Ileo-neorectal anastomosis: medium- and long-term follow-up of 37 patients

BACKGROUND: Ileo-neorectal anastomosis (INRA), an alternative restorative procedure, was developed to reduce the pouch-related complication rate with an (at least) equal functional result. METHODS: For this surgical outcome, data of all INRA patients, including bowel function and complications, were prospectively recorded. The reservoir capacity was determined repeatedly by physiologic tests. The anal sphincter complex was assessed by manometry and ultrasound examination. Evaluation of the neorectal mucosa was performed by endoscopy. RESULTS: An INRA procedure was carried out in 39/53 selected patients (47 ulcerative colitis and 6 familial adenomatous polyposis). Fourteen UC cases were converted to ileal pouch anal anastomosis or proctectomy only, because of impossibility to completely remove the rectal mucosa or short of length of the rectal stump. The median operation time for INRA was 323 min (range 240-518), with 1,400 ml blood loss (400-4,500). The reservoirs were permanently defunctioned in 2 patients--one because of reclassification into Crohn's disease, and one with pouchitis refractory to medical treatment. In 18 out of 37 cases, web-like stenoses occurred at the mucosa-anal level, which were treated by single (9) or repeated (5) dilatation or surgical stenoplasty (2). No pouch-related complications like pelvic sepsis, fistula or sexual dysfunction occurred. Thirteen patients had episodes of 'pouchitis', successfully treated with antibiotics, and 7 other cases, with functioning reservoirs, also had proximal 'non-specific' (i.e. no histological criteria of Crohn's disease found) small bowel inflammation. The median bowel frequency decreased from 15x/24 h initially to 7x/24 h at 2 years. Continence was perfect in 24/37 cases. Twelve out of 37 cases had occasional nocturnal soiling and passive nocturnal fecal incontinence was reported by 2/37 patients. The neorectal compliance volume recovered from 12.5 ml kPa after subtotal colectomy and 11 ml/kPa at 6 months after INRA to a neorectal compliance of 24 ml/kPa at 2 years' follow-up (p < 0.002; Wilcoxon signed rank test). CONCLUSION: The INRA procedure shows a low complication rate and reasonable functional results, there was however a considerable conversion rate in these first 53 cases and a high incidence of reclassification to CD.