key: cord-0809877-28gjl943
authors: Jethwa, Hannah; Wong, Richard; Abraham, Sonya
title: Covid-19 vaccine trials: ethnic diversity and immunogenicity
date: 2021-05-11
journal: Vaccine
DOI: 10.1016/j.vaccine.2021.05.017
sha: e0ac083ac7dc1bbaaed4f7bf81397bf9956e4af2
doc_id: 809877
cord_uid: 28gjl943

nan

In recent years it has become increasingly recognised that immunogenicity of vaccines can vary significantly between individuals, with age, comorbidities and medication use being some of the key factors determining immune response 1 . Responses can differ further with ethnic groups from the same location demonstrating varied vaccine responses and differing antibody decline rates, suggesting a genetic influence 1 . Furthermore, twin studies estimate a 36-90% and 39-90% degree of heritability for humoral responses and cellular responses, respectively, depending on the vaccine 1 .

There is evidence to suggest that Covid-19 affects individuals from black and South Asian ancestries disproportionately, with higher rates of hospitalisation and increased mortality compared to Caucasian counterparts 2 . In a UK study, after age, sex and geography were accounted for, death rates for individuals of black African descent were 3.5 times higher than for Caucasians, compared to 2.7 and 1.7 times higher for those of Pakistani or black Caribbean descent 2 . Similar data was reported in studies from the United States, with significantly increased death rates among black/African-American and Hispanic/Latino people compared to those from Caucasian or Asian descent (92.3, 74.3, 45.2 and 34.5 deaths per 100,000 population, respectively) 2 . Furthermore, the severity of Covid-19 in different ethnic populations could be further exacerbated by the underrepresentation of recruits from ethnically diverse backgrounds for interventional treatment trials 3, 4 . The success of candidate interventions could be statistically proven or dismissed for one population but not the other.

As such it is important that phase III clinical trials, in particular, assessing vaccine efficacy and safety include a diverse mix of ethnicities and that such sub-analyses are included in trial data. This is especially pertinent for candidate vaccines against diseases with either a higher prevalence or increased risk of morbidity and/or mortality in ethnic minority groups such as Covid-19 in order for vaccine efficacy data to be applicable to these individuals. Furthermore the paucity of ethnic diversity in vaccine trials means efficacy may differ between countries, resulting in health disparities.

This commentary assesses clinical trial data of the three Covid-19 vaccines: BNT162b2 (Pfizer/BioNtech/Fosun Pharma), AZD1222 (Oxford/Astra Zeneca) and mRNA-173 (Moderna) currently approved for use by the UK Medicines and Healthcare products Regulatory Agency, and asks whether the results provide sufficient evidence of efficacy in the whole population.

As of 20 th April 2021 the World Health Organization (WHO) reported 91 and 184 Covid-19 vaccines in clinical and pre-clinical development, respectively 5 . The majority of candidate vaccines in the clinical phase constitute Covid-19 viral protein subunits, with other common vaccines containing nonreplicating viral vectors, DNA, RNA and inactivated Covid-19 virus 5 .

In a recent UK report, the National Institute of Health Research analysed ethnicity data provided by a total of 622,978 participants in Covid-19 vaccine studies nationally and identified that only 5.72% (1,509 participants) came from ethnic minority backgrounds 6 . The protocol for the Oxford vaccine Phase 2/3 UK study (COV002), for example, had no mention of ethnicity and out of a total of 560 participants in Phase 2 only 28 (5%) individuals were from ethnic minority backgrounds 7 . This reflects a significant underrepresentation of such groups, which constitute 13.8% of the UK population 6 ; although sub-analyses of vaccine efficacy can be performed depending on ethnicity, low numbers from minority groups may reduce the statistical significance of this data and thereby render the results less accurate. An interim report of the Phase 3 trials in UK & Brazil showed that participation from ethnic minority backgrounds had marginally increased to 8.6% in the UK, and significantly (33.4%) in Brazil (Table 1) 9 . The overall study population included 83%, 28%, 9% and 4.3% from white, Hispanic/Latino, black/African-American, Hispanic/Latino and Asian backgrounds, respectively 9 .

In the above two studies, although the significant proportion of Hispanic/Latino participants in the trials reflects the high proportion of such individuals in the United States such numbers are disproportionate when considering national populations in other continents such as Europe, Asia and Africa. Furthermore, the above studies have very few participants from Asian origins, and clarification regarding whether such individuals are from South Asian or East Asian decent is required to enable local health authorities to review whether trial data is applicable nationally. These comments also apply for the AZD1222 vaccine, despite trials in Brazil including a large cohort of local ethnic groups, the statistical power remains poor for those ethnically from East and South Asia. The producers of the AZD1222 have lauded the vaccine as advantageous to developing nations, especially those without ubiquitous freezer storage, yet Phase 3 trials have yet to begin in South (India) or East Asia (Japan), and no detailed participation data was provided for their South African trial 8 .

Although the immunogenicity to vaccinations is likely to be dependent on multiple factors, for example vaccine composition, it is likely that there will be some variability to each of the candidate vaccines depending on race. One mechanism that produces variations in vaccine response are polymorphisms in major histocompatibility complex (MHC) genes, though polymorphisms in pattern recognition receptors (PRRs) or single-nucleotide polymorphisms (SNPs) have also been shown to be associated with variations in immunogenicity 1 .In recent years, multiple studies reviewing immunogenicity to influenza A vaccines have demonstrated this and, in particular, polymorphisms in the IGHV1-69 gene have been demonstrated to modulate anti-influenza antibody repertoires 10 . This is one of the genes responsible for antibody production and it exists in 14 variations, with certain polymorphisms being more effective at mounting a response to the influenza A virus compared to others; ethnicity dramatically varied the frequency of various polymorphisms, with those from African, Asian and European descent showing marked differences 10 . Furthermore there are variations in immune response to the 2007 H5N1 influenza virus itself depending on which polymorphisms in the IGHV1-69 gene were present, as well as how many copies of the genes were present.

Other gene polymorphisms that vary between ethnicities are also likely to play a part. A study by Kurupati et al. (2016) , for example, demonstrated that gene expression profiling arrays revealed marked differences in 1,368 probes at baseline between Caucasians and African-Americans 11 ; furthermore genes expressed differently between these two ethnicities irrespective of age were enriched for myeloid genes, whereas those that differed between younger participants were enriched for those specific for B-cells 11 .

Although we are still in the early stages of studies reviewing immunogenicity of candidate vaccines against Covid-19, it is key that clinical trials take the potential differences between ethnicities into account during study development as well as during post-study data analysis. This will not only allow local authorities to be better informed regarding the potential effects of the vaccine on their local population, but this may also guide which vaccines may be more appropriate for different patient cohorts. Indeed, the Phase 3 protocols for the three approved UK vaccines did not propose to analyse the immune response of individuals from different ethnicities separately in subgroups. However, Moderna did add a sentence to describe their intention to enrol a representative sample of racial and ethnic minority participants 12 One limitation to consider with ethnic diversity in clinical trials, however, is social and cultural behaviours that may differ globally irrespective of underlying race, as it is likely that the complex interactions between host, pathogen and environmental factors are key. For example, a Western diet of African-American participants in clinical trials may affect immunogenicity compared to individuals native to Africa and factors such as smoking habits, use of herbal remedies and levels of stress between countries may also differ between populations of the same ethnic group. As such, future data is required not only on ethnicity and immunogenicity to medicinal products such as vaccinations but also on different population groups internationally. Certain genetic mutations within geographic populations may also, for example, impact immunogenicity, highlighting the importance of both ethnicity as well as geographic distribution of trial participants 10, 17 .

Continued evaluation of vaccine efficacy on an international level is pertinent, as modernisation of third world countries are likely to lead to changes in environmental factors, for example diet, which may impact factors relating to immunogenicity. Future trials in vaccines could consider obtaining genetic and ethnicity data /environmental data and linking with immunogenicity.

Factors that influence the immune response to vaccination

Evidence mounts on the disproportionate effect of COVID-19 on ethnic minorities

The under-representation of BAME patients in the COVID-19 Recovery trial at a major London NHS Trust

Dexamethasone in hospitalized patients with Covid-19

World Health Organization. Draft landscape of COVID-19 candidate vaccines

NIHR research ethnicity data provides insight on participation in COVID-19 studies

Safety and immunogenicity of ChAdOx1 nCoV-19 vaccine administered in a prime-boost regimen in young and old adults (COV002): a single-blind, randomised, controlled, phase 2/3 trial

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. The

Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine

IGHV1-69 polymorphism modulates anti-influenza antibody repertoires, correlates with IGHV utilization shifts and varies by ethnicity

Race-related differences in antibody responses to the inactivated influenza vaccine are linked to distinct pre-vaccination gene expression profiles in blood

Available online at

World Map of Moderna studies linked to Covid-19 (2020)

World Map of ChAdOx1 Studies linked to Covid-19 (2020)

World Map of BNT162b2 Studies linked to Covid-19 (2020)

COVID-19 vaccine tracker

Host genetic factors can impact vaccine immunogenicity and effectiveness