Industry is anxiously awaiting guidance from FDA on the development of companion diagnostics for drugs that will inform efforts to move towards personalized medicine. The agency has vowed to release a draft guidance on companion diagnostics by the end of the year. But the initial document will address development of companion diagnostics for drugs already on the market, not simultaneous drug/diagnostic development for new drugs.
According to Vicki Seyfert-Margolis, FDA’s senior advisor for science innovation and policy:
“One challenge is that multiple centers have to be involved – CDRH and CDER/CBER. The challenge is internal to determine the mechanism for how the applications will come in and how the review is done. I think we are well on our way to figuring out that process between the different centers. With respect to how one designs and qualifies a diagnostic versus a clinical trial, the study design aspects are another challenge.”
“We are also concerned that we will have markers and use patient selection strategies in a clinical trial and then that marker may turn out to be a disease prognostication marker and you may have selected a set of patients that have different metabolisms. There are a myriad of possibilities one could think of that may or may not completely relate to the drug. We just have to be very thoughtful. I think we will evaluate a lot on a case-by-case basis, but we are trying to at least get some information out about what the paths are.”
And then there’s the issue of the “open kimono.”
Ms. Seyfert-Margolis: “One thing I think would be a big win in helping drive personalized medicine and co-development and companion diagnostics will be to open the data. We need to open the data in-house and also strive to get industry to be more transparent and work with us on this. If we took rheumatoid arthritis, for example, and looked across all the trials that have been done with TNF-alphas and evaluated who are the true responders, who are the non-responders, is there anything we can find in all that data that might point to some marker or some diagnostic that could be used, then we could begin to get at least hints about things that could open up new scientific areas for predictive diagnostics.
Amen. Sounds like a good way to involve the Reagan/Udall Foundation.
