Institute produces study in response to Potti scandal

The Institute of Medicine of the National Academies published a report on research integrity and procedure Friday that was partly in response to the scandal surrounding former Duke oncologist Dr. Anil Potti.

The report addresses problems with the application of genomics and other molecular bioscience research in clinical trials. Titled “Evolution of Translational Omics: Lessons Learned and the Path Forward,” it notes that the lack of a defined review process allows flawed research to pass unnoticed and potentially enter the clinical phase. The report provides recommendations for new practices that can evaluate and verify the data and computational research supporting proposed clinical treatments.

At a meeting in October 2010, the IOM decided to investigate translational research procedures as well as specific clinical trials based on Potti’s work. The IOM pursued this examination at the request of the National Cancer Institute, which originally intervened in the Potti research in summer 2010. Although the Duke case is alluded to, Potti is not named explicitly in this report. Duke is currently investigating Potti for alleged research misconduct.

Translating molecular bioscience research into reliable, usable data has proven challenging, the report states. The research often involves complex computational procedures and without access to the computer code and systems, outsiders are unable to assess and review the data. These issues came to life following “premature use” of data from genomics-based clinical trials.

In his research, Potti used genomics-based tests to predict individualized cancer treatments. The University suspended Potti’s clinical trials summer 2010 in response to questions regarding the underlying research. Potti resigned from Duke in November 2010, taking sole responsibility for inconsistencies in his data. Potti and his co-authors have since retracted 10 papers based on his research. There are currently two lawsuits pending against Potti, Duke and others related to clinical trials conducted based on Potti’s genomic research.

“For three years, the Duke investigators rebuffed external criticisms about publications describing their tests. But in July 2010, more than 30 outside scientists raised concerns about the validity of the tests,” the report states.

The report suggests that computational procedures used to manage large data sets should be confirmed and made public or patented. Processes that are candidates for clinical trials should also be defined by explicit measurements, a breakdown of computational procedures and intended clinical use. Additionally, all test validation should take place in a Clinical Laboratory Improvement Act-certified laboratory, and researchers should consult with the Food and Drug Administration before moving on to clinical trials.

The report was sponsored by numerous national institutions, including the American Society for Clinical Pathology, Centers for Disease Control and Prevention, College of American Pathologists and Department of Veterans Affairs.