Summarizing historical information on controls in clinical trials

Infliximab, a chimeric monoclonal antibody directed against tumor necrosis factor alpha, is an established treatment for Crohn's disease but not ulcerative colitis. Two randomized, double-blind, placebo-controlled studies--the Active Ulcerative Colitis Trials 1 and 2 (ACT 1 and ACT 2, respectively)--evaluated the efficacy of infliximab for induction and maintenance therapy in adults with ulcerative colitis. In each study, 364 patients with moderate-to-severe active ulcerative colitis despite treatment with concurrent medications received placebo or infliximab (5 mg or 10 mg per kilogram of body weight) intravenously at weeks 0, 2, and 6 and then every eight weeks through week 46 (in ACT 1) or week 22 (in ACT 2). Patients were followed for 54 weeks in ACT 1 and 30 weeks in ACT 2. In ACT 1, 69 percent of patients who received 5 mg of infliximab and 61 percent of those who received 10 mg had a clinical response at week 8, as compared with 37 percent of those who received placebo (P<0.001 for both comparisons with placebo). A response was defined as a decrease in the Mayo score of at least 3 points and at least 30 percent, with an accompanying decrease in the subscore for rectal bleeding of at least 1 point or an absolute rectal-bleeding subscore of 0 or 1. In ACT 2, 64 percent of patients who received 5 mg of infliximab and 69 percent of those who received 10 mg had a clinical response at week 8, as compared with 29 percent of those who received placebo (P<0.001 for both comparisons with placebo). In both studies, patients who received infliximab were more likely to have a clinical response at week 30 (P< or =0.002 for all comparisons). In ACT 1, more patients who received 5 mg or 10 mg of infliximab had a clinical response at week 54 (45 percent and 44 percent, respectively) than did those who received placebo (20 percent, P<0.001 for both comparisons). Patients with moderate-to-severe active ulcerative colitis treated with infliximab at weeks 0, 2, and 6 and every eight weeks thereafter were more likely to have a clinical response at weeks 8, 30, and 54 than were those receiving placebo. (ClinicalTrials.gov numbers, NCT00036439 and NCT00096655.) Copyright 2005 Massachusetts Medical Society.

Current methods for meta-analysis still leave a number of unresolved issues, such as the choice between fixed- and random-effects models, the choice of population distribution in a random-effects analysis, the treatment of small studies and extreme results, and incorporation of study-specific covariates. We describe how a full Bayesian analysis can deal with these and other issues in a natural way, illustrated by a recent published example that displays a number of problems. Such analyses are now generally available using the BUGS implementation of Markov chain Monte Carlo numerical integration techniques. Appropriate proper prior distributions are derived, and sensitivity analysis to a variety of prior assumptions carried out. Current methods are briefly summarized and compared to the full Bayes analysis.