Delayed second dose of the BNT162b2 vaccine: innovation or misguided conjecture?

We strongly support vaccination against COVID-19 with the Pfizer-BioNTech COVID-19 mRNA vaccine BNT162b2 when adhering to the 3-week dosing schedule that was found highly effective in the phase 3 randomised clinical trial—regarded as the gold standard. However we do not support the second dose being delayed to 12 weeks, as implemented by UK Chief Medical Officers.1, 2, 3 The latter followed recommendations by the Joint Committee on Vaccination and Immunisation (JCVI), based on unplanned, retrospective analysis and unwarranted assumptions.

The UK is currently the only country to have adopted a maximal 12 weeks delay. How science-led is the UK strategy? Is it innovative and world-leading, or scientifically fallacious, resulting in an unproven dosing schedule introduced without fully informed patient consent? What are the potential risks, for individuals and the population?

The idea of protecting more of the population by delaying the second dose is predicated on a joint statement by the JCVI and Public Health England (PHE) that a first dose provides 89–90% efficacy (protection).4, 5 This is contrary to the clinical data or efficacy generated from real-life clinical observational data from Israel.6, 7 How did the JCVI arrive at their estimate of 89%?

The JCVI performed an unplanned, retrospective analysis of the randomised clinical trial data. They compared COVID-19 cases in the vaccine group versus the control group from a 6-day window (15–21 days), selected retrospectively after examining the data. The resulting 89% efficacy (95% CI 52–97) was based only on roughly 20 events. Retrospective analyses in therapeutic trials can be hypothesis-generating but should not be used to treat individuals. The JCVI then made a major assumption that the 89% effectiveness persists from day 21 to day 85 in the absence of the second dose,4, 5 for which no empirical evidence was adduced. In a further major, incorrect assumption, the JCVI stated “There is currently no strong evidence to expect that the immune response from the Pfizer-BioNTech and AstraZeneca vaccines differ substantially from each other”. No scientific data on mRNA vaccines exists to support this assumption. The available quality peer-reviewed, published immunology data would refute the assumptions documented by the JCVI and PHE.

mRNA vaccines had never been used therapeutically in humans. The JCVI assumption that mRNA vaccines (BNT162b2 and Moderna's mRNA-1273 SARS-CoV-2 vaccine) would behave similarly to the AZD1222 viral vector DNA vaccine developed by the University of Oxford and AstraZeneca is not supported by published evidence.8, 9, 10 Phase 1/2 trial data of AZD1222 show a substantial specific anti-virus spike protein T-cell responses at day 7, which peaks at day 14. This response is not seen with BNT162b2. Furthermore, there are marked quantitative differences in the production and duration of neutralising antibodies (NAbs). The mRNA vaccines show marked falls in NAb titres (compared with the DNA vaccine) in the period before the scheduled second dose (day 22 and day 29 for BNT162b211, 12 and mRNA-1273, respectively), something we have specifically highlighted as occurring in all age groups. Inevitably, NAb titres will continue to fall during days 21–85, leading to very reduced immunity and increased risk to individuals of infection, especially in frail older people.

An efficacy of 52·4% was reported out to day 22 for BNT162b2, and efficacy of 50–60% has been reported in observational cohort studies from Israel covering the same period.6, 7 UK's delayed second dose strategy for BNT162b2 is, in our view, a misguided conjecture. It will yield some protection for the individual after a first dose: how much, and for how long, is unknown and without patient consent.

The population risk is that the UK's delayed second dose could strongly favour the emergence of consequential SARS-CoV-2 variants resulting from sub-optimal or partial immunity. The Government's Scientific Advisory Group for Emergencies has also documented concern about emergence of variants as a result of the delayed second dose.

Sub-optimal vaccination will create selective pressure facilitating the emergence of vaccine-resistant variants, which could result in a persisting pandemic. New vaccines, covering such variants, can be made but will require time for testing, mass production, and distribution.

We have no concerns regarding the second dose of AZD1222 at 12 weeks, as this is supported by evidence. However, if escape variants arise due to sub-optimal dosing with BNT162b2, they will likely be resistant to other vaccines that target the same viral spike protein.

In conclusion, we would strongly recommend that the UK Government reverts to the two doses in a 3-week schedule (94% efficacy) for BNT162b2; or, as recently supported by WHO and the US Centers for Disease Control and Prevention, adopt no more than a 6-week delay to the second dose “in exceptional circumstances”.15, 16