Management and outcome of pouch-vaginal fistulas following restorative proctocolectomy

PURPOSE: The aim of this study was to assess the short-term and long-term outcomes of surgical repair of patients with pouch-vaginal fistulas after restorative proctocolectomy. METHODS: A descriptive study was undertaken of all patients developing pouch-vaginal fistulas following restorative proctocolectomy between 1978 and 2003 in a single tertiary referral institution. Kaplan-Meier survival analysis was used to evaluate the time to first pouch-vaginal fistula recurrence and pouch-vaginal fistula-free survival at last follow-up. RESULTS: Sixty-eight patients (mean age, 32.2 years; standard deviation, 10.7) were identified with a median follow-up of 5.5 (range, 0.2-25.5) years. The origin of the pouch-vaginal fistulas was the pouch-anal anastomosis in 52 (76.5 percent) patients, pouch body/top in 9 (13.2 percent), or cryptoglandular or other source in 7 (10.3 percent). Associated early complications in patients with pouch-vaginal fistulas included pelvic sepsis in 20 (29 percent) patients, anastomotic separation in 6 (24 percent), anastomotic stricture in 16 (24 percent), small bowel obstruction in 17 (25 percent), hemorrhage in 2 (3 percent), or pouchitis in 12 (18 percent). Surgery was undertaken in 59 (87 percent) patients with 14 (20.6 percent) of them undergoing pouch excision/diversion or seton drainage. Forty-five (66 percent) patients underwent primary repair. First recurrence of pouch-vaginal fistula occurred in 27 of 45 (60 percent) patients with a median pouch-vaginal fistula-free interval of 1.6 years (95 percent confidence interval, 0.6-2.7). Fourteen (51.9 percent) patients with recurrent pouch-vaginal fistulas heated following one or more repeat procedures. The diagnosis of Crohn's disease was made in eight (12 percent) patients, with pouch-vaginal fistulas persisting or recurring in all patients with Crohn's disease within five years of the primary treatment. Median pouch-vaginal fistula-free survival was 1.4 years for patients with Crohn's disease and 8.1 years for patients with ulcerative colitis or familial adenomatous polyposis. The pouch-vaginal fistula-free survival improved with repeated local or abdominal repairs for patients with ulcerative colitis. The overall pouch failure rate for patients with pouch-vaginal fistulas was 35 percent (median pouch survival, 4.2 years). CONCLUSIONS: Pouch-vaginal fistulas can persist and recur indefinitely, even after repeated repairs. Repair in those patients with Crohn's disease uniformly failed within five years from primary repair. Patients with recurrent pouch-vaginal fistulas and ulcerative colitis should be offered salvage surgery because successful closure following initial failure occurs in approximately 50 percent.