Concerns about SARS-CoV-2 evolution should not hold back efforts to expand vaccination

Abstract

When vaccines are in limited supply, expanding the number of people who receive some vaccine, such as by halving doses or increasing the interval between doses, can reduce disease and mortality compared with concentrating available vaccine doses in a subset of the population. A corollary of such dose-sparing strategies is that the vaccinated individuals may have less protective immunity. Concerns have been raised that expanding the fraction of the population with partial immunity to SARS-CoV-2 could increase selection for vaccine-escape variants, ultimately undermining vaccine effectiveness. We argue that, although this is possible, preliminary evidence instead suggests such strategies should slow the rate of viral escape from vaccine or naturally induced immunity. As long as vaccination provides some protection against escape variants, the corresponding reduction in prevalence and incidence should reduce the rate at which new variants are generated and the speed of adaptation. Because there is little evidence of efficient immune selection of SARS-CoV-2 during typical infections, these population-level effects are likely to dominate vaccine-induced evolution.

Fig. 1. Potential vaccine-induced evolution.

a | Vaccination increases the transmission advantage of an escape variant compared with wild type (WT). Here, vaccine escape is complete, allowing the variant to replace the wild type in vaccinated hosts. b | If residual immune protection from vaccination slows the transmission of the variant, the variant cannot spread as readily in the vaccinated population, reducing prevalence and incidence. c | Within hosts, ‘intermediate’ immune pressure could in theory maximize the rate of adaptation. After two doses of vaccine, strong immune responses will likely inhibit viral replication and the emergence of escape mutations. Some have proposed that with just one dose, the rate of within-host adaptation could be high (triangle at top of the curve). We suggest that selection during COVID-19 infections is inefficient (triangle to lower right of curve). Adapted with permission from refs^,, AAAS.