FDA fully approves Paxlovid. Now we can learn how well it really works.

During the Covid-19 pandemic, the FDA used its authority to allow new medications and vaccines to reach the public, without its usual lengthy approval process. That power—emergency use authorization (EUA)—saved a lot of lives by getting new vaccines and therapeutics to people in months, instead of years.

Yesterday, the FDA approved Paxlovid, Pfizer’s anti-viral pill that was shown to decrease the rates of severe and fatal Covid-19.

That won’t affect much in terms of your ability to get the drug, although it may have some unfortunate unintended consequences, including over-prescribing. Drugs that have EUA behind them can only be given for the narrow indicated purposes to the precise populations. With full FDA approval, though, off-label prescribing will be permitted by licensed clinicians.

People may be excited that this means they can now get the drug more easily, (here are guidelines up until yesterday) but I’m not so enthusiastic. I don’t think it’s anywhere close to clear that this drug helps people beyond the high-risk groups for whom the drug was already authorized for use prior to yesterday. There are side effects and adverse effects (like rebound) to consider.

In fact, we don’t really know how well this drug works now. We know it had a major impact on high-risk unvaccinated people with no prior history of infection. But that’s a population that is almost non-existent today.

Most people who take Paxlovid do not resemble the people who volunteered to be in Pfizer’s blockbuster trial. There are signs that the drug helps high-risk vaccinated people, but the effect is smaller and the studies on that, while well conducted— are not from randomized trials. We have no trial data telling us how well Paxlovid works “in the real world” in the year 2023. All we know (other than from observational data which may or may not hold up in well-run clinical trials) is how it did in the pre-vaccine era.

That’s where the FDA approval is good news. Now that Paxlovid has approved, researchers who are independent from Pfizer will be able to do randomized trials. Why? Because under EUA, outside researchers could not get the drug! But now they’ll be able to (assuming their study designs are approved by their own internal ethics boards).

What do we want to know that we can now learn in independent randomized trials with blinded placebo controls? Experts have many questions and there’s much to learn.

Here are some that I have which randomized trials (rather than observational studies based on medical chart reviews) might really answer well:

How well does Paxlovid work in the vaccinated?

Is 10 days of Paxlovid better than 5?

Does “rebound” really happen more often among Paxlovid takers?

How well does Paxlovid work in the immunocompromised? (This might be harder to test, because giving half the volunteers placebo would be unethical).

Does Paxlovid prevent Long Covid?

There are so many questions that Pfizer did not provide answers to. Why? Probably because after its initial trial succeeded so spectacularly, it had no interest in learning anything more. Many people who had heard of the drug wanted it, and Pfizer didn’t want to change that. That’s too bad, because the patient population of 2023 has very different immune histories than the trial volunteers from 2020. The drug likely works less strongly today, because (mostly) we are less vulnerable.

Now, independent researchers will be able to do these studies, whether Pfizer likes it or not. I’m eager to get the results. I predict some polar outcomes. I predict the drug will turn out to be less useful for most “supposedly” high-risk people than many currently believe. On the other hand, I predict longer courses (10 days, for example) for the severely immunocompromised will have important benefits that we have not yet appreciated.

Hopefully we won’t have to wait too long for answers. At least now, it’s possible that we’ll finally get some new high-quality intel on this.