Scientific Abstract

Proposal No. IBD-0063
Principal Investigator: Joseph Romagnuolo, M.D.
Current Applicant Organization:  Medical University of South Carolina (Charleston, U.S.A.)
Original Applicant Organization: University of Calgary (Canada)
Project Title: Capsule endoscopic findings and incidence of disease in first-degree relatives of Crohn’s patients with abnormal intestinal permeability
Period of Award: October 1, 2003 - March 31, 2007

Background and Significance: Approximately 10% of healthy relatives of patients with Crohn’s disease (CD) have increased lactose-mannitol (L/M) intestinal permeability. Although hypothesized to reflect a primary defect and to play an etiological role in the development of CD, evidence to this effect is lacking. Existing imaging tests are too insensitive to identify early macroscopic mucosal abnormalities in the small bowel; wireless capsule endoscopy (CE) images of the small bowel may have higher accuracy than conventional options.

Hypothesis: First-degree relatives of CD patients with elevated L/M permeability will have a higher prevalence of mucosal abnormalities by CE and a higher three-year incidence of clinical CD than those without.

Methods and Analysis: Approximately 400 first-degree relatives under age 35 years will be identified through the Calgary CD database (n=900) and undergo L/M testing (10% anticipated to be abnormal). Eighty subjects will then be recruited (40 with and 40 age- and sex-matched without L/M abnormalities), and will undergo baseline history, physical, laboratory blood testing (including NOD-2 genetic testing). Each will then undergo CE, SBFT and colonoscopy with ileal and colonic biopsies. A GI pathologist will read the specimens, and blood, tissue, and stool will be banked for further testing. Patients will undergo repeat L/M annually, and if positive, will undergo repeat CE. If symptoms suspicious for CD develop, a re-evaluation with CE, SBFT, and colonoscopy will be performed. All patients will have a repeat CE at 36-months. Prevalence and 95% CIs will be calculated and comparisons made using the Fisher Exact test. Survival analysis (Kaplan-Meier) and multivariate Cox proportional hazard model will be used.

Sample Size Calculation: A precision of ± 15% would be reasonable for the prevalence estimate of CE mucosal abnormalities (estimated to be one-third). To achieve this 95% CI width, 38 patients are required in that group, rounded up to 40. The conventional power estimates (alpha=0.05) of the one-sample estimate of prevalence of CE abnormalities in those with a positive L/M, and of the incidence of CD for the entire cohort are 95% and 72%, respectively, compared with an alternative hypothesis of 0%.