A Bayesian Analysis of Disease Modification Using Doubly Randomized Delayed-Start and Matched-Control Design Paradigms

• Activity Number: 550
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Ibrahim Turkoz* and Marc Sobel
• Companies: Janssen R&D and Temple University
• Keywords: Bayesian Hierarchical Models ; Doubly Randomized ; Matched-Control ; Delayed-Start

Abstract

Randomized Delayed-Start designs introduce major drawbacks for evaluating disease modifying agents. A high percentage of subjects in the control arm are expected to drop out before entering into the delayed-start period due to the long duration of treatment period necessary to show the effects on disease progression. This results in major drawbacks for existing analytic approaches. The proposed study design resolves issues of randomized delayed-start trials and makes it feasible to establish a disease modification indication. The proposed innovation is a hybrid of randomized and epidemiologic designs that introduces a run-in period and the second randomization for delayed-start. The run-in period is used for a matched control analysis. In the delayed-start period, the objective is to show that the late initiation of treatment does not permit the same level of recovery as experienced by those subjects who have been on the drug from the start. The "failure to catch up" concept is evaluated using hierarchical priors. Bayesian methodology is a natural fit for establishing disease modification compared to the traditional non-inferiority margins or parallel lines approach.