What we lose when we approve unproven drugs

Last month, an expert panel at the US Food and Drug Administration (FDA) voted 14-1 to recommend the FDA withdraw Makena (a synthetic hormone for preventing premature birth) from the market, because it failed to show benefit in its postmarketing trial.

Preterm birth (defined as delivery before 37 weeks) impacts one in ten births in the US, and accounted for 16% of infant deaths in 2020. However, there are significant racial disparities. According to the CDC, the rate of preterm birth among Black women was 14.8% in 2021, much higher than the rate among white (9.5%) or Hispanic women (10.2%).

Given the devastating consequences of preterm birth, it’s tempting to want to take any action possible to fix it. If there is a chance that a drug could reduce premature births and racial disparities in infant death and disability, shouldn’t we take that chance? Those who approve drugs for serious conditions are not immune to the desire for hope in treating these conditions — it’s what drove approval of the recent ALS drug, for example.

But the Makena case shows how the logic of “What do we have to lose?” falls apart when you look at the evidence. From approval to withdrawal, the story of Makena shows some of the major pitfalls of the FDA’s drug approval process.

What is Makena?

Makena is a compound version of a synthetic hormone called hydroxyprogesterone caproate that was first developed in the 1950s. For decades, doctors thought the drug might reduce preterm births, but there was no evidence to show it worked. Bristol-Myers Squibb marketed the drug as Delalutin until the FDA issued a warning in the 1970s about the lack of effectiveness and safety issues, and BMS took it off the market.

A 2003 trial brought the hormone back into the public eye. The National Institute of Child Health and Human Development conducted a randomized trial of hydroxyprogesterone caproate for women with a history of spontaneous preterm birth. They found a significant difference in rates of preterm birth and low birth weight, although there were no differences in infant mortality or morbidity.

Seems promising, but researchers noted some concerning methodological issues. For one, the placebo group had higher rate of preterm birth before the trial, indicating that the randomization was not done correctly. Another issue was that a large proportion of trial participants were from a single site, so the study results may have been confounded (ie. something particular about that site is driving the results, not the effect of the drug). The FDA statistical review noted these limitations and concluded that the evidence wasn’t strong enough to approve Makena based on a single study.

Accelerated approval

Despite these concerns, K-V Pharmaceuticals (which owned Makena at the time) was still able to gain approval for the compound through the FDA’s accelerated approval process in 2011. Accelerated approval allows for drugs to be greenlit without evidence that they improve meaningful clinical outcomes, as long as they prove effective on a “surrogate endpoint” (a metric that is “reasonably likely to predict clinical benefit”).

In the case of Makena, the drug was approved based on reduction in delivery before 37 weeks, a surrogate endpoint for infant mortality associated with premature birth. The company was required to conduct a “confirmatory” trial to ensure that the drug was indeed effective.

Cut to 2019 when the new trial results came out. The PROLONG trial showed no significant difference in frequency of preterm birth or infant mortality for those taking Makena. In 2020, an FDA advisory panel voted to withdraw Makena from the market, but Covis pharmaceuticals, the private equity-backed new owner of the drug, demanded a hearing with the FDA before withdrawal. That hearing didn’t happen until October 2022.

What’s at stake?

Between the time it took to conduct the follow-up trial and the delay for the hearing, Makena was on the market for 11 years. That’s more than a decade of giving pregnant people a drug with an unclear benefit.

It’s also an immense cost to the system. Although hydroxyprogesterone caproate was first developed 80 years ago, after the FDA approved it, K-V pharmaceuticals marked up Makena 75 times the generic price. Medicaid, which covers about half of all births in the US, spent $700 million on the drug since 2018. Advocates have pointed out that this $700 million could have been used on other interventions to improve birth outcomes, such as funding birth centers and cultural competency training for medical professionals.

Makena is just one example of the waste generated by unproven drugs greenlit through the accelerated approval pathway. According to a recent report from the Office of the Inspector General, Medicare and Medicaid spent more than $18 billion from 2018 to 2021 for accelerated approval drugs with incomplete follow-up trials past their original due dates.

What’s next?

Covis’ “hail mary” hearing didn’t work, but that doesn’t mean Makena is off the market instantly. The FDA will deliberate for a few months before deciding whether to remove the drug. The Makena saga provides an important lesson for the FDA and the drug approval process: An ounce of prevention is worth a pound of cure.

We could save so much time and money by ensuring that evidence supports the approval of drugs in the first place, rather than approve them and wait for the evidence. Makena is not an anomaly in terms of drugs approved through the accelerated pathway that later prove ineffective. A 2019 study found that out of 93 cancer drug indications granted accelerated approval, only 19 (20%) had confirmatory trials that demonstrated a benefit in overall survival. The case of Makena should urge the FDA to raise the bar on evidence for accelerated approval.