This blog post is part of a series of posts giving an insight into the role of statisticians working on trials within the Imperial Clinical Trials Unit (ICTU) and how we can incorporate methods work into trials. To view the other posts in the series you can navigate here.
A brief introduction of COV-Boost Trial
Due to concerns that existing COVID-19 vaccines may be less effective against new variant strains and uncertainty on the most effective booster vaccine schedule in March 2021 the COV-Boost trial was set up as the world’s first COVID-19 vaccine booster study. There was an urgent need to investigate a combination of booster vaccines with the vaccines already received by the population to examine their safety profile and immune responses and compare these.
The Joint Committee for Vaccination and Immunisation and UK Chief Medical Officers needed timely information regarding the effects of different booster vaccinations to previous and new variants of SARS-CoV-2 in order to inform national policy for autumn and winter 2021.
The COV-Boost trial is a randomised, phase II UK multi-centre trial which is in follow up as of January 2023. Initially it aimed to establish the reactogenicity and immunogenicity of thirteen experimental arms against ancestral and novel variants of SARS-CoV-2 in participants who have received two doses of AstraZeneca (referred to as ChAd/ChAd in the flow chart) or Pfizer (referred to as BNT/BNT in the flow chart) early in the UK NHS deployment campaign (Stage 1). In addition, this study aims to examine the safety and efficacy of boosting using vaccines designed specifically to target novel variants of the SARS-CoV-2 virus when these vaccines become available (Stage 2).
As this is the world’s first randomised controlled trial for COVID-19 vaccine boosters and its design is complex due to logistical considerations, the number of questions it aimed to examine regarding vaccines, differences between those primed with Pfizer and those with AstraZeneca, and important participant sub-groups (</>= 70 years).
A total of 2883 adults aged 30 years or older were randomised to a booster dose of seven vaccines plus three at half dose and 3 control arms: AstraZeneca, Novavax, Novavax ½ dose, Pfizer, Pfizer ½ dose, Valneva, Valneva ½ dose, Janssen, Moderna, CureVac or the control of meningococcal vaccine. To minimise the risk of dosing errors rather than every site having to randomise to all 13 arms, instead sites were split into groups which offered smaller subsets of the different potential boosters labelled A, B and C.
For more information, you can visit the COV-Boost website. Reporting of solicited adverse events
In COVID-19 vaccine trials participants are usually asked to report local and systemic adverse events that occur within 7 days of receipt of the vaccine. This is so the vaccines can be compared based on their reactogenicity profiles.
In this trial, participants were asked to complete electronic symptom diaries for the 7 days (and longer if symptoms persist at day seven, until resolution or stabilisation) following their booster vaccine.
Diary forms collected information on the timing and severity of the following solicited AEs:
To date, typically in vaccine trials, to present reactogenicity data stacked bar charts are used to display the proportion of participants with events by severity. However, due to the complexity of this trial design it was not straightforward to present and summarise comparisons for 17 local and systemic reactions across 13 vaccine arms. So, for the first time in a vaccine trial, we used Radial or Radar graphs. These presented the proportion of severe local and systemic adverse events in the first 7 days post-vaccination in groups A, B, and C. As a summary measure, we used the most severe category of solicited adverse events experienced over first 7 days. It was decided to use this method because we wanted to provide a fingerprint plot of reactogenicity analysis and summarise the totality of the burden across all solicited adverse events.
First published paper and summary of findings
The first report of our trial on the 28 day safety and immune response outcomes was published in The LANCET on 2nd December 2021. We have received good feedback on the use of radial graphs from clinicians and researchers who have read our first published paper in the LANCET.
Of the 2883 participants randomised in June 2021, 2878 received the COVID-19 or control booster vaccines. Participants were 30 years or older, with approximately half aged 70 years or older. The average age of participants who received Astra Zeneca was 53 years in the younger age group (<70) and 76 years in the older age group (>=70). Average ages for Pfizer were 51 and 78 years, respectively.
The above figure shows the most severe category of local solicited adverse events experienced over first 7 days by study arms and priming vaccine for Group A. Fatigue and malaise were the most reported systemic reactions as seen in the radial graphs. Although some schedules were more reactogenic than others, all vaccines showed acceptable side effect profiles.
The trial demonstrated the potential of all vaccines tested boosted immunity following Astra Zeneca and of 6 vaccines boosted following Pfizer. However, there are large variations in antibody and cellular immune responses between vaccines.
Leila Janani and Annie Wright
Acknowledgements:
As mentioned in this blog series the running of this trial wouldn’t be possible without having such a collaborative group. We would like to acknowledge the roles of others in the team: Professor Saul Faust (chief investigator), Dr Alasdair Munro, our sponsor (University Hospital Southampton NHS Foundation Trust), our funders (NIHR, the Vaccine Task Force and DHSC and CEPI), The Oxford Vaccine Group, the external contract research organisation (PHARMExcel) and the COV-Boost Study Group.
Comments