSTRACT

Aim: Primary perineal wound closure afer conventional abdominoperineal resection (cAPR) for rectal cancer has been the standard of care for many years. Since the introduction of neoadjuvant radiotherapy and the extralevator APR (eAPR), oncological outcome has been improved, but at the cost of increased rates of perineal wound healing problems and perineal hernia. Tis has progressively increased the use of biological meshes, although not supported by sufcient evidence. Te aim of this study is to determine the efectiveness of pelvic foor reconstruction using a biological mesh afer standardized eAPR with neoadjuvant (chemo)radiotherapy compared to primary perineal wound closure.

Method: In this multicentre randomised controlled trial, patients with a clinical diagnosis of primary rectal cancer who are scheduled for eAPR afer neoadjuvant (chemo)radiotherapy will be considered eligible. Exclusion criteria are prior radiotherapy, sacral resection above S4/S5, allergy to pig products or polysorbate, collagen disorders, and severe systemic diseases afecting wound healing, except for diabetes. Afer informed consent, 104 patients will be randomised between standard care using primary wound closure of the perineum and the experimental arm consisting of suturing a biological mesh derived from porcine dermis in the pelvic foor defect, followed by perineal closure similar to the control arm. Patients will be followed for one year afer the intervention and outcome assessors and patients will be blinded for the study treatment. Te primary endpoint is the percentage of uncomplicated perineal wound healing, defned as a Southampton wound score of less than II on day 30. Secondary endpoints are hospital stay, incidence of perineal hernia, quality of life, and costs.

Conclusion: Te BIOPEX-study is the frst randomised controlled multicentre study to determine the additive value of using a biological mesh for perineal wound closure afer eAPR with neoadjuvant radiotherapy compared to primary perineal wound closure with regard to perineal wound healing and the occurrence of perineal hernia.

68 BIOPEX-study protocol

INTRODUCTION

Conventional abdominoperineal resection (cAPR) for distal rectal cancer is associated with relatively high rates of positive resection margins and iatrogenic tumour perforation. Pooled analysis of fve European randomised controlled trials revealed that cAPR was an independent risk factor for local recurrence (19.7% versus 11.4%) and associated with a decreased 5-year survival rate (59% versus 70%) compared to low anterior resection.1 Following the total mesorectal excision (TME) plane all the way down to the pelvic foor, as performed in cAPR, results in the typical coning of the specimen and thereby compromising resection margins. Extralevator APR (eAPR) reduces the rate of positive resection margins and tumour perforation and improves oncological outcome based on a systematic review of non-randomised studies.2 Te extralevator approach entails en bloc resection of the distal rectum, sphincter complex and levator muscles. Besides a change in surgical technique, neoadjuvant radiotherapy has also contributed to better local regional control in patients with distal rectal cancer.

Both wider excisions as a result of eAPR as well as increased use of neoadjuvant (chemo)radiotherapy have signifcantly increased perineal wound healing problems, which have been reported in up to 59%.3-6 Furthermore, perineal hernia is more likely to occur with a reported incidence of up to 20%.7 4 Perineal wound complications may only consist of superfcial infection or minor dehiscence, but are ofen more serious with major dehiscence or deep abscesses. Impaired perineal wound healing is a signifcant clinical problem being associated with increased hospital stay, the need for re-operation, and intensive wound care for several months. Tere is even a small proportion of patients in whom a presacral or perineal sinus remains. Quality of life may be severely impaired because of pain, smelly wound discharge, and interference with basic daily activities such as sitting and walking.

Internationally there is a trend towards the use of additional closure techniques afer eAPR in an attempt to overcome this clinical problem. One of these techniques is the reconstruction of the pelvic foor with a biological mesh, but available literature data are scarce. A recently published review included 12 studies, of which six were only conference abstracts, with total numbers of included patients between 2 and 24.8 Te percentage of minor wound complications in series with at least 5 patients ranged between 17% and 63%, and of major complications between 9% and 27%. Two larger studies not included in this review reported a perineal wound complication rate of 50% and 11% in 30 and 35 patients respectively.9,10 Autologous tissue faps might also be an option and seems to be equally efective as a biological mesh.11 However, the routine use of faps afer eAPR for rectal cancer is controversial. Te rectus abdominis muscle (RAM) fap cannot be combined with a laparoscopic approach, will add donor site morbidity, substantially increases operating time and ofen implicates assistance by a plastic surgeon.12 Te gluteal fap does not disturb abdominal wall integrity, but seems not to prevent perineal hernia formation.7

Te popularity of the use of biological meshes for eAPR is probably the result of its successful application in repair of complex and infected hernias of the abdominal wall as well as for example pelvic organ prolapses.13,14 However, the costs of a biological mesh ranges between 500 and 1800 euro per patient depending on the size of acellular porcine collagen implant and company. Based on the limited available data and costs of biological meshes, primary perineal wound closure afer eAPR remains the standard of care in many institutes. An unpublished survey among Dutch

69 Chapter 4

surgeons performed in 2012 revealed that 66% exclusively used primary perineal wound closure afer APR, although the extralevator approach is not yet completely implemented at a national level. Using simple direct closure of the perineal wound afer eAPR is supported by data from relatively large cohort series (104 to 282 patients), which reported perineal wound complication rates between 17% and 35%, with a proportion of patients who received radiotherapy between 31% and 60%.5,15-17 However, standardized perineal wound assessment was not used in any of these studies and complication rates may be underestimated because perineal wound healing was not a primary outcome parameter. Tis underlines the need for a randomised comparison of diferent perineal closure techniques afer eAPR.

METHODS

Objective Te primary objective of this study is to determine the efectiveness of pelvic foor reconstruction using a biological mesh afer eAPR with neoadjuvant radiotherapy compared to primary perineal closure. It is hypothesized that biological mesh closure is superior to primary perineal closure in a cost-efective manner in terms of a higher primary perineal wound healing rate and a lower perineal hernia rate at reduced costs.

Design Tis is a multicentre, single blinded, randomised controlled trial, in which eligible patients will be randomised between pelvic foor reconstruction using a biological mesh (intervention arm) and primary closure of the perineal defect (standard arm) in a 1:1 ratio (Figure 1). Randomisation will be performed pre-operatively by a central automated randomisation using the trial website, with stratifcation for age, sex and laparoscopic surgery. Te allocation of treatment is blinded to the patient.

Te trial will be conducted in fve academic centres and eight teaching hospitals, including two national referral centres for locally advanced rectal cancer. Tis is a multicentre, single blinded, randomised controlled trial, in which eligible patients will be randomised between pelvic foor reconstruction using a biological mesh (intervention arm) and primary closure of the perineal defect (standard arm) in a 1:1 ratio (Figure 1). Randomisation will be performed pre-operatively by a central automated randomisation using the trial website, with stratifcation for age, sex and laparoscopic surgery. Te allocation of treatment is blinded to the patient.

Te trial will be conducted in fve academic centres and eight teaching hospitals, including two national referral centres for locally advanced rectal cancer. Participating centres will perform standardized perineal dissection according to the extralevator approach. Te procedures in the experimental arm in centres without experience in biological mesh reconstruction of the pelvic foor will be supervised by one of the principal investigators (P.T. and W.B.) until a standardized biological mesh placement is ensured. Furthermore, a preceding investigators’ meeting will be organized at the start of the trial and a refreshment course during the inclusion period, both

70 BIOPEX-study protocol

including a workshop using fresh frozen cadavers. Te perineal wound healing will be evaluated at 7 and 30 days postoperatively using the Southampton wound scoring system by an independent nurse or physician not aware of the intervention to which the patient was allocated.18 In addition, full colour photographs will be taken from the perineal wound and assessed by an expert panel (P.T. and W.B.) blinded for treatment allocation. During routine outpatient clinic visits for oncological follow-up at 3, 6, 9 and 12 months, the perineal wound will be inspected and scored accordingly with respect to healing and herniation (see Figure 1). In addition, CT-scan of the pelvis as usually performed during oncological follow-up at 1 year, will be reviewed with respect to presacral or perineal sinuses and perineal herniation. Quality of Life questionnaires will be administered to the patient at each follow-up interval. Patients without primary perineal wound healing who are treated in a nonhospital setting will get a wound diary in which the number of wound dressing changes, the number of visits by a home care nurse if applicable, and the wound materials used are registered per week until fnal closure of the perineal wound or end of study period. In addition, the nature and severity of any wound event, all medical or surgical interventions and readmissions will be collected. 4

Figure 1 Flow-diagram eAPR: abdominoperineal resection

Study population Patients are eligible for this study when they meet the following inclusion criteria: age of 18 years or older, planned eAPR for primary rectal cancer, pre-operative (chemo)radiotherapy, life expectancy of

71 Chapter 4

more than 2 years, ability to return for all scheduled and required study visits, and written informed consent for study participation. Exclusion criteria are: previous pelvic irradiation for other cancers (i.e. prostate cancer), total pelvic exenteration or sacral resection above level S4/S5, sensitivity to porcine derived products or polysorbate, severe systemic diseases afecting wound healing except diabetes (i.e. renal failure requiring dialysis, liver cirrhosis, and immune compromised status like HIV), collagen disorders (i.e. Marfan), enrolment in trials with overlapping primary endpoint or otherwise expected infuence on wound healing (i.e. systemic therapy like anti-angiogenic agents). An interim review will be performed at 50 included patients (of the total of 104 patients). At 6 weeks afer inclusion of these patients, the trial’s safety data will be evaluated. Te steering committee will be supplied with the number of (serious) adverse events in both groups at this time point. If there is a skewed distribution of the number of (serious) adverse events between the two groups, an efcacy analysis can be performed at the discretion of the steering committee. Following these interim analyses, the steering committee will advise upon continuation of the trial.

Intervention strategies Te perineal phase of the APR will be performed according to the principles of an extralevator approach, which means that the levator muscles will be laterally transected in order to leave a muscular cuf around the tumour. Te coccyx will not be routinely resected, but only if indicated based on surgical exposure or oncological principles. Te extent of excision of perineal skin and ischioanal fat will be as limited as oncological justifed. Te eAPR specimens will be classifed according to Phil Quirke and photographed in a standardized fashion. Patient positioning (prone or supine), the surgical approach for the abdominal phase (open or laparoscopic), and the use of an omentoplasty are lef to the discretion of the operating surgeon, because there is no conclusive evidence on which of these technical aspects is the most optimal with regard to perineal wound healing.

Closure of the perineum in the control arm consists of stitching the ischioanal and subcutaneous fat using interrupted Vicryl sutures in one or two layers. Subsequently, the skin will be closed using interrupted sutures according to the preference of the surgeon. Placement of a transabdominal or transperineal drain will be performed according to local protocols.

Te intervention in the experimental arm consists of suturing an acellular biological mesh derived from porcine dermis in the pelvic foor defect (Strattice™, 6x10 cm). Te mesh will be sutured at each side of the coccyx or distal sacrum with Prolene or Polydioxanone sutures (PDS). Laterally, the mesh is attached to the remaining of the levator complex and, anteriorly, to the transverse perineal muscle. A suction drain will be inserted and positioned on top of the mesh. Te perineal subcutaneous fat and skin will be subsequently closed in layers similar to primary perineal closure as performed in the standard arm.

Outcome parameters Primary endpoint is the percentage of uncomplicated perineal wound healing defned as a Southampton wound score less than II at 30 days postoperatively. Secondary endpoints of this study are the need for wound care, either in hospital or out of hospital and specifed to frequency, duration, and type of wound dressing, the need for surgical re-intervention or re-admission for perineal

72 BIOPEX-study protocol

wound problems, total hospital stay during one year of follow-up, incidence of asymptomatic and symptomatic perineal hernia, quality of life (EQ-5D-5L, EORTC-30, EORTC-CR29, SF-36), and (in)direct medical and non-medical costs (clinical report forms, hospital information systems, modifed health and labour questionnaire). In addition, perineal wound healing according to the Southampton wound grading will be repeated at 3, 6, 9 and 12 months postoperatively in order to determine diferences in wound healing between the study arms over time. Diferences in the incidence of persistent perineal or presacral sinuses, both clinically and by CT imaging, will be evaluated at 1 year of follow-up.

Sample size calculation Te principal analysis will consist of an intention-to-treat comparison of the proportions of patients with primary uncomplicated perineal wound healing between both study arms. Te hypothesis is to test superiority of the biological mesh assisted closure over primary closure of the perineum. Te currently available literature is difcult to interpret with regard to perineal wound healing for the two interventions to be studied, hampering defning the expected diference in healing rate. Data on primary closure are mostly derived from cohort series in which patients were mostly operated upon 4 before 2005.3-7,10,19-23 At that time, eAPR was only applied in a minority of centres. Furthermore, there has been an enormous increase in the use of neoadjuvant radiotherapy in the past decade. Tese two developments in the treatment of distal rectal cancer have had a signifcant impact on perineal healing rate. In contrast, literature on biological mesh assisted perineal closure has been published recently only in patients undergoing eAPR and with high rates of neoadjuvant radiotherapy.7,9,10,20-23 In addition, these studies had a specifc focus on perineal wound healing, which results in more detailed data with a relative overreporting of perineal wound complications compared to the historical data on primary closure in which perineal healing was most ofen a secondary endpoint that was retrospectively studied. Both the low percentage of eAPR in the primary closure group and relative over reporting of perineal wound complications in the biological mesh led us to conclude that the available literature may underestimate the potential diference between both closure techniques.

Given the before mentioned lack of high quality data in the literature, we defned a clinically relevant diference in primary uncomplicated perineal healing which justifes the routine use of an expensive biological implant. Tis diference is considered to be 25%. A total number of 98 patients (49 per group) are needed to be able to detect a 25% increase in primary perineal wound healing by insertion of a biological mesh from 60% to 85%, applying a Chi2-test with a two-sided 0.05 signifcance level and with 80% power. With an estimated drop-out of 5%, a total number of 104 patients are required (52 per group).

Data-analysis Te primary endpoint, the percentage of uncomplicated wound healing defned as a Southampton wound score of less than II at 30 days postoperatively, will be compared between the two study groups using a two-sided Chi-square test with a signifcance level of 0.05. Te primary endpoint will be further explored by comparing wound scores as categorical variable using the Mann Whitney U test. Furthermore, diferences in time to healing between the two groups will be analysed as

73 Chapter 4

a censored continuous variable using Kaplan-Meier survival analysis. Treatment efects will be expressed as a relative risk with 95% confdence interval. Any binary secondary outcome measures (e.g. hernia rate, re-operation rate, infection rates, etc.) will be analysed in the same way as the primary outcome. Quality of life data (e.g. EQ-5D-5L, EORTC-30, EORTC-CR29, SF-36) will be graphically represented across all time points and analysed using a repeated measures analysis of variance. Te scoring profles from a patient on subsequent EQ-5D-5L assessments up to month 12 will further be used to derive health utilities based on the most recent crosswalk value sets at the time of analysis.24 Te health utilities will then be used to calculate quality adjusted life years (QALY) as the product sum of the utilities and the lengths of time preceding the repeated measurements.

Te most recent Dutch guideline for costing in health care research will be applied to assign unit costs to observe health care resources in order to estimate the costs of care for each study group.25 Costs diferences will be ofset against the diference in numbers of patients with uncomplicated perineal wound healing. Cost diferences will also be ofset against the diference in quality adjusted life years to calculate the extra costs per QALY gained. Both economic analyses will be performed from a societal perspective, with a time horizon set at 12 months. In addition, the analyses will be repeated from a hospital care provider perspective to improve the study’s generalizability to countries with other health care systems. No discounting of efects and costs will be performed. Sensitivity analyses will be performed to account for sampling variability and key study parameters (unit costs of the closure procedure using the biological mesh, alternative utility weights of observed health states). Results will be displayed graphically with cost-efectiveness planes and acceptability curves.

Ethics and Safety Te medical ethical committee of the Academic Medical Centre, Amsterdam, the Netherlands, has approved the study protocol (MEC 2012-360, NL42094.018.12). Tis study will be conducted according to the principles of Good Clinical Practice.

DISCUSSION

Te BIOPEX study is the frst randomised controlled trial comparing a new perineal closure technique afer eAPR with primary perineal closure. Tere is one other registered ongoing trial on perineal closure afer eAPR, but not including a study arm consisting of what we consider to be the standard of care, namely primary wound closure. Tis is a Swedish study randomizing between gluteus maximus myocutaneous fap and acellular porcine collagen implant (NEAPE; trials.gov identifer NCT01347697). Pelvic foor reconstruction using a musculocutaneous fap such as the RAM fap is ofen indicated in large irradiated perineal defects, for example afer salvage surgery for residual or recurrent anal cancer.26-28 But routinely using a musculocutaneous fap for perineal reconstruction afer APR for rectal cancer, except for exenterative procedures, seems to be over- treatment.12 Te gluteal fap as used in the Swedish trial does not disturb abdominal wall integrity, but seems not so attractive by signifcantly increasing surgical trauma in the irradiated perineum with postoperative mobilisation problems. If the gluteal fap is used as a fasciocutaneous fap, it is

74 BIOPEX-study protocol

associated with an unacceptably high rate of perineal hernia development according to Christensen et al..7 Te biological mesh has been recently introduced and appears to be a valid alternative for the autologous tissue faps.7,9,10,19-23 Since the perineal wound is by defnition contaminated, a biological mesh, being more resistant to infection, is preferred over less expensive synthetic meshes. Clinical outcome of a biological mesh appeared to be comparable to fap assisted perineal closure in a non- randomised comparison based on systematic review of the literature.11 Furthermore, a biological mesh is associated with lower costs per patient compared to a RAM fap.21 Others have suggested an omentoplasty for perineal reconstruction, but this functions as flling of dead space rather than giving strength to a neo-pelvic foor. A biological mesh difers essentially from a tissue fap by not flling the pelvic defect. Combining an omentoplasty with a biological mesh may therefore be of additive value. Although a high level of evidence is lacking, an omentoplasty is recommended afer APR, because it adds well vascularised, non-irradiated tissue to the pelvic cavity.16 Omentoplasty is not routine care in some of the participating centres of the BIOPEX study, which is the reason that we decided not to include this into the study protocol as a required part of the reconstruction.

Te costs associated with routinely adding a biological mesh to the surgical procedure is supposed to ultimately translate into cost saving by reducing hospital stay, reducing the need for operative 4 reintervention, reducing the need for intensive wound care and reducing costs related to secondary repair of perineal hernia. Tere is even a subgroup of patients with a persisting perineal sinus afer one year that is unlikely to heal eventually.29 Non-healing at one year afer pre-operative radiotherapy has been reported in up to 26% and may also be reduced by biological mesh assisted closure.19 Te percentage of surgical reintervention for perineal complications is reported between 8% and 16% and is expected to be halved from 10% to 5% by the use of a biological mesh.3,19,26 Furthermore, it is expected that perineal hernia incidence will be reduced by at least 10%, based on a diference of 20% as recently published by Christensen et al..7 Although this was a comparison between gluteal fap and biological mesh, gluteal fap is similar to primary closure regarding risk of perineal hernia because it does not reconstruct the pelvic foor. Primary use of a biological mesh is therefore expected to reduce by more than half the costs related to secondary perineal hernia repair requiring hospital admittance for about three days with operative intervention.

In conclusion, pelvic foor reconstruction using biological meshes has been suggested to improve perineal wound healing and may potentially save net health care costs compared to simple primary closure of the perineum afer eAPR, but proof of which is urgently needed.

75 Chapter 4

REFERENCE LIST

1. den Dulk M, Putter H, Collette L, et al. Te abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of fve European randomised clinical trials on rectal cancer. Eur J Cancer 2009; 45(7): 1175-83.

2. Stelzner S, Koehler C, Stelzer J, Sims A, Witzigmann H. Extended abdominoperineal excision vs. standard abdominoperineal excision in rectal cancer--a systematic overview. Int J Colorectal Dis 2011; 26 (10): 1227-40.

3. Bullard KM, Trudel JL, Baxter NN, Rothenberger DA. Primary perineal wound closure afer preoperative radiotherapy and abdominoperineal resection has a high incidence of wound failure. Dis Colon Rectum 2005; 48(3): 438-43.

4. El-Gazzaz G, Kiran RP, Lavery I. Wound complications in rectal cancer patients undergoing primary closure of the perineal wound afer abdominoperineal resection. Dis Colon Rectum 2009; 52(12): 1962-6.

5. Nissan A, Guillem JG, Paty PB, et al. Abdominoperineal resection for rectal cancer at a specialty center. Dis Colon Rectum 2001; 44(1): 27-35.

6. West NP, Anderin C, Smith KJ, Holm T, Quirke P. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010; 97(4): 588-99.

7. Christensen HK, Nerstrom P, Tei T, Laurberg S. Perineal repair afer extralevator abdominoperineal excision for low rectal cancer. Dis Colon Rectum 2011; 54(6): 711-7.

8. Butt HZ, Salem MK, Vijaynagar B, Chaudhri S, Singh B. Perineal reconstruction afer extra-levator abdominoperineal excision (eLAPE): a systematic review. Int J Colorectal Dis 2013; 28(11): 1459-68.

9. Dalton RS, Smart NJ, Edwards TJ, Chandler I, Daniels IR. Short-term outcomes of the prone perineal approach for extra-levator abdomino-perineal excision (elAPE). Surgeon 2012; 10(6): 342-6.

10. Han JG, Wang ZJ, Wei GH, Gao ZG, Yang Y, Zhao BC. Randomised clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. Am J Surg 2012; 204(3): 274-82.

11. Foster JD, Pathak S, Smart NJ, et al. Reconstruction of the perineum following extralevator abdominoperineal excision for carcinoma of the lower rectum: a systematic review. Colorectal Dis 2012; 14(9): 1052-9.

12. Singh B, Lloyd G, Nilsson PJ, Chaudhri S. Laparoscopic extralevator abdominal perineal excision of the rectum: the best of both worlds. Tech Coloproctol 2012; 16(1): 73-5.

13. Bellows CF, Smith A, Malsbury J, Helton WS. Repair of incisional hernias with biological prosthesis: a systematic review of current evidence. Am J Surg 2013; 205(1): 85-101.

14. Peppas G, Gkegkes ID, Makris MC, Falagas ME. Biological mesh in hernia repair, abdominal wall defects, and reconstruction and treatment of pelvic organ prolapse: a review of the clinical evidence. Am Surg 2010; 76(11): 1290-9. Bebenek M. Abdominosacral amputation of the rectum for low rectal cancers: ten years of experience. Ann Surg Oncol 2009; 16(8): 2211-7.

15. de Broux B, Parc Y, Rondelli F, Dehni N, Tiret E, Parc R. Sutured perineal omentoplasty afer abdominoperineal resection for adenocarcinoma of the lower rectum. Dis Colon Rectum 2005; 48(3): 476-81.

16. de Campos-Lobato LF, Stocchi L, Dietz DW, Lavery IC, Fazio VW, Kalady MF. Prone or lithotomy positioning during an abdominoperineal resection for rectal cancer results in comparable oncologic outcomes. Dis Colon Rectum 2011; 54(8): 939-46.

76 BIOPEX-study protocol

17. Bailey IS, Karran SE, Toyn K, Brough P, Ranaboldo C, Karran SJ. Community surveillance of complications afer hernia surgery. BMJ 1992; 304(6825): 469-71.

18. Chadwick MA, Vieten D, Pettitt E, Dixon AR, Roe AM. Short course preoperative radiotherapy is the single most important risk factor for perineal wound complications afer abdominoperineal excision of the rectum. Colorectal Dis 2006; 8(9): 756-61.

19. Han JG, Wang ZJ, Gao ZG, Xu HM, Yang ZH, Jin ML. Pelvic foor reconstruction using human acellular dermal matrix afer cylindrical abdominoperineal resection. Dis Colon Rectum 2010; 53(2): 219-23.

20. Peacock O, Pandya H, Sharp T, et al. Biological mesh reconstruction of perineal wounds following enhanced abdominoperineal excision of rectum (APER). IntJ Colorectal Dis 2012; 27(4): 475-82.

21. Vaughan-Shaw PG, King AT, Cheung T, et al. Early experience with laparoscopic extralevator abdominoperineal excision within an enhanced recovery setting: analysis of short-term outcomes and quality of life. Ann R Coll Surg Engl 2011; 93(6): 451-9.

22. Wille-Jorgensen P, Pilsgaard B, Moller P. Reconstruction of the pelvic foor with a biological mesh afer abdominoperineal excision for rectal cancer. Int J Colorectal Dis 2009; 24(3): 323-5.

23. EQ-5D-5L Value Sets http://www.euroqol.org/about-eq-5d/valuation-of-eq-5d/eq-5d-5l-value-sets.html. 2014. 4 24. Hakkaart-van Rooijen L. Handleiding voor kostenonderzoek www.euroqol.org/about-eq-5d/valuation-of- eq-5d/eq-5d-5l-value-sets.html.

25. Butler CE, Gundeslioglu AO, Rodriguez-Bigas MA. Outcomes of immediate vertical rectus abdominis myocutaneous fap reconstruction for irradiated abdominoperineal resection defects. J Am Coll Surg 2008; 206(4): 694-703.

26. Lefevre JH, Parc Y, Kerneis S, et al. Abdomino-perineal resection for anal cancer: impact of a vertical rectus abdominis myocutaneus fap on survival, recurrence, morbidity, and wound healing. Ann Surg 2009; 250(5): 707-11.

27. Nisar PJ, Scott HJ. Myocutaneous fap reconstruction of the pelvis afer abdominoperineal excision. Colorectal Dis 2009; 11(8): 806-16.

28. Artioukh DY, Smith RA, Gokul K. Risk factors for impaired healing of the perineal wound afer abdominoperineal resection of rectum for carcinoma. Colorectal Dis 2007; 9(4): 362-7.

77

CHAPTER 5

Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer: a single blinded, multicentre randomised controlled trial (the BIOPEX-study)

G.D. Musters, C.E.L. Klaver, R.J.I. Bosker, J.W.A. Burger, P. van Duijvendijk, B. van Etten, A.A.W. van Geloven, E.J.R. de Graaf, C. Hof, J.W.A. Leijtens, H.J.T. Rutten, B. Singh, R.J.C.L.M. Vuylsteke, J.H.W. de Wilt, M.G.W. Dijkgraaf, W.A. Bemelman, and P.J. Tanis

Submitted Chapter 5

ABSTRACT

Aim: Perineal wound complications, including infection, dehiscence, and pelvic abscess, frequently occur afer extralevator abdominoperineal resection (eAPR) with preoperative radiotherapy for low rectal cancer. Cohort studies have suggested that biological mesh closure of the pelvic foor improves perineal wound healing. Tis study aimed to determine whether biological mesh closure was superior in reducing perineal wound complications compared to primary wound closure.

Methods: In this single blinded, superiority, multicentre randomised controlled trial, patients were enrolled from 12 hospitals across the Netherlands and United Kingdom and randomly assigned to primary closure (standard arm) or biological mesh closure of the pelvic foor (intervention arm). A non- cross-linked porcine acellular dermal mesh (Strattice™, LifeCell, Acelity™ Company, Branchburg, New Jersey, USA) was sutured to the pelvic foor remnants in the intervention arm, followed by a layered closure of the ischioanal and subcutaneous fat and skin similar to the control intervention. Randomisation was performed by the principle investigator using a randomisation website with concealed block sizes in a 1:1 ratio, with stratifcation for age (18-59 or ≥60 years), gender, and surgical approach (laparoscopic or open). Te outcome of the randomisation was concealed from the patient and perineal wound assessor. Te primary endpoint was the rate of uncomplicated perineal wound healing defned as a Southampton wound score of less than two at 30 days postoperatively, and was assessed with an intention to treat analysis using a Chi2-test. Te trial was registered on ClinicalTrials.gov with the number NCT01927497.

Results: Between February 1, 2013 and September 1, 2014, 104 patients were enrolled and randomly assigned to primary closure (n=54; 2 dropouts) and biological mesh closure (n=50; 3 dropouts). Uncomplicated perineal wound healing rate at 30 days postoperative was 66% (33/50; 2 not evaluable) afer primary closure, which did not signifcantly difer from 64% (30/47) afer biological mesh closure (RR 0.967; 95% CI 0.721-1.296; p=0.8229). Surgical complications within 90 days of surgery occurred in 40% (21/52) afer primary closure and in 42% (20/47) afer biological mesh closure (p=0.8268).

Conclusion: Perineal wound healing afer eAPR in patients who had undergone preoperative radiotherapy for low rectal cancer was not superior when a biological mesh was used to close the pelvic foor, when compared to primary perineal wound closure. A signifcantly lower one-year perineal hernia rate afer biological mesh closure is a promising secondary fnding that needs longer follow-up to determine its clinical relevance.

80 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

INTRODUCTION

Perineal wound complications following abdominoperineal resection (APR) for rectal cancer are a frequent source of morbidity.1 Perineal wound infection, dehiscence, and pelvic abscesses ofen require prolonged post-operative wound care, which may last for several months and may result in a chronic perineal sinus or fstula. Furthermore, the perineal wound can be associated with prolonged pain, sitting problems and require daily wound care, all of which can interfere with a patients’ quality of life. As rectal cancer is one of the most common cancers in the Western world with up to one third of these patients undergoing an APR, it is not too surprising that perineal wound complications are an important health care problem.2

Treatment related morbidity has become increasingly important since the oncological outcome of rectal cancer has been optimized with improved surgical techniques. In a standard APR approach, the total mesorectal excision plane is followed down to the pelvic foor. Due to the anatomical narrowing of the distal mesorectum, this results in a typical tapering of the resected specimen at the level of the primary tumour, with a risk of a positive resection margin and perforation. A pooled analysis of fve European randomised controlled trials showed that standard APR was an independent risk factor for local recurrence and associated with a signifcantly lower fve year survival rate compared to low anterior resection.3 Using an extralevator approach (eAPR), an en bloc resection which includes the levator muscles, results in a cylindrical specimen. Although there 5 is still controversy about the role of the extralevator approach and the extent of the resection, the surgical quality of APR has improved over time.4,5 In addition, the increased use of preoperative radiotherapy has contributed to improved locoregional control.6

However, whilst both preoperative radiotherapy and wider surgical excision have improved the oncological outcome, it has come at the expense of increased perineal wound complications.7,8 Closure of the perineum afer eAPR can either be performed as a primary layered closure of the ischioanal and subcutaneous tissues, or by using a biological mesh or autologous tissue fap. Tere is currently no consensus as to the best surgical technique for perineal wound closure afer eAPR in terms of short and long-term wound complications and quality of life. Available data from cohort studies are for the most part of low quality. Despite the lack of evidence, biological meshes are increasingly being used for pelvic foor closure afer eAPR. Several recent reviews concluded that there is a need for a properly designed prospective study to address this important question.9,10 Tis is the frst multicentre randomised controlled trial aimed to determine the efectiveness of pelvic foor reconstruction using a biological mesh in improving perineal wound healing compared to primary perineal wound closure afer eAPR in patients with low rectal cancer who have undergone preoperative radiotherapy.

81 Chapter 5

METHODS

Study design Te BIOPEX-study is a multicentre, parallel-group, single blinded, superiority, randomised controlled trial, performed in one non-teaching hospital, six teaching hospitals and four university hospitals in the Netherlands, and one university hospital in the United Kingdom. In this investigator initiated study, eligible patients were randomised between primary closure of the perineal defect (standard arm) and pelvic foor reconstruction using a biological mesh followed by primary perineal closure (intervention arm) (Figure 1). An independent observer, unaware of the intervention to which the patient was allocated, evaluated perineal wound healing using the Southampton wound scoring system (appendix) and perineal herniation afer 7 and 30 days, and 3, 6, 9 and 12 months postoperatively.11 An independent observer also took photographs of the perineal wound on clinical follow-up, which were evaluated by the trial coordinators (G.M, C.K.) in relation to the Southampton wound score assigned by the independent observer. When a discrepancy arose, the trial coordinator contacted the local independent observer to reach consensus. At twelve months postoperatively, computed tomography (CT-scan) of the pelvis was performed, to assess for the presence of a presacral sinus, perineal sinus, and perineal herniation. Quality of Life questionnaires were taken at each follow-up interval. In addition during follow-up, the nature and severity of any wound event, all medical or surgical interventions, re-operations and oncological outcome were recorded. Te study protocol was approved by the ethical review board of the Academic Medical Centre, Amsterdam, the Netherlands. Te study protocol has been previously published and was available online at the start of the trial.12

Patients Eligible patients were preoperatively approached for participation into the BIOPEX-study at the surgical outpatient clinics. A patient was classifed as eligible when the following criteria were fulflled; age older than 18 years, primary rectal cancer, life expectancy of more than two years, preoperative radiotherapy for rectal cancer, planned for an eAPR, and able to return for outpatient follow-up visits. Preoperative radiotherapy could consist of either short course radiotherapy (5x5 Gy) or long course chemoradiotherapy. Exclusion criteria were the need for a total exenteration, sacral resection above the level of S4/5, previous pelvic irradiation, severe systemic or collagen disorders which could afect wound healing (i.e. renal failure requiring dialysis, liver cirrhosis, immunocompromised status, and Marfan syndrome), sensitivity to porcine derived products or polysorbate, and enrolment in other trials which could infuence wound healing. Written informed consent was obtained for all participating patients before randomisation.

Randomisation and masking Afer written informed consent, patients were randomly assigned by the local physician to primary perineal wound closure (standard arm) or pelvic foor reconstruction using a biological mesh followed by primary perineal closure (intervention arm) in a 1:1 ratio. Randomisation was performed by a central automated randomisation website pre-operatively, with random concealed block sizes (2, 4, and 6) and stratifcation for age (18-59 or ≥60 years), gender, and surgical approach (laparoscopic or open). Te allocation of the treatment was blinded to the patient and perineal wound assessor.

82 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

Procedures All patients received preoperative antibiotics, according to the local trial site protocol. Patient positioning for the perineal phase of the eAPR (supine or prone), the surgical approach (laparoscopic or open), and the use of omentoplasty were lef to the discretion of the operating surgeon. In all patients, the principles of an extralevator approach were adhered to, in which the levator muscles were laterally transected in order to leave a muscular cuf around the resected tumour. Te coccyx was not routinely resected, except for surgical exposure or oncological reasons. Excision of the perineal skin and ischioanal fat was limited if oncologically justifed. Te quality of the resected specimen was evaluated by the local trial site pathologists and photographs were sent to the trial coordinator (G.M.).

In the standard arm, the perineal wound was closed in layers by stitching together the ischioanal and subcutaneous fat using interrupted Vicryl® (Ethicon Inc, Johnson and Johnson, USA) sutures. Te skin was closed using interrupted sutures according to the preference of the surgeon’s. Placement of an abdominal drain and/or perineal drain was lef to the discretion of the local surgeon. For pelvic foor reconstruction in the experimental arm, an acellular non-cross linked biological mesh derived from porcine dermis (6x10 cm, Strattice™, LifeCell, Acelity™ Company, Branchburg, New Jersey, USA), was fxed using interrupted Prolene® (Ethicon Inc, Johnson and Johnson, USA) or polydioxanone (PDS) sutures. Te mesh was sutured posteriorly either side of the coccyx or sacrum. Laterally, the biological mesh was attached to the remnant of the levator muscle and anteriorly to 5 the transverse perineal muscles. A suction drain was placed on top of the mesh and the ischioanal and subcutaneous fat and skin were closed, similar to the control arm. To ensure a standardized technique of biological mesh closure, a course on fresh frozen cadavers was undertaken prior to the start of the trial and a refreshment course was delivered afer 52 included patients had been included in the study. In addition, in centres with limited experience in biological mesh closure at the start of the trial, the technique for pelvic foor reconstruction was proctored by an experienced principle investigator (P.T.) at their trial site hospital.

Outcome Te primary endpoint was the percentage of uncomplicated perineal wound healing defned as a Southampton wound score of less than two at 30 days postoperatively. Secondary endpoints were perineal wound healing according to the Southampton wound score, symptomatic and asymptomatic perineal hernia, postoperative pain according to the visual analogue scale at 3, 6, 9, and 12 months. Other secondary outcomes were the presence of persistent perineal or presacral sinus, need for readmission or re-interventions related to pre-sacral abscesses or perineal wound problems, and length of hospital stay during one year of follow-up. Generic quality of life was assessed using the Short Form-36 version 2 (SF36v2) and a fve level version of the fve dimensional EuroQol (EQ-5D- 5L) preoperatively, and at 3 and 12 months afer surgery. Health utilities were derived from patients’ scoring profles on the EQ-5D-5L based on available crosswalk value sets Also the gastrointestinal quality of life was assessed using the EORTC (QLQ-C30/CR29) preoperatively and at 3, 6, 9, and 12 months postoperatively.

83 Chapter 5

All serious adverse events defned as death, a life threatening event, need for hospitalization, prolongation of hospitalization, signifcant disability or incapacity were reported to the trial coordinators (G.M., C.K.) and the ethical review board. Although this was considered to be a low- risk trial because of the use of biological meshes for pelvic foor reconstruction in routine clinical practice outside a trial setting, an interim-analysis on safety was nevertheless performed afer 52 included patients. Tis analysis showed a non-skewed distribution in serious adverse events between both groups.

Statistical analyses Given the lack of high quality data in the current literature, we defned a clinically relevant diference in primary uncomplicated perineal wound healing as a diference of 25%, which would justify the routine use of a costly biological mesh in addition to primary perineal closure. Applying a Chi2-test with a two-sided 0.05 signifcance level and a power of 80% with an estimated drop-out of 5%, a total number of 104 patients (52 per group) was needed to detect an increase in uncomplicated wound healing from 60% to 85%. According to distribution, descriptive data were reported as median with interquartile range (IQR) or mean with standard deviation (SD). Categorical data were analysed with the Chi2-test or Fisher’s exact test and continuous variables were analysed using the Mann-Whitney-Wilcoxon test. Te primary endpoint was assessed using Chi2-test and presented with absolute incidences and relative risk (RR) with 95% confdence interval (95% CI). A post- hoc multivariable cox regression (without a time variable) on the primary endpoint was performed to assess the efect of wound closure, corrected for baseline characteristics (excluding the three stratifcation parameters age, gender and type of surgery associated with randomisation); results are presented as RR and 95% CIs. A p-value of 0.05 was considered signifcant. Perineal hernia rate was assessed with a Fisher’s exact test for comparison of overall proportions at 12 month follow- up, and a diference in time to perineal hernia was assessed with a log rank test. To evaluate the repeated measurements in pain score, a generalized mixed model with Poisson distribution was used. Other repeated measurements were analysed using a generalized estimating equations (GEE model) with a multinomial ordinal distribution and a logit link function. All questionnaires were analysed according to the manuals and presented as domain and summarised scores. Questionnaire outcome comparisons were analysed using a mixed linear model corrected for multiple testing using the Bonferroni correction method. All the data were analysed in accordance to the intention- to-treat principle. No data monitoring committee was installed. Te trial was registered on a trial registration website under the registration code NCT01927497 (clinicaltrials.gov).

Role of funding Tis trial was an investigator initiated trial and the funder of the study had no infuence on study design, data collection, data analysis, data interpretation, or writing of the manuscript. Te trial coordinators and principle investigator had full access to all the data in the study and the corresponding author had fnal responsibility for the decision to submit for publication.

84 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

5

Figure 1 Flow-diagram

RESULTS

Between the frst of February 2013 and the f rst of September 2014, 117 eligible patients were approached to participate in the BIOPEX-study. Of those 117 patients, 104 patients consented to the trial of which 54 patients were randomly assigned to primary perineal wound closure (standard arm) and 50 patients to biological mesh closure of the pelvic foor (intervention arm). Afer randomization two patients in the primary perineal closure group and three patients in the biological mesh closure group were excluded from further analysis due to protocol violations (i.e. no preoperative radiotherapy, no eAPR performed) or death between randomisation and surgery (fgure 1), resulting in 52 patients in the standard arm and 47 patients in the intervention arm.

Te baseline characteristics of the included patients are described in table 1. Te mean age of the whole group was 65 years (SD 11), and 74% (73/99) of the patients were male. Surgery started with the abdominal phase frst in 31 (60%) of the 52 patients in the primary closure group, which

85 Chapter 5 was signifcantly less than 41 (87%) of the 47 patients in the biological mesh group (p=0.0020). Laparoscopic surgery for the abdominal phase was performed in 62% (32/52) of the patients that underwent primary closure, and in 64% (30/47) of the patients who underwent biological mesh closure (p=0.8140). An omentoplasty was placed in the pelvis in 71% (37/52) of the patients undergoing primary closure, which was signifcantly higher compared to 51% (24/47) in the biological mesh group (p=0.0338). Te coccyx was resected in 19% (10/52) of the patients in the primary closure group, which did not signifcantly difer from 15% (7/47) in the biological mesh group (p=0.5677). Te additional resections performed during eAPR are displayed in the appendix. Tumour or rectal perforation occurred in six percent (3/52) of the patients in the primary closure group, and in six percent (3/47) of the patients in the biological mesh group (p=1.0000). A perineal drain was placed in 31% (16/52) of the patients in the primary closure group, which is signifcantly lower than 81% (38/47) in the biological mesh group (p<0.0001). Te mean duration of the surgery afer primary closure was 223 minutes (SD 68), which is signifcantly less than 275 minutes (SD 81) for the biological mesh group (p=0.0010).

Te number of patients in whom the perineal wound was assessed at the diferent follow-up intervals is shown in fgure 1. At 30 days postoperatively, the percentage of patients with uncomplicated perineal wound healing (Southampton wound score <2) was 66% (33/50, two patients could not be evaluated) afer primary perineal closure, which did not signifcantly difer from 64% (30/47) afer biological mesh closure (RR 0.967; 95% CI 0.721-1.296; p=0.8229). Tere was also no statistical signifcant diference between the type of perineal wound infection according to the Southampton wound score between primary closure and biological mesh closure at postoperative day 30 (p=0.1405-0.6808; table 2). Te post-hoc power for the primary endpoint, with an increase in uncomplicated perineal wound healing from 60% to 85%, applying a Chi2-test with a two-sided 0.05 signifcance level with 50 and 47 patients in the randomised groups, respectively, was 79%.

Perineal wound healing was uncomplicated during one year of follow-up at any postoperative time interval in 52% (27/52) of the patients afer primary closure, and in 55% (26/47) of the patients afer biological mesh closure (RR 1.065; 95% CI 0.738-1.537; p=0.7351). During the complete follow-up period, there was no signifcant diference between the type of perineal wound infection according to the Southampton wound score and randomisation groups (p=0.3044; appendix). Te post-hoc multivariable regression on perineal wound healing within 12 months corrected for diferences in baseline characteristics between the randomisation groups did not reveal any signifcant association for biological mesh (RR 0.873; 95% CI 0.440-1.733; p=0.6980), abdominal phase frst (RR 1.268; 95% CI 0.631-2.549; p=0.5048), omentoplasty (RR 1.314; 95% CI 0.676-2.554; p=0.4213), and perineal drain (RR 1.255; 95% CI 0.610-2.583; p=0.5366).

Postoperative pain did not signifcantly difer between the randomisation groups at all follow- up visits (p=0.8780; appendix).Te median postoperative stay was seven days (IQR 6-11) afer primary closure and eight days (IQR 7-13) afer biological mesh closure (p=0.2465). Postoperative complications (within 90 days) were not signifcantly diferent between the randomisation groups for surgical complications (p=0.8268) and non–surgical complications (p=0.6659; table 2, appendix). Tere was also no signifcant diference between both groups for surgical (p=0.4403) and percutaneous re-interventions (p=0.3432) within 12 months (table 2). None of the biological meshes had to be explanted.

86 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

Table 1 Baseline characteristics

Group A Group B Primary Biological mesh wound closure closure (n=52) (n=47) Hospital Non-teaching hospital (n,%) 4 (8) 7 (15) Teaching hospital (n,%) 28 (54) 20 (43) University hospital (n,%) 20 (38) 20 (43) Gender Male (n,%) 38 (73) 35 (74) Female (n,%) 14 (27) 11 (23) Age Years ± SD 64 (12) 65 (12) Body-mass index kg/m2 ± SD 26 (3) 27 (5) ASA-classifcation ASA 1 (n,%) 37 (71) 29 (62) ASA 2 (n,%) 14 (27) 16 (34) ASA 3 (n,%) 1 (2) 2 (4) Previous surgery Abdominal surgery (n,%) 11 (21) 6 (13) Pelvic surgery (n,%) 2 (4) 3 (6) Anorectal surgery (n,%) 4 (8) 4 (9) Comorbidity Diabetes (n,%) 4 (8) 5 (11) 5 Respiratory (n,%) 4 (8) 4 (9) Cardiac (n,%) 6 (12) 10 (21) Vascular (n,%) 2 (4) 1 (2) Nicotine addiction (n,%) 6 (12) 3 (6) Medication Immunosuppressant’s (n,%) 2 (4) 0 Weight loss before surgery >10% of total body weight (n,%) 5 (10) 3 (6) Obstruction For which diverting stoma (n,%) 4 (8) 4 (9) Tumour location Distance between lower border tumour and anal 1 (0-3) 2 (0-3) verge in cm on MRI (IQR) Neoadjuvant radiotherapy Short course (5x5 Gy) (n,%) 10 (19) 10 (21) Long course chemoradiotherapy (n,%) 42 (81) 37 (79) Quality of resected specimen High quality (n,%) 31 (60) 29 (62) Moderate quality (n,%) 4 (8) 3 (6) Poor quality (n,%) 3 (6) 4 (9) Not reported (n,%) 14 (27) 11 (23) Radical surgical resection Circumferential Resection Margin >1mm (n,%) 48 (92) 43 (91) γpTNM-stage Stage 1 (n,%) 19 (37) 24 (51) Stage 2 (n,%) 15 (29) 10 (21) Stage 3 (n,%) 12 (23) 10 (21) Stage 4 (n,%) 6 (12) 3 (6) γpTNM= Pathological tumour staging afer neoadjuvant (chemo)radiotherapy. n=number of patients. Data are presented according to distribution in means with standard deviation (SD) or in medians with interquartile range (IQR). ASA-classifcation= American Society of Anaesthesiologists classifcation. MRI= Magnetic resonance imaging. High quality specimen= mesorectum intact and cylindrical specimen. Moderate quality specimen= small defects in mesorectum or conical specimen. Poor quality= dissection of the rectal wall or perforation. 87 Chapter 5

Table 2 Perineal wound healing

Group A Group B Primary Biological P-value closure mesh closure (n=52)1 (n=47)2 Normal perineal wound healing 7 days postoperative (n, %) 35/50 (70) 34/46 (74) 0·6701 (Southampton wound score <2) 30 days postoperative (n, %) 33/50 (66) 30/47 (64) 0·8229 3 months postoperative (n, %) 42/52 (81) 39/47 (83) 0·7759 6 months postoperative (n, %) 43/49 (88) 39/44 (87) 0·8955 9 months postoperative (n, %) 44/49 (90) 41/43 (95) 0·4419 12 months postoperative (n, %) 49/50 (98) 41/43 (95) 0·5940 Severity of infection Erythema plus other signs of 0 2/47 (4) 0·1405 (at 30 days) infammation (n, %) Clear or haemoserous discharge (n, %) 7/50 (14) 8/47 (17) 0·6808 Pus discharge (n, %) 7/50 (14) 2/47 (4) 0·1608 Deep or severe wound infection (n, %) 3/50 (6) 5/47 (11) 0·4780 Surgical complications Overall (n, %) 20 (38) 20 (43) 0.6878 (within 90 days) Non-surgical complications Overall (n, %) 3 (6) 3 (6) 1.0000 (within 90 days) Perineal hernia Total (n, %) 13 (25) 4 (9) 0.0351 Within 12 months Asymptomatic (n, %) 4 (8) 2 (4) Surgical re-interventions Total (n, %) 5 (10) 2 (4) 0.4403 For perineal wound problems Perineal hernia correction (n, %) 2 (4) 1 (2) within 12 months Abscess drainage (n, %) 1 (2) 1 (2) Gluteus fap (n, %) 2 (4) 0 Percutaneous re-intervention Abscess drainage (n, %) 1 (2) 3 (6) 0.3432 For perineal wound problems within 12 months Surgical complications are; urinary retention, ileus, trocard hernia, postoperative bleeding, presacral fstula, stoma dysfunction, pneumonia, perineal hernia <90 days, (appendix). Non-surgical complications are; atrial fbrillation, heart decompensation, urinary tract infection, cholecystitis, the fu (appendix). 1= number of evaluable patients for each group difers for diferent time intervals postoperatively (Figure 1).

Perineal herniation occurred in 25% (13/52) of the patients afer primary closure, which was signifcantly more than the nine percent (4/47) in the biological mesh group (RR 1.220; 95% CI 1.019-1.460; p=0.0351; table 2). Te median duration between surgery and a perineal hernia was nine months (IQR 5-12). As assessed by clinical examination and/or CT-scan, perineal hernias occurred throughout the follow-up period in the primary closure group but, mainly at the end of the 12 month follow-up in the biological mesh group (appendix; p=0.0329 (log-rank test)). Of the 17 patients with a perineal hernia, six patients were asymptomatic, three patients were operated for their perineal hernia and the remaining eight patients were symptomatic but conservatively treated.

88 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

During follow-up, adjuvant chemotherapy was given to six patients in the primary closure group, and in eight patients in the biological mesh group (p=0.4343). A local recurrence occurred in three patients afer primary closure and in one patient afer biological mesh closure (p=0.6192). Metastasis to liver or lung occurred in 14 patients. In total, three patients died, all in the biological mesh group, but unrelated to the intervention. One patient died afer a Bricker procedure for a bladder fstula 11 months afer eAPR, one patient died due to septic shock of unknown origin three months afer eAPR during adjuvant chemotherapy, and one patient died due to an intracerebral bleeding ten months afer eAPR.

Te response rates of the quality of life questionnaires (SF-36v2, EQ 5D-5L, EORTC QLQ C30, and EORTC QLQ CR29) varied between 76% (75/99) at nine months postoperatively to 93% (92/99) for the preoperative questionnaires. Te quality of life questionnaires showed no signifcant diferences in the main outcome scores. In addition, the subscales also did not show any statistically signifcant diferences afer correction for multiple testing. Te responses and subscales are shown in the appendix.

DISCUSSION

Biological mesh closure of the pelvic foor following eAPR in patients who have undergone 5 preoperative radiotherapy resulted in an uncomplicated perineal wound healing rate of 64% at 30 days postoperatively, which did not signifcantly difer from 66% in the primary perineal wound closure group. No other signifcant diferences in perineal wound healing or quality of life were observed at other time intervals postoperatively between both randomisation groups up to one year afer surgery. Biological mesh closure signifcantly increased operation time with an absolute mean diference of 52 minutes. Te one year perineal hernia rate was signifcantly lower in the biological mesh group compared to the control group, namely 9% versus 25%, respectively.

Te increasing use of biological meshes to close the defect in the pelvic foor in routine practice may be a result of a number of publications suggesting a positive efect on perineal wound healing. We performed an update of our previously published systematic review with meta-analysis.8 Te pooled perineal wound complication rate afer eAPR with preoperative radiotherapy was 38% (95% CI 24- 54; I2 72%; 3 studies, n= 188)13-15 afer primary closure, and 11% (95% CI 6-19; I2 0%; 3 studies, n= 104)14,16,17 afer biological mesh closure (appendix). However, these were all non-randomised cohort studies with almost exclusively retrospective data. Furthermore, surgical and patient characteristics greatly difered among the primary closure studies, which may explain the considerable statistical heterogeneity (I2).

Te considerable diference in perineal wound complication rates between the systematic review of the literature and the present randomised controlled trial is most likely related to several methodological shortcomings of the previously published cohort studies. Besides potential confounding by indication, perineal wound complications were not prospectively evaluated using a standardised wound scoring in the cohort studies, but mostly by retrospective chart review. In

89 Chapter 5

addition, patients and wound assessors were not blinded. All these factors may have resulted in underreporting of perineal wound complications in the biological mesh studies.

Despite that the anus is ofen closed just before disinfection and subsequent incision, the operating feld is considered to be contaminated. Because there is an increased risk of infectious complications if a synthetic mesh is used in a contaminated feld, a biological mesh for closing the pelvic foor defect afer eAPR seems more appropriate.18 A biological mesh can be derived from human, bovine or porcine tissues and are processed to leave only the extracellulair matrix behind. Te extracellular matrix promotes neo-vascularisation and provides a scafold for host cell migration, which makes it suitable for use in a contaminated feld. Biological meshes can be divided in cross linked and non-cross linked meshes. Cross links are covalent bonds between the collagen and may increase the life span of the mesh by limiting the efect of host cell and bacterial collagenase. However, they may decrease cellular infltration and neo-vascularization, which may make them less useful in a contaminated feld compared to non-cross linked biological meshes.19 Te cost of a biological mesh ranges between 1100 and 1800 euro per patient depending on the size of acellular porcine collagen implant and the commercial company. In addition, fxating the biological mesh signifcantly increased the duration of surgery. Terefore, biological meshes may also increase direct medical costs.

One explanation for not fnding a signifcant impact on perineal wound healing might be that any positive efect of biological mesh reconstruction of the pelvic foor is nullifed by creating a dead space between the biological mesh and the closed perineum. In this dead space, fuid may accumulate and develop into an abscess whitin a contaminated feld. A perineal drain was most ofen placed in the biological mesh group but did not show a signifcant efect in the post hoc analysis on perineal wound healing. Terefore, a perineal drain might be insufcient to collapse the dead space between the biological mesh and skin. Te flling of this dead space with a muscle or fasciocutaneous transposition fap might therefore, increase wound healing. However, the downside of a tissue fap is donor side morbidity and the risk of fap necrosis.

In our updated systematic review, a perineal hernia afer eAPR occurred in 4% (95% CI 1-11; I2 67%; 6 studies, n= 411)13,14,20-22 of the patients afer primary wound closure, and in 7% (95% CI 4.5-11; I2 0%; 8 studies, n= 287)14,16,23-28 of the patients afer biological mesh closure. In contrast, the one year perineal hernia rate of 25% afer primary wound closure in the present RCT was signifcantly higher than 9% afer biological mesh closure. Te remarkably low rate of perineal hernia afer primary wound closure in the updated systematic review is likely related to the primary oncological design of these cohort studies, without a focus on perineal wound complications, and retrospective data collection. Biological mesh studies are ofen primarily concerned with perineal wound outcome, which explains the similar hernia rate compared to our RCT (7% versus 9%). Follow-up in the updated systematic review ranged between 8 and 60 months. Te 12 months’ follow-up in the present study may underestimate the long-term perineal hernia rate. Degradation of the non-cross- linked biological mesh starts at around 6 months, while full degradation may take up to one year.29 As a result, the perineal hernia rate will continue to rise in both groups. Perineal hernia repair is

90 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

considered a non-contaminated procedure and can be performed with a less expensive synthetic mesh. Although promising, it seems to be too early to conclude on the cost-efectiveness of biological mesh closure afer eAPR at 12 months’ follow-up considering the reduced perineal hernia rate.

Newer techniques need to be investigated to resolve the increasing clinical problem of perineal wound complications. Currently, only one single centre, open label randomised controlled trial is being conducted in which patients are randomised between a porcine biological mesh and gluteus maximus myocutaneous fap closure of the perineal wound afer eAPR (NEAPE; clinicaltrial.gov identifer: NCT01347697). Tis study does not include a primary closure group. In a systematic review of cohort studies, pelvic foor closure afer eAPR with tissue faps was found to be equally efective as a biological mesh closure, but results of the NEAPE trial are awaited.9 Primary perineal wound closure is ofen not considered to be a feasible option afer eAPR, because it is thought that simply closing the remaining fatty tissue and skin would be insufcient, and will almost certainly result in perineal herniation. Te control arm of the present RCT demonstrates that primary layered closure of the perineum is still a valid option, resulting in a perineal hernia in only a quarter of the patients afer one year. Terefore, further research should still include primary perineal wound closure as a control intervention, because presently there is no other perineal closure technique that has convincingly been shown to be superior. Te intervention arm of future studies may include a tissue fap, depending on the results of the NEAPE trial. 5 Limitations of this study could be that some participating centres did not assess all patients for participation in the BIOPEX-study, or did not report all eligible non-participating patients to the trial coordinators. As a result, some selection bias might have occurred. In addition, eligible non- participating patients that were reported to the trial coordinators did not gave informed consent to assess their medical records and could not be compared with the included patients. Terefore, comparability between the included patients and the eligible non-participating patients could not be assessed. Furthermore, this trial was designed to look for a signifcant diference of at least 25% on perineal wound healing, and may be underpowered. However, the observed percentages of wound healing (64% vs. 66%) do not indicate any diference at all in favour of the biological mesh. Finally, to objectively assess perineal wounds with limited inter and intra-observer variability is challenging, especially when wound scoring systems are limited and not validated for every type of wound. However, the Southampton wound score was the best available method for this purpose.

We conclude that biological mesh closure of the pelvic foor is not superior in perineal wound healing afer eAPR in patients with rectal cancer who have undergone preoperative radiotherapy when compared to primary perineal wound closure. Furthermore, biological mesh closure did not improve the quality of life, but did increase the duration of surgery. However, biological mesh closure resulted in a signifcantly lower perineal hernia rate afer one year.

91 Chapter 5

SUPPLEMENTARY APPENDIX

Inclusion per hospital and reasons for declining participation

Hospital Time interval* Eligible patients Included patients (months) (n=117) (n=104) 1 University hospital 4 23 20 2 Teaching hospital 5 15 14 3 Non-teaching hospital 6 14 11 4 University hospital 1 11 10 5 Teaching hospital 3 9 9 6 Teaching hospital 9 11 9 7 Teaching hospital 4 9 9 8 University hospital 17 6 6 9 Teaching hospital 6 6 6 10 University hospital 5 6 4 11 Teaching hospital 3 4 3 12 University hospital 5 2 2 13 Teaching hospital 11 1 1

*= time interval between ethical approval from ethical review board and the frst included patient.

Te reasons why eligible patients did not participate: • Does not want to give a reason, or didn’t want their medical records used for research (n=6) • Patient feels that to many things are happening at once and therefore declines (n=3) • Does not want a biological mesh, if randomized (n=2) • Does not want to participate due to blinding of the randomization (n=1) • Religion (n=1)

92 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

Additional surgical resection outcome

Group A Group B Primary wound Biological mesh P-value closure (n=52) closure (n=47) Surgical approach Laparoscopic (n,%) 32 (62) 30 (64) 0.8140 Conversion to open (n,%) 2 (4) 0 0.1571 Intra-operative radiotherapy 10 Gy (n,%) 2 (4) 4 (9) 0.4190 Drain Abdominal drain (n,%) 40 (77) 34 (72) 0.6002 Perineal drain (n,%) 16 (31) 38 (81) <0.0001 Perineal phase Prone positioning (n,%) 25 (48) 30 (64) 0.1152 Local antibiotics Gentamicin sponge (n,%) 9 (17) 3 (6) 0.1273 Subcutaneous closure Two layers (n,%) 35/46 (76) 33/43 (77) 0.1942 One layer (n,%) 11/46 (24) 10/43 (23) Skin closure Lef open (n,%) 3 (6) 5 (11) 0.4718 Vicryl® sutures (n,%) 15 (29) 18 (38) 0.4762 Prolene® sutures (n,%) 19 (37) 13 (28) Other sutures (n,%) 15 (29) 11 (23) Length of perineal wound Centimetres (IQR) 11 (10-12) 9 (9-11) 0.0622 5 Blood loss Millilitres (IQR) 300 (100-750) 225 (100-678) 0.3304 Duration of surgery Minutes (SD) 223 (68) 275 (81) 0.0010 Postoperative antibiotics Less than 10 days (n,%) 11 (21) 9 (19) 1.0000 n= number of patients. According to distribution data are presented in median with interquartile range (IQR) or mean with standard deviation (SD). Conversion to an open procedure was performed in two patients in the primary wound closure group due the size of the tumour in one patient and due to intra-abdominal adhesions in the other patient. Other sutures= ethylene terephthalate sutures, synthetic monoflament suture. Vicryl® (Ethicon Inc, Johnson and Johnson, USA), Prolene® (Ethicon Inc, Johnson and Johnson, USA)

93 Chapter 5

Additional surgical resection at primary surgery

Group A Group B Primary wound Biological mesh closure (n=52) closure (n=47) Additional resection Number of patients with one additional resection (%) 17 (33) 14 (30) Number of patients with two additional resections (%) 4 (8) 3 (6) Os coccygeus resection (n,%) 10 (19) 7 (15) Partial prostate resection (n,%) 4 (8) 2 (4) Partial vaginal wall resection (n,%) 3 (6) 2 (4) Iliac lymph node excision (n,%) 3 (6) 2 (4) Partial presacral fascia resection (n,%) 1 (2) 0 Partial obturator fascia resection (n,%) 0 1 (2) Seminal vesicle excision (n,%) 3 (6) 3 (6) Para-aortic lymph node resection (n,%) 0 1 (2) Adnexal excision (n,%) 1 (2) 1 (2) Coloproctectomy (n,%)* 0 1 (2) n= number of patients. *= Coloproctectomy was performed due to a double tumour in a patient with familial polyposis coli syndrome.

94 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

Postoperative complications (<90 days)

Group A Group B Primary wound Biological mesh P-value closure (n=52) closure (n=47) Surgical complications Urinary retention (n,%) 9 (17) 6 (13) 0.5291 Ileus (n,%) 7 (13) 5 (11) 0.7635 Port site hernia (n,%) 0 3 (6) 0.1033 Postoperative bleeding (n,%) 1 (2) 1 (2) 1.0000 Presacral abscess (n,%) 4 (8) 5 (11) 0.7322 Perineal fstula (n,%) 0 2 (4) 0.2228 Stoma dysfunction (n,%) 2 (4) 0 0.4961 Perineal hernia <90 days (n,%) 1 (2) 0 1.0000 Non-surgical complications Atrial fbrillation (n,%) 1 (2) 0 1.0000 Heart failure (n,%) 1 (2) 0 1.0000 Pneumonia (n,%) 0 1 (2) 0.4747 Urinary tract infection (n,%) 0 1 (2) 0.4747 Cholecystitis (n,%) 0 1 (2) 0.4747 Flu (n,%) 1 (2) 0 0.4747 5 Hospital stay Days in ICU (IQR) 1 (1-2) 1 (1-3) 0.6495 Total hospital stay (IQR) 7 (6-11) 8 (7-13) 0.2465 Re-admission (n, %) 17 (33) 10 (21) 0.2174 Duration re-admission days (IQR) 6 (3-25) 11 (4-22) 0.6889 n= number of patients. According to distribution, data are presented in median with interquartile range (IQR).

Postoperative pain score (VAS)

Group A Group B Primary wound Biological mesh P-value* closure1 (n=52) closure2 (n=47) Visual analogue scale 30 days postoperative (IQR) 3 (1-7) 3 (2-6) 0.878 While sitting 3 months postoperative (IQR) 1 (0-3) 1 (0-2) 6 months postoperative (IQR) 0 (0-1) 0 (0-1) 9 months postoperative (IQR) 0 (0-1) 0 (0-1) 12 months postoperative (IQR) 0 (0-0) 0 (0-1) n= number of patients. Data are presented as medians with interquartile range (IQR). *= assed using a general linear model, according to Poisson distribution.1= patients in which visual analogue scale is not assessed; 30days (n=16), 3 months (n=14), 6 months (n=9), 9 months (n=12), 12 months (n=50). 2= patients in which visual analogue scale is not assessed; 30 days (n=3), 3 months (n=6), 6 months (n=3), 9 months (n=8), 12 months (n=43).

95 Chapter 5

Kaplan-Meier analysis; probability of a perineal hernia afer eAPR

96 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer 0 0 12m 1 (2) 1 (2) (n=43) Biological Biological 0 0 0 12m 1 (2) (n=50) Primary Primary 0 0 9m 1 (2) 1 (2) (n=43) Biological Biological 0 0 0 9m 5 (10) (n=49) Primary Primary 0 6m 2 (5) 2 (5) 1 (2) (n=44) Biological Biological - - 0 6m 12m 1 (2) 3 (6) 2 (4) 0.5513 1.0000 0.4624 (n=49) Primary Primary 0 - - 3m 1 (2) 1 (2) 9m 6 (13) (n=47) 0.1830 0.4673 0.2092 Biological Biological 0 3m - 3 (6) 2 (4) 5 (10) 6m (n=52) 5 Primary Primary 0.6170 0.4949 1.0000 1.0000 2 (4) 2 (4) 30dy 8 (17) 5 (11) (n=47) Biological Biological 3m p-value 0.4674 0.4747 0.7522 0.2443 1.0000 0 3 (6) 30dy 7 (14) 7 (14) (n=50) Primary Primary 30dy 0.3611 0.1405 0.6808 0.1608 0.4780 7dy 1 (2) 2 (4) 1 (2) 8 (17) (n=46) Biological Biological 7dy 1 (2) 1 (2) 2 (4) 7dy (n=50) 11 (22) Primary Primary 0.9139 1.0000 0.5712 0.6056 1.0000 -test or Fisher’s exact test. Corrected for repeated measurements within a patient the overall p-value=0.6765(according to generalized linear model) linear generalized to p-value=0.6765(according overall the a patient within measurements repeated for Corrected exact test. Fisher’s or -test 2 Overall ammation f in of signs other plus Erythema discharge haemoserous or Clear Pus discharge infection wound severe or Deep erent types of infectious complications over time over types complications infectious of erent Di f (n, %) ammation f in of signs other plus Erythema (n, %) discharge haemoserous or Clear (n, %) Pus discharge (n, %) infection wound severe or Deep patients. of n= number (p-value) infection wound Perineal Chi with tested are p-values

97 Chapter 5 0.5176 p-value* 0.86 12m (0.17) (n=38) Biological 0.84 12m (0.13) (n=41) Primary 0.9852 0.8336 0.4256 0.5099 0.8537 0.4038 0.5885 0.2745 0.4371 0.0609 p-value* 9m 0.85 (0.17) (n=35) Biological 12m 53 (9) (n=38) 46 (12) 76 (26) 55 (43) 79 (26) 83 (18) 79 (17) 80 (36) 66 (17) 64 (23) 9m Biological Biological 0.85 (0.12) (n=40) Primary 6m 0.86 (0.14) (n=36) 12m 53 (9) 8 (14) (n=40) 47 (10) 75 (21) 64 (39) 82 (24) 82 (24) 76 (38) 70 (19) 69 (18) Biological Primary Primary 6m 0.84 (0.12) (n=40) Primary 3m 51 (8) (n=44) 41 (12) 71 (24) 42 (41) 68 (30) 73 (23) 77 (16) 75 (41) 62 (18) 61 (18) Biological Biological 3m 0.81 (0.19) (n=41) Biological 3m 52 (9) (n=44) 43 (10) 70 (26) 44 (42) 70 (22) 74 (22) 78 (15) 69 (44) 66 (19) 70 (16) Primary Primary 3m 0.81 (0.15) (n=44) Primary 30 51 (9) 47 (8) (n=43) 0.66 Pre-op 84 (21) 62 (73) 76 (21) 81 (19) 76 (18) 78 (37) 65 (21) 62 (19) 30dy (0.23) Biological Biological (n=39) Biological 0.65 30dy (0.20) (n=43) Primary (n=47) Pre-op 50 (10) 44 (10) 78 (24) 49 (45) 70 (27) 74 (24) 77 (18) 69 (41) 66 (20) 66 (16) Primary Primary 0.84 (0.14) (n=41) Pre-op Biological

0.82 (0.15) (n=45) Pre-op Primary SF-36 version 2 SF-36 (mean, SD) component Mental (mean, SD) component Physical SD) (mean, functioning Physical SD) (mean, Physical Role SD) (mean, Bodily Pain Social SD) (mean, functioning SD) health (mean, Mental SD) (mean, emotional Role SD) (mean, Vitality SD) (mean, General health perception Questionnaires n= number of patients returning the questionnaires. Data are mean scores with (worst outcome) standard and deviation 100 e T (SD). (best outcome). main P-values are scores calculated using are linear mixed printed was of life ect e f model, quality of and time cant f in A signi with correction). (Bonferroni structure. cant f autoregressive Bold. Time signi and statistically a 0.005 of results than less randomisation p-value was considered Values testing were can for multiple used ects.*=Adjusted e f range as between f xed 0 with time were seen. SF-36 or 2 version randomisation ect Short e f 2™ version cant Form-36 of f survey© however randomisation no health 1992, signi 2003 demonstrated, health Survey for (3rd the ed.) SF-36v2 Health manual 4 and QM User’s using: (Dutch). Analysed Netherlands Standard, a Inc. Qualitymetric Trust LAB, Outcome medical assessment estimation. score Missing with f ware so scoring ed f certi 5d-5L QLQ utility Health SD) (mean, n= number of Data patients the returning are questionnaires. mean scores with deviation standard (SD). P-values are calculated using linear mixed model, with autoregressive structure. can Values range between -0.329(worst outcome) and 1(best outcome). Time and results randomisation were used as f xed ects.*=p-value e f cant. f of signi A statistically cant f 0.05 ect signi e f was of time considered and quality of life was demonstrated, however no cant f ect signi e f of or randomisation with EQ-5D-5L user guide. randomisation the using time 5D 5L. Analysed were EQ5D5L= EuroQol seen.

98 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer 0.0975 0.8117 0.8129 0.4402 0.9443 0.7657 0.7544 0.9043 0.9566 0.4216 0.9944 0.8791 0.7572 0.6939 0.9598 p-value* 12m 0 (0) 2 (8) 6 (17) 4 (14) 5 (10) (n=38) 75 (20) 84 (22) 77 (26) 92 (12) 89 (16) 84 (22) 21 (20) 14 (22) 14 (20) 11 (22) Biological Biological 12m 1 (6) 2 (7) 3 (12) 5 (12) 6 (15) (n=40) 82 (14) 87 (16) 83 (27) 89 (14) 90 (16) 88 (23) 20 (20) 11 (18) 11 (23) 13 (20) Primary Primary 9m 2 (9) 4 (11) 9 (20) (n=35) 77 (20) 85 (23) 76 (29) 88 (15) 88 (17) 87 (22) 24 (23) 12 (20) 10 (19) 16 (22) 10 (25) 10 (23) Biological Biological P-values are calculated using linear mixed model, model, mixed linear using calculated are P-values 9m 31 1 (3) 2 (7) 3 (9) 8 (18) 3 (10) 5 (18) (n=40) 81 (12) 86 (16) 83 (20) 90 (13) 91 (15) 90 (17) 18 (18) 15 (16) 17 (24) Primary Primary 6m 1 (5) 3 (9) 7 (20) 4 (14) 7 (18) 9 (23) (n=38) 75 (14) 84 (20) 78 (25) 86 (13) 90 (13) 85 (22) 23 (20) 14 (20) 17 (22) Biological Biological 6m 0 (3) 1 (5) 2 (7) 8 (23) 5 (14) (n=42) 80 (14) 85 (18) 79 (21) 89 (14) 92 (14) 88 (18) 20 (20) 12 (17) 11 (19) 12 (23) Primary Primary 3m

2 (7) 5 9 (18) 9 (13) 4 (13) 8 (18) (n=41) 71 (18) 80 (22) 65 (30) 90 (12) 91 (12) 75 (28) 30 (20) 25 (31) 14 (24) 17 (23) Biological Biological 3m 3 (12) 7 (18) 5 (18) (n=44) 76 (14) 81 (22) 70 (28) 84 (17) 88 (16) 81 (21) 28 (21) 22 (23) 22 (30) 11 (21) 13 (29) 11 (20) Primary Primary 30dy 8 (15) 9 (21) (n=39) 53 (23) 65 (23) 38 (27) 77 (21) 80 (17) 62 (29) 51 (26) 47 (33) 15 (25) 38 (32) 32 (32) 10 (20) 11 (26) Biological Biological 30dy 9 (21) 6 (15) (n=43) 59 (21) 61 (27) 36 (29) 76 (24) 83 (22) 62 (33) 54 (27) 10 (20) 50 (26) 12 (21) 36 (35) 27 (38) 10 (10) Primary Primary 5 (15) 8 (18) 9 (19) 6 (18) 9 (21) 6 (15) (n=43) Pre-op 78 (18) 88 (20) 83 (24) 82 (15) 90 (16) 89 (19) 21 (22) 13 (20) 21 (24) Biological Biological 3 (10) 6 (16) 7 (18) (n=49) Pre-op 75 (18) 85 (20) 75 (28) 81 (18) 84 (18) 26 (21) 17 (22) 25 (29) 16 (24) 10 (23) 18 (30) 89 (150 Primary Primary

(mean, SD) (mean, (mean, SD) (mean, (mean, SD) (mean, (mean, SD) (mean, (mean, SD) (mean, EORTC-C30 Global health status functioning Physical functioning Role functioning Emotional Cognitive functioning Social SD) (mean, functioning SD) (mean, Fatigue SD) (mean, vomiting and Nausea SD) (mean, Pain Dyspnoea SD) (mean, SD) (mean, Insomnia SD) (mean, loss Appetite SD) (mean, Constipation SD) Diarrhoea (mean, SD) di f culties (mean, Financial n= number of patients the Data returning are questionnaires. mean scores with deviation standard (SD). e T main user guide. scores EORTC are the using printed in Analysed centre. Bold. can data Values range EORTC between UNIT, 0 life of CR30 quality (worst EORTC outcome). 100 (best and outcome) seen. were time with randomisation or randomisation of ect e f cant f signi no with however structure. autoregressive Time and demonstrated, was randomisation results life were of used quality and as time f xed of ect e f ects. e f cant f signi *=Adjusted A for correction). multiple testing (Bonferroni a cant f p-value signi of less than 0.003 was considered statistically

99 Chapter 5 0.3744 0.3744 0.3860 0.4196 0.2771 0.1286 0.2137 0.4017 0.6972 0.5361 0.4471 0.1518 0.2184 p-value* 12m 8 (9) 27 (4) 7 (16) 5 (20) 4 (18) 4 (10) (n=38) 82 (18) 72 (27) 67 (38) 20 (16) 12 (18) 79 (33) 42 (32) Biological Biological 12m 21 (3) 7 (15) 9 (21) 4 (17) 7 (20) 3 (12) 10 (9) (n=41) 79 (16) 71 (32) 96 (12) 26 (19) 73 (33) 17 (41) Primary Primary 9m 1 (6) 9 (11) 4 (11) 7 (18) 6 (14) (n=35) 75 (28) 76 (19) 72 (25) 67 (38) 21 (21) 10 (19) 10 (18) 28 (33) Biological Biological 9m 1 (5) 12 (9) 5 (14) 5 (20) 3 (12) 6 (14) (n=40) 82 (19) 85 (11) 69 (23) 89 (17) 28 (20) 11 (21) 77 (32) Primary Primary P-values are calculated using linear mixed model, model, mixed linear using calculated are P-values 1 6m 1 (6) 8 (15) 8 (17) 9 (25) 8 (20) (n=36) 77 (29) 79 (18) 74 (27) 93 (15) 28 (20) 10 (13) 79 (28) 17 (25) Biological Biological 6m 3 (6) 5 (12) 5 (14) 8 (16) (n=39) 79 (22) 80 (16) 68 (33) 87 (17) 25 (18) 12 (10) 14 (25) 74 (34) 13 (25) Primary Primary 3m 9 (22) 7 (18) 4 (17) (n=41) 74 (29) 78 (19) 78 (23) 86 (26) 23 (16) 18 (16) 10 (17) 19 (27) 76 (34) 25 (50) Biological Biological 3m 0 (0) 4 (13) 8 (22) 62 (40 (n=44) 82 (20) 81 (20) 74 (36) 94 (13) 27 (22) 20 (14) 11 (20) 11 (21) 12 (26) Primary Primary 30dy 4 (13) 9 (20) (n=39) 76 (28) 71 (20) 87 (19) 100 (0) 33 (19) 18 (24) 22 (23) 28 (31) 14 (25) 84 (27) 33 (58) Biological Biological 0 (0) 30dy 4 (13) (n=43) 79 (24) 73 (21) 80 (23) 92 (15) 36 (21) 32 (16) 18 (24) 19 (29) 21 (31) 17 (29) 57 (39) Primary Primary 0 (0) 7(15) 9 (12) 6 (17) 8 (18) (n=43) Pre-op 89 (15) 71 (17) 70 (29) 91 (24) 23 (18) 17 (20) 10 (19) 23 (32) Biological Biological 3 (9) 0 (0) 9 (21) 8 (17) 6 (14) (n=47) Pre-op 90 (14) 75 (14) 72 (24) 92 (15) 21 (17) 14 (16) 13 (20) 31 (31) Primary Primary (mean, SD) (mean, (mean, SD) (mean, (mean, SD) (mean, (mean, SD) (mean, EORTC-CR29 Body image Anxiety men Sexual interest women Sexual interest SD) (mean, problems Micturition SD) (mean, pelvic pain and Abdominal SD) (mean, feeling Bloated Dry SD) (mean, mouth SD) (mean, loss Hair SD) (mean, taste with Trouble SD) (mean, problems related Stoma SD) (mean, Impotence SD) (mean, Dyspareunia between can 0 range (SD). deviation (worst e in T Data with mean are scores printed Bold. are scores main standard Values the guide. questionnaires. user n= returning of number patients EORTC the using Analysed centre. data EORTC UNIT, life of quality CR29 EORTC outcome). (best 100 and outcome) structure. results and were Time with randomisation used autoregressive were time as with f xed ects. e f randomisation or for *=Adjusted testing a multiple randomisation p-value of ect of e f less cant f than signi no 0.003 however was demonstrated, considered statistically was life of quality and time of ect f e cant f signi A correction). (Bonferroni cant f signi seen.

100 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

Methods updated systematic review8 Search strategy update systematic review8: Electronic databases; Medline, EMBASE (search date 25-09-2015) Search: PubMed 148 hits (limitation: data search 26-03-2014-25-09-2015): ((“rectal neoplasms”[MeSH Terms] OR (“rectal”[All Fields] AND “neoplasms”[All Fields]) OR “rectal neoplasms”[All Fields] OR (“rectal”[All Fields] AND “neoplasm”[All Fields]) OR “rectal neoplasm”[All Fields]) OR (“rectal neoplasms”[MeSH Terms] OR (“rectal”[All Fields] AND “neoplasms”[All Fields]) OR “rectal neoplasms”[All Fields]) OR (“rectal neoplasms”[MeSH Terms] OR (“rectal”[All Fields] AND “neoplasms”[All Fields]) OR “rectal neoplasms”[All Fields] OR (“rectal”[All Fields] AND “cancer”[All Fields]) OR “rectal cancer”[All Fields]) OR ((“administration, rectal”[MeSH Terms] OR (“administration”[All Fields] AND “rectal”[All Fields]) OR “rectal administration”[All Fields] OR “rectal”[All Fields]) AND (“tumour”[All Fields] OR “neoplasms”[MeSH Terms] OR “neoplasms”[All Fields] OR “tumor”[All Fields]))) AND (((“abdomen”[MeSH Terms] OR “abdomen”[All Fields] OR “abdominal”[All Fields]) AND (“perineum”[MeSH Terms] OR “perineum”[All Fields] OR “perineal”[All Fields]) AND resection[All Fields]) OR (Abdominoperineal[All Fields] AND resection[All Fields]) OR (abdominoperineal[All Fields] AND excision[All Fields]) OR ((“abdomen”[MeSH Terms] OR “abdomen”[All Fields] OR “abdominal”[All Fields]) AND (“perineum”[MeSH Terms] OR “perineum”[All Fields] OR “perineal”[All Fields]) AND excision[All Fields]) OR (“colorectal surgery”[MeSH Terms] OR (“colorectal”[All Fields] AND “surgery”[All Fields]) OR “colorectal surgery”[All Fields]) OR ((“perineum”[MeSH Terms] OR “perineum”[All Fields]) AND (“surgery”[Subheading] OR “surgery”[All Fields] OR “surgical procedures, operative”[MeSH Terms] OR (“surgical”[All Fields] AND “procedures”[All Fields] AND “operative”[All Fields]) OR “operative surgical procedures”[All Fields] OR “surgery”[All Fields] OR “general surgery”[MeSH Terms] OR (“general”[All Fields] AND “surgery”[All Fields]) OR “general surgery”[All Fields])) OR ((“perineum”[MeSH Terms] OR “perineum”[All Fields] OR “perineal”[All Fields]) AND (“surgery”[Subheading] OR “surgery”[All Fields] OR “surgical procedures, operative”[MeSH Terms] OR (“surgical”[All Fields] AND “procedures”[All Fields] AND “operative”[All Fields]) OR “operative surgical procedures”[All Fields] OR “surgery”[All Fields] OR “general surgery”[MeSH Terms] OR (“general”[All Fields] AND “surgery”[All Fields]) OR “general 5 surgery”[All Fields]))) AND ((“wound healing”[MeSH Terms] OR (“wound”[All Fields] AND “healing”[All Fields]) OR “wound healing”[All Fields]) OR (primary[All Fields] AND (“wound healing”[MeSH Terms] OR (“wound”[All Fields] AND “healing”[All Fields]) OR “wound healing”[All Fields] OR “healing”[All Fields])) OR (“wound healing”[MeSH Terms] OR (“wound”[All Fields] AND “healing”[All Fields]) OR “wound healing”[All Fields] OR “repair”[All Fields]) OR (“reconstructive surgical procedures”[MeSH Terms] OR (“reconstructive”[All Fields] AND “surgical”[All Fields] AND “procedures”[All Fields]) OR “reconstructive surgical procedures”[All Fields] OR “reconstruction”[All Fields]) OR (“wounds and injuries”[MeSH Terms] OR (“wounds”[All Fields] AND “injuries”[All Fields]) OR “wounds and injuries”[All Fields] OR “wound”[All Fields]) OR (“wound healing”[MeSH Terms] OR (“wound”[All Fields] AND “healing”[All Fields]) OR “wound healing”[All Fields] OR “healing”[All Fields]))

EMBASE 53 hits (Limitations: date search: 2014-2015, journal articles): (exp rectum tumor/ OR rectal neoplasm.mp. OR rectal neoplasms.mp. OR rectal cancer.mp. OR exp rectum cancer/) AND (Abdominal perineal resection.mp. OR abdominoperineal resection.mp. OR exp rectum abdominoperineal resection/ OR abdominoperineal excision.mp.OR abdominal perineal excision.mp. OR perineum surgery.mp. OR perineal surgery.mp. OR exp colorectal surgery/) AND (exp wound healing/ OR primary healing.mp. OR repair.mp. OR reconstruction.mp.OR exp wound/ OR exp healing/)

Afer screening of title and abstract: ten articles1,7,14,21,25,32-36 Excluded afer reading full text: seven articles Reason for exclusion • a modifed eAPR was performed (n=1)35 • no separate data on perineal wound infection and eAPR, biological mesh or primary wound closure was provided (n=1)1 • no separate data on perineal wound infection and primary perineal wound closure, biological mesh assisted perineal wound closure or perineal myocutaneous fap closure was provided (n=3)34,36,7 • not all patients had rectal cancer (n=1)33 • the type of APR was not specifed (n=1)32

Statistical analysis Te updated results were combined with our previous published systematic review with meta-analysis.8 Te averages described are a weighted average of the logit proportions by the use of the generic inverse variance method. Te logit proportions that were obtained were back-transformed to the summary estimate and 95% confdence intervals in a summary proportion. Heterogeneity was described using the I2-test. Te Newcastle Ottawa quality assessment scale for cohort studies was used to assess the risk of bias.36 101 Chapter 5 + +

+ +

(range) Months Months Follow-up 31 (19-56) 44 (18-68) (range) Months Months Follow-up 44 (18-68) 28 (14-56) 0 4 (5) n(%) hernia Incisional Incisional n(%) 2 (11) 3 (10) hernia Incisional Incisional n(%) PWH 8 (57) 73 (88) n(%) PWH 12 (63) Perineal wound wound Perineal n(%)* 6 (43) cPWP 10 (12) Perineal wound wound Perineal n(%)* 7 (37) cPWP 12 (48) - n(%)* 5 (36) dPWP - - n(%)* dPWP 1 (7) n(%)* sPWP - 10 (12) n(%)* 7 (37) sPWP n(%) - 10 (71) 83 (100) n(%) Concomitant chemotherapy 19 (100) Concomitant chemotherapy n(%) - 10 (71) Neoadjuvant treatment Neoadjuvant 83 (100) n(%) Radiotherapy Neoadjuvant treatment Neoadjuvant 19 (100) Radiotherapy - Yes - (Yes/no) Yes antibiotics Preoperative Preoperative (Yes/no) antibiotics Preoperative Preoperative No Yes series No Yes series Consecutive Consecutive Consecutive Consecutive Study Study design Cohort Cohort Study Study design Cohort Cohort Study characteristics Study n 14 Study characteristics Study n 83 of 102 83 of 19 of 102 19 of 136 25 of year 2015 2014 year 2014 2014 25 21 14 14 er eAPR with eAPR primaryer rectalf perineal a cancer for closure problems wound Perineal Included studies author First et al. Han et al. Musters perineal Deep dPWP: days. 30 within dehiscence) and (infections problems wound perineal cial f Super sPWP: resection. abdominoperineal Extralevator eAPR: follow-up. Median +: abscesses) 30 within and perineal presacral abscesses) abscesses dehiscence, (perineal days. and cPWP:wound (infections, problems presacral perineal wound problems combined healing. wound perineal PWH: primary present. was complication one than more if once counted only was *: A patient 30 days. within rectal cancer for eAPR er Biological the perineumf a of closure assisted mesh Included studies author First et al. Dinnewitzer et al. Han +: Median follow-up. eAPR: Extralevator resection. abdominoperineal sPWP: cial f Super perineal wound problems (infections and dehiscence) within perineal 30 wound problems (perineal and days. abscesses) presacral within 30 dPWP: days. cPWP: Deep perineal combined wound problems (infections, dehiscence, perineal abscesses and healing. presacral wound perineal PWH: primary present. was complication one than more if once counted only was *: A patient 30 days. within abscesses)

102 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

Assessment risk of bias

Newcatle Ottawa quality assessment scale37 cohort studies Included Studies Selection Comparability Outcome Total (0-4) (0-2) (0-3) (0-9) Dinnewitzer et al.25 ** - *** 5 Han et al.14 *** - ** 5 Musters et al.21 *** ** ** 7

Southampton wound score system

Class Grade Description Appearance Normal wound healing 0 Normal healing I Normal healing with mild bruising or erythema A Some bruising B Considerable bruising C Mild erythema Minor wound II Erythema plus other signs of infammation infammation A At one point 5 B Around sutures C Along wound D Around wound III Clear of haemoserous discharge A At one point only (<2cm) B Along wound (>2cm) C Large volume D Prolonged (>3days) Wound Infection IV Pus A At one point only (<2cm) B Along wound (>2cm) Major wound V Deep or severe wound infection with or without tissue breakdown, complication hematoma requiring aspiration

103 Chapter 5

REFERENCE LIST

1. Asplund D, Prytz M, Bock D, Haglind E, Angenete E. Persistent perineal morbidity is common following abdominoperineal excision for rectal cancer. Int J Colorectal Dis 2015; 30(11): 1563-70.

2. Weitz J, Koch M, Debus J, Hohler T, Galle PR, Buchler MW. Colorectal cancer. Lancet 2005; 365(9454): 153-65.

3. den Dulk M, Putter H, Collette L, et al. Te abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of fve European randomised clinical trials on rectal cancer. Eur J Cancer 2009; 45(7): 1175-83.

4. Zhou X, Sun T, Xie H, Zhang Y, Zeng H, Fu W. Extralevator abdominoperineal excision for low rectal cancer: a systematic review and meta-analysis of the short-term outcome. Colorectal disease 2015; 17(6): 474-81.

5. van Leersum N, Martijnse I, den Dulk M, et al. Diferences in circumferential resection margin involvement afer abdominoperineal excision and low anterior resection no longer signifcant. Annals of surgery 2014; 259(6): 1150-5.

6. McCarthy K, Pearson K, Fulton R, Hewitt J. Pre-operative chemoradiation for non-metastatic locally advanced rectal cancer. Cochrane Database Syst Rev 2012; 12: CD008368.

7. Prytz M, Angenete E, Bock D, Haglind E. Extralevator Abdominoperineal Excision for Low Rectal Cancer- Extensive Surgery to be Used With Discretion Based on 3-Year Local Recurrence Results: A Registry- based, Observational National Cohort Study. Annals of surgery 2015.

8. Musters GD, Buskens CJ, Bemelman WA, Tanis PJ. Perineal wound healing afer abdominoperineal resection for rectal cancer: a systematic review and meta-analysis. Dis Colon Rectum 2014; 57(9): 1129-39.

9. Foster JD, Pathak S, Smart NJ, et al. Reconstruction of the perineum following extralevator abdominoperineal excision for carcinoma of the lower rectum: a systematic review. Colorectal disease 2012; 14(9): 1052-9.

10. Butt HZ, Salem MK, Vijaynagar B, Chaudhri S, Singh B. Perineal reconstruction afer extra-levator abdominoperineal excision (eLAPE): a systematic review. Int J Colorectal Dis 2013; 28(11): 1459-68.

11. Bailey IS, Karran SE, Toyn K, Brough P, Ranaboldo C, Karran SJ. Community surveillance of complications afer hernia surgery. BMJ 1992; 304(6825): 469-71.

12. Musters GD, Bemelman WA, Bosker RJ, et al. Randomized controlled multicentre study comparing biological mesh closure of the pelvic foor with primary perineal wound closure afer extralevator abdominoperineal resection for rectal cancer (BIOPEX-study). BMC Surg 2014; 14: 58.

13. De Broux E, Parc Y, Rondelli F, Dehni N, Tiret E, Parc R. Sutured perineal omentoplasty afer abdominoperineal resection for adenocarcinoma of the lower rectum. Dis Colon Rectum 2005; 48(3): 476-81.

14. Han JG, Wang ZJ, Qian Q, et al. A prospective multicenter clinical study of extralevator abdominoperineal resection for locally advanced low rectal cancer. Dis Colon Rectum 2014; 57(12): 1333-40.

15. Nissan A, Guillem JG, Paty PB, et al. Abdominoperineal resection for rectal cancer at a specialty center. Dis Colon Rectum 2001; 44(1): 27-35.

16. Han JG, Wang ZJ, Gao ZG, Xu HM, Yang ZH, Jin ML. Pelvic foor reconstruction using human acellular dermal matrix afer cylindrical abdominoperineal resection. Dis Colon Rectum 2010; 53(2): 219-23.

104 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

17. Vaughan-Shaw PG, King AT, Cheung T, et al. Early experience with laparoscopic extralevator abdominoperineal excision within an enhanced recovery setting: analysis of short-term outcomes and quality of life. Ann R Coll Surg Engl 2011; 93(6): 451-9.

18. Choi JJ, Palaniappa NC, Dallas KB, Rudich TB, Colon MJ, Divino CM. Use of mesh during ventral hernia repair in clean-contaminated and contaminated cases: outcomes of 33,832 cases. Annals of surgery 2012; 255(1): 176-80.

19. Novitsky YW. Biology of biological meshes used in hernia repair. Surg Clin North Am 2013; 93(5): 1211-5.

20. Bebenek M. Abdominosacral amputation of the rectum for low rectal cancers: ten years of experience. Ann Surg Oncol 2009; 16(8): 2211-7.

21. Musters GD, Sloothaak DA, Roodbeen S, van Geloven AA, Bemelman WA, Tanis PJ. Perineal wound healing afer abdominoperineal resection for rectal cancer: a two-centre experience in the era of intensifed oncological treatment. Int J Colorectal Dis 2014; 29(9): 1151-7.

22. Toshniwal S, Perera M, Lloyd D, Nguyen H. A 12-year experience of the Trendelenburg perineal approach for abdominoperineal resection. ANZ J Surg 2013; 83(11): 853-8.

23. Christensen HK, Nerstrom P, Tei T, Laurberg S. Perineal repair afer extralevator abdominoperineal excision for low rectal cancer. Dis Colon Rectum 2011; 54(6): 711-7.

24. Dalton RS, Smart NJ, Edwards TJ, Chandler I, Daniels IR. Short-term outcomes of the prone perineal approach for extra-levator abdomino-perineal excision (elAPE). Surgeon 2012; 10(6): 342-6. 25. Dinnewitzer A, Meissnitzer M, Meissnitzer T, et al. Dynamic magnetic resonance imaging evaluation 5 of pelvic reconstruction with porcine dermal collagen mesh following extra-levator abdominoperineal excision for primary rectal cancer. Int J Colorectal Dis 2015; 30(4): 491-6.

26. Han JG, Wang ZJ, Wei GH, Gao ZG, Yang Y, Zhao BC. Randomized clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. Am J Surg 2012; 204(3): 274-82.

27. Jensen KK, Rashid L, Pilsgaard B, Moller P, Wille-Jorgensen P. Pelvic foor reconstruction with a biological mesh afer extralevator abdominoperineal excision leads to few perineal hernias and acceptable wound complication rates with minor movement limitations: single-centre experience including clinical examination and interview. Colorectal Dis 2014; 16(3): 192-7.

28. Peacock O, Simpson JA, Tou SI, et al. Outcomes afer biological mesh reconstruction of the pelvic foor following extra-levator abdominoperineal excision of rectum (APER). Tech Coloproctol 2014; 18(6): 571-7.

29. Pascual G, Sotomayor S, Perez-Lopez P, Bujan J, Bellon JM. Long term behavior of biological prostheses used as abdominal wall substitutes. Histol Histopathol 2014; 29(1): 139-49.

30. van Hout B, Janssen MF, Feng YS, et al. Interim scoring for the EQ-5D-5L: mapping the EQ-5D-5L to EQ- 5D-3L value sets. Value in health 2012; 15(5): 708-15.

31. Aaronson NK, Ahmedzai S, Bergman B, et al. Te European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993; 85(5): 365-76.

32. Davidge KM, Raghuram K, Hofer SO, et al. Impact of fap reconstruction on perineal wound complications following ablative surgery for advanced and recurrent rectal cancers. Annals of surgical oncology 2014; 21(6): 2068-73.

105 Chapter 5

33. Harries RL, Luhmann A, Harris DA, Shami JA, Appleton BN. Prone extralevator abdominoperineal excision of the rectum with porcine collagen perineal reconstruction (Permacol): high primary perineal wound healing rates. International journal of colorectal disease 2014; 29(9): 1125-30.

34. Ortiz H, Ciga MA, Armendariz P, et al. Multicentre propensity score-matched analysis of conventional versus extended abdominoperineal excision for low rectal cancer. Te British journal of surgery 2014; 101(7): 874-82.

35. Park EJ, Baik SH, Kang J, et al. Short-term outcomes of the modifed extralevator abdominoperineal resection for low rectal cancer (with videos). Surgical endoscopy 2015.

36. Perdawood SK, Lund T. Extralevator versus standard abdominoperineal excision for rectal cancer. Techniques in coloproctology 2015; 19(3): 145-52.

37. Te Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses. 27-09-2015 2008.

106 Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer

5

107

CHAPTER 6

Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection; a systematic review

G.D. Musters, J.W.A. Burger, C.J. Buskens, W.A. Bemelman, and P.J. Tanis

World J Surg 2015; 39(11): 2786-94 Chapter 6

ABSTRACT

Aim: Use of topical antibiotics to improve perineal wound healing afer abdominoperineal resection (APR) is controversial. Te aim of this systematic review was to determine the impact of local application of gentamicin on perineal wound healing afer APR.

Methods: Te electronic databases Pubmed, EMBASE and Cochrane library were searched in January 2015. Perineal wound outcome was categorized as infectious complications, non-infectious complications and primary perineal wound healing.

Results: From a total of 582 articles, eight studies published between 1988 and 2012 were included: four randomised controlled trials (RCTs), three comparative cohort studies and one cohort study without control group. Gentamicin was administered using sponges (n=3), beads (n=4), and by local injection (n=1). Tere was substantial heterogeneity regarding underlying disease, defnition of outcome parameters and timing of perineal wound evaluation among the included studies, which precluded meta-analysis with pooling. Regarding infectious complications, three of six evaluable studies demonstrated a positive efect of local application of gentamicin: one of four RCTs and both comparative cohort studies. Only two RCTs reported on non-infectious complications, showing no signifcant impact of gentamicin sponge. All three comparative cohort studies demonstrated a signifcantly higher percentage of primary perineal wound healing afer local application of gentamicin beads, but only one out of three evaluable RCTs did show a positive efect of gentamicin sponges.

Conclusion: Currently available evidence does not support perineal gentamicin application afer APR.

110 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection

INTRODUCTION

Impaired perineal wound healing afer abdominoperineal resection (APR) is a clinically signifcant problem. Perineal wound infection, wound dehiscence and deep pelvic abscess are frequently encountered complications, which ofen require intensive and long lasting wound care and interfere with quality of life. Furthermore, modifed approaches for distal rectal cancer have increased the incidence of perineal wound problems afer APR, especially the use of neoadjuvant radiotherapy.24 Surgical site infections, like perineal wound infection, may result in prolonged hospitalization, readmission, increase in health care costs and even an increase in mortality.25 Te cornerstone in the prevention of wound infection is the use of systemic and/or oral prophylactic antibiotics. Despite prophylactic antibiotics, perineal wound infection have been reported to occur in more than half of the patients who undergo an APR afer neoadjuvant radiotherapy for rectal cancer.26 For the antibiotics to be efective, a high antibiotic concentration at the designated site is warranted. However, systemic administration hardly efects perineal wound healing. Tis is most likely due to dead space with retention of fuids and surgically impaired blood supply. Perineal application leads to high local concentrations with low serum levels and hence low systemic adverse reactions.27 In general, local gentamicin application is regarded as safe, easy to use and inexpensive.28 However, there is still no consensus on the clinical value of topical gentamicin to improve perineal wound healing in patients undergoing APR. Terefore, we conducted a systematic review of the literature on the efect of local gentamicin in preventing perineal wound infection and improving perineal wound healing afer APR. 6 METHODS

Search strategy All studies evaluating the impact of local application of gentamicin on perineal wound healing afer APR were considered eligible. Inclusion was not restricted to study design (i.e. randomised controlled trial (RCT), prospective or retrospective cohort study (with or without control group) and underlying disease (i.e. rectal cancer, anal cancer, and infammatory bowel disease). Animal studies, systematic reviews, non-English articles, case series (<5 patients), and congress abstracts were excluded. Studies were identifed by searching electronic databases and scanning of reference lists. Tis search was applied in PubMed (1971-2015), EMBASE (1982-2015) and Cochrane library (2005) databases in January 2015. Te following medical subject heading (MESH) terms were used; gentamicin’s, aminoglycosides, anti-bacterial agents, perineum, abdomen, colorectal surgery, surgical procedures operative, general surgery. Other search terms were abdominal perineal resection, abdominoperineal resection, abdominoperineal excision, abdominal perineal excision, perineum surgery, perineal surgery, gentamicin, gentamycin, aminoglycoside antibiotics, and local antibiotics. No language or publication date restrictions were applied. Te candidate publication titles and abstracts were screened to exclude nonrelated publications. Secondly, the full text of the remaining manuscripts was read to determine whether they were eligible for inclusion.

111 Chapter 6

Validity assessment and assessment of eligibility Methodological quality and risk of bias were assessed for the included studies. For cohort studies, the Newcastle Ottawa quality assessment scale for cohort studies was used to assess risk of bias.29 Te quality items scored were as follows: representativeness of the exposed cohort, selection of the non-exposed cohort, ascertainment of exposure, the absence of outcome of interest at the start of the study, validity of the design or analysis, assessment of outcome, duration of follow-up, and lost to follow-up. For RCTs the Jadad scoring system was used to assess the risk of bias.30

Data extraction and data analysis Data were extracted from the included studies by two independent investigators (G.M., P.T.). Disagreements were resolved by discussion between the two reviewers. For each article, the following data were collected: year of publication, study design, number of patients, underlying disease, comorbidities, type of APR, neoadjuvant radiotherapy, perineal closure technique, perineal drainage, systemic antibiotic use, administration method and dosage of locally applied gentamicin, defnitions of outcome parameters, follow-up period, type and incidence of perineal wound complications, primary perineal wound healing rate, reoperation rate, and hospital stay. Meta- analysis was not intended because of the heterogeneity among studies regarding study population, variety in gentamicin application and non-uniform defnitions of outcome parameters.

RESULTS

Te systematic search resulted in a total of 582 articles. Afer removal of duplicates among the diferent databases, 527 articles remained. Afer screening of publication titles and abstracts, 17 publications were retrieved for full text review. Of these 17 publications, 10 studies were excluded, because gentamicin was not placed in the perineal wound (n=3)31-33, no separate data for perineal and abdominal wound outcome were provided (n=4)34-37, surgery was not an APR procedure (n=1)38, use of bactaracin spray instead of gentamicin (n=1)39, and the type of local antibiotic was not described (n=1)40. One additional study was included afer crosschecking references of the included studies. In total, eight studies were included of which four were RCTs, three were comparative cohort studies and the remaining study was a non-comparative cohort study. Afer exclusion of patients who underwent non-APR surgery from two cohort studies41,42, a total number of 602 patients were included in this systematic review. Te risk of bias of the included studies is displayed in appendix 1.

Patient characteristics Patient characteristics of each included study are displayed in Table 1. APR was performed for cancer in 68% (n=381) of the patients, which was primary rectal cancer in 82% (n=311; 5 studies)42-46, anal cancer in 4% (n=15; 2 studies)43,46, locally recurrent rectal or anal cancer in 1% (n=4; 1 study 43, sarcoma in one patient43, and unspecifed type of cancer in 13% (n=50; 2 studies) of the patients.39,47 In three studies, patients with infammatory bowel disease were also included with a specifed number of patients in two studies (total n=46)41,43,47. In the remaining patients, APR was performed for incontinence43, villous adenoma47, or unspecifed diseases.41,48 Comorbidities related

112 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection

to risk of wound infection were described in two studies.44,45 Use of neoadjuvant radiotherapy was described in three studies and was given to 21% (n=115) of the included patients.43-45 Type of neoadjuvant radiotherapy was only specifed in one study (5x5 Gy).44 One study described prior pelvic radiotherapy in four patients.43

Surgery and prophylactic antibiotics Te diferent types of surgery being performed in the included studies are specifed in Table 1. None of the studies described the use of an omentoplasty. Primary perineal wound closure was performed in all studies, of which four studies mentioned closure in multiple layers43-46. Routine pelvic drainage was described in seven studies41,42,44-48, and a perineal drain was selectively placed in one study.19 Type of drainage difered among study groups in one study, with irrigation-suction drainage in the control group, and either continuous or intermittent drainage in the local gentamicin beads group.48

Considering the type of local administration of gentamicin at the perineal wound, a collagen carrier was used in three studies43-45, gentamicin beads were used in four studies41,42,46,48, and gentamicin was locally injected in one study (Table 2).47 Dosage of locally applied gentamicin was provided in four studies. A total dose of 210 mg45, 200 mg43, and 160 mg47 was used in three studies. In the remaining study, the concentration of gentamicin was measured in the wound fuid with a mean value of 138.1 µg/l (n=7, standard deviation 58.4) at day one postoperatively.46 Four studies described placement of the local antibiotic in the sacral cavity44-46,48, while a more superfcial application in the perineal wound was reported in the other four studies.41-43,47 Seven of eight studies described routine use of diferent regimens of prophylactic pre-operative antibiotics, and fve studies also continued prophylactic antibiotics postoperatively, only in a subgroup of patients in one of these fve 6 studies (Table 2).

Perineal wound outcome Perineal wound outcome parameters were defned in six out of eight included studies, all being diferent among these six studies (Table 3). Time points for wound evaluation were provided in fve studies, which were also not uniform.

Perineal wound outcome was categorized as infectious complications, non-infectious complications and primary wound healing for the purpose of this review (Table 3). Regarding infectious complications, three of six evaluable studies demonstrated a positive efect of local application of gentamicin. Tis signifcant diference was found in one out of four RCTs45, and both comparative cohort studies.44,48 Te impact on infectious complications was found in two out of three studies using gentamicin sponge44,45 and in one out of two studies using gentamicin beads.48

Regarding non-infectious complications (i.e. seroma, hematoma, and dehiscence without infection), only two studies uniformly reported this outcome parameter. Tese two RCTs did not report a signifcant impact of gentamicin sponge on this perineal wound outcome parameter.43,45 One additional comparative cohort study reported a signifcant diference in the incidence of persistent perineal fstula in favour of gentamicin beads48, although a fstula may be also considered as a long- term consequence of an initial infectious problem.

113 Chapter 6

Primary perineal wound healing was the most consistently reported outcome parameter among seven included studies, including six comparative studies. However, the RCT of Collin et al. demonstrates that this outcome parameter is highly dependent on the time point of evaluation, ranging from 82% afer 7 days to 47% afer 1 month in the control group.43 All three comparative cohort studies demonstrated a signifcantly higher percentage of primary perineal wound healing afer local application of gentamicin42,44,48, but only one out of three evaluable RCTs did show a positive efect of local gentamicin.46 Te positive efect of local gentamicin on primary wound healing was found in one out of three studies using sponges44 and in all three comparative studies using beads.42,46,48 Signifcant improvements in perineal wound outcome afer local gentamicin were only observed in homogenous cancer populations, while the two RCTs containing subgroups of patients with IBD were both negative.43,47

Reoperation and hospital stay Te reoperation rate for perineal wound infection was described in three studies and ranged between 2-5% for the local application of gentamicin and ranged between 0-9% when no topical antibiotics were applied.43,45,46 Hospital stay was described in six studies (Table 3).41,43,44,46-48 Out of fve comparative studies, three studies reported a signifcant reduction in hospital stay (p<0.05).44,46,48 Te mean reduction in hospital stay was at least seven days.44,46,48

DISCUSSION

Te present systematic review of the literature on perineal application of gentamicin afer APR reveals that only a limited number of randomised and comparative cohort studies have been published, which are all of relatively low quality. Five of eight included studies were published more than 20 years ago. Most RCTs did not perform a power calculation and can be considered underpowered. Substantial heterogeneity was observed among the studies regarding underlying disease, surgical techniques, type of administration of gentamicin, antibiotic dosage, and peri-operative use of systemic antibiotics, defnitions of outcome parameters and timing of perineal wound evaluation. In addition, the non-randomised comparative studies have a substantial risk of bias. Taking all these methodological shortcomings in mind, we conclude that there is no convincing evidence for routine perineal application of gentamicin afer APR. Statistically signifcant diferences in favour of local gentamicin were mainly observed in comparative cohort studies, while the majority of RCTs were negative for both wound infection and primary wound healing.

In contrast to the present study, a recent systematic review showed that local application of gentamicin in all clean and clean contaminated wounds signifcantly reduces surgical site infections.49 For patients undergoing colorectal surgery, several studies reported a positive efect of local gentamicin.33,36 However, gentamicin was mostly applied to the abdominal wound in these studies, without reporting on the use of topical antibiotics at the perineal wound with corresponding outcome. Tere are several potential reasons for a diference in wound healing between abdominal and perineal wounds, for example with regard to the degree of bacterial contamination, tension on wound edges, and the received radioactivity dose in case of neoadjuvant treatment.50,51 In addition,

114 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection

fuid may collect in the dead space above the closed perineum, and may become secondarily infected, with subsequent drainage through the perineum. Te abdominal pressure on the perineal wound afer (partially) resecting the pelvic foor may also negatively infuence perineal wound healing. Because of these diferences, our systematic review specifcally focused on the perineal wound.

Tere are several patients and surgery related factors infuencing perineal wound healing afer APR. Although the majority of patients were diagnosed with cancer, infammatory bowel disease was also included in three studies. Infammatory bowel disease patients signifcantly difer from cancer patients regarding perineal wound healing due to malnutrition, chronic pelvic infammation, use of immunosuppressive medication, and pre-existing perineal fstulas and sinuses.52 On the other hand, the use of pre-operative radiotherapy in rectal cancer patients may afect wound healing signifcantly.24 Surgical techniques may also infuence the occurrence of perineal wound infection and subsequent perineal wound healing. Te omentum for instance, is a major contributor to the local immune response and has the ability to promote angiogenesis.53,54 Terefore, placing an omentoplasty in the pelvic cavity has been suggested to improve primary perineal wound healing.55 Besides an omentoplasty, several types of musculocutaneous or perforator faps can be used for closure of the pelvic defect in selected patients undergoing extensive resection (i.e. exenterations) or salvage surgery for anal cancer.56-58 None of the patients in the included studies underwent fap reconstruction and the use of omentoplasty was not specifed. Terefore, the additive value of local gentamicin in combination with an autologous tissue fap or omentoplasty is unknown. Furthermore, none of the included studies described laparoscopic surgery, the use of vessel sealing equipment, or patient positioning which have been shown to be related to blood loss, operative time, and risk of surgical site infection. Tese procedural characteristics may be of infuence on perineal 6 wound healing and may hamper extrapolation of results of the present systematic review to current clinical practice.47,59

Among the diferent application techniques, sponges, beads and injection difer in local gentamicin concentration and release over time.27 Besides diferences in local antibiotic concentration, the collagen from which the gentamicin is delivered can have an infuence on perineal wound healing itself.60 Tis is most likely due to faster haemostasis and the protective efect on seroma and hematoma formation, which ultimately could afect primary perineal wound healing.60

Gentamicin is most efective against gram-negative bacteria and some gram-positive bacteria, with the exception of anaerobe bacteria. Nelson et al. described in a systematic review of 182 RCTs that the addition of antibiotics against anaerobe bacteria in colorectal surgery reduces surgical site infections.61 Based on these data, an antibiotic that is efective against anaerobic bacteria is currently most ofen being added as pre-operative prophylaxis.62 However, antibiotics efective against anaerobe bacteria are most ofen not locally applied in a sustained matter and combined with gentamicin. Terefore, the local application of gentamicin in combination with an antibiotic targeting anaerobe bacterium might enhance the efectiveness against perineal wound problems.

In conclusion, a restricted number of low quality randomised and non-randomised studies do not convincingly show a positive efect of local gentamicin on perineal wound infection or primary wound healing rate. APR for benign and malignant disease is essentially diferent and treatment

115 Chapter 6

of distal rectal cancer has changed signifcantly since publication of most of the included studies. Tis underlines the need for new RCTs on locally applied gentamicin in more representative and homogeneous patient groups. At present, available literature does not support the routine use of local gentamicin application for perineal wound healing following APR.

116 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection

n (%) Type of surgery of Type APR 38 (73) APR 14 (27) APR Intersphincteric 44 (88) APR 6 (12) APR Intersphincteric 19 (100) APR Standard 21 (100) APR Standard 49 (100) APR Miles 48 (100) APR Miles 22 (100) APR Loyd-Davies 22 (100) APR Loyd-Davies APR 34 (83) APR 7 (17) Proctocolectomy 37 (97) APR 1 (3) Proctocolectomy sacral rectum Abdominal 80 (100) amputation sacral rectum Abdominal 26 (100) amputation 67 (100) APR 42 (100) APR Abdominosacral 24 (100) proctectomy ------3(6) n (%) 8 (16) 4 (15) contamination Intra-operative Intra-operative (94) (93) e e ------0 1 (2) n (%) 19 (100) 21 (100) 37/39 41/44 Radiotherapy ------n (%) 7 (37) 8 (38) 14 (29) 10 (21) Smoking ------Comorbidity 4 (8) n (%) 3 (16) 5 (24) 7 (15) Diabetes - - 0 0 0 0 0 0 0 0 NS IBD n (%) 5 (10) 13 (25) 16 (39) 12 (32) 6 a disease - - NS Underlying Underlying n (%) 39 (75) 44 (88) 25 (61) 25 (66) Cancer 19 (100) 21 (100) 49 (100) 48 (100) 22 (100) 22 (100) 67 (100) 42 (100) ------32/20 29/21 12/07 13/08 11/11 12/10 22/19 19/19 (male/ (male/ Gender female) c c c c b b b b b b b,d b, d b, - - - Age 67 64 (years) 69 ± 9 71 ±10 Patients 65 (29-87) 67 (35-85) 54 (21-80) 62 (17-83) 62 (44-83) 63( 41-90) 67 (44-84) 64 (46-83) n 52 50 19 21 49 48 22 22 41 38 80 26 67 42 25 No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Local Local genta Year 2012 2008 2001 1991 1988 1988 1988 1988 RCT RCT RCT RCT Prospective Prospective Study design Study comparative comparative comparative comparative comparative cohort study cohort cohort study cohort study cohort Cohort study Retrospective Retrospective 24 22 19 21 25

20 23 18 er APR f a healing perineal on wound local of gentamicin the impact evaluating studies non-comparative and comparative of characteristics Study Table 1 Table Included studies author First Collin et al. De Bruin et al. et al. Gruessner Rosen et al. et al. Moesgaard Sachweh et al. Sachweh et al. Mühleder et al. Lütje b= age ± mean colitis). Ulcerative disease and disease (Crohn’s bowel ammatory f IBD= in recurrences. and cancer rectal cancer, anal trial. a=cancer: controlled RCT=randomised NS=not resection. abdominoperineal APR= patients. cancer for calculated only was group this e= unknown. distribution = d (range). age median c= (range). or deviation standard ed. f speci

117 Chapter 6

------2 days 2 days incision Postoperative Not specif ed Not specif ed Not n (%): 11 (21) n (%): 11 (22) day of operation of day operation of day Metro 3x500 mg; 3 days 3x500 mg; Metro Cefa 2 x 2 g, Metro 2x 500 mg; 2x 500 mg; Cefa 2 x g, Metro 2x 500 mg; Cefa 2 x g, Metro er f a 1x 2 g 12 hours Latamoxef Genta 3 x 80 mg, Metro 3 x 500 mg; 3 x 500 mg; Metro 3 x 80 mg, Genta 3 x 500 mg; Metro 3 x 80 mg, Genta Systemic antibiotics Systemic - - - Preoperative Latamoxef 1x 2g Latamoxef Prophylaxis not speci f ed not Prophylaxis speci f ed not Prophylaxis Augmentin 1x 1000/200 mg Augmentin 1x 1000/200 mg Augmentin Cefa 1x 2 g, Metro 1x 500 mg Cefa 1x 2 g, Metro Cefa 1 x 2 g, Metro 1 x 500 mg Cefa 1 x 2 g, Metro 1 x 500 mg Cefa 1 x 2 g, Metro 1 x 500 mg Cefa 1 x 2 g, Metro Genta 1x 80 mg, Metro 1x 500 mg Metro 1x 80 mg, Genta 1x 500 mg Metro 1x 80 mg, Genta Paro 4x 500 mg, Metro 3x 500 mg orally 3x 500 mg Metro 4x 500 mg, Paro ------Location Sacral cavity Sacral cavity Sacral cavity Sacral cavity Perineal wound Perineal wound Perineal wound Perineal wound Perineal

------Type Local antibiotics 3 per patient 3 per patient 400 mg metro 400 mg Genta-PMMA, Genta-PMMA, 1 chain per patient 1 chain per patient 1 chain 2 chains per patient 2 chains Genta sulphate sponge sponge sulphate Genta 2.0 mg/m2 (Collatamp) 2.0 mg/m2 Genta PMMA 30-beadsGenta Injection 160 mg genta / genta 160 mg Injection Genta sponge (Garacol), (Garacol), sponge Genta Genta sponge (Septcoll), (Septcoll), sponge Genta Genta-PMMA, 30 beads,Genta-PMMA, 30 beads,Genta-PMMA, (Septopal), 1 chain per patient 1 chain (Septopal), No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Local (Yes/no) n 52 50 19 21 49 48 22 22 41 38 80 26 67 42 25 24 22 19 21 25

20 23 18 Use of local and systemic antibiotics in patients undergoing APR with primary with APR perineal closure undergoing in patients antibiotics local systemic of and Use Included studies author First Collin et al. De Bruin et al. et al. Gruessner Rosen et al. et al. Moesgaard et al. Sachweh et al. Mühleder et al. Lütje Table 2 Table Genta=gentamicin. paromomycin. Paro= metronidazole. . Metro= cefazolin Cefa=

118 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection c c d d d d - - stay Days 25±18 27±12 15±8* 20±7* Hospital Hospital 11 (6-52) 10 (4-27) - - 2 (4) 3 (6) 1 (2) 0 (0) 1 (5) 2 (9) n (%) Perineal Perineal Reoperation n(%) 9 (43) 43 (88) 36 (75) 10 (45) 16 (84)* 19 (86)* 7 d: 43/52 (83) 7 d: 40/49 (82) 1 m: 28/52 (54) 3 m: 36/51 (71) 1 m: 23/49 (47) 3 m: 30/48 (63) Primary healing healing Primary 12 m: 43/48 (90) 12 m: 38/43 (88) - - - - 3 (6) 2 (4) 7 d: 2/52 (4) 7 d: 3/49 (6) Non-infectious Non-infectious 1 m: 14/52 (27) 3 m: 12/51 (24) 12 m: 5/48 (10) 1 m: 12/49 (24) 3 m: 17/48 (35) 12 m: 5/43 (12) complications n(%) complications Perineal wound outcome wound Perineal Infectious Infectious Deep: 1 (5) Sacral 2 (4) Deep: 6 (29) 3 m: 3/51 (6) 3 m: 1/48 (2) Abscess 1 (5) Abscess 2 (9) Abscess 7 d: 6/52 (12) 7 d: 6/49 (12) 12 m: 0/48 (0) 12 m: 0/43 (0) Perineal 1 (2)* Perineal Perineal 5 (10) Perineal 1 m: 10/52 (19) 1 m: 14/49 (29) f cial: 6 (29) Super f cial: 2 (11)* Super complications n(%) complications n 52 50 19 21 49 48 22 22 6 Study group Study Control Control Control Control Genta beads Type of genta of Type Genta sponge Genta sponge Genta sponge - 14 days 8 weeks 7-10 days 7, 14 days 1,3,12 months Time interval(s) Time =

= = no = no = = = cellulitis, = cellulitis, = skin and and = skin Perineal wound evaluation wound Perineal nition outcome parameters outcome nition De f Infectious complications Infectious and/or swelling redness, open and/or discharge purulent infected wound complication Non-infectious persistent wound, clean open stula / f sinus healing wound Perineal non- or infectious without complication infectious infection cial f Super deep tissue evidenceno of infection infection Deep with breakdown subcutaneous + sacral/deepinfection wound abscess healing wound Primary hematoma, seroma, of signs local dehiscence, wound no sepsis and or infection dressing. wound need for healing wound Primary without 14 days within healing dehiscence/need for or secretion dressing or local treatment CCS RCT RCT RCT Study Study design 22 21

20 23 er APR with primary perineal closure and local application of gentamicin gentamicin of local primary with er APR and perineal application f a closure outcome wound Perineal Table 3 Table Included studies author First Collin et al. De Bruin et al. et al. Gruessner Rosen et al.

119 Chapter 6

c f d g - - 44 20.4 / 16.6* e 25 (10-72) 24 22 (10-100) ------,b - - 8 (31) 62 (93)* 70 (88)* 23/39 (59) 23/24 (96) - - - - 1 (1) 3 (4)* 8 (31) stula stula f Persistent stula f Persistent stula f Persistent a - - - = median hospital stay (range);d = mean hospital stay ± standard deviation;

S: 2 (5) S: 0 (0) 18 (69) 10 (13)* IAA: 1 (2) IAA: 1 (2) IAA: PWI: 19 (46) PWI: 18 (48) 41 38 80 26 67 42 25 Genta Control Control Control injection Genta beads Genta beads Genta beads - - - 1,3 months At discharge At Daily during during Daily hospitalization = = positive blood = positive = no infection < infection = no = presence of pus, either either pus, of = presence Perineal wound infection infection wound Perineal (PWI) or spontaneously discharging drainage requiring (IAA) abscess Intra-abdominal or drainage surgical f edveri by aspiration ultrasound-guided (S) Septicaemia culture - - - healing Wound < 3 days fever no 14 days, CS CCS CCS RCT 24 19 25 18 Continuation of Table 3 Table of Continuation et al. Moesgaard et al. Sachweh et al. Mühleder et al. Lütje RCT=randomised controlled trial; CCS = cohort comparative study; CS = cohort study; *= cant f erence f statistically signi di between the 25% two versus 2%; drainage intermittent study b groups; = a 75% drainage continuous = versus 97%; drainage intermittent continuous drainage c e = continuous drainage, distribution unknown; f = intermittent drainage, distribution unknown; g = distribution unknown. PWI= perineal wound s=septicaemia. infection. abscess. IAA=intra-abdominal

120 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection

Appendix 1 Assessment risk of bias

Newcastle Ottawa quality assessment scale6 cohort studies Included Studies Jadad et al.7 Selection Comparability Outcome Total (0-5) (0-4) (0-2) (0-3) (0-9) Collin et al.20 2 - - - - De Bruin et al.21 - ** - * 3 Gruessner et al.22 1 - - - - Rosen et al.23 1 - - - - Lütje et al.18 - * - * 2 Sachweh et al.25 - * - * 2 Mühleder et al.19 - * - * 2 Moesgaard et al.24 2 - - - -

6

121 Chapter 6

REFERENCE LIST

1. Musters GD, Buskens CJ, Bemelman WA, Tanis PJ. Perineal Wound Healing Afer Abdominoperineal Resection for Rectal Cancer: A Systematic Review and Meta-analysis. Dis Colon Rectum 2014; 57(9): 1129-39. 2. Perencevich EN, Sands KE, Cosgrove SE, Guadagnoli E, Meara E, Platt R. Health and economic impact of surgical site infections diagnosed afer hospital discharge. Emerg Infect Dis 2003; 9(2): 196-203. 3. Nissan A, Guillem JG, Paty PB, et al. Abdominoperineal resection for rectal cancer at a specialty center. Dis colon rectum 2001; 44(1): 27-35. 4. Ruszczak Z, Friess W. Collagen as a carrier for on-site delivery of antibacterial drugs. Adv Drug Deliv Rev 2003; 55(12): 1679-98. 5. Friberg O, Dahlin LG, Levin LA, et al. Cost efectiveness of local collagen-gentamicin as prophylaxis for sternal wound infections in diferent risk groups. Scand Cardio vasc J 2006; 40(2): 117-25. 6. Wells GA, Shea B, O’Connell D, et al. Te Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Research institute. http://www.ohri.ca/programs/clinical_ epidemiology/oxford.asp. Accessed 1 januari 2015. 7. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomised clinical trials: is blinding necessary? Control ClinTrials 1996; 17(1): 1-12. 8. Bennett-Guerrero E, Pappas TN, Koltun WA, et al. Gentamicin-collagen sponge for infection prophylaxis in colorectal surgery. N Engl J Med 2010; 363(11): 1038-49. 9. Pochhammer J, Zacheja S, Schafer M. Subcutaneous application of gentamicin collagen implants as prophylaxis of surgical site infections in laparoscopic colorectal surgery: a randomised, double-blinded, three-arm trial. Arch Surg 2015; 400(1): 1-8. 10. Rutten HJ, Nijhuis PH. Prevention of wound infection in elective colorectal surgery by local application of a gentamicin-containing collagen sponge. Eur J Surg Suppl 1997; (578): 31-5. 11. Aubrey DA, Jenkins NH, Morgan WP, Tomas M. Te use of gentamicin-PMMA chains in abdominal surgery: a pilot study of prophylaxis against wound infection. Pharmatherapeutica 1986; 4(8): 536-40. 12. Morawiec Z, Pawlak M, Kolacinska A, Dziki A. Local antibiotic therapy in rectal cancer surgery. Pol Pr zegl Chir 2012; 84(12): 613-7. 13. Nowacki MP, Rutkowski A, Oledzki J, Chwalinski M. Prospective, randomised trial examining the role of gentamycin-containing collagen sponge in the reduction of postoperative morbidity in rectal cancer patients: early results and surprising outcome at 3-year follow-up. Int J Colorectal Dis 2005; 20(2): 114-20. 14. Spence RA, Anderson JR, Parks TG. Te use of gentamicin-PMMA chains in colorectal surgery. Br J ClinPract 1984; 38(7-8): 252-4. 15. Brehant O, Sabbagh C, Lehert P, Dhahri A, Rebibo L, Regimbeau JM. Te gentamicin-collagen sponge for surgical site infection prophylaxis in colorectal surgery: a prospective case-matched study of 606 cases. Int J Colorectal Dis 2013; 28(1): 119-25. 16. Irvin TT, Goligher JC. A controlled clinical trial of three diferent methods of perineal wound management following excision of the rectum. Br J Surg 1975; 62(4): 287-91. 17. Meyer L, Bereuter M, Marusch F, et al. Perineal wound closure afer abdomino-perineal excision of the rectum. Tech Coloproctol 2004; 8: s230-s4. 18. Lutje HC, Penschuck C, Aydin V. Local antibiotic treatment of sof-tissue infections with gentamicin- PMMA chains. Reconstr Surg Traumatol 1988; 20: 112-9. 19. Muhleder H, Togel K, Urbanek H, Leibl W. Gentamicin-polymethylmethacrylate chains in colonic surgery. Reconstr Surg Traumatol 1988; 20: 105-11.

122 Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection

20. Collin A, Gustafsson U-M, Smedh K, Pahlman L, Graf W, Folkesson J. Efect of local gentamicin-collagen on perineal wound complications and cancer recurrence afer abdominoperineal resection: A multicentre randomised controlled trial. Colorect Disease 2013; 15(3): March. 21. De Bruin AFJ, Gosselink MP, Wijfels NAT, Coene P-P, Van Der Harst E. Local gentamicin reduces perineal wound infection afer radiotherapy and abdominoperineal resection. Techniques in Coloproctology 2008; 12(4): December. 22. Gruessner U, Clemens M, Pahlplatz PV, Sperling P, Witte J, Rosen HR. Improvement of perineal wound healing by local administration of gentamicin-impregnated collagen feeces afer abdominoperineal excision of rectal cancer. American Journal of Surgery 2001; 182(5): 2001. 23. Rosen HR, Marczell AP, Czerwenka E, Stierer MO, Spoula H, Wasl H. Local gentamicin application for perineal wound healing following abdominoperineal rectum excision. Am J Surg 1991; 162(5): 438-41. 24. Moesgaard F, Lykkegaard NM. Failure of topically applied antibiotics, added to systemic prophylaxis, to reduce perineal wound infection in abdominoperineal excision of the rectum. Acta Chir Scand 1988; 154(10): 589-92. 25. Sachweh D. Te local application of antibiotics in sof-tissue surgery. Reconstr Surg Traumatol 1988; 20: 63-72. 26. Chang WK, Srinivasa S, MacCormick AD, Hill AG. Gentamicin-collagen implants to reduce surgical site infection: systematic review and meta-analysis of randomised trials. Ann Surg 2013; 258(1): 59-65. 27. Bullard KM, Trudel JL, Baxter NN, Rothenberger DA. Primary perineal wound closure afer preoperative radiotherapy and abdominoperineal resection has a high incidence of wound failure. Dis colon rectum 2005; 48(3): 438-43. 28. Schwenk W, Haase O, Neudecker J, Muller JM. Short term benefts for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005; (3): CD003145. 29. Wiatrek RL, Tomas JS, Papaconstantinou HT. Perineal wound complications afer abdominoperineal resection. Clin Colon Rectal Surg 2008; 21(1): 76-85. 6 30. Goldsmith HS, Grifth AL, Kupferman A, Catsimpoolas N. Lipid angiogenic factor from omentum. JAMA 1984; 252(15): 2034-6. 31. Walker FC, Rogers AW. Te greater omentum as a site of antibody synthesis. BrJ Exp Pathol 1961; 42: 222-31. 32. Terranova O, Sandei F, Rebufat C, Maruotti R, Pezzuoli G. Management of the perineal wound afer rectal excision for neoplastic disease: a controlled clinical trial. Dis colon rectum 1979; 22(4): 228-33. 33. Anderin C, Martling A, Lagergren J, Ljung A, Holm T. Short-term outcome afer gluteus maximus myocutaneous fap reconstruction of the pelvic foor following extra-levator abdominoperineal excision of the rectum. Colorectal Dis 2012; 14(9): 1060-4. 34. Hurwitz DJ. Te distal gluteus maximus advancement musculocutaneous fap for coverage of trochanteric pressure sores. Ann Plast Surg 1988; 20(2): 198-200. 35. Peterson SL, Stranahan PL, Schmaltz D, Mihaichuk C, Cosgrif N. Comparison of healing process following ligation with sutures and bipolar vessel sealing. Surg Technol Int 2002; 10: 55-60. 36. Showalter SL, Kelz RR, Mahmoud NN. Efect of technique on postoperative perineal wound infections in abdominoperineal resection. AmJ Surg 2013; 206(1): 80-5. 37. Stemberger A, Grimm H, Bader F, Rahn HD, Ascherl R. Local treatment of bone and sof tissue infections with the collagen-gentamicin sponge. Eur J Surg Suppl 1997; (578): 17-26. 38. Nelson RL, Glenny AM, Song F. Antimicrobial prophylaxis for colorectal surgery. Cochrane Database Syst Rev 2009; (1): CD001181. 39. Song F, Glenny AM. Antimicrobial prophylaxis in colorectal surgery: a systematic review of randomised controlled trials. Br J Surg 1998; 85(9): 1232-41.

123

CHAPTER 7

Is there a place for a biological mesh in perineal hernia repair?

G.D. Musters, O. Lapid, J. Stoker, B.F. Musters, W.A. Bemelman, and P.J. Tanis

Submitted Chapter 7

ABSTRACT

Aim: Tis study aimed to determine the outcome of perineal hernia repair with a biological mesh afer abdominoperineal resection (APR).

Methods: All consecutive patients who underwent perineal hernia repair with a porcine acellular dermal mesh between 2010 and 2014 were included. Follow-up was performed by clinical examination and MRI.

Results: Fifeen patients underwent perineal hernia repair afer a median of 25 months from APR. Four patients had a concomitant contaminated perineal defect, for which a gluteal fasciocutaneous fap was added in three patients. Wound infection occurred in three patients. Afer a median follow-up of 17 months (IQR 12-24), a clinically recurrent perineal hernia developed in 7 patients (47%): 6 of 11 patients afer a non-cross-linked mesh and 1of 4 patients afer a cross-linked mesh (p=0.57). Routine MRI at a median of 17 months revealed a recurrent perineal hernia in 7 of 10 evaluable patients, with clinical confrmation of recurrence in 5 of these 7 patients. No recurrent hernia was observed in the three patients with combined fap reconstruction for contaminated perineal defects.

Conclusion: A high recurrence rate was observed afer biological mesh repair of a perineal hernia following APR.

126 Perineal hernia repair with a biological mesh

INTRODUCTION

Te reported incidence of perineal hernia ranges between 1-13% afer abdominoperineal resection (APR) with primary perineal wound closure for rectal cancer.1 True incidences of perineal hernia might even be higher because of the underreporting of (asymptomatic) perineal hernia. In addition, literature data may not be representative for current practice with the increasing use of the extralevator APR, which has been associated with perineal hernia rate of up to 26%.2

Patients with a symptomatic perineal hernia most ofen present with perineal discomfort, pain and urinary dysfunction, and rarely with bowel obstruction. Treatment of perineal hernia is most ofen conservative, consisting of supportive undergarments. When surgical intervention is indicated, mesh repair is preferred over primary suturing as a basic treatment principle in hernia surgery.3

Several types of meshes have been used for perineal hernia repair, including biological meshes.3 Biological meshes are suitable for contaminated felds and are supposed to result in fewer bowel adhesions compared to synthetic meshes.4,5 Because of these characteristics, a biological mesh may be considered to be of additional value in the potentially contaminated area of the perineum and the possibility of small bowel lying on top of it. However, current literature on perineal hernia repair with a biological mesh mainly consists of case reports. Terefore, the aim of this study was to determine the postoperative outcome and mid-term follow-up, including magnetic resonance imaging (MRI), of a consecutive cohort of perineal hernia repair using a biological mesh.

PATIENTS AND METHODS

All consecutive patients who underwent a perineal hernia repair between March 2010 and April 2014 at the Academic Medical Centre, Amsterdam, were included. All perineal hernia reconstructions 7 were performed or supervised by two colorectal surgeons (P.T. and W.B.) and performed in prone position using a transperineal approach, except for one patient who underwent a transabdominal approach. Transperineal repair started with resection of the redundant skin and hernia sac. Subsequently, a porcine acellular dermal mesh was sutured to the sacrococcygeal ligaments dorsally, the remnants of the levator muscle laterally, and the transverse perineal muscle anteriorly with interrupted polypropylene 2.0 sutures (Figure 1). During the study period, we switched from a cross-linked biological mesh (Permacol™ 10x10 cm) to a non-cross-linked mesh (Strattice™ 6x10 cm), related to institutional uniformity in the use of a biological mesh for all indications. Tereafer, a 10 French vacuum drain was placed on top of the mesh and the subcutaneous tissue and skin were closed. In most cases the mesh was covered top side with omentum because an omentoplasty was performed at the index operation as a routine. If there was insufcient sof tissue to cover the mesh bottom side, a gluteal fasciocutaneous transposition fap was used for perineal closure.

Postoperatively, the drain was removed afer 3-7 days. Patients were fully mobilized at postoperative day one, except for those with a gluteal fasciocutaneous fap, who were mobilized afer three days and were allowed to sit afer seven days. Follow-up was at least 12 months and consisted of clinical examination during each visit to the outpatient clinic. In addition, 10 patients underwent MRI as

127 Chapter 7

routine follow-up between seven and 41 months from perineal hernia repair. MRI was performed on 1.5-3 Tesla scanners (Philips Healthcare, Best, the Netherlands and Siemens Avanto, Erlangen, Germany) with an axial and/or sagittal, coronal T2 weighted sequence, and a sagittal and/or coronal dynamic sequence with Valsalva manoeuvre. Te MRI was evaluated by a gastrointestinal radiologist with extensive experience in pelvic MRI (J.S.).

Data extraction Patient records were retrospectively searched for patient and treatment characteristics. Baseline data were extracted on underlying disease, (chemo)radiotherapy, extent of primary APR, the use of an omentoplasty and the method of perineal wound closure. Operative reports of perineal hernia repair were searched for the description of the operative approach (transperineal, transabdominal), the type of biological mesh used (cross-linked, non-cross-linked), and the use of an additional gluteal transposition fap. Outcome parameters were hospital stay, perineal wound infection, clinical and radiological recurrence of perineal hernia, and perineal re-interventions. Clinical recurrent hernia was defned as a midline swelling in standing position with absence of the anal clef and a palpable pelvic foor defect. A radiological hernia was defned as descent of small bowel or omentoplasty below the line between the coccyx and the perineal body (Figure 2) or below visible remnants of a mesh.

Statistical analysis According to distribution, descriptive data were reported as median with interquartile range (IQR) or mean ± standard deviation (SD). Categorical data were analysed with the Chi-square-test or Fisher’s exact test. All analyses were performed with IBM SPSS statistics, version 20.0.0 (IBM Corp., Armonk, NY, United States).

RESULTS

Patient characteristics. Between March 2010 and April 2014, 15 patients were operated for a perineal hernia afer APR. Te patient characteristics are displayed in Table 1. An extralevator approach was performed in fve of 11 rectal cancer patients, but in none of them the coccyx was removed. Te perineum was closed primarily in 13 patients, closed with a biological mesh (Permacol™) in one patient, and primary Vacuum Assisted Closure (VAC) was applied in the remaining patient. A postoperative perineal wound infection occurred in four of 14 patients with intentionally primary wound healing, of whom one underwent reoperation for a presacral abscess and one was treated with VAC therapy.

A symptomatic perineal hernia developed afer a median of 14 months (IQR 5-34). A total of fve patients were referred from another hospital. Two of them underwent surgical intervention prior to referral. One patient underwent laparoscopic biological mesh (Permacol™) repair of a perineal hernia 13 months afer APR, followed by hysterectomy with McCall culdoplasty 17 months later, and developed a recurrent perineal hernia. Te other patient underwent hysterectomy and sacrocolpopexy using a prolene mesh (gynemesh®) afer 18 months because of sexual dysfunction

128 Perineal hernia repair with a biological mesh

and feeling of pressure in the perineum and developed recurrent symptoms with radiological descent of small bowel beyond the coccyx on imaging.

Figure 1 Technique of perineal hernia repair in prone position 7

Perineal hernia repair Primary perineal hernia repair was performed in 14 patients afer a median of 25 months (IQR 17- 55) from primary APR. Te remaining patient was operated upon for recurrent perineal hernia 51 months from primary APR and 38 months from frst perineal hernia repair. Tree female patients had a concomitant chronic perineal wound and one male patient was admitted from the emergency department with a perineal necrosis and infection. For the pelvic foor reconstruction, a cross-linked biological mesh in the frst three patients and a non-cross-linked biological mesh was used in the following 11 patients. In the remaining patient, the ‘meshoma’ of the cross-linked biological mesh was removed (Figure 3) and a new reconstruction of the pelvic foor was performed using a non- cross-linked biological mesh. Details of the perineal hernia and the surgical repair are summarized in Table 2. Te median duration of the operation was 103 minutes (IQR 75-210). A postoperative perineal wound infection requiring antibiotic therapy occurred in three patients, with percutaneous drainage in one of them. Te median postoperative hospital stay was three days (IQR 2-6).

129 Chapter 7

Figure 2 Explant of a cross-linked biological mesh

Outcome perineal hernia reconstruction. Te overall clinical recurrence rate of primary perineal hernia repair using a biological mesh, including the patient who underwent initial repair at the referring hospital, was seven of 15 patients (47%). Recurrent perineal hernia was diagnosed afer a median of 17 months (IQR 12-24). Te median follow-up afer hernia repair in patients not having a recurrent perineal hernia was 20 months (IQR 13-44). A recurrent perineal hernia occurred in six out of 11 non-cross-linked biological mesh repairs, and in one out of four primary repairs using a cross-linked biological mesh (p=0.57). None of the three patients treated with an additional gluteal fasciocutaneous fap had a recurrent perineal hernia afer 13, 13 and 21 months of follow-up.

Routine MRI was performed in 10 patients afer a median of 17 months (IQR 11-30) from perineal hernia repair (Figure 4). Five patients did not undergo MRI for the following reasons; palliative setting with metastatic disease in two patients, short follow-up (<6 months) in two patients, and a pacemaker in one patient. Nine out of 10 patients had perineal complaints at the time of MRI, and a clinical recurrent hernia was diagnosed by physical examination in fve of these patients. Te main complaints consisted of a pressing sensation in seven patients, of perineal pain in two patients, and dyspareunia in the remaining patient. A radiological recurrent perineal hernia was diagnosed on MRI in seven out of the 10 patients. In one of these seven patients, herniation along both lateral borders of a cross-linked mesh was visible afer 41 months, but herniation could not be confrmed clinically. In the other six patients, remnants of a non-cross-linked biological mesh could be identifed afer an interval between 12 and 28 months (Figure 4b,c), with radiological herniation in all these patients and clinical recurrence in fve patients. In the three other patients, an intact

130 Perineal hernia repair with a biological mesh

cross-linked mesh was visible afer 36 months, and an intact non-cross-linked biological mesh was visible at 7 and 10 months postoperatively without signs of recurrent hernia (Figure 4a).

Figure 3 MRI images of two patients afer perineal hernia repair. A: sagittal plane, T2 sequence, sufcient hernia repair with visible biological mesh between the coccyx and the perineal body. B, C: transverse and sagittal images, T2 sequence, of a patient with a recurrent hernia, revealing remnants of the biological mesh along the posterior and right lateral borders of the perineal defect.

DISCUSSION

To our knowledge, we report the largest cohort of patients who underwent a biological mesh reconstruction of a perineal hernia. A relatively high recurrence rate of 47% was found, which may even become higher with extended follow-up. Current literature on perineal hernia repair afer APR is limited. A systematic review of the literature identifed 39 relevant papers between 1939 and 2011, describing only 76 patients who underwent surgical repair of a perineal hernia.3 A biological mesh was used in only fve of those patients, reported by three diferent authors.6-8 Tese case reports are mainly focused on surgical technique and follow-up is insufcient to draw any conclusion on long- term outcome. 7

Te 47% recurrence rate found in our study is considerably higher than a recently reported 5% recurrence rate afer transperineal synthetic mesh reconstruction of uncomplicated perineal hernia repair in 21 patients with a median follow-up of 24 months.9 Terefore, we recently changed our approach to synthetic mesh repair for primary and recurrent perineal hernias in the absence of contamination, combined with an omentoplasty if not already present. We still prefer the transperineal approach, but others efectively use a (laparoscopic) trans-abdominal route of mesh placement.10,11

Te 93% radiotherapy rate may have contributed to the high failure rate. A biological mesh provides a scafold for ingrowth by host tissue, leading to integration of the mesh.12 Radiotherapy disrupts cellular cytokine reactions and reduces nitric oxide and metalloproteinase.13,14 As a result, inadequate sof tissue regeneration and disorganized deposition of collagen occurs.13,15 Te amount of cross- linking might be another explanation for the high recurrence rate, because a non-cross-linked biological mesh was used in the majority of patients.16 Cross-links are covalent bonds between the collagen and are supposed to resist collagen degradation by host or bacterial collagenase.17,18 Tereby, cross linking increases tensile strength which is supposed to result in a lower rate of recurrent hernia

131 Chapter 7

compared to non-cross-linked meshes, especially in a contaminated feld. On the other hand, cross- linking may restrict early cellular infltration and may elicit an unintended infammatory response, which could lead to fbrosis and could limit tissue remodelling.19,20 Encapsulation of a cross-linked mesh similar to a synthetic mesh was observed in one of our patients. Furthermore, fstula formation afer a cross-linked biological mesh repair of a perineal hernia has been described.21 Currently available literature does not allow for a defnitive answer on the preferred type of biological mesh.

If no omentoplasty has been performed at time of primary APR, small bowel will lie on top of the mesh following perineal hernia repair. Direct contact of bowel loops to a mesh might lead to adhesion or fstula formation. Besides small bowel descent, an omentoplasty also minimizes dorsal displacement of the internal genital organs and bladder, which is associated with sexual and bladder dysfunction, respectively. Tis is the reason why we consider adding an omentoplasty to the hernia repair if not done so primarily in order to restore pelvic anatomy as much as possible.

In this study, hernia repair was combined with a gluteal fasciocutaneous fap (VY or SGAP) to fll and close a chronic perineal defect in three patients. A gluteal fap does probably not add any strength to the pelvic foor reconstruction.22 However, it is important to adequately cover the mesh with well-vascularized subcutaneous tissue to prevent seroma and abscess formation below the mesh, and to close the perineal skin without tension. Loss of perineal tissue related to extended primary resection or infectious complications requiring debridement may require autologous tissue fap reconstruction. All three patients with combined biological mesh and fap reconstruction were still without signs of recurrent hernia at last follow-up. Probably, the use of a biological mesh for pelvic foor reconstruction should be restricted to such patients with contaminated perineal defects.

Adequate follow-up to determine success of hernia repair is essential and routine imaging may by the most objective measure. However, imaging might overestimate the clinically relevant recurrence rate as shown by the present data. MRI to evaluate perineal hernia repair using a biological mesh has frst been described by Kavanagh et al. in a single case.23 We performed MRI during follow-up to assess the basis for perineal complaints, with specifc focus on mesh ingrowth and remodelling and potential mechanisms of technical failure. Tis revealed that partial ‘disappearance’ of the biological mesh without tissue regeneration seemed to be the main reason for failure.

Limitations of this study are its retrospective design and the limited number of patients included. Failure may be associated with restricted experience, because of the rarity of this problem. However, our group also conducted a multicenter randomized controlled trial on pelvic foor reconstruction using a biological mesh following extralevator APR during the study period, which contributed to our expertise in this feld.24 Also, follow-up is still relatively short, and imaging was not performed at standardized follow-up intervals.

Despite these limitations, this study shows that perineal hernia reconstruction with a biological mesh results in a high recurrence rate in patients who underwent APR for cancer. Terefore, a biological mesh seems not to be the frst choice implant for perineal hernia repair in the absence of contamination and the presence of an omentoplasty.

132 Perineal hernia repair with a biological mesh

Table 1 patient, primary treatment and tumour characteristics

Characteristic n=15 Gender Male (n, %) 9 (60) Age Mean years ± SD 62 ± 11 BMI Median (IQR) 27 (24-29) Comorbidity Diabetes (n, %) 3 (20) Respiratory (n, %) 0 Cardiac (n, %) 2 (13) Primary disease Rectal cancer (n, %) 11 (73) Anal cancer (n, %) 4 (27) Radiotherapy Chemoradiotherapy (n, %) 7 (47) Short course 5x5 Gy (n, %) 6 (40) Long course without chemotherapy 1 (7) Type of APR Conventional APRa (n, %) 5 (33) Extralevator APRb (n, %) 5 (33) Ischio-anal APR (n, %) 4 (27) Intersphincteric APR (n, %) 1 (7) Surgery Laparoscopic approach (n, %) 13 (87) Omentoplasty at time of APR (n, %) 13 (87) pTNM-stage Stage I-II (n, %) 9 (60) Stage III (n, %) 6 (40)

APR=abdominoperineal resection. SD=standard deviation. IQR=interquartile range 7

133 Chapter 7

Table 2 perineal hernia repair using a biological mesh with subsequent outcome (n=15)

Number Status Perineal hernia repair Follow-up 6 Uncomplicated perineal Transperineal Strattice™ n=2 no recurrent hernia 13m / 18m healing, primary perineal n=2 recurrent hernia 14m / 24m, no hernia further intervention n=1 recurrent hernia 17m, prolene mesh repair 22m, recurrent hernia 33m n=1 recurrent hernia 9m, prolene mesh repair 10m, second recurrent hernia 14m, prolene mesh repair 16m, no recurrent hernia 25m 2 Uncomplicated perineal Transperineal Permacol™ n=1 no recurrent hernia, exenteration healing, primary perineal because of recurrent rectal cancer 24m hernia n=1 laparoscopic prolene mesh fxation of biological mesh anteriorly 10m, no recurrent hernia 46m (radiological recurrence) 1 Omental prolapse with Transperineal Strattice™ 44m Recurrent hernia 28m, prolene mesh perineal necrosis repair 29m, no recurrent hernia 39m 1 Primary Permacol™ at Transperineal Strattice™ 17m Recurrent hernia 12m, no further time of APR, primary intervention perineal hernia 1 Surgical abscess drainage Transperineal repair No recurrent hernia 21m with secondary healing, posterior vaginal wall + perineal defect with Strattice™ + bilateral VY uterine prolapse gluteal fap 94m 1 Primary VAC therapy Transperineal Strattice™ + No recurrent hernia 13m, metastatic with secondary healing, unilateral VY gluteal fap disease omental prolapse with 8m chronic wound 1 Secondary VAC therapy, Transperineal repair No recurrent hernia 13m, metastatic omental prolapse with posterior vaginal wall + disease chronic wound Strattice™ + unilateral SGAP fap 17m 1 Hysterectomy + Transabdominal Permacol™ No recurrent hernia 36m sacrocolpopexy for + revision omentoplasty + sexual dysfunction bladder interposition 53m and feeling of pressure, recurrent symptoms 1 Laparoscopic perineal Handassisted laparoscopic No recurrent hernia 14m hernia repair with omentoplasty + Permacol™ 13m, transperineal Strattice™ 51m hysterectomy + culdoplasty 30m, recurrent perineal hernia No.=number of patients. VAC=vacuum assisted wound closure. APR=abdominoperineal resection. SGAP=superior gluteal artery fasciocutaneous perforator fap; (x)m= time interval in months from primary APR or perineal hernia repair.

134 Perineal hernia repair with a biological mesh

REFERENCE LIST

1. Musters GD, Buskens CJ, Bemelamn WA, Tanis PJ. Perineal wound healing afer abdominoperineal resection for rectal cancer; a systematic review and meta-analysis. Dis colon Rectum 2014; 57(9): 1129-39.

2. Sayers AE, Patel RK, Hunter IA. Perineal hernia formation following extralevator abdominoperineal excision. Colorectal Dis 2014;17(4): 351-5.

3. Mjoli M, Sloothaak DA, Buskens CJ, Bemelman WA, Tanis PJ. Perineal hernia repair afer abdominoperineal resection: a pooled analysis. Colorectal Dis 2012; 14(7): 400-406.

4. Burns NK, Jafari MV, Rios CN, Mathur AB, Butler CE. Non-cross-linked porcine acellular dermal matrices for abdominal wall reconstruction. Plast Reconstr Surg 2010; 125(1): 167-76.

5. Rosen MJ. Biologic mesh for abdominal wall reconstruction: a critical appraisal. AmSurg 2010; 76(1): 1-6.

6. de Campos FG, Habr-Gama A, Araujo SE, et al. Incidence and management of perineal hernia afer laparoscopic proctectomy. Surg Laparosc Endosc Percutan Tech 2005; 15(6): 366-70.

7. Kathju S, Lasko LA, Medich DS. Perineal hernia repair with acellular dermal graf and suture anchor fxation. Hernia 2011; 15(3): 357-60.

8. Skipworth RJ, Smith GH, Anderson DN. Secondary perineal hernia following open abdominoperineal excision of the rectum: report of a case and review of the literature. Hernia 2007; 11(6): 541-5.

9. Martijnse IS, Holman F, Nieuwenhuijzen GA, Rutten HJ, Nienhuijs SW. Perineal hernia repair afer abdominoperineal rectal excision. Dis colon Rectum 2012; 55(1): 90-5.

10. Abbas Y, Garner J. Laparoscopic and perineal approaches to perineal hernia repair. Tech Coloproctol 2014; 18(4): 361-4.Gu Q, Wang D, Gao Y, et al. Expression of MMP1 in surgical and radiation-impaired wound healing and its efects on the healing process. J Environ Pathol Toxicol Oncol 2002; 21(1): 71-8.

11. Allen SK, Schwab K, Day A, Singh-Ranger D, Rockall TA. Laparoscopic repair of post-operative perineal hernia using a two-mesh technique. Colorectal Dis 2014; 17:70-73. 12. Novitsky YW. Biology of biological meshes used in hernia repair. Surg Clin North Am 2013; 93(5): 1211-5. 7 13. Gu Q, Wang D, Gao Y, et al. Expression of MMP1 in surgical and radiation-impaired wound healing and its efects on the healing process. JEnvironPatholToxicolOncol 2002; 21(1): 71-8.

14. Herskind C, Bamberg M, Rodemann HP. Te role of cytokines in the development of normal-tissue reactions afer radiotherapy. Strahlenther Onkol 1998; 174 ( 3): 12-5.

15. Johnson LB, Jorgensen LN, Adawi D, et al. Te efect of preoperative radiotherapy on systemic collagen deposition and postoperative infective complications in rectal cancer patients. Dis colon Rectum 2005; 48(8): 1573-80.

16. Deeken CR, Melman L, Jenkins ED, Greco SC, Frisella MM, Matthews BD. Histologic and biomechanical evaluation of crosslinked and non-crosslinked biologic meshes in a porcine model of ventral incisional hernia repair. J Am Coll Surg 2011; 212(5): 880-8.

17. de Castro Bras LE, Shurey S, Sibbons PD. Evaluation of crosslinked and non-crosslinked biologic prostheses for abdominal hernia repair. Hernia 2012; 16(1): 77-89.

18. Oliver RF, Grant RA, Cox RW, Hulme MJ, Mudie A. Histological studies of subcutaneous and intraperitoneal implants of trypsin-prepared dermal collagen allografs in the rat. Clin Orthop Relat Res 1976; (115): 291-302.

135 Chapter 7

19. Cornwell KG, Landsman A, James KS. Extracellular matrix biomaterials for sof tissue repair. Clin Podiatr Med Surg 2009; 26(4): 507-23.

20. Liang HC, Chang Y, Hsu CK, Lee MH, Sung HW. Efects of crosslinking degree of an acellular biological tissue on its tissue regeneration pattern. Biomaterials 2004; 25(17): 3541-52.

21. Eriksen M, Bulut O. Chemotherapy-induced enterocutaneous fstula afer perineal hernia repair using a biological mesh: a case report. Int Med Case Rep J 2014; 7: 11-3.

22. Christensen HK, Nerstrom P, Tei T, Laurberg S. Perineal repair afer extralevator abdominoperineal excision for low rectal cancer. Dis colon Rectum 2011; 54(6): 711-7.

23. Kavanagh DO, Imran H, Almoudaris A, Ziprin P, Faiz O. Dynamic Magnetic Resonance Imaging Demonstrates the Integrity of Perineal Reconstruction following Cylindrical Abdominoperineal Excision with Reconstruction of the Pelvic Floor Using Porcine Collagen. Case Rep Med 2012; 2012: 752357.

24. Musters GD, Bemelman WA, Bosker RJ, et al. Randomized controlled multicentre study comparing biological mesh closure of the pelvic foor with primary perineal wound closure afer extralevator abdominoperineal resection for rectal cancer (BIOPEX-study). BMC Surg 2014; 14: 58.

136 Perineal hernia repair with a biological mesh

7

137

CHAPTER 8

Surgery for complex perineal fstula following rectal cancer treatment using biological mesh combined with gluteal perforator fap

G.D. Musters, O. Lapid, W.A. Bemelman, and P.J. Tanis

Tech Coloproctol 2014; 18(10): 955-959 Chapter 8

ABSTRACT

Tree patients with complex perineal fstula afer extensive pelvic surgery and radiotherapy underwent surgical treatment by combining a biological mesh for pelvic foor reconstruction and a unilateral superior gluteal artery perforator (SGAP) fap for flling of the perineal defect. All patients had both faecal and urinary diversion. Two fstulas originated from the small bowel, necessitating parenteral feeding, and one from the bladder. Symptoms included severe sacral pain and skin maceration. Afer laparotomy with complete debridement of the pelvic cavity, the pelvic foor was reconstructed by stitching a biological mesh at the level of the pelvic outlet. Subsequently, patients were turned to prone position and perineal reconstruction was completed by rotating a SGAP fap into the defect between the biological mesh and the perineal skin. Operating time ranged from 10 to 12.5 h, and hospital stay lasted from 9 to 23 days. Te postoperative course was uneventful in all three patients. Reconstruction of large pelvic defects with an alternative to a recuts abdominis musculocutaneous fap and may be preferable afer extensive pelvic surgery with ostomy.

140 Surgery for complex perineal fstula following rectal cancer treatment

INTRODUCTION

Radical surgery and (chemo) radiotherapy as combined treatment modalities in patients with pelvic malignancies may result in long-term complications such as a persisting presacral sinus and fstula formation.1 Tese complications may be prevented by the use of an omentoplasty and/or musculocutaneous faps at the time of primary resection, by flling the dead space of the irradiated pelvis with well-vascularized tissue.2 Te vertical rectus abdominis musculocutaneous (VRAM) fap is most ofen used for primary perineal reconstruction in patients undergoing abdominoperineal excision (APE) or pelvic exenterative procedures.3,4

If secondary pelvic complications occur, a VRAM fap may not be available for reconstructive procedures, due to the presence of a colostomy and urostomy through both rectus muscles, and the omentum may not be present or sufcient anymore. We describe a novel combined surgical technique of a biological mesh and superior gluteal artery perforator (SGAP) fap for patients with complex perineal fstula afer extensive pelvic surgery.

Figure 1 Small bowel fstula and presacral cavity A: Small bowel enclosed by a fbrotic sheet with an enterocutaneous fstula. B= afer removal of the bladder and debridement of the presacral cavity. Te entrance of the perineal fstula is shown. C= prone position; excised perineal fstula with removal of the os coccyges and surrounding radiation fbrosis. 1= fbrotic sheet, 2= fstula, 3= small Bowel, 4= abscess cavity, 5= fstula entrance, 6= biological mesh, 7= sacrum.

Operative technique 8 All procedures started with a midline laparotomy. Tereafer, adhesiolysis was performed, and small bowel was removed from the pelvic cavity (fgure 1A). In case of a small bowel fstula, a segmental resection was performed with primary anastomosis. Subsequently, an extensive debridement of the pelvic cavity was performed with excision of all radiation induced scar tissue and granulation tissue (fgure 1B). On indication, completion proctectomy and/or cystectomy was performed to eliminate sources of mucus production and chronic infammation. Te abdominal phase was then fnalized by closure of the inferior pelvic aperture using a porcine derived mesh of 10 x 10 centimetre (StratticeTM, LifeCell Corp., One Millennium Way, Branchburg, NJ, USA). Te biological mesh was attached to the remainders of the levator muscle and fbrous tissue along the inferior rim of the pubic bone, the pelvic side wall and the sacrum at the level of S4 using interrupted prolene sutures (fgure 2A). Subsequently, the edges of the mesh were stitched to the surrounding tissue with interrupted vicryl sutures resulting in a smooth foor supporting the descending small bowel loops.

141 Chapter 8

Tereafer the patient was turned into the prone position for the perineal phase of the procedure. Te fstula tract and surrounding scar tissue at the level of the perineum were excised. Te incision was extended in the midline and the coccygeal bone was resected for wider access to the dead space below the biological mesh (fgure 1C).

Tereafer, the gluteal perforators were identifed using a hand Doppler and marked on the skin. Subsequently the skin, subcutaneous fat and gluteal fascia were dissected from lateral to medial with careful identifcation of the available perforators (fgure 2B). Afer enabling a 180-degree arc of rotation, the skin was marked and de-epithelialized and rotated into the defect (fgure 2C). Te fap was fxed in the space beneath the biological mesh using interrupted vicryl sutures. A 10 French vacuum drain was positioned in between the fap and the biological mesh, and at the donor site (fgure 3A). Finally, the subcutaneous fat and skin were closed in layers.

Figure 2 SGAP fap with biological mesh placement A: Intra-abdominal placed biological mesh sutured with interrupted vicryl sutures. B= superior gluteal artery perforator fap with arterial supply (perforators) to the fap shown. C= superior gluteal artery perforator fap being de-epithelialized. 1= biological mesh, 2= gluteal perforators of the SGAP, 3= de-epithelialized part of the SGAP.

RESULTS

Tree patients underwent combined biological mesh and SGAP fap perineal reconstruction at the Academic Medical Centre, Amsterdam, the Netherlands. Te frst patient developed enteric fstula to both the perineum and ventral abdominal wall afer treatment of locally recurrent rectal cancer including pelvic re-irradiation. Te high-output perineal fstula required parenteral feeding and resulted in severe sacral pain with the necessity of fstula care every 1.5 hours. Te second patient developed an entero-perineal fstula afer cystoprostatectomy for bladder cancer (10-year interval) and subsequent abdominoperineal excision with neoadjuvant radiotherapy for rectal cancer (7-year interval), necessitating parenteral feeding (fgure 4A). Te third patient had a history of Bricker urostomy for a ureteral fstula two years afer complicated low anterior resection with neoadjuvant radiotherapy for rectal cancer, ultimately resulting in a chronic presacral sinus with a fstula to the bladder and perineum (fgure 4B). Te patient characteristics are summarized in Table 1.

Details of the surgical procedure for each patient are displayed in Table 2. Postoperatively, pressure on the SGAP fap was minimized by instructing the patient to lay on either lef or right side at an anti-decubitus mattress. Patients were mobilized afer three days and were allowed to sit afer seven days for an increasing period of time. Te vacuum drain was removed afer an output of less than 50 cc over 24 hours. Te postoperative course was uneventful in all three patients. Hospital stay was

142 Surgery for complex perineal fstula following rectal cancer treatment

23, 14 and 9 days respectively. Te relatively long hospital admission of the frst two patients was used to gradually switch from parenteral to enteral feeding and to mobilize afer being bedridden for several months preoperatively. Figure 3B shows the postoperative result afer unilateral SGAP fap.

During follow-up, the frst patient died because of disseminated disease four months afer reconstructive surgery. Te second patient developed a local recurrence of his bladder carcinoma afer 14 months with small bowel obstruction and a recurrent perineal fstula, for which and end ileostomy was constructed. Te follow-up in the third patient is still uneventful without clinical signs of perineal hernia afer 16 months.

Table 1 Patient and primary treatment characteristics

Patient 1 Patient 2 Patient 3 Age (years) 69 66 63 Sex Female Male Male Previous medical history - Cystoprostatectomy for bladder cancer Hypertension Appendectomy Bronchitis Lynch syndrome Rectal cancer TNM stage ypT3N0M0 ypT3N0M0 ypT4N0M0

Primary surgery LAR, ileostomy APE LAR, ileostomy Preoperative 5x5 25x2 radiotherapy (gy) 25x2 with concurrent capecitabine Postoperative Presacral abscess Perineal wound infection Anastomotic leakage complications Presacral abscess Presacral abscess

Follow-up events Local recurrence - Enterocutaneous fstula, -Ureteral fstula, Bricker small bowel resection deviation - Renal insufciency, - Entero-perineal fstula, small percutaneous bowel resection nephrostomy - Transient ischemic attack - Deep vein trombosis/ -Atrial fbrillation 8 infection related to TPN - Bacterial arthritis Redo pelvic surgery Total exenteration, Debridement of persistent Breakdown of anastomosis VRAM fap reconstruction presacral abscess with end colostomy, anal excision, omentoplasty Re-irradiation 17x1.8 Gy with concurrent No No chemotherapy, Postoperative brachytherapy (30 Gy) LAR= low anterior resection. APE=abdominoperineal excision. TPN= total parenteral nutrition VRAM=vertical rectus abdominis musculocutaneous fap.

143 Chapter 8

Table 2 Treatment for complex perineal fstula with postoperative outcome

Patient 1 Patient 2 Patient 3 Perineal fstula High-output entero- High-output entero- Fisula between bladder, perineal fstula perineal fstula pre-sacral sinus and perineum with gluteal fstula Time interval* (months) 78 81 92 Surgical interventions for Segmental jejunum Segmental jejunum Cystoprostatectomy, fstula treatment resection with primary resection with primary intersfncteric rest anastomosis anastomosis proctectomy Perineal reconstruction Biomesh + SGAP Biomesh + SGAP Biomesh + SGAP Duration of surgery (hours) 12.5 11 10 Total hospital stay (days) 23 14 9 Postoperative complications No No No

*= time between primary surgery for rectal cancer and reconstructive surgery for perineal fstula

Figure 3 Postoperative result A: Postoperative wound with drains positioned at the donor site and between the biological mesh and superior gluteal artery perforator fap. B= postoperative result of the superior gluteal artery perforator fap. 1= Drain going to the donor site, 2=drain positioned between biological mesh and SGAP fap.

DISCUSSION

A number of musculocutaneous and perforator faps can be used for perineal closure afer pelvic surgery. Te gracilis fap, lateral thigh fap, VRAM fap and the gluteal fap have all been described for this purpose.5-8

However, if a total pelvic exenteration has been done with a colostomy and urostomy placed bilaterally through the rectus muscles, the use of a VRAM fap is not advocated due to potential damage of the epigastric vessels and stoma related problems. Other alternatives for the perineal

144 Surgery for complex perineal fstula following rectal cancer treatment

reconstruction are the gracilis fap and lateral thigh fap. Although promising, the literature on the use of these faps is scarce, especially in oncological and irradiated patients. Terefore, we decided to reconstruct the pelvic foor using a biological mesh in addition to the SGAP fap. Te gluteal fap can be performed as a mycocutaneous fap or a perforator fap (IGAP/SGAP).8 Te gluteal perforator fap lies outside the feld of irradiation and transfers solely well vascularised skin and subcutaneous tissue into the perineal wound. Since no muscle is used, patients experience less pain and morbidity at the donor site compared to the gluteal myocutaneous fap.6 However, using only a gluteal fap has been associated with a relatively high rate of perineal hernia formation.9 Te presented patients all had a wide pelvic foor defect without pelvic organs that may prevent herniation. Terefore, the neo- pelvic foor was strengthened with a biological mesh.

In conclusion, a reconstructive option for irradiated pelvic foor reconstruction afer treatment of a complex perineal fstula following redo pelvic surgery and ostomies through both rectus muscles is a combined technique of a biological mesh and SGAP fap.

Figure 4 Perineal fstula A=mascerated perineal skin and buttocks due to pancreatic fuid and bile containing output of the enterocutaneous fstula. B=sagittal-plane, T2 weighted MRI image, with a presacral abscess with a small not sufcient omentoplasty 8 visible.1= bladder with thickened wall, 2= omentoplasty, 3=presacral sinus

145 Chapter 8

REFERENCE LIST

1. Chadwick MA, Vieten D, Pettitt E, Dixon AR, Roe AM. Short course preoperative radiotherapy is the single most important risk factor for perineal wound complications afer abdominoperineal excision of the rectum. Colorectal Dis 2006; 8: 756-761. 2. Hultman CS, Sherrill MA, Halvorson EG, Lee CN, Boggess JF, Meyers MO, Calvo BA, Kim HJ. Utility of the omentum in pelvic foor reconstruction following resection of anorectal malignancy: patient selection, technical caveats, and clinical outcomes. Ann Plast Surg 2010; 64: 559-562. 3. Butler CE, Gundeslioglu AO, Rodriguez-Bigas MA. Outcomes of immediate vertical rectus abdominis myocutaneous fap reconstruction for irradiated abdominoperineal resection defects. J Am Coll Surg 2008; 206: 694-703. 4. Lefevre JH, Parc Y, Kerneis S, Shields C, Touboul E, Chaouat M, Tiret E. Abdomino-perineal resection for anal cancer: impact of a vertical rectus abdominis myocutaneus fap on survival, recurrence, morbidity, and wound healing. Ann Surg 2009; 250: 707-711. 5. McCraw JB, Massey FM, Shanklin KD, Horton CE. Vaginal reconstruction with gracilis myocutaneous faps. Plast Reconstr Surg 1976; 58: 176-183. 6. Nelson RA, Butler CE. Surgical outcomes of VRAM versus thigh faps for immediate reconstruction of pelvic and perineal cancer resection defects. Plast Reconstr Surg 2009; 123: 175-183. 7. Ng RW, Chan JY, Mok V, Li GK. Clinical use of a pedicled anterolateral thigh fap. J Plast.Reconstr. Aesthet Surg 2008; 61: 158-164. 8. Unal C, Yirmibesoglu OA, Ozdemir J, Hasdemir M. Superior and inferior gluteal artery perforator faps in reconstruction of gluteal and perianal/perineal hidradenitis suppurativa lesions. Microsurgery 2011; 31: 539-544. 9. Christensen HK, Nerstrom P, Tei T, Laurberg S. Perineal repair afer extralevator abdominoperineal excision for low rectal cancer. Dis Colon Rectum 2011; 54: 711-717.

146 Surgery for complex perineal fstula following rectal cancer treatment

8

147

CHAPTER 9

Early reconstruction of the leaking ileal pouch-anal anastomosis: a novel solution to an old problem

T.J. Gardenbroek*, G.D. Musters*, C.J. Buskens, C.Y. Ponsioen, G.R.A.M. D’Haens, M.G.W. Dijkgraaf, P.J. Tanis, and W.A. Bemelman

*Both authors contributed equally to this work

Colorectal Dis 2015; 17(5): 426-432 Chapter 9

ABSTRACT

Aim: In this study, we determined the efectiveness and direct medical costs of early surgical closure of the anastomotic defect afer short course of Endo-sponge® therapy of the presacral cavity, compared to conventional treatment in patients with anastomotic leakage afer ileal pouch-anal anastomosis (IPAA).

Methods: Patients with anastomotic leakage afer IPAA undergoing early surgical closure of the anastomotic defect afer short Endo-sponge® treatment were prospectively followed and compared with a consecutive cohort of IPAA patients with an anastomotic leak treated with the creation of a loop ileostomy and occasionally drainage of the presacral cavity.

Results: A total of 15 patients were treated with early surgical closure and 29 patients were treated conventionally. In the early surgical closure group, the Endo-sponge® treatment was continued for a median of 12 days (IQR 7-15), with a median of 3 (IQR 2-4) Endo-sponge® changes. Secondary anastomotic healing was achieved in all patients (n=15) in the early surgical closure group, which was signifcantly higher compared to 52% (n=16) in the conventional treatment group (P=0.003). Closure of the anastomotic defect was achieved afer a median of 48 days (25-103) in the early surgical closure group compared to 70 days (IQR 49-175) in the conventional treatment group (P=0.013). A functional pouch was seen in 93% and 86% of the patients respectively. No signifcant diferences in direct medical costs were found.

Conclusion: Early surgical closure afer a short period of Endo-sponge® treatment is highly efective to treat anastomotic leakage afer IPAA without increasing costs.

150 Early reconstruction of the leaking ileal pouch-anal anastomosis

INTRODUCTION

Restorative proctocolectomy (RPC) with ileal pouch-anal anastomosis (IPAA) is the treatment of choice when surgery is necessary for patients with ulcerative colitis (UC) and in some patients with familial adenomatous polyposis (FAP). Te Achilles’ heel of this procedure is anastomotic leakage. Even in the presence of a protective stoma, anastomotic leakage occurs in 4% of IPAA patients, while the non-diverted pouches have a risk of leakage of up to 15%.1–3

Anastomotic leakage in the non-diverted pouches will most ofen present with pelvic sepsis. Anastomotic leakage in diverted pouches ofen presents subclinically as a presacral abscess. Tis may develop into a presacral sinus with chronic low grade infection or even fstula formation to the perineal or gluteal area.4 Te presacral sinus may postpone or even preclude stoma reversal. If it is possible to close the stoma, pouch function might be compromised due to the chronic pelvic infammation and its subsequent fbrosis.5,6

Te goal of treating anastomotic leakage is the prevention of chronic sepsis and concomitant chronic sinus formation. Conventional management of anastomotic leakage afer IPAA is faecal diversion by an ileostomy, which can be combined with transanal or percutaneous drainage of the presacral abscess cavity.

Weidenhagen et al. described a novel approach to treat the presacral abscess with Endo-sponge® vacuum assisted drainage.7 By changing the Endo-sponge® two times per week and tapering the size of the Endo-sponge® systematically, the abscess cavity gradually collapses. Tis technique is labour-intensive, expensive and takes several weeks until closure is achieved.7–13 In addition, a small sinus may persist in the end with a risk of recurrent sinus afer stoma reversal. We have proposed a modifcation of the technique. Instead of repetitive changes of the Endo-sponge® aiming at gradual collapse of the cavity, the Endo-sponge® is used to clean the cavity during two or three Endo- sponge® placements. When the cavity is clean and is surrounded by healthy granulation tissue, the anastomotic defect is closed surgically.14 Tis innovative approach might overcome the drawbacks of the Weidenhagen technique.

Te aim of this study was to determine whether early surgical closure is feasible, results in a higher chance of secondary healed anastomosis, reduces the time to anastomotic healing and leads to a higher percentage of functional pouches in patients with anastomotic leakage afer IPAA for UC or 9 FAP compared with conventional management. In addition, we analysed the direct medical cost of both treatment approaches.

METHODS

All consecutive patients with anastomotic leakage afer IPAA for UC or FAP in the period January 2003 to January 2014 were included. Anastomotic leakage was defned as a symptomatic or asymptomatic leakage with an anastomotic defect confrmed by radiological imaging or during re-laparotomy for abdominal sepsis. Patients undergoing early surgical closure afer Endo-sponge®

151 Chapter 9

treatment of the anastomotic leakage between January 2010 and January 2014 were prospectively recorded and compared with a retrospectively identifed cohort of IPAA patients with anastomotic leakage who underwent conventional treatment between January 2003 and December 2009. Since 2010, the anastomotic leakages were treated with a short course of Endo-sponge® followed by surgical closure of the anastomotic defect, when conservative treatment was not feasible. Te Institutional Review Board of the Academic Medical Centre in Amsterdam, Te Netherlands, granted exemption from approval for this study.

Early surgical closure In patients with anastomotic leakage afer IPAA without ileostomy, the frst Endo-sponge® placement was combined with ileostomy creation under general anaesthesia. Subsequent Endo- sponge® changes were carried out under light sedation in the endoscopy room. First, the abscess cavity was examined and rinsed with saline (0.9%) using a fexible gastroscope (GIF-100 Video Gastroscope; Olympus, 9.8-mm diameter, Olympus Corp., Tokyo, Japan). Next, one or more open-pored polyurethane Endo-sponge(s)® (B. Braun Medical B.V., Melsungen, Germany) were placed via a plastic overtube under the guidance of the gastroscope into the deepest point of the abscess cavity (Fig. 1). Tereafer, the Endo-sponge(s)® was connected to a low-vacuum suction bottle (Redyron® TRANS PLUS suction device; Melsungen, Germany). Te Endo-sponge(s)® was endoscopically changed every 3–4 days to prevent tissue ingrowth or if the vacuum diminished. When the abscess cavity was considered clean and the edges of the anastomosis were mobile, the anastomotic defect was closed surgically. Under general anaesthesia, the anastomotic dehiscence was transanally sutured with polydioxane 3-0 (PDS; Ethicon Endo-Surgery Inc., Cincinnati, Ohio, USA) over a vacuum drain by using the Lone Star Retractor System™ (Lone Star Medical Products®, Houston, Texas, USA). Te trans-anastomotic drain was removed on the third postoperative day and antibiotics were given for 10–14 days postoperatively. All reconstructed anastomoses were evaluated by endoscopic inspection afer 2 weeks or CT scan with intraluminal contrast. Stoma closure was scheduled when anastomotic healing was confrmed. If there was a persisting anastomotic defect during radiological or endoscopic evaluation, continuation of Endo-sponge® treatment with a second attempt of defect closure was considered.

Conventional treatment Afer radiological confrmation of the anastomotic leakage, a diverted ileostomy was created, if not done so primarily. Depending on the size of the cavity, additional transanal or percutaneous drainage of the abscess cavity was performed, sometimes followed by regular irrigation via the drain, or a wait and see policy was adopted. During follow-up, secondary healing of the anastomosis was regularly checked endoscopically or by an x-ray with contrast enema of the pouch. Stoma closure was scheduled if anastomotic healing was confrmed.

Outcomes Te main outcome parameter was the percentage of secondary healed anastomosis at 6 months, without a persistent abscess or anastomotic defect assessed by imaging. Secondary outcome parameters were the time from diagnosis of the leakage to secondary anastomotic healing and the percentage of functional pouches at last date of follow-up. In addition, the size of the initial

152 Early reconstruction of the leaking ileal pouch-anal anastomosis

abscess cavity, the time to intervention, duration of Endo-sponge® treatment, the number of Endo- sponge® treatments, the number of patients with a chronic presacral sinus longer than one year, and the direct medical costs of both treatment modalities were calculated. Secondary healing of the anastomosis was defned as no signs of contrast leakage during contrast enema or abdominal CT scan with intravenous, oral and rectal contrast and an intact anastomosis during endoscopic inspection. Patients were classifed as having a functional pouch when continuity was restored by ileostomy closure and no long-term complications were reported (such as a persisting high defecation frequency, urgency incontinence or re-creation of a stoma).

Cost analysis Te direct medical cost per patient was calculated for both treatment groups. Tis was calculated from the diagnosis of anastomotic leakage to reversal of the stoma or up to 2 years of follow-up if ileostomy closure had not yet been performed. Te direct medical cost was calculated as the product of the sum of healthcare volumes involved and their unit cost. Tis included the cost of the index hospital admission, re-admission(s), re-operation(s), visits to and telephone contacts with the outpatient department and the emergency room, and endoscopic examination or radiological imaging from diagnosis of anastomotic leakage to ileostomy closure. In addition, we accounted for subsequent ileostomy care cost. For the early surgical closure group, the cost of the Endo-sponge® treatment, Endo-sponge® changes and surgical closure of the defect was also included.

Te total medical cost consists of all patients who were successfully treated with early surgical closure and the successfully treated patients in the conventional treatment group. In addition, a separate cost analysis was performed up to 2 years afer diagnosis of the leakage for those patients in whom closure of the defect was not achieved at the end of follow-up or where an ileostomy was present owing to pouch failure. Unit costing was based on the Dutch costing manual for healthcare research or was determined in cooperation with the hospital administration and pharmacy.15 Te unit costs are determined for the year 2010, afer price-indexing (based on general consumer price indices; www.cvz.nl access date 20 April 2013) of unit costs stemming from diferent calendar years.

Statistical analysis Descriptive data are reported as median with interquartile range (IQR) or mean ± standard deviation according to distribution. Categorical data were analysed with Fisher’s exact test or Chi-square-test. Continuous variables were analysed using the Mann-Whitney-Wilcoxon test. P-value < 0.05 was 9 considered statistically signifcant.

For the cost analysis, the costs of both treatment modalities were compared by calculating the 95% confdence intervals (CI) for the mean diferences afer correction for bias, because of skewed distributions. Accelerated, non-parametric bootstrapping was used with stratifcation for Endo- sponge® treatment, drawing 10,000 samples of the same size as the original samples separately for each group and with replacement. All analyses were performed with IBM SPSS Statistics, version 20.0 (IBM Corp., Armonk, NY, United States).

153 Chapter 9

Figure 1 One of two Endo-sponges® in situ in the presacral cavity

RESULTS

Between January 2003 and January 2014, 393 patients underwent RPC and IPAA. Of these, 46 (11.7%) were diagnosed with an anastomotic leakage. Five additional patients were referred from other hospitals for treatment of anastomotic leakage. Of the 51 patients, 15 underwent early surgical closure and 29 patients received conventional treatment. Te remaining seven patients were excluded because they underwent other closure techniques including the original Weidenhagen technique (n = 4) and direct suturing of the defect without preceding Endo-sponge® therapy (n = 3).

Tese procedures were performed in the period between the use of the conventional method and the adoption of the early surgical closure technique. Tere was no diference between the location of the leakage or the size of the abscess cavity between the groups (Table 1).

Early surgical closure In the early surgical closure group, a diverting ileostomy was created during the RCP in four (27%) of the 15 patients. In nine (60%), an ileostomy was constructed afer diagnosis of the anastomotic leakage. In two (13%) patients early closure afer Endo-sponge® treatment was attempted without an ileostomy. In one (7%) of these two patients the early closure failed initially. Afer creation of a diverting stoma and a second period of Endo-sponge® treatment surgical closure was successful. In the conventional treatment group, three (10%) patients were primarily diverted during the RPC and the other 26 (90%) underwent secondary diversion afer diagnosis of leakage. Te Endo-sponge® was placed afer a median of 2.0 (IQR 0–8) days following the diagnosis of anastomotic leakage. Te Endo-sponge® treatment was continued for a median of 12 (IQR 7–15) days with a median of three (IQR 3–4) Endo-sponge® changes. Afer a median of 15 (IQR 11–31) days the anastomotic defect was surgically closed (Table 2). In two patients the surgical closure afer Endo-sponge® treatment failed initially. A second period of Endo-sponge® treatment followed by surgical closure was successful in both of them.

154 Early reconstruction of the leaking ileal pouch-anal anastomosis

Table 1 Patient characteristics

Conventional Early surgical p-value treatment closure n = 29 n = 15 Age at surgery Years (IQR) 36 (22–47) 37 (25–56) 0.603¶ Sex Males (n, %) 15 (52) 12 (80) 0.068¶ Diagnosis UC (n, %) 20 (69) 9 (60) 0.552¶ FAP (n, %) 9 (31) 6 (40) Indication for surgery Refractory (n, %) 14 (48) 7 (47) 0.764¶ Dysplasia/carcinoma (n, %) 8 (28) 3 (20) Preventive (n, %) 7 (24) 5 (33) Colectomy procedure Open (n, %) 14 (48) 7 (47) 0.137¶ Total laparoscopic (n, %) 5 (48) 4 (27) Hand-assisted laparoscopic (n, %) 10 (17) 4 (27) Anastomosis Stapled (n, %) 26 (90) 11 (73) 0.161¶ Abscess size Length x height x width (cm3)* 141 (15–222) 161 (74–257) Length (cm)† 4 (2–6) 4 (3–6) 0.291** Height (cm)‡ 7 (3–8) 6 (5–8) Width (cm)§ 6 (3–7) 6 (4–7) Primary diversion 3 (10) 4 (27) 0.207** Mortality 1 (3) 0 1.00** Data are presented as median (IQR) or n (%). *Measurement of the abscess size was possible in 18 patients in the conventional treatment group and 11 patients in the early surgical closure group. All dimensions of the abscess cavity were measured from an abdominal CT scan at their maximal size.†Te length of the cavity was measured on the sagittal or axial plane.‡Te height was measured on the sagittal or axial plane.§Te width of the cavity was measured on the coronal or axial plane. ¶Fisher’s exact test (two-tailed). **Mann Whitney U test.

Table 2 Treatment of anastomotic defect afer IPAA

Conventional Early surgical p-value treatment closure n = 28* n = 15 Time between diagnosis of leakage and surgical closure 15 (11–31) 9 of defect (days) Time between diagnosis of leakage and observed closure 70 (49–175) 48 (25–103) 0.013‡ of defect (days) Time between pouch surgery and observed closure of 84 (59–183) 59 (39–137) 0.081‡ defect (days) Ileostomy reversed at end of follow-up 25 (89) 14 (93) 1.000‡ Long-term pouch dysfunction (%) 4 (14) 1 (7.1) 1.000† Data are presented as median (IQR) or n (%). *In one patient the pouch was excised due to adeno- carcinoma.†Fisher’s exact test (two-tailed).‡Mann–Whitney test.

155 Chapter 9

Anastomotic healing Secondary anastomotic healing afer 6 months was achieved in all 15 patients in the early surgical closure group compared with 16 (52%) of the 29 patients in the conventional treatment group (P=0.003). Secondary anastomotic healing was achieved in the early surgical group afer a median of 48 (IQR 25–103) days compared with 70 (49–175) days in the conventional treatment group (P=0.013). A chronic presacral sinus was present in none of the patients in the early surgical closure group vs two (7%) patients in the conventional treatment group (P=0.542).

Follow-up Patients in the early surgical closure group were followed for a median of 25 (IQR 12–39) months vs a median 104 (IQR 60–122) months in the conventional treatment group (P<0.001). One pouch in the early surgical closure group had to be diverted again, due to persisting high defecation frequency and urge incontinence. Overall early surgical closure of the anastomotic defect resulted in a functional pouch in 14 (93%) of the 15 patients. In the conventional treatment group, the pouch was excised in one FAP patient due to the development of an adenocarcinoma; this patient was therefore excluded from the functional analysis. In the remaining 28 patients, a presacral sinus persisted in four (14%) of the 28 patients afer reversal of the ileostomy. For one of these, the persisting leakage resolved afer extensive surgical treatment. Te presacral sinus in the other three patients was observed during endoscopy but appeared to be not clinically relevant. Terefore, no treatment was instigated. One additional patient developed a dysfunctional pouch afer stoma reversal in the conventional treatment group due to high defecation frequency. No diversion was performed in this patient. Tus, 24 (86%) of 28 patients in the conventional treatment group had a functioning pouch at the end of follow-up.

Cost analysis Te total treatment costs per patient were €27.627 (95 per cent C.i. €23.239 to €32.690) for the early closure group and €33.441 (95 per cent C.i. €24.955 to €46.728) for the conventional treatment group (P=0.529). When patients with no observed secondary healed anastomosis and patients without ileostoma reversal are added with a follow-up of two years, the total direct medical costs were €27.879 (95 per cent C.i. 23.716 to 32.580) for the early surgical closure group, and €37.687 (95 per cent C.i. 27.592 to 51.007) for the conventional treatment group (P=0.304).

DISCUSSION

Early surgical closure of a leaking ileo-anal anastomosis following a short Endo-sponge® treatment resulted in a signifcantly higher percentage of secondary healed anastomoses at 6 months compared with conventional treatment. Furthermore, early surgical closure signifcantly reduced the time to anastomotic healing by a median of 22 days compared with conventional treatment. Early surgical closure resulted in a high percentage of functioning pouches at the end of follow-up. Treatment costs per patient in both study groups were no diferent.

156 Early reconstruction of the leaking ileal pouch-anal anastomosis

Closing anastomotic leakage surgically is a paradigm shif, because traditionally this has been considered to fail by defnition. Te success of early surgical closure of the anastomotic defect as shown in the present study is probably explained by the use of Endo-sponge® treatment, which resolves the pelvic sepsis behind the anastomosis. Subsequent surgical closure of the anastomotic defect is of great importance to prevent infux of mucus and debris in the cleaned cavity jeopardizing closure of the cavity. Te short period of negative pressure via a vacuum drain is thought to be useful for drainage of contaminated fuids in the cavity enabling expansion of the pouch with collapse of the presacral cavity. Not all presacral cavities, however, are suitable for Endo-sponge® treatment. Te surgeon should be cautious of higher anastomotic leakage because this could potentially result in small bowel adhesion, erosions and even fstula formation. When a large presacral abscess is present, more than one Endo-sponge® can be inserted.

Diversion of the pouch might be an essential component of this treatment strategy, even when the patient is not septic. When the Endo-sponge® treatment is started without a diverting ileostomy, the sponge has to drain intestinal contents and air as well, and might lose adequate vacuum in the cavity. Furthermore, if the anastomotic repair is not perfect the cavity will expand again due to the pressure of the intestinal contents. Closure failed initially in one of two patients without faecal diversion. Overall the frst results of early anastomotic reconstruction afer a short period of Endo-sponge® treatment are very promising and compare favourably with the alternative strategies. Endo-sponge® treatment until closure as developed by Weidenhagen has a mixed rate of success (56–94%)7,11, with a median time to closure of more than 340 days in patients undergoing low anterior resection or RPC4, but most of the available literature focuses on patients with anastomotic leakage afer rectal cancer surgery. Patients with rectal cancer signifcantly difer from UC or FAP patients with regard to age, comorbidity and the type of neorectum and neoadjuvant (chemo)radiotherapy. Prolonged treatment using the Weidenhagen technique is associated with high cost due to the large number of Endo-sponge® changes.7-11 Te present study is the frst to compare the cost of standard and Endo- sponge® treatment. No statistically signifcant diference in direct medical cost in the two treatment strategies was found, but early surgical closure seems to be less expensive than the conventional treatment. Tis is most likely to be due to the higher cost of stoma equipment and outpatient clinic visits.

Early surgical closure treatment was successful in almost all patients compared with only about one-half of patients having conventional treatment. Chronic presacral abscess or presacral sinus 9 formation afer leakage is responsible for delayed anastomotic healing or the inability to close the ileostomy and even when healing of the defect occurs and the stoma is closed, function may be poor owing to fbrosis.5,6 Te time interval between diagnosis of anastomotic leakage and the duration of subsequent treatment is likely to be an essential factor afecting long-term function. Endo-sponge® treatment according to Weidenhagen has been shown to be more successful if started within 6 weeks of the initial surgery rather than afer this interval.11

Tis study is limited by its partially retrospective design. Te defnition of the moment of anastomotic healing in both groups was therefore difcult. In the prospective group all closures were checked endoscopically two weeks afer surgical closure, while in the conventional group intermittent imaging was done to see whether the cavity was healed. Tis could have caused an overestimate

157 Chapter 9

of the time to heal. Due to the diference in length of follow-up, the pouch failure rate of the early closure group might have been shown to be more favourable because it is well known that failure increases over time. Most failures however, were caused by a persistent sinus that would have been prevented with early closure.

In conclusion, the study shows that early surgical closure afer a short period of Endo-sponge® treatment is efective in treating anastomotic leakage of the ileo-anal anastomosis by preventing chronic sinus formation, thereby avoiding a permanent stoma.

158 Early reconstruction of the leaking ileal pouch-anal anastomosis

REFERENCE LIST

1. Gullberg K, Liljeqvist L. Stapled ileoanal pouches without loop ileostomy: a prospective study in 86 patients. Int J Colorectal Dis 2001; 16(4): 221-7.

2. Wasmuth HH, Trano G, Endreseth B et al. Long-term surgical load in patients with ileal pouch-anal anastomosis. Colorectal Dis 2009; 11(7): 711-8.

3. Weston-Petrides GK, Lovegrove RE, Tilney HS et al. Comparison of outcomes afer restorative proctocolectomy with or without defunctioning ileostomy. Arch Surg 2008; 143(4):406-12.

4. van Koperen PJ, van der Zaag ES, Omloo JM, Slors JF, Bemelman WA. Te persisting presacral sinus afer anastomotic leakage following anterior resection or restorative proctocolectomy. Colorectal Dis 2011; 13(1): 26-9.

5. Breen EM, Schoetz DJ, Jr., Marcello PW et al. Functional results afer perineal complications of ileal pouch- anal anastomosis. Dis Colon Rectum 1998; 41(6): 691-5.

6. Farouk R, Dozois RR, Pemberton JH, Larson D. Incidence and subsequent impact of pelvic abscess afer ileal pouch-anal anastomosis for chronic ulcerative colitis. Dis Colon Rectum 1998; 41(10): 1239-43.

7. Weidenhagen R, Gruetzner KU, Wiecken T, Spelsberg F, Jauch KW. Endoscopic vacuum-assisted closure of anastomotic leakage following anterior resection of the rectum: a new method. Surg Endosc 2008; 22(8): 1818-25.

8. Cresti S, Ouaissi M, Sieleznef I et al. Advantage of vacuum assisted closure on healing of wound associated with omentoplasty afer abdominoperineal excision: a case report. World J Surg Oncol 2008; 6: 136.

9. D’hondt M, De HG, Malisse P, Vanden Boer J, Knol J. Chronic pelvic abscedation afer completion proctectomy in an irradiated pelvis: another indication for ENDO-sponge treatment? Tech Coloproctol 2009; 13(4): 311-4.

10. Mees ST, Palmes D, Mennigen R et al. Endo-vacuum assisted closure treatment for rectal anastomotic insufciency. Dis Colon Rectum 2008; 51(4):404-10.

11. van Koperen PJ, van Berge Henegouwen MI, Rosman C et al. Te Dutch multicenter experience of the endo-sponge treatment for anastomotic leakage afer colorectal surgery. Surg Endosc 2009; 23(6): 1379-83.

12. Weidenhagen R, Gruetzner KU, Wiecken T, Spelsberg F, Jauch KW. Endoluminal vacuum therapy for the treatment of anastomotic leakage afer anterior rectal resection. Rozhl Chir 2008; 87(8): 397-402.

13. van Koperen PJ, van Berge Henegouwen MI, Slors JF, Bemelman WA. Endo-sponge treatment of anastomotic leakage afer ileo-anal pouch anastomosis: report of two cases. Colorectal Dis 2008; 10(9): 943-4. 9 14. Verlaan T, Bartels SA, van Berge Henegouwen MI et al. Early, minimally invasive closure of anastomotic leaks: a new concept. Colorectal Dis 2011; 13 Suppl 7: 18-22.

15. CVZ. (2012) Handleiding voor kostenonderzoek (manual for cost research).

159

CHAPTER 10

Ileostomy closure by colorectal surgeons results in less major morbidity: results from an institutional change in practice and awareness

G.D. Musters, J.J. Atema, H.L. van Westreenen, C.J. Buskens, W.A. Bemelman, and P.J. Tanis

Int J Colorectal Dis 2016 Chapter 10

ABSTRACT

Aim: Previous institutional analysis of ileostomy closure revealed substantial morbidity. Tis subsequent study aimed at determining if a change in clinical practice resulted in reduced complication rates.

Methods: Between June 2004 and January 2014, all consecutive adult patients undergoing ileostomy closure were retrospectively identifed. Postoperative outcome afer change in clinical practice consisting of routine participation of a colorectal surgeon, stapled side to side anastomosis and increased clinical awareness (cohort B), was compared with our previously published historical control group (cohort A). Te primary outcome was major morbidity, defned as Clavien-Dindo grade three or higher. Independent risk factors of major morbidity were identifed using multivariable analysis.

Results: In total, 165 patients underwent ileostomy closure in cohort A, and 144 patients in cohort B. At baseline, more primary diverting ileostomies were present in cohort A (94% vs. 82%; p=0.001) with a similar rate of loop and end-ileostomy between the two cohorts (p=0.331). A signifcant increase in colorectal surgeon participation (89% vs. 53%; p<0.001) and stapled side to side anastomosis was observed (63% vs. 16%; p<0.001). Te major morbidity rate was 11% in cohort A, which signifcantly reduced to 4% in cohort B (p=0.03). Surgery being performed or supervised by a colorectal surgeon (OR 0.28; 95% CI 0.11-0.67) and loop ileostomy compared to end-ileostomy (OR 0.18; 95% CI 0.07- 0.52) were independently associated with lower major morbidity.

Conclusion: Ileostomy closure appears to be more complex surgery then generally considered, especially end-ileostomy closure. Postoperative outcome could be signifcantly improved by a change in surgical practice.

162 Morbidity afer ileostomy closure

INTRODUCTION

Defunctioning ileostomies are constructed for several reasons, but mostly to protect a low colorectal, coloanal or an ileal pouch-anal anastomosis (IPAA). Although there is still debate about the value of routine diversion of such anastomoses, a defunctioning ileostomy has been reported to signifcantly reduce clinical anastomotic leakage rates and mitigate its consequences.1 In patients who develop leakage of a non-diverted anastomosis, a secondary defunctioning ileostomy can be constructed if breakdown of the anastomosis is not indicated. Less frequent indications for diverting ileostomy are infammatory bowel disease, intestinal ischemia, oncological diseases, functional problems, or surgical complications other than distal leaking anastomosis.

Although a defunctioning ileostomy has clear advantages, stoma related morbidity should be taken into account as well. Closure of a defunctioning ileostomy can result in substantial morbidity, with a reported complication rate of 17% in a systematic review of more than 6000 patients.2 A retrospective cohort study of ileostomy closures performed at our own university hospital revealed a morbidity rate of 20%, with postoperative small bowel obstruction as the most frequent complication.3 As a result, increasing the surgical expertise present at closure and improving the perioperative clinical awareness of potential risks changed our clinical practice.

Te aim of this study was to evaluate if these changes in practice and increased awareness of the associated risks of the procedure led to a reduced major morbidity afer ileostomy closure. Secondly, we aimed to identify independent risk factors of major morbidity following ileostomy closure.

METHODS

Patients and management All consecutive adult patients undergoing ileostomy closure between June 2004 and January 2014 were identifed from a surgical administrative database. Patients were included in the present analysis independent of the underlying disease or the indication for constructing an ileostomy. Patients who underwent ileostomy closure between June 2004 and June 2010 (cohort A) were compared with patients operated upon between July 2010 and January 2014 (cohort B). Cohort A comprises of patients who were also described in a previous study.3

Antibiotic prophylaxis was routinely given pre-operatively. Bowel continuity was restored by either a hand-sewn anastomosis using a running PDS 3.0 suture in an end-to-end, side-to-end, or end-to- side confguration, or by a side-to-side stapled anastomosis. Te fasciae of the posterior and anterior 10 rectus sheath were separately closed by running or interrupted sutures using Vicryl or PDS. Te skin was partially closed with approximating interrupted transcutaneous sutures or with a purse string intracutaneous suture. Oral intake was started at day one postoperatively if tolerated. No routine imaging was performed to evaluate incisional hernia at the previous stoma site.

163 Chapter 10

Evaluation of our results in 2010 revealed considerable morbidity afer ileostomy closure, and resulted in several management changes.3 Firstly, the operating team was more strictly selected based on specifc expertise. Before June 2010, a surgical resident predominantly performed ileostomy closure and was not being routinely supervised by a consultant. When a resident was supervised, the supervising surgeon could be any type of surgical consultant. Afer June 2010, ileostomy closure was preferably performed or supervised by a colorectal surgeon, being either a consultant or fellow. Tere is not an ofcial colorectal subspecialisation within gastrointestinal surgery in the Netherlands, but these consultants and fellows almost exclusively perform colorectal surgery in daily practice at the AMC. Secondly, side-to-side stapled anastomoses were constructed whenever possible, instead of hand sewn end-to-end anastomoses. Besides these changes in surgical practice, the overall awareness of the risk of major morbidity afer ileostomy reversal increased. Since the present study involved a retrospective analysis of data, Dutch regulations do not require written informed consent.

Data collection Patient and treatment characteristics were retrospectively collected from patient records. Operative reports, radiology reports, and patient charts were used for collection of data on patient’s demographics, primary treatment characteristics, operative technique of stoma closure, 30-day postoperative complications, hospital stay, and out-patient follow-up with respect to long term stoma site related complications.

Defnition of outcome Te primary endpoint was major morbidity within 30 days postoperatively, defned as all complications classifed as Clavien-Dindo grade three or higher. Tis includes a complication for which a surgical, endoscopic or radiological intervention is required (grade three), a life threatening complication for which intensive care management is needed (grade four), or when the patient dies (grade fve).4

Statistical analysis According to distribution, descriptive data were reported as median with interquartile range (IQR) or mean ± standard deviation (SD). Categorical data were analysed with the Chi-square-test or Fisher’s exact test and continuous variables were analysed using the Mann-Whitney-U test or the independent T-test, according to the distribution. Univariable logistic regression analysis was performed to identify risk factors for major morbidity following ileostomy closure. All predictors with a P-value of less than 0.10 in univariable analysis were candidate variables for inclusion in a multivariable model. Multivariable regression analysis was used to identify independent risk factors for major morbidity afer stoma reversal. All analyses were performed with IBM SPSS statistics, version 20.0.0 (IBM Corp., Armonk, NY, United States).

164 Morbidity afer ileostomy closure

Table 1 Patients and stoma characteristics

Cohort A Cohort B p-value June 2004- June 2010 July 2010- Jan 2014 (n=165) (n=144) Gender Males (n, %) 93 (56.4) 83 (57.6) 0.821 Age Mean age (years, ± SD) 48.9 (±16.5) 51.4 (±15.0) 0.474 BMIa Mean BMI (± SD) 24.9 (±5.9) 24.3 (±5.0) 0.646 Smoking Number of patients (%) 30 (18.2) 21 (14.6) <0.001 Unknown (n, %) 52 (31.5) 10 (7.0) ASA-classifcation ASA I (n, %) 32 (19.4) 32 (22.2) 0.024 ASA II (n, %) 103 (62.4) 101 (70.1) ASA III (n, %) 30 (18.2) 11 (7.6) Primary diagnosis Colorectal Cancer (n, %) 61 (36.9) 60 (41.7) 0.003 Ulcerative colitis (n, %) 59 (35.8) 24 (16.7) Morbus Crohn (n, %) 9 (5.5) 9 (6.3) Familial adenomatous polyposis (n, %) 11 (6.7) 13 (9.0) Other (n, %) 25 (15.2) 38 (26.4) Primary surgery Laparoscopic surgery (n, %) 47 (28.5) 77 (53.5) <0.001 Open surgery (n, %) 77 (46.7) 66 (45.8) Unknown (n, %) 3 (1.8) 1 (0.7) Low anterior resection (n, %) 52 (31.5) 45 (31.3) 0.004 IPAA (n, %) 79 (47.9) 47 (32.6) Sigmoid resection (n, %) 12 (7.2) 14 (9.7) Diversion without resection (n, %) 11 (6.6) 14 (9.7) Colonic resection (n, %) 10 (6.0) 16 (11.1) Small bowel resection (n, %) 1 (0.6) 1 (0.7) Pull through with colo-anal 0 7 (4.8) anastomosis (n, %) Indication for Diversion for primary disease or 155 (93.9) 118 (81.9) 0.001 ileostomy during primary surgery (n, %) Secondary diversion for anastomotic 10 (6.1) 26 (18.1) leakage (n, %) Type of ileostomy Loop-ileostomy (n, %) 152 (92.1) 128 (88.9) 0.331 End-ileostomy (n, %) 13 (7.9) 16 (11.1)

BMI = body Mass Index, ASA= American Society of Anaesthesiology, IPAA= ileal pouch-anal anastomosis. 10

165 Chapter 10

RESULTS

Patients and stoma characteristics Between June 2004 and January 2014, 309 patients underwent ileostomy closure and were included in the present study. Cohort A consisted of 165 patients and cohort B consisted of the remaining 144 patients who were operated afer July 2010.

Patient and stoma characteristics of both cohorts are displayed in Table 1. Te primary diagnoses difered signifcantly between the two cohorts; ulcerative colitis was more frequently diagnosed in cohort A (35.8% vs 16.7%), while more patients had a stoma for other indications in cohort B (15.2% vs 26.4%; p=0.003). Furthermore, patients in cohort A were more ofen classifed as grade III according to the American society of anaesthesiologists (ASA) (18.2% vs 7.6%; p=0.024). Primary surgery was more ofen performed using an open approach (71.5% vs 46.5%; p<0.001) and more ofen with diversion during the initial surgery (93.9% vs 81.9%; p=0.001) in cohort A, compared with cohort B.

Ileostomy closure characteristics Time to ileostomy closure was comparable between both cohorts; 22.8 (SD ±18.45) weeks in cohort A, versus 21.6 (SD ±14.4) weeks in cohort B (p=0.09; Table 2). Stoma closure was performed or supervised by a colorectal surgeon in 53.3% (88/165) in cohort A, which was signifcantly lower compared to 88.9% (128/144) of the patients in cohort B (p<0.001). Also, stapled anastomoses were more ofen constructed in cohort B (66.0% vs 10.9%; p<0.001). Stoma site closure was diferent from the protocol in nine patients, four patients in cohort A and fve patients in cohort B: local gentamicin was used in fve patients and an absorbable mesh was used for fascial closure in four patients. A loop-ileostomy was closed in 92.1% of the patients (152/165) in cohort A, and in 88.9% of the patients (128/144) in cohort B (p=0.331).

Morbidity afer ileostomy closure Tirty-day postoperative morbidity afer ileostomy closure in the two cohorts is shown in Table 3. In total, 15 patients developed a wound infection (4.9%), with a similar rate among the two cohorts. Wound infection rates were 2% (2/101) afer purse string closure, 6.3% (9/144) using approximating interrupted transcutaneous sutures, and 5.9% (2/34) afer primary closure (p=0.277, missing 2). Anastomotic leakage rate was 6.7% (11/165) in cohort A, which was signifcantly higher than a 2.1% (3/144) leakage rate in cohort B (p=0.05). Major morbidity rate was also signifcantly higher in cohort A (10.9%; 18/165) compared to cohort B (4.2%; 6/144) (p=0.03). Major morbidity occurred in 9.3% (17/183) of the patients with a hand sewn anastomosis, and in 5.3% (6/113) of the patients with a stapled anastomosis (p=0.214). Similarly, anastomotic leakage (5.5% vs. 2.7%; p=0.83) and postoperative ileus (9.8 vs. 3.5%; p=0.07) showed non-signifcant diferences between hand sewn and stapled anastomosis, respectively. Two patients died within 30 days due to a complicated ileostomy reversal, one patient in each group. Cause of death was bleeding from the epigastric vessels, which resulted in a low fow state with cardiac ischemia and an intra-abdominal haematoma with bowel ischemia, and one patient died due to sepsis because of anastomotic leakage.

166 Morbidity afer ileostomy closure

Table 2 Ileostomy closure characteristics

Cohort A Cohort B p-value June 2004- June 2010 July 2010- Jan 2014 (n=165) (n=144) Time to stoma reversal Mean weeks (± SD) 22.8 (±18.45) 21.6 (±14.4) 0.088 Colorectal surgeon Yes (n, %) 88 (53.3) 133 (92.4) <0.001 performing or No (n, %) 77 (46.7) 11 (7.6) supervising surgery

Type of constructed End-to-end anastomosis (n, %) 117 (70.9) 40 (27.8) <0.001 Anastomosis Side-to-side anastomosis (n, %) 27 (16.4) 91 (63.2) Side-to-end or end-to-side 8 (4.8) 6 (4.2) anastomosis (n, %) Unknown (n, %) 13 (7.9) 7 (4.9) Anastomotic technique Sewn (n, %) 141 (85.5) 42 (29.2) <0.001 Stapled (n, %) 18 (10.9) 95 (66.0) Unknown (n, %) 6 (3.6) 7 (4.9) Skin closure Purse string intracutaneous 25 (15.2) 76 (52.8) <0.001 suture (n, %) Primary closure (n, %) 20 (12.1) 14 (9.7) Approximating interrupted 101 (61.2) 43 (29.9) transcutaneous sutures (n, %) Unknown (n, %) 19 (11.5) 11 (7.6)

Table 3 Postoperative morbidity afer ileostomy closure

Cohort A Cohort B p-value June 2004- June 2010 July 2010- Jan 2014 (n=165) (n=144) Hospital stay Median days (±IQR) 5.0 (3-6) 5.0 (4-7) 0.86 Wound infection Number of patients (%) 7 (4.2) 8 (5.5) 0.59 Ileus Number of patients (%) 15 (9.1) 7 (4.8) 0.15 Abscess Intra-abdominal (n, %) 4 (2.4) 3 (2.1) 0.84 Anastomotic leakage Number of patients (%) 11 (6.7) 3 (2.1) 0.05 Major morbidity Clavien-Dindo ≥3 (n, %) 18 (10.9) 6 (4.2) 0.03 Clavien-Dindo grade 1-2 (n, %) 12 (7.2) 11 (7.6) 0.90 Clavien-Dindo grade 3 (n, %) 14 (8.4) 1 (0.69) <0.01 Clavien-Dindo grade 4 (n, %) 3 (1.8) 4 (2.8) 0.48 10 Clavien-Dindo grade 5 (n, %) 1 (0.61) 1 (0.69) 0.92 Reoperation Number of patients (%) 11 (6.7) 3 (2.0) 0.06 Readmission Number of patients (%) 9 (5.5) 5 (3.5) 0.40

167 Chapter 10

In univariable analysis, primary laparoscopic surgery and a loop ileostomy were signifcantly associated with a lower risk of major morbidity, compared to primary open surgery and end- ileostomy, respectively (Table 4). Furthermore, major morbidity occurred signifcantly less afer ileostomy closure when performed or supervised by a colorectal surgeon. In multivariable analysis, loop-ileostomy and surgery performed or supervised by a colorectal surgeon remained independent favourable factors for the risk of major morbidity.

Table 4 Uni- and multivariable analysis of risk factors for major morbidity afer ileostomy closure

Univariable analysis Multivariable analysis OR 95% CI p-value OR 95% CI p-value Male gender 0.61 0.27-1.42 0.255 - - - Age (years) 1.01 0.98-1.04 0.544 - - - ASA Classifcation I ------II 0.66 0.24-1.81 0.417 - - - III 1.34 0.38-4.72 0.646 - - - Smoking 0.36 0.81-1.59 0.177 - - - BMI > 30 kg/m2 1.04 0.29-3.71 0.951 - - - Malignant disease (versus benign disease) 1.64 0.66-4.087 0.286 - - - Primary laparoscopic surgery (versus open) 0.36 0.13-0.99 0.047 0.46 0.16-1.30 0.455 Secondary diversion for anastomotic leakage 2.16 0.75-6.18 0.153 - - - (versus primary diversion) Loop ileostomy (versus end ileostomy) 0.20 0.76-0.54 0.001 0.18 0.07-0.52 0.001 Skin closure (versus primary closure) ------Purse string intracutaneous suture 0.31 0.07-1.31 0.111 - - - Approximating interrupted transcutaneous sutures 0.50 0.14-1.73 0.274 - - - Colorectal surgeon (versus any surgeon or resident) 0.27 0.12-0.64 0.003 0.28 0.11-0.67 0.005 Stapled anastomosis (versus hand-sewn) 0.55 0.21-1.43 0.220 - - - S-S anastomosis (versus S-E, E-S, E-E anastomosis) 0.82 0.33-2.01 0.658 - - - BMI = body Mass Index, ASA= American Society of Anaesthesiology, S-S= side to side, S-E=side to end, E-S= end to side, E-E=end to end

Follow-up Te median follow-up afer ileostomy reversal was 71.0 (IQR 14-200) weeks. Afer stoma reversal, a stoma was constructed again in 6.9% (10/144) of the patients in cohort A and in 10.9% (18/165) in cohort B (p=0.23). Tis secondary stoma was constructed due to anastomotic leakage (n=9), fstula (n=4), incontinence (n=3), construction of a pouch (n=3) progression of the primary disease (n=4), ileus (n=2), bowel perforation (n=2), and unknown reason (n=1). A clinical diagnosed stoma site hernia occurred in 6.3% (9/144) of the patients in cohort A, and in 4.2% (7/165) of the patients in cohort B (p=0.43). A stoma site hernia occurred afer a median of 9.5 (IQR 4.7-17.5) months and in 50.0% (8/16) of the patients a surgical correction was performed.

168 Morbidity afer ileostomy closure

DISCUSSION

Tis analysis shows that increased awareness of the risks of morbidity afer ileostomy closure and adapting surgical practice signifcantly improved postoperative outcome. Stoma reversal being performed or supervised by a colorectal surgeon and closure of a loop-ileostomy instead of an end- ileostomy were independently associated with a lower risk of major 30-day postoperative morbidity, with Odds Ratios of 0.19 and 0.32, respectively.

Increasing the role of a colorectal surgeon in performing ileostomy closure resulted in a signifcantly decreased major morbidity rate. Tis might be the result of a diference in surgical technique and surgical volume between colorectal surgeons and general surgeons, surgeons with another specialisation or unsupervised residents. Tis is a remarkable fnding, because surgical diferentiation and volume is ofen considered to be only relevant in high complex and/or low volume surgical procedures. However, ileostomy closure should probably be regarded as redo bowel surgery, requiring meticulous dissection and adequate tissue handling within a previously exposed operating feld with scar tissue. Adequate exposure of the operative feld may be challenging while operating through a small hole, especially in obese patients. Several problems may be encountered during stoma closure, such as restricted length of the mesentery, bleeding from epigastric vessels in the rectus sheath, and atrophy of the eferent loop. Closure of the stoma site requires adequate recognition of the abdominal wall structures and layered closure. Furthermore, experience with larger abdominal wall defects and additional reconstruction techniques are required. Terefore, restricting ileostomy reversal a to a selective group of specialized surgeons might decrease complication rates.

Literature regarding complex colorectal surgery already showed that higher surgical volume results in fewer defnitive stomas, betters overall survival in colorectal cancer patients and lower costs.5,6 Whether specialized high volume surgeons are performing the procedure themselves or have a supervising role does not seem to infuence the outcome. A trainee who is supervised by a colorectal surgeon can obtain similar quality results.7 However, literature on surgical and hospital volume is not conclusive and the defnition of high volume difers substantially between studies.8

End-ileostomy was an independent predictor of major morbidity. A possible explanation might be that identifcation and mobilization of the distal limb is more complicated because of its intra- abdominal localisation. Tis may require an extension of the stoma site opening, or even a midline laparotomy. Furthermore, a laparoscopic approach during index surgery seemed to be a predictor of reduced morbidity afer stoma closure in univariable analysis. However, primary laparoscopic surgery did not turn out to be an independent predictor in multivariate analysis anymore. Tis 10 might be the result of the increased use of laparoscopic surgery over time, without a benefcial impact itself on stoma reversal morbidity.

Postoperative ileus has been reduced from 9% to 5% over time, which might be the result of the increased use of a stapler device with construction of a side-to-side anastomosis. Because the distal limb is not functional for some time, the anastomosis is generally made to a relative small calibre distal limb if restored in an end-to-end confguration. Perioperative oedema might therefore

169 Chapter 10

compromise the luminal diameter causing an early bowel obstruction. With the stapled anastomosis, a greater calibre anastomosis is made which potentially reduces the risk of postoperative small bowel obstruction.9 Although subgroup analysis did not defnitively confrm this hypothesis, a non-signifcant trend in favour of stapled anastomosis was observed.

Besides the intraoperative measures, we postulated that increased awareness during the whole hospital admission has likely contributed to the improved outcome. It is hard to provide objective data to support this hypothesis, but this is something that is very well known from quality improvement programs using auditing. Auditing in itself is able to improve outcome, just because of getting insight into the process and being aware of items that need special attention.

Considering the potential risks associated with closure of a defunctioning ileostomy, one should critically look at its application as a routine afer colo-anal and ileo-anal anastomoses. An anastomotic leakage can signifcantly increase short- and long term morbidity, reduce the quality of life and might even increase the risk of local cancer recurrence.10;11 Preventing this severe complication seems to be attractive, but the question is what the role is of primary diversion. Besides the risk of complications associated with the presence of a stoma (i.e. dehydration, parastomal hernia, prolaps), the reversal of the stoma is associated with an additional risk of morbidity which has been described in literature to occur in up to 17% of the patients.12 In addition, a primary defunctioning stoma is not reversed in a substantial number of patients afer LAR and also afer IPAA surgery.13;14 A defnitive stoma might reduce the quality of life and might lead to an increase in medical costs. Terefore, a more selective approach for a defunctioning ileostomy is advocated.15 However, more research is needed to enable more appropriate patient selection.

Tis study is limited due to the retrospective and non-randomised design, which might have resulted in the omission of data and selection bias. Although a randomised controlled trial would be methodologically superior to a comparative cohort study, randomization between unselected trainees and surgeons versus colorectal surgeons performing ileostomy reversal would be unethical. Another risk of bias in this retrospective cohort study was the limited number of events for a multivariate analysis, which might have led to an underestimation of the independent variables. In addition, the primary disease of ulcerative colitis signifcantly difered between both groups. However the efect of this signifcant baseline diference on the primary outcome might be limited because ileostomy reversal was usually performed when the patient no longer used immunosuppressive medications. Despite these limitations, the present study clearly illustrates that ileostomy reversal is probably a more complex procedure than ofen considered, with a not negligible risk of major morbidity. Increasing the role of a colorectal surgeon might improve outcome afer ileostomy reversal.

170 Morbidity afer ileostomy closure

REFERENCE LIST

1. Huser N, Michalski CW, Erkan M, et al. Systematic review and meta-analysis of the role of defunctioning stoma in low rectal cancer surgery. Ann Surg 2008; 248(1): 52-60.

2. Chow A, Tilney HS, Paraskeva P, Jeyarajah S, Zacharakis E, Purkayastha S. Te morbidity surrounding reversal of defunctioning ileostomies: a systematic review of 48 studies including 6,107 cases. Int J Colorectal Dis 2009; 24(6): 711-723.

3. van Westreenen HL, Visser A, Tanis PJ, Bemelman WA. Morbidity related to defunctioning ileostomy closure afer ileal pouch-anal anastomosis and low colonic anastomosis. Int J Colorectal Dis 2012; 27(1): 49-54.

4. Dindo D, Demartines N, Clavien PA. Classifcation of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004; 240(2): 205-213.

5. Damle RN, Macomber CW, Flahive JM, Davids JS, Sweeney WB, Sturrock PR et al. Surgeon volume and elective resection for colon cancer: an analysis of outcomes and use of laparoscopy. J Am Coll Surg 2014; 218(6): 1223-1230.

6. Archampong D, Borowski D, Wille-Jorgensen P, Iversen LH. Workload and surgeon’s specialty for outcome afer colorectal cancer surgery. Cochrane Database Syst Rev 2012; 3: CD005391.

7. Maslekar S, Sharma A, Macdonald A, Gunn J, Monson JR, Hartley JE. Do supervised colorectal trainees difer from consultants in terms of quality of TME surgery? Colorectal Dis 2006; 8(9): 790-794.

8. van Erning FN, van Steenbergen LN, van den Broek WT, Rutten HJ, Lemmens VE. No diference between lowest and highest volume hospitals in outcome afer colorectal cancer surgery in the southern Netherlands. Eur J Surg Oncol 2013; 39(11): 1199-1206.

9. Markides GA, Wijetunga IU, Brown SR, Anwar S. Meta-analysis of handsewn versus stapled reversal of loop ileostomy. ANZ J Surg 2015; 85(4): 217-24.

10. Chambers WM, Mortensen NJ. Postoperative leakage and abscess formation afer colorectal surgery. Best Pract Res Clin Gastroenterol 2004; 18(5): 865-880.

11. McArdle CS, McMillan DC, Hole DJ. Impact of anastomotic leakage on long-term survival of patients undergoing curative resection for colorectal cancer. Br J Surg 2005; 92(9): 1150-1154.

12. Mennigen R, Sewald W, Senninger N, Rijcken E. Morbidity of Loop Ileostomy Closure afer Restorative Proctocolectomy for Ulcerative Colitis and Familial Adenomatous Polyposis: a Systematic Review. J Gastrointest Surg 2014; 18(12): 2192-2200.

13. den Dulk M, Smit M, Peeters KC, et al. A multivariate analysis of limiting factors for stoma reversal in patients with rectal cancer entered into the total mesorectal excision (TME) trial: a retrospective study. Lancet Oncol 2007; 8(4): 297-303.

14. Phang PT, Hain JM, Perez-Ramirez JJ, Madof RD, Gemlo BT. Techniques and complications of ileostomy takedown. Am J Surg 1999; 177(6): 463-466. 10

15. Pata G, D’Hoore A, Fieuws S, Penninckx F. Mortality risk analysis following routine vs selective defunctioning stoma formation afer total mesorectal excision for rectal cancer. Colorectal Dis 2009; 11(8): 797-805.

171

CHAPTER 11

Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection for rectal cancer

G.D. Musters, W.A. Borstlap, W.A. Bemelman, C.J. Buskens, and P.J. Tanis

Colorectal Dis 2015 Chapter 11

ABSTRACT

Aim: Tis study aimed to determine the clinical outcome of salvage surgery without restoring continuity for symptomatic chronic presacral sinus afer low anterior resection (LAR) for rectal cancer.

Methods: Out of a prospective cohort of 46 patients with chronic presacral sinus (>1 year afer LAR), 27 underwent completion proctectomy with omentoplasty between January 2005 and July 2014.

Results: Te initial treatment for rectal cancer included neoadjuvant radiotherapy in 26 (96%) patients. Besides a chronic presacral sinus, a secondary fstula was present in 15 (56%) patients. Defnitive salvage surgery was performed afer a median of 40 (12-350) months from the primary resection. Te median hospital stay afer single and multiple-stage salvage surgery was 11 and 17 days. Postoperative complications occurred in 44% of patients. Re-intervention rate was 33% with a range of 1 to 10 interventions per patient. During a median follow-up of 20 (4-45) months from salvage surgery, healing of the chronic presacral sinus occurred in 78% of patients, with a healing rate afer single and multiple-stage procedures of 88% and 64% respectively (p=0.19).

Conclusion: Patients with a symptomatic chronic presacral sinus afer LAR for rectal cancer, in whom restoration of continuity is not intended, can be efectively managed by completion proctectomy with complete debridement of the sinus and fstula tracts followed by an omentoplasty to fll the presacral cavity, preferably as a single stage procedure.

174 Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection

INTRODUCTION

Anastomotic leakage is one of the most dreaded complications of low anterior resection (LAR). A diverting loop ileostomy reduces the risk of a symptomatic leakage, but subclinical leaks can occur and may be diagnosed or become symptomatic well afer the initial postoperative period.1,2 Anastomotic leakage afer LAR is ofen treated with a defunctioning stoma if not already in place, combined with drainage of the presacral abscess via a trans-abdominal, trans-anal or percutaneous approach. A diverted anastomosis does not always heal, however despite adequate exclusion of faeces and drainage and a chronic presacral sinus may develop.

Te reported incidences of chronic presacral sinus of a few percent may be an underestimate of this difcult clinical problem.3,4 Te increased use of neoadjuvant radiotherapy is likely to reduce the rate of secondary healing afer anastomotic leakage, because of its impact on wound healing.5,6 As a result, fewer stomas will be reversed with a subsequent impact on morbidity and quality of life. Up to 20% of defunctioning stomas afer LAR are never reversed.6 Furthermore, secondary complications of a chronic presacral sinus, such as fstula formation and necrotizing fasciitis, may occur.3,7,8

Based on our initial experience with salvage surgery for chronic presacral sinus, basic treatment principles were defned to increase the success rate.9 As a result an increasing numbers of patients with a complicated chronic presacral sinus have been referred to our academic hospital. For most of these patients, restoring continuity is no longer achievable or advisable. Te purpose of the present study was to determine the outcome in the group of patients who had developed a chronic presacral sinus and who were then treated by completion abdominoperineal proctectomy with end colostomy and complete flling of the presacral cavity by an omentoplasty.

METHOD

All 46 consecutive patients with a chronic presacral sinus afer LAR for rectal cancer between January 2005 and July 2014 at our university hospital were prospectively registered in a database. A chronic presacral sinus was defned as a sinus diagnosed at least one-year afer primary LAR. Te presence and features of the sinus were investigated by radiological imaging and/or endoscopy (Figure 1). Te initial cohort of 22 patients included up to March 2012 has already been described in a previous study.9

Te chronic presacral sinus was symptomatic in all 46 patients with one or more of the following symptoms: continuous purulent discharge from the anus and/or fstula tracts, urinary or faecal discharge from fstula tracts, sacral pain, pain radiating to the leg, difculty in sitting and/or walking, chronic anaemia, the need for daily anal irrigation, the need for use of a pad, impaired physical condition with the inability to perform daily activities or being confned to bed. A sinus-related 11 fstula was defned as having one or more tracks originating from the presacral sinus either with an external opening in the perianal area, the gluteal region or at the level of the major trochanter of the femur (Figure 2) or with a communication to an adjacent organ such as the bladder or small bowel. Emergency presentations were mostly related to fstula tracts along the piriformis muscles with abscesses in the buttocks or legs, potentially giving rise to necrotizing fasciitis (Figure 3).

175 Chapter 11

Based on our initial experience, we defned basic treatment principles consisting of either constructing a new anastomosis or intersphincteric completion proctectomy end colostomy and flling of the presacral cavity.9 For the purpose of the present study, the subgroup of patients who underwent single or multiple-stage salvage surgery including completion proctectomy and omentoplasty were selected from the enlarged cohort of 46 patients with extended follow-up. We further optimized the approach of salvage surgery based on monitoring of our results.9

In our initial experience, a trans-perineal intersphincteric completion proctectomy without abdominal exploration was performed where the anastomosis had already been dismantled and an end-colostomy had been formed. Even if an omentoplasty had been performed at the time of leakage of the anastomosis (Figure 4), resecting the rectal stump only was associated with a high risk of persisting presacral abscess formation. In such multiple stage procedures, the omentum does not completely fll the presacral cavity and an additional abdominal exploration is subsequently required to mobilize the omentum and bring it down to the level of the external sphincter as a ‘pull through’ procedure. Alternatively, it may be necessary to use a muscle transposition fap. We therefore modifed our surgical strategy to a one stage abdominoperineal approach with simultaneous resection of the disrupted anastomosis including the rectal stump, debridement of the sinus and fstula tracts, and a ‘pull through’ omentoplasty (Figure 4).

Figure 1 Presacral sinus Sagittal (a) and transverse (b) image of a pelvic MRI, showing a colorectal anastomosis in connection with a presacral sinus. Endoscopic image (c) of a side-to-end colorectal anastomosis with a posterior located defect (*), which is the entrance to the presacral sinus (d).

176 Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection

Data collection Patient and treatment characteristics were retrospectively collected from patient records. Operation reports, pathology reports, endoscopic reports, radiology reports and patient charts were searched for demographics, primary treatment characteristics, tumour characteristics, symptoms related to the presacral sinus, secondary complications, hospital stay, preceding interventions or re- interventions (radiological, endoscopic and surgical), and disease status at the last date of follow-up. Te primary endpoint was healing of the sinus and secondary fstula tracts. Healing was defned as the absence of discharge and symptoms related to a persisting presacral sinus during follow- up. Healing rate of a one-stage abdominoperineal approach was compared with multiple-stage salvage procedures.

Statistical analysis Te median with range or interquartile range (IQR) was used to express non-parametric data. Categorical data were analysed using the Fisher’s exact test and continuous variables were analysed using the Mann-Whitney-Wilcoxon test. All analyses were performed with IBM SPSS Statistics, version 20.0.0 (IBM Corp., Armonk, NY, United States). Te Institutional Review Board (IRB) declared exemption from approval for this study.

Figure 2 Two main types of fstula’s originating from a chronic presacral sinus Two main types of fstula’s originating from a chronic presacral sinus: a gluteal fstula (a) with sagittal MRI images 11 showing the gluteal fstula tract in two other patients (b, c), and a trochanteric fstula (d) with a fstula tract/abscess along the piriformis muscle as shown on a transverse CT image (e).

177 Chapter 11

RESULTS

Patients Twenty-seven patients with a chronic presacral sinus afer LAR for rectal cancer underwent single or multiple stage salvage surgery including completion proctectomy, debridement of the sinus and fstula tract(s) if present, and cavity flling using an omentoplasty (Table 1). Tirteen of the 27 had been described in our previous report9, but four of these had undergone further surgery since the publication because of non-healing of the sinus. Te details of the remaining 19 patients, who were excluded for the present analysis, are summarized in Table 2.

Table 1 Demographics of 27 patients treated for presacral sinus afer low anterior resection.

n=27

Gender Male (%) 17 (63) Age Median age (years, IQR) 66 (59-72) Comorbidity Smoker (n, %) 3 (11) Diabetes Mellitus (n, %) 2 (7) Referral Number of patients referred (n, %) 23 (85) Median time between primary surgery and referral (months, IQR) 32 (18-71) Neoadjuvant radiotherapy Short course 5x5 Gy (n, %) 21 (78) Long course with concomitant chemotherapy (n, %) 5 (19) No radiotherapy (n, %) 1 (4) Low anterior resection Laparoscopic surgery (n, %) 11 (41) (LAR) Primary anastomosis with diverting ileostomy (n, %) 19 (70) Primary anastomosis with diverting colostomy (n, %) 3 (11) Primary anastomosis without faecal diversion (n, %) 5 (19) Side to end anastomosis (n, %) 15 (56) End to end anastomosis (n, %) 6 (22) Unknown type of anastomosis (n, %) 6 (22) Clinical signs of an anastomotic leakage (n, %) 10 (37) Other postoperative complicationsa (n, %) 5 (19)

Primary tumour γp T0-2 (n, %) 10 (37) pathological stage γp T3-4 (n, %) 17 (63)

γp N0 (n, %) 17 (63)

γp N+ (n, %) 10 (37)

γp M0 (n, %) 27 (100) Median tumour distance from anal verge (cm, IQR) 6 (4-8) Median tumour size (cm, IQR) 3 (2-4) Adjuvant therapy Chemotherapy (n, %) 5 (19) Time interval First sign of presacral abscess afer LAR (months, IQR) 1 (0-32)

Data are expressed as median with interquartile range (IQR) or percentage. a= abscess (n=2), urinary tract infection (n=1), postoperative bleeding (n=1), and surgical site infection (n=1).

178 Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection

Table 2 Patients with chronic presacral sinus not included in the present analysis

Bowel continuity or stoma Management n=19 Restored bowel continuity Repeated colo-anal anastomosis 10 Observation, old anastomosis in situ, patient refused further surgery 2 Diverting ileostomy Observation, old anastomosis in situ, mild symptoms, patient refused 1 further surgery Observation, old anastomosis in situ, Endo-sponge, 1 End-ileostomy Observation, patient condition does not allow for further surgery 1 End-colostomy Observation, patient went back to referring hospital 1 Observation, patient refuses further surgery 1 End-colostomy with stapled Observation, patient has mild symptoms and refuses further surgery 1 eferent loop in situ Other Marsupialization 1

Initial treatment for rectal cancer included neoadjuvant (chemo)radiotherapy in 26 (96%) patients. In 10 (37%) of the 27 patients anastomotic leakage occurred in the early postoperative period. Overall, the median interval from the operation to the diagnosis of anastomotic leakage was one (0-32) month. Te anastomoses had been diverted in all patients during the period between LAR and salvage surgery, by primary diverting stoma in 23 and a secondary diverting stoma in four. Te diverting stoma was still in situ at the time of salvage surgery in 17 of the 27 patients. In total, 23 of 27 (85%) patients were referred with a median time between diagnosis of anastomotic leakage and referral of 17 months (IQR 10-38).

Presacral sinus and salvage surgery All patients had a chronic presacral sinus associated with the anastomotic defect and an additional fstula originating from the presacral sinus had developed in 15 (56%) of the 27 patients, including a fstula to the perineum (n=4), the gluteal or trochanteric region (n=6), the small bowel (n=3), the large bowel (n=1) and the bladder (n=1). Other secondary complications related to the chronic presacral sinus consisted of hydronephrosis due to fbrosis around the ureter (n=2), signifcant anal blood loss (n=1) and sepsis (n=1). Interventions that had been performed before salvage surgery are summarized in Table 3.

Te median interval between primary LAR and salvage surgery for symptomatic chronic presacral sinus was 40 (12-350) months. Salvage surgery was performed as a single stage abdominoperineal procedure in 16 (59%) of 27 patients and as a multiple stage procedure in 11 (41%) patients. Te rate of single stage salvage procedures increased from 5 of 13 (38%) patients up to March 2012, to 11 of 14 (79%) patients since (p=0.03). Salvage surgery was open in 22 patients and hand assisted laparoscopic combined with open surgery through a Pfannenstiel incision in fve patients. 11

179 Chapter 11

Figure 3 Extension of infection from a chronic presacral sinus along the piriformis muscles Extension of infection from a chronic presacral sinus along the piriformis muscles as shown by PET-CT imaging (a) with the risk of subsequent sof tissue infltration including air bubbles in the right gluteal (b) and lef trochanteric (c) area as demonstrated by coronal CT images of two other patients, which may result in necrotizing fasciitis of the leg as demonstrated by a sagittal CT image of a fourth patient (d).

Outcome Te median hospital stay was 13 (IQR 8-17) days. It was 11 (IQR 8-14) days for one-stage procedures and 17 (IQR 11-21) days for multiple stage procedures (p=0.09). One patient was admitted to the intensive care unit for 37 days due to postoperative sepsis and respiratory failure. Another patient was readmitted at the referring hospital afer an uncomplicated postoperative course and died due to sepsis of unconfrmed origin 32 days afer single stage salvage surgery. Overall, 15 postoperative complications occurred in 12 (44%) of the 27 patients. Tese consisted of abdominal wound infection (n=4), urinary retention (n=2), acute renal failure (n=1), ileus (n=1), delirium (n=1), respiratory insufciency (n=1), pneumonia (n=1), small bowel perforation (n=1), gastroparesis (n=1), urinary tract infection (n=1), and urosepsis (n=1). Te perineal wound was lef to heal by secondary intention in all patients.

In nine patients, further surgical and non-surgical re-interventions were required in 20 and seven patients (Table 4). An autologous tissue fap was performed afer failed omentoplasty in three patients. Te frst underwent completion exenteration because of a bladder fstula with flling of the wound by a superior gluteal artery perforator (SGAP) fasciocutaneous transposition fap.10 In the second, a vertical rectus abdominus muscle (VRAM) fap was performed and in the third a gracilis and VRAM fap both failed and a subsequent gluteal VY plasty resulted in healing of the presacral sinus.

180 Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection

Te median follow-up from salvage surgery was 20 months (IQR 4-45). Successful closure of the presacral sinus occurred in 21 (78%) of 27 patients afer a median of four (IQR 1-31) months. Tis occurred in 14 (88%) of 16 patients when a one-stage procedure was performed and in 7 (64%) of 11 patients afer multiple-stage surgery (p=0.19). Te sinus eventually healed in all three patients undergoing an autologous tissue fap afer failed omentoplasty. Of the 15 fstulas originating from the presacral sinus, all except for one were healed at the end of follow-up.

Table 3 Interventions preceding salvage surgery

n=27 Surgical interventions Per patient (n, range) 2 (0-4) All interventions performeda Per patient (n, range) 3 (0-22) Endoscopic Endo-sponge® treatment (n) 3 Radiological Percutaneous drainage (n) 8 Percutaneous nephrostomy (n) 14 Surgery Trans anal drainage (n) 9 Closure of diverting ileostomy (n) 15 Closure of diverting colostomy (n) 8 Diverting colostomy (n) 18 Diverting ileostomy (n) 5 Diverting jejunostomy (n) 1 Sigmoid resection (n) 1 Abscess drainage (n) 18 Muscle fap (n) 2 New colo-anal anastomosis (n) 1 Lords Procedure (n) 1 Resection of rectal stump (n) 1 Coccyx resection (n) 1 Segmental ileum resection (n) 2 Debridement necrotizing fasciitis (n) 1 Abdominal wall closure (n) 1 Appendectomyb (n) 1 Other Hyperbaric oxygen therapy (n) 1 Data are medians with range or number of interventions a=more than one procedure could be performed during the same operation. b=infamed due to presacral infection. 11

181 Chapter 11

Table 4 Re-interventions afer intersphincteric APR with omentoplasty

n=27 Surgical re-interventions Percentage of patients 19% Total number per patient (range) 1-7 All re-interventionsa Percentage of patients 33% Total number per patient (range) 1-10 Endoscopic Endo-sponge® treatment (n) 1 Radiological Percutaneous drainage (n) 4 Surgery Surgical drainage (n) 4 Fistula excision (n) 2 Revision end colostomy (n) 4 Muscle fap (n) 4 Removal of muscle fap (n) 2 Debridement sinus (n) 1 Bladder excision (n) 1 Ureter-jejuno-cutaneostomyb (n) 1 Ileo-coecal resectionc (n) 1 Other VAC therapy (n) 1 Hyperbaric oxygen therapy (n) 1

VAC=vacuum assisted wound closure. APR=abdominoperineal resection. a= more than one procedure could be performed during the same operation. b= for ureter stenosis. c= due to stenosis resulting in ileus.

DISCUSSION

Chronic presacral sinus afer LAR for rectal cancer is a complex problem with a risk of secondary complications if not treated appropriately.9 Growing experience in salvage surgery for the condition at our institute has improved the results. In patients in whom restoration of continuity was no longer possible, perineal wound healing was achieved in 78% of patients by intersphincteric resection of the anorectal stump and flling of the presacral cavity with well-vascularized tissue. When this was performed as a single-stage abdominoperineal procedure, healing occurred in 88%.

Pain and bowel dysfunction may indicate the presence of a chronic presacral sinus, but ofen remain misinterpreted until pelvic imaging is performed or a secondary fstula has arisen. Initial management at the referring hospital mostly consisted of faecal diversion and drainage (Table 3). Tese measures may reduce symptoms but do not result in healing because of fbrosis due to the long-lasting pelvic infammation and prior radiotherapy.11,12 Radiotherapy is probably the major factor preventing healing despite defunctioning, giving rise to complication as long as 30 years afer the initial treatment. All but one patient in the present series had had neoadjuvant radiotherapy. In

182 Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection

recent years, patients were ofen referred much earlier as we became a national referral centre for such low volume revision surgery, which explains the increase from 38% to 79% in the proportion of referrals having salvage surgery. Early aggressive surgery seems justifed, because of the high success and acceptable complication rate, given the complexity of the condition.

Tis presacral dead space afer removal of the anorectal stump in an irradiated pelvis will only heal by flling it with well-vascularized tissue. Myocutaneous faps such as the VRAM fap and gracilis fap have been described for this purpose9,13, but they may cause morbidity at the donor site and can undergo necrosis.13 For this reason omentoplasty was preferred. Te omentum is well vascularized and outside the feld of irradiation; it improves the local immune response and promotes angiogenesis, increasing the chance of healing.14-16 In addition omentoplasty prevents prolapse of small intestine into the pelvis, which may decrease the risk of obstruction or fstula formation. Filling of the pelvic cavity by an omentoplasty can be difcult in thin patients or when an omentoplasty has already been placed on to the rectal stump at the time of dismantling the anastomosis. Te omentum then needs to be mobilized and brought down to the pelvic foor, usually requiring complete mobilization along the greater curvature of the stomach to achieve additional length. Bringing it through the mesentery results in the shortest route to the pelvis (Figure 4).

Te study is limited by the small number of patients, despite it being the largest published cohort of such patients to our knowledge. Tere is also a large element of selection bias whereby owing to the severity of their condition, the patients referred may not be representative of the whole population of chronic presacral sinus some of whom may be asymptomatic. In some cases with minor symptoms minimally invasive treatment such as fbrin glue injection along the track or stapler marsupialization have been reported to have a success rate in the short term.17,18 Clearly adequate follow-up in such patients is required, because secondary complications may occur even several years later.

In conclusion, a chronic presacral sinus may become a serious septic clinical problem almost exclusively in radiated patients. When restoration of continuity is not intended and the sinus is symptomatic or associated with secondary infection, there is a high chance of successful closure afer a one-stage abdominoperineal intersphincteric completion proctectomy, debridement of the sinus and omentoplasty. If the omentoplasty fails, various plastic fap procedures are available. Te complexity of this rare long-term complication requires centralized treatment in a specialist colorectal unit.

11

183 Chapter 11

Figure 4 Persisting infectious problems afer dismantling of the anastomosis and omentoplasty, but with the rectal stump still being in situ Persisting infectious problems afer dismantling of the anastomosis and omentoplasty, but with the rectal stump still being in situ as shown on sagittal MRI (a). Intersphincteric completion proctectomy with debridement of a gluteal fstula (b). Omentoplasty mobilized along the splenic hilum and greater curvature of the stomach, with the right gastroepiploic artery as feeding pedicle, tunnelled beneath the transverse colon and terminal ileum to the pelvic cavity (c). ‘Pull-trough’ omentoplasty visible within the external sphincter afer intersphincteric resection of the rectal stump (d).

184 Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection

REFERENCE LIST

1. Tan WS, Tang CL, Shi L, Eu KW. Meta-analysis of defunctioning stomas in low anterior resection for rectal cancer. Br J Surg 2009; 96(5): 462-72. 2. Snijders HS, Bakker IS, Dekker JW, et al. High 1-year complication rate afer anterior resection for rectal cancer. J Gastrointest Surg 2014; 18(4): 831-8. 3. Arumainayagam N, Chadwick M, Roe A. Te fate of anastomotic sinuses afer total mesorectal excision for rectal cancer. Colorectal Dis 2009; 11(3): 288-90. 4. van Koperen PJ, van der Zaag ES, Omloo JM, Slors JF, Bemelman WA. Te persisting presacral sinus afer anastomotic leakage following anterior resection or restorative proctocolectomy. Colorectal Dis 2011; 13(1): 26-9. 5. Musters GD, Sloothaak DA, Roodbeen S, van Geloven AA, Bemelman WA, Tanis PJ. Perineal wound healing afer abdominoperineal resection for rectal cancer: a two-centre experience in the era of intensifed oncological treatment. Int J Colorectal Dis 2014. 6. den Dulk M, Smit M, Peeters KC, et al. A multivariate analysis of limiting factors for stoma reversal in patients with rectal cancer entered into the total mesorectal excision (TME) trial: a retrospective study. Lancet Oncol 2007; 8(4): 297-303. 7. Killeen S, Souroullas P, Ho TH, et al. Outcomes of asymptomatic anastomotic leaks found on routine postoperative water-soluble enema following anterior resection for cancer. World J Surg 2013; 37(11): 2700-4. 8. Marinatou A, Teodoropoulos GE, Karanika S, et al. Do anastomotic leaks impair postoperative health- related quality of life afer rectal cancer surgery? A case-matched study. Dis Colon Rectum 2014; 57(2): 158-66. 9. Sloothaak DA, Buskens CJ, Bemelman WA, Tanis PJ. Treatment of chronic presacral sinus afer low anterior resection. Colorectal Dis 2013; 15(6): 727-32. 10. Musters GD, Lapid O, Bemelman WA, Tanis PJ. Surgery for complex perineal fstula following rectal cancer treatment using biological mesh combined with gluteal perforator fap. Tech Coloproctol 2014; 18(10): 955-9. 11. Krol R, Hopman WP, Smeenk RJ, van Lin EN. Increased rectal wall stifness afer prostate radiotherapy: relation with fecal urgency. Neurogastroenterol Motil 2012; 24(4): 339-e166. 12. O’Brien PC. Radiation injury of the rectum. Radiother Oncol 2001; 60(1): 1-14. 13. Borel Rinkes IH, Wiggers T. Gracilis muscle fap in the treatment of persistent, infected pelvic necrosis. Eur J Surg 1999; 165(4): 390-1. 14. Goldsmith HS, Grifth AL, Kupferman A, Catsimpoolas N. Lipid angiogenic factor from omentum. JAMA 1984; 252(15): 2034-6. 15. Konturek SJ, Brzozowski T, Majka I, Pawlik W, Stachura J. Omentum and basic fbroblast growth factor in healing of chronic gastric ulcerations in rats. Dig Dis Sci 1994; 39(5): 1064-71. 16. Walker FC, Rogers AW. Te greater omentum as a site of antibody synthesis. BrJ Exp Pathol 1961; 42: 222-31. 17. Swain BT, Ellis CN. Fibrin glue treatment of low rectal and pouch-anal anastomotic sinuses. Dis Colon Rectum 2004; 47(2): 253-5. 11 18. Alsanea N, Alabbad S. Use of the endostapler for the treatment of non-healing sinus secondary to a dehisced colorectal anastomosis. Tech Coloproctol 2010; 14(3): 249-51.

185

Summary and future perspectives

Summary and future perspectives

188 Summary and future perspectives

SUMMARY AND FUTURE PERSPECTIVES

Treatment related morbidity is one of the important factors that should be taken into account in determining patient tailored treatment strategies. Although several techniques have been described, minimally invasive techniques need to be explored to limit post procedural morbidity. In the beginning of this thesis we describe the results of a minimally invasive technique in the removal of rectal adenomas. In the following chapters of this thesis, predictive factors, prevention, and surgical treatment strategies for complications afer rectal cancer surgery are being described.

In chapter 1 of this thesis we discuss the safety and efectiveness of endoscopic mucosal resection (EMR) for large rectal adenomas. Sixty-four patients underwent EMR for rectal adenomas larger than 3 cm. In this multicentre cohort study, EMR was a feasible treatment option in 97% of the patients. However, the recurrence rate of a rectal adenoma at the site of removal was 25% within two years. Currently, a multicentre randomised controlled trial is being fnalised in which EMR is being compared to transanal endoscopic microsurgery (TEM) in relation to safety, efciency and medical costs. Tis study will provide insight in the optimal technique to remove large rectal adenomas (NTR1422; trialregister.nl).

Perineal wound infection is one of the most common clinical complications afer abdominoperineal resection (APR). Chapter 2 describes a systematic review with meta-analysis and determines the impact of the extralevator abdominoperineal resection (eAPR) and neoadjuvant (chemo)radiotherapy on perineal wound healing. A total of 32 cohort studies were included and showed that neoadjuvant (chemo)radiotherapy signifcantly increases the risk of perineal wound complications. Several treatment approaches for patients undergoing APR have changed over time. Chapter 3 describes a two-intuitional experience with APR for rectal cancer. In this chapter, the eAPR and intra-operative perforations were found as independent predictors of impaired perineal wound healing. Over time, the use of neoadjuvant (chemo)radiotherapy increased from 70% to 91% and the use of an extralevator approach increased from 9% to 19%. As a result, the rate of perineal wound infections increased from 18% to 31% of the patients. To overcome this frequent clinical problem, standard primary perineal wound closure technique afer eAPR are being replaced by other closure techniques. One of those techniques is the placement of a biological mesh in the pelvic foor. A subgroup analysis described in chapter 3 showed that biological mesh assisted wound closure following eAPR afer neoadjuvant radiotherapy results in a perineal wound problem rate of 7%, whereas afer primary perineal wound closure a perineal wound problem rate occurred in 30% of the patients. Although promising, these results are all based on cohort studies and do not provide sufcient evidence to justify their daily clinical use. As a result, a single blinded multicentre randomised controlled trial was started. Te study protocol of this multicentre randomised controlled trial is being described in chapter 4. In this trial, patients with distal rectal cancer undergoing an eAPR afer neoadjuvant (chemo)radiotherapy are randomised between the standard primary perineal wound closure and biological mesh assisted perineal wound closure. Chapter 5 describes the results of the 104 patients that were randomised. Afer a follow- up period of one year, no statistical signifcant diference in perineal wound healing was found between both perineal wound closure techniques. In addition, no signifcant diference between

189 Summary and future perspectives

surgical complications, non-surgical complications or quality of life was demonstrated. Terefore, perineal wound healing afer neoadjuvant (chemo)radiotherapy and eAPR for distal rectal cancer seem not to improve perineal wound healing afer biological mesh assisted perineal wound closure in comparison to primary perineal wound closure. However, a signifcant diference in perineal herniation between both randomisation groups was found. Tis diference however might be the result of a limited follow-up period of only one year. A longer follow-up period might give better insight in the efect of biological mesh degradation and perineal herniation. Tis extended follow- up is especially important, because in chapter 7 we describe the perineal hernia recurrence rate of patients undergoing perineal hernia reconstruction with a biological mesh following APR. In this cohort study, 15 patients were followed for a median duration of 17 months and showed that 47% of the patients developed a recurrent perineal hernia. Based on the largest published cohort so far in literature, a perineal hernia recurrence rate afer synthetic meshes of only 5% has been described. Terefore we concluded that a biological mesh seems not to be the frst choice implant for perineal hernia repair afer APR in the absence of contamination.1

Although biological meshes do not seem to reduce perineal wound healing problems, other techniques need further exploration to overcome this problem. Chapter 6 describes a systematic review of the local application of gentamicin in the pelvis, which may serve as a treatment option. In this systematic review, eight studies were identifed of which four were randomized controlled trials. Te underlying disease, the type of APR performed, the location of the local gentamicin application, and the method for applying gentamicin in the perineal wound difered among the included studies. As a result, we were not able to perform a pooled analysis and concluded that the impact of gentamicin on perineal wound healing is inconclusive. Terefore, its use in daily clinical practise is not supported by the currently available literature. Besides the local application of gentamicin, an omentoplasty might also have an efect on perineal wound healing.2 An omentoplasty has an efect on the local immune response and flls up the pelvic cavity afer rectal excision. A downside of this technique might be the increase in operation time, especially when laparoscopic surgery is performed. However, no randomised controlled trials have been conducted to support its use.2 Another technique that seems promising is the use of a musculocutaneous or fasciocutaneous fap for perineal wound closure.3 Te limited evidence available seems to demonstrate a reduction in perineal wound problems. However this is at the cost of an increased operation time, the requirement of a specialist surgeon, donor site morbidity and fap failure. Currently, a single centre, randomized controlled trial is being conducted to investigate if perineal wound healing increases afer gluteal plasty in comparison to biological mesh assisted perineal wound closure afer eAPR (ClinicalTrials. gov Identifer: NCT01347697).

In patients with a complex perineal fstula afer extensive pelvic surgery, surgical treatment is highly complex and needs combination of techniques to solve the problem. In chapter 8 we describe a case series in which the pelvic foor is reconstructed with a biological mesh and the sof tissue defect below the biological mesh is flled with a unilateral superior gluteal artery perforator fap. Te postoperative course was uneventful in all three patients. Terefore, considering all reconstructive techniques and combining remaining options afer previous extensive pelvic surgery can solve difcult infectious problems.

190 Summary and future perspectives

Anastomotic leakage afer low pelvic anastomosis has various treatment options, but not always successful. Against all surgical principles, it was hypothesized that an anastomotic defect could be closed transanally afer the presacral abscess cavity was cleaned using a short course of Endo- sponge® treatment. Chapter 9 describes the outcome of this new technique and compares it with the conventional treatment in patients with an anastomotic leakage afer ileal pouch-anal anastomosis for ulcerative colitis and familial adenomatous polyposis. Secondary anastomotic healing was achieved in all patients in the early surgical closure group, which was signifcantly higher compared to 52% in the conventional treatment group, without a signifcant diference in direct medical costs. Although promising, more research is needed to evaluate if this success rate can also be achieved in rectal cancer patients who underwent low anterior resection, especially afer neoadjuvant radiotherapy. To answer this question, a prospective cohort study including 30 patients has been performed and the results of this novel technique in rectal cancer patients will become available in 2017(NTR4671; trialregister.nl).

A diverting stoma is generally considered to be a routine part of a low pelvic anastomosis. Tis is because several studies have shown a decrease in (clinical) anastomotic leakage rate.4 However, morbidity and even mortality related to a diverting stoma and its reversal afer low anterior resection (LAR) is probably a neglected aspect of this clinical topic. Afer the publication of the results of ileostomy reversal at our clinical practice, several changes were made in order to improve outcome. In chapter 10 we described that supervision by a specialist colorectal surgeon has increased, the anastomosis technique has been modifed and the awareness of postoperative complications have increased. Tese changes have led to a signifcant reduction in major morbidity afer ileostomy closure from 11% to 4%. Reversal of a stoma supervised or performed by a colorectal surgeon turned out to be independently associated with reduced major morbidity. Reversing an end ileostomy was an independent risk factor for major morbidity. Tis implicates that the reversal of a diverting ileostomy may not be considered a residents procedure and that it should only be performed under direct supervision of a colorectal surgeon.

When treatment of an anastomotic leakage afer LAR fails, an anastomotic defect with a presacral abscess may persist over one year. Te outcome of surgical treatment in symptomatic chronic presacral sinus patients is scarcely being described in literature. In chapter 11 we describe a treatment option, which is derived from a previous publication. Patients with a symptomatic chronic presacral sinus afer LAR for rectal cancer, in whom restoration of continuity is not intended, can be managed by completion proctectomy with complete debridement of the sinus and fstula tracts followed by an omentoplasty to fll the presacral cavity. Tis technique resulted in healing of the chronic presacral sinus in 78% of patients, afer previous treatment failed.

191 Summary and future perspectives

REFERENCE LIST

1. Martijnse IS, Holman F, Nieuwenhuijzen GA, Rutten HJ, Nienhuijs SW. Perineal hernia repair afer abdominoperineal rectal excision. Diseases of the colon and rectum 2012; 55(1): 90-5.

2. Killeen S, Devaney A, Mannion M, Martin ST, Winter DC. Omental pedicle faps following proctectomy: a systematic review. Colorectal disease 2013; 15(11): e634-45.

3. Foster JD, Pathak S, Smart NJ, et al. Reconstruction of the perineum following extralevator abdominoperineal excision for carcinoma of the lower rectum: a systematic review. Colorectal disease 2012; 14(9): 1052-9.

4. Wu SW, Ma CC, Yang Y. Role of protective stoma in low anterior resection for rectal cancer: a meta- analysis. World J Gastroenterol 2014; 20(47):18031-7.

192 Summary and future perspectives

193

Samenvatting en toekomstige perspectieven

Samenvatting en toekomstige perspectieven

196 Samenvatting en toekomstige perspectieven

SAMENVATTING EN TOEKOMSTIGE PERSPECTIEVEN

Morbiditeit is een belangrijke factor waarmee rekening gehouden moet worden bij het bepalen van de behandeling van het rectumcarcinoom en de voorlopers hiervan. Ondanks dat diverse technieken zijn ontwikkeld voor het verwijderen van grote rectumadenomen, moeten minimaal invasieve technieken verder worden onderzocht om de postprocedurele morbiditeit te verminderen. In het begin van dit proefschrif worden de resultaten van een dergelijke minimaal invasieve techniek beschreven. In de daaropvolgende hoofdstukken worden studies beschreven naar voorspellende factoren, preventie en chirurgische behandelingen van complicaties na rectumcarcinoom chirurgie.

In hoofdstuk 1 van dit proefschrif wordt de veiligheid en efectiviteit van de endoscopische mucosale resectie (EMR) voor grote rectumadenomen beschreven. In totaal werden 64 patiënten geïncludeerd die voor een groot rectumadenoom (>3cm) een EMR ondergingen. In deze multicenter cohort studie, was het in 97% van de patiënten mogelijk het gehele adenoom endoscopisch te verwijderen. Echter, het recidiefpercentage van een rectumadenoom op de plek van verwijdering is 25% binnen twee jaar. Momenteel wordt een multicenter studie afgerond waarin wordt gerandomiseerd tussen EMR en transanale endoscopische microchirurgie (TEM). Deze studie zal beter inzicht geven in de beste verwijderingstechniek voor grote rectumadenomen (NTR1422; trialregister.nl).

Wanneer een abdominoperineale resectie (APR) is vereist, is een perineale wondinfecties één van de meest voorkomende complicaties. In hoofdstuk 2 is een systematisch literatuuronderzoek met meta-analyse uitgevoerd om te achterhalen wat de impact van de extralevator abdominoperineale resectie (eAPR) en neoadjuvante (chemo)radiotherapie is op perineale wondgenezing. In totaal, werden er in dit onderzoek 32 cohort studies geïncludeerd. Hieruit blijkt dat neoadjuvante (chemo)radiotherapie leidt tot signifcant meer perineale wondproblemen. In hoofdstuk 3 blijkt uit de ervaringen van twee ziekenhuizen dat de eAPR en een intra-operatieve perforatie van de darm onafankelijke voorspellers zijn voor het krijgen van meer perineale wondproblemen. In de afgelopen decennia is voor de behandeling van het rectumcarcinoom veel veranderd. Zo is het gebruik van neoadjuvante (chemo)radiotherapie toegenomen van 70% naar 90% en is de extralevatoire techniek toegenomen van 9% naar 19%. Deze veranderingen zijn verantwoordelijk voor een stijging in het aantal perineale wondproblemen van 18% naar 31%. Een oplossing die voor dit frequente probleem momenteel in de literatuur wordt aangedragen, is het inhechten van een biologische mat op de plek waar de bekkenbodemspieren zich bevonden. Uit een subgroep analyse van hoofdstuk 2 blijkt dat het inhechten van een biologische mat na neoadjuvante (chemo)radiotherapie en eAPR in 7% van de patiënten perineale wondproblemen geef, vergeleken met 30% als de wond primair wordt gesloten. De biologische matten worden wereldwijd toegepast voor deze indicatie, terwijl het gebaseerd is op cohort onderzoek van onvoldoende kwaliteit. Dit heef ertoe geleid dat er een multicenter gerandomiseerde gecontroleerde studie is opgezet waarvan het protocol wordt beschreven in hoofdstuk 4. In deze studie worden patiënten met een distaal rectumcarcinoom welke na neoadjuvante (chemo)radiotherapie een eAPR ondergaan gerandomiseerd tussen standaardzorg waarbij het perineum primair wordt gesloten en reconstructie van de bekkenbodem met een biologische mat. In hoofdstuk 5 worden de resultaten van deze studie beschreven waarin 104 patiënten zijn geïncludeerd. In tegenstelling tot de systematische literatuuronderzoek van hoofdstuk 2, is er over een periode van één jaar, geen statistisch signifcant

197 Samenvatting en toekomstige perspectieven

verschil gevonden tussen primaire perineale wondsluiting en het gebruik van een biologische mat. Tevens werd er geen verschil gevonden in chirurgische complicaties, niet chirurgische complicaties of kwaliteit van leven tussen de patiënten met of zonder een biologische mat. Hieruit hebben wij geconcludeerd dat het inhechten van een biologische mat geen signifcante verbetering geef in perineale wondgenezing na neoadjuvante (chemo)radiotherapie en eAPR. Het verschil in resultaten tussen hoofdstuk 2 en hoofdstuk 3 zijn meest waarschijnlijk een gevolg van de methodologische tekortkomingen in de geïncludeerde studies van hoofdstuk 2. Wel werd in hoofdstuk 5 signifcant minder perineale hernia’s geconstateerd in de groep waarbij een biologische mat werd geplaatst. Mogelijk is dit verschil te verklaren door de beperkte follow-up van één jaar. Bij een langere follow- up duur is de biologische mat volledig geïncorporeerd en zou dan mogelijk een andere inkijk kunnen geven in het percentage perineale hernia’s. Deze langere follow-up is vooral belangrijk omdat in hoofdstuk 7 het perineale hernia recidiefpercentage wordt beschreven na biologische mat reconstructie voor een perineale hernia. In deze cohort studie werden 15 patiënten gevolgd voor een mediaan van 17 maanden en daarvan ontwikkelde in 47% van de patiënten een recidief perineale hernia. Terwijl gebaseerd op het grootste gepubliceerde cohort onderzoek tot nu toe het huidige recidiefpercentage na een synthetische mat maar op 5% ligt. Uit deze studie valt dan ook op te maken dat een perineale hernia reconstructie met een biologische mat niet de eerste keus is in een niet gecontamineerde wond.1

Gezien het uitblijven van het beoogde efect van de biologische mat op perineale wondgenezing zullen er nieuwe technieken gevonden moeten worden om het percentage perineale wondinfecties te verminderen. Hoofdstuk 6 beschrijf in een systematisch literatuuronderzoek, de lokale applicatie van gentamicine in de perineale wond als mogelijke techniek om perineale wondproblemen te verminderen. In deze literatuurstudie werden acht studies geïncludeerd, waarvan vier gerandomiseerde studies. Het onderliggende ziektebeeld, type APR die werd uitgevoerd, de locatie van de gentamicine applicatie en de methode van aanbrengen van de lokale gentamicine, verschilde tussen de geïncludeerde studies. Hierdoor kon geen gepoolde analyse worden uitgevoerd en werd geconcludeerd, dat het standaard gebruik van lokale gentamicine in de perineale wond momenteel niet ondersteund kan worden met voldoende bewijs. Behoudens de applicatie van lokale gentamicine zou mogelijk toekomstig onderzoek ook gericht kunnen worden op het gebruik van een omentumplastiek in het kleine bekken.2 Van een omentumplastiek wordt momenteel verondersteld dat het efect heef op de lokale immuunrespons. Tevens zou het opvullen van de holte in het kleine bekken na resectie wondgenezing kunnen verbeteren. Een nadeel van een omentumplastiek is de verlengde operatietijd, vooral bij een laparoscopische procedure. Momenteel zijn er geen gerandomiseerde studies uitgevoerd om het eventuele efect van een omentumplastiek te bewijzen.2 Een andere mogelijke techniek om de perineale wond te sluiten is door gebruik te maken van een musculocutane of fasciocutane fap. In de beperkte literatuur die beschikbaar is, lijkt deze fap het aantal perineale wondinfecties te verminderen.3 Echter vereist het gebruik van deze fap een langere operatietijd en een gespecialiseerde chirurg. Tevens kan een dergelijke fap wondproblemen geven op de plaats waar hij is uitgenomen en kan er necrose van de fap ontstaan. Momenteel wordt er een gerandomiseerde studie verricht waarbij patiënten met een distaal rectumcarcinoom worden gerandomiseerd tussen het plaatsen van een musculocutane fap of een biologische mat na eAPR (ClinicalTrials.gov Identifer: NCT01347697).

198 Samenvatting en toekomstige perspectieven

Bij patiënten met een perineale fstel na chirurgie van een vergevorderd rectumcarcinoom is het aantal wetenschappelijke studies waarin een oplossing voor perineale wondproblemen wordt beschreven zeer beperkt. Zo is de behandeling van een perineale fstel na uitgebreide bekkenchirurgie erg complex en is vaak een combinatie van verschillende technieken nodig. In hoofdstuk 8 beschrijven we een mogelijke oplossing voor perineale fstels in deze bijzondere patiëntengroep. In dit onderzoek zijn een beperkt aantal patiënten behandeld met een biologische mat en daar bovenop een gluteus plastiek, hetgeen in deze patiëntengroep geleid heef tot een ongecompliceerd postoperatief beloop. De combinatie van deze technieken biedt een handvat in de behandeling van deze moeilijke infectieuze complicaties.

Voor naadlekkage in het kleine bekken bestaan verschillende behandelingsopties met allen een beperkte succeskans. In tegenstelling tot de standaard chirurgische principes, was de hypothese dat een naadlekkage transanaal gesloten kon worden nadat de presacrale holte was schoongemaakt met Endo-sponge®. In hoofdstuk 9 wordt deze techniek beschreven in een groep patiënten met colitis ulcerosa en familiaire adenomateuze polyposis syndroom, welke na een ileal pouch-anale anastomose een naadlekkage hebben gekregen. In deze groep patiënten was het anastomose defect genezen bij alle patiënten na Endo-sponge® behandeling en transanaal chirurgisch sluiten, in vergelijking met een slagingspercentage van 52% bij de conventionele behandeling. Waarbij de direct medische kosten tussen de twee behandelingen niet verschilde. Ondanks deze veelbelovende resultaten, is er meer onderzoek nodig om de efectiviteit van deze behandeling ook te bewijzen in rectumcarcinoom patiënten die na neoadjuvante (chemo)radiotherapie een laag anterieure resectie hebben ondergaan. Momenteel loopt er een prospectief cohort onderzoek waarin 30 patiënten met een rectumcarcinoom zijn geïncludeerd. De resultaten van deze studie zullen in 2017 beschikbaar komen (NTR4671; trialregister.nl).

Een deviërend stoma wordt beschouwd als een routinebehandeling bij een kleine bekken anastomose. Een aantal studies hebben laten zien dat een deviërend stoma de symptomen van een naadlekkage vermindert.4 Echter is de morbiditeit en mortaliteit gerelateerd aan een deviërend stoma en het ophefen ervan na een laag anterieure resectie (LAR) onderbelicht in de klinische praktijk. Na de publicatie van eerder onderzoek uit onze klinische praktijk, zijn er diverse klinische veranderingen doorgevoerd om de uitkomst te verbeteren. In hoofdstuk 10 wordt beschreven dat supervisie door een gespecialiseerd colorectaal chirurg is toegenomen, de anastomose techniek is aangepast en de alertheid op postoperatieve complicaties zijn verbeterd. Deze veranderingen hebben ervoor gezorgd dat de stoma gerelateerde morbiditeit signifcant is verlaagd van 11% naar 4%. Daarbij is gebleken dat bij het ophefen van een ileostoma, supervisie of uitvoering van de operatie door een colorectaal chirurg een onafankelijke voorspeller is voor het verlagen van de morbiditeit. Daarnaast blijkt dat het ophefen van een eindstandig ileostoma ten opzichte van een dubbelloops ileostoma een onafankelijke voorspeller is voor een toename in de morbiditeit. Deze bevindingen suggereren dat het ophefen van een ileostoma niet meer gezien kan worden als een ingreep voor de chirurg in opleiding, maar uitgevoerd zal moeten worden onder directe supervisie van een colorectaal chirurg.

199 Samenvatting en toekomstige perspectieven

Als de behandeling van een naadlekkage faalt dan kan het defect van de anastomose en het hieruit volgende presacrale abces voor langer dan een jaar blijven bestaan. Een zogenoemde presacrale sinus heef een beperkt aantal behandelingsmogelijkheden en de uitkomst van dergelijke behandelingen worden nauwelijks beschreven in de huidige literatuur. Een eerdere publicatie stelt een potentieel efectieve chirurgische behandeling voor in de vorm van een abdominoperineale proctectomie en een eindstandig colostoma met een omentumplastiek in het kleine bekken. In hoofdstuk 11 wordt de uitkomst beschreven van deze mogelijke behandelingsoptie. Deze techniek heef geleid tot een genezingspercentage van 78%, terwijl eerdere behandeling voor de naadlekkage in deze patiënten faalde. Ondanks deze veelbelovende resultaten, zal nog meer onderzoek gedaan moeten worden om de efectiviteit vast te stellen.

200 Samenvatting en toekomstige perspectieven

REFERENTIES

1. Martijnse IS, Holman F, Nieuwenhuijzen GA, Rutten HJ, Nienhuijs SW. Perineal hernia repair afer abdominoperineal rectal excision. Diseases of the colon and rectum 2012; 55(1): 90-5.

2. Killeen S, Devaney A, Mannion M, Martin ST, Winter DC. Omental pedicle faps following proctectomy: a systematic review. Colorectal disease 2013; 15(11): e634-45.

3. Foster JD, Pathak S, Smart NJ, et al. Reconstruction of the perineum following extralevator abdominoperineal excision for carcinoma of the lower rectum: a systematic review. Colorectal disease 2012; 14(9): 1052-9.

4. Wu SW, Ma CC, Yang Y. Role of protective stoma in low anterior resection for rectal cancer: a meta- analysis. World J Gastroenterol 2014; 20(47):18031-7.

201

Ph.D. portfolio

Samenvatting en toekomstige perspectieven

204 Ph.D. portfolio

PH.D. PORTFOLIO

Name Ph.D. student: G.D. Musters Ph.D. period: 2012-2015 Name Ph.D. supervisor: Prof. dr. W.A. Bemelman 1. Ph.D. training Year Workload General courses - Clinical epidemiology 2013 0.9 - Developing a systematic review 2013 0.7 - Practical biostatistics 2013 1.1 - Basiscursus regelgeving klinisch onderzoek (BROK) 2012 0.9 Specifc courses - Advanced topics in Biostatistics 2014 2.1 Seminars, workshops and master classes - Weekly department research seminars 2012-2015 2 - Master class prof. dr. J. Powell 2012 0.1 - ICH and EU directives on good clinical practise 2012 0.2 - Complex abdominal wall repair and perineal foor reconstruction, 2014 1.8 Barcelona, Spain Poster presentations - Perineal wound problems afer abdominoperineal resection for 2014 0.5 rectal cancer; a two-institutional experience in the era of intensifed oncological treatment. UEGW - Early reconstruction of anastomotic leakage: a novel solution to an 2013 0.5 old problem. Digestive disease week Oral presentations - De Trials van de 3D groep. Chirurgendagen 2013 0.5 - Vroege reconstructie van naaddehiscentie met behulp van 2013 0.5 vacuümtherapie met endospons na ileo-pouch anale anastomose. Chirurgendagen (Inter)national conferences - United European Gastroenterology, Wenen 2014 0.25 - NVvH chirurgendagen, Veldhoven 2013 0.25 - European society of coloproctology, Belgrado 2013 0.25 - United European Gastroenterology, Amsterdam 2013 0.25 - NVvH chirurgendagen, Veldhoven 2012 0.25 Teaching, tutoring and mentoring - Tutoring 2nd year medical students 2014 1 - Tutoring 3rd year medical students 2013 1

205

List of publications

Samenvatting en toekomstige perspectieven

208 List of publications

LIST OF PUBLICATIONS

2016 Musters GD, Atema JJ, van Westreen HL, Buskens CJ, Bemelman WA, Tanis PJ. Major morbidity afer ileostomy closure: results from an institutional change in practice and awareness. Int J Colorectal Dis 2016.

2015 Musters GD, Buskens CJ, Bemelman WA, Tanis PJ. Local application of gentamicin in the prophylaxis of perineal wound infection afer abdominoperineal resection, a systematic review. World J Surg 2015; 39(11): 2786-94.

2015 Vennix S, Musters GD, Mulders IM, Swank HA, Belgers EJ, Consten EC, van Geloven AAW, Gerhards MF, Govaert MJ, Grevenstein WM, Hoofdijk AG, Kruyt PM, Nienhuijs SW, Boermeester MA, van Dieren S, Lange JF, Bemelman WA. Laparoscopic peritoneal lavage or sigmoidectomy for perforated diverticulitis with generalised purulent peritonitis: a multicentre randomized trial (Te LOLA arm of the Ladies Trial). Lancet 2015;386(10000): 1269-77.

2015 Musters GD, Borstlap WA, Bemelman WA, Buskens CJ, Tanis PJ. Intersphincteric completion proctectomy with omentoplasty for chronic presacral sinus afer low anterior resection for rectal cancer. Colorectal Dis 2015.

2015 Westerterp M, Musters GD, van Geloven AAW, Rutten HJT, Tanis PJ. Huidige praktijk van de abdominoperineale resectie; een survey onder Nederlandse chirurgen. NTvH 2015; 24 (1): 37-40.

2015 Borstlap WA, Stellingwerf ME, Molla Z, Musters GD, Buskens CJ, Tanis PJ, Bemelman WA. Preoperative iron therapy as treatment of anaemia in patients with colorectal carcinoma; a systematic review. Colorectal Dis 2015;17(12):1044-54.

2015 Klaver CEL, Musters GD, Bemelman WA, Punt CJA, Verwaal VJ, Dijkgraaf MGW, Allbers AGJ, van der Bilt JDW, Boerma D, Bremers AJA, Burger JWA, Buskens CJ, Evers P, van Ginkel RJ, Grevenstein WMU, Hemmers PHJ, de High IHJT, Lammers LA, van Leeuwen BL, Meijerink WJHJ, Nienhuijs SW, Pon J, Radema SA, van Ramshorst B, Snacebjornsson, Tuynman JB, te Velde EA, Wiezer MJ, de Wilt JHW, Tanis PJ. Adjuvant hyperthermic intraperitoneal chemotherapy (HIPEC) in patients with colon cancer at high risk of peritoneal carcinomatosis; the COLOPEC randomized multicenter trial. BMC cancer 2015;15(1):428.

2015 Gardenbroek TJ, Musters GD, Buskens CJ, Ponsioen CY, D’Haens GR, Dijkgraaf MG, Tanis PJ, Bemelman WA. Early reconstruction of the leaking ileal pouch-anal anastomosis: a novel solution to an old problem. Colorectal Dis 2015; 17(5): 426- 432.

2014 Barendse RM, Musters GD, Fockens P, Bemelman WA, de Graaf EJ, van den Broek FJ, van der Linde K, Schwartz M, Houben M, van Milligen de Wit A, Witteman B, Winograd R, Dekker E; TREND study group. Endoscopic mucosal resection of large rectal adenomas in the era of centralization: Results of a multicenter collaboration. United European Gastroenterol J 2014; 2(6): 497-504.

209 List of publications

2014 Musters GD, Sloothaak DAM, Roodbeen S, van Geloven AAW, Bemelman WA, Tanis PJ. Perineal wound problems afer abdominoperineal resection for rectal cancer; a two- institutional experience in the era of intensifed oncological treatment. Int J Colorectal Dis 2014; 29(9): 1151-1157.

2014 Musters GD, Buskens CJ, Bemelman WA, Tanis PJ. Perineal wound healing afer abdominoperineal resection for rectal cancer; a systematic review and meta-analysis. Dis Colon Rectum 2014; 57(9): 1129-1139.

2014 Musters GD, Bemelman WA, Bosker RJ, Burger JW, van Duijvendijk P, van Etten B, van Geloven AA, de Graaf EJ, Hof C, de Korte N, Leijtens JW, Rutten HJ, Singh B, van de Ven A, Vuylsteke RJ, de Wilt JH, Dijkgraaf MG, Tanis PJ. Biological mesh closure of the pelvic foor afer extralevator abdominoperineal resection for rectal cancer (BIOPEX- study). BMC Surg 2014; 27: 58.

2014 Musters GD, Lapid O, Bemelman WA, Tanis PJ. Surgery for complex perineal fstula following rectal cancer treatment using biological mesh combined with gluteal perforator fap. Tech Coloproctol 2014; 18(10): 955-9.

2014 Musters GD, Stassen LPS, Havenga K, Bemelman WA. Vroeg herstel van naaddehiscentie na low anterior resectie, de CLEAN-studie. NTvH 2014; 23: 44-46.

2013 Musters GD, Rutten HJT, Bemelman WA, Tanis PJ. Kostenefectiviteit bekkenbodem- reconstructie met biologische mat, de BIOPEX study. Ned Tijdschr Geneeskd 2013; 157: A5999.

2013 Musters GD, Boele van Hensbroek van P, Ponsen KJ, Luitse JSK, Goslings JC. Locking compression Plates are more difcult to remove than conventional non- locking plates. Eur J Trauma Emerg surg 2013;39: 159-162.

2011 Musters GD, Scherpenzeel KM, Strackee SD, Goslings JC. Te Natural history of a missed scaphoid fracture. J Med Cases 2011; 2: 246-247.

2011 Musters GD, Kleipool RP, Bipat S, Maas M. Features of the popliteal lymph nodes seen on musculoskeletal MRI in a Western population. Skeletal Radiol 2011; 40(8): 1041-1045.

2009 Verhoef JJ, Stalpers LJ, Claes A, Hovinga KE, Musters GD, Van der top WP, Richel DJ, Leenders WP, van Furth WR. Tumour control by whole brain irradiation of anti-VEGF- treated mice bearing intracerebral glioma. European journal of cancer 2009; 45: 3074-3080.

210 List of publications

211

Dankwoord

Samenvatting en toekomstige perspectieven

214 Dankwoord

DANKWOORD

Met ontzettend veel trots en met heel veel plezier zal ik altijd blijven terugdenken aan mijn onderzoeksperiode. Dit heb ik te danken aan alle mensen die hebben bijgedragen aan mijn proefschrif. Hierbij wil ik deze mensen dan ook ontzettend bedanken voor hun bijdrage, waarbij ik een aantal personen in het bijzonder wil noemen.

Allereerst mijn promotor prof. dr. W.A. Bemelman, beste Willem, heel erg bedankt voor het vertrouwen en de mogelijkheden van de afgelopen jaren. Tevens heel erg veel dank voor de congressen en alle gezellige etentjes aan de Vinkeveens plassen. Het was een zeer leerzame periode. Je bent een geweldige onderzoeker en een fantastische begeleider. Ik heb veel bewondering en ontzag voor de manier waarop jij klinische zorg levert. Ik kijk er dan ook naar uit om in de toekomst nog veel van je te leren.

Dr. P.J. Tanis, beste Pieter, je bent een geweldige copromotor! Ik had mij geen betere begeleiding kunnen wensen. Je bodemloze kennis, je wetenschappelijk inzicht en je chirurgische vaardigheden maken je een inspirerende chirurg. Ik heb met heel veel plezier met je samengewerkt en zeer veel van je geleerd, hetgeen ik hopelijk in de toekomst kan blijven doen.

Prof. dr. E. Dekker, beste Evelien, heel erg bedankt voor je enthousiaste begeleiding. Ik heb veel van je geleerd en ben trots dat je mijn copromotor hebt willen zijn.

Dr. C.J. Buskens, beste Christianne, heel erg bedankt voor je enthousiasme en je soms andere kijk op het onderzoek. Je aanvullingen op mijn manuscripten hebben mij erg geholpen.

Geachte leden van de leescommissie, heel erg bedankt voor uw bereidheid om zitting te nemen in mijn promotiecommissie.

Beste mede-onderzoekers van de BIOPEX-studie, LADIES-trial, TREND-studie en CLEAN-studie, ontzettend bedankt voor al jullie inzet en hulp. Zonder jullie waren deze fantastische studies niet mogelijk. Beste Christine, heel erg bedankt voor al je steun en hulp bij de TREND-studie.

Beste onderzoekers van G4, bedankt voor de mooie tijd en voor de vrijdagmiddag lunches en ontelbare borrels.

Beste Bemelman onderzoekers, beste Tjibbe Gardenbroek, Didi Sloothaak, Joline de Groof, Salomeh Sahami, Sandra Vennix, Hilco Swank, Irene Mulder, Emma Bruns, Daniël Lambrichts, Renée Barendse, Charles van Rossem, Matthijs Bolmers, Lotje Klaver en Wernard Borstlap, heel erg bedankt voor jullie steun en geweldige tijd op G4.

215 Dankwoord

Beste mannengang, beste Robert-Jan de Muinck keizer, Jasper Atema, Ben Bennedetto en Sytse van Beek, bedankt voor de mooie tijd.

Beste Coos, Els, Jacqueline, Ingrid, Joke, Indra, en Fenne bedankt voor al jullie hulp en inzet, zonder jullie is onderzoek doen op G4 niet mogelijk.

Beste collega’s uit Tergooi ziekenhuizen, bedankt voor de fantastische assistententijd. Dr. A.A.W van Geloven, beste Nanette, ontzettend bedankt voor al je steun, samenwerking en hulp.

Mijn beste paranimfen, Jim Boersma en Tijs Siegenbeek van Heukelom, bedankt dat jullie mij begeleiden in de afsluiting van mijn proefschrif. Het is meer dan logisch dat jullie naast mij staan.

Al mijn vrienden, bedankt voor jullie steun tijdens mijn onderzoeksperiode.

Beste Boudewijn, Evelien, Max en Luuk Musters, bedankt voor jullie vertrouwen en steun.

Lieve ouders, woorden schieten te kort om te beschrijven hoeveel jullie voor mij hebben betekend. Bedankt voor jullie grenzeloze steun, begrip en liefde.

Lieve Judith, bedankt voor al je steun en begrip van de afgelopen jaren. Het promotieonderzoek is af en ik kijk uit naar wat de toekomst ons zal bieden.

216 Dankwoord

217 Samenvatting en toekomstige perspectieven

218 About the author

ABOUT THE AUTHOR

Gijsbert Musters was born on the 2nd of April 1986 in Huizen the Netherlands. In 2006 he graduated from high school and was admitted to the medical school at the University of Amsterdam. Afer graduation, he started with his research at the department of surgery at the Academic Medical Centre in Amsterdam. His thesis was under the supervision of his promotor prof. dr. W.A. Bemelman, and co-promotors dr. P.J. Tanis and prof. dr. E. Dekker. During his Ph.D. program, he also started working as a non-training resident in surgery at the Tergooi ziekenhuizen in Hilversum and Blaricum under the guidance of dr. J.P. Eerenberg. Afer ten months of clinical work, Gijsbert went back to the department of surgery at the Academic Medical Centre to complete his Ph.D. thesis. Currently, he is a surgical trainee at the Academic Medical Centre and will continue his training until he certifes as a surgeon.

219