Skip to main content

Safety and efficacy of the ChAdOx1 nCoV-19 (AZD1222) Covid-19 vaccine against the B.1.351 variant in South Africa

Our take —

This phase Ib/II clinical trial in South Africa, available as a preprint and thus not yet peer reviewed, estimated that the ChAdOx1 nCoV-19 vaccine (Oxford/AstraZeneca) was not shown to be effective against the B.1.351 South African strain, with an estimated efficacy of preventing symptomatic COVID-19 of 10.4% against the B.1.351 variant. Because this variant was highly prevalent among cases, the overall efficacy for symptomatic COVID-19 was 21.9%. The study could not determine if the vaccine was protective against severe disease due to insufficient numbers. No serious adverse events were reported in the sample. Further research is necessary, with larger sample sizes, including groups with comorbidities or immunocompromised individuals, to fully understand the effectiveness of the vaccine against this emergent variant.

This preprint was later published in a scientific journal here.

Study design

Randomized Controlled Trial

Study population and setting

This was a Phase Ib/II randomized controlled trial across 7 study sites in South Africa. Participants were randomized to the vaccine (0.33 – 0.5mL ChAdOx1-nCoV19 AstaZeneca and Oxford vaccine) or placebo, with an additional booster dose administered 21 to 35 days later. A cohort of 70 HIV-negative people were initially enrolled for safety testing, and an additional 1956 HIV-negative individuals were enrolled for a total of 2021 in the safety study, 1010 received placebo, and 1011 received the vaccine. The safety study collected reactogenicity within 7 days of full vaccination and unsolicited adverse event reporting for 28 days following vaccination. Of these 2021 initial people, 1467 were included in the efficacy study (717 placebo and 750 vaccine), with sampling for SARS-CoV-2 infection occurring at routine visits and also if COVID-19 symptoms were reported. The primary endpoint for the efficacy study was symptomatic COVID-19 occurring >14 days after the second shot among participants seronegative at baseline. Serostatus at randomization was evaluated with IgG assays, and antibody neutralization studies were conducted within 2 weeks of the 2nd dose in 107 randomly selected vaccine recipients who were seronegative at enrolment.

Summary of Main Findings

In the safety study, adverse events (both mild and serious) were similar among vaccine and placebo groups with no statistically significant differences. For vaccine efficacy, there were 42 cases with either mild (15 received vaccine and 17 received placebo) or moderate (4 who received vaccine, 6 receiving placebo) COVID-19. There were no reported severe disease or hospitalization in placebo or vaccine groups. The incidence of COVID-19 for those who received the placebo >14 days after the booster was 93.6 cases per 1000 person-years, and 73.1 cases per 1000 person-years for the vaccinated group. In their neutralization study, there was a strong antibody response after the first dose, with 9 of 19 seronegative (47%) showing no detectable neutralization to an RBD triple-mutant pseudovirus, while 79% (15 of 19) showed no detectable neutralization response to the B.1.351 pseudovirus. 41 of 42 samples among endpoint cases were sequenced, of which 39 (95.1%) were B.1.351. The vaccine efficacy about patients with the B.1.351 variant was 10.4% (95% CI: -76.8 to 54.8%), and all COVID-19 (regardless of severity or variant) was 21.9% (95% CI: -49.9 to 59.8%).

Study Strengths

The study was a double-blind randomized controlled trial, which reduced selection and measurement bias in this sample. They were able to examine not only safety, but also efficacy, and immunogenicity through its neutralization study. Additionally, the study included a diverse range of people, including being 70.5% black African, which are often underrepresented in clinical trial studies.


The primary limitation is the generalizability of the RCT results. The study had a diverse range of participants included in the study, however the age range was primarily <60 years of age, and may reflect individuals with stronger immune systems who can respond to the vaccine. Similarly, a small minority of the study population had respiratory disorders, and given that COVID-19 is expected to have pronounced severe disease among those with prior respiratory conditions, these results may not reflect certain immunocompromised or comorbid populations. Additionally, the follow-up of this study was from 140-171 days, and it is not clear if the same level of efficacy can be seen long-term beyond this. Finally, there were relatively few individuals in each arm of the study and larger sample sizes are needed for more precise estimates of efficacy.

Value added

The study is the first to explicitly interrogate the impact of the vaccine on the B.1.351 South African strain of COVID-19, as new strains become a growing concern. It found no significant efficacy against symptomatic COVID-19 caused by the B.1.351 variant, which has critical implications for vaccine rollout and planning in particularly in Southern Africa.

This review was posted on: 19 February 2021