Immune correlates analysis of antibody responses against SARS-CoV-2 variants in the ENSEMBLE vaccine efficacy trial

Abstract

In Latin American sites of the ENSEMBLE trial of the Ad26.COV2.S vaccine vs. placebo, binding antibodies and neutralizing antibodies measured 4 weeks post-vaccination (∼peak) against circulating lineages (Ancestral, Gamma, Lambda, Mu, Zeta) were assessed as a correlate of risk of, and correlate of protection against, lineage-specific COVID-19. Comparison of lineage-matched controlled vaccine efficacy (VE) curves showed similar relationships across lineages of lineage-specific antibody with VE against lineage-matched COVID-19, supporting a “variant-invariant correlate of protection model” that undergirds immunobridging inferences of efficacy against new variants based on variant-matched neutralizing antibody titers. Lambda departed from this model at undetectable/just-detectable titers: at ∼ peak Reference-specific titers of 2.7 arbitrary units (AU)/mL (just-detectable) and 30 AU/ml, VE against Ancestral COVID-19 was 53.0% (95% CI: 30.7%, 67.9%) and 84.5% (73.6%, 93.0%), respectively; at the same Lambda-specific titers, VE against Lambda COVID-19 was 12.3% (−54.1%, 50.3%) and 91.1% (68.9%, 98.0%). Additional research is needed for Omicron variants.