Adaptive Randomization of Aerobic Exercise During Chemotherapy in Colon Cancer

This Bayesian, two-stage, response-adaptive, dose-ranging trial of aerobic exercise during chemotherapy will include 219 stage II-III colon cancer survivors recruited from three diverse racial, and geographic regions (Baton Rouge, LA; Boston, MA; Oakland, CA). In stage 1 of the study, patients are randomized equally to 1 of 5 groups: attention control (stretching) or one of four doses of moderate-intensity aerobic exercise at 75 min/wk, 150 min/wk, 225 min/wk, or 300 min/wk. In stage 2 of the study (starting after 33% of subjects are enrolled), randomization probabilities will be optimized, contingent upon stage 1 efficacy (measured by chemotherapy toxicities and relative dose intensity) and feasibility (measured by exercise adherence) using a longitudinal model to maximize the use of incomplete data, predictive probabilities to guide decisions, and a dose-response model to borrow information across adjacent doses.

0