Fistula and abscess

Several different types of fistula can develop including enteroenteric, enterovesical, enterovaginal, enterocutaneous, perianal, and perirectal. Similarly, abscesses can occur in various sites: enteroparietal, interloop, intramesenteric, and retroperitoneal. Fistulae rarely heal with corticosteroid therapy. However, immunomodulator therapy will promote fistula closure in 30–40% of cases, especially if distal obstruction and active disease are absent. Abscesses can be managed by initial percutaneous drainage and delayed resection or primary resection without preoperative drainage. Jawhari reported his experience with this latter approach of primary resection without initial drainage in 28 Crohn's disease patients with intra-abdominal abscesses; 9 (32%) experienced a recurrent abscess, 4 (14%) developed an enterocutaneous fistula, and 6 (21%) now suffer from short bowel syndrome (Jawhari 1998). Experience suggests that when an abscess is not amenable to percutaneous drainage, appropriate treatment requires laparotomy with thorough abscess drainage, resection of the disease-affected intestine, and the liberal use of fecal diversion. If abscess management risks significant loss of non-diseased bowel, resection is delayed for several months following adequate surgical drainage with proximal diversion (grade C).

Obstruction

Bowel obstruction can be acute or chronic and arise from single or multiple sites of stricturing. Although obstructive symptoms might improve with high-dose corticosteroids, the response is often temporary and symptoms typically recur as the medication is tapered. Moreover, high-grade obstructive lesions usually do not respond to medical therapy and early operative intervention is recommended before symptoms worsen or perforation occurs.

Perforation and hemorrhage

Free perforation, although rare, usually occurs during an acute exacerbation of chronic disease, particularly in the presence of distal obstruction, or during a bout of toxic colitis when transmural ulceration has developed. The transmural ulcers of Crohn's disease more typically result in formation of inflammatory adhesions between the diseased segment and surrounding structures that seal most perforations. However, the resultant abscess may subsequently rupture, spill its contents, and create a communication between the bowel lumen and the peritoneal cavity. Alternatively, the deep ulcers can erode into moderate-sized vessels of the mucosa or submucosa leading to massive intestinal hemorrhage. Emergent operative treatment should be individualized but must be considered in patients with: no other obvious sources of bleeding; life-threatening hemorrhage; failure to attain stabilization after initial transfusion with 4 to 6 units of packed red blood cells; significant rebleed during hospitalization; or coexisting indication for intestinal resection (grade C).

Internal bypass

Intestinal resection was the procedure of choice when surgery for Crohn's disease was first employed. Early mortality rates associated with resection approached 20%, and led Berg to perform a staged procedure for patients with terminal ileal Crohn's disease, especially those complicated by localized intraperitoneal sepsis. In the first stage, the small bowel was transected proximal to the diseased ileum, the distal ileal limb was oversewn, and an anastomosis was constructed between the proximal bowel limb and the transverse colon. During the second stage, the bypassed segment was resected. In many instances disease of the bypassed segment had resolved leading clinicians to conclude the second stage operation was unnecessary. However, as the risk of malignancy in the bypassed segment became recognized, this staged approach to ileocecal Crohn's disease was largely abandoned.

Bypass operations are still considered reasonable or desirable options for select cases. For instance, continuity bypass is a preferred method of management of symptomatic gastroduodenal Crohn's disease that is refractory to medical treatment. Similarly, in patients with severe ileocecal disease complicated by an extensive, contained perforation adherent to the common iliac vessels or cava, an exclusion ileotransverse colon bypass procedure may be performed. The proximal end of the excluded ileal segment should be exteriorized as a small mucus fistula to vent mucosal secretions that could cause blow-out of the ileal stump. This is usually a successful procedure, at least initially, and the patient may remain asymptomatic for months or years. Elective resection of the bypassed segment is recommended approximately 6 months later, even in the asymptomatic patient. The only possible exception to “early” resection of the bypassed segment is if the disease duration is short (e.g., < 5 years) and the risk of cancer consequently low.

External bypass

Ileostomy alone is used infrequently in current times. Even for free perforation of the small bowel, resection of the perforated segment with exteriorization of the proximal bowel as an end stoma is standard practice. The blow-hole colostomy with ileostomy for toxic colitis and toxic megacolon is a bypass procedure utilized during the 1960s and 1970s. Fortunately, the procedure is infrequently necessary these days, in part because of an overall decline in the incidence of toxic colitis. The principal reason, however, is that toxic colitis rarely worsens to the catastrophic variety of toxic megacolon, for which the blow-hole procedure is most valuable. Physician awareness and access to medical care have improved such that far fewer patients progress to this most severe degree.

Resection margins

Historically, surgeons debated whether the small bowel resection margin needed to be microscopically-free of inflammation and, if not, how extensive the macroscopically-free margin needed to measure. While some early studies reported cumulative recurrence rates were greater for patients with microscopic evidence of disease at the resection margins, most recent series have shown no relationship between microscopic inflammation at resection margins and recurrence rates (grade C).

Recently, a randomized, controlled trial of patients (n = 152) undergoing resection of small bowel Crohn's disease randomized patients to resection margins that were macroscopically clear of disease for either 2 cm (limited group; n = 82) or 12 cm (extended group; n = 70); 131 patients underwent anastomoses (Fazio 1996). After a median follow-up of 56 months, disease recurred (reoperation for recurrence) in 29 patients; 25% of patients in the limited group and 18% in the extended group. Patients were also studied to assess the effect of microscopic changes at the macroscopically normal line of resection. The margins were characterized as normal, nonspecific changes, suggestive but not diagnostic of Crohn's disease, or diagnostic for Crohn's disease. Recurrence rates were not significantly different between the 4 categories.

The presence of residual microscopic Crohn's disease at the resection margins does not increase recurrence rates significantly, compared with normal margins, and extended resection margins confer no advantage to patients in reducing cumulative recurrence rates. Therefore, most surgeons favor conservative resection margins dividing the intestine approximately 2 to 5 cm proximal to overt disease (grade B and C).

Anastomotic technique

The neo-terminal ileum tends to be the usual site of recurrence following ileocolic resection for Crohn's disease. Opinions vary as to the preferred technique of anastomosis, whether handsewn or stapled, and which anastomotic configuration is least associated with recurrent inflammation. The presence of a suture line, the size of the anastomosis, and reflux of colonic contents into the ileum have been suggested as potential causative factors. In an effort to resolve the controversy, Cameron randomized Crohn's disease patients undergoing first-time ileocolic resection to either an end-to-end or side-to-end anastomosis (Cameron 1992). The clinical recurrence rate did not significantly differ between the two groups. In a similar study of patients undergoing resection for primary or recurrent ileal disease, patients treated with a side-to-side anastomosis were compared to a historical cohort of patients with side-to-end anastomoses (Scott 1995). The cumulative recurrence rates for both groups were nearly identical. While many common techniques exist, inverting, stapled anastomoses of any standard configuration are generally preferred because of safety, ease, and comparable recurrence rates (grade B and C).

Laparoscopy

As experience with laparoscopy for Crohn's disease increases, the role of this approach broadens as the contraindications lessen and benefits emerge. Singh reported 24 patients who underwent laparoscopy because of Crohn's disease, only 2 required laparotomy with the remainder undergoing a purely laparoscopic or laparoscopic-assisted procedure (Singh 1998). Although there was little difference in the median stay for patients treated laparoscopically or by laparotomy, it seemed the extent or severity of the disease process, rather than the approach used, influenced the length of the stay. Similarly, Wu reported his Crohn's disease experience with laparoscopic-assisted ileocolic resections attempted in 46 patients and conventional open ileocolic resections performed in another 70 individuals (Wu 1997). He found that co-morbid preoperative conditions such as abscess, phlegmon, or recurrent disease at a previous ileocolic anastomosis were not contraindications to a successful laparoscopic-assisted ileocolic resection. Moreover, the operative morbidity and length of stay were significantly less for the group undergoing laparoscopic-assisted resection.