This year, the Food and Drug Administration rejected the only medicine capable of treating the rare and fatal lung disease known as idiopathic pulmonary fibrosis. Pirfenidone, which has been available in Japan since 2008 and was just approved in Europe, was spurned by the FDA because the drug only showed efficacy in a single big trial--not the two large studies the FDA now requires. The decision to ban the drug is one of a rash of recent decisions that shows the FDA is making it more and more difficult for promising drugs to reach severely ill patients.
Last week, the FDA revoked an approval for the cancer drug Avastin because it said that evidence supporting its use in breast cancer wasn't strong enough. (The drug was judged ineffective because it merely stalls the spread of tumors.) European regulators, looking at the same data, made the opposite decision.
It wasn't supposed to be this way. In 1997, Congress passed the FDA Modernization Act, which gave the FDA broad discretion to reduce the quantity and rigor of clinical data needed to approve drugs targeting grave illnesses. The purpose of the law was to save lives by reducing the cost and time needed to launch such medicines.
But the FDA has steadily disregarded many of the law's provisions. Longer, larger trials that require drug makers to evaluate "hard" endpoints (like how long a cancer patient lives) rather than "surrogate" endpoints (like a drug's ability to shrink tumors) give FDA reviewers more statistical confidence. Reviewers prefer these drawn-out trials because they insulate the FDA from critics who say that it isn't focused enough on safety. But bigger trials increase the time needed to develop a drug, keeping it out of the hands of patients.
The Modernization Act also allowed the FDA to conduct drug trials in which patients are treated with an experimental medicine in a single group or "arm," and the trial can be completed in less than a year. But the FDA doesn't often opt for such trials. The agency commonly requests more complex, "multi-arm" and "placebo controlled" studies, which can take three years to finish and are much more expensive. Each patient enrolled in a trial adds over $30,000 to drug-development costs.
Of 76 cancer drugs approved since 2005, the FDA gave only 13 "accelerated approval"--another process created under the Modernization Act to expedite drug development. From 2001 to 2003, 78% of the novel cancer drugs approved were granted accelerated approval. Since then only 32% got the designation.
What's more, the clinical trial requirements that the FDA is imposing on cancer drugs with accelerated approval are now as burdensome as the requirements imposed on regular drugs. So, practically speaking, having "accelerated approval" doesn't mean anything.
Europeans are now approving novel drugs an average of three months more rapidly than we do. Of 82 novel drugs that were submitted for approval in both the U.S. and Europe between 2006 to 2009, 11 were approved only in Europe. One is for relapsed ovarian cancer, another for bone cancer.
To reverse these discouraging trends, Congress should reaffirm the provisions of the Modernization Act. It should spell out in legislation that the FDA "shall"--rather than "may"--approve drugs for severe conditions on the basis of a single study, or a more lenient statistical orthodoxy than "two, randomized, placebo controlled trials."
While Congress may not want to get into the business of establishing the FDA's analytical methods, it can call on the agency to convene an advisory panel to cultivate principles that are more permissive when it comes to very bad diseases. And it could go a step further, empowering patient groups with a mechanism to seek review of FDA decisions.
Congress also needs to modernize the way the FDA's review process is organized in order to increase efficiency and enable more cooperation. The science embedded in the most novel drugs is increasingly complex, requiring collaboration across many disciplines, including clinical medicine, pharmacology and statistical modeling.
But in recent years, the FDA has carved out each scientific discipline into its own distinct office. In addition, new work rules allow drug reviewers to spend two days each week working from home. The result is that FDA scientists don't collaborate well. Reviewers rarely meet as full teams, so they struggle to resolve internal debates and provide timely feedback to drug makers. The FDA's scientists should be organized around areas of therapeutic expertise--not broken into discrete offices based on what degree they have.
Finally, the FDA should be required to disclose its reasons for rejecting a drug.
The next Congress will reauthorize a user fee program that funds the FDA's review process. It should use this legislation to revive the FDA's fundamental mission: giving very sick Americans the best medical options available.
Scott Gottlieb, M.D., is a resident fellow at AEI.