054-2018.R1Systematic Review

Systematic review of the literature regarding the

Mmanagement of early pouch-related septic complications in ulcerative colitis: systematic

review Guy H T Worley, Jonathan P Segal, Janindra Warusavitarne, Susan K Clark, Omar D Faiz

St Mark's Hospital and Academic Institute, Harrow, Middlesex, UK and Department of Surgery and Cancer, Imperial College London, London, UK

Corresponding author: Guy Worley guy.worley@nhs.net

Guarantor of the article: Omar Faiz omar.faiz@nhs.net

Keywords: ulcerative colitis, restorative proctocolectomy, ileal pouch, surgery,

complications, sepsis, leak

Abstract word count: 21620

Word count: 2,9943

Tables: 3

Figures: 2

Appendices: 1

References: 26

Date: 927th February May 2018

AbstractIntroductionIt is well establishedreported that ileoanal pouch-related septic complications (PRSC)

increase the risk of pouch failure. There are a number of publications that describe the

management of early PRSC in ulcerative colitis (UC) in small series. This article aims to

systematically review and summarise the relevant contemporary data on this subject and

provide an algorithm for the management of early PRSC.

MethodA systematic review was undertaken in accordance with PRISMA guidelines. Studies

published between 2000 and 2017 describing the clinical management of PRSC in patients

with UC within 30 days of primary ileoanal pouch surgery were included. A qualitative

analysis was undertaken due to the heterogeneity and quality of studies included.

Results1157 abstracts and 266 full text articles were screened. Twelve studies were included for

analysis involving a total of 207 patients. The studies described a range of techniques

including image-guided, endoscopic, surgical and endocavitational vacuum methods. Based

on the evidence from these studies, an algorithm was created to guide the management of

early PRSC.

ConclusionThe results of this review case series included in this review suggest that although successful

salvage of early pouch related septic complications is improving over time there is little in-

formation available relating to methods of salvage and outcomes. a paucity of research cor-

relating the method used with functional outcome. Novel techniques may offer increased

chance of salvage but comparative studies with longer follow-up are required.

2

What does this paper add to the literature?This is a systematic review of the published literature regarding the management of short

term pouch-related septic complications for patients with ulcerative colitis. The appropriate

management of post-operative sepsis is crucial to give the best chance of avoiding pouch

failure. We present an algorithm for management based on this review.

IntroductionSince the first described publication in 1978, restorative proctocolectomy or the ileoanal

pouch procedure has been frequently performed as an alternative to total colectomy and

come a popular option for those patients requiring colectomy forpermanent ileostomy in

patients with ulcerative colitis (UC). , avoiding the need for permanent ileostomy (1). The

incidencerate of pelvic sepsis after ileoanal pouch formation approachesis up to 15%, and

septic complications account for the majority of pouch failures. Many papers describe the

rate and sequelae of early pouch related septic complications (PRSC) but very few empirical

studies describe their clinical management in detail (2–6).

This systematic review of the literature aims to identify and summarise empirical studies

describing the management of early PRSC.

MethodThis systematic review was undertaken in adherence with the Preferred Reporting In

Systematic Reviews and Meta-Analyses (PRISMA) guidelines and Cochrane Handbook for

Systematic Reviews of Interventions(7).

Information sourcesMEDLINE (1946 to present), Embase (1974 to present) (both searched via Ovid), and Web of

Science (1950 to present) were searched systematically as per the terminology in Appendix

1. The Cochrane Database of Systematic Reviews was searched by 'Topic' and articles were

hand searched from the references and citing articles of papers which met the inclusion

criteria. The last literature search was 21st November 2017.

3

SearchAll databases were searched with the intention of high sensitivity and wide capture. The

following terms were included as 'exploded' Medical Subject Headings (MeSH) or Keywords

terms depending on database, and also as text word searches, using the suffix '.tw.'. These

terms were then combined with the Boolean operators 'AND' or 'OR' as follows, and the

search filtered by year (2000 to 2017 to focus on contemporary studies), by publication type

(Journal Article) and to remove duplicates. No language filter was employed during the

initial literature search but only abstracts in English were screened.

Inclusion criteriaAll articles from peer-reviewed journals which described empirical studies on patients with

UC after primary pouch surgery, with an abstract written in English, were included. Studies

were required to describe techniques in managing pouch-related septic complications (i.e.

anastomotic leak, abscess, sinus, fistula and peritonitis) within 30 days of the primary

procedure.

Exclusion criteriaReview articles, published abstracts, conference posters, audits, editorials and letters were

excluded. Two researchers (GW and JPS) independently reviewed the articles retrieved by

the search strategy and resolved disagreements by consensus.

Data collection and abstractionData were extracted and tabulated in Microsoft Excel (Version 15.18). Data collected

included: first author, institution city, year of publication, sample size, study design, gender,

age, BMI, diagnosis, pouch configuration, anastomotic technique, operative stages, de-

functioning ileostomy at pouch creation, categorisation of leak, diagnostic criteria for PRSC,

signs and symptoms of PRSC, diagnostic imaging, diagnostic endoscopy, management

techniques, time to diagnosis and treatment, duration of treatment, time to closure of

defect, time to discharge, length of follow up and functional outcomes.

Synthesis of resultsA descriptive synthesis was employed. Meta-analysis was not appropriate due to the extent

4

of heterogeneity within the study designs and reported outcome variables.

ResultsStudy selectionA total of 1155 results were obtained from the original database search, with duplicates

removed as one of the search filters. Two results were obtained from hand searching. Of

these 1157 screened abstracts 798 were excluded, leaving 339 for full text review. Further

duplicates were removed from this selection in citation software (Mendeley v 1.14. 2008-

2015) leaving 266 full text articles to be assessed for eligibility. After review 254 articles

were excluded, leaving 12 for qualitative analysis (see PRISMA flow diagram, figure 1). When

the studies are referred to in the results they are referenced in square brackets according to

the numbers allocated in table 1, i.e. [2,5]. Other references are included in parentheses.

Study characteristicsThe 12 included papers were published between 2007 and 2015. Six papers were from

Europe and six from North America. Four papers were case reports of novel techniques

[2,3,4,10], five were observational studies [5,8,9,11,12] and one was an interventional study

[6]. Participants ranged from 1 to 141, with a total of 207 adult participants included across

all studies (table 1).

Synthesis of evidenceDiagnosis and InvestigationThe signs and symptoms associated with acute pouch sepsis were described in five papers

[3-5,8,12]. Raval et al. stated that 130 of their series of 141 had symptomatic leaks, 67%

exhibiting fever, 38% with abdominoperineal pain and 6% perineal abscess. Of 94 patients

without a primary de-functioning ileostomy, high frequency of defaection was present in

29%. Kirat et al. reported fever in only 19% and abdominal pain in 56% of their 27 patients.

One article [4] described the use of digital examination of the anal anastomosis in the

diagnosis of pouch leaks.

Mennigen et al. reported that three of 12 patients had very mild or absent symptoms,

delaying diagnosis until 20, 26 and 50 days following surgery [11]. Raval et al reported 11 of

5

141 patients had asymptomatic leaks diagnosed on ‘pouchogram’ before ileostomy closure.

Sagap et al. categorised symptoms as mild (clinical suspicion without prominent symptoms),

moderate (prominent clinical and/or biochemical features) and severe (generalised

peritonitis or shock requiring rapid surgery or ITU care).

Computed tomography (CT) was the most common diagnostic imaging modality. Five

studies [2,3,5,9,11] described using CT diagnosis and two [2,11] augmented CTs with

contrast enemas. Three studies described routine use of water-soluble contrast enema

radiography or 'pouchogram' to diagnose leaks [5,10,12]. Pouchoscopy was used to

diagnose or confirm a leak in six studies [2,5,6,8,9,11]. Mennigen et al. reported routine

pouchoscopy before discharge.

Time from pouch formation to diagnosis of leak was presented in seven studies [2,4,9-12].

Three case studies [2,4,10] quoted 14, 10 and 5 days respectively, and the case series papers

[1,9,11,12] presented medians of 25, 19, 14 and mean 11.5 days respectively. The mean

time to diagnosis from these six papers was 18 days.

Categorisation/site of leak Only Raval et al. presented a categorisation of leaks: either ileo-anal anastomosis (IAA) with

(21%) or without (33%) abscess; pouch-cutaneous fistula (10%); pouch-vaginal fistula (12%);

radiologic leak (8%) or pouch body leak (16%). Kirat et al reported larger abscesses in their

group undergoing CT-guided percutaneous drainage (7.7cm) than those undergoing trans-

anastomotic drainage (5.1cm). In Mennigen et al's series of 12, five had IAA dehiscence with

associated abscess, three had pouch body leaks, two had fistulas and three had abscesses

without evidence of leak. Kirat et al's series only described leaks from the 'tip' (blind end) of

the J pouch, whereas the cumulative 22 patients from the Dutch trials [3,8,11] only dealt

with low leaks, i.e. the IAA or just above.

De-functioning ileostomy and pouch catheter useThe rate of primary de-functioning ileostomy varied between 26% and 100%. All of Kirat et

al's 27 patients with leak from the blind end of the pouch were primarily de-functioned,

compared to four of 15 in Gardenbroek et al's series of anastomotic leaks. Not all patients

6

without a primary ileostomy were subsequently de-functioned: Raval et al. reported 63% of

141 patients were de-functioned with an ileostomy, but 42 of the remaining 47 leaks were

managed with a pouch catheter and drainage. Five required an ileostomy and washout

within 24 hours for generalised peritonitis. Gardenbroek et al. managed two patients with

short-term Endo-SPONGE® treatment without de-functioning at all. In one of these patients

the initial closure was successful, and the second required formation of an ileostomy and a

second period of Endo-SPONGE® treatment before successful closure of the defect.

Mennigen et al. managed one patient without ileostomy when there was an abscess but no

leak present at abdominal washout.

Percutaneous radiological drainagePercutaneous CT-guided drainage of peri-pouch abscess was reported in five papers.

Thosani et al. described the use of CT-guided percutaneous drainage after primary closure

of the anastomotic defect with an over the scope clip (OTSC). Kirat et al reported that two of

18 patients developed pouch-cutaneous fistula along the trans-gluteal CT drainage track.

Both healed with conservative management and drainage of the associated sepsis

Trans-anal drainageTrans-anastomotic drainage of peri-pouch abscesses was reported in three papers. Raval et

al. stated the indication for trans-anastomotic drainage was anastomotic dehiscence. Only

Varadarajulu et al. described peri-pouch abscess drainage endoluminally without a pre-

existing anastomotic defect by endoscopic ultrasound guided drainage [9]. Kirat et al

compared the non-randomised use of percutaneous vs trans-anastomotic drainage of

abscesses and found no difference in outcome.

Endo-cavitational vacuum therapyFive papers reported the use of vacuum therapy. All specified that this is only appropriate

for low pelvic leaks due to technical limitations and to avoid exerting suction on abdominal

small bowel. Two approaches were described – long and short term. The long-term method

involved Endo-SPONGE® change every three to four days, reducing sponge size until a very

small cavity remained, often with a persisting sinus. The median time from diagnosis to

7

closure was 70 days. Gardenbroek et al. and Verlaan et al's short-term technique (three to

six sponge changes to clean and encourage healing before early surgical closure of the

defect) resulted in a time of 48 (from diagnosis) and 14 days (from start of treatment) to

resolution [8,11].

Abdominal re-operationFive studies described re-operation within 30 days [4,5,11,12], excluding surgery only to

form a de-functioning ileostomy. The indication for abdominal salvage was peritonitis,

abscess not amenable to radiological drainage or high leak [2,7]. Maruthachalam et al.

described an initial ileostomy formation and pelvic drainage which failed, so the IAA was

disconnected and pouch brought out as a mucus fistula [4]. The pouch was re-connected 12

months later and the technique resulted in good anorectal physiology at 18 months follow-

up. Kirat et al. reported the rate of intervention required after failed local drainage. Trans-

anal drainage failed in 13 patients (24.5%) of whom three had a re-do pouch, one had their

pouch mobilised and an ileostomy formed, and six had their pouches excised. Of three failed

CT-guided drainages (17%), two had a subsequent re-do pouch performed.

Sagap et al. reported that abdominal re-operation was significantly associated with failure

on univariate logistic regression, demonstrating 41% failure rate. Fifty-five of 157 patients

required a laparotomy, either after failed drainage or as a primary procedure in 22. None of

the studies described pouch excision within 30 days of the primary procedure.

Quality of life and functional outcomeNine studies stated their length of follow-up (table 2). Three papers reported functional

outcomes [4,5,11]. Mennigen et al's paper reported no statistically significant difference in

Öresland, SIBDQ and GIQLI functional scores. The median frequency in the PRSC group was

seven, compared to six for controls. Kirat et al's article also reported no significant

difference in frequency, urgency, incontinence, seepage or pad use; neither was there

significance difference between Cleveland Clinic scores, quality of health or sexual function.

Maruthachalam et al's case report on pouch salvage by mucus fistula resulted in good

capacity and compliance.

8

Pouch failureSix papers presented failure rate [1,2,5-7,11]. Sagap et al. reported 75.8% of pouches were

salvaged. On multivariate analysis – fistula, trans-anal drainage, delayed ileostomy closure,

new ileostomy diversion and hypertension were associated with failure. Based on these

factors they developed a predictive score.

Raval et al. reported 84% with a functioning pouch from their sample, using multiple

treatment techniques. Stapled anastomoses fared better than hand-sewn (93% vs 72%

salvaged). There was no statistically significant difference in outcome between patients who

had no ileostomy when PRSC was diagnosed, or those who had PRSC diagnosed before or

after ileostomy closure (83%, 85%, 85% respectively). They reported improved rates of

salvage from 67% between 1981 to 1984 to 88% between 2001 to 2003.

Of Kirat et al’s series of 27 patients with leaks from the blind end of the J pouch, one patient

had excision and ileostomy, one patient was still awaiting stoma closure and the remaining

25 had functioning pouches at 3.2 year mean follow-up.

Mennigen et al. reported four pouch failures out of 12 patients with PRSC. Each had at least

three procedures before failure and three had abdominal salvage attempts. Kirat et al.

reported 75.5% and 83% success for trans-anal and percutaneous CT-guided drainage

respectively, with a mean follow-up of 3.1 years. In Gardenbroek et al’s comparison of

short-term Endo-SPONGE® treatment 14 of 15 patients had a functional pouch compared

with 24 of 28 with the long-term technique. The median follow-up in their intervention

group was significantly shorter at 25 compared to 104 months.

DiscussionIt is well established that PRSC increases the risk of pouch dysfunction and failure (2–6).

However, these complications are not homogenous and neither is the clinical management.

Only one study reported included here described presented a categorisation of the nature

of pouch leaks. Such a classification might ; a widely adopted categorisation of leaks may

significantly improve the quality of reporting and homogeneity in the literature for pouch

surgery. The International Study Group of Rectal Cancer categorisation of leaks is

inadequate for ileoanal pouches as it does not account for the anatomic site, which guides

the approach to treatment, albeit the vast majority are from the pouch-anal anastomosis

9

(8).

The definition of early sepsis is either Whether defining early sepsis as within 30 days, or

until planned closure of ileostomy, (i.e. 3 months),. bBoth definitions refer to sepsis that is a

direct result of the operative procedure. This is in comparison, compared towith a second

peak of PRSC in patients who develop de novo sepsis several years after pouch formation.

Some of these cases are attributed to Crohn’s disease, but recent literature suggests that

this diagnosis is over used (9).

Many factors influence the As described by Raval et al, there may be varying degrees of the

magnitude of the pouch septic insult. These include the , depending on time to time to

diagnosis of the leak, leak diagnosis,the presence or otherwise of a defunctioning

ileostomy, anatomical site and size. A leak from the tip of the J is generally a very different

clinical scenario to a leak from the IAA. Further studies are required to follow up cases of

PRSC with prospective data on different sepsis characteristics and management to

accurately assess the correlation with outcome. The rate of successful salvage has improved

over time, and now ranges between 75 and 85% (10). The need for laparotomy or trans-anal

drainage, however, was consistently associated with failure in this review.

The clinical burden of insidious leaks is unknown. not yet known. Patients who are

asymptomatic until reversal of stoma ileostomy may well have had ongoing low-grade

sepsis. Roughly 8% of Raval et al’s series were asymptomatic, with an abnormal

pouchogram before reversal of ileostomy. It is not known how many of their patients

developed sepsis after ileostomy closure with a normal pouchogram, but it has been

demonstrated in other studies that pouchogram has poor sensitivity for the detection of

anastomotic leaks (11,12). Only Mennigen et al’s paper describes the routine use of

endoscopy to review the integrity of the pouch before discharge. Only one study describes

the use of routine imaging in the early post-operative period before discharge, as opposed

to before ileostomy closure. Chronic low grade sepsis can be asymptomatic or misdiagnosed

as chronic pouchitis (13), and it may be appropriate to incorporate early prophylactic pouch

imaging to detect silent leaks, with the intention of acute anastomotic repair or vacuum

therapy to prevent the influence of insidious septic exposure on the pouch. There is very

little mention of the routine examination of pouches in the post-operative period, and the

10

sensitivity of digital examination to detect ileo-anal anastomotic defects in pouches is not

known.

Managing PRSC without a diverting ileostomy continues to be a contentious. issue. It is

becoming more popular to undertake pouch surgery without a de-functioning ileostomy

(14) and several of the studies included in this review even describe the successful

management of PRSC without a de-functioning stoma [2, 6, 10]. As with selecting patients

for primary pouch formation without ileostomy, the management of leaks without a de-

functioning ileostomy must be assessed on a case by case basis, taking in to account patient

factors (no steroid use, good nutritional status, non-smoker) and operative factors (tension

free anastomosis, good perfusion etc). One of the arguments in favour of forming an ileal

pouch without de-functioning is that a leak is likely to be more clinically apparentobvious,

and thereby manifested earlier. The current progression of techniques towards trans-anal

dissection, single circular stapled anastomosis and the use of Endo-SPONGE® for leak

control go hand in hand with the ‘modern two-stage’ approach to pouch formation; i.e.

colectomy with ileostomy followed by proctectomy and pouch formation without ileostomy.

The novel use of the Endo-SPONGE® device to clean the cavity and promote healing before

closure in Dutch studies shows promising results in terms of early healing, but as yet there

are no published follow up data describing functional outcomes. Further studies are

required to investigate the subsequent functional outcome depending on the time taken to

resolve sepsis and the technique used. As stated previously, a published categorisation of

pouch leaks would add homogeneity to future research.

Based on the evidence collated here and the experience of our own institution we present

an algorithm for the management of early PRSC (figure 2). Trans-gluteal image guided

drainage should be avoided if possible because of potential fistulation along the drain tract.

For a pelvic abscess cavity without a visible staple line defect or communicating sinus,

controversy exists as to the best management. If transcutaneous drainage is performed in

the context of a communication with the pouch this may result in an extrasphincteric fistula

and should be avoided. If a small communication is found, it is possible to slightly dilate it

and introduce an Endo-SPONGE®. Algorithms have been published previously following

review of the techniques for all pouch complications, but they either do not focus

specifically on early PRSC or do not include the use of Endo-SPONGE® therapy (10,15,16).

11

The limitations of this review lie with the quality of the studies reported. The relative lack of

published data available in this area should prompt future research aiming to establish the

most effective management of early PRSC by correlating interventions with outcome

measures.

ConclusionThe results of this review case series included in this review suggest that although successful

salvage of early pouch related septic complications is improving. However, over time there

is a paucity of research investigating salvage techniques and outcomes. correlating the

method used with functional outcome. Novel techniques may offer an increased chance of

salvage but comparative studies with longer follow-up are required.

Author ContributionsGW conceived the study, performed the literature search, extracted and analysed the data,

drafted the manuscript and approved the final version for publication. JS performed the

literature search, extracted and analysed the data, critically appraised the manuscript and

approved the final version for publication. JW and SKC critically appraised the manuscript

and approved the final version for publication. ODF conceived the study, critically appraised

the manuscript and approved the final version for publication.

Conflicts of InterestNone

12

13

References1. Parks AG, Nicholls RJ. Proctocolectomy without ileostomy for ulcerative colitis. Br Med J. 1978

Jul;2:85–88.

2. Tekkis PP, Lovegrove RE, Tilney HS et al. Long-term failure and function after restorative

proctocolectomy - a multi-centre study of patients from the UK national ileal pouch registry. Color Dis.

2010 May;12:433–441.

3. Tulchinsky H, Hawley PR, Nicholls J. Long-Term Failure After Restorative Proctocolectomy for Ulcerative

Colitis. Ann Surg. 2003;238:229–234.

4. Hueting WE, Buskens E, van der Tweel I et al. Results and complications after ileal pouch anal

anastomosis: A meta-analysis of 43 observational studies comprising 9,317 patients. Dig Surg.

2005;22:69–79.

5. Forbes SS, O’Connor BI, Charles Victor J, Cohen Z, McLeod RS. Sepsis is a major predictor of failure after

ileal pouch-anal anastomosis. Dis Colon Rectum. 2009;52:1975–81.

6. Fazio VW, Tekkis PP, Remzi F et al. Quantification of risk for pouch failure after ileal pouch anal

anastomosis surgery. Ann Surg. 2003;238:605-14-7.

7. Moher D, Liberati a, Tetzlaff J, Altman DG, Grp P. Preferred Reporting Items for Systematic Reviews

and Meta-Analyses: The PRISMA Statement (Reprinted from Annals of Internal Medicine). Phys Ther.

2009;89:873–880.

8. Rahbari NN. Definition and grading of anastomotic leakage following anterior resection of the rectum :

A proposal by the International Study Group of Rectal Cancer. 2010;339–351.

9. Lightner AL, Fletcher JG, Pemberton JH, Mathis KL, Raffals LE, Smyrk T. Crohn’s Disease of the Pouch: A

True Diagnosis or an Oversubscribed Diagnosis of Exclusion?. Dis Colon Rectum. 2017;60:1201–1208.

10. Raval MJ, Schnitzler M, O’Connor BI et al. Improved outcome due to increased experience and

individualized management of leaks after ileal pouch-anal anastomosis. Ann Surg. 2007;246:763–770.

11. Malcolm PN, Bhagat KK, Chapman MA, Davies SG, Williams NS, Murfitt JB. Complications of the ileal

pouch: is the pouchogram a useful predictor?. Clin Radiol. 1995;50:613–617.

12. da Silva GM, Wexner SD, Gurland B, Gervaz P, Moon S Do, Efron J, et al. Is routine pouchogram prior to

ileostomy closure in colonic J-pouch really necessary?. Colorectal Dis. 2004;6:117–120.

13. van der Ploeg VA, Maeda Y, Faiz OD, Hart AL, Clark SK. The prevalence of chronic peri-pouch sepsis in

patients treated for antibiotic-dependent or refractory primary idiopathic pouchitis. Colorectal Dis.

2017 Sep;19:827–831.

14. Sahami S, Buskens CJ, Fadok TY et al. Defunctioning Ileostomy is not Associated with Reduced Leakage

in Proctocolectomy and Ileal Pouch Anastomosis Surgeries for IBD. J Crohn’s Colitis. 2016 Jul; 10: 779-

785

15. Theodoropoulos GE, Choman EN, Wexner SD. Salvage procedures after restorative proctocolectomy: a

systematic review and meta-analysis. J Am Coll Surg. 2015;220:225–242

16. Ouro S, Thava B, Shaikh I, Clark SK. Management of pouch dysfunction in a tertiary centre. Colorectal

14

Dis. 2016;18:1167–1171.

17. Sagap I, Remzi FH, Hammel JP, Fazio VW. Factors associated with failure in managing pelvic sepsis after

ileal pouch-anal anastomosis (IPAA) - A multivariate analysis. Surgery. 2006;140:691–704.

18. Van Koperen PJ, Van Berge Henegouwen MI, Slors JFM, Bemelman WA. Endo-sponge treatment of

anastomotic leakage after ileo-anal pouch anastomosis: report of two cases. Colorectal Dis.

2008;10:943–944.

19. Maruthachalam K, Kumar R, Hainsworth P. Parking the Pouch: Pouch Salvage After Anastomotic Leak

Following Restorative Proctocolectomy. Report of a Case. Dis Colon Rectum. 2008;51:1724–1726.

20. Kirat HT, Kiran RP, Oncel M, Shen B, Fazio VW, Remzi FH. Management of Leak From the Tip of the ‘J’ in

Ileal Pouch-Anal Anastomosis. Dis Colon Rectum. 2011;54:454–459.

21. Mennigen R, Senninger N, Bruewer M, et al. Pouch function and quality of life after successful

management of pouch-related septic complications in patients with ulcerative colitis. Langenbeck’s

Arch Surg. 2012;397:37–44.

22. 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? Int J Colorectal Dis.

2011;26:1469–1474.

23. Verlaan T, Bartels S a. L, van Berge Henegouwen MI et al. Early, minimally invasive closure of

anastomotic leaks: a new concept. Color Dis. 2011;13(SUPPL. 7):18–22.

24. Varadarajulu S. Endoscopic ultrasound-guided drainage of a pelvic abscess via a J-pouch. Endoscopy.

2012;44:E92–3.

25. Srinivasamurthy D, Wood C, Slater R et al. An initial experience using transanal vacuum therapy in

pelvic anastomotic leakage. Tech Coloproctol. 2013;17:275–281.

26. Gardenbroek TJ, Musters GD, Buskens CJ et al. Early reconstruction of the leaking ileal pouch-anal

anastomosis: a novel solution to an old problem. Colorectal Dis. 2015;17:426–432.

27. Thosani N, Sethi S, Hovsepian D, Kochar R, Welton M, Banerjee S. Rendezvous EndoSeel Technique for

Non-operative Closure of Anastomotic Leak After Ileoanal Pouch Operation. Dig Dis Sci. Springer US;

2015;60:3545–3548.

15

Table 1 Study Characteristics. If the study included more than one pathology, then the number of patients with UC is included in parenthesis.

Author Country Year of Publication

Patients studied

(n)

Female Gender

(%)

Average Age Study summary

Average Follow Up (Months)

Functional outcomes reported?

Quality of life

outcome reported?

1.Sagap (17) USA 2006157

(85 UC + 24 I/UC)

41 38

Factors associated with

failure in managing pouch

sepsis

66 No No

2.Raval (10) Canada 2007 141(131 UC) 43 36

Individualised management of

leaks36 No No

3.Van Koperen (18) Holland 2007 2 (2 UC) 50 34

Long course Endo-SPONGE®

therapyNR No No

4.Maruthachalam (19) UK 2008 1 N/A 59

Interim pouch salvage as a

mucus fistula18 Yes No

5.Kirat (20) USA 2011 27 (22 UC) 48 37

Management of leaks from the

tip of the J pouch

38 Yes Yes

6.Mennigen (21) Germany 2011 12 (12 UC) 50 34.5

Function after successful

management of PRSC

23.5 Yes Yes

7.Kirat (22) USA 2011

71(61 UC post-op

histology)

30 37

CT vs trans-anastomotic drainage of

pouch abscesses

45 Yes Yes

8.Verlaan (23) Netherlands 2011 6 (5 UC) 17 50

Early closure of anastomotic defect after short course

Endo-SPONGE® therapy

4.5 No No

9.Varadarajulu (24) USA 2012 1 NR NREUS guided drainage of

pouch abscessNR No No

10.Srinivasamurthy (25) UK 2012 6

(1 UC) N/A 50Endo-SPONGE® therapy in pelvic

leaks41 Yes No

11.Gardenbroek (26) Netherlands 2015 15

(9 UC) 20 37

Early closure of anastomotic defect after short course

Endo-SPONGE® therapy in

comparison with standard

treatment

25 No No

12.Thosani (27) USA 2015 1 N/A 50

Over the scope clip closure of anastomotic

defect

1.5 No No

16

Table 2 Functional Outcomes

Author F/U Score Used

Score outcome

Daily Stool Frequency Urgency In-

continenceSeepage Pads

Day Night Day Night

Maruthachalam 1.5y - - NR NR ‘complete continence’ NR NR NR NR

Kirat

Trans-anal

Drainage

Mean 4y - - 7.6 3.6 7.5% 2.5% 35.1% NR 17.9% NR

CT Drainage

Mean 3y - - 7.9 3.3 6.2% 6.2% 31.2% NR 37.5% NR

MennigenSepsis

1yÖresland(0-16; 0

optimum)

Mean 8.2 7 NR NR NR NR NR NR

Controls Mean 6.6 6 NR NR NR NR NR NR

KiratSalvage

Mean 3.2 y - - 8.4 4.6 15.4% 0 18.2% 25% 8.3% 8.3%

ControlsMean 5.6y - - 8.8 10 4.3% 2.1% 23.9% 30.4% 19.6% 21.7%

Srinivasamurthy NR - - 6 NR

‘remain[s] continent throughout an 8-h shift at work’

NR NR NR NR

Table 3 Quality of Life Outcomes

Author Subgroup Questionnaire Outcome P Value

Kirat Transanal Drainage CGQL 0.7 0.2 0.9CT Drainage 0.7 0.1

Mennigen

Sepsis SIBDQ(1-7; 7 optimum)

5.00.20

Controls 5.5Sepsis GIQLI

(0-144; 144 optimum)95.8 0.12

Controls 107.3

Kirat Salvage CGQL 0.8 0.3 0.67Controls 0.8 0.2

17

Figure 1: PRISMA flow diagram

18

Figure 2. Algorithm for the management of early pouch sepsis. *See 'presentation' in results section. ☩As per

Gardenbroek et al (16). ☨Percutaneous drains should be removed as quickly as possible.

19

Appendix 1.Search strategy:1. exp Colitis, Ulcerative/2. ulcerative coliti*.tw.3. 1 or 24. exp Colonic Pouches/5. exp Proctocolectomy, Restorative/6. ("restorative proctocolectomy" or "RPC" or "ile* pouch anal anastomosis" or"IPAA" or "ile* pouch" or "j pouch" or "s pouch" or "w pouch" or "h pouch").tw.7. 4 or 5 or 68. 3 and79. exp Anastomotic Leak/10. exp Sepsis/11. exp Abscess/ or exp Abdominal Abscess/12. exp Postoperative Complications/13. ("leak*" or "collection" or "sac* collection" or "abscess" or "fistula*").tw.14. ("pouch related septic complications" or "PRSC").tw.15. "pouch sepsis".tw.16. 9 or 10 or 11 or 12 or 13 or 14 or 1517. 8 and 1618. remove duplicates from 1719. 18 and 2000:2017.(sa_year).20. 19 and "Journal: Article" [Publication Type]

20