Don’t Get Too Excited About a Coronavirus Vaccine
There is widespread anticipation of vaccines to prevent COVID-19 infections so life can “get back to normal.” Some three dozen vaccines, made with a variety of technology platforms, or approaches (naked RNA, weakened or killed viruses, hybrid viruses, subunit vaccines, etc.), are now in clinical trials. Many of these vaccine development programs have been assisted by a White House crash program, “Operation Warp Speed,” which was launched in May.
At the announcement, President Trump said the goal would be to have 300 million doses of a vaccine available by the end of this year. Optimism abounds. According to a June survey of 50 health-care executives and investors, nearly all expected a vaccine would eventually be approved in the U.S., and more than half expected approval to occur late this year or in the first quarter of 2021.
Unfortunately, those predictions may indicate a triumph of hope over experience.
In principle, testing a vaccine might seem simple—you give the vaccine to one group of subjects, a placebo to a second, then compare how many in each group develop COVID-19. In practice, however, the process is diabolically complex. Vaccine development raises many questions: how well does the vaccine work? (None is 100% effective.) How much of the active material (antigen) should be in each dose? Does one dose suffice to elicit immunity, or do you need two (as is the case for the second-generation shingles vaccine) or more (three were required for the first hepatitis B vaccine)? Is the addition of an “adjuvant,” a chemical booster, necessary to elicit a sufficient immune response? How well does the vaccine work in the elderly, who are highly vulnerable to Covid-19 infection but who tend to mount a poor immune response? Does immunity last long enough to make immunizing billions of people worthwhile?
In addition, there are potential, unknown safety issues, particularly with the new technologies being used to make most of the COVID-19 vaccine candidates. For example, one of the leading candidates introduces a messenger RNA sequence (the molecule that tells cells which proteins to synthesize) which codes for, or expresses, a disease-specific antigen (part of the coronavirus’s spike protein). Once produced within the body, the antigen is recognised by the immune system, priming it to fight off coronavirus infection. Other vaccines use a debilitated cold virus that contains a coronavirus antigen, which, when injected, stimulates the immune system. In theory, both of these techniques should work—but unexpected side effects are always a concern with new technologies.
And consider the massive logistical effort required just to organize clinical trials for the necessary large-scale testing. For example, Moderna’s trial of their messenger RNA vaccine will be conducted at 89 sites nationwide. From my experience as an FDA official reviewing the results of clinical trials, various glitches are inevitable: a few investigators will be much slower than others to recruit and inject patients, some of the paperwork will be tardy and/or incomplete, and some patients who don’t meet the entry criteria will be mistakenly entered into the study. All of this causes delays.
Finally, the medical, ethical, and regulatory bar is high for a vaccine intended for vast numbers of healthy people. The medical community and FDA officials still recall the mistaken approval of a vaccine to prevent swine flu in the 1970s, which resulted in four hundred and fifty people developing a serious adverse reaction: the rare, paralytic Guillain-Barré syndrome. What made the situation even worse for regulators (and, of course, for those afflicted) was that the predicted epidemic never materialized—so the vaccine wasn’t even needed.
Regulation is, therefore, especially conservative (risk averse) for vaccines that will be administered to large numbers of healthy people. Before approval, the first successful rotavirus vaccine (RotaTeq) was tested on 72,000 healthy infants; the first human papillomavirus vaccine (Gardasil) on more than 24,000 people; and the newest shingles vaccine (Shingrix) on about 29,000 subjects.
Once burned, twice shy, the old saying goes, and FDA officials will not want to repeat the mistake of the 1976 swine flu vaccine. Before signing off, they will ensure that the evidence of safety and efficacy is solid—which may dampen the optimism about a speedy approval.
PRACTICAL AND POLICY HURDLES TO VACCINE APPROVAL
Despite the practical hurdles (and in view of current presidential poll numbers), we can expect intense pressure on the FDA from the White House to deliver an “October Surprise,” in the form of a vaccine approval, even if it’s premature. The FDA may already have capitulated to similar pressure, when, in March, regulators allowed at least 160 antibody tests for COVID-19 to enter the U.S. market without validation or FDA scrutiny, as the agency rushed to get them to the public. Many of these tests were fraudulent or otherwise flawed, and the FDA subsequently mandated stricter review.
Interestingly, FDA Commissioner Stephen Hahn, formerly the Chief Medical Executive of the MD Anderson Cancer Center in Houston, seems in several ways to have created an insurance policy for himself as a bulwark against such pressure in the future. In a lengthy policy statement, “Development and Licensure of Vaccines to Prevent COVID-19: Guidance for Industry,” published on June 30, his agency specifies, in excruciating detail, the numerous criteria for approval of coronavirus vaccines.
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 (as shown in clinical trials), and a vaccine must be at least 50% more effective than a placebo in preventing the disease. (The 50% benchmark is routinely applied to flu vaccines.) In other words, vaccines that are less effective need not apply. The clinical trials would also need to demonstrate safety, of course.
These criteria are extremely important, because the emphasis on “traditional approval” means that regulators do not intend to consider cutting corners via “accelerated approval” based on “surrogate endpoints,” which are outcomes in a clinical trial that are not themselves a direct measurement of clinical benefit, but are thought to predict clinical benefit. A surrogate endpoint can be a laboratory measurement, radiographic image, or physical sign; for a COVID-19 vaccine, it could be the ability to elicit antibodies to the virus, but without showing actual prevention of infection. (Under certain circumstances, the FDA can grant an “accelerated approval” based solely on surrogate endpoints, but that applies primarily to therapeutics, not vaccines, and the language quoted above—that regulators intend to grant “traditional approval via direct evidence of vaccine efficacy”—appears to foreclose that possibility for COVID-19 vaccines.)
The published guidance to industry enables FDA Commissioner Hahn to fall back on that policy—especially if he is pressured by his superiors to adopt a lower standard. He went out of his way to emphasize the FDA’s independence and integrity on July 21st, tweeting, “Americans should know that we are steadfast in maintaining our regulatory independence & ensuring our decisions for treatments & vaccines for #COVID19 are based on science & data. This is a commitment that the American public can have confidence that I will continue to uphold.”
Less than a week later, on July 27th, Dr. Hahn repeated assurances yet again that the FDA “will not cut corners in order to evaluate a vaccine,” adding that the agency will only grant approvals, “based upon the data and upon the gold standard that we have at FDA regarding the safety and efficacy of a vaccine.”
The not-so-subtle message to his bosses at the Department of Health and Human Services and the White House is clear: don’t bother pressuring me for premature approval. None will be granted until the data warrant them.
The FDA’s guidance to industry contains myriad details to guide vaccine developers, but the most strategically useful to Dr. Hahn is the criterion for efficacy in clinical trials, because it’s a firm benchmark based on the actual prevention of COVID-19 infection. The other details do, however, offer insights into the complexity of vaccine development. They include:
- Testing must be in place to monitor every batch of vaccine at all stages of production, and there must be consistency in manufacturing;
- The production process must be “validated,” which “would typically include a sufficient number of commercial-scale batches that can be manufactured routinely,” and that meet pre-specified criteria for purity, potency, yield, etc.;
- Storage conditions, including “use-by” dates, must also be validated;
- Appropriate preclinical (animal) studies should be conducted, especially to evaluate concerns that the vaccine could actually increase the severity of subsequent COVID-19 infection, which has been seen with other vaccines;
- Clinical trials to demonstrate vaccine efficacy “will likely need to enroll many thousands of participants, including many with medical comorbidities for trials seeking to assess protection against severe COVID-19.”
- Clinical trials should include sufficient numbers of individuals who are elderly or who have medical comorbidities, in order to show that the vaccine is safe and effective for them.
The important message here is that, although regulators intend to streamline regulation and work with industry to facilitate the development of Covid-19 vaccines, they won’t be stampeded by their political masters into exposing Americans to inadequately tested, potentially dangerous products.
There’s an old saying in pharmaceutical development that there are a thousand ways for things to go wrong, which can range from the product just not working or unexpected side effects to a shortage of glass vials of the required quality. In any case, politics and science are not bound by the same time constraints, which means the Trump campaign will have to come up with an alternative October Surprise.