In patients with inflammatory bowel disease (IBD), surgical intervention is sometimes required due to medically refractory colitis or the development of neoplasia, and restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA) is the most common procedure for those with colitis. However, despite a preoperative diagnosis of ulcerative colitis (UC), a phenotype of the J pouch which has features similar to Crohn’s disease (CD) may develop after the surgery. This study aims to assess the frequency and risk factors of fistula development in IBD patients with a J pouch.
This is a retrospective single-center study of IBD patients treated by total proctocolectomy with IPAA and who subsequently underwent pouchoscopies at the University of Chicago between January 2007 and December 2019. We reviewed the endoscopic findings in different anatomic areas of the pouch (the afferent limb, inlet, “tip of the J”, proximal and distal pouch, anastomosis, rectal cuff, anal canal, and perianal area). Endoscopic phenotypes were classified into 5 main categories: (1) afferent limb, (2) inlet involvement, (3) diffuse inflammation of the pouch body, (4) cuffitis, and (5) J pouch with fistulas. This analysis included pouches with any type of fistulas that developed ≥6 months after ileostomy takedown. We assessed clinical data and pathological findings of the resected colon including deep, focal inflammation, and granulomas generally thought to be unique features of CD. Deep inflammation included ulcerations, fissures, and lymphoid aggregates observed below the level of the mucosa. Fisher’s test was used for univariate analysis to assess factors contributing to fistula development. Logistic regression analysis was performed as a multivariable analysis including variables with a P-value < 0.10 in the univariate analysis.
We reviewed 1,359 pouchoscopies from 426 IBD patients who underwent proctocolectomy with IPAA and identified 18.8% of patients (78/414) with a J pouch and fistulas ≥ 6 months from ileostomy takedown. Ninety-one percent of patients were preoperatively diagnosed with UC. The most frequent type of fistula was perianal fistula (60.3%), followed by enterovaginal or rectovaginal fistula (24.4%), and fistula from the pouch (7.7%). Among 325 IBD patients with available pathological reports of resected colon, 86 patients (26.5%) showed deep inflammation and 88.4% (76/86) of patients were diagnosed with UC rather than CD by our pathologists. On univariate analysis, we did not find any significant clinical factors for fistula development in patients with a J pouch. As for pathological data, our univariate and multivariable analyses showed that deep inflammation of the resected colon was significantly associated with fistula development (OR = 3.52; 95%CI = 1.84-6.72; P < 0.001), but focal inflammation and granulomas were not. Notably, this association was not observed in patients with other endoscopic phenotypes of the J pouch.
In this tertiary cohort of patients with a preoperative diagnosis of UC, we found that deep inflammation of the resected colon is a specific predictor for fistula formation in the J pouch. These findings suggest that in such patients a careful monitoring and early treatment strategy may be beneficial.