Department of Biostatistics Seminar
Individual Response to Treatment: Theranostics Between Hope and Hype
Stephen Senn, Ph.D.
Department of Statistics, University of Glasgow
Friday, December 3, 12:00-1:00 pm, Light Hall, Room 214
An oft-expressed hope is that pharmacogenomics will offer the possibility both of reducing variability in clinical trials and of individualising treatment using theranostics: diagnostic and therapeutic strategies. However, I consider some salient case-studies to show that random within-patient variability, which cannot be genetic in origin, is often underestimated as a possible explanation of varying response in clinical trials.
Since patients vary by more than their genes, patient-by-treatment interaction provides an upper bound to gene-by-treatment interaction. That being so that neglected and despised clinical trial the cross-over could be used to detect treatments for which individual response is important and hence genetic variability might be. The sort of design required is one in which patients are repeatedly treated with the same treatment. Unfortunately statisticians prefer to use such designs in a hopeless quest for dealing with treatment carry-over rather than for examining patient-by-treatment interaction and the latter is a neglected subject in clinical trials.
I also consider whether, since we currently show extreme reluctance to individualising treatment using something as a simple as bathroom scales, whether we are not rash in assuming that we will be prepared to do so using gene chips. I suggest that a change in attitude in drug development is needed.