There is widespread anticipation of vaccines to prevent COVID-19 infections so that Americans can get their lives back to some semblance of normal. More than 50, made with a variety of technology platforms, are now in clinical trials, 11 in large-scale, late safety/efficacy testing. Vaccines have also assumed unprecedented political importance. It's widely thought that a pre-election announcement of a COVID-19 vaccine approval would boost President Donald Trump's election prospects.
There are several possible routes for vaccines to emerge from the regulatory maze. The two most likely are a conventional biologics license application (BLA) or an emergency use authorization, which must meet a lower bar. Both require a judgment by scientists at the Food and Drug Administration that the product is generally safe and effective, and that the benefits outweigh potential risks.
However, there's an old saying at the agency (where I spent 15 years as a medical reviewer and office director) that all approval decisions are based on incomplete data, some more incomplete than others. In other words, you can always be surer of your decision with more patients treated for a longer time, but at some point, you have to make a judgment call.
An FDA policy statement, "Development and Licensure of Vaccines to Prevent COVID-19: Guidance for Industry," published on June 30, specified in great detail the criteria for FDA approval of COVID-19 vaccines, but the overarching principle is simple:
"The goal of development programs should be to pursue traditional approval via direct evidence of vaccine efficacy" in protecting humans from COVID-19 – through clinical trials – and a vaccine must be at least 50% more effective than a placebo in preventing the disease.
Of course, the clinical trials would also need to demonstrate that the vaccine is safe, which is a context-dependent judgment. (You accept greater risks when the need for the product is great and the benefits are significant.)
The FDA's guidance left the door open for emergency use authorizations (EUAs) for COVID-19 vaccines in certain circumstances but cautioned that, "Issuance of an EUA for a COVID-19 vaccine prior to the completion of large randomized clinical efficacy trials could reduce the ability to demonstrate effectiveness of the investigational vaccine in a clinical disease endpoint efficacy trial to support licensure." In other words, echoing the FDA aphorism about data always being incomplete, an early EUA might create uncertainty about a vaccine's ability to actually prevent infections, as opposed to simply eliciting antibodies, which might be an indirect indicator of efficacy.
The FDA said in the guidance document that it might consider issuing an EUA for a COVID-19 vaccine as an interim measure after studies have demonstrated safety and efficacy, but before a biologics license application (BLA) has been submitted or formally reviewed. It added, "Any assessment regarding an EUA would be made on a case by case basis considering the target population, the characteristics of the product, the preclinical and human clinical study data on the product, and the totality of the available scientific evidence relevant to the product."
The agency has made it clear that it is amenable to issuing EUAs. In an interview published by the Financial Times, FDA head, Dr. Stephen Hahn, said his agency was prepared to authorize a vaccine before Phase 3 clinical trials (the final, definitive stage of testing) were complete, if regulators become convinced that the benefits outweigh the risks.
And on Oct. 6, the FDA issued guidance hours after disclosing some of its advice to vaccine makers in briefing documents for an advisory committee meeting on Oct. 22. An important requirement in the guidance (which will probably delay an approval long enough to prevent a pre-election "October Surprise") is: "Data from Phase 3 studies should include a median follow-up duration of at least two months after completion of the full vaccination regimen to help provide adequate information to assess a vaccine's benefit-risk profile." (It appears that many of the vaccines will require two doses.)
That will provide more confidence in the safety of an EUA-authorized vaccine, which takes us back to the FDA apothegm about the incompleteness of data at decision-time and raises the question, when would enough be enough for an EUA but not for the conventional, full approval of a BLA?
The likely scenarios that would justify an EUA would be if the data from tens of thousands of subjects showed compelling evidence of safety and efficacy, but: (1) the trials had not yet reached the target number of subjects, especially in certain important sub-populations, such as the elderly, minorities, and people with common comorbidities; and/or (2) there weren't sufficient, medium-term data on adverse events (i.e., safety); and/or (3) there were uncertainties about the ability of the vaccine manufacturer to reproducibly produce batch after batch that meet purity, potency, and sterility standards.
The last of these is a factor that has been largely omitted from public discussions of COVID-19 vaccines, and yet, for several reasons, is extremely important. First, the facilities that produce many of these vaccines at large-scale are brand new and relatively unproven, while in some cases, production will occur abroad. Historically, there have been notorious examples – extremely rare, fortunately – of the distribution of contaminated or otherwise substandard vaccines.
The FDA's career scientists and physicians are competent to evaluate the manufacturing and clinical data, and they understand the urgent need for COVID-19 vaccines. Whether the agency's political appointees will be able to resist the intense political pressure for pre-election vaccine approvals remains to be seen.
Henry I. Miller, a physician and molecular biologist, was a research fellow at the National Institutes of Health and the founding director of the FDA's Office of Biotechnology.