6.03 Enterocutaneous Fistula Treatment (ECF) with Fibrin Glue Injection – Does it work?

J. S. Merkow1, A. Paniccia1, M. Gipson1, J. Durham1, L. Wilson1, J. Vogel1  1University Of Colorado Denver,Aurora, CO, USA

Introduction:  ECF is a challenging problem that often requires complex surgery for resolution.  A reliable nonsurgical cure for ECF would be welcome.  Fibrin glue treatment of anal fistula tract is an accepted and commonly used therapy with several studies demonstrating its efficacy.  The value of fibrin glue treatment for ECF has been relatively poorly evaluated.  The PURPOSE of this study was to describe our experience with fibrin glue therapy of ECF and determine characteristics associated with therapeutic success or failure. 

Methods:  Clinical data was extracted from a chart review of patients with ECF who underwent fibrin glue injection at the University of Colorado Hospital from 2003 to 2014.  Eligible patients had clinical and radiologic evidence of a fistula originating from the small or large intestine, between the ligament of Trietz and the upper rectum. Low output vs. high output fistula were <200 cc/day vs. >200cc/day.  Complete success was defined as 100% closure of the ECF.  Partial success was defined as decreased output reported by patient or physician after the gluing and during follow up.  Demographic and clinical data were recorded. 

Results: There were 38 patients with a median age of 55 years (IQR 45-62) with 22 (58%) male and 16 (42%) female patients. The median BMI was 24 (IQR 22-29), albumin 2.5 g/dL (IQR 1.9-2.9), and hemoglobin 9.9 g/dL (IQR 8.6-11).  Average ECF duration was 5 months (IQR 2-19).  20 (52%) patients had low output fistulas compared to 1 (3%) with a high output fistula. ECF origin was 17 (45%) small bowel, 16 (42%) colon, and 1(13%) rectum.  Etiology of fistula formation was iatrogenic in 21 (55%) patients, of which 18 (47%) occurred after a surgical procedure.  Other causes of fistula formation were infection/abscess (9, 24%) and pancreatitis (8, 21%).  The median number of gluing procedures was 1 (IQR 1-2).  Median follow-up after glue therapy was 17 months (IQR 5-52). Complete success occurred in 12 (34%) patients and partial success in 6 (17%).  Complete and partial therapeutic success was 23% and 18% for small bowel and 50% and 12% for colorectal fistula.  Of patients with complete success, 83% closed within 1 week of therapy and 75% required only a single fibrin treatment.  There were no complications associated with the use of fibrin therapy.  Analysis of factors including immunosuppression, albumin, obesity, IBD, cancer, output volume, repeated gluing procedures and fistula duration did not predict successful fibrin glue therapy.  

Conclusion: Fibrin glue therapy was a complete success in one-third of patients with an ECF that originated from the small or large bowel.  There were no complications associated with this therapy.  Further studies on a larger sample will be required to identify factors associated with successful fibrin glue therapy of ECF.  In the meantime, with little to lose and much to gain, we advocate a trial of fibrin glue therapy for ECF prior to surgical intervention.