P171. Prevalence of Crohn's disease in patients with spondyloarthropathies: interim analysis of the SPaCE study
U. Kopylov1, B. Adelson1, M. Starr2, C. Watts2, E. Seidman1, 1McGill University Health Center, Center for IBD research, Montreal, Canada, 2McGill University Health Center, Rheumatology, Montreal, Canada
Inflammatory bowel disease (IBD) is clinically associated with spondylarthropathies (SpA) in 5–15% of cases. Colonoscopic studies have shown a high prevalence (30–44%) of asymptomatic inflammation characteristic of Crohn's disease in SpA. Capsule endoscopy (CE) has been established as providing a higher detection rate for small bowel Crohn's disease, compared to ileo-colonoscopy (IC) and other imaging modalities.
The aim of our study was to evaluate the accuracy of CE for detection of small bowel lesions in patients with SPA.
This is a prospective cross-sectional study. The study cohort included patients (>18 y) followed by MUHC Rheumatology Center for known SpA or ankylosing spondylitis. All patients were evaluated by CE, followed (within 4 weeks) by ileocolonoscopy (IC), performed blinded to CE results. NSAIDs were discontinued for a minimum of 4 weeks. The presence of small bowel mucosal inflammation on CE was quantified by the Lewis score (LS).
We report the preliminary results of the first 24 SPaCE study patients (age 39±12, 54% male). 86% were HLA-B27 positive. Sixty three percent of the patients had gastrointestinal complaints, and 33.3% had elevated CRP levels. Recent therapy for SpA included etanercept (1 patient), methotrexate (1 patient) and NSAIDs in 58.3% of the patients.
IC with biopsies demonstrated endoscopic and pathologic findings consistent with CD in 18.2% of the patients. CE identified small bowel lesions consistent with CD in 33% of the patients (mean LS 1083±702), including all patients with IC-detected CD, and equivocal IC findings in additional 8.3% of the patients (LS of 168 and 225), respectively. In patients with equivocal CE results, CD was not demonstrated by IC.
Gastrointestinal symptoms had a positive predictive value of 40% and 24%, and negative predictive value of 78.7% and 88.9% for findings on CE and IC, respectively.
Following the diagnosis of small bowel inflammatory lesions, ant-TNF treatment was initiated in 4/8 patients.
Small bowel lesions consistent with Crohn's disease are common in patients with spondyloarthropathies. These preliminary results suggest that CE of the SB is superior to IC in detecting CD in patients with known SpA. The findings obtained by CE are of clinical importance and resulted in a treatment change in 50% of the patients. Gastrointestinal complaints are a poor indicator for the presence Crohn's disease. These results merit verification on completion of the study.