Over the past 10 years ileal pouch-anal anastomosis has become the operation of choice for most patients with ulcerative colitis. Although pouch surgery in ulcerative colitis does have a moderate complication rate, so do the alternatives of proctocolectomy and ileostomy.1 The very real advantage of avoiding a stoma along with the inherent curability of ulcerative colitis by excisional surgery help to explain the attractiveness of this operation.

In Crohn’s disease the situation is notably different as the condition is not curable by surgery, and all operations in Crohn’s disease are followed by a higher complication rate and fairly frequent recurrence. Indeed, recurrence is often seen when a permanent stoma is avoided and an anastomosis constructed.

Patients with Crohn’s disease are just as averse to a stoma as those with ulcerative colitis, and so for very good reasons stomas are often avoided even though both patient and doctor are well aware that this puts the patient at increased risk of recurrence and further surgery. Thus, it would be standard practice to offer a patient with terminal ileal Crohn’s disease right hemicolectomy and some patients with large bowel Crohn’s disease colectomy and ileorectal anastomosis. As a consequence of restorative surgery, clinical recurrence is seen within 10 years in 50–80% of these patients.2  3 Figure 1 makes the point quite dramatically: the upper curve represents recurrence rates after colectomy and ileorectal anastomosis, whereas the lower shows how infrequent it can be when a permanent stoma is used instead of an anastomosis. Yet despite this high rate of recurrence, restorative surgery in these circumstances is generally accepted by the surgical and medical community. Does such a policy result in undue loss of small intestine? One study of 82 patients who underwent colectomy and ileorectal anastomosis found that 46 (56%) subsequently underwent revisional surgery with an average cumulative loss of small intestine of 44 cm in the fixed state.4 Although these losses at first glance might seem to be quite high, short bowel syndrome is an uncommon problem in patients with Crohn’s disease and there is as yet no sign of waning enthusiasm for either right hemicolectomy or colectomy and ileorectal anastomosis because of this anxiety.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Cumulative recurrence rate after total proctocolectomy (TPC) or colectomy and ileorectal anastomosis (IRA) for Crohn’s disease at St Mark’s Hospital 1947–87.3 Reproduced with permission.

Ileal pouch-anal anastomosis has, for some time, been considered to be quite inappropriate in patients with Crohn’s disease. To some extent this has arisen owing to poor results when compared with pouches in ulcerative colitis and familial adenomatous polyposis, but to some extent it has also arisen because of a tendency to compare apples with oranges.

Kock pouches were originally constructed for both Crohn’s disease and ulcerative colitis. Nils Kock selected patients with disease limited to the colon, in many of whom the preoperative diagnosis had in fact been ulcerative colitis but who later turned out to have Crohn’s disease.5 Of the 280 patients treated with Koch pouches, 49 had Crohn’s disease. One patient died postoperatively and 27% developed complications as inpatients. In eight (16%) the pouch had to be removed some time later, in four because of recurrent disease. Of the remaining 40 patients, another 17 developed recurrent disease, six in the ileal segment proximal to the reservoir, five in the reservoir alone and six at both sites. Fourteen of these patients had surgical removal of their recurrent Crohn’s disease. Overall, 37 of these 40 patients had continent ileostomies and only three needed to wear an ileostomy appliance.

For restorative operations in patients with Crohn’s disease these results were really very good. But Kock, comparing apples with oranges, was worried because the in-hospital complication rate was double that for patients with colitis and excision of the pouch had been necessary in 16% compared with only 2% in ulcerative colitis. He concluded that, “this procedure should be performed in patients with Crohn’s disease only exceptionally”.

In another report6 15 of the 168 patients studied had Crohn’s disease, eight (53%) of whom subsequently had to have their pouch removed. However, 11 of the 15 patients had been referred after proctocolectomy or construction of a continent ileostomy which had been performed elsewhere, so there may well have been a selection bias for those patients with more severe disease. The authors6 were generally against construction of Kock pouches in Crohn’s disease, particularly in small bowel Crohn’s disease, but did “concede that it may have a role in selected patients with colonic Crohn’s disease, under appropriately controlled conditions”.

Three papers in the early 1990s described a combined experience of 44 patients who had undergone ileo-anal pouch surgery for Crohn’s disease.7-9 These patients represented about 6% of the total number performed in those institutions over the previous two to 10 year period. Fifteen (34%) subsequently had their pouch excised or permanently defunctioned.

All of these papers report some cases diagnosed as having Crohn’s disease at the time or shortly after surgery and others being diagnosed some time later. These late diagnoses almost always represent a biased group selected because of complications, as may also have been the case in Fazio and Church’s study.6 Deutschet al described five patients diagnosed with Crohn’s disease immediately after surgery on the basis of examination of the resected rectum. Three of the patients had a continuing functional pouch. Of a further four diagnosed, on average two and a half years later, two had the pouch excised, one had a pouch in situ with a pouch-vaginal fistula, and another had problematic anal fissures. Clearly, this latter group represents patients with late Crohn’s disease with serious clinical problems and ignores a potential group of unknown size with unsuspected Crohn’s disease who did not have problems.

It is against this background that the paper by Paniset al in theLancet should be considered.10This described 31 patients apparently with Crohn’s disease who had undergone pouch surgery with comparable short and longer term results to a group of 71 patients with ulcerative colitis similarly operated upon over the same period. The authors were at particular pains to point out that the patients selected had neither anoperineal nor small bowel Crohn’s disease, and concluded that pouch surgery could be recommended in some patients with Crohn’s disease.

The key issue is whether the patients really did have Crohn’s disease, or whether some of them had what other authors term indeterminate colitis, where results for pouch surgery have been shown to be equivalent to those for ulcerative colitis.11 Only 18 of their patients were considered preoperatively to have Crohn’s disease, whereas 13 were at that time classified as having indeterminate colitis, only being reclassified as Crohn’s disease after resection. Overall, only eight (26%) patients had epithelioid granulomas, two had “chronic ileitis” and the remaining 21 patients were classified as having Crohn’s disease because they had at least four of the following features: skip segments on the resected bowel (n=20), lymphoid aggregates (n=19), crypt abscesses (n=19), fissuring ulceration (n=4), submucosal fibrosis (n=21), and mucus secretion (n=17). Were some of these cases really cases of indeterminate colitis? The authors must have known that their paper would stir up considerable controversy, so it was a missed opportunity not to have had an external panel of pathologists agree that these really were all cases of Crohn’s disease.

As it is, we really do not know the place of pouch surgery in Crohn’s disease. We do know, however, that operating on patients with Crohn’s disease is not the same as operating on patients with ulcerative colitis as complications and recurrence are frequent, especially when an anastomosis is used. Nevertheless, restorative operations such as right hemicolectomy and colectomy and ileorectal anastomosis are part of accepted surgical practice in patients with Crohn’s disease, short bowel syndrome is uncommon and lengths of small bowel lost through this policy are not that different to those used to construct an ileo-anal pouch. The results of pouch surgery do not seem any worse than for any other restorative operation in Crohn’s disease. There are fears that inordinate lengths of small bowel might be lost if pouches were used more frequently in these patients. But pouches, like stricturoplasty sites, may just as likely be less prone to Crohn’s disease than the afferent limb of neoterminal ileum after right hemicolectomy or colectomy and ileorectal anastomosis. If Kock’s experience is anything to go by, then six of 17 recurrences were completely outside the pouch and 14 were able to have their diseased gut removed and the pouch preserved.5

We should stop comparing pouch surgery in Crohn’s disease with pouch surgery in completely different conditions, such as ulcerative colitis or familial adenomatous polyposis. Rather, we should consider it in the context of other restorative operations in Crohn’s disease: something not to be dismissed out of hand in those few patients without small bowel or anal disease, so long as they have an experienced colorectal surgeon on hand.