Acute Severe Ulcerative Colitis Is Associated With an Increased Risk of Acute Pouchitis

Abstract

Background

Pouchitis occurs in up to 80% of patients after total proctocolectomy (TPC) with ileal pouch–anal anastomosis (IPAA) and has been associated with microbial and host-related immunological factors. We hypothesized that a more robust immune response at the time of colectomy, manifested by acute severe ulcerative colitis (ASUC), may be associated with subsequent acute pouchitis.

Methods

This was a retrospective cohort analysis of all patients with UC or indeterminate colitis complicated by medically refractory disease or dysplasia who underwent TPC with IPAA at Mount Sinai Hospital between 2008 and 2017 and at least 1 subsequent pouchoscopy. Acute pouchitis was defined according to the Pouchitis Disease Activity Index. Cox regression was used to assess unadjusted relationships between hypothesized risk factors and acute pouchitis.

Results

A total of 416 patients met inclusion criteria. Of the 165 (39.7%) patients who underwent urgent colectomy, 77 (46.7%) were admitted with ASUC. Acute pouchitis occurred in 228 (54.8%) patients a median of 1.3 (interquartile range, 0.6-3.1) years after the final surgical stage. On multivariable analysis, ASUC (hazard ratio [HR], 1.50; 95% confidence interval [CI], 1.04-2.17) and a greater number of biologics precolectomy (HR, 1.57; 95% CI, 1.06-2.31) were associated with an increased probability of acute pouchitis, while older age at colectomy (HR, 0.98; 95% CI, 0.97-0.99) was associated with a decreased probability. Time to pouchitis was significantly less in patients admitted with ASUC compared with those not (P = .002).

Conclusion

A severe UC disease phenotype at the time of colectomy was associated with an increased probability of acute pouchitis.

Key Messages.

• What is already known? Pouchitis is common in patients with ulcerative colitis after total proctocolectomy with ileal pouch–anal anastomosis.

• What is new here? In a retrospective cohort analysis of 416 patients, acute severe ulcerative colitis at the time of colectomy was significantly associated with subsequent acute pouchitis.

• How can this study help patient care? The risk of pouchitis may be driven by immune activation and disease severity precolectomy, and these variables may be used for preoperative risk stratification and counseling.

Introduction

Pouchitis occurs in up to 80% of patients with ulcerative colitis (UC) after total proctocolectomy (TPC) with ileal pouch–anal anastomosis (IPAA).¹ The pathogenesis of pouchitis remains unclear but is thought to involve a complex interaction between the microbiome and mucosal immune system.² Fecal stasis in the pouch promotes dysbiosis, triggers colonic metaplasia of the ileal mucosa, and activates the innate and the adaptive mucosal immune systems.^3,4 Manipulation of the pouch microbiome via antibiotics is considered first-line therapy for pouchitis, with induction of remission in up to 80% of patients.⁵

In addition to microbial factors, a number of host-related immunological factors have been associated with an increased risk of pouchitis. These include history of primary sclerosing cholangitis (PSC), extensive UC, anti-tumor necrosis factor (TNF) therapy precolectomy, and polymorphisms in the interleukin-1 receptor and NOD2/CARD15.^6-8 Recently, a colectomy risk score (CRS) used to stratify UC severity and predict colectomy was noted to be an independent predictor of pouchitis, with a significantly greater cumulative incidence of pouchitis in patients with a high-risk CRS compared with a low-risk CRS.^9,10 In addition, a strong correlation was noted between the CRS and Pouchitis Disease Activity Index (PDAI) score, with significantly greater PDAI scores at pouchitis diagnosis in patients with a high-risk CRS compared with a low-risk CRS.¹⁰

These data suggest that a more robust immune response may be associated with an increased risk of pouchitis. However, to date, formal assessments of this hypothesis are lacking. Herein, we tested this hypothesis by performing a retrospective cohort analysis at our high-volume inflammatory bowel disease center, comparing the incidence of acute pouchitis in patients admitted with acute severe UC (ASUC) at the time of colectomy with the incidence of acute pouchitis in patients admitted without ASUC.

Methods

Study Population and Outcome

This was a retrospective cohort analysis of all patients with UC or indeterminate colitis complicated by medically refractory disease or dysplasia who underwent TPC with IPAA at Mount Sinai Hospital between January 2008 and December 2017 and at least 1 subsequent pouchoscopy. All surgical procedures were performed at Mount Sinai Hospital by 1 of 15 colorectal surgeons. Patients <18 years of age or with a baseline diagnosis of Crohn’s disease (CD) were excluded. The primary outcome was development of acute idiopathic pouchitis. This study was approved by the Mount Sinai Hospital Institutional Review Board.

Definitions

Acute severe UC was defined by the Truelove and Witts criteria as ≥6 bloody bowel movements per 24 hours and 1 of the following admission clinical characteristics: temperature >37.8°C, heart rate >90 beats/min, hemoglobin <10.5 g/dL, or erythrocyte sedimentation rate >30 mm/h.¹¹ Acute pouchitis was defined as a total PDAI score ≥7 with symptoms of increased stool frequency from postoperative baseline, rectal bleeding, fecal urgency, abdominal cramps, or fever (temperature ≥37.8°C) in the setting of an abnormal pouchoscopy with mucosal edema, granularity, friability, loss of vascular pattern, mucosal exudate, and/or ulceration and histologic inflammation.¹² Pouchitis was defined as a modified PDAI score ≥5 using clinical and endoscopic subscores only if there was no histology associated with the pouchoscopy or if the histology was reported in a manner that did not permit histologic PDAI subscore calculation. The PDAI score was calculated retrospectively for each patient using clinical, pouchoscopy, and pathology notes. A colectomy was defined as urgent if it was performed during the admission for the indication of medically refractory UC. Time to colectomy was defined as the time from admission for UC to colectomy.

Data Collection and Variables

Clinical information was abstracted from the electronic medical record using a standardized data collection sheet. Collected patient demographics and disease characteristics included age, sex, disease duration, precolectomy biologic use, disease extent at colectomy, colectomy indication, pathologic diagnosis (UC or indeterminate colitis), admission with ASUC, and number of surgical stages. All postoperative pouchoscopy procedure notes were reviewed and results recorded.

Analytic Approach

Descriptive statistics were performed to describe baseline characteristics and are reported as proportions or medians (with interquartile range [IQR]) for categorical and continuous variables, respectively. Univariable Cox regression was used to assess unadjusted relationships between hypothesized risk factors and acute pouchitis. Multivariable Cox regression for the primary outcome of acute pouchitis was performed a priori with selection of the following clinically relevant variables: age (analyzed as a continuous variable), sex, ASUC admission, number of precolectomy biologics (analyzed as a categorical variable), number of surgical stages, and disease extent. A subgroup analysis excluding patients with indeterminate colitis was performed. Hazard ratios (HRs) and 95% confidence intervals (CIs) are reported. All analyses were performed using SAS v9.4 (SAS Institute). Two-sided P values <.05 were considered statistically significant.

Results

A total of 416 patients 18 years of age and older underwent TPC with IPAA at Mount Sinai Hospital and at least 1 subsequent pouchoscopy between 2008 and 2018 for symptoms or dysplasia surveillance. The median age at colectomy was 35.4 (IQR, 26.1-49.0) years and 224 (53.8%) patients were male. Biologics were used in 266 (63.9%) patients precolectomy, and disease extent was reported as extensive in 329 (79.1%). Urgent colectomy was performed in 165 (39.7%) patients, of whom 77 (46.7%) were admitted with ASUC. The median time to urgent colectomy was 4 (IQR, 2-9) days. Detailed patient demographics and clinical characteristics for the cohort are provided in Table 1. The median time to first pouchoscopy was 1.0 (IQR, 0.4-2.2) years and the median duration of follow-up for all patients was 3.9 (IQR, 2.0-6.4) years.

Table 1.

Clinical and demographic characteristics of patients who underwent TPC with IPAA (N = 416).

Age at colectomy, y	35.4 (26.1-49.0)
BMI at colectomy, kg/m2	22.5 (19.3-25.6)
Male	224 (53.8)
Precolectomy disease duration, y	5.0 (1.9-12.0)
Cyclosporine precolectomya	69 (16.6)
Tacrolimus precolectomya	6 (1.4)
Biologics precolectomya	266 (63.9)
• Anti-TNF (adalimumab, infliximab)	207 (77.8)
• Vedolizumab	56 (21.1)
• Ustekinumab	3 (1.1)
Number of biologics precolectomy
• 1	202 (69.2)
• 2	61 (22.9)
• 3	3 (1.1)
Disease type
• Ulcerative colitis	397 (95.4)
• Indeterminate colitis	14 (3.4)
Disease extent
• Left sided	87 (20.9)
• Extensive	329 (79.1)
Colectomy indication
• Medically refractory disease	367 (88.2)
• Dysplasia	49 (11.8)
ASUC admissions	77 (18.5)
Number of surgical stages
• 1	38 (9.1)
• 2	157 (37.7)
• 3	221 (53.1)
Anastomosis
• Stapled	294 (70.7)
• Handsewn	122 (29.3)
Rectal cuff length, cm	2 (1-3)
Postoperative complications
• Leak	35 (8.4)
• Abscess	46 (11.1)

Values are median (interquartile range) or n (%).

Abbreviations: ASUC, acute severe ulcerative colitis; BMI, body mass index; IPAA, ileal pouch–anal anastomosis; TNF, tumor necrosis factor; TPC, total proctocolectomy.

^aWithin 1 year of precolectomy.

Acute pouchitis occurred in 228 (54.8%) patients a median of 1.3 (IQR, 0.6-3.1) years after final surgical stage, and all patients were prescribed antibiotics. Of the 77 patients admitted with ASUC, 67 (87.0%) had at least 1 biologic exposure and 47 (61.0%) developed pouchitis a median of 0.9 (IQR, 0.5-1.9) years after final surgical stage. On univariable analysis, ASUC (HR, 1.87; 95% CI, 1.34-2.62), a greater number of biologics precolectomy (HR, 1.76; 95% CI, 1.24-2.50), staged surgery (HR, 2.14; 95% CI, 1.28-3.57), and extensive colitis (HR, 1.74; 95% CI, 1.15-2.62) were associated with an increased probability of acute pouchitis, while older age at colectomy (HR, 0.98; 95% CI, 0.97-0.99) was associated with a decreased probability of acute pouchitis. On multivariable analysis, ASUC (HR, 1.50; 95% CI, 1.04-2.17) and a greater number of biologics precolectomy (HR, 1.57; 95% CI, 1.06-2.31) were associated with an increased probability of acute pouchitis, and older age at colectomy (HR, 0.98; 95% CI, 0.97-0.99) was associated with a decreased probability of acute pouchitis. Full univariable and multivariable results are provided in Table 2. On subgroup analysis excluding patients with indeterminate colitis, the results of the multivariable analysis remained largely unchanged, with ASUC (HR, 1.53; 95% CI, 1.06-2.22) and a greater number of biologics precolectomy (HR, 1.29; 95% CI, 1.03-1.60) associated with an increased probability of acute pouchitis and with older age at colectomy (HR, 0.98; 95% CI, 0.97-0.99) associated with a decreased probability of acute pouchitis. Time to pouchitis was significantly less in patients admitted with ASUC compared with those not (P = .002) (Figure 1).

Table 2.

Univariable and multivariable results: Primary outcome acute pouchitis.

Variable	HR (95% CI)	P Value
Univariable results
Colectomy agea	0.98 (0.97-0.99)	.01b
Sex	1.23 (0.94-1.60)	.13
PSC	1.61 (0.88-2.95)	.13
ASUC	1.87 (1.34-2.62)	.0002b
Number of biologics precolectomy (ref = 1)c	1.76 (1.24-2.50)	.002b
Time to colectomy	1.03 (0.99-1.06)	.07
Extensive colitis (ref = left sided)	1.74 (1.15-2.62)	.01b
Stapled anastomosis (ref = handsewn)	0.99 (0.75-1.31)	.94
Number of surgical stagesc	2.14 (1.28-3.57)	.004b
Postoperative leak	0.64 (0.38-1.08)	.09
Postoperative abscess	0.81 (0.53-1.23)	.32
Multivariable results
Colectomy agea	0.98 (0.97-0.99)	.01b
Sex	1.15 (0.84-1.56)	.39
ASUC	1.50 (1.04-2.17)	.03b
Number of biologics precolectomyc	1.57 (1.06-2.31)	.02b
Extensive colitis	1.42 (0.90-2.22)	.13
Number of surgical stagesc	1.91 (0.98-3.69)	.06

Abbreviations: ASUC, acute severe ulcerative colitis; CI, confidence interval; HR, hazard ratio; PSC, primary sclerosing cholangitis.

^aAnalyzed as a continuous variable.

^bStatistically significant.

^cAnalyzed as a categorical variable.

Figure 1.

Kaplan-Meier estimates of developing acute pouchitis in patients with vs without acute severe ulcerative colitis (ASUC).

Discussion

In this single-center study of patients with UC or indeterminate colitis who underwent TPC with IPAA and at least 1 subsequent pouchoscopy, acute pouchitis occurred in approximately 55% of patients a median of 1.3 (IQR, 0.6-3.1) years after final surgical stage. We observed that markers of increased UC disease severity, specifically ASUC and greater use of biologics precolectomy, were associated with the development of acute pouchitis. Similar to previously published data, older age was protective against pouchitis.^13,14

Multiple observations support the significant impact of host-related immunological factors on the development of acute pouchitis. The first, and perhaps most striking, is the significantly lower incidence of acute pouchitis in patients with familial adenomatous polyposis compared with patients with UC.^15,16 Despite undergoing the same TPC with IPAA procedure as patients with UC and theoretically harboring the same risk of fecal stasis and microbial dysbiosis, patients with familial adenomatous polyposis lack the underlying immune dysregulation of patients with UC and accordingly develop less episodes of pouchitis. The second is the significant association between extensive UC and pouchitis. Extensive colitis has been shown to be associated not only with acute pouchitis, but also specifically with diffuse pouchitis and chronic pouchitis, presumably a result of the robust immune response that persists even after TPC with IPAA.^17-19 The third is the significantly greater incidence of pouchitis in patients with PSC compared with those without.^20,21 Theorized to be immune mediated, PSC may heighten the mucosal immune response in patients with UC and result in an increased risk of pouchitis.²²

Precolectomy anti-TNF therapy has also been associated with pouchitis development. Anti-TNF therapies are thought to alter the ileal microbiome and mucosal immune system in a manner that predisposes it not only to acute pouchitis, but also to de novo CD and diffuse pouchitis.^17,23 In a prospective study of 400 patients with UC who underwent TPC with IPAA, there was a significant association between precolectomy anti-TNF therapy and the development of de novo CD.²⁴ In a recent study of 426 patients who underwent TPC with IPAA and subsequent pouchoscopy, precolectomy anti-TNF therapy was significantly associated with the development of severe, diffuse inflammation of the pouch body. Whether exposure to anti-TNF therapy is a surrogate for more aggressive disease or whether anti-TNF therapies precondition the small bowel, change the intestinal microbiome, and predispose to pouch inflammation remains unknown and merits further exploration.

Our results extend the current literature by identifying ASUC and a greater number of biologics precolectomy, the majority of which were anti-TNF therapies, to be significantly associated with the development of acute pouchitis. Both ASUC and prior anti-TNF exposure are surrogates for severe disease and immune activation, and our results suggest that pouchitis may be driven by the propensity to inflammation that exists with both. This is in line with the increased risk of acute pouchitis noted with other immunological factors discussed previously—PSC, extensive UC—and with the known mucosal immune activation that occurs in both UC and pouchitis as manifested by overlap in specific myeloid and T cell populations.²⁵

This study had several strengths. First, all patients underwent TPC with IPAA, subsequent pouchoscopy, and long-term follow-up at Mount Sinai Hospital. Second, strict definitions were utilized for ASUC and pouchitis. Third, chart review was standardized and investigators were blinded to the outcome. Limitations of this study include its retrospective nature (specifically with retrospective calculation of the PDAI) and susceptibility to selection bias. Certain variables within the Truelove and Witts criteria were not available for all patients admitted with UC, precluding disease severity assessment and potentially underestimating the number of patients with ASUC. In addition, we were unable to objectively report on surgeon experience, and certain details such as body mass index at the time of pouch construction were not collected.

Conclusions

UC disease severity and immune activation as reflected by ASUC and increasing medical therapy refractoriness were associated with acute pouchitis in this single-center study. These data provide additional variables that can be used for preoperative risk stratification and counseling and need to be validated in a large multicenter prospective study. Future steps should incorporate the comparison of genetic and immunological factors in patients admitted with and without ASUC who undergo TPC with IPAA and subsequently develop pouchitis as a means to determine novel factors associated with disease pathogenesis and develop risk stratification guidelines.

Author Contribution

M.K. is the submission’s guarantor and takes responsibility for the integrity of the work as a whole, from inception to published article. M.K.: contributed to study concept and design, acquisition of data, analysis and interpretation of data, drafting of the manuscript, critical revision of the manuscript for important intellectual content, and statistical analysis. H.P. contributed to acquisition of data and critical revision of the manuscript for important intellectual content. H.M.W. contributed to critical revision of the manuscript for important intellectual content. M.P. contributed to acquisition of data and critical revision of the manuscript for important intellectual content. S.K. contributed to critical revision of the manuscript for important intellectual content. P.S. contributed to critical revision of the manuscript for important intellectual content. A.G. contributed to critical revision of the manuscript for important intellectual content. M.D. contributed to critical revision of the manuscript for important intellectual content. S.M. contributed to critical revision of the manuscript for important intellectual content. J.F.C. contributed to study concept and design and critical revision of the manuscript for important intellectual content. All authors approved the final version of the manuscript.

Funding

M.K. receives grant support from NIH-DK127241-01A1.

Conflict of Interest

M.K. has served as a consultant for GoodRx and AbbVie. M.C.D. has served as a consultant for AbbVie, Arena Pharmaceuticals, Boehringer Ingelheim International, Bristol-Myers Squibb, Celgene, Eli Lilly, F. Hoffman-La Roche, Genentech, Gilead, Janssen, Pfizer, Prometheus, Takeda, and UCB; has received research grants from Pfizer, AbbVie, Janssen, and Prometheus; has ownership interest in Trellus; and has licensing fees from Takeda. S.M. has received research grants from Genentech and Takeda; payment for lectures from Takeda, Genentech, and Morphic; and consulting fees from Takeda, Morphic, Ferring, and Arena Pharmaceuticals. J.F.C. has received research grants from AbbVie, Janssen Pharmaceuticals and Takeda; received payment for lectures from AbbVie, Amgen, Allergan, Inc. Ferring Pharmaceuticals, Shire, and Takeda; has received consulting fees from AbbVie, Amgen, Arena Pharmaceuticals, Boehringer Ingelheim, BMS, Celgene Corporation, Eli Lilly, Ferring Pharmaceuticals, Galmed Research, Genentech, Glaxo Smith Kline, Janssen Pharmaceuticals, Kaleido Biosciences, Imedex, Immunic, Iterative Scopes, Merck, Microbia, Novartis, PBM Capital, Pfizer, Protagonist Therapeutics, Sanofi,Takeda, TiGenix, and Vifor; and holds stock options in Intestinal Biotech Development.

Data Availability

The data underlying this article will be shared on reasonable request to the corresponding author.