As expected, on Thursday the FDA’s panel of outside advisors recommended that Pfizer / BioNTech’s vaccine candidate, BNT162b2, be granted an emergency use authorization. The vote was 17 to 4 with one abstention.

The committee that met (all day) yesterday is called the Vaccines and Related Biological Products Advisory Committee, or VRBPAC. Among its members are clinicians, epidemiologists, infectious disease doctors, and statisticians. Historically, the FDA has almost always followed the advice of its outside experts. The NY Times reported that the agency plans to meet today and to issue an EUA tomorrow, Saturday December 12. This morning the FDA itself issued a press release essentially promising that an EUA would be forthcoming as quickly as possible.

The US will be the fourth nation to authorize Pfizer’s vaccine for emergency use. It follows Bahrain and the UK, which authorized the shot last week, and Canada, which gave the nod on Wednesday.

On the same day the VRBPAC met, the New England Journal of Medicine published Pfizer’s detailed pre­sen­ta­tion of their Phase III results. This represents the peer-reviewed, scientific record of the data the company provided to the FDA on November 20 and which the agency has been poring over, taking apart, and validating in the weeks since. The FDA issued a favorable report on the submission earlier this week.

The issues that came up in the VRBPAC’s deliberations yesterday included three of a medical nature and one involving the protocol for going forward with Pfizer’s Phase III trial (and others).

Medical issues

The first issue was the possibility of serious allergic reactions to the shot. Two people in Britain suffered such reactions after being vaccinated, and the UK regulators issued a warning that anyone with a history of strong allergic reactions should not take the Pfizer vaccine at this time. Under questioning, a Pfizer representative said that people with such a history had been excluded from its US trial, so they had no data on this point. One committee member expressed concern that news of allergic reactions could dampen the public’s will­ing­ness to take the vaccine. (By one estimate over 1% of the US population, more than 3.6 million Ameri­cans, were prescribed EpiPens in 2015.) An FDA representative downplayed concern for the issue, saying that the EUA could require that shots be given at a facility with access to treatment for anaphylaxis.

The second medical issue involved people who are or may become pregnant, or are breastfeeding. They were excluded from the study, as is often the case in medical trials. This could be a big issue, because 75% of health-care workers are women — an early cohort slated to receive the vaccine — and most are in their child-bearing years. By one estimate 330,000 people in this population are pregnant at any one time. A Pfizer representative said that the company is conducting animal studies and expects to have preliminary data on safety and ef­fi­cacy in January; human trials could follow. An FDA representative said that it would be acceptable to offer the vaccine to pregnant people and leave the decision up to their own calculation of the risks.

The third issue involved a paucity of data on the vaccine’s safety in 16-17-year-olds. The question the FDA asked the committee to vote on was whether to recommend use of the vaccine in those 16 and older. The data were thin because few people in that age group develop symptomatic illness from SARS-CoV-2, and symptomatic illness was the trigger for being counted in the study. In the end this issue may have accounted for most of the “no” votes in the final tally.

Trial protocol issue

What happens to Pfizer’s Phase III trial once an EUA is issued? How will it affect other ongoing trials, and those yet to begin? The problem is that, if trial participants who had received a placebo are offered the auth­or­ized vaccine (as Pfizer feels obligated to do), then the trial becomes “unblinded” and any chance to gather more data on vaccine safety and efficacy is compromised. The FDA wants companies to gather data for as long as possible.

One way out was detailed by a clinical researcher at Stanford: the so-called “crossover.” (We touched on this idea in last Friday’s update.) Here’s the concept: The trial remains blinded. For everyone in the trial, once they become eligible to receive the auth­or­ized vaccine, they get a series of two shots like everyone else. But those who had unknowingly received the placebo get the actual vaccine; those in the vaccine arm get two shots of placebo. That way everyone can feel confidant they are protected, but some data collection in the study would still be possible.

Pfizer representatives pushed back against the idea, saying it would be impractical to implement.

Next steps

The CDC’s Advisory Committee on Immunization Practices, which two weeks ago issued preliminary priorities for who should get a vaccine first, will meet to decide whether it recommends that the vaccine be deployed for use. ACIP has scheduled a meeting for today, and plans to vote during a meeting set for Sunday.

The VRBPAC will gather again next Thursday, December 17, to consider issuing a similar recommendation for Moderna’s candidate, mRNA-1273. We should see the FDA’s analysis of that company’s submission before then.

Other coverage of yesterday’s VRBPAC meeting:

• Associated Press
• NBC News
• Ars Technica