Vaccines and SARS-CoV-2 variants: the urgent need for a correlate of protection

Immune-escape variants have raised concerns about the effectiveness of vaccines as the world scales up SARS-CoV-2 immunisation. COVID-19 vaccines have shown up to 95% efficacy¹ in preventing clinical cases and up to 100% efficacy² in preventing severe disease or admission to hospital in settings with pre-existing variants. New variants, especially 501Y.V2 (B.1.351), which escape natural-induced and vaccine-induced immunity, have created uncertainty on whether the vaccines are effective in preventing both mild and severe COVID-19.

Preliminary reports show that the 501Y.V2 variant has complete immune-escape in South African convalescent serum samples,³ and reductions in neutralising activity in vaccinee serum samples for all four vaccines tested.4, 5, 6, 7 Although these reductions were small for the BBIBP-CorV,⁴ BNT162b2,⁵ and mRNA-1273⁶ vaccines, preliminary data suggest they were substantial, including with a complete immune escape for the AZD1222 vaccine.⁷ Concerningly, the clinical trial efficacy of AZD1222 was 70% in the UK and Brazil,⁸ but 22% according to preliminary data from South Africa.⁷ For NVX-CoV237 the efficacy was 89% in the UK but 49% in South Africa,⁹ whereas for Ad26.COV2-S the efficacy was 72% in the USA but 57% in South Africa.¹⁰ Extrapolating vaccine efficacy against pre-existing variants to new variants could be seriously misleading.

Adequate genomic surveillance, standardised variant nomenclature, and a repository of variants and vaccinee serum samples¹¹ are needed to deal with the challenges of repeatedly emerging new SARS-CoV-2 variants, but there is a particularly pressing need to establish a correlate of protection so that vaccine efficacy results obtained with pre-existing variants can be translated to newly emerging variants because it is impractical and time consuming to repeat clinical trials with each new immune-escape variant. Furthermore, repeating clinical trials for each variant might take so long that even newer variants could emerge while these clinical trials are underway.

Because the immune responses required to prevent mild disease might be different to severe disease, correlates of protection might need to be stratified on the basis of disease severity. There are four key requirements to achieve this aim. First, all SARS-CoV-2 vaccine developers with existing or completed efficacy trials should commit to transparency and open data sharing. Second, an expert committee (preferably under WHO) should be appointed to review existing and planned analyses to identify correlates of protection for each efficacious vaccine. Third, studies with multiple vaccines to fast-track the identification of an animal model, assay, or marker as a correlate of protection should be initiated to address gaps in the correlate research plans. Finally, a central database should be created to collate data for each of the efficacious vaccines, thereby providing larger sample sizes to assess multiple variables as correlates of protection and to test if a correlate identified in one trial is valid in other trials.

The identification of a correlate of protection is too important and urgent to be left to uncoordinated separate studies by individual investigators or vaccine developers. South Africa, at the forefront of dealing with the challenge of its vaccine roll-out during the spread of a predominant 501Y.V2 variant, has to make vaccine decisions without adequate efficacy data. A correlate of protection for mild and severe SARS-CoV-2 infection will go a long way to providing an evidence base for these decisions and overcome the obstacles that new variants are placing on the vision of global SARS-CoV-2 control with the widespread implementation of effective immunisation.

© 2021 Bloomberg/Getty Images