Clinical development of variant-adapted BNT162b2 COVID-19 vaccines: the early Omicron era

ABSTRACT

Introduction

The Omicron BA.1 variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and subsequent sub-lineages exhibit partial escape from neutralizing antibodies elicited by vaccines containing or encoding wild-type spike protein. In response to the emergence of Omicron sub-lineages, variant-adapted vaccines that contain or encode for Omicron spike protein components have been developed.

Areas covered

This review presents currently available clinical immunogenicity and safety data on Omicron variant-adapted versions of the BNT162b2 messenger RNA (mRNA) vaccine and summarizes the expected mechanism of action, and rationale for development, of these vaccines. In addition, challenges encountered during development and regulatory approval are discussed.

Expert opinion

Omicron-adapted BNT162b2 vaccines provide a wider breadth and potentially more durable protection against Omicron sub-lineages and antigenically aligned variants when compared with the original vaccine. As SARS-CoV-2 continues to evolve, further vaccine updates may be required. To facilitate this, a globally harmonized regulatory process for the transition to updated vaccines is needed. Next-generation vaccine approaches may provide broader protection against future variants.

KEYWORDS:

• COVID-19
• SARS-CoV-2
• BNT162b2
• variant of concern
• Omicron
• bivalent vaccines

1. Rationale for variant-adapted COVID-19 vaccines

Following the emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in December 2019, multiple variants have been reported. Those with an increased risk to public health are categorized as variants of concern (VOCs), variants of interest (VOIs), or variants under monitoring (VUMs) [1]. To date, five VOCs have been identified: Alpha, Beta, Gamma, Delta, and Omicron [1]. These VOCs have been associated with high transmission and infectivity, driven by mutations in the viral genome that enhance viral fitness and mediate evasion of neutralizing antibodies [2]. As original coronavirus disease 2019 (COVID-19) vaccines were developed based on genomic sequences from the Wuhan-Hu-1 wild-type or closely related viruses, the continuing evolution and emergence of VOCs has the potential to impact vaccine effectiveness.

The Omicron BA.1 variant, which was first detected in November 2021 [3], represented a new VOC that was phylogenetically distinct from prior lineages [4]. Compared with the Wuhan-Hu-1 wild-type virus, BA.1 contains 53 mutations, including 30 non-synonymous substitutions in the spike (S) protein [5]. This resulted in enhanced intrinsic transmissibility versus previous VOCs [6,7] and partial escape from neutralizing antibodies [8]. Several other Omicron sub-lineages have since emerged, and intra-variant diversity is increasing. Sub-lineages BA.4 and BA.5 replaced BA.1 and BA.2; descendants of BA.5, including BQ.1 and BQ.1.1, and direct BA.2 descendent sub-lineages, such as BA.2.75 and CH.1.1, continue to circulate [1,9]. As of 14 May 2023, recombinant variants, such as XBB, which is a recombinant of two BA.2 sub-lineages [10], and its descendants XBB.1, XBB.1.5, XBB.1.16, and XBB.1.9.1 are now the most prevalent globally [9]. In particular, XBB.1.5 seems to have a growth advantage over other circulating Omicron sub-lineages and has become dominant, accounting for 44% of all submitted sequences [9,11].

BA.4 and BA.5 are descended from sub-lineage BA.2, which is antigenically distinct from BA.1 [12,13]. The antigenic distance of BA.4 and BA.5 from BA.1 is similar to that between the Delta variant and wild-type SARS-CoV-2 [14]. Both BA.4 and BA.5 exhibit increased fusogenicity compared with BA.1 or BA.2, leading to increased infectivity of human lung cells [15,16]. This suggests that these sub-lineages may be more likely to manifest as a lower respiratory tract infection than BA.1 or BA.2. BA.4 and BA.5 may have growth advantages over BA.1 and BA.2 [17]; however, severity of disease is similar to the BA.1 lineage, ie, reduced compared with the Delta variant [18].

BA.4 and BA.5 demonstrate a greater ability to resist neutralization by vaccine-elicited sera compared with BA.1 and BA.2 [19], likely due to the unique F486V mutation in the S protein of these sub-lineages [20]. Although they are descended from BA.5, sub-lineages BQ.1 and BQ.1.1 are antigenically distinct [21]. BQ.1 and BQ.1.1 exhibit further increased resistance to neutralization by vaccine-elicited antibodies compared with BA.5 [22], due to an additional set of mutations at positions K444T (BQ.1 and BQ.1.1), N460K (BQ.1 and BQ.1.1), and R346T (BQ.1.1 only) [22]. The antigenic distance between XBB and descendent sub-lineages and earlier Omicron sub-lineages is even greater than that of BQ.1 and BQ.1.1; XBB sub-lineages also exhibit increased resistance to neutralization [23,24]. Although a booster dose of the original BNT162b2 vaccine elicits robust neutralizing activity across most Omicron sub-lineages, albeit reduced compared with the wild-type virus, neutralization efficiency versus BA.4 and BA.5 is further reduced compared with previous Omicron sub-lineages, and is lowest against BQ.1, BQ.1.1, XBB, and XBB.1 [22–24]. Booster vaccination with the original BNT162b2 vaccine provides a marginal increase in protection against symptomatic disease caused by BA.1 and BA.2, although this wanes after a few months [25–27]. Protection against severe disease caused by these sub-lineages after booster vaccination remains high [27–31]; however, the long-term durability of protection against severe disease beyond 6 months is currently unknown [30].

The high transmissibility of Omicron sub-lineages and their descendants, which has led to a large increase in the number of COVID-19 cases and hospitalizations, and the waning of responses to vaccination and limited durability of protection, have led regulatory and public health authorities to advise updating the authorized and approved COVID-19 vaccines to include Omicron S proteins [32]. Bivalent messenger RNA (mRNA) booster vaccines containing Omicron BA.1 have been rolled out in Europe, and those containing BA.4/BA.5 components are being administered in many countries, including the United States and across Europe. However, SARS-CoV-2 seems to be continually evolving through the accumulation of random mutations, leading to antigenic drift, as observed with influenza viruses. The emergence of recombinant variants, such as XBB and its descendants, demonstrates that a form of antigenic shift may also occur, resulting in viruses that are antigenically distant from previously circulating variants [33]. Similar to influenza, if antigenic drift or shift continues to occur, future adaptations of COVID-19 vaccines are likely to be required, at least until next-generation vaccines with improved breadth and increased durability of protection are developed. In this Review, we discuss the development of, and regulatory approval pathways for, variant-adapted versions of the BNT162b2 vaccine developed by BioNTech and Pfizer. An initial comprehensive literature search was performed on 22 September 2022, supplemented by BNT162b2-specific data known to authors. Additional publications were included on a case-by-case basis following the initial literature search, owing to the rapidly evolving nature of this topic. A final literature search was carried out on 15 March 2023 to ensure all relevant evidence was captured.

2. Development of variant-adapted BNT162b2 vaccines

2.1. Proposed mechanism of action of variant-adapted vaccines

Bivalent variant-adapted vaccines may provide improved protection against COVID-19 through several mechanisms. Inclusion of an Omicron component is expected to result in increased neutralizing antibody responses against Omicron sub-lineages, while the original wild-type component is anticipated to provide similar titers against prior wild-type virus to the original vaccine. It may also be hypothesized that bivalent vaccines could activate naïve B cells and elicit the formation of new memory B cells against Omicron sub-lineages, as well as recalling old memory B-cell responses derived from previous vaccination and/or infection. This would result in an expanded memory B-cell repertoire, thus broadening the breadth of neutralizing antibody responses to different and expanded epitopes. This hypothesis is based on data demonstrating robust recall responses in individuals double- or triple-vaccinated with original BNT162b2 who had breakthrough BA.1 infection, resulting in the expansion of memory B cells against epitopes shared across variants [19,34]. In addition, BA.4/BA.5 breakthrough infection in individuals triple-vaccinated with mRNA vaccines has been shown to result in broad and robust cross-neutralizing activity against BA.1, BA.2, wild-type virus, and previous VOCs Beta and Delta [35,36]. Furthermore, and unlike natural infection, vaccination may initiate the formation of new germinal centers [37,38], resulting in further maturation of B cells and increasing immunological breadth. Experimental data from individuals vaccinated with bivalent vaccines are required to confirm these mechanisms.

Bivalent vaccines are also expected to preserve the breadth of T-cell responses, providing durable protection against severe disease. T-cell immunity is required for early, broad, and durable protection against SARS-CoV-2 [39]. Conservation of the cell-mediated immune response in the lungs may be associated with the prevention of severe disease [40]; thus, T-cell responses are required to provide more durable protection against severe disease and deaths.

In summary, bivalent variant-adapted vaccines are expected to provide a wider breadth of protection and potentially increased duration of protection against infection and disease caused by previous VOCs, Omicron sub-lineages, and other antigenically aligned lineages of SARS-CoV-2 that are related to Omicron or wild-type virus, when compared with original vaccines.

2.2. Development of BNT162b2 variant-adapted vaccines

In response to the emergence of Omicron BA.1, which exhibited partial immune escape, in January 2022, BioNTech and Pfizer initiated the clinical development of a monovalent BA.1 vaccine candidate (monovalent BNT162b2 Omicron BA.1; 30 μg and 60 μg doses) and a bivalent vaccine candidate targeting both the original wild-type virus and BA.1 (bivalent BNT162b2 Original/Omicron BA.1; 30 μg and 60 μg doses) (Figure 2) [41,42]. Moderna, the manufacturer of the mRNA-1273 vaccine, chose to develop a bivalent vaccine only, containing original vaccine and BA.1 components. These vaccines were developed with the aim of eliciting greater protection against BA.1 and other sub-lineages that may arise, compared with the original vaccines. A phase II/III clinical study (NCT04955626) was initiated to explore the monovalent BNT162b2 Omicron BA.1 30 μg vaccine in individuals 18–55 years of age and both monovalent BNT162b2 Omicron BA.1 and bivalent BNT162b2 Original/Omicron BA.1 (both doses) vaccines in participants >55 years of age (Figure 1a,b).

Figure 1. Omicron-adapted BNT162b2 vaccine study designs. a) Monovalent BNT162b2 Omicron BA.1 vaccine (30 μg dose) in participants 18–55 years of age [41,42]; b) Monovalent BNT162b2 Omicron BA.1 vaccine (30 μg and 60 μg doses) and bivalent Original/Omicron BA.1 vaccine (30 μg and 60 μg doses) in participants >55 years of age [41,42]; c) Bivalent Original/Omicron BA.4–5 vaccine in participants >12 years of age [43].

Figure 2. Timeline for clinical development and approval of original and variant-adapted BNT162b2 vaccines in adults and older children.

*Approval of BNT162b2 Original/BA.4–5 initially based on preclinical/non-clinical data and data for other BNT162b2 vaccines.

CMA, conditional marketing authorization; EMA, European Medicines Agency; EUA, emergency use authorization; FDA, United States Food and Drug Administration; FMA, full marketing authorization; R&D, research and development.

Subsequent to the initiation of this trial, neutralization data were generated from BA.1 convalescent serum of triple-mRNA-vaccinated individuals that showed unexpectedly low neutralizing antibody titers against BA.4 and BA.5 [19,44]. This suggested that a vaccine based on BA.1 may not provide sufficiently improved protection against these later sub-lineages. Conversely, BA.4 and BA.5 convalescent serum from triple-mRNA-vaccinated individuals had strong neutralizing activity against all previous Omicron sub-lineages, as well as against BA.4 and BA.5 [36]. By this time, sub-lineages BA.4 and BA.5 were increasing in prevalence and were predicted to become dominant. Therefore, BioNTech and Pfizer also began development of a bivalent vaccine targeting the original wild-type virus and BA.4/BA.5 (BNT162b2 Original/Omicron BA.4–5; 30 μg and 60 μg doses). Development of this vaccine was undertaken at risk.

Preclinical and clinical data on the monovalent and bivalent BA.1 BNT162b2 vaccines, and preclinical data on the BNT162b2 Original/Omicron BA.4–5 vaccine, were presented to the United States Food and Drug Administration (FDA) in June 2022 in the context of a Vaccines and Related Biological Products Advisory Committee (VRBPAC) meeting. BioNTech and Pfizer requested that the BNT162b2 Original/Omicron BA.4–5 vaccine be considered, owing to the increasing prevalence of these sub-lineages. Based on the data presented, the FDA subsequently specified that updated vaccines should include a BA.4/BA.5 component [45]. Preclinical data showing a reduced neutralizing antibody response to the U.S.A-WA1/2020 reference strain (a proxy for the wild-type virus) with the monovalent BA.1 vaccine [42], compared with the bivalent BA.1 vaccine, led to the recommendation for a bivalent approach [45]. Also taking into account these findings, the International Coalition of Medicines Regulatory Authorities (ICMRA) and World Health Organization (WHO) recommended implementation of bivalent rather than monovalent vaccines [46]. Following the FDA and ICMRA/WHO recommendations, a phase II/III clinical study (NCT05472038) was initiated to evaluate the BNT162b2 Original/Omicron BA.4–5 vaccine in participants >12 years of age [43] (Figure 1c).

The variant-adapted BNT162b2 vaccines were evaluated in line with FDA and European Medicines Agency (EMA) guidance (Table 1). The guidance recommends that the effectiveness of a modified vaccine against a particular VOC is established on the basis of efficacy of the original vaccine manufactured by the same process and immunogenicity bridging studies demonstrating superiority of the neutralizing antibody response elicited by the adapted vaccine when compared with the original vaccine [47,48]. Superiority of the geometric mean ratio (GMR) of neutralizing antibody titers elicited by the adapted vaccine versus the original vaccine against the particular VOC/sub-lineage in the adapted vaccine requires the lower bound of the 95% confidence interval (CI) for the GMR (adapted vaccine:original vaccine) to be >1. Non-inferiority of seroresponse is another key outcome measure, defined as achieving a ≥4-fold rise in 50% neutralizing titer from baseline (prior to primary vaccine series), which requires the lower bound of the 95% CI for the percentage difference to be >−5 [47,48]. Other important outcome measures include comparison of neutralizing geometric mean titers (GMTs) to the VOC/sub-lineages versus GMTs to the reference strain, and the local and systemic reactogenicity profile of the adapted vaccine [47,48].

Table 1. Regulatory authority guidance on clinical evaluation of variant-adapted COVID-19 vaccines [47,48].

Agency	FDA	EMA
Subjects	For primary vaccination: previously unvaccinated persons For booster vaccination: persons who previously completed primary vaccination with the parent vaccine	For primary vaccination: unvaccinated subjects with no serological evidence of prior SARS-CoV-2 infection For booster vaccination: Subjects with fully documented prior vaccination with the parent vaccine
Primary endpoints	Superiority of variant-adapted versus parent vaccine against variant, based on >1.5 or >1-fold margin for GMT ratio Non-inferiority of variant-adapted versus parent vaccine against variant, based on <5% margin for seroresponse* rate difference (alternatively, superiority based on >0% or >10% margin)	Lower bound of 95% CI for difference in seroconversion^† rate against variant ≤−10%, for variant-adapted versus parent vaccine Lower bound of 95% CI for GMT ratio against variant ≥0.67, for variant-adapted versus parent vaccine
Key secondary endpoints	Seroresponse* rates and GMTs, including against other clinically relevant variants Comparisons of immune responses post-variant-adapted booster with responses post-primary series of parent vaccine	GMTs and seroconversion^† rates elicited by parent vaccine versus variant-adapted vaccine virus, and vice versa
Safety	Safety assessments for ≥7 days post-vaccination (solicited events) and during immunogenicity evaluation period (serious/unsolicited events) Studies should plan for longer-term assessment of serious and medically attended events	In the absence of safety concerns for parent vaccine, safety data collected during immunogenicity trials with the variant vaccine suffice for approval

*Determination of appropriate seroresponse definition should be based on available data.

^†At least 4-fold increase in titer from pre- to post-vaccination (nominal value for pre-vaccination applied to subjects seronegative at baseline; actual values from previous trials for seropositive subjects, or data from a matched population).

CI, confidence interval; COVID-19, coronavirus disease 2019; EMA, European Medicines Agency; FDA, United States Food and Drug Administration; GMT, geometric mean titer; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

2.3. Clinical and preclinical immunogenicity data for variant-adapted vaccines

The immunogenicity of the monovalent BNT162b2 Omicron BA.1, the bivalent BNT162b2 Original/Omicron BA.1, and the bivalent BNT162b2 Original/Omicron BA.4–5 vaccine was investigated in several preclinical studies in mice. The variant-adapted candidate vaccines were administered as boosters following a primary series with the original BNT162b2 vaccine, and neutralizing titers and breadth of responses were evaluated [36,42]. The vaccines were shown to broaden cross-neutralizing activity against Omicron sub-lineages relative to the original vaccine [36,42].

The vaccines were also assessed in clinical studies in humans. At 1 month post-vaccination, 60 μg monovalent BNT162b2 Omicron BA.1 and both doses of bivalent BNT162b2 Original/Omicron BA.1 met superiority criteria and seroresponse non-inferiority criteria for neutralizing antibody responses against BA.1 compared with the original BNT162b2 vaccine when administered as a fourth dose booster in participants >55 years of age without evidence of prior infection up to 1 month post-vaccination [49]. Monovalent BNT162b2 Omicron BA.1 also met superiority and seroresponse non-inferiority criteria for neutralizing antibody responses versus the original vaccine when administered as a fourth dose booster in participants 18–55 years of age without evidence of prior infection up to 1 month post-vaccination [42].

In individuals >55 years of age without evidence of prior infection, geometric mean fold rises (GMFRs) in neutralizing antibody activity against BA.1 at 1 month post-vaccination with monovalent BNT162b2 Omicron BA.1 as a fourth dose booster were 13.5 for the 30 μg dose and 19.6 for the 60 μg dose; GMFRs in neutralizing antibody activity against the reference strain were 5.2 and 5.0, respectively [49]. GMFRs against BA.1 at 1 month post-vaccination with bivalent BNT162b2 Original/Omicron BA.1 as a fourth dose booster were 9.1 for the 30 μg dose and 10.9 for the 60 μg dose; GMFRs in neutralizing antibody activity against the reference strain were 4.3 and 5.6, respectively. GMTs to the reference strain were boosted with both the monovalent and bivalent vaccines [49].

In individuals >55 years of age without evidence of prior infection, responses to BA.4 and BA.5 at 1 month post-vaccination with bivalent BNT162b2 Original/Omicron BA.1 as a fourth dose booster were modestly improved, but were lower than those to BA.1 [49]. Serum 50% neutralizing GMTs 1 month post-vaccination with bivalent BNT162b2 Original/Omicron BA.1 for the 30 μg and 60 μg doses, respectively, were 110 and 114 against the BA.4/BA.5 S protein and 771 and 900 against BA.1 [49].

In a longer-term follow-up of older individuals (≥60 years of age), both doses of monovalent BNT162b2 Omicron BA.1 and of bivalent BNT162b2 Original/Omicron BA.1 resulted in moderately stronger neutralizing antibody responses against BA.1 compared with the original vaccine, at 3 months post-vaccination [50].

Early phase II/III clinical data in adults have demonstrated greater neutralization of BA.4/BA.5 with the bivalent Original/Omicron BA.4–5 vaccine compared with the original BNT162b2 vaccine, when administered as a fourth dose booster. In participants 18–55 years of age at 1 month post-booster, the bivalent Original/Omicron BA.4–5 vaccine elicited a 9.5-fold increase in BA.4/5 neutralizing antibody titers from pre-booster levels; the corresponding increase in adults >55 years of age was 13.2-fold [51]. By comparison, the original BNT162b2 booster elicited a 2.9-fold increase in BA.4/5 neutralizing antibody titers from pre-booster levels in adults >55 years of age. Therefore, neutralizing antibody titers against Omicron BA.4/5 were approximately 4-fold higher for the bivalent vaccine versus the original vaccine [51].

Both adults with and without prior infection had a substantial increase in BA.4/BA.5 neutralizing antibody titers, albeit greater in those without prior infection. In participants >55 years of age without prior infection, neutralizing antibody titers against the recently emerged Omicron sub-lineages BA.4.6, BA.2.75.2, BQ.1.1, and XBB.1 increased 4.7- to 22.2-fold from pre-booster levels with the Original/Omicron BA.4–5 booster at 1 month post-booster, compared with 1.3- to 2.5-fold with the original BNT162b2 booster [52]. Among all participants, regardless of prior infection, an increase of 3.2- to 4.8-fold was observed with the bivalent vaccine versus the original vaccine against BA.4.6, BA.2.75.2, BQ.1.1, and XBB.1. Neutralizing antibody titers against the wild-type virus with the bivalent Original/Omicron BA.4–5 vaccine were maintained at a similar level to the original vaccine [52].

Taken together, these findings suggest that the Omicron-adapted BNT162b2 vaccines may provide a wider breadth of protection against the Omicron sub-lineage included in the vaccine, when compared with the original vaccine. The BNT162b2 Original/BA.4–5 vaccine may also provide protection against VOCs, such as BQ.1.1 and XBB.1.

2.4. Clinical safety data for variant-adapted vaccines

In individuals >55 years of age, the reactogenicity profiles of monovalent BNT162b2 Omicron BA.1 and bivalent BNT162b2 Original/Omicron BA.1 as fourth dose boosters at both doses were similar to that of the original vaccine as a third dose booster [49,53]. Similar findings were observed for the monovalent vaccine in adults 18–55 years of age [42]. Most frequent adverse reactions in both age groups were injection-site pain, fatigue, headache, myalgia, chills, and arthralgia [49,53]. No new adverse reactions were identified for either vaccine. Early clinical data at 7 and 30 days post-vaccination indicated that the bivalent Original/Omicron BA.4–5 vaccine was well tolerated, with a safety profile similar to that of the original vaccine [51].

Original BNT162b2 is well tolerated and has an extensive safety database based on clinical trials and real-world surveillance data and reporting systems. The most common adverse events (occurring in ≥1 in 10 recipients) with the original vaccine observed during clinical studies and post-authorization were headache, diarrhea, arthralgia, myalgia, injection-site pain, fatigue, chills, pyrexia, and injection-site swelling [53]. There is an increased risk of myocarditis and pericarditis following vaccination with BNT162b2, which is highest in younger males; however, the overall frequency of such events is very rare [53]. Monitoring for these events post-approval is ongoing.

3. Regulatory challenges

Approvals of Omicron-adapted vaccines were supported by pre-existing regulatory tools, such as the rolling review, which allowed data to be assessed as soon as they became available [54,55]. Post-approval roll-out utilized similar mechanisms to those used for annual influenza vaccine updates. Nevertheless, several challenges were encountered during the approval and roll-out processes.

Firstly, differences in guidance across regulatory bodies affected the development of variant-adapted vaccines. Although major regulators had already outlined the regulatory framework for the potential approval of variant-adapted vaccines [47,48,56], a harmonized decision-making process for triggering an update of COVID-19 vaccines was not defined at the global level. The resulting uncertainty led to multiple discussions around the timing, composition, and requirements for Omicron-adapted vaccines throughout the first half of 2022.

Eventually, major regulators agreed on the key principles around Omicron-adapted vaccines in June 2022 [57], but initial guidance on target populations, composition, and sub-lineages for inclusion was not fully harmonized. In fact, while the FDA advised that updated vaccines must contain a BA.4/BA.5 component [45], the EMA and other regulators did not specify which sub-lineage should be included. The FDA recommended that variant-adapted vaccines should be developed for adult and pediatric populations [58]; however, it was initially unclear whether adapted vaccines for pediatric populations would be required in other regions. The WHO has not yet released a full policy statement on variant-adapted vaccines [59,60], although it has stated that adapted vaccines will be considered for policy recommendations once they have received approval by a stringent national regulatory authority. Policy recommendations will address different use-case scenarios and include consideration of programmatic aspects [59].

Secondly, time taken to reach to regulatory decisions affected development timelines. Between the generation of clinical data for the monovalent and bivalent BNT162b2 Omicron BA.1 vaccines and regulatory decision-making, the BA.4/5 sub-lineages had emerged and became dominant, leading regulatory authorities to consider BA.4/5-based vaccines as the best strategy moving forward. Similarly, the monovalent vaccine was developed rapidly in response to the emergence of BA.1 before bivalent approaches were considered. By the time regulatory authorities had decided to recommend bivalent adapted vaccines, studies of both monovalent and bivalent vaccines were already underway. The development of both vaccines was undertaken by manufacturers at risk.

Overall, while technical timelines for updating and manufacturing Omicron-adapted mRNA-based vaccines would have allowed availability by the first half of 2022, the lack of global harmonized guidance regarding decision-making, vaccine composition, and other requirements meant that Omicron-adapted vaccines were not made available until August/September 2022.

4. Regulatory approvals

Based on the clinical and preclinical data described above, bivalent BNT162b2 Original/Omicron BA.1 as a 30 μg booster in individuals >12 years of age received a positive opinion from the EMA’s Committee for Medicinal Products for Human Use (CHMP) [61], and subsequently received conditional authorization for use in the European Union [53,62]. This vaccine has also been approved by the UK Medicines and Healthcare products Regulatory Agency (MHRA) [63].

An application for Emergency Use Authorization (EUA) to the FDA for the 30 μg bivalent BNT162b2 Original/Omicron BA.4–5 vaccine was completed [64], and the EUA for BNT162b2 was amended to authorize the bivalent vaccine for use as a single booster dose in individuals >12 years of age [65]; a 10 μg dose was subsequently authorized for use in children 5–11 years of age based on data with the higher dose [53,66,67]. A submission to the EMA for the 30 μg bivalent BNT162b2 Original/Omicron BA.4–5 vaccine was also completed [68], and the vaccine subsequently received full marketing authorization for use in individuals >12 years of age [53,62]; the 10 μg dose has been recommended for approval for use in children 5–11 years of age [69,70]. A 3 μg dose of the bivalent BNT162b2 Original/Omicron BA.4–5 vaccine was approved by the FDA as the third dose of a three-dose primary series, or as a single booster dose after completion of a primary series of original BNT162b2, for children 6 months to 4 years of age [71].

The initial approvals of the BNT162b2 Original/Omicron BA.4–5 vaccine, both in the United States (under EUA) and in the European Union (initially under conditional marketing authorization), were granted before clinical data for this specific vaccine were available. These approvals were based on available clinical evidence generated with the BNT162b2 Original/Omicron BA.1 vaccine, and immunogenicity shown in non-clinical models, as well as an overview of safety data collected with multiple BNT162b2-based variant-adapted vaccine candidates, which showed a safety and reactogenicity profile consistent with the original vaccine. The emerging data from the ongoing clinical trials investigating the BNT162b2 Original/Omicron BA.4–5 vaccine (described above) have since confirmed the expected safety profile and the predicted immunogenicity [51,52].

The updated Omicron BNT162b2 vaccines were made available for autumn/winter 2022/2023 booster vaccination campaigns globally. The booster vaccines were primarily implemented for high-risk individuals, such as the elderly, immunocompromised, and those with underlying medical conditions [62]. This is in line with the expansion to the FDA EUA for individuals >12 years of age in March 2022 to include a second booster dose specifically for people ≥50 years of age and for people with certain immunocompromising conditions [72], and with guidance from the WHO Strategic Advisory Group of Experts on Immunization (SAGE), which emphasizes the importance of vaccinating those at risk of severe COVID-19, including older adults and those with underlying conditions, with additional booster doses [73].

Consistent with the increasing confidence associated with these COVID-19 vaccines and a continuously improving understanding of their immunogenicity, the EMA Emergency Task Force (ETF) released a statement noting that it is reasonable to expect bivalent Original/Omicron BA.4–5 vaccines to elicit priming against SARS-CoV-2. The ETF considers that it would be acceptable to use these vaccines to deliver a primary series should this become necessary to support vaccination campaigns [74]. In April 2023, the FDA removed the EUA for the original mRNA COVID-19 vaccines, stating that the bivalent Original/Omicron BA.4–5 BNT162b2 vaccine should be used for all doses administered to individuals ≥6 months of age going forward [75]. In May 2023, the WHO Technical Advisory Group on COVID-19 Vaccine Composition (TAG-CO-VAC) stated that the original antigen should no longer be included in future variant-adapted vaccines, as variants that are antigenically related to the wild-type virus are no longer in circulation [76].

5. Conclusions

Based on clinical and preclinical data, bivalent Omicron-adapted BNT162b2 vaccines are anticipated to increase neutralizing antibody responses specifically to the Omicron sub-lineage in the vaccine, as well as to phylogenetically aligned sub-lineages. The broader and more robust immune response elicited by the bivalent BNT162b2 Original/BA.1 and Original/BA.4–5 vaccines may provide improved and potentially more durable protection against symptomatic and severe disease caused by Omicron sub-lineages and antigenically similar VOCs compared with the original vaccine. In line with this, initial data from the United States indicate a 2.4-times-lower risk of death due to COVID-19 in vaccinated individuals who received an Original/Omicron BA.4–5 vaccine booster compared with those who did not [77], with greater vaccine effectiveness against severe Omicron infection, hospitalization, and death compared with a booster dose of the original vaccine [78]. A study in Finland showed a reduction in the risk of hospitalization and death due to COVID-19 in individuals ≥65 years of age who received either an Original/Omicron BA.1 or Original/Omicron BA.4–5 booster versus those who did not [79]. Although no benefit was detected in chronically ill adults 18–64 years of age, the sample size for this group was smaller [79]. In a study in Israel, hospitalization rates in people ≥65 years of age were lower in the BNT162b2 Original/Omicron BA.4–5 booster-vaccinated cohort compared with the non-boosted cohort, for up to 120 days post-vaccination [80].

Recent data have shown that bivalent mRNA vaccines incorporating a BA.4/5 component provide greater neutralization of XBB sub-lineages than the original vaccines [23,24]. In the majority of individuals, a fifth dose booster of bivalent Original/Omicron BA.4–5 vaccine results in enhanced neutralization of XBB.1.5 and CH.1.1 [81,82]. Neutralization of XBB and related sub-lineages by bivalent BA.4/5 vaccine-elicited sera is lower than that of earlier and BA.5 descendent sub-lineages [23,24,83], although vaccine-elicited SARS-CoV-2-specific T cells recognize the XBB sub-lineage, potentially leading to protection from severe disease [84]. The effectiveness of the Original/Omicron BA.4–5 vaccines against symptomatic XBB/XBB.1.5 infection has been estimated to be similar to that against BA.5 [85]. The Original/Omicron BA.4–5 vaccines have been associated with a lower risk of infection or severe infection with the BQ.1/BQ.1.1 and XBB/XBB.15 sub-lineages [78], and provided protection against hospitalization or death due to COVID-19 during periods of BQ.1/BQ.1.1 and XBB/XBB.1.5 circulation, as well as during BA.4/5-dominant periods [78,80,86–90].

However, in studies in the United Kingdom, reduction in the risk of hospitalization and death from COVID-19 in those who received a fifth dose booster of bivalent Original/Omicron BA.1 vaccine was similar to that observed with previous original vaccine boosters [91,92]. This is likely due to differences in circulating lineages by the time this vaccine was rolled out. In addition, waning effectiveness of the bivalent Original/Omicron BA.4–5 vaccines has been observed from 2–4 months post-booster [93], likely due to the antigenic difference between the vaccine and circulating sub-lineages. This highlights the limitations of variant-adapted vaccines. Antigenic drift will inevitably lead to further changes in the genome of SARS-CoV-2 over time, which may eventually result in a mismatch between the vaccine and the circulating virus. Antigenic drift and shift also have the potential to grant additional immuno-evasive properties. As emerging variants that are more antigenically distant from the variant-adapted vaccine displace previously circulating viruses, further variant-adapted vaccines will be needed. Consequently, in May 2023, the WHO TAG-CO-VAC recommended an update to COVID-19 vaccines based on an XBB.1 descendent sub-lineage, such as XBB.1.5 or XBB.1.16, in order to better match the vaccines to the current dominant circulating variant. A monovalent formulation was recommended, as variants that are antigenically related to the wild-type virus are no longer in circulation [76]. Ongoing updates to COVID-19 vaccines are expected as SARS-CoV-2 continues to evolve.

Regulatory harmonization is required to ensure timely access to updated vaccines. There remains a need for a globally harmonized regulatory pathway or decision-making model for transitioning to a variant-adapted vaccine [94]. Implementation of a model for regulatory approval of updated COVID-19 vaccines that somewhat resembles the method for approval of influenza vaccines that have been updated to reflect changes in circulating strains would support the timely availability of future variant-adapted vaccines. The rationale for requiring an updated vaccine remains to be determined but could potentially be based on the emergence of VOCs through antigenic drift or shift that have mechanisms to escape protection offered by current bivalent vaccines.

6. Expert opinion

Further real-world evidence generated during autumn/winter 2022/2023 booster vaccination campaigns will provide additional data to further assess the effectiveness of Omicron-adapted vaccines. Vaccine effectiveness should continue to be monitored over time in order to understand how the vaccines work in a real-world setting and to assess the durability of protection and determine when further booster doses or further vaccine updates may be required.

SARS-CoV-2 variant epidemiology continues to change rapidly, with a new VOC arising every few months; variants with enhanced transmissibility and/or mechanisms of immune escape are likely to displace prior prevalent lineages. Experience with the development and approval of Omicron-adapted vaccines has shown that mRNA-based vaccines can be rapidly adapted to address emerging variants and sub-lineages. However, lack of regulatory alignment and uncertainty in the decision-making process can delay availability. In January 2023, the FDA presented a proposal for a recommendation and vaccine composition update process, to take place on an annual basis in order to ensure that updated COVID-19 vaccines, as needed, will be made available for the upcoming autumn/winter season [95]. Similarly, the EMA has stated that COVID-19 vaccination campaigns are expected to be conducted annually [96]. The WHO TAG-CO-VAC convened in May 2023 and will convene again approximately 6 months later; this frequency will be adjusted in future if, and as, necessary [76]. While the precise timing and details of the process must still be defined and harmonized across regulators, these proposals represent an important first step towards standardization of vaccine updates. The proposals suggest that, in future, COVID-19 vaccine updates may follow a similar model to that adopted for influenza vaccines. This may support potential synergies between COVID-19 and influenza vaccination programs, such as concomitant administration or combined influenza–COVID-19 vaccines, particularly as high-risk groups for severe outcomes from infection with these viruses overlap.

Next-generation vaccine approaches are in development that could address VOCs with enhanced immune escape mechanisms that may arise in future. BioNTech and Pfizer are undertaking a clinical trial of a next-generation mRNA-based candidate vaccine with an enhanced spike protein design that includes an Omicron variant [97]. A WHO meeting on advancing the development of pan-sarbecovirus vaccines noted that, if they can be developed and deployed, these vaccines would be expected to protect against future variants and future crossovers of related coronaviruses to humans [98]. However, several challenges to the development of these vaccines exist. For example, the reliability of animal models that allow for the evaluation of vaccines against infections caused by the large number of viruses in the sarbecovirus subgenus against which protection is desired is untested, and there are no well-defined immune correlates of protection [98]. Research into the antigenic breadth of sarbecoviruses and their potential for antigenic drift will be required, and conserved epitopes must be identified [98]. Several pan-sarbecovirus vaccines are currently in development, with different compositions and mechanisms of action, including the inclusion of immune-stimulating adjuvants, receptor-binding domain mosaics displayed on nanoparticles, and mRNA vaccines expressing chimeric spike protein domains [99]. BioNTech and Pfizer have developed a vaccine candidate composed of a T-cell antigen mRNA encoding for SARS-CoV-2 non-spike proteins conserved across a range of variants. This vaccine is designed to enhance T-cell immunity and potentially broaden and improve durability of protection against COVID-19; a phase I study is underway [100].

Article highlights

• This review presents an overview of the clinical development and approval of Omicron variant-adapted versions of the BNT162b2 mRNA COVID-19 vaccine.

• Bivalent BNT162b2 vaccines containing an Omicron BA.1 or BA.4/5 component in addition to the original wild-type component have been shown to provide improved neutralization of Omicron sub-lineages while maintaining neutralizing antibody activity against wild-type SARS-CoV-2 at a level similar to that of the original BNT162b2 vaccine.

• The reactogenicity and safety profile of the bivalent BNT162b2 vaccines is consistent with that of the original vaccine.

• Challenges, such as the ongoing emergence of new SARS-CoV-2 variants with immune escape mechanisms and differences across regulatory bodies in decision-making and guidance on vaccine composition, affected the time to approval and roll-out, highlighting a need for a global harmonized regulatory process prior to future vaccine updates.

Declaration of interest

A Muik, S Pather, F Mensa, and R Rizzi are employees at BioNTech SE (Mainz, Germany). A Muik is an inventor on patents and patent applications related to RNA technology and COVID-19 vaccines. A Muik, S Pather, F Mensa, and R Rizzi have securities from BioNTech SE. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or material discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

All authors substantially contributed to the manuscript conception, interpretation of relevant literature, writing, and review process, and approved the final version for submission.

Acknowledgments

Medical writing support, including assisting authors with the development of the outline and initial draft and incorporation of comments was provided by Rachel Wright, PhD, and editorial support was provided by Ian Norton, PhD, both of Scion, London, UK, supported by BioNTech SE according to Good Publication Practice guidelines (Link).

Additional information

Funding

This manuscript was funded by BioNTech SE.