To us – and we suspect to most of the exceptional public servants employed by the FDA – Sunday, August 23, 2020 will be a day they would rather forget. On that day the Agency issued an “emergency use authorization” (“EUA”) for what is known as “convalescent plasma” for the treatment of COVID-19, notwithstanding a relative dearth of scientific support for the effectiveness of this treatment for this indication. Beyond the plasma treatment itself, which at least doesn’t seem dangerous, the FDA’s action – and particularly its gross overstatement of the treatment’s effectiveness – stoked fears that this action was driven by overt political pressure. That has frightening implications. Is this the first step down the path to Lysenkoism for the Agency?
In the not-too-distant future, when the FDA is called upon to evaluate the safety and effectiveness of an actual COVID-19 vaccine, will the Agency follow the Russian example? That could cause a true disaster, should a rushed, unproven vaccine prove ineffective, while at the same time having crowded out more efficacious competitors, and poisoned public opinion against COVID-19 vaccination generally.
We note that the FDA’s apparent scientific surrender came one day after the president’s now notorious tweet that claimed the Agency harbored unidentified “deep state” personnel that were delaying agency action to cause “political” damage. Just about everyone outside of the president’s base finds that accusation “baseless” or worse.
First, a little background. Despite the hype (where have we seen this before?), convalescent plasma is neither a cure nor a breakthrough. It’s simply blood plasma (the not-red part of blood) donated by people who have recovered from COVID-19. Such donations (for other purposes) have been routine for decades, and it is a “longstanding treatment strategy for infectious diseases that involve the respiratory system.” The FDA has provided some background:
Passive antibody therapy, including convalescent plasma, has been proposed or used to treat a wide variety of infectious diseases for more than a century, including several respiratory viral illnesses such as influenza, Respiratory Syncytial Virus (RSV), Severe Acute Respiratory Syndrome (SARS), and Middle East Respiratory Syndrome (MERS). . . . Passive immune therapy . . . is thought to be most effective when administered prophylactically. . . . However, well-controlled studies in this field are rare.
“Clinical Memorandum,” concerning “COVID-19 Convalescent Plasma,” at p. 3 (amended Aug. 23, 2020) (footnotes omitted). In this case, the plasma contains antibodies that the once-infected donors’ immune systems created to combat COVID-19.
Even as to COVID-19, this treatment has been around for (relatively) some time. Tens of thousands of people infected with COVID-19 have already been treated with it, but not in any scientifically controlled fashion. This extensive use is why, unlike the previously promoted off-label use of hydroxychloroquine, the plasma therapy does appear to be safe. High risks would already have been identified. But the FDA’s 8/23 emergency use authorization does not cite to controlled studies, or to any COVID-19-related study of any sort, that compare this treatment to placebo.
The FDA’s more comprehensive memorandum mentions a couple of more general controlled studies, but dismisses favorable results as “low quality” and potentially biased:
A systematic review of passive antibody therapy for SARS coronavirus (SARS-CoV-1) and severe influenza found a trend towards reduction in mortality, but noted that studies were commonly of low or very low quality, lacked control groups, and were at risk of bias.
Clinical Memorandum, at 6 (footnote omitted); id. at 7 (describing randomized studies as “underpowered”). That no reliable studies exist is not surprising, since use of convalescent plasma during the pandemic has been therapeutic – to try to treat the damn disease. But the FDA’s sudden willingness to authorize use in an admitted data vacuum is troubling, particularly given the timing.
The Mayo Clinic, which has been coordinating existing efforts, and also studying the statistics, currently concludes only:
The relationships between reduced mortality and both earlier time to transfusion and higher antibody levels provide signatures of efficacy for convalescent plasma in the treatment of hospitalized COVID-19 patients. This information may be informative for the treatment of COVID-19 and design of randomized clinical trials involving convalescent plasma.
Joyner, et al., “Effect of Convalescent Plasma on Mortality among Hospitalized Patients with COVID-19: Initial Three-Month Experience,” at 4 (Aug. 12, 2020) (non-peer-reviewed), available here.
The FDA concluded that this Mayo Clinic data lacked both “precise performance characteristics” and “gold-standard methodology,” which “were not available at the time of this review.” Clinical Memorandum, at p.10. As to the effectiveness of convalescent plasma to treat COVID-19, the FDA’s evaluator concluded:
Considering the totality of the scientific evidence summarized above, I agree that current data support the conclusion that CCP [convalescent plasma] to treat hospitalized patients with COVID-19 meets the “may be effective” criteria for issuance of an EUA. Adequate and well-controlled randomized trials remain nonetheless necessary for a definitive demonstration of CCP efficacy and to determine the optimal product attributes and the appropriate patient populations for its use.
Clinical Memorandum, at p.11.
However, that caution-filled go-ahead was nowhere in evidence at the 8/23 press conference. Rather, everyone who spoke – including the FDA commissioner himself − grossly overstated the effectiveness of convalescent plasma by claiming that it reduced the COVID-19 death rate by 35%. That simply ain’t so, and belatedly, days after the damaging misperception was created, the commissioner admitted it. Given the number of people who have already been treated with convalescent plasma, something as dramatic as a 35% overall decrease in mortality would have stuck out like a sore thumb and would have been detected long ago by even rudimentary statistical analysis. Instead conference participants inflated what was merely a limited relative risk differential found when use of plasma was compared to no equivalent treatment. Specifically, we think that the 35% was cherry-picked from the Mayo Clinic data, most likely from this aspect of Mayo’s analysis:
[T]here is a consistent signal of a protective effect of the high antibody levels across the strata. The pooled, or common, relative risk for 7-day and 30-day mortality were 0.65 (0.47 to 0.92) and 0.77 (0.63 to 0.94).
Effect of Convalescent Plasma, at 12. Thirty-five percent is the difference between the 7-day mortality relative risk and one (1 – 0.65). But that was only one (and the most optimistic) “alternative analytical approach” in the Mayo Clinic non-peer-reviewed data, and was limited to: (1) higher versus lower doses in (2) persons who were not that severely affected by COVID-19, and (3) who received the transfusion within three days of diagnosis. Id. at 12-13.
The FDA thus engaged in nationally televised statistical exaggeration that, if attempted by any of our clients, would have resulted in a warning letter or worse. The Agency should be ashamed of its performance, particularly given the overt political pressure immediately preceding the action, the magnitude of the misstatement, and the Agency’s significant delay in issuing a correction. The scientifically-based verdict remains:
Randomized controlled trials are required to show definitive evidence of safety and efficacy and to determine the optimal product attributes and appropriate patient populations for the use of COVID-19 Convalescent Plasma.
Clinical Memorandum, at p.14.
This is hardly the first time there has been evidence of the FDA making regulatory decisions based on political, rather than medical, considerations. This Blog has chronicled the fraught origin of the now-dead FDA “final rule” that sought to change drug reporting requirements to eliminate preemption as applied to generic drugs. That was a blatant attempt to mollify the plaintiffs’ bar in accordance with the political leanings of the prior administration – and we did not hesitate to call it such:
[F]rom our perspective, anyway, it appears the FDA suffered what is sometimes called “agency capture” by pro-plaintiff elements. . . . [T]he FDA did what the other side wanted and issued a proposal to change its regulations designed to thwart preemption in product liability litigation.
What happened the other day, however, is much worse. While the FDA “final rule” sought to jigger agency processes to help the plaintiffs’ bar thwart preemption, that was the extent of it. That was bad enough, as we pointed out repeatedly, but the “final rule” only attempted to revise certain internal procedures to assist a politically powerful interest group then allied with that administration. The FDA’s core mission – using science to insure the safety and effectiveness of prescription medical products – was unaffected by that unsuccessful “final rule.”
Not this time.
On August 23, 2020, the leaders of the FDA at the very least gave the appearance of jettisoning science, and did so seemingly for political reasons. They did this, not just in the absence of sound data, but over the objections of many scientists, including Dr. Anthony Fauci, and the National Institutes of Health as a whole. Further, the Agency did so at a politically chosen time, and in a politically charged setting. The FDA approved, for unlimited emergency COVID-19 treatment, a substance supported by no generally accepted methodology – let alone the FDA’s longstanding “gold standard” of controlled clinical trials.
Perhaps even worse, with the FDA having let this genie out of the bottle, it will likely be impossible ever to conduct the type of statistically valid controlled trials that the FDA usually requires. As we’ve said before (most recently here), once a treatment modality becomes the medical standard of care, blinded trials become medically unethical. That’s a big problem with off-label uses if they, over time, achieve standard-of-care status. It’s a bigger problem here, where the FDA itself is responsible for the hype that may well push a not-ready-for-primetime treatment prematurely towards general medical acceptance.
Maybe, in this instance, everything will turn out OK from a medical standpoint. Notwithstanding the FDA’s EUA, convalescent plasma remains “an unapproved biological product” that is subject to restricted investigational use, as detailed here. As we recently observed in our comment on the Russian “Sputnik V” vaccine, “[n]ot every cut corner, no matter how brazen, results in a head-on collision.”
But until now, that has never been how the FDA has done things. We always thought of science as the FDA’s unshakable bottom line, but we’re afraid this is no longer the case, at least when the Agency is under COVID-19 political pressure.