When:
Monday, April 2, 2018
3:00 PM - 4:00 PM CT
Where: 680 N. Lake Shore Drive, Stamler Conference Room, Suite 1400, Chicago, IL 60611 map it
Audience: Faculty/Staff - Student - Post Docs/Docs - Graduate Students
Contact:
Tameka Brannon
Group: Department of Preventive Medicine
Category: Lectures & Meetings
An adaptive Bayesian design for early phase platform trials of targeted cancer therapies
With the advances in discovery of novel molecular targets, there are increasing interests in evaluating a targeted therapy across different disease subtypes characterized by the unique expression of certain molecular targets, biomarkers or genotypes. At the same time, patients with a certain biomarker expression could potentially benefit from different experimental drugs; and therefore, evaluating the relative efficacy of these drugs for a given biomarker is an important objective in the early phase drug development. In this paper, we consider the design of a platform trial in phase I drug development, in which multiple therapies are compared in patients with different biomarker profiles, with an objective to identify the best drug at an efficacious and safe dose for a given disease subtype. Specifically, we use the continual reassessment method (CRM) to estimate the maximum tolerated dose of a drug and adopt hierarchical Bayesian modeling (HBM) to estimate efficacy of a drug administered at multiple doses. Using the CRM and HBM as basis of inference, we study various algorithms that prescribe drugs/doses for the study subjects using both adaptive and non-adaptive randomization. Through simulations, we compared the performance of the proposed approaches and demonstrate that overall our approaches will correctly prioritize the candidate drugs, identifying the drug that works better for patients with a particular biomarker.