Adaptive randomized trial design for patients with recurrent glioblastom

Purpose: To evaluate whether the incorporation of Bayesian principles into clinical trials for glioblastoma would be feasible and allow for more efficient trials.
Patients and Methods: We generated an adaptive randomization procedure that was retrospectively applied to primary patient data from four separate phase II clinical trials in patients with recurrent glioblastoma. We then compared adaptive randomization to more conventional trial designs using realistic hypothetical scenarios consistent with survival data reported in the literature.

Results: If our phase II trials had run as one multi-arm adaptively randomized trial, bevacizumab would have been identified as an efficacious therapy and required 30 fewer patients than a balanced randomized design. More generally, Bayesian adaptive randomization trial design for patients with glioblastoma would result in trials with fewer overall patients needed to achieve similar statistical power and more patients randomized to effective treatment arms. For a trial with a control arm, two ineffective arms and an effective arm with hazard ratio 0.66, a median of 50 patients would be randomized to the effective arm compared with 35 in a balanced randomized design.

Conclusions: Given the need for control arms in phase II trials, an increasing number of experimental therapeutics for patients with glioblastoma and a relatively short time for events, Bayesian principles are attractive for glioma clinical trials.