The gradual rollout of the Pfizer and Moderna mRNA vaccines and the likelihood of additional approvals in the coming months provide hope for bringing the covid-19 pandemic under control in the USA during 2021. However, it is difficult to immunise many millions of people, particularly as mRNA vaccines are logistically complex to transport, store, and administer. With infection, hospitalisation, and death rates in the US as high as ever, we need to successfully immunise as many people as possible, as soon as possible. Accordingly, there are proposals to change the FDA-approved vaccine regimens to try to increase the number of people protected, specifically by halving the number of vaccine doses given to each person. I think doing this would be imprudent.

The mRNA vaccines are given in two doses either three (Pfizer) or four (Moderna) weeks apart. Placebo-controlled efficacy trials proved that these procedures confer ~95% protection. The data were reviewed by the FDA’s independent experts, leading to Emergency Use Authorizations last month. The argument to now use only one vaccine dose is based on data sets in the FDA briefing documents showing that rates of infection in the vaccine and placebo groups start to diverge ~10-14 days after the first dose, before the second dose is given. The Moderna document states: “There appears to be some protection against covid-19 disease following one dose.” The Pfizer document says that vaccine efficacy after one dose was 52.4% (95% Confidence Intervals 29.5%, 68.4%), while the corresponding value for Moderna was 80.2% (95% Confidence Intervals 55.2%, 92.5%). However, there is an important caveat in the Moderna document: “The small, non-random samples and short median follow-up limits the interpretation of these results”. And both briefing documents state that the available data “cannot support a conclusion on the efficacy of a single dose of the vaccine”. Given the importance of making evidence-based decisions, this is not a comment that should be ignored. 

Another notable comment appears in the Moderna document: “These data do not provide sufficient information about longer term protection beyond 28 days after a single dose.” In short, we do not know for how long any benefit of a single dose might persist. It is widely thought that mRNA vaccines confer protection by inducing virus neutralizing antibodies (NAbs). An important measure of vaccine performance is the NAb titer in the recipients’ sera. NAb responses to the Pfizer and Moderna vaccines are weak after the first dose but strongly boosted by the second. About half of the recipients have no detectable NAbs after the first dose of the Pfizer vaccine, but the mean titer in adults aged 18-55 increases from 14.4 to 360.9 after the second dose (Figure 6 of the FDA briefing document). For Moderna, the corresponding titer increase is from 18.2 to 343.8. The boosting effect of the second dose can be seen graphically here, as can the variable and transient nature of the NAb response to the first dose. Overall, the second mRNA vaccine dose increases antibody titers by ~20 to 30-fold, and perhaps as much as 100-fold when technical aspects of the assay read-out are factored in.

Antibody titers to the mRNA vaccines gradually diminish over time. We do not yet know what the decrease will be over the long-term, and nor do we know when the extent of the drop will reduce vaccine efficacy. It is reasonable to assume, however, that the stronger the initial degree of protection (i.e., NAb titers), the longer it will be sustained. I am not confident a single dose mRNA vaccine would be effective for long enough; a regimen that only worked for a short period would be of limited value to individuals or populations. A similar problem arises if the interval between the two doses were extended to allow more people to receive their partially protective first dose sooner. Some flexibility in timing the second dose would be appropriate, but vaccine failures will surely increase with the length of the delay. Would uptake take a hit if there are multiple reports of vaccine failures among at-risk individuals? That would be contrary to the goal of the one-shot concept. It would also be highly problematic if individual recipients of a single vaccine dose felt protected and increased their infection risk by behaving differently. Furthermore, there is a risk that poorly protective antibody response could drive the emergence of resistant variants in infected vaccine recipients. Already, viral variants with mutations that are suspected to confer some resistance to human antibodies are circulating. Large populations of inadequately immunized individuals provide an ideal environment for the enrichment and further evolution of these variants. The consequences to the overall vaccine effort would be very serious if this happened.

It could be hard to persuade the public to switch to an inferior, one-dose regimen. Asking individuals to reduce their own protection for the sake of protecting a stranger requires quite a degree of altruism. Clearly, some people would choose to do that, but many will not be so selfless about helping others. Predictable controversies are best avoided, particularly given the need to elevate public confidence in vaccines and maximize their use. 

In the USA, the FDA has come out strongly against any changes to the approved ways to administer the Pfizer and Moderna vaccines. I agree with the position it has taken. The UK has taken a different perspective, recommending delaying the second dose of the Pfizer and AstraZeneca and even allowing different vaccines to be used for each dose in “extremely rare occasions”. The clinical trial evidence to support these new policies is flimsy or non-existent, so there are real risks involved. The WHO have said that there is no scientific evidence for  delay of more than six weeks in administering the second dose of the Pfizer vaccine, although they understand why the UK might consider this given the rise in covid-19 cases. 

Are there better options that the UK should consider instead of the dose-delay strategy? An alternative way to double the number of people immunized is to halve the amount of vaccine given in each of the two doses. Here, there is some support based on dose-ranging studies in the Phase 1 trials of the Pfizer and Moderna vaccines. Extrapolating from the data for the approved doses (30 µg and 100 µg, respectively) suggests that halving the dose would reduce the NAb titers by ~2-3 fold, a much smaller decrease than is the case when only a single full-dose is given. While there can be no certainty that efficacy would be unaffected, this strategy seems to be a safer option than dropping, or seriously delaying, the second dose. It should be noted, however, that the FDA has also rejected this option for use in the USA, for reasons that I agree with.

One scenario where a one-dose regimen could be valuable involves people who have recovered from Covid-19 and who now wish to be vaccinated. A single vaccine dose may provide a sufficient boost to their infection-induced and already substantially protective antibodies. The data on this idea may already exist from the completed clinical trials, as some previously infected participants were vaccinated. If not, it could be tested by monitoring the antibody responses in convalescent people who are now receiving the approved vaccines. As 20 million Americans and nearly 3 million Britons are already known to be infected, a substantial number of vaccine doses could be saved if each of them required only one. Testing this idea clinically by measuring antibody titers in a few tens of people would not be onerous.

It is in everyone’s interests to successfully immunize as many people as we can, and as soon as the vaccines become available. Creative thinking on how to improve the process is certainly appropriate. However, it would be best to consider all aspects of what is involved, followed by a data-driven and transparent decision-taking process. The mounting controversy in the U.K. over poorly explained and arguably ill-judged decisions on how best to use the approved vaccines is something best avoided.

John P. Moore is a professor of microbiology and immunology involved in vaccine research at Weill Cornell Medicine in New York City.

Competing interests: none declared