Computer-simulated tests eyed at FDA to cut drug approval costs

Computer simulations may get a role alongside human testing as part of an effort to bring new medications and medical devices to market more quickly and cheaply.

The FDA outlined a proposal Friday to help integrate computer modeling and virtual testing as part of the regulatory approval process for manufacturers – a step the agency said could save money while helping find cures for puzzling conditions such as Alzheimer’s disease.

The plan is one step toward President Donald Trump’s goal of lowering drug prices. The average cost of developing a new medication is about $2.56 billion, according to a 2014 estimate from the Tufts Center for the Study of Drug Development, and much of that goes to fulfilling the FDA’s rigorous demands for proving safety and effectiveness.

“The price of new technology affects the ability of people to access these new treatments,” FDA Commissioner Scott Gottlieb wrote in a blog post. “We therefore need to be mindful of the costs of our regulatory processes, to the degree that these costs also affect the availability of new innovations, and the way that they are ultimately priced.”

While the White House hasn’t been able to reach consensus on an executive order to help make drugs more affordable, Gottlieb has gone after the issue from a few angles. He said last month the FDA would target high-priced products by prioritizing the approval of genetic competitors.

The idea of computer-simulated experiments, sometimes called in silico trials, has been around for years but the FDA hasn’t provided guidance to allow the drug industry to use it in testing. The European Union formed a two-year project in 2013 to study computer-simulated trials, and one of the researchers involved wrote last year that more research is needed to ensure its reliability.

Computer simulation is already being used to help find certain patients who may need adjusted drug dosages compared to the general patient population. Its use could be significantly expanded to represent some aspects of a patient group that would receive a placebo instead of an experimental medication, Gottlieb wrote.

The FDA has begun using computer modeling to build databases to help researchers predict how new treatments for Parkinson’s and Alzheimer’s disease will perform. And the agency is developing a family of “virtual patients” for testing new devices, Gottlieb said in the blog post.

The move to computer simulation is part of an Innovation Initiative that Gottlieb said he will expand on at a later date but “will be aimed at making sure our regulatory processes are modern and efficient” as part of the agency’s implementation of the 21st Century Cures Act. Congress passed the legislation last year that aimed to enhance the FDA’s authority to take a regulatory approach that focuses oversight on the riskiest areas and eases the burden on others.