The Lancet 2020.12.08
The AstraZeneca/University of Oxford adenovirus vector vaccine against COVID-19, ChAdOx1/AZD1222, was safe and effective, if marred by dosing irregularities, according to peer-reviewed interim data from the product’s phase III trials in Great Britain and Brazil.
Overall vaccine efficacy was 70.4% (95.8% CI 54.8%-80.6%), reported Oxford’s Andrew Pollard, PhD, and colleagues in The Lancet.
But it was only 62.1% (95% CI 41.0%-75.7%) among patients who received two standard doses of the vaccine — about three-quarters of the roughly 5,800 individuals assigned to receive the active vaccine.
Safety was called acceptable, with severe adverse events comparable between groups, including three cases of transverse myelitis leading to a highly publicized pause in the study. While independent clinical review found one transverse myelitis case in the vaccine group and one in the control group were unlikely to be related to the vaccine, “a relationship remained possible in the third case,” Pollard and colleagues wrote.
Notably, the vaccine also appeared to show promise in blocking asymptomatic transmission of SARS-CoV-2 in a subset of participants from the U.K. who completed weekly COVID-19 tests. Overall vaccine efficacy against asymptomatic transmission was 27%.
These results, initially published via press release, were similar to interim analyses of trial data from the mRNA COVID-19 vaccines developed by Pfizer/BioNTech and Moderna.
The Half-Dose Happy Accident?
A manufacturing error resulted in 1,367 people in the British trial, all adults under age 55, receiving a half-dose of the vaccine in their first injection, while the rest of the group received a full dose. All participants received the full dose in their second shot.
For reasons that remained unclear, this half-dose group showed higher vaccine efficacy than those receiving two full doses. Risk of clinical illness was reduced 90% (95% CI 67.4%-97.0%), and incidence of asymptomatic transmission was reduced 59%, both relative to control participants who received a meningococcal vaccine.
The authors characterized efficacy in this group as “intriguingly high,” citing other vaccines where a “priming dose” is given prior to a “booster dose.”
They speculated about the reasons, including “higher levels of [neutralizing] antibody, lower levels of anti-vector immunity with lower vector-derived antigen content of the first dose, or differential antibody functionality or cellular immunity, including altered avidity or immunodominance.”
Pollard and colleagues also cautioned about the wide confidence intervals around the point estimates, saying that “further data are needed to confirm these preliminary findings, which will be done in future analyses of the data accruing in these ongoing trials.”
In an accompanying editorial, Maria Deloria Knoll, PhD, and Chizoba Wonodi, DrPH, both of Johns Hopkins School of Public Health in Baltimore, noted “the observed differences in efficacy by dose were not consistent with results from previous immunogenicity trials of this vaccine” where participants received either two low doses or two standard doses.
They also warned this mixed dose regimen might raise issues down the road.
“Disparity between immunogenicity and efficacy findings could imply that clear-cut immunological correlates of clinical protection might not exist for COVID-19 vaccines, meaning efficacy cannot be extrapolated to other unevaluated ages or populations,” the editorialists wrote. “Bridging trials… or immunogenicity equivalence trials… that are faster and easier might be infeasible, posing challenges for future vaccine development, evaluation, and regulatory approval.”
Indeed, media reports indicated these differing efficacy estimates would likely raise some eyebrows at the FDA, if the manufacturer were to apply for emergency use authorization.
AstraZeneca is performing a separate phase III trial in the U.S. that was halted for several weeks while the adverse events in Britain were investigated; no results have yet been released. Those data are likely to be critical for the vaccine’s prospects for authorization.
Data was included from April 23 to Nov. 4 for 11,636 participants in the U.K. and Brazil, out of 23,848 overall in the two countries and in a third trial conducted in South Africa. Adults (ages 18 and older) received either two doses of AZD1222 vaccine or two doses of MenACWY vaccine as control. Primary outcome was symptomatic COVID-19 in seronegative participants with a positive nucleic acid amplification test more than 14 days after the second dose. Researchers noted timing of priming and booster doses varied, due to both edits in the protocol and time to manufacture vaccine, meaning “doses could not be administered at a 4-week interval.”
Over 85% of participants in the U.K. and Brazil trials were ages 18-55, with older participants recruited later. About 61% of participants were women. Nearly all were white in the U.K., while two-thirds were white in Brazil.
More than 21 days after the first dose, 10 were people hospitalized for COVID-19 in the control group, including two with severe COVID-19 and one death. Five cases in the primary analysis were in adults ages 55 and older, but vaccine efficacy in older age groups could not be assessed, the authors said.
Serious adverse events were comparable between groups, though there were two additional transverse myelitis cases not considered to be related to the vaccine. There were four non-COVID deaths (three in the control group, one in the vaccine group) unrelated to the vaccine (road traffic accident, blunt force trauma, homicide, and fungal pneumonia).
Previous data on the vaccine indicated local and systemic reactogenicity was “less in intensity” in older adults with lower doses after the second dose.