Skip to main content

Preliminary Efficacy of the NVX-CoV2373 Covid-19 Vaccine Against the B.1.351 Variant

Our take —

This preprint, which had not yet been peer reviewed, offered evidence that the Novavax NVX-CoV2373 is well tolerated and provides protection against symptomatic SARS-CoV-2 infection from the B.1.351 variant. Claims that prior infection with pre-B.1.351 viruses did not reduce the risk of COVID-19 due to re-infection with B.1.351 variants are unsupported.

Study design

Randomized Controlled Trial

Study population and setting

This study, available as a preprint thus not yet been peer-reviewed, summarized the results of a phase IIa/b multicenter randomized placebo-controlled trial of Novavax’s NVX-CoV2737 conducted in South Africa. The NVX-CoV2737 vaccine contains full-length, pre-fusion SARS-CoV-2 spike protein (AA sequence from the original Wuhan isolate) and a saponin-based adjuvant.

The primary study population included 2,684 predominantly (95%) Black-African participants aged 18-84 years (18-64 years for a subset of medically stable people living with HIV [PLWH]) who were randomized in an approximately 1:1 ratio to receive two injections spaced 21 days apart of either NVX-CoV2737 or saline placebo. The primary endpoint assessed was vaccine efficacy (defined as prevention of diagnostically confirmed symptomatic mild, moderate, or severe COVID-19 in initially seronegative participants) seven or more days after receiving the second vaccine dose. Various separate analyses were also conducted on seropositive participants who received either vaccine or placebo. Monitoring for safety was conducted throughout the study, and is planned to continue for a year.

Summary of Main Findings

Fifteen cases of COVID-19 were detected amongst the 1,357 two-dose vaccine recipients, and 29 cases of COVID-19 were detected amongst the 1,327 two-dose placebo recipients, producing an overall estimated vaccine efficacy of 49.4% (95% CI 6.1 to 72.8%).

The majority (38/41) of available SARS-CoV-2 full-genome sequences obtained from positive, symptomatic participants (n = 44) in this trial belonged to the B.1.351 lineage, suggesting that viruses of this lineage were predominant in overall circulation.

No vaccine-related serious adverse events were recorded at the time of writing, although adverse events overall were slightly more common amongst the vaccine recipients versus the placebo recipients.

Study Strengths

Participants with an appropriate range of ages were included, and some had preexisting comorbidities (e.g. hypertension, diabetes mellitus, etc).


As these are phase IIa/b trials, sample sizes are small, and any finding will need to be confirmed in larger phase III trials. Small sample sizes resulted in very large confidence intervals for the vaccine efficacy calculations (e.g., vaccine efficacy of 49.4% [95% CI 6.1 to 72.8%]).

The authors’ claim that prior infection with pre-B.1.351 viruses did not reduce the risk of COVID-19 due to re-infection with B.1.351 variants is based solely on the initial serological status of the participants as determined by an in-house IgG ELISA. This is insufficient evidence for such claims. As history of prior or current symptomatic COVID-19 excluded participation in the study, these recorded seropositive cases likely represent either asymptomatic infection or false positives from their in-house ELISA. However, it is unclear as the time since infection, severity of infection, and variant causing infection are all unknown or unreported for these cases. Additionally, the authors report approximately 30% of the total study population was seropositive at the start of the study. This is questionably high given that South Africa has currently reported about 1.5 million total infections out of a population of 59 million, which equates to about 2.5% of the population known to have been infected overall. To more convincingly demonstrate that previous infection with pre-B.1.351 SARS-CoV-2 provides no protection from infection with B.1.351, prior natural infection would need to be demonstrated by more than serology (e.g., a history of diagnostic nucleic acid amplification testing [NAAT], symptoms, etc.).

Finally, while inclusion of the PLWH participants was advantageous and various attempts are made to evaluate the efficacy amongst this subgroup, the study was underpowered for these purposes (as acknowledged by the authors) and thus inclusion of these participants adds very little meaningful data.

Value added

This study provides the first efficacy data for Novavax’s NVX-CoV2373 vaccine.

This review was posted on: 22 March 2021