The durability of natural infection and vaccine-induced immunity against future infection by SARS-CoV-2

The durability of vaccine-mediated immunity to SARS-CoV-2, the durations to breakthrough infection, and the optimal timings of booster vaccination are crucial knowledge for pandemic response. Here, we applied comparative evolutionary analyses to estimate the durability of immunity and the likelihood of breakthrough infections over time following vaccination by BNT162b2 (Pfizer-BioNTech), mRNA-1273 (Moderna), ChAdOx1 (Oxford-AstraZeneca), and Ad26.COV2.S (Johnson & Johnson/Janssen). We evaluated anti-Spike (S) immunoglobulin G (IgG) antibody levels elicited by each vaccine relative to natural infection. We estimated typical trajectories of waning and corresponding infection probabilities, providing the distribution of times to breakthrough infection for each vaccine under endemic conditions. Peak antibody levels elicited by messenger RNA (mRNA) vaccines mRNA-1273 and BNT1262b2 exceeded that of natural infection and are expected to typically yield more durable protection against breakthrough infections (median 29.6 mo; 5 to 95% quantiles 10.9 mo to 7.9 y) than natural infection (median 21.5 mo; 5 to 95% quantiles 3.5 mo to 7.1 y). Relative to mRNA-1273 and BNT1262b2, viral vector vaccines ChAdOx1 and Ad26.COV2.S exhibit similar peak anti-S IgG antibody responses to that from natural infection and are projected to yield lower, shorter-term protection against breakthrough infection (median 22.4 mo and 5 to 95% quantiles 4.3 mo to 7.2 y; and median 20.5 mo and 5 to 95% quantiles 2.6 mo to 7.0 y; respectively). These results leverage the tools from evolutionary biology to provide a quantitative basis for otherwise unknown parameters that are fundamental to public health policy decision-making.

The unprecedented development of efficacious vaccines against SARS-CoV-2 has represented a triumph in the global effort to control the ongoing COVID-19 pandemic. Vaccines have been shown to provide short-term protection from major adverse health outcomes of hospitalization and death (1–4). However, protection against breakthrough infection wanes (5), and breakthroughs have been extensively documented (6, 7). In response, the Food and Drug Administration advisory committee has recommended a booster of the Pfizer-BioNTech and Moderna vaccines at least 5 mo after completion of the primary series to people ≥12 and ≥18 y of age, respectively (8). A booster dose of the Johnson & Johnson/Janssen vaccine has been authorized on a faster timescale—as early as 2 mo after the single dose to individuals 18 y of age and older (8). Nevertheless, the optimal timing of boosting remains challenging to assess. Consequently, rigorous prediction of the durability of immunity conferred by vaccination against the SARS-CoV-2 virus is essential to personal and public health decision-making, having major implications regarding policy decisions about COVID-19 vaccination around the world (9, 10).Short-term longitudinal studies of SARS-CoV-2-neutralizing antibodies in vaccinated individuals (11–13) can provide information crucial to our understanding of the durability of vaccine-mediated immunity. Peak antibody responses following vaccination versus natural responses have also been quantified (14), facilitating analytical comparison of initial immune responses. For endemic viruses, longitudinal data on reinfection can provide reinfection probabilities associated with antibody level. However, longitudinal data on SARS-CoV-2 reinfection are not available during the short term associated with pandemic spread. Nevertheless, longitudinal reinfection data for a diversity of coronaviruses have been collected (15–20). SARS-CoV-2 reinfection probabilities have been obtained from them by phylogenetic analysis, using continuous ancestral and descendent state estimation (21). These estimates, produced before reinfection was commonplace, proved accurate (predicting an 18% probability of reinfection at ∼270 d [ref. 21] that was validated by a subsequent empirical finding of 18% reinfection by 275 to 300 d after primary infection [ref. 22] and, likewise, predicting a 34% probability of reinfection at ∼450 d after primary infection [ref. 21] that was validated by a subsequent empirical finding of 34% breakthrough infection 420 to 480 d after primary vaccination [ref. 23]). Similar analyses pairing antibody response and rates of waning for each vaccine with infection probabilities can enable quantification of the durability of vaccine-mediated immunity against breakthrough infections. The aim of this study is to leverage data on antibody response to each vaccine and corresponding probabilities of infection to estimate the durability of vaccine-mediated immunity against breakthrough SARS-CoV-2 infection for four well-studied vaccines: mRNA-1273, BNT162b2, ChAdOx1, and Ad26.COV2.S.