Pouch failures following restorative proctocolectomy in ulcerative colitis

Since its introduction by Parks and Nicholls in 1978 [1], restorative proctocolectomy has become the procedure of choice in patients with ulcerative colitis (UC) requiring surgery. Despite the evolved details of the operation and centralization of the surgery to centres having the number of operations on an acceptable level, poor functional results and pouch-related complications may compel excision of the pouch and the construction of a permanent stoma. The long-term failure rate for RPC is reported to be 10–15% [2‐4]. The most common causes of pouch failure have been pelvic sepsis, poor pouch function, pouchitis and Crohn’s disease [4‐6]. The factors leading to pouch excision are still poorly understood. This prompted us to explore the long-term risk of pouch failure, covering more than three decades of RPC cases in a high-volume hospital. We report the reasons for pouch failure, factors associated with the risk and the morbidity of pouch excision.

All consecutive patients with UC undergoing RPC at Tampere University Hospital between 1985 and 2016 were identified from the patient records using the ICD-9 and ICD-10 codes for UC and NCSP (Nordic Classification for Surgical Procedures) codes for the types of operations. A database to form an RPC registry was collected from patient files including details on patient history, surgical technique, postoperative morbidity and follow-up. On patients with pouch failure, we collected details on patient history, reasons for failure, treatment prior to failure and on the pouch excision procedure. Seventy-five patients had moved elsewhere in Finland. Data on the pouch excision of these patients were retrieved from the register of the National Institute of Health and Welfare (NIHW). The patients were identified by the NSCP procedure codes. A widely used prognostic model, the Charlson Comorbidity Index, was used to categorise pouch failure patients’ comorbidities, a method based on the diagnostic codes [7]. Each comorbidity category had an associated weight (from 1 to 6, 6 being the most severe) based on the adjusted risk of mortality or resource use, and the sum of all the weights yields a single comorbidity score for a patient. A score of zero indicates that no comorbidities were found. The higher the score, the more likely the outcome to result in mortality or higher resource use. Complications were classified according to the Clavien-Dindo (C-D) classification, consisting of seven grades (I, II, IIIa, IIIb, IVa, IVb and V) [8]. Crohn’s disease was diagnosed in endoscopy, either by pouch endoscopy or endoscopy from stomal entrance, histology, capsule imaging or magnetic resonance imaging (MRI) and clinical picture.

All pouches in our RPC registry are of J-type and the anastomosis is either handsewn or stapled. Pouch failure was defined as the need for a permanent ileostomy with or without pouch excision. Early complications were defined as occurring within 30 days of the operation.

The data were analysed using SPSS (Version 25.0 IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.). For categorical variables, the results were given as frequencies and percentages, and for continuous variables as means and standard deviations or as medians. Chi-square or Fisher’s exact test was used to assess differences in categorical variables. The binary logistic regression analysis was used to determine the predictors of pouch failure. Results are shown as odds ratios (ORs) with their 95% confidence intervals (CIs). The cumulative probability of pouch excision was estimated by the Kaplan-Meier method. Statistical significance was set at p ≤ 0.05.

This study investigated the frequency and reasons for pouch failure after RPC for UC in the second biggest IBD surgery centre in Finland in a period spanning more than three decades. The overall frequency of pouch failure was 10.8%. In a previous report with a large number of patients and in a meta-analysis, similar pouch failure percentages of 9.7 [5] and 8.6% [9], respectively, were seen with a minimum follow-up time of 5 years. The Charlson Comorbidity Index among pouch failure patients was 0 in 89%, which means that failure patients were otherwise mainly healthy. In a Danish national registry study with a median follow-up of 11.4 years, a 5-year risk of 9.1%, 10-year risk of 12.1% and 20-year risk of 18.2% for pouch failure were found [2].

The reasons leading to pouch failure were often multifactorial: one complication leading to another and subsequently to functional difficulties. The reasons reported in different studies may vary according to how pouch problems are categorised or recorded. The most common reasons for failure in our study were septic events in 21(40%) (fistula 23%, leakage 15%, perforation 2%), pouchitis in 13%, chronic pouchitis in 21%, Crohn’s disease in 13% and functional reasons: poor function was seen in 16 (30%), incontinence in 12 (23%) and stricture in 12 (23%) patients. In a British multicentre study comprising 94 pouch failure patients, the causes were similar: septic complications in 33 (35.1%), poor function in 29 (30.8%), pouchitis in 9 (9.6%) and Crohn’s disease in 5.3% [4]. Septic complications, deterioration of the functional results and Crohn’s disease have also been reported earlier in 1997 [6] and in large studies in this millennium [5]. We have previously reported our functional results for those with pouch in place and 70% of the patients reported good function of the pouch as measured with the Oresland score [10]. Usually before the pouch failure, the function is devastating and the patient is too tired to consider new procedures to save the pouch.

Here, in subjects with pouch failure, the diagnosis of Crohn’s disease was made before in two patients and after failure in five. Altogether, in patients with Crohn’s disease diagnosed after RPC, failure was seen in 47%. This is similar to the findings obtained in large studies [11, 12]. No granulomas were found in our Crohn’s series. A similar result was reported in the Mayo Clinic study with 35 Crohn’s disease pouch failure patients; in only seven cases was pathological confirmation given [13]. In the whole RPC series, the frequency of Crohn’s disease was only 3%. The histologic variability of Crohn’s cannot give an exact diagnosis in every case, but the diagnosis is a combination of the clinical picture, X-ray findings and histology. Nevertheless, in cases with pouch failure, the diagnosis of Crohn’s disease should be borne in mind. In our series there was no significant age or gender difference for pouch failure for Crohn or UC diagnosis.

In our study 15.2% of men and 4.7% of women suffered pouch failure. In earlier studies, there has been no significant difference in gender distribution [14‐16] except in the Danish national registry study, which reported a risk in women 39% greater than that in men over time [2]. In early failures, pelvic sepsis due to anastomotic complications played a major role but there was no difference between genders in this report. Nor was there any difference between genders in median time from RPC to failure. We found no explanation for this gender distribution in our study.

Leakage was found to be an independent risk factor for pouch failure OR 3.65 (95%CI 1.21–1.83) p = 0.022. This has been reported in several reports before and the prompt diagnose and treatment of leakage cannot be stressed enough [17].

The differences in the uni- and multivariate analyses for the use of covering stoma and type of anastomoses are explained with the changes in practice through the years: handsewn anastomoses were performed mainly in 1985–2005 and covering ileostomy when considered necessary in 1985–2005 and as a routine after that. The year of IPAA has also been taken into account there and it obviously affects the follow-up time.

In our study, all but one of the pouch failure patients (52, 98%) underwent pouch excision. In earlier reports, this percentage has been smaller: in the study by Foley 40% [18], Tulchinsky and Meagher 60% [5, 19] and McRae 84% [20]. In another Finnish study, all 52 failed pouches were also excised [21]. In the present study, 8 (15%) out of 52 patients underwent a loop ileostomy and six patients had an end stoma built before the pouch excision operation. A pouch left in place may burden patients with incontinence. There is a need for guidelines on the management of pouch failure patients; the issues to consider are morbidity of the pouch left in place, morbidity after pouch excision, danger of dysplasia and follow-up of the redundant pouch.

In this study, morbidity of the excision was frequent but most of the complications were mild and thus acceptable. There are very few studies on morbidity after pouch excision. Karoui et al. concluded that early and especially late morbidity is high and pouch excision RPC patients should be initially counselled on this [22]. Prudhomme et al. studied perineal wound healing and divided the patients into groups by diagnosis. Crohn’s disease seemed to be a major risk for perineal wound healing [23]. Bengtsson et al. [24] analysed their diverted pouch failure patients who did not undergo excision (n = 22) and found no dysplasia or cancer in histopathology in pouch biopsies; none of the patients requested further surgery and the majority had satisfactory ileostomy function. Altogether, dysplastic transformation in the ileal pouch has been estimated to be rare [25]. The present study supports this view as no dysplasia was found in the postoperative histology. One patient died of mucinotic cancer after pouch excision; this cancer was not found preoperatively despite extensive investigations.

The study was limited due to its retrospective nature and lack of systematic follow-up of the pouch failure patients.

Although most of the RPC patients operated on for UC had satisfactory functional outcome, 10.8% experienced pouch failure in the long run. The most common reasons for failure were deterioration of functional results after septic complications. Pouch excision is a major surgery and associated with moderate morbidity, necessitating precise planning and patient counselling. The possibility of Crohn’s disease should be considered in UC patients with pouch failure who have been operated on.