Lipkovich Dmitrienko 2014

From MultXpert

Jump to: navigation, search

Lipkovich, I., Dmitrienko, A. (2014). Biomarker identification in clinical trials. Clinical and Statistical Considerations in Personalized Medicine. Carini, C., Menon, S., Chang, M. (editors). Chapman and Hall/CRC Press, New York.


This chapter discusses statistical methods used in the discovery of biomarkers and biomarker signatures in clinical trials. We view the problem of biomarker discovery as a special case of a more general problem of identifying a subgroup of subjects who experience enhanced treatment benefit compared to the general population. The subgroup is defined based on one or more pre-specified biomarkers. We begin with a review of common approaches to subgroup search in the context of personalized medicine and then focus on a novel biomarker discovery/subgroup search method (Subgroup Identification based on Differential Effect Search, or SIDES). SIDES is based on recursive partitioning and has been used in prospective and retrospective subgroup analysis in multiple drug development programs. Key components of SIDES will be discussed, including the subgroup search algorithm, selection of search parameters, efficient subgroup search with biomarker screening, and control of the Type I error rate using a resampling-based method. A case study based on a Phase III trial in subjects with colorectal cancer will be used to illustrate biomarker identification strategies.


Download Lipkovich Dmitrienko 2014.