Defining the optimal response criteria for the Crohn's disease activity index for induction studies in patients with mildly to moderately active Crohn's disease.

Objectives: The Crohn's Disease Activity Index (CDAI) is used to judge efficacy in clinical trials. We explored the effect of CDAI response definitions for induction on study efficiency.

Methods: We analyzed primary CDAI data from induction studies in patients with mildly to moderately active Crohn's disease, not receiving concomitant aminosalicylates, corticosteroids, or immunomodulator therapy, and without fistulizing or stricturing complications. The 12 definitions of clinical response included: CDAI decrease from baseline by 50, 70, 100, or 150 points; decrease by 25% from baseline and by 70 or 100 points; CDAI <100 or 150 points; CDAI <150 points plus decrease by 70 or 100 points; CDAI <150 points at any time sustained for the duration of the trial; or decrease in the CDAI by 70 points for the last two consecutive visits. Response definitions were ranked according to ability to optimize the effect difference between treatment arms. The effect of time, baseline disease activity (CDAI 200-299 or > or =300 points), and previous surgical resections on response definitions were evaluated and ranked. Multivariate analysis on additional factors of age (<40 or > or =40 yr), gender and duration of disease (<2 or > or =2 yr) were performed to determine predictors of response when applied to these CDAI definitions.

Results: Treatment effect differences in placebo-controlled studies were maximized by response definitions that incorporated either a decrease CDAI > or =70 points for the last two consecutive visits or decrease in baseline CDAI > or =100 points, and remained optimal when evaluated for the composite effect of time, baseline activity, and prior resections. A decrease in baseline CDAI > or =100 points had some advantages over a decrease CDAI > or =70 points over two visits in terms of study efficiency, as it produced a lower control response rate and was not influenced by any of the baseline factors.

Conclusion: Clinical trial efficiency for induction studies in patients with mildly to moderately active Crohn's disease can be improved by using either a decrease in CDAI by > or =70 points for the last two consecutive visits or a decrease in baseline CDAI by > or =100 points as the primary end point for the trial. These findings are valid for patients with ileocecal Crohn's disease not refractory to aminosalicylates, corticosteroids, immunomodulators, and biologics, and patients who do not have stricturing or penetrating complications. It is unclear if these CDAI response criteria would similarly increase study efficiency in trials that recruited patients with moderately to severely active disease, patients refractory to aminosalicylates, corticosteroids, immunomodulators, and biologics, and patients with stricturing or penetrating complications.