The relative efficacy, safety, and immunogenicity of monovalent mRNA-1273 (ancestral WA1) versus bivalent mRNA-1273.222 (ancestral plus omicron BA.4/BA.5) boosting have not been studied prospectively where HIV and SARS-CoV-2 seroprevalence are high. CoVPN 3008 (NCT05168813) was a randomized, double-blinded trial in seven African countries during omicron waves, enrolling people with HIV (PWH) or comorbidities associated with severe COVID-19 who had previously received mRNA-1273. Participants were randomized 1:1 to intramuscular monovalent or bivalent booster and routinely PCR-tested for SARS-CoV-2. Between 10/2022 and 03/2023, 3,988 participants (3,086 PWH) received bivalent (n=2012) or monovalent (n=1976) booster. By month 12, 293 COVID-19 cases and 221 asymptomatic infections occurred, including one severe case. COVID-19 rates did not differ between arms (hazard ratio [HR] 0.92, 95% CI 0.73-1.15, P=0.45). Combined COVID-19 and asymptomatic infection was 15% lower with the bivalent (239 vs. 275; HR 0.85, 0.71-1.01, P=0.07), and asymptomatic infection alone was 24% lower (HR 0.76, 0.59-1.00, P=0.05). Neutralizing antibody titers to BA.4/5 and XBB.1.5, measured in 100 PWH per arm, boosted higher in bivalent recipients (P=0.026 and 0.002). Adverse events were similar. In predominantly PWH, the bivalent booster was well tolerated, elicited superior immunogenicity against omicron variants, and appeared to protect against asymptomatic infection but not COVID-19.
