This Could Happen Anywhere: Evolution of Translational 'Omics' and Lessons Learned from the Duke Saga
InformationDave DeMets, PhD
November 28, 2012
130 Recorded video views
21 Live stream views
DeMets' presentation will highlight the 2012 Institute of Medicine (IOM) report recommending ways to strengthen "omics" based test development and evaluation in response to flawed clinical trials undertaken at Duke University. DeMets was one of the IOM committee members authoring the report.
Based on a set of highly publicized reports claiming to be able to identify cancer risk and predict the response of individual patients to specific treatments using genomic data, Duke University initiated clinical trials for selecting patients and assigning treatment strategies guided by microarray genetic analysis, i.e., personalized cancer treatment.
Those trials were suspended and later terminated by the National Cancer Institute after outside biostatisticians outlined a series of problems with the analysis and data behind these sentinel papers. In addition, many of the original publications describing the research or the subsequent clinical trials were retracted.
DeMets' talk will highlight the series of problems with the analysis and data, summarize the IOM report, and discuss lessons learned from the series of events at Duke University. While these issues arose at Duke University, DeMets argues these or similar errors could occur elsewhere if the IOM process is not followed.