Jessica Flanigan argues that human challenge studies are ethical and should be used widely in both the COVID-19 and future pandemics. Human challenge studies involve intentionally exposing healthy volunteers to an infectious agent, in this case the SARS-CoV-2 virus, in a controlled environment. She argues that challenge studies would speed vaccine development, including vaccines for viral variants, and pose acceptable benefits and risks to participants. Further, she argues that prohibiting volunteers from participating in a SARS-CoV-2 challenge study is paternalistic and wrong. In my essay, I take issue with each of these claims.
First, Flanigan tells us that SARS-CoV-2 challenge studies “enable researchers to test the safety and efficacy of vaccines and other therapeutics more quickly, which has substantial public health benefits.” On these grounds she asserts that had challenge studies been permitted when they were first proposed, an effective vaccine would have been developed more quickly. As such, “the invisible graveyard of people whose deaths could have been prevented by faster access to vaccines” could have been avoided. But even though we now have safe and effective vaccines in hand, challenge studies could still save lives, we are told. They could “speed the pace of mass vaccination” by, for instance, telling us the efficacy of a single dose of vaccine. And challenge studies may be pivotal “to test new vaccines as new variants of Covid-19 emerge.”
Flanigan is quick to point out that critics of challenge studies make empirical claims without citing evidence. (Referring to the impact on public trust, she says: “Like the previous objections, this is largely an empirical claim, and the people who make it don’t cite much evidence from existing challenge trials to support this claim.”) It is surprising, therefore, that she presents no evidence to support her empirical claims. Indeed, experience with human challenge studies and available evidence do not support them.
In contrast to the narrative presented, traditional vaccine development has been bold, efficient, and exceptionally successful. The first cluster of cases of COVID-19 were reported in December 2019. The genome of the virus was released on January 10, 2020. Three days later, the first vaccine candidate had been developed, and it was first tested in human beings 63 days later. By September 2020, there were hundreds of vaccine candidates, 27 of which were in early human trials and six of which were in large phase 3 trials. In November 2020, results from the first of these trials showed two vaccines to be safe and highly effective. As of March 2021, there are now 12 vaccines—including mRNA, viral vector, protein subunit, and inactivated virus vaccines—that have limited or full approval and are in current use.
SARS-CoV-2 challenge studies could not have speeded the development of a vaccine and, consequently, they would not have saved any lives. Why? Human challenge studies require 12 to 18 months of background work before a vaccine can even be tested. For example, the two most recent influenza challenge models took 15 months and 19 months to develop. The first step involves the production of a Good Manufacturing Practice challenge stock that is both stable (meaning that the virus has not acquired mutations) and produces infections consistently in animal models. The second step involves developing the challenge model, that is, the protocol defining the route of administration and viral dose to reliably produce infection while avoiding serious illness. Healthy volunteers are exposed to increasing doses of virus, with each dosing step taking about 1 month to complete. Had work on SARS-CoV-2 challenge studies begun when first proposed, we would still not be ready to test vaccines with them.
Further, if a new vaccine is shown to prevent infection in a challenge study, all that means is that it prevents disease due to a single strain given by an artificial route of inoculation. One or more large trials are still required to demonstrate its safety and effectiveness against the many variants that are currently circulating. Because vaccines are administered to millions of healthy people, safety is a primary concern. Only a trial involving tens of thousands of people can reliably rule out rare serious adverse events. The fact that challenge studies involve perhaps 100 healthy volunteers 18–29 years of age means they provide little information on safety. And studying healthy twentysomethings tells us nothing about how a vaccine will perform in people with comorbidities or the elderly, the groups at highest risk of serious illness, hospitalization, and death.
Flanigan shrugs off this concern, noting that “both approaches struggle with external validity…[and] data generated from either kind of trial design can still be useful to researchers and potential consumers.” But this falsely equivocates between a case in which we have data from a hundred unrepresentative people and one in which we have data from a much bigger, heterogenous sample of individuals.
SARS-CoV-2 challenge studies are also unlikely to play a role in developing modified vaccines for viral variants of concern. For each viral variant, the steps to create a challenge stock and a challenge model would have to be completed afresh, and this will take 12 to 18 months. Two things will almost certainly have happened by then: traditional vaccine development methods will have produced a vaccine, and new variants of concern will have emerged. Recent events with the B.1.351 variant first described in South Africa are illustrative. Moderna has announced that it has already developed and manufactured a booster vaccine for the B.1.351 variant. As the booster modifies a vaccine of proven safety and effectiveness, regulators have indicated that only immunogenicity studies will be required for approval and, consequently, the booster vaccine is likely to be rolled out by September 2021.
In her second argument, Flanigan asserts that SARS-CoV-2 challenge studies pose acceptable benefits and risks to healthy volunteers. She says that volunteers may benefit by participation, that many volunteers “are altruistically motivated, and they would find it meaningful to participate in such beneficial research.” Further, such studies may fulfil the ethical requirement of equipoise “if researchers are uncertain about whether participating in the trial is safer than abstaining and risking unvaccinated exposure.”
But challenge studies do not confer medical benefits on participants. In challenge studies, healthy volunteers are not given treatment; rather, they are exposed to an infectious agent that causes disease. This is important as the concept of equipoise applies to treatments (such as drugs, surgery, or behavioral therapy) administered on the basis of evidence that suggests they may benefit research participants. Thus, equipoise cannot obtain in the challenge studies.
Because challenge studies offer no prospect of direct benefit to volunteers, it is essential to limit the risk to which they may be exposed for scientific ends. This is why scientists go to extraordinary lengths to protect healthy volunteers in challenge studies. Scientists limit challenge studies to infectious agents that are well understood. Further, most challenge strains cause only asymptomatic infection or mild disease. For example, influenza challenge strains are limited to isolates that cause mild disease. Where an infectious agent may cause serious illness, the availability of rescue medication is required. For example, malaria challenge strains are exquisitely sensitive to a single drug, chloroquine, which is administered at the first sign of infection. Thousands have volunteered in challenge studies since the 1970s, and not one volunteer has died as a result of their participation.
SARS-CoV-2 challenge studies are controversial precisely because they would abandon these safeguards. SARS-CoV-2 infection carries with it the risk of serious illness, disability, and death in young, healthy adults. Viral pneumonia is the most common manifestation. But the virus also infects the blood vessels, the brain, and the heart, causing blood clots, strokes, and myocarditis (infection of the heart). A minority of those with COVID-19 develop disabling symptoms that are long lasting—so called, “long covid.” Finally, there is no curative treatment for COVID-19. Remdesivir will be given to participants in the UK challenge study, but the Solidarity trial concluded that it does not prevent COVID-19 deaths.
In her final argument, Flanigan claims that prohibiting SARS-CoV-2 challenge studies is paternalistic. The grassroots organization 1Day Sooner has recruited over 38,000 people from 166 countries who say they would be willing to volunteer for a challenge study. In not allowing them to participate, Flanigan says we are failing to respect their autonomy. “Just as adult patients are generally capable of consenting to other medical procedures…they are capable of consenting to participate in challenge trials.” Flanigan further argues that “the freedom to make a risky choice is especially morally urgent…when granting someone the freedom to take a risk is likely…to promote the greater good.”
But paternalism is not morally problematic if justified. Consider the case of a living heart donor who wishes to have her heart removed so that a brilliant COVID-19 researcher who urgently requires a transplant may live. The donor is motivated by the desire to save the life of another, understands that she will die when her heart is removed, and is competent. Indeed, not only will the donor’s actions save a life, the COVID-19 researcher may, through her work, save countless other lives. Ought we allow the living heart donation?
The answer is “no,” of course. Even though the donor’s choice is autonomous and promotes the greater good, it wrongly interferes with the duties of others. The donor’s physicians have obligations not to harm her and to promote her medical interests—referred to as a “duty of care.” The living heart donor’s freedom to give someone else her heart is rightly restricted because it would violate the physician’s duty of care. In human challenge studies, physician-researchers also have a duty of care to research participants. Knowingly exposing volunteers to a virus that may cause serious illness, disability, and death violates the physician-researcher’s duty of care to the participant. Absent rescue medication, restricting the freedom of volunteers for SARS-CoV-2 challenge studies is justified.
In closing, I would like to respond to Flanigan’s assertion that the more than 500,000 deaths from COVID-19 in America are “due largely to the needless delays associated with administrative incompetence and excessive caution.” Many people in the United States died needlessly during the COVID-19 pandemic. They died due to a degraded public health infrastructure, leadership that sidelined public health experts, and federal, state, and local governments that failed to implement mandatory mask wearing, social distancing, and closure of venues where transmission is known to occur. These failures were not due to excessive caution. They were motivated, at least in part, by the very same libertarian ideals that inform the author’s argument.
I am grateful to Jessica du Toit (Western University), Hayden P. Nix (Western University), and Dr. Stanley M. Spinola (Indiana University School of Medicine) for their helpful comments on an earlier draft of this paper.
I receive consulting income from Cardialen, Eli Lilly & Company, and the Research Triangle Institute International.