Histological assessment of the distal 'doughnut' in patients undergoing stapled restorative proctocolectomy with high or low anal transection.

A non-randomized prospective study of 38 patients, 32 with ulcerative colitis and six with familial adenomatous polyposis (FAP), who underwent high or low anal transection during stapled restorative proctocolectomy was undertaken. The median (range) height of the staple line 6 months after operation was 5.2 (3.2-6.0) cm after high transection compared with 2.9 (1.8-3.6) cm after low transection. Nineteen of 20 patients after high anal transection had columnar epithelium in the distal 'doughnut' versus 16 of 18 after low transection. Active colitis was present in 12 of 19 'doughnuts' in patients with high anal transection and columnar mucosa and in seven of 16 after low transection. Nine patients (high transection two, low transection seven; P < 0.05) had striated muscle in the stapled distal 'doughnut'. Dysplasia was found in the resected colon in one patient with ulcerative colitis and adenocarcinoma in two colectomy specimens (ulcerative colitis, one; FAP, one). No dysplasia or carcinoma was seen in any of the 'doughnuts' from patients with ulcerative colitis. Four patients with FAP (high transection, two; low transection, two) had microadenoma in the distal 'doughnut'. Despite attempts to place a stapled pouch-anal anastomosis below the anal transition zone, it was not possible to remove columnar mucosa completely from the remaining anal canal in most patients (16 of 18). High anal transection and pouch-anal anastomosis should be the preferred option in restorative proctocolectomy, as a dentate-line anastomosis may not fully eliminate columnar epithelium and may involve resection of some of the external sphincter.