Raising the Bar

It's time for a total overhaul of the way the FDA licenses drugs

Last year, I wrote about a new drug called brexpiprazole—brand name “Rexulti”—that had just been approved by the Food and Drug Administration, making it the first licensed antipsychotic drug for dementia sufferers in the US.

But there’s just one problem with brexpiprazole, and it’s kind of a big problem: The drug doesn’t work.

And what’s worse, the FDA knew it didn’t work when they approved it.

In three pre-approved trials, the drug brought just a 5.3-point improvement for users—on a 174-point scale. Seventeen points is the minimum for an improvement to be considered clinically significant. More worryingly, the drug increased the user’s risk of death fourfold.

So: no significant improvement, but a four-times-greater risk of death. Got it.

Brexpiprazole wasn’t the first drug of its kind to produce such dismal results during trials. In fact, same thing happened in every previous trial of antipsychotics for dementia sufferers that took place. The difference is, those drugs weren’t approved—but brexpiprazole was.

Why?

The FDA has so far refused to comment on the licensing process for brexpiprazole, which is no surprise. Absent specifics, we can still get a good idea of how such an obviously nonsensical approval could take place, and it has to do with falling standards at the regulator and growing corporate influence over it. Those two things are related, of course.

Critics have pointed out that standards at the FDA have declined significantly in the last two decades, and the decline has accelerated. Well-meaning legislation like the 21st Century Cures Act, passed in 2016, has allowed drugs to reach market quicker, but at the cost, it seems, of rigor and certainty in the approval process. A far greater proportion of drugs are now approved by the FDA on the basis of just a single study than ever were before.

Transparency has suffered too—a significant proportion of drug trials are not disclosed on the FDA’s website, as they should be—at the same time as lobbying by drug companies has become more confident and aggressive. This is surely no coincidence. In the case of brexpiprazole, the FDA was lobbied by multiple patient-advocacy groups funded by Otsuka, one of the two manufacturers of the drug. Otsuka and its partner Lundbeck have forecast annual sales of $1 billion in the US for brexpiprazole, with a monthly dose costing $1400.

It’s clear there’s something very wrong with the way the FDA regulates pharmaceuticals, and that brexpiprazole is not a rare instance of failure. The true scale of the problem, however, has not been fully appreciated until now.

A few weeks ago, an investigative report from The Lever showed that, between 2013 and 2022, almost 73% of all drugs approved by the FDA didn’t meet the agency’s very own four foundational standards, which are, by the agency’s own admission, the bare minimum for a drug to be considered fit for approval.

Fifty-five of the 429 drugs approved over that period met just one of those four standards, and 39 met none of them.

These issues were particularly bad when it came to cancer drugs. Just 2.4% of the 123 cancer drugs approved between 2013 and 2022 met all four of the foundational standards. Twenty-nine drugs—or 23%—met none.

The report was carried out over two years by four experts—three physicians and one postdoctoral fellow at Harvard—and they sifted through “government reports, internal FDA documents, investigators’ notes, congressional testimony, court records, and interviews with more than 100 researchers, legal scholars, current and former federal officials, patients, and their families.” The experts’ findings were then vetted by a 14-member advisory committee that included an FDA insider and an FDA advisor.

The experts themselves and the committee were shocked by their findings.