Waning immunity to SARS-CoV-2: implications for vaccine booster strategies

As we approach the end of the second year of the COVID-19 pandemic, of the many challenges facing immunologists and vaccinologists, one of the toughest has been the characterisation of the durability of protective immunity. The duration of immune protection has crucial implications for the implementation of booster vaccination programmes¹—including the need for and timing of additional doses—which are a source of intense debate among both scientists and policy makers. In a Personal View in The Lancet Respiratory Medicine,² Gregory Milne and colleagues review emerging data on cellular and humoral immunity to SARS-CoV-2, in response to natural infection and vaccination, and offer their views on what the evidence means in terms of the longevity of protective immunity. Health policy makers have the unenviable task of developing strategies to reduce the burden of disease in the face of many points of uncertainty and controversy, including those related to the emergence of SARS-CoV-2 variants of concern and the equitable distribution of vaccines.

Booster programmes require adequate vaccine stocks, national supply and implementation logistics, and a rationale for age or risk-group prioritisation, but also the challenge of yet more uncharted immunology for which the existing evidence base is thin: in a world in which booster recipients will be drawn from those with variable prior immunity—which might be based on previous infection, often overlaid with vaccination with an mRNA or adenovirus vector-based spike vaccine—what form should a booster dose take, and when might it most advantageously be given?

Added to this is consideration of the substantial geopolitical and ethical implications of booster vaccination programmes, including the questions of whether it is appropriate to give what has been described as an extra life jacket to the privileged few, when many in the world have yet to receive any life jacket at all.³ The intricacy of these deliberations is demonstrated by the divergent booster programme strategies of countries such as the UK (cautious, stratified booster roll-out), Israel (hard-line revision of the green passport approach, which affords full social access specifically and only to those who have received a third dose), and the USA (ongoing reappraisals, but moving towards a highly targeted booster programme).

The basic immunology evidence base that underpins the current deliberations is evaluated by Milne and colleagues.² As with so much that we have had to confront in the pandemic, decisions on booster programmes represent real-life choices that sit at the interface between immunological research and epidemiology data on SARS-CoV-2-related morbidity and mortality. Until recently, we had no clear consensus on the serum half-life of protective, neutralising antibodies after either natural infection or vaccination. Initial assumptions were that the poor durability of protective antibodies following seasonal infection by the human common cold coronaviruses might be replicated with SARS-CoV-2⁴—the fact that we succumb to winter colds caused by the same viruses year after year is a stark warning about the possibility of rapidly waning protection. The underlying mechanism of reinfection with common cold coronaviruses is generally thought to involve coronavirus adaptations that subvert innate pathways such as those leading to production of type I interferons. As longitudinal data have emerged from cohorts infected by SARS-CoV-2, interpretation of findings on waning immunity has become a source of ongoing debate. There is substantial heterogeneity between individuals in antibody levels after infection.5, 6 However, with estimates of a serum half-life for neutralising antibodies of more than 200 days, along with evidence of well sustained T-cell and B-cell memory⁶ and an improving B-cell repertoire due to affinity maturation (the concept that the immune response develops through a progressively more focused and tightly binding antibody repertoire),⁷ immune protection might be expected to last for about 1–2 years after infection.

Since the end of 2020, data on immunity from countries with vaccination programmes have been overlaid with the effects of diverse spike vaccine platforms, each with distinct profiles of mean decline in neutralising antibodies and vaccine response kinetics. During this time, variants of concern have also emerged, notably the increasing dominance of the delta variant, which is associated with a 5–10-times fall in neutralising antibodies.⁸ Although basic immunology findings predict that the vast majority of people would nevertheless have sufficient protection to avert any risk of infection,⁹ real-world data have been more salutary. Data from individuals who received the Pfizer–BioNTech mRNA vaccine BNT162b2, especially those over the age of 60 years, who were vaccinated early in the programme, show susceptibility to breakthrough infection caused by the delta variant. Breakthrough infections are correlated with diminished antibody titres, especially at 6 months or more after the second vaccine dose.1, 10 A third booster dose is sufficient to flatten the curve of breakthrough cases, even in areas where the delta variant is dominant.¹⁰ On the one hand, this can be read as a good news story, of calm, determined, rational immune monitoring managing to largely mitigate the next public health disaster. On the other hand, it is a sobering reminder of what a formidable pathogen we face, such that all traditional, textbook immunology assurances that T-cell and B-cell memory priming will provide lasting protection look a little thin.

Finally, there is a need to confront the detail of protective immune repertoires and immune imprinting. It is already known that the immune response elicited by first-generation vaccines results in wide variability in neutralisation of current variants of concern, and that imprinting of the immune repertoire by previous exposure to the virus can differentially shape the protective response.¹¹ Decoding these patterns and optimising protection against future variants will depend on fine mapping of cross-protective epitopes and making informed choices about which spike sequences to adopt for future booster dose programmes. The optimum COVID-19 vaccination strategy for the next few years is currently far from obvious, but probably not just a matter of using the sequence of the most recent variant of concern.

© 2021 Jim West/Science Photo Library

DMA and RJB receive support from the UK Medical Research Council (MR/S019553/1, MR/R02622X/1, MR/V036939/1, and MR/W020610/1) and the National Institute for Health Research (NIHR) Imperial Biomedical Research Centre Institute for Translational Medicine and Therapeutics; Cystic Fibrosis Trust Strategic Research Centre (2019SRC015); NIHR Efficacy and Mechanism Evaluation Fast Track (NIHR134607); NIHR Long Covid (COV-LT2-0027); Innovate UK (SBRI 10008614); and Horizon 2020 Marie Skłodowska-Curie Innovative Training Network European Training Network (no. 860325). Both authors are members of the Global T cell Expert Consortium and have consulted for Oxford Immunotec, outside of the submitted work.