SARS-CoV-2 Vaccine Strain Selection: Guidance From Influenza

Abstract

When first approved, many hoped that the SARS-CoV-2 vaccine would provide long-term protection after a primary series. Waning of immunity and continued appearance of new variants has made booster inoculations necessary. The process is becoming increasingly similar to that used for annual updating of the influenza vaccine. The similarity has become even more apparent with selection of BA.4/BA.5 as the Omicron strain of the updated bivalent (Original + Omicron) COVID-19 vaccines. It is hoped that, if COVID-19 develops winter seasonality, SARS-CoV-2 vaccines will require only annual review to determine if updates are necessary. Recommendations on whom should receive the booster would be based on conditions at that time.

On 28 June 2022, the Vaccines and Related Biological Products Advisory Committee (VRBPAC) of the US Food and Drug Administration (FDA) voted to recommend inclusion of a severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron component in coronavirus disease 2019 (COVID-19) booster vaccines [1]. Among the discussion topics for this meeting, in addition to whether a change in the current COVID-19 strain composition of the vaccine was necessary, was whether the Omicron lineage should be BA.1 or BA.4/BA.5 and whether the vaccine should be monovalent (Omicron) or bivalent (prototype plus Omicron). After review of antibody data from vaccinated and infected individuals, the committee felt that a change was warranted and it preferred a bivalent vaccine containing a BA.4/BA.5 spike as the Omicron component. Accordingly, on 31 August 2022, the FDA approved an amendment to the Emergency Use Authorization of the Moderna and the Pfizer-BioNTech COVID-19 mRNA vaccines to authorize bivalent formulations to contain both the original strain of SARS-CoV-2 and one “in common between the BA.4 and BA.5 lineages” [2].

RECOGNITION OF THE NEED FOR BOOSTER INOCULATIONS

This action, which was not universally accepted, was in response to the continued emergence of variants of the SARS-CoV-2 virus and the reduction in effectiveness of the COVID-19 vaccines due to immune escape and waning in protection over time [3]. The problem had been recognized a year before. Shortly after vaccination, there appeared to be a surprisingly robust protective effect both on clinical disease and even on transmission [4]. There was hope that this vaccine would be like many of our vaccines, that a primary series would elicit long-term protection. The only non-mRNA vaccine to be authorized in the United States during this period, Janssen/Johnson and Johnson's vectored preparation, was said to be “one and done,” in part as a way to appeal to the vaccine hesitant.

In retrospect, there should have been more attention paid to applying the lessons learned concerning reinfections from other respiratory viruses, such as influenza, to SARS-CoV-2. There were even clues in human coronavirus behavior that reinfection was the norm. In the first 8 years of an observational study of the HIVE cohort in Michigan, the mean time to reinfection with any of the 4 seasonal or “common-cold” coronaviruses was 505 days and, for the same type, 983 days [5]. Given this experience with natural infection, it could have been expected that long-lasting immunity would be unlikely to follow immunization. This pattern of regular reinfection is characteristic of other respiratory viruses even when there is little issue of antigenic variation over time.

By summer 2021, infections were occurring in those who received the primary series although by far the greatest number of cases were still occurring in the unvaccinated [6]. Three variants of concern, Alpha, Beta, and Delta had spread globally, adding worry about immune escape [7]. A VRBPAC meeting was called on 17 September 2021 to consider recommending booster inoculations with the Pfizer-BioNTech COVID-19 vaccine for the immune competent [8]. Data were presented from the United States but also from Israel, which had vaccinated a large part of its population with the Pfizer-BioNTech vaccine. At the time, Israel had experienced an Alpha wave, and was then in a Delta wave. Antibody was waning in all age groups but there was still reasonable cross-neutralization of the Wuhan virus with those circulating. The primary series continued to protect against severe disease, but that protection was waning at a slower rate than against milder disease, particularly in older individuals. As a result, Israel had already initiated booster inoculations on 30 July 2021, starting with the elderly. At the time of the VRBPAC meeting, the vaccination program was already showing success [9, 10].

The sponsor, Pfizer-BioNTech, asked for approval for those 16 years of age and older. Immunologic data were presented covering the entire age group. At the time, data on the need for the booster to prevent severe disease in younger individuals outside Israel were limited. In addition, the data on myocarditis/pericarditis as a vaccine-induced adverse event were just emerging but it was already clear that it was mainly occurring in adolescent and young adult males [11, 12]. There was also a question raised about the methodology used in observational studies, the basis for much of the real-world assessment of effectiveness [13]. As a result, the committee voted unanimously to authorize a booster dose for those 65 years of age and older and those in special situations.

The FDA subsequently, on 20 September 2021, authorized the Pfizer-BioNTech vaccine for these population groups. Authorizations followed for Moderna and for the Janssen vaccines, really a second dose for the latter [14]. As more data on the need for the boosters to prevent severe disease as well as minor illnesses accumulated, approvals for the booster were extended to younger individuals.

EFFECT OF OMICRON VARIANTS

The first Omicron variants began to spread globally before the end of 2021 [15]. There was always difficulty in separating out how much of the decline in vaccine effectiveness was due to waning of immunity and how much to antigenic change in the circulating viruses. However, the latter appeared to be handled by the titers produced [16]. The emergence of the original Omicron variant (B.1.1.529) represented a major change in the sequence and antigenicity of the spike protein. It was clear that there was little protection against this virus with the primary series but that the booster restored some degree of protection [15, 17]. However, that protection was reduced, as was already shown in data from Israel. Infections were frequently observed even in those who had received the boosters. As a result, a fourth inoculation was approved in Israel [18]. The United States followed on 29 March 2022, with the authorization for the booster dose of both mRNA vaccines for immunocompromised and older individuals.

The Omicron variants have continued to evolve, starting with BA.1 and now with BA.4/5 and additional subvariants. These lineages are antigenically different from each other, but not as different as all Omicron viruses are from previous strains [19]. This meant that, with further waning, as summer ended in the United States, many months had elapsed for the vaccinated population since their previous dose, one that still contained the ancestral strain. The process of developing booster vaccines containing the Omicron strain started well before the VRBPAC meeting of June 28 [20]. With the knowledge that new variants would continue to emerge, the aim was production of the greatest breadth of immunity. That consideration and the fact that it had proven so effective explained the continued inclusion of the Wuhan strain in the bivalent vaccine. The selection of the Omicron variant representing BA.4/BA.5 was also based on producing broad immunity. BA.1 was incorporated into bivalent vaccines being used outside the United States but was not chosen because that variant is no longer circulating and there is a suggestion that immunization with it may not achieve the broad-spectrum protection that is desired [21–23]. As a result, a decision was made to proceed with the selection of a bivalent vaccine containing the original vaccine and the Omicron variant BA.4/BA.5, despite the lack of clinical immunogenicity data in humans [21].

THE STRAIN SELECTION PROCESS

It had been predicted, on the basis of waning of immunity and continued emergence of variants, that the vaccine would need updating and the process could resemble that for influenza [24]. Table 1 summarizes the current steps used by the World Health Organization in determining whether and how the influenza vaccine will be updated. Global surveillance identifies currently circulating influenza variants. Their antigenic relatedness to previously circulating viruses is determined using animal and human sera. Strain selection by the committee is based on a combination of immunologic, virologic, and molecular data. The new strain is not tested in humans in advance of being included in the updated vaccine because it is similar in many respects to the strain it replaces. A similar process has now been used for COVID-19 vaccine authorization.

The model of influenza strain selection has successfully worked on a global basis for many years but is clearly not perfect, especially because of the delay between recognition of the updated virus and availability of vaccine. That delay sometimes results in the inability to get the most recent virus into the vaccine. The mRNA technique has substantially reduced the time between strain selection and vaccine availability. The timing of seasonal influenza vaccination programs is well established, and it cannot be given too early in the autumn because of the documented waning of vaccine effectiveness. This experience provides a model for eventual seasonal vaccination against SARS-CoV-2. Thus far, no seasonality of SARS-CoV-2 has been established. In contrast, it is clear with the common-cold coronaviruses, which spread almost exclusively from December to April (Figure 1) [25]. Scheduling the Omicron booster for the northern hemisphere autumn is being done in the hope that protection will last through the winter season when incidence is expected to increase. If, as is hoped, such predictability emerges for SARS-CoV-2 circulation, the selection of viruses for the COVID-19 vaccines will more completely resemble that of influenza.

Notes

Acknowledgment. The authors acknowledge Jackie M. Katz for review of the methods used by World Health Organization for influenza strain selection.

Financial support. This work was funded in part by the National Institute of Allergy and Infectious Diseases (grant number 75N9302.1C.00015).

References

Author notes