BLOG: How the FDA system can fail
Click Here to Manage Email Alerts
How do some drugs and devices make it through the FDA approval system, only to show significant, undiscovered flaws? Why do other products never gain approval when they otherwise should?
After years of dialogue with industry insiders and interaction with the FDA, I would propose the following as the most common causes of failure.
Bad luck. Statistically speaking, about one in 20 trials should fail to meet its primary endpoint if it is aiming to refute the null hypothesis with 95% confidence. That’s about the same probability as flipping a coin and landing on heads four times in a row. It’s not common, but it happens. Companies hope that having multiple endpoints with at least some promising outcomes will mitigate a trial failure. Unfortunately, this often requires conducting additional trials, which costs money and brings us to our next failure point.
Not enough money. The cost of bringing a new drug to market has been estimated at near $1 billion. Many small companies simply aren’t capitalized to have any form of failure along the way. Viewed as too risky by future investors, they will fail to raise more money and eventually fold. This is particularly unfortunate for products that would serve a niche but clinically important need, like treatment of Acanthamoeba or fungal keratitis, where we are unlikely to see funding for new treatments.
Escalating commitment bias. This is almost the opposite problem. Sometimes, early safety concerns warn that a product should not proceed through trials without significant modification. Companies throw good money after bad by continuing trials instead of pausing to get it right. If you’ve ever been on a losing streak but continued gambling instead of cutting your losses, or if you’ve stayed in a bad relationship simply because you didn’t want to break free, you too have fallen victim to the human frailty of escalating commitment bias. In retrospect, the futility is usually obvious, but many companies, especially those with only one product, can be tempted to push on when they should instead fail quickly.
Inadequate trial design. Trials have so many ways of failing, including enrolling not enough patients to meet their endpoints or having the wrong kinds of patients. Recently, the FDA has focused on enrolling both men and women as well as different races and ethnicities, so the true makeup of the population is represented in a trial. Only through doing this can we get a real estimate of how a product will fare in treating disparate populations.
Not examining the right measures. Early trials of accommodating lenses defined accommodation very differently from today. As a result, we have had no lens achieve an accommodating claim since the Crystalens (Bausch + Lomb) in 2003. What are the right measures is a matter of debate. In the case of Beovu (brolucizumab, Novartis), which was discovered after approval to have a higher than expected rate of retinal vasculitis, we have learned that anti-VEGF drug trials should include careful examinations of pathology in the peripheral retina.
Setting too high a bar. Many dry eye companies failed to meet endpoints of treating both signs and symptoms of this disease when the FDA modified its requirements after the initial approval of Restasis (cyclosporine, Allergan) in 2002. Some argued that the bar was set too high. However, both Xiidra (lifitegrast, Novartis) and Cequa (cyclosporine 0.09%, Sun) have achieved approval since the new requirements were put in place, proving they can be met, though at significant cost.
A hostile FDA panel. A new drug or device may be subject to review by an FDA advisory committee, a panel of volunteer experts who review data and give non-binding advice to the FDA. But panels are made up of individuals who have their own biases. There are many stories of unfair treatment of products and companies by panels, who are made up of well-intentioned but human participants.
A crooked drug company hiding actual results. This virtually never happens. In general, our industry colleagues have the same ethics we physicians do: to innovate and serve patients with truly superior new offerings. Additionally, transparency in the approval process, double-masked trial designs, a multitude of investigators who have eyes on real patient results and fierce enforcement for violators make it unlikely that nefarious motives would be rewarded in our U.S. system.
Fortunately, the FDA has become much more streamlined in the past several years, allowing approval of many life-changing drugs and devices with relatively few failures. The system, which is run by intelligent and patient-focused civil servants and volunteers, works far more often than it doesn’t. The continued collaboration between a well-funded industry, a rigorous but benign governmental agency, and knowledgeable clinicians ready to try new treatments will be responsible for many of the life-changing advances our patients will enjoy now and in the future.
Collapse