Ileoneorectal anastomosis: early clinical results of a restorative procedure for ulcerative colitis and familial adenomatous polyposis without formation of an ileoanal pouch

Abstract

Objective: To evaluate a new surgical procedure, ileoneorectal anastomosis (INRA), in patients with ulcerative colitis (UC) and familial adenomatous polyposis (FAP).

Summary background data: Surgical treatment in UC and FAP aims to resect diseased colonic mucosa and restore oroanal continuity. The ileopouch anal anastomosis achieves this but has a 15% to 35% complication rate, a 10% failure rate, and an unpredictable functional outcome. An alternative surgical technique, INRA, has been developed in which the rectal mucosa is replaced by a vascularized ileal mucosa graft.

Methods: Eleven patients underwent an INRA procedure with a temporary diverting ileostomy. Clinical history, repeat endoscopy, histologic examination, and rectal compliance measurements were carried out before and after surgery.

Results: The INRA procedure was technically successful in all patients. Endoscopy showed ingrowth of ileal mucosa in the neorectum, with 100% coverage after 6 weeks. No patient had pelvic sepsis, neorectal-anal or -vaginal fistula, autonomic nerve damage, or fecal incontinence. Neorectal function improved with time. The median 24-hour defecation frequency decreased from 15 (range 9 to 25) to 7 (range 4 to 10) at 11 months follow-up, and the median maximum tolerated volume increased to 157 (range 130 to 225) ml. Anal manometry and electrosensitivity were not affected by the surgery. Histologic biopsy samples after 1 year showed a normal small intestinal mucous membrane, without inflammation or fibrosis.

Conclusion: The combination of a low complication rate and good neorectal function at 1 year is a substantial improvement that justifies extension of the clinical application in patients with UC and FAP.

Figure 1. Histologic examination of the neorectal resection specimen of a pig, showing a composite bowel wall with (1) perirectal fatty tissue, (2) muscularis propria of the rectum, (3) small intestinal submucosa and mucosa (hematoxylin and eosin stain).

Figure 2. Excision specimen 1 year after an INRA procedure in a pig, showing a normal ileorectal anastomosis with a compliant neorectum.

Figure 3. Rectal mucosectomy, surgical technique. Dentate line (1), circular and longitudinal rectal muscle layers (2), rectal mucosa (3).

Figure 4. Histology of rectal mucosectomy specimens. (A) A patient with UC. The dissection plane is invariably located beneath the submucosa, which is thickened, whereas the mucosa has been destroyed by the long-lasting colitis (hematoxylin & eosin stain, ×250). (B) A patient with FAP. The dissection plane is invariably located beneath the submucosa, and the polyps are completely removed.

Figure 5. Ileal mucosal sling formation, surgical technique. Ileal mucosa (1), vessels penetrating ileal wall (2), meshlike graft incisions (3).

Figure 6. The ileoneorectal anastomosis: anastomosis (3) between the ileum and proximal rectal muscular wall with the ileal mucosa sling (2) plugged on the denuded rectal muscle. The distal end is sutured to the dentate line (1). The central vascular pedicle is shown, as well as the mucosal fenestrations for drainage and increased luminal diameter.

Figure 7. Defecation frequency of INRA patients over time.

Figure 8. Maximum tolerated volume of the neorectum and its development over time after the INRA procedure.

Figure 9. Histology of the neorectal mucosa during follow-up after the INRA procedure. (1) At 1 week: normal length of the villi with a dense mixed inflammatory infiltrate and ulcerations. (2) At 6 weeks: subtotal villous atrophy with disappearance of the goblet cells and brush border. (3) At 3 months: partial villous atrophy with regeneration of goblet cells, enterocytes, and brush border, disappearance of the inflammatory infiltrate. (4) After 6 to 12 months: normal small intestinal mucous membrane without inflammatory infiltrate, normal length of villi, intact goblet cell population and brush border (hematoxylin and eosin stain, ×250).