“Some Novel Clinical Trial Designs For Dose-Finding In Oncology" By Philip Boonstra PhD, Biostatistics, University of Michigan

When studying new pharmaceutical treatments for cancer, one of the challenges is finding the right dose of the drug: too much drug may be harmful for the patient and too little drug may not be
efficacious against the cancer. Thus, before conducting the large randomized controlled trials comparing a new treatment against a standard of care, which we sometimes read about in the news,
it is often required to first conduct a so-called “early phase” clinical trial that explores different dose levels of the new treatment, so as to characterize this tradeoff between toxicity to the patient and
efficacy against the cancer. In this talk, I will discuss some novel clinical trial designs that I work on as a cancer biostatistician. These designs try to balance between quickly identifying a safe and
efficacious dose while also treating as many patients as possible at that dose level. I will show how fundamental statistical ideas, such as likelihood and Bayes Theorem, are relevant to these designs. I
will end my talk with an advertisement for biostatistics graduate studies (especially, but not just, at University of Michigan!). Biostatisticians need to be good at thinking quantitatively but also must
have a passion for biomedical science and be effective communicators, all of which means your liberal arts training at Hope provides an ideal foundation for graduate education in biostatistics.