More and more studies have shown that genetic determinants may mediate variability among persons in the response to a drug, and thus some therapeutics may benefit only a subset of patients. Genomic technologies, such as DNA sequencing, mRNA transcript profiling, and comparative genomic hybridization, are providing biomarkers to predict who are most likely to respond to a given drug, and thus brings opportunity to conduct targeted clinical trials with eligibility restricted to the subset of patients. In this paper, we evaluate the relative cost of a targeted design versus an untargeted design for a randomized phase III clinical trial comparing a new treatment to a control. Our investigation indicates that the effectiveness of the targeted design critically depends upon the difference of treatment effect between patient subsets, the proportion of targeted patients in the population, the diagnostic assay performance, and the relative cost of screening versus drug expenses.
Hsiao et al. (2015)