Major CRO taps 'big data' analytics to divine diabetes drug R&D success

Covance ($CVD), the large contract research organization (CRO), has formed an alliance with the 'big data' analytics outfit GNS Healthcare to work on computer models to aid predictions on the success of clinical trials and to zero in on patients likely to benefit from experimental drugs. The models are a work in process, GNS CEO Colin Hill tells FierceBiotech IT, but such predictive capabilities could be "game-changing" in an industry unhappy with its abysmal win-to-loss ratio in clinical trials.

Cambridge, MA-based GNS has crafted computer models for major pharma and biotech groups with its reverse engineering and forward simulation platform, but the collaboration with Princeton, NJ-based Covance is the company's first focused on the ambitious idea of helping a CRO offer its customers a forecasting option for clinical development, according to Hill. To start, the partners aim to develop computer models for trials of Type II diabetes meds, which is one of the busiest R&D areas in pharma.

Starting with diabetes makes sense. The disease afflicts millions of Americans, and companies typically invest huge sums for trials of new drugs against the metabolic disease, in part because of the large size and scope of the programs. When the programs fail--as some inevitably do--all the millions invested in R&D go up in smoke. In general, 90% of drug programs crash and burn for various reasons at some point during clinical trials. And pharma groups are scrambling to improve on this score.

"This would be something that becomes part of the Covance offering that really gives them a differentiator compared with other companies and ultimately is something that the pharma industry needs and has needed for years," Hill said. "If you're in Big Pharma and you're trying to develop a drug for Type II diabetes, the fact that the CRO that is going to run the trial for you has a predictive model of showing insights of what parameters are likely to lead to greater success, that's game-changing."

Hill hopes the computer models are ready for action at Covance in 2013. (Terms of the collaboration weren't revealed.) To produce the models, GNS plans to use Covance's ample safety and efficacy data from years of drug-development programs.

- here's the release