The rapid development and approval of COVID-19 vaccines were crucial in combating the pandemic. However, the emergence of SARS-CoV-2 variants of concern (VoCs) with mutations affecting transmissibility and immune evasion necessitated evaluating vaccine efficacy (VE) against these specific variants. South Africa experienced distinct waves of infection: the ancestral SARS-CoV-2 (WT) wave in 2020, followed by the Beta variant-driven second wave in late 2020 and early 2021, and the Delta variant-driven third wave in mid-2021. Previous studies showed varying VE for AZD1222 (ChAdOx1 nCoV-19), ranging from high VE against WT in some settings to a significantly reduced VE against the Beta variant in South Africa. This study aimed to provide a final analysis of the VE of the AZD1222 primary series against various VoCs prevalent in South Africa and assess its safety profile, addressing the knowledge gap regarding its effectiveness against emerging variants in the South African context.

Prior research highlighted the variable efficacy of COVID-19 vaccines against different SARS-CoV-2 variants. An interim analysis of four AZD1222 studies across Brazil, South Africa, and the UK reported an overall VE of 66.7% against symptomatic infection. However, the VE against the Alpha variant was reported as 74.6% in the UK. A pivotal phase 3 trial in the US, Chile, and Peru showed a VE of 74.0% against symptomatic COVID-19 predominantly caused by the ancestral virus. In contrast, an earlier interim analysis of the COV005 study revealed a low overall VE of 21.9% against mild-to-moderate COVID-19 in South African participants, largely due to the prevalence of the Beta variant.

COV005 is a phase 1b/2, double-blind, randomized, placebo-controlled, multicenter trial (NCT04444674; PACTR202006922165132). Healthy adults aged 18-65 were recruited between June 24, 2020, and November 9, 2020, and randomized (1:1) to receive two doses of either placebo (saline) or AZD1222. The primary endpoint was VE against PCR-positive symptomatic COVID-19 occurring >14 days post-second dose, along with safety assessment. Secondary objectives included variant-specific VE and reactogenicity. Nasopharyngeal swabs were tested using Emergency Use Authorization SARS-CoV-2 assays. Whole-genome sequencing was performed on a subset of samples to determine the SARS-CoV-2 variant. VE was calculated as 1 minus the relative risk (incidence in vaccinated/incidence in placebo). 95% confidence intervals were calculated using the Clopper-Pearson exact method. Adverse events (AEs) and serious adverse events (SAEs) were recorded.

A total of 960 placebo and 935 vaccinated participants were included in the final analysis. The study captured infections across WT, Beta, and Delta variant waves. Overall VE against any SARS-CoV-2 infection was 38.3% (95% CI 15.1, 55.4). High VE was observed against WT (90.6%, 95% CI 35.4, 99.8) and Delta (77.1%, 95% CI 30.4, 94.4) variants. However, VE against the Beta variant was low (6.7%, 95% CI -41.1, 38.5). Overall VE against mild and moderate COVID-19 was 45.2% (95% CI 19.7, 63.1). The mean time between the second dose and infection was 5.5 months in the placebo group and 4.5 months in the vaccine group. For Delta variant, the mean time was 9.3 and 10 months respectively. Adverse events were generally mild. One vaccinee experienced a severe Grade 3 fever considered related to AZD1222, resolving within 48 hours. No severe cases of COVID-19 were documented before unblinding.

The high VE against WT and Delta variants aligns with previous findings, demonstrating good durability of protection against these variants, even months after vaccination. The lower VE against the Delta variant compared to WT might be attributed to waning immunity, viral escape from neutralizing antibodies, and the higher transmissibility of the Delta variant. The low VE against the Beta variant confirms previous findings and highlights the ability of some VoCs to evade vaccine-induced immunity, particularly neutralizing antibody activity. However, real-world data suggests continued protection against severe disease through cellular immunity, emphasizing that vaccine decisions shouldn’t solely depend on the effectiveness against non-severe illness. The absence of severe cases in this study during the Delta wave, despite its severity in the broader population, may be due to the exclusion of older adults and those with underlying conditions from the trial.

This final analysis of the COV005 trial confirms the efficacy of the AZD1222 primary series against WT and Delta variants but highlights its limited effectiveness against the Beta variant. The findings underscore the need for continued monitoring of vaccine effectiveness against emerging variants and the importance of considering cellular immunity in addition to humoral immunity when assessing overall vaccine protection. Future research should focus on understanding the long-term effectiveness of AZD1222 and other vaccines against Omicron and subsequent variants, as well as the role of booster doses in maintaining protection against diverse VoCs.

The study population was limited to healthy adults aged 18-65, limiting the generalizability of the results to other age groups and those with comorbidities. The study was censored before the emergence of the Omicron variant, preventing assessment of vaccine efficacy against this dominant variant. Further research is needed to fully understand the impact of the AZD1222 vaccine across different age groups, health conditions, and against the various currently circulating variants of concern.