Fistula formation is common in patients with Crohn’s disease. Fecal drainage from fistulas is a distressing symptom. Surgical diversion of the colon to a stoma often leads to fistula closure, but this invasive option also produces substantial patient distress. Tumor necrosis factor inhibitors, such as infliximab, have been used with some success in patients with inflammatory bowel disease. Sands and colleagues investigated the use of infliximab for closure of fistula tracts in patients with Crohn’s disease.

This multinational trial was underwritten by Centocor, the manufacturer of infliximab. Eligible patients were adults with at least one perianal or enterocutaneous fistula that had been draining for at least three months. Initial enrollment included 306 patients who were given intravenous infusions of infliximab in a dosage of 5 mg per kg at weeks zero, 2, and 6. Use of other standard medications for Crohn’s disease, such as 5-aminosalicylates, oral steroids, or azathioprine, continued during the trial.

Adverse events, withdrawn consent, lack of efficacy, or noncompliance led to exclusion of 24 patients after the initial induction phase. The remaining 282 participants were re-evaluated at weeks 10 and 14, and those who maintained a reduction of at least 50 percent in the number of draining fistulas at both visits were classified as responders (195 patients, 69 percent). Responders were randomized to maintenance therapy with infliximab or placebo every eight weeks until 54 weeks of follow-up. Loss of response during maintenance treatment was defined as reappearance of a draining fistula, need for an additional treatment or change to other disease-related medications, need for surgical treatment related to Crohn’s disease, or perceived lack of efficacy.

The time to loss of response during maintenance therapy was significantly longer in patients randomized to infliximab (40 weeks) than in patients receiving placebo (14 weeks). At 54 weeks of follow-up, a complete response (no draining fistulas) was sustained more often in those still receiving infliximab (36 percent) than in those assigned to placebo (19 percent). Multivariate analysis did not identify any patient characteristics that predicted a sustained response in patients randomized to infliximab maintenance therapy.

Patients without an initial response to infliximab also were studied to see if further active treatment would eventually increase response rates. No significant improvement occurred with prolonged infliximab treatment, compared with placebo, in initial nonresponders.

Antibody formation to infliximab occurred in 32 percent of patients on active treatment. Presence of antibodies did not decrease the efficacy of infliximab, but it did increase the likelihood of an infusion reaction. Patients who also were taking oral steroids and immunomodulators were less likely to develop antibodies to infliximab. Only one infusion reaction was considered serious. Two opportunistic infections occurred in patients receiving active treatment (i.e., cytomegalovirus infection and cutaneous nocardia infection).

The authors conclude that more than one half of patients with Crohn’s disease and draining fistulas respond to initial treatment with infliximab, and that maintenance therapy every eight weeks over one year approximately doubles the likelihood of maintaining fistula closure.