Share this article on:

Postsurgical Recurrences in Crohn's Disease: Why, When and How to Prevent Them

Mamula, Petar; Baldassano, Robert N.

Section Editor(s): Braegger, Christian M.D., ESPGHAN; Büller, Hans M.D., ESPGHAN; Thomas, Adrian M.D., ESPGHAN; Bishop, Warren P. M.D., ESPGHAN; Haber, Barbara Anne M.D., ESPGHAN; Lichtman, Steven N. M.D., ESPGHAN; Shneider, Benjamin L. M.D., ESPGHAN

Journal of Pediatric Gastroenterology & Nutrition: May 2000 - Volume 30 - Issue 5 - p 557
Selected Summary

Center for Pediatric Inflammatory Bowel Disease

Division of Gastroenterology and Nutrition

The Children's Hospital of Philadelphia

University of Pennsylvania Medical School

Philadelphia, Pennsylvania, U.S.A.

Side-to-side Stapled Anastomosis May Delay Recurrence in Crohn's Disease. Hashemi M, Novell RN, Lewis AAM. Dis Colon Rectum 1998;41:1293–6.

Back to Top | Article Outline


From 1984 through 1997, 69 patients underwent resection and primary anastomosis for Crohn's disease (17 jejunal or ileal, 72 ileocolic, and 7 colocolic). Of these, 42 patients underwent functional stapled end-to-end (anatomic side-to-side) anastomosis, and 27 patients underwent end-to-end sutured anastomosis. The primary outcome studied for surgical treatment was symptomatic recurrence requiring reoperation. In the group that underwent stapled anastomosis, one patient (2%) required reoperation for recurrent symptoms at 46 months, and four (8%) had complications (fistula, abscess, wound infection, and ileus). The median follow-up for this group was 23 months (range: 1–66 months). In the group with sutured anastomosis, 14 patients (43%) required 15 further resections for symptomatic recurrence at a median of 46 months. Complication occurred in six patients (17%): two fistulas, two anastomotic leaks, one stricture, and one pulmonary embolus. The median follow-up period was 52 months (range: 1–125 months). The authors concluded that functional end-to-end stapled anastomosis may be associated with fewer complications than sutured anastomosis and may delay reoperation for symptomatic recurrence.

Back to Top | Article Outline


The surgery for Crohn's disease has undergone numerous changes, and it has significantly improved since 1932 when the classic study published by Crohn et al. described 14 patients with this disease, then called “regional ileitis.” The initial high mortality rate has decreased with advances in perioperative care, fluid management, anesthesia, nutritional therapy, and total parenteral nutrition (Dis Colon Rectum 1999;42:979–88). Most patients with Crohn's disease must have surgery at some time in their lives. This, together with the early recognition of inability to cure the disease surgically has prompted many physicians to look into the causes and patterns of postsurgical recurrence. Most of the published studies are retrospective, rather than prospective, some using crude rates of recurrence as opposed to more appropriate actuarial and life-table methods. This has contributed to the conflicting data about the factors affecting risk of recurrence. At the same time, the very definition of recurrence varies between different studies. Symptomatic recurrence based on radiologic or endoscopic evidence is used in some studies, whereas surgical recurrence as a clear outcome end point, either at the site of the anastomosis or anywhere else in the intestine, is used in others (Baillieres Clin Gastroenterol 1998;12:167–77).

Crohn's disease can affect any part of the intestine, and the lesions are often patchy and occasionally very extensive. The usual surgical procedures for Crohn's disease include segmental resection and stricturoplasty. The resection and types of anastomoses after resection will be reviewed in more detail later. Stricturoplasty is performed to preserve as much intestine as possible and prevent short bowel syndrome. In the past, bypass of the intestine was tried, but it was abandoned because of stump “blow out,” septic complications, abscesses, high recurrence rates, and reports of carcinoma developing in the bypassed segments (Ann Surg 1989;210:621–5). Stricturoplasty, initially used by Indian surgeons for multiple tuberculous strictures, has been successfully applied in patients with Crohn's disease and was first reported by Lee and Papaioannou (Ann R Coll Surg 1982;64:229–33). Several stricturoplasty techniques have evolved over time, depending on the length of the narrowed segment (Dis Colon Rectum 1993;36:71–2). A study by Fazio et al. (Dis Colon Rectum 1993;36:355–61) followed up on the outcome of 452 stricturoplasties in 116 patients for a median of 3 years (range: 6 months to 7 years) and concluded that stricturoplasty has morbidity, mortality, and complication rates comparable to resection. The authors stated that it should not be performed in case of phlegmon, abscess, perforation, or fistula, but it can serve as a useful adjunct to other current surgical options. They also discussed the concern about development of carcinoma in the segment of bowel affected by the long-standing inflammation. So far, there appears to be no increased incidence of carcinoma after stricturoplasty. In a study of small bowel Crohn's disease (Br J Surg 1989;76:335–8) recurrence rates (defined by site-specific, operation-free interval) after stricturoplasty and resection were similar.

In the past, a large number of studies were performed to evaluate the factors influencing the recurrence rate after resection for Crohn's disease. These risk factors can be categorized into several groups:

• Patient characteristics: age, gender, tobacco use, and age at diagnosis

• Disease pattern: location of disease, extent of disease, duration of disease, extraintestinal manifestations, presence of granulomas, laboratory parameters, and number of involved sites

• Medical treatment: blood transfusions, mesalamine (5-ASA), metronidazole, budesonide, and 6-mercaptopurine

• Surgical factors: indication for operation, extent of resection, microscopic involvement at the resection margin, type of anastomosis, fecal stream diversion, and use of ileocolic nipple valve anastomosis

An excellent article by Borley et al. (Br J Surg 1997;84:1493–502) reviewed more than 100 studies focusing on assessment of recurrence risk. The recurrence rates defined by reoperation were 11% to 27% at 5 years, 22% to 44% at 10 years, 32% to 51% at 15 years, and 46% to 55% at 20 years. The recurrence rates defined by symptomatic, endoscopic and radiologic recurrence, in addition to reoperation, were obviously higher: 17% to 55% at 5 years, 32% to 76% at 10 years, 45% to 79% at 15 years, 72% at 20 years and, 79% at 25 years.

In the first two groups of risk factors (except for the use of tobacco, which increases the risk of recurrence) the results are conflicting, and there appear to be no consistent statistically significant data supporting either increased or decreased risk of recurrence after surgery. Although the scope of this summary prevents us from going into the details of particular studies, we will mention several studies involving the risk factors in the third and fourth groups.

Medical treatment has been shown to reduce the risk of recurrence: Metronidazole reduces endoscopic recurrence at 1 year (Gastroenterology 1995;108:1617–21), and 5-ASA reduces endoscopic recurrence (postoperative colonoscopy performed 6, 12, 24, and 36 months after surgery) in combination with side-to-side or end-to-side anastomosis at 3 years (Dis Colon Rectum 1996;39:335–41). Perioperative treatment with 6-mercaptopurine may be beneficial in reducing recurrence in children (J Pediatr Gastroenterol Nutr 1997;25:93–7). However, all these studies had rather short-term follow-up, and the results must therefore be confirmed in long-term studies. Also of interest is the effect of newly introduced anticytokine therapies on the risk of recurrence.

Surgical factors have also been extensively studied. In 1996 Fazio et al. published a well-designed study (Ann Surg 1996;224:563–73) of 131 patients followed up for a median of 55 months, and randomized depending on width of resection and microscopic involvement. The study concluded that “bigger is not better”: Recurrence rates are not influenced by microscopic involvement at the margin of resection, and frozen sections at the time of the surgery are not necessary.

Rutgeerts has performed elegant studies introducing a widely accepted endoscopic scoring system (Gastroenterology 1990;99:956–63). He has also shown that most endoscopic recurrences in the preanastomotic area preceding clinical symptoms occur before 1 year (72%) and does not significantly differ from the 3-and 10-year recurrence rates (79 and 77%, respectively). This indicates that such features occur by one year or not at all (Gut 1984;25:665–72). In 1991 he investigated the role of fecal stream in endoscopic recurrence. In all 5 patients showing no histologic inflammation six months after the diversion, there was prompt return of severe inflammation after the reanastomosis was performed. This supported the role of a combination of the fecal stream and coloileal reflux in post surgical recurrence. The new approach, ileocolic nipple valve anastomosis (Dis Colon Rectum 1990:33:987–90), was developed to treat the coloileal reflux but was associated with technical difficulties (deterioration of the valve with time).

Going back to the initial interest of this summary, the anastomotic technique has been the focus of research for the past 10 years. There are several anatomic types of anastomoses: end-to-end, end-to-side, side-to-end, and side-to-side, and they can be either hand-sewn or stapled. Stapled anastomoses are performed with a circular stapler in an anatomic end-to-end fashion or a linear stapler in a functional end-to-end, but actual anatomic side-to-side, fashion. The previously mentioned study regarding the reduction of postoperative recurrence by the use of mesalamine (5-ASA) also showed decreased risk of recurrence in the mesalamine-treated group with side-to-end, side-to-side, and end-to-side anastomoses groups when compared with the end-to-end anastomosis group. On the other hand, a prospective study (Br J Surg 1992;215:546–51) with a mean 4-year follow-up of 86 patients randomized into end-to-end and side-to-end sutured anastomotic group did not show a significant difference.

The weaknesses of the reviewed study are the retrospective nature of the data collection, with an unclear definition of recurrence and unequal group size; the small number of patients; and the relatively short follow-up. However, this study raises interesting questions. Is it possible that the wider lumen achieved by stapled side-to-side anastomosis plays a role in the decreased recurrence by a decrease in the luminal pressure and ischemia? Is the circulation better preserved with this technique? The technique appears to be easier to perform and faster than suturing, with less anastomotic leaks and complications, and it involves a shorter hospital stay. A prospective randomized study on 68 patients (Dig Surg 1998;15:679–82) showed statistically significant shorter operation times for the group that underwent stapled anastomosis with no difference in anastomotic dehiscence and 7.5% recurrence at the anastomosis site in this group compared with a 16.7% recurrence rate in the group with sutured anastomosis (median follow up of 4.5 years). In adult patients undergoing surgery for Crohn's disease, stapling is used on regular basis, when technically possible, and has rendered hand-sewn anastomoses almost obsolete.

In pediatric patients, the surgical recurrence is also a subject of interest in several retrospective studies (Gut 1998;43:634–8;Am Surg 1998;41:1577–80;Gut 1991;32:491–5;Br J Surg 1990;77:891–4). The true incidence of recurrence is unknown, and different types of anastomoses have not been studied. Different methods, types of surgery, and a mixed population of patients make it difficult to compare the data from these studies in meta-analysis fashion. However, several observations are common. Children with extensive disease have higher risk of recurrence, children with early recurrence have a tendency to have longer lasting disease, significant growth and in some cases catch-up growth is achieved after surgery (especially if the surgery is performed before puberty or in patients with delayed bone age), and quality of life is significantly improved after surgery. In one adult trial, authors went so far as to randomize patients to surgery strictly to improve quality of life (Am J Surg 1997;174:339–41). In this group, improvement of life quality was noted with shorter hospital stays, no anastomotic leaks, no late complications, and only one reoperation in 21 patients (mean follow-up of 50 months).

In conclusion, side-to-side stapled anastomosis may be the procedure of choice for bowel resection in pediatric patients with Crohn's disease. Larger prospective and well-randomized studies with a longer follow-up are needed. Early perioperative medical treatment could contribute to the reduction of postoperative recurrence and also warrants further investigation. Traditionally, surgery is the last resort for patients with Crohn's disease, unless it is performed emergently. Early involvement of pediatric surgeons in the treatment process, with a more liberal approach in carefully selecting pediatric patients with localized Crohn's disease could have a beneficial effect, resulting in improved growth, better disease control, and better quality of life.

© 2000 Lippincott Williams & Wilkins, Inc.