Hydroxychloroquine vs. REGN-COV2

If you have COVID-19, your doctor can prescribe for you hydroxychloroquine. Its only FDA-authorized use is for malaria, and the FDA has cautioned against the use of hydroxychloroquine for COVID-19. But it's up to you and your doctor, whose only legal worry should be the prospect of malpractice liability.

As of this writing, your doctor cannot prescribe for you REGN-COV2 (though Trump has promised to issue an Emergency Use Authorization). That's the antibody treatment that President Trump received. Preliminary data suggests that this drug is promising. Another antibody treatment has produced even more promising though still preliminary results, but you can't yet get that one either.

What accounts for the asymmetric treatment of hydroxychloroquine and REGN-COV2? Hydroxychloroquine is approved to treat malaria, and in the United States, once a drug is approved, doctors may prescribe it off-label for other uses. Doctors are advised to weigh the risks and benefits of off-label uses, but the matter is ultimately left in their discretion. Some commentators have observed that much off-label use is not based on scientific evidence and others have called for government regulation to quantify risks and benefits of off-label use. Yet another says that off-label use "is the death of the idea of regulation."

But the idea of regulation is alive and well. We still prevent pharmaceutical companies from selling drugs like REGN-COV2, which have not been established to be safe and effective for any use (but which a rational doctor might well prescribe in some cases). One argument in defense of this regime is that once a drug is approved, it has been established that it is safe, so there is no harm from off-label use. But that can't be quite right. The FDA compares the costs and benefits of a drug. That cost-benefit analysis will  focus especially on the condition for which the drug is approved. Just because the drug is safe enough to be used for that condition does not mean that the cost-benefit analysis would come out the same way as to other indications.

The discrepant treatment of off-label drugs on the one hand and promising-but-not-yet-proven drugs on the other requires some justification. The status quo could reflect a compromise between those who would prefer less FDA regulation and those who prefer more, but this seems unlikely, because the discrepancy is rarely noted. Another theory is enforceability. Once we allow a drug on the market, it's too difficult to enforce off-label use prohibitions, so we don't even try. But if that were the reason, we would expect at least a nominal little-enforced prohibition. Many doctors would follow the law, even if violations were difficult to detect.

Maybe the discrepancy arises from the need to encourage clinical trials. I recently argued here that many shy from this justification for FDA regulation more generally, because it means that we are denying people treatments that we (or at least their doctors) think are more likely beneficial than not on the ground that we need test subjects. But at least a utilitarian should not be embarrassed by this argument. Eugene Volokh recognizes the possibility as a possible constraint on a right to medical self defense (p. 27), and others are even more unafraid to make it. It might seem that the need to encourage clinical trials should lead both to prohibition of off-label use and insistence on keeping promising drugs off the market until we can be sure of their efficacy. But drug companies may have reduced incentives to conduct controlled trials of off-label use anyway, and thus, there is less reason to prevent such uses.

REGN-COV2 may present an exceptionally strong case for keeping the drug off the market so that more clinical trials can be conducted. Supplies of the drug are scarce, and it appears that it will take a while to ramp up production. With scarcity, not everyone can obtain the drug anyway, and so we might as well use whatever drug supply exists for clinical trials. If we don't use it all up, but the drug proves effective, we should still find uses for what has been produced so far. Indeed, scarcity is a standard justification for randomization in a variety of contexts, such as charter schools with only so many seats for students. Though interest in hydroxychloroquine caused some shortages, in general, supply is likely to be less constrained for off-label uses of a drug already on the market, especially in comparison to an exotic drug like an antibody that is expensive to manufacture.

Where does this leave someone suffering from COVID-19 who wants the Presidential treatment? You may be able to receive REGN-COV2 under the Right to Try Act, which would more accurately be named the Right to Ask Act, because it authorizes a patient "who has exhausted approved treatment options" to request the drug. Presumably, that is how President Trump received the drug. It's not clear exactly how Trump exhausted approved treatment options, but the Right to Try Act essentially removes FDA oversight from the process. Still, the drug companies may refuse and moreover have strong incentives to do so.

Can a drug company use the Right to Try Act to do an end run around the requirement of FDA approval (or emergency use authorization)? No. The drug must be currently "under investigation in a clinical trial," and the patient must be "unable to participate in a clinical trial involving the eligible investigational drug." These requirements solidify the theory that the goal of our drug regulatory system is to coerce people into clinical trials. If the goal were just doing the best thing for each patient, it shouldn't matter that some other patients are currently participating in a clinical trial for which one is eligible. But if we accept the goal of coercing patients into participating in clinical trials, then the provisions advance that goal, both by requiring the pharmaceutical company to be running a clinical trial before it authorizes a "Right to Try" request and by limiting the emergency use to someone who cannot participate in the trial.