The Ethics of Human Challenge Trials Using Emerging Severe Acute Respiratory Syndrome 2 Variants

Abstract

The world’s first coronavirus disease 2019 human challenge trial using the D614G strain of severe acute respiratory syndrome 2 (SARS-CoV-2) is underway in the United Kingdom. The Wellcome Trust is funding challenge stock preparation of the Beta and Delta variant for a follow-up human challenge trial, and researchers at hVIVO are considering conducting these trials. However, little has been written thus far about the ethical justifiability of human challenge trials with SARS-CoV-2 variants of concern. We explore 2 specific characteristics of some variants that may initially be thought to make such trials unethical and conclude that SARS-CoV-2 variant challenge trials can remain ethical.

controlled human infection model, COVID-19, human challenge trials, research ethics, vaccine ethics

In many countries, virus variants that partly evade authorized vaccines are reversing progress on coronavirus disease 2019 (COVID)-19. As we write, the Beta and Delta variants are of particular concern. Although authorized messenger ribonucleic acid (mRNA) vaccines protect against the Beta variant, non-mRNA vaccines that represent the majority of the prepurchased COVID-19 vaccine doses [1], as well as doses ordered by COVAX, show reduced efficacy against this variant. An AstraZeneca study [2] of 2000 people in South Africa found that “vaccine efficacy against [the Beta] variant, analyzed as a secondary endpoint, was 10.4%.” This is particularly worrisome, because AstraZeneca’s is the most widely purchased vaccine in the world. South Africa has sold their entire portfolio.

The Delta variant, which is now the dominant strain of severe acute respiratory syndrome 2 (SARS-CoV-2) worldwide due to its increased transmissibility, also partially evades vaccines. Although data from England shows that 2 doses of a Pfizer vaccine are highly protective against the Delta variant, approximately half of adults in Israel infected in the recent Delta variant outbreak were fully inoculated with the Pfizer vaccine [3], and a single dose is just 30% effective at preventing symptomatic disease against the Delta variant [4]. The US Food and Drug Administration recognizes that the emergence of vaccine-resistant strains warrants immediate preparation of strain-specific booster doses from all authorized vaccine candidates [5]. However, how can we test variant-specific booster doses, or regular doses in new regimens, for their efficacy against such variants in human participants? This Perspective recommends a challenge design.

THE LIMITS OF CONVENTIONAL DESIGNS FOR VARIANT-SPECIFIC TESTING

Immunogenicity studies that demonstrate adequate levels of neutralizing antibodies against variants of concern may be indicative of booster efficacy, per regulatory guidance in the United States [5] and Europe [6]. However, although some data show that neutralizing antibodies are predictive of protection against SARS-CoV-2 infection, this may not be the case for the Beta variant and future variants more resistant to neutralization [7]. Stronger data describing which immune responses correlate with protection against these variants would facilitate determining the efficacy of boosters by surrogate endpoint, because no correlate of protection has yet been confirmed for these variants [8].

Conventional clinical efficacy trials for boosters would face limitations. First, because efficacy studies require a critical mass of trial participants to encounter a specific strain, these trials may not produce interpretable results if a different strain predominates during the trial. Second, Delta has mysteriously waved in the United Kingdom (UK) and India, and subdued case rates may cause these trials to take many months, costing lives in the meantime. Third, withholding vaccination from thousands of placebo arm participants in an area of high community spread, without the high quality of informed consent that is realistic in a smaller challenge trial, is ethically fraught and unlikely to be approved by surrounding communities.

THE ROLE OF HUMAN CHALLENGE TRIALS WITH SEVERE ACUTE RESPIRATORY SYNDROME 2 VARIANTS

Challenge trials could assist in developing variant-specific vaccines—to be administered either as boosters for the already-vaccinated, or first doses for the unvaccinated—by helping discern a correlate of protection against different variants of SARS-CoV-2 [9, 10]. The more evidence that is gathered to establish a robust correlate, the likelier are variant-specific vaccines to be authorized by surrogate endpoint in a laboratory, accelerating their path to market.

Second, the efficacy of variant-specific vaccines in terms of curbing infection and infectiousness can also be tested directly in a challenge trial by administering volunteers a vaccine candidate and then deliberately exposing them to a variant of concern. Su et al [11] have outlined a plan to assess the efficacy of authorized and updated vaccines against SARS-CoV-2 variants. If unvaccinated volunteers are needed for such a trial but are difficult to find in the UK, trialists could open recruitment internationally. Ahmid et [12] al have argued that such recruitment can be conducted ethically so long as several robust protections are in place to avoid exploitation of low-income volunteers.

Third, challenge trials can be used to demonstrate whether a third dose of already-available mRNA vaccines actually protects against new variants. Third doses are thought to be more effective by increasing neutralizing antibody response, yet no neutralizing antibodies have been confirmed to correlate with protection against variants.

More broadly, variant challenge trials could offer unprecedented detail regarding variant kinetics and pathogenesis: “human challenge studies [on new strains] have the advantage that the course of developing immunity, viral shedding, local suppression of SARS-CoV-2 shedding in the upper respiratory tract and other parameters can be measured in a controlled setting” [13].

It might seem that virus variants would rapidly overtake one another, reducing the usefulness of challenge trials for formerly predominant variants. However, the risk of waiting so long that data become irrelevant is higher in longer field studies, where cases would be to an unpredictable mix of variants, making interpretation harder. Nevertheless, to prevent a process of constantly testing no-longer interesting strains, the entire challenge process should be reviewed for expediency within bounds of scientific validity and ethics, and not ones of habit or of red tape. Natural exposure trials could remove the need for time-consuming vaccine culture [14], and variant challenge studies could shorten dose-escalation studies by utilizing data from the ongoing UK dosing study.

In sum, variant human challenge studies could provide data on vaccines’ (including variant-specific vaccines’) and vaccine regimens’ efficacy in curbing infection and infectiousness rates against specific variants by discerning a correlate of variant protection, directly confirming the exact benefits of a third mRNA dose on specific variants, and broadly advancing our understanding of variant pathogenesis.

HUMAN CHALLENGE TRIALS AGAINST SEVERE ACUTE RESPIRATORY SYNDROME 2 VARIANTS CAN REMAIN ETHICAL

Human challenge trials involving variants that evade authorized vaccine protection could have momentous global public health value, yet little has been written about their specific ethics. Human challenge trials with the D614G mutation strain have been endorsed by the World Health Organization and the UK Medicines and Healthcare Products Regulatory Agency, and the Delta and Beta variants are being prepared for challenge trials [13, 15–17]. Elsewhere, researchers have made the case that COVID-19 human challenge trials can be acceptably safe for volunteers and that these trials can command fully free and informed consent [15]. Several decisions would need to be made in advance of a variant challenge study, including on volunteer payment and which of several designs to assess vaccine safety should complement challenge trials [18], but these decisions should not delay trial preparation.

Below we explore 2 ethical considerations that may initially motivate opposition to variant challenge trials and argue that although the trials’ risk/reward ratio will depend on the exact strain and type of trial, vaccine-resistant variant challenge trials can be ethically justified.

Transmissibility

The Alpha, Beta, and Delta variants appear to be approximately 50% more transmissible than the strain first identified in Wuhan [19]. However, increased transmissibility, as such, does not increase risk to human challenge participants. It only increases the social value of any challenge trial that speeds the authorization and rollout of vaccines, boosters, or regimens that better protect against a more transmissible and hence, for society, more dangerous variant.

Greater transmissibility does underline the importance of biosafe quarantine facilities to prevent inadvertent community infection. The WHO Working Group has outlined specific standards for biosecure facilities to prevent such transmission.

Virulence

Abu-Raddad et al [20] find that people infected with the Beta variant are 25% more likely to become severely ill and 57% more likely to die than people infected with the Alpha variant. Sheikh et al [21] find that people infected with the Delta variant are 85% more likely to be hospitalized than those infected with the Alpha variant.

However, the risks of severe adverse events in a COVID-19 human challenge trial (exposing only young and healthy volunteers to “earlier” strains) are so low—Manheim et al [22] find that the risk of death for a trial participant is approximately 0.00025%—that even a 2-fold increase in risks for a variant challenge volunteer would remain acceptably safe. For instance, trial risks would remain far lower than common comparator risks like live right liver lobe donation, which involves a 0.4% risk of death and a 1.1% risk of lasting disability [23]. Although Delta is currently landing even considerable numbers of young people in hospitals, the denominator is large thanks to its high transmissibility; a 2-fold or even a wholly unrealistic 100-fold greater risk for young healthy volunteers would remain well within postulated upper bounds of risk in clinical trials, such as the organ donation comparator stated above and certainly a 1% cap on risk of severe adverse events, the other standard echoed by critics of challenge trials [24]. If challenge volunteers are vaccinated before the trial, risks will be further reduced considerably. The availability of effective therapies such as bamlanivimab plus etesevimab would increase the safety of challenge trials [25]. To reduce risks more, researchers can also use low-dosage challenge trials with virus variants [26].

CONCLUSIONS

Vaccine-resistant viral variants threaten to get us back to square one in the fight against the pandemic [27]. Human challenge trials can accelerate the authorization of variant-specific vaccine boosters and otherwise advance our understanding of and ability to fight these variants successfully. Such trials can remain ethical, and researchers should begin preparations for them immediately.

Notes

Acknowledgments. We are grateful to Ginny Schmitt, Randall Kincaid, Garth Rapaport, Josh Morrison, and David Manheim for their insights and feedback on this paper. Being mentioned does not imply any endorsement of our work.

Financial support. A.R. reports funding from 1Day Sooner. N.E. reports funding from the National Science Foundation (Award Numer 2039320; “RAPID: Responsible conduct of accelerated SARS-CoV-2 vaccine research through challenge trials”) and Open Philanthropy (no award numbers).

Potential conflicts of interest. N.E. serves on the Board of Advisors of 1Day Sooner, an unpaid position. A.R. reports affiliation with 1Day Sooner throughout the writing of the article. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.