Reconstructive Surgery for Failed Ileal Pouch-Anal Anastomosis

PURPOSE: Salvage procedures for failed ileal pouch-anal anastomoses frequently require total reconstruction with a combined abdominal and perineal approach. The aim of this study was to determine the indications for surgery and the outcomes in this group of patients. METHODS: All patients who underwent combined abdominal and perineal ileal pouch-anal anastomosis reconstruction at the Mount Sinai Hospital between 1982 and 2000 were reviewed. Data were collected prospectively in the inflammatory bowel disease database. RESULTS: Sixty-three reconstructive procedures were performed in 57 patients, with a mean age of 33.9 (±10.4) years at the time of reconstruction. There were 14 males. The mean follow-up was 69.1 months. The initial indication for ileal pouch-anal anastomosis was ulcerative colitis in 98 percent. The primary indication for reconstruction was pouch-vaginal fistula in 21 patients, long outlet in 14, pelvic sepsis in 14, ileoanal anastomotic stricture in 5, pouch-perineal fistula in 2, and chronic pouchitis in 1. The mean operative time was four hours (±1.1), the average blood loss was 500 mL (±400), and the average length of stay was 10.3 days (±4.6). All patients had a diverting ileostomy. Forty-two (73.6 percent) of the patients have a functioning pouch. Seven (12.3 percent) patients have had their pouch excised. The ileostomy has not yet been closed in 8 (14 percent) patients; 3 of these patients are awaiting closure, whereas the remaining 5 have a permanently defunctioning ileostomy. Eighty-nine percent have ten or fewer bowel movements per day. No patients are incontinent of stool during the day, whereas two patients are incontinent at night. Seventeen percent complain of frequent urgency. Despite this, more than 80 percent rate their physical and psychological health as good to excellent. CONCLUSION: Reconstructive pouch surgery has a high success rate in experienced hands. The functional results in those whose pouch is in use are good.