DAUK Evidence Summary: What do I need to know about the ‘Oxford Vaccine’?

DAUK’s Zoe Brandon has produced the following as a summary of the latest Lancet publication on the ChAdOx1 nCoV-19 vaccine phase 3 trials to help you understand the evidence.

What is the ChAdOx1 nCoV-19 vaccine?

Developed by Oxford University in partnership with AstraZeneca, the ChAdOx1 nCoV-19 vaccine is a replication-deficient chimpanzee adenovirus that has been engineered to express the coronavirus spike glycoprotein gene. This method of vaccine delivery makes use of our lack of pre-existing immunity to chimpanzee adenoviruses, conferring immunogenic properties that stimulate the immune system to mount a response against both the adenovirus and the coronavirus spike protein it carries.

Efficacy of ChAdOx1 nCOV-19

Phase 2 trials, the results of which were published in November, demonstrated the ability of ChAdOx1 nCoV-19 to stimulate spike-specific humoral and cellular immune responses in trial participants. In comparison, this latest paper published in the Lancet describes the interim results of phase 3 trials for the vaccine, representing the first peer-reviewed publication of SARS-Cov-2 vaccine efficacy to date. The purpose of these phase 3 trials is to assess the safety of the vaccine, and to determine whether the ChAdOx1 nCOV-19 vaccine protects against symptomatic, virologically confirmed COVID-19.

11,636 people in total participated in the phase 3 trials in Brazil and the UK; participants in each trial were randomly allocated to receive either a prime-boost regimen of the ChAdOx1 nCOV-19 vaccine, or a control product (two doses of the meningitis MenACWY vaccine, or MenACWY followed by saline). Across both studies, 5807 people received the vaccine, of which 0.5% went on to develop symptomatic COVID-19 at least 14 days after receiving the second dose of vaccine. In comparison, 5829 people were allocated to the control group, of which 1.7% developed symptomatic COVID-19. Vaccine efficacy was calculated to be 70.4% from these results.

Though all participants in the vaccine arm of the trials received an initial dose followed by a booster dose, this initial dose varied between cohorts. In the Brazil trial, all subjects in the vaccine arm received two standard doses, whilst in the UK trial, the vaccine arm comprised two cohorts with different dosage regimens: low-dose followed by standard dose (LD/SD), and standard dose followed by standard dose (SD/SD). Though it was originally planned that all subjects in the vaccine group would receive two standard doses, the batch of vaccines administered to the initial cohort was retrospectively found to contain a lower dose of ChAdOx1 nCOV-19 viral particles than intended. The reason for this was unexpected interference between a non-active ingredient of the vaccine and the spectrophotometry assay used to measure the viral dose being used. This was amended for subsequent cohorts, who received the standard dose for both initial and booster doses.

When participants in the vaccine arm are split into subgroups according to dose regimen, the analysis shows a significant increase in vaccine efficacy in the LD/SD group within the UK trial (reported as 90.0%), compared to the SD/SD group taken as a whole across both trials (62.1%). Age was considered as a possible confounding factor in these results, given that LD/SD participants were all 18-55 years, in comparison to the SD/SD group which included participants in the 18-55, 56-69, and ≥70 age groups. The interval between doses also varied between the LD/SD and SD/SD groups: in the LD/SD group, only 0.8% had a second dose within 8 weeks of the first dose – this was because the LD/SD group had already received the priming dose when, on the basis of antibody response findings in the UK phase 1/2 trial, the phase 3 trial protocol was modified to a two-dose regimen. Subgroup analysis restricting the UK SD/SD group to participants aged 18-55 with less than 8 weeks’ interval between vaccine doses did not substantially alter the finding: vaccine efficacy in the restricted UK SD/SD group did increase slightly (from 60.3% to 65.6%), but with a higher p-value and wider confidence intervals implying more uncertainty about the result. In the discussion section of the paper, the study authors suggest several possible mechanisms that could explain this unexpected finding.

In addition to symptomatic COVID-19 infection, the study looked at whether there was a difference in the number of COVID-19 cases occurring more than 21 days after the first dose of vaccine, compared to controls. This analysis was restricted to participants that received two standard doses of the vaccine. Overall, ten cases requiring hospitalisation occurred >21 days after the first dose, notably all of which were in the control group. Of these participants, two had severe COVID-19 (WHO score ≥6), and one case was fatal.

Preliminary results as to whether the vaccine reduces incidence of asymptomatic infection, which was detected in the UK trial through weekly NAAT testing of participants, are also published in this paper. In the UK trial, the incidence rate of asymptomatic infection was slightly reduced in the vaccine group compared to the control group, with vaccine efficacy calculated as 27.3%. As the study authors note however, the wide confidence interval for this result suggests a need for a greater sample size in order to draw reliable conclusions.

One limitation of this study is the lack of efficacy data for older adults, with only 8.3% of participants aged 56-69 years, and 3.8% of participants aged ≥70 years. The study did not perform subgroup analysis for vaccine efficacy stratified by age group due to the limited data, however this has been identified as a priority once further data becomes available.

Safety profile of ChAdOx1 nCOV-19

Safety analysis for the vaccine included data from 23,848 participants enrolled in four trials. This included participants from the earlier phase 1/2 studies (which evaluated safety and immunogenicity of the vaccine) in the UK and South Africa, in addition to participants in the vaccine arm of phase 3 trials. A similar number of adverse events were recorded in trial participants that received the vaccine and those who did not: 84 adverse events classified as either serious or of special interest occurred in vaccine recipients, versus 91 adverse events in participants who received MenACWY or saline. Of the adverse events in the vaccinated participants, only two were considered to be possibly related to the vaccine, including one case of transverse myelitis, and one case of fever >40°C.


Overall, these interim results suggest that the Oxford-AstraZeneca vaccine has an acceptable safety profile and is effective at reducing symptomatic COVID-19 infection. The authors also identify several unanswered questions that additional data and subgroup analysis will help to resolve as the study continues, including the duration of protection offered by the vaccine, whether vaccine efficacy is affected by age, ethnicity, and dose regimen, and the ability of the vaccine to prevent asymptomatic spread of the virus.