Editor,—Lim and colleagues recently presented three cases of diversion colitis which seemed to act as a trigger for instream ulcerative colitis (Gut1999;44:279–282). We would like to present a fourth case which has recently come to our attention and which adds to the literature.

In 1994, a 75 year old woman presented with a five year history of faecal soiling and urinary incontinence. A former smoker of 50 years, the patient had had a traumatic forceps delivery and episiotomy when she was 30 years old. Her mother had a history of unspecified colitis. Rigid sigmoidoscopy and barium enema were normal, but anorectal ultrasound showed defects of the internal and external sphincters. Anterolateral repair of the anal sphincter did not control the symptoms and in September 1997 an end colostomy was performed for her disabling faecal incontinence.

In July 1998, the patient presented with blood and mucus per rectum, and an anterior mucosal prolapse was diagnosed and repaired. However, her symptoms persisted and in October 1998 flexible sigmoidoscopy showed a granular, congested, and oedematous mucosa with contact bleeding throughout the rectosigmoid stump. Histology showed a mixed inflammatory cell infiltrate with distortion of the crypt architecture and cryptitis, and a diagnosis of diversion colitis was made; the rectal symptoms responded quickly to topical steroid enemas.

Four months later, the patient developed increased stomal frequency and bleeding into the stoma bag. Colonoscopic examination of the instream colon, via the colostomy, revealed an active distal colitis with a granular, oedematous, congested, and friable mucosa. Histology showed a mixed inflammatory cell infiltrate in the lamina propria with cryptitis, crypt abscess formation, and a reduction in the number of goblet cells. Ulcerative colitis was diagnosed and treated with oral mesalazine and topical steroid enema per stoma. Symptoms quickly improved and there has been no further bleeding from the rectum or stoma.

Although the histological features of ulcerative colitis and diversion colitis are indistinguishable, the clinical history in this case suggests that ulcerative colitis developed after true diversion colitis. It is possible that this patient's colitic process represented an idiopathic ulcerative colitis, but it seems much more likely that the colostomy with faecal diversion was the initiating factor and that, as speculated in the cases studied by Lim and colleagues, diversion colitis is a risk factor for ulcerative colitis. Hypotheses on the pathogenesis of both diversion colitis and ulcerative colitis should take into account cases such as these.