P222. Incidence and risk factors for free bowel perforation in patients with Crohn's disease
J.W. Kim1, B.D. Ye1, S.-K. Yang1, S.-K. Park1, S.H. Park1, D.-H. Yang1, K.W. Jung1, K.-J. Kim1, J.-S. Byeon1, S.-J. Myung1, J.-H. Kim1, 1Asan Medical Center, Gastroenterology, Seoul, South Korea
Free bowel perforation (FP) in Crohn's disease (CD) can be life-threatening especially among previously stable patients in clinical remission. The aim of this study was to investigate the incidence and risk factors of FP in CD patients.
We analyzed the medical records of CD patients registered at Asan Medical Center from March 1991 to May 2012. The incidence of FP, defined as a sudden spontaneous penetration of the bowel resulting in intestinal contents flowing into the peritoneal cavity, was investigated. After matching FP cases to controls (1:4) by year at diagnosis and gender, clinical characteristics were compared using conditional logistic regression. Medication history during the last 3 months before FP (cases) or corresponding follow-up period (controls) was investigated.
Among 1601 patients who were managed for CD at out center, there were 1136 (71.0%) males, and the mean age at diagnosis of CD was 24.8 years (standard deviation, 8.7 years). Over a median follow-up time of 5.5 years (interquartile range [IQR], 2.4–9.0), 55 patients (3.4%) developed FP, with the incidence of 5.34 per 1000 person-years (95% confidence interval [CI], 4.1–7.0). Among FP cases, there were 42 males (76.4%), and the median age at diagnosis of CD was 29.4 years (IQR, 20.8–35.6 years). The disease locations were ileum in 21 (38.2%), ileocolon in 32 (58.2%) and colon in 2 (3.6%). The disease behaviors before FP were non-stricturing and non-penetrating in 43 (78.2%), stricturing in 6 (10.9%), and penetrating in 6 (10.9%). On multivariable analysis comparing FP cases to matched controls, previous steroid (odds ratio [OR], 8.6; 95% confidence interval [CI], 3.5–21.3) or anti-TNF therapy (OR, 5.9; 95% CI, 1.3–25.9), higher body mass index (BMI) (OR, 1.2 per increase in 1 kg/m2; 95% CI, 1.1–1.4) were independently associated with increased FP. Immunomodulator therapy (OR, 0.4; 95% CI, 0.2–0.9) and disease location at colon (OR, 0.2; 95% CI, 0.1–0.4) were associated with decreased odds for FP.
The incidence of FP in a large cohort of Korean patients with CD was 3.4%, which was similar to Western reports. Disease location at small bowel, prior steroid or anti-TNF therapy, higher BMI and non-use of immunomodulator were associated with increased FP.This study was supported by a grant of the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A120176).