This week's Biostatistics Seminar will feature Dr. Lei Shen, Research Advisor, Global Statistical Sciences and Advanced Analytics, Eli Lilly & Company. His colleagues joining him on this Pittsburgh trip are Gary Sullivan, Senior Director and a Pittsburgh native, and Meg Gamalo, who is a Bayesian expert and a PhD graduate of the Statistics Department.
Dr. Shen's seminar is titled “Subgroup Identification for the Development of Tailored Therapeutics ".
Abstract: The topic of precision medicine has garnered increasing amount of attention in recent years, a trend that is expected to continue given the demand for both improved outcome and efficiency by patients and healthcare providers. One of the most important contributions to this effort by the pharmaceutical industry is the development of what is often referred to as tailored therapies. While many challenges remain, such development has been made more feasible by recent advances in statistical methods across the entire development program, which includes—among other aspects—effective identification of patient subgroups with enhanced response to a treatment as well as successful prospective demonstration of this response that is required for regulatory approval. In this presentation, I will review relevant methodology in the following areas: (1) a number of recent statistical methods for subgroup identification, (2) a general framework to synthesize these methods and some key insight it provides, (3) a rigorous process to determine a priori the optimal subgroup identification method for a given application, (4) a Bayesian method that effectively handles multiplicity and utilizes prior knowledge, and (5) design of prospective tailoring trials and important considerations for the regulatory process when multiple populations are of interest, such as the overall population and a sub-population defined by a biomarker. Taken together, these methodological advances can greatly improve the success of tailored therapy development.