The Lancet JANUARY 09, 2021

Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK

Merryn Voysey, Sue Ann Costa Clemens, Shabir A Madhi, et al. (Oxford Vaccine Trial Group)

Bottom Line

This interim analysis of phase 3 trials demonstrated that the ChAdOx1 nCoV-19 vaccine has an acceptable safety profile and is effective in preventing symptomatic COVID-19, with variations in efficacy observed depending on the dosage regimen.

Key Findings

1. The overall vaccine efficacy against symptomatic COVID-19 occurring more than 14 days after the second dose was 70.4% (30 [0.5%] of 5807 in the vaccine group vs 101 [1.7%] of 5829 in the control group).
2. Subgroup analysis revealed that participants receiving a low dose followed by a standard dose achieved 90.0% efficacy, whereas those receiving two standard doses showed 62.1% efficacy.
3. No COVID-19-related hospitalizations or severe disease occurred in the vaccinated group from 21 days after the first dose, compared to 10 hospitalizations (including two severe cases and one death) in the control group.
4. The safety profile was favorable, with no treatment-related deaths and a balanced occurrence of serious adverse events (84 in the vaccine group vs 91 in the control group).

Study Design

Design
RCT
Double-Blind
Sample
11,636
Patients
Duration
3.4 mo
Median
Setting
Multicenter, UK/Brazil/South Africa
Population Adults aged 18 years and older with no prior laboratory-confirmed SARS-CoV-2 infection.
Intervention ChAdOx1 nCoV-19 vaccine (AZD1222) administered as two doses (either LD/SD or SD/SD).
Comparator Meningococcal conjugate vaccine (MenACWY) or saline placebo.
Outcome Symptomatic COVID-19 occurring more than 14 days after the second dose in seronegative participants.

Study Limitations

The interim analysis was based on a relatively short median follow-up period of 3.4 months.
The varied dosage regimens (low-dose/standard-dose vs standard-dose/standard-dose) were not initially planned, resulting from an unintended dosage quantification error in one of the cohorts.
There was limited data regarding efficacy in older populations and against emerging variants of concern at the time of the analysis.
The evaluation of asymptomatic infection was hampered by insufficient data, limiting conclusions on the vaccine's impact on viral transmission.

Clinical Significance

This trial provided critical evidence supporting the use of the ChAdOx1 nCoV-19 vaccine as a scalable tool for global pandemic control, demonstrating robust protection against severe disease and hospitalization, which are the most clinically consequential outcomes of SARS-CoV-2 infection.

Historical Context

Published during the height of the COVID-19 pandemic, this study represents a pivotal moment in vaccine development, marking one of the first major randomized controlled trial reports for a viral-vectored platform aimed at rapid, equitable global distribution and supporting mass vaccination strategies in LMICs.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

How does the ChAdOx1 nCoV-19 vaccine utilize a non-replicating chimpanzee adenovirus vector to induce an immune response against SARS-CoV-2?

Key Response

The vaccine uses a modified chimpanzee adenovirus (ChAdOx1) that has been engineered to be replication-deficient in humans. It carries the genetic sequence for the SARS-CoV-2 spike protein. Once injected, the vector enters human cells and delivers the DNA template; the host cells then synthesize the spike protein, which is presented on the cell surface to prime the immune system to recognize and mount an immune response (both antibody and T-cell mediated) against the actual virus.

Resident
Resident

The interim analysis noted a higher efficacy (90%) in a subgroup that received a low-dose followed by a standard-dose (LD/SD) compared to those receiving two standard doses (SD/SD, 62%). How should this finding influence clinical discussions regarding vaccine dosing schedules?

Key Response

While the LD/SD regimen showed higher efficacy, it was initially a result of a measurement error in a specific cohort. Clinicians must recognize that although the higher efficacy is intriguing, the standard-dose regimen is what was primarily validated across the broader trial population. This finding suggests that dose-sparing or heterologous dosing might enhance immunogenicity, but clinical decisions should follow the authorized dosing protocols approved by regulatory bodies until prospective trials confirm the LD/SD benefit.

Fellow
Fellow

What is the clinical significance of anti-vector immunity in the context of the ChAdOx1 platform, and how does this affect the potential for multi-dose booster strategies using the same viral vector?

Key Response

Anti-vector immunity occurs when the host's immune system develops neutralizing antibodies against the adenovirus vector itself rather than the target spike protein. Using a chimpanzee adenovirus (ChAdOx1) instead of a common human adenovirus (like Ad5) reduces the likelihood of pre-existing immunity. However, repeated use of the same vector for booster shots may lead to diminished efficacy as the body clears the vector more rapidly upon subsequent exposures, necessitating the consideration of 'mix-and-match' (heterologous) prime-boost strategies.

Attending
Attending

Given that the trials in the UK, Brazil, and South Africa were conducted during periods of differing viral prevalence and emerging variants, how does the 'pooled' efficacy of 70.4% translate to real-world effectiveness in a rapidly evolving pandemic landscape?

Key Response

Pooled analysis provides a robust sample size but hides geographical and temporal heterogeneity. The 70.4% figure represents a weighted average across different populations and viral strains (like the early B.1.1.7 or B.1.351 variants). In practice, an attending must communicate that vaccine efficacy is a dynamic metric; while the vaccine remains highly effective at preventing severe disease and hospitalization, its protection against symptomatic infection may fluctuate significantly depending on the dominance of specific variants of concern.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

Critique the statistical validity of combining data from four separate trials (COV001, COV002, COV003, and COV005) with different recruitment criteria, placebo controls (meningococcal vaccine vs. saline), and dosing intervals into a single interim analysis.

Key Response

Combining heterogeneous trials increases power but introduces significant 'noise' and potential bias. The use of a meningococcal vaccine (MenACWY) as a control in some trials was intended to maintain blinding by mimicking local reactogenicity, but it complicates the safety comparison against an inert saline placebo. Furthermore, the varying intervals between doses (ranging from 4 to 26 weeks) make it difficult to isolate the effect of the dosing interval from the effect of the dose concentration itself without a pre-planned factorial design.

Journal Editor
Journal Editor

As a reviewer, how would you evaluate the publication of a 'serendipitous' finding (the LD/SD subgroup) that was not a part of the original randomized protocol but showed a significantly higher efficacy than the intended dose?

Key Response

A journal editor would flag this as a post-hoc or unplanned subgroup analysis. While the results are statistically significant, they are prone to 'p-hacking' or chance findings because the subgroup was not randomized to that specific dose a priori. The editorial concern is whether the paper over-emphasizes a finding born from a manufacturing error. A rigorous review would require the authors to clearly label this as exploratory and call for prospective validation before it influences public health policy.

Guideline Committee
Guideline Committee

In light of the finding that efficacy was higher with a longer interval between doses (up to 12 weeks), how should global guidelines balance the need for rapid mass vaccination versus optimizing individual-level immune protection?

Key Response

The data suggests that a longer interval (around 12 weeks) between the first and second dose of ChAdOx1 improves overall efficacy. For a guideline committee (like the WHO SAGE or JCVI), this supports a 'first-dose-first' strategy during supply shortages. By extending the interval, more people can receive an initial dose—which provides substantial protection against severe disease (approx. 76% efficacy after one dose)—while also potentially yielding a more robust and durable secondary immune response once the second dose is eventually administered.

Clinical Landscape

Noteworthy Related Trials

2020

BNT162b2 mRNA Vaccine Trial

n = 43,548 · NEJM

Tested

BNT162b2 mRNA COVID-19 vaccine

Population

Individuals 16 years of age or older

Comparator

Placebo

Endpoint

Incidence of symptomatic COVID-19

Key result: The vaccine showed 95% efficacy in preventing COVID-19 infection starting 7 days after the second dose.
2021

mRNA-1273 Vaccine Efficacy Trial

n = 30,420 · NEJM

Tested

mRNA-1273 SARS-CoV-2 vaccine

Population

Adults at high risk of SARS-CoV-2 exposure

Comparator

Placebo

Endpoint

Incidence of symptomatic COVID-19

Key result: The vaccine demonstrated 94.1% efficacy in preventing symptomatic COVID-19 illness.
2021

Ad26.COV2.S Vaccine Trial

n = 43,783 · NEJM

Tested

Ad26.COV2.S (single-dose)

Population

Adults 18 years and older globally

Comparator

Placebo

Endpoint

Incidence of moderate to severe–critical COVID-19

Key result: The vaccine was 66.9% effective in preventing moderate to severe–critical COVID-19 at 14 days and 66.1% at 28 days.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis