Abstract:
|
Our increased understanding of genomics and related fields has led to: 1) the identification of subtypes of cancer based on various biomarkers and 2) to the development of drugs to target specific subtypes of cancer. Clinical trials test whether these novel treatments are effective in a population. Deciding which patients to enroll in a confirmatory, or phase III, trial can be difficult, but phase II trial results are informative. We propose using logic regression, a tree-based classification algorithm, on such results. The method identifies a subgroup of patients that have differential treatment benefit by finding a Boolean, or logical, statement of binary baseline covariates most strongly associated with an outcome. We have developed two adaptations of this method for binary and continuous outcomes along with two decision rules to determine who to enroll in a phase III trial. In simulation studies, we evaluate the performance of the two adaptations under various scenarios, including various subgroup and full population effect sizes, the prevalence of subgroups, the number of covariates considered and sample sizes.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.