Safety, effectiveness and immunogenicity of heterologous mRNA-1273 boost after prime with Ad26.COV2.S among healthcare workers in South Africa: The single-arm, open-label, phase 3 SHERPA study

Abstract

Limited studies have been conducted on the safety and effectiveness of heterologous COVID-19 vaccine boosting in lower income settings, especially those with high-HIV prevalence., The Sisonke Heterologous mRNA-1273 boost after prime with Ad26.COV2.S (SHERPA) trial evaluated a mRNA-1273 boost after Ad26.COV2.S priming in South Africa. SHERPA was a single-arm, open-label, phase 3 study nested in the Sisonke implementation trial of 500000 healthcare workers (HCWs). Sisonke participants were offered mRNA-1273 boosters between May and November 2022, when Omicron sub-lineages were circulating. Adverse events (AE) were self-reported, and co-primary endpoints (SARS-CoV-2 infections and COVID-19 hospitalizations or deaths) were collected through national databases. We used Cox regression models with booster status as a time-varying covariate to determine the relative vaccine effectiveness (rVE) of the mRNA-1273 booster among SHERPA versus unboosted Sisonke participants. Of 11248 SHERPA participants in the rVE analysis cohort (79.3% female, median age 41), 45.4% had received one and 54.6% two Ad26.COV2.S doses. Self-reported comorbidities included HIV (18.7%), hypertension (12.9%) and diabetes (4.6%). In multivariable analysis including 413161 unboosted Sisonke participants, rVE of the booster was 59% (95%CI 29-76%) against SARS-CoV-2 infection: 77% (95%CI 9-94%) in the one-Ad26.COV2.S dose group and 52% (95%CI 13-73%) in the two-dose group. Severe COVID-19 was identified in 148 unboosted Sisonke participants, and only one SHERPA participant with severe HIV-related immunosuppression. Of 11798 participants in the safety analysis, 228 (1.9%) participants reported 575 reactogenicity events within 7 days of the booster (most commonly injection site pain, malaise, myalgia, swelling, induration and fever). More reactogenicity events were reported among those with prior SARS-CoV-2 infections (adjusted odds ratio [aOR] 2.03, 95%CI 1.59-2.59) and less among people living with HIV (PLWH) (aOR 0.49, 95%CI 0.34-0.69). There were 115 unsolicited adverse events (AEs) within 28 days of vaccination. No related serious AEs were reported. In an immunogenicity sub-study, mRNA-1273 increased binding and neutralizing antibody titres and spike-specific T-cell responses 4 weeks after boosting regardless of the number of prior Ad26.COV2.S doses, or HIV status, and generated Omicron spike-specific cross-reactive responses. mRNA-1273 boosters after one or two Ad26.COV2.S doses were well-tolerated, safe and effective against Omicron SARS-CoV-2 infections among HCWs and PLWH. Trial registration: The SHERPA study is registered in the Pan African Clinical Trials Registry (PACTR): PACTR202310615330649 and the South African National Clinical Trial Registry (SANCTR): DOH-27-052022-5778.

Fig 1. Flow chart of the SHERPA analysis cohort nested in the Sisonke implementation trial.

Fig 2. Antibody response elicited after mRNA-1273 heterologous boost.

Plasma from a subset of participants were assessed to determine the antibody response elicited upon heterologous boost (mJnJ) after 1 Ad26.COV2.S dose (JnJ x1; n = 19) or 2 Ad26.COV2.S doses (JnJ x2; n = 29). The neutralization activity was measured using a SARS-CoV-2 pseudovirus-based neutralization assay with (panel A) D614G, (panel B) Omicron BA.4 and (panel C) D614G in people living with HIV (PLWH). The plasma neutralization titer is measured as an ID₅₀. Black horizontal bars represent medians. The threshold of detection for the neutralization assay is an ID₅₀ of 20. Antibody binding was measured using an in-house SARS-CoV-2 assay using the D614G full spike protein (panel D and E). An EC50 was used to measure the binding titers of the samples. ADCC activity was measured by detecting the crosslinking ability of the antibodies present in the serum (panel F and G). Relative light units were measured which correlate with the levels of FcγRIIIa signalling. For all assays, statistical significance was measured with the Kruskal-Wallis test with Dunn’s multiple comparisons test. Significance is shown as: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. Medians and fold changes are depicted under each graph. Samples were run in duplicate for all assays. Nucleocapsid ELISA was used to identify prior SARS-CoV-2 infections. Only individuals who were seropositive at baseline were included in the analysis as there was only 1 nucleocapsid-seronegative participants in the JnJ x1 group.

Fig 3. T-cell response elicited after mRNA-1273 heterologous boost.

(A) Clinical characteristics of the T-cell sub-study cohort. Frequency of spike-specific CD4+ (B) and CD8+ (C) T-cells producing any of the measured cytokines (IFN-γ, IL-2 or TNF-α) in response to stimulation with a peptide pool spanning ancestral spike protein at baseline (BL), week 4 (W4) and week 24 (W24), in participants vaccinated with one or two prior doses of Ad26.COV2.S and an mRNA-1273 booster. The median response and proportion of participants with a spike-specific CD4+ or CD8+ T-cell responses are indicated below the graphs. (D) Spike-specific CD4+ and CD8+ T-cell cytokine responses stratified according to number of Ad26.COV2.S doses received. (E) Spike CD4+ and CD8+ T-cells in HIV-uninfected participants (pink) or PLWH (red). (F) Spike-specific CD4+ T-cell responses to spike from ancestral virus and Omicron sub-lineages BA.1, BA.4/5 and XBB.1. (G) Fold change of Omicron sub-lineage responses compared to ancestral spike. Bars represent the medians. Differences between responses at different time points post boost were assessed using a two-tailed Wilcoxon paired test. Differences between participants vaccinated with one or two doses of Ad26.COV2.S and responses to different SARS-CoV-2 variants were calculated using a Mann-Whitney U test. The number of participants included in each analysis is indicated on the graphs.