Search in the Abstract Database

Abstracts Search 2017

* = Presenting author

P493 Efficacy of anti-TNF for internal fistula in Crohn's disease – results from a retrospective multicenter cohort study

Kobayashi T.*1, Hishida A.2, Tanaka H.3, Bamba S.4, Yamada A.5, Toshimitsu F.6, Shinichiro S.7, Kamata N.8, Yoshino T.9, Hibi T.1 AIM Jr. Internal Fistula Study Group

1Kitasato University Kitasato Institute Hospital, Center for Advanced IBD Research and Treatment, Tokyo, Japan 2Nagoya University Graduate School of Medicine, Department of Preventive Medicine, Nagoya, Japan 3Sapporo Kosei General Hospital, IBD Center, Sapporo, Japan 4Shiga University of Medical Science, Division of Gastroenterology, Shiga, Japan 5Toho University Sakura Medical Center, Department of Internal Medicine, Chiba, Japan 6Tokyo Medical and Dental University, Department of Gastroenterology and Hepatology, Tokyo, Japan 7Osaka University Graduate School of Medicine, Department of Gastroenterology and Hepatology, Osaka, Japan 8Osaka City University Graduate School of Medicine, Department of Gastroenterology, Osaka, Japan 9Tazuke Kofukai Medical Research Institute Kitano Hospital, IBD Center, Osaka, Japan


Fistulizing Crohn's disease is considered as a refractory disease phenotype which could be resistant to anti-TNF treatment. Internal fistula is especially a condition likely to require surgery, while few reports showed successful conservative management with medical treatments such as anti-TNF. Therefore, we performed a multicenter retrospective cohort study to investigate the outcome of anti-TNF for internal fistula in Crohn's disease.


Data were retrospectively collected from all Crohn's disease patients diagnosed with internal fistula then treated with anti-TNF (infliximab or adalimumab) between January 2002 and November 2015 at 20 institutions. Surgery was defined as a primary endpoint, and secondary endpoints were fistula closure and physician's assessment (closed, improved, no change, or worsened). Cumulative rate of surgery was evaluated by Kaplan-Meier analysis. Prognostic factors associated with surgery, fistula closure, and physician's assessments were assessed.


A total of 93 Crohn's disease cases were included in the study with mean follow-up period of 1452.8 days. Infliximab was used in 69 (74.2%) patients and 49 (52.7%) were on concomitant immunomodulators. Most had single (n=67, 72.0%), while the remainder (n=26, 28%) had complex draining fistulas. Of these fistulas, majority were entero-entero/colonic (n=69, 76.7%), and others were enterovesical (n=16, 17.8%) or enterovaginal (n=5, 5.6%). Coexisting stricture was present in 55 (59.1%) patients. Cumulative surgery rate was 17.7, 27.5, 37.3, and 47.2% at 1, 2, 3, and 5 years from induction of anti-TNF, respectively, and there was a trend for the increased risk of surgery by the increment of CDAI scores (univariate HR 1.41 (0.93–1.84); p=0.073 & multivariate HR 1.46 (0.99–2.13); p=0.055, by 100 points increase in CDAI) by Cox regression analysis. Fistula closure was confirmed in 27.0% in 5 years. Number of fistula (OR 0.21 (0.05–0.77)), enterovesical fistula (OR 4.4 (1.34–14.43)) and stricture (OR 0.38 (0.15–0.99)) were associated with closure of fistula by univariate logistic model but none of them was significant by multivariate analysis. Anal lesion, infliximab, and concomitant immunomodulator use independently reduced the risk of worsening defined by the physician's assessment.


Anti-TNF can be effective and surgical treatment may be avoided in selected patients with internal fistula. Further prospective trials with larger sample sizes are necessary to confirm our results.