Changes in U.S. Food and Drug Administration (FDA) procedures meant to speed approvals for medications may have resulted in less exacting standards, a new analysis suggests.
Congressional acts that changed the way the FDA evaluates drugs have led to less rigorous evaluations, with drug approvals being based on fewer and/or earlier-stage clinical trials that may not be randomized, controlled, blinded or based on traditional measures of efficacy, experts noted in the article published in JAMA.
For example, the proportion of new drug approvals supported by at least two so-called pivotal trials - the clinical trials the FDA primarily relies on for its approval decisions - decreased from 80.6% in 1995-1997 to 52.8% in 2015-2017.
Those changes may lead to less confidence in the FDA’s approval process, said lead author Jonathan Darrow of Harvard Medical School in Boston and Brigham & Women’s Hospital Division of Pharmacoepidemiology & Pharmacoeconomics.
If drugs approved with less evidence turn out to be problematic it may lead to “an erosion of the ‘FDA approved’ brand,” Darrow said.
Darrow said the FDA didn’t always have a stringent review process. In 1962, when researchers determined that thalidomide, an anti-nausea drug given to pregnant women, caused birth defects, the agency’s mandate to test the efficacy and safety of new drugs was stepped up.
But over the last four decades, a number of laws have led to looser standards in some cases, Darrow said. One example is the Orphan Drug Act, which was aimed at fostering research on drugs for diseases that impact fewer than 200,000 Americans.
“If the standards are different than they were in the past, it’s important for patients and physicians to be aware of that,” Darrow said. “Patients and physicians need to focus on the evidence and not the fact of FDA approval. How big are the benefits, and how certain are we of the benefits?”
Darrow is even more concerned about the FDA’s most basic standard for drug approval. “It just has to be better than nothing,” he said. Because the comparison is to no treatment at all, it’s possible that an older drug might be more effective than a newly approved one, Darrow said.
Despite those issues, more drugs have been approved in recent years, especially in the category of biologics, which are often used to treat autoimmune diseases. The average annual number of new drug approvals, including biologics, was 34 from 1990-1999, 25 from 2000-2009 and 41 from 2010-2018. The proportion of drugs approved with the Orphan Drug Act designation increased from 18% in 1984-1995 to 41% in 2008-2018.
One major bright spot is that the median annual number of generic drugs approved rose from 284 prior to the Generic Drug User Fee Act of 2012 to 488 from 2013-2018, Darrow said.
The new report provides both good and bad news, said Dr. Albert Wu, an internist and a professor of health policy and management at the Johns Hopkins School of Public Health in Baltimore.
In the good news category, Wu points to the increased number of orphan drugs that have been approved. “Who doesn’t love breakthrough treatments for rare and incurable diseases?” Wu said in an email.
“Over the last 40 years, there has been a series of reforms which have given the FDA more control over the approval process - the goals being to speed the development and marketing of new medications for devastating diseases like spinal muscular atrophy,” Wu said. “They have done this by shortening the time it takes to review a new drug, accepting less data for a drug to be approved, and requiring fewer outcome measures.”
Then there’s the bad news. “Faster in this case means that less data gets collected,” Wu said. “Some drugs may get accelerated approval, on thinner evidence, and wind up not being any better than existing and often cheaper alternatives.”
Moreover, changes in the way the FDA works have led indirectly to higher drug prices, Wu said, pointing to the fact that the agency has slowed approvals of certain generic drugs that would be competing with new ones.
Our Standards: The Thomson Reuters Trust Principles.