Background: This study was performed to assess if using endoscopic ultrasound (EUS) to assess and guide combination medical and surgical therapy during fistula healing will lead to a high rate of durable fistula closure and a low or absent incidence of perianal abscess formation in patients with Crohn's perianal fistulas.
Methods: This is a retrospective analysis of 21 patients who presented with a symptomatic Crohn's perianal fistula. Patients were enrolled in a clinical practice protocol of serial EUS exams. All patients underwent a baseline rectal EUS and were placed on maximal medical treatment with 6-mercaptopurine (6-MP) or azathioprine, Cipro, and infliximab (5 mg/kg at 0, 2, and 6 wk and then every 8 wk). Patients were also assessed at baseline by a colorectal surgeon who was aware of the EUS findings. Seton placement and incision and drainage were performed when appropriate. Serial EUS examinations were performed, and the findings were used to guide therapy (i.e., the presence of fistula healing on EUS was used to guide seton removal, discontinuation of infliximab, and Cipro).
Results: In the 21 patients enrolled, the median duration of active perianal symptoms was 9 wks (1-36). 10 patients (48%) had previous perianal surgery and 5 (24%) had received infliximab previously. The fistulas treated included 8 trans-sphincteric, 2 superficial, 3 recto-vaginal, and 7 with multiple and horseshoe fistulas. 13 patients (62%) had associated abscesses at presentation. Eighteen of 21 patients (86%) had complete cessation of drainage initially. Median time to cessation of drainage was 10.6 weeks (range, 4-32 wk). Sixteen of 21 patients (76%) maintained long-term cessation of drainage. The median length of follow-up was 68 weeks (range, 35-101 wk). No abscess developed during treatment in any patient. EUS evidence of persistent fistula activity was seen in 10 patients (48%). Of the 11 patients (52%) in whom EUS showed no persistent fistula activity, 7 (64%) have maintained fistula closure off of infliximab and Cipro. Median duration from last infliximab infusion was 47 weeks (range, 20-80 wk). The remaining 4 patients continued infliximab to maintain remission of their luminal disease. Only 1 patient with a horseshoe fistula showed complete healing on EUS.
Conclusion: In conclusion, using EUS to guide therapy for Crohn's perianal fistulas with infliximab, an immunosuppressive, and an antibiotic is associated with a high short and long-term fistula response rate. EUS may identify a subset of patients who can discontinue infliximab without recurrence of fistula drainage.