There are growing signs that the Trump administration is preparing to license a COVID-19 vaccine before the November presidential election, alarming scientists who warn that rushing a vaccine into use before large-scale clinical trials have concluded could be dangerous and might dent the public confidence needed for a successful inoculation program.
The U.S. Centers for Disease Control and Prevention has told state health officials to be prepared to begin distributing a coronavirus vaccine to health care workers and other high-priority groups by Nov. 1, days before Americans head to the polls.
Over the past several days, two top U.S. health officials—Anthony Fauci, who heads the National Institute of Allergy and Infectious Diseases, and Stephen Hahn, commissioner of the Food and Drug Administration—have also told news organizations that a vaccine might be made available to some high-priority groups, like health care workers, even before clinical trials are concluded, if the preliminary data from those trials is overwhelmingly positive.
It is thought that President Donald Trump and his advisers are eager to have a vaccine approved in order to blunt criticism of the administration’s COVID-19 response and boost Trump’s chances of reelection.
Vaccine A and Vaccine B
The CDC said it was planning to distribute one or two vaccines but did not identify them, calling them only Vaccine A and Vaccine B in the planning documents sent to state health agencies, according to the New York Times, which obtained copies of the documents. The paper said that descriptions in the documents seemed to match a vaccine candidate being tested by biotechnology company Moderna and another being developed by Pfizer in conjunction with German biotech firm BioNTech.
The Financial Times reported last month that members of the Trump administration had discussed granting emergency use authorization for the COVID-19 vaccine being developed by a team from the Jenner Institute at the University of Oxford, in England, and manufactured by pharmaceutical giant AstraZeneca. In a statement, the company said it has not held any such discussions and that it would be “premature to speculate on that possibility.”
Pfizer has said previously that its vaccine—which is currently undergoing late-stage clinical trials in the U.S., Germany, Argentina, and Brazil—was “on track” for regulatory approval “as early as October.” Moderna’s vaccine is also undergoing late-stage human testing in the U.S.
But what will we know from these clinical trials by November? The answer is far from clear. And there’s a good case to be made that not enough evidence will have been gathered to be certain that a vaccine works and is safe.
Airfinity, a London-based science information and analytics firm, forecasts that the AstraZeneca and Oxford team will be the first to report preliminary data from a Phase III trial, probably around Sept. 19. It estimates that Pfizer will have preliminary data around Oct. 8, and Moderna around Oct. 23. Preliminary results for a fourth vaccine candidate, produced by Chinese firm Sinovac, may also be ready around that date.
The company bases these predictions on when the clinical trials begin, the rate at which they report enrolling volunteers, and the infection rates in the areas where the testing is being conducted. But these are just preliminary readouts; AstraZeneca’s vaccine is the only candidate that Airfinity predicts will have final results ready before Election Day.
Rasmus Bech Hansen, Airfinity’s chief executive officer, says that preliminary readouts from clinical trials are sometimes sufficient to give a clear indication that a drug will be effective, allowing pharmaceutical firms to apply for immediate regulatory approval without having to complete expensive further clinical testing. Conversely, sometimes the early results show that a medicine will clearly fail. But most often, he says, the preliminary readouts simply provide enough positive indications to know that it is worth completing the full clinical trial.
To receive a U.S. license, the FDA has said, a vaccine must provide 50% more protection than a placebo. To determine that, the companies conducting the trials give the vaccine to one group of volunteers and a placebo to a control group and then see who actually gets COVID-19 over the subsequent weeks and months.
But if there aren’t enough infections in the control group, proving that the vaccine works is difficult. In both Moderna’s and Pfizer’s clinical trials, which involved 30,000 people, about 50 infections will need to be reported between both the control group and the inoculated group in order to get an early readout of the results, Hansen says. The trials are not expected to produce final results, however, until three times that number of infections have been recorded among the participants.
That is why it has been essential to conduct trials in places where infection rates remain high. Even in a place like Florida, with about 115 new cases per 100,000 people at the moment, one would only expect about 35 total cases in a cohort of 30,000 trial volunteers—which is why it can take a long time to get enough infections among the trial participants to draw conclusions about whether the vaccine works.
It is possible for the preliminary readouts to produce results that appear promising but aren’t statistically significant. This is why scientists say they would be particularly wary of any move to grant emergency use authorization to a vaccine on the basis of the initial efficacy results from earlier stage human testing, as opposed to the larger Phase III trials.
These early tests often involve 1,000 or fewer people and are mostly designed to provide an indication that a vaccine is safe, not whether it is effective. Nonetheless, these early studies do provide clues to how effective a vaccine might be. One is by studying what is called the “immunogenicity” of the vaccine: What kind of immune response does it seem to produce. This is most often gauged by looking for antibodies to the virus in blood samples taken from the inoculated volunteers. Those results have already been reported for all three leading vaccines, but the lack of standardized antibody tests makes interpreting the results difficult.
‘Vaccine efficacy measure’
Unusually for a Phase I trial, a few of the COVID-19 early human trials—most notably the Oxford Jenner Institute’s study—had an explicit goal of providing an “vaccine efficacy measure.” This is done the same way as in the larger, later-stage clinical trials: Just wait and see how many people in the inoculated group contract COVID-19 compared with the control group. The AstraZeneca-Oxford team is expected to report these figures for its Phase I trial sometime in the next few weeks.
But Ira Longini, a biostatistician at the University of Florida’s Emerging Pathogens Institute in Gainesville who has also written a book on vaccine trial design, told the trade publication Pharmaceutical Technology that the Phase I trials of the Oxford vaccine, with 1,000 volunteers and an “efficacy measure” end point at 30 total COVID-19 infections among this study participants, is too small to produce “clinical significance” and should not be used to authorize a vaccine. In fact, other researchers told the publication that depending on how those 30 infections break down between the control group and the inoculated volunteers, the results might not have any statistical significance.
Steve Pritchard, a spokesman for Oxford University, told Fortune that the idea of using the Phase I trial to demonstrate the effectiveness of the Jenner Institute’s vaccine was based on the idea that there would still have been enough COVID-19 transmissions in the U.K., where the trial took place, to get a good indication of the vaccine’s effectiveness. But this “has changed since the protocol’s inception. We are in Phase III clinical trials currently with higher expected transmission in the U.K. and other populations in the months ahead.”
Safety concerns are another reason why many researchers would prefer that full Phase III trial results are known before any vaccine receives authorization. Some rare adverse reactions to the inoculations may only show up once the vaccines have been studied in a large, diverse population. Others may not even show up until months or years after a person has been vaccinated. Hansen says that vaccines have “an unfortunate history of not being safe,” but that the need for long-term safety studies needed to be balanced against the risks of the pandemic.
But many researchers worry that the Trump administration’s rush to approve a vaccine before the election may backfire—undermining public confidence in the vaccine and maybe any future ones too. “Any negative publicity at all about vaccine efficacy or safety would muddy the waters and undermine confidence in one of the most significant missions humankind has ever faced,” Danny Altmann, a professor of immunology at Imperial College’s Hammersmith Hospital in London, wrote in an opinion piece for U.K. medical journal The BMJ.