On March 31, 2020, three researchers proposed a big idea to speed up the development of a vaccine for the SARS-CoV-2 coronavirus (which causes Covid-19): directly exposing human volunteers to the virus.
Bioethicist Nir Eyal and epidemiologists Marc Lipsitch and Peter Smith suggested conducting “human challenge trials” on willing volunteers. In such trials, volunteers are dosed with either the candidate vaccine or a placebo and then directly exposed to the coronavirus. That way, researchers would know that 100 percent of participants had been exposed, making it easier to judge the vaccine’s effectiveness. In normal vaccine trials, the treatment group is given the vaccine and the control group a placebo, and the volunteers in both groups go about their regular lives, where they may or may not encounter the virus in the wild.
Doctors are generally not in the business of purposefully exposing people to dangerous viruses, so the proposal was met with some controversy. But Eyal, Lipsitch, and Smith are hardly alone in their thinking. “I proposed challenge trials to Dr. Fauci’s office in early March for SARS-CoV-2, well before March 31,” says Matthew Memoli, director of the Laboratory of Infectious Diseases Clinical Studies Unit at the National Institute of Allergy and Infectious Diseases (NIAID). In the subsequent two months, the research community and its regulators have begun serious discussions that move the world closer to actual challenge trials for the coronavirus.
A letter authored by Reps. Bill Foster (D-IL), a former research physicist, and Donna Shalala (D-FL), a former secretary of Health and Human Services, and signed by 33 other members of Congress from both parties, expressed support for the use of human challenge trials. The World Health Organization has released a working paper titled “Key criteria for the ethical acceptability of COVID-19 human challenge studies,” and National Institutes of Health Director Francis Collins instructed a vaccine working group at the NIH to write a paper on the issue.
Collins and NIAID Director Anthony Fauci also co-authored a research paper in the journal Science with Lawrence Corey and John Mascola addressing the possibility of challenge studies.
Meanwhile, the advocacy group 1 Day Sooner has signed up more than 23,000 volunteers in 102 countries who have offered (in a non-binding way) to participate in such studies.
While there is momentum behind the idea, logistical hurdles remain, and preparatory work has yet to begin in earnest. The main task that could be done right now is mass-production of the virus itself: actual SARS-CoV-2 viruses prepared for dosing into volunteers. “We should be working on making viruses that could be used for this, I think,” Memoli says.
Memoli has been conducting human challenge studies of influenza vaccines since 2011. He emphasizes that he speaks only for himself, not for NIH or NIAID as a whole. “If we can’t make [the viruses] then we know we can’t do this, and if we make them then we have them” in case they’re needed, he explains.
Beyond that, preparatory work could be done to make sure a challenge trial is logistically doable. This includes a preliminary study to determine the specific dose of the SARS-CoV-2 virus to which volunteers would be exposed. Both the preliminary study and the wider challenge study require large medical facilities close to (or part of) hospitals. They need nurses and other medical staff, personal protective equipment (PPE) for caretakers, and funding for all of the above.
There are also ethical and regulatory barriers. A challenge trial would have to be approved by the Food and Drug Administration (FDA) and by the institutional review boards (IRB) of whatever institutions are involved in the study.
Whether or not challenge trials actually take place, there is a lot of preparatory work that needs to be done right now if trials are even hypothetically capable of saving time on the road to a vaccine.
Why conduct a human challenge study?
Clinical trials of vaccines and other treatments usually proceed in four phases. While the exact role of each phase differs subtly based on the lab or company, the division is usually something like this:
- Phase I trials are the first step and involve safety testing to make sure the vaccine isn’t actively dangerous;
- Phase II is a small-scale efficacy trial, meant to test if the vaccine is protective against infection;
- Phase III is an expanded, larger-scale efficacy trial including more varying groups of people, like the elderly or people with preexisting conditions, whose immune systems might operate differently; and
- Phase IV studies evaluate drugs after they’re officially on the market and being used.
In most Phase II and Phase III studies for vaccines, participants are exposed to the disease being targeted only in the course of their regular lives. That creates a potential problem in a situation like the Covid-19 outbreak.
If you’re a healthy person who’s practicing social distancing, you probably aren’t going to get infected — or if you are, it’ll probably take some time for the infection to occur. That makes it hard to do a real comparison between the vaccinated and unvaccinated groups. Maybe they both end up the same because no one in either group was exposed to the pathogen. If that’s the case, the trial hasn’t told us anything useful. Alternatively, adequate exposure to the virus might happen only after many months or a year, slowing down the testing process.
That’s the case for using challenge trials, in which participants are immediately exposed to the virus, with no delays or uncertainty about the share of volunteers who have actually been exposed and thus have had their vaccinations tested adequately. Eyal, Lipsitch, and Smith have argued for a challenge trial (or trials) to replace Phase III, the last step before deployment. But they’ve also been clear that challenge trials can’t be used on the elderly or immunocompromised, which limits their value as a Phase III tool.
Advocates of human challenge trials for SARS-CoV-2 generally agree that only young, healthy participants with the lowest possible mortality risk from the virus should be purposefully infected; it would be unacceptably risky to purposely infect an elderly person or someone with hypertension. That’s why Memoli argues that a challenge trial might be more appropriate as a replacement for Phase II, with Phase III involving normal trials among various groups.
Using challenge trials with only young, healthy people raises a problem that Fauci, Collins, Corey, and Mascola note in their Science paper: “Partial efficacy in young healthy adults does not predict similar effectiveness among older adults with major cofactors associated with COVID-19 disease, nor would it prove reduction of transmissibility to major susceptibility groups.”
In other words, a vaccine that works on young adults may not work on everybody else, particularly people who are most vulnerable to infection. That’s another reason Memoli suggests using a challenge trial as a Phase II test, to be followed by a traditional Phase III that would include a broader set of research participants. We would first see if the vaccine works on young people, then test it on everyone if it does. “It would be highly unlikely that it’s protective in old people and kids if it’s not protective in young healthy people,” Memoli says.
The inverse — “If it’s protective in young healthy adults, it’s protective in kids and seniors” — is not necessarily true, but if a vaccine works in young adults and not in other groups, that gives researchers valuable information that could help them alter the dose or add adjuvants so that it does work in those other populations.
A challenge study could also help enhance researchers’ understanding of the underlying virus, even if the vaccine being tested doesn’t work. “You know exactly who these people were, everything about them,” Memoli says. “You could be measuring gene response hours after giving the virus. … If the vaccine works or doesn’t work, you can actually get data that allows you to determine why the vaccine worked or didn’t work.” Larger-scale traditional studies generally don’t produce data that granular or useful.
The nuts and bolts of a human challenge study
Let’s say there’s a vaccine that’s made it out of Phase I testing. It’s safe for human consumption, and now researchers have to figure out if it works. A couple of vaccine candidates seem near or at this point, so this is a likely scenario in the next few weeks or months. What needs to happen, then, for a human challenge trial to work?
As Memoli explains, you would first need to develop a model of the virus with which to expose volunteers. The process of choosing the dose you want to use and building a model of what to expect in a challenge trial is a bit like conducting a challenge trial in miniature.
First, researchers give a very weak dose of the virus to a small group (say, five people). If enough of them (say, three out of five) get sick, then researchers get 10 to 15 more people and dose them. If enough of them get sick (Memoli suggests 70 percent as a threshold), then researchers know they’ve picked a strong enough dose.
But the lowest dose sometimes doesn’t get enough people sick to be useful. So then researchers up the dose a bit and dose five more people to see if enough get sick. This process continues until they eventually have a sample of 20 people who’ve all gotten the same dose — and enough of whom have subsequently fallen ill — to clear the threshold for how potent the dose has to be.
This whole process typically takes about 50 people, Memoli says. A riskier, sped-up version would involve testing multiple doses simultaneously or closely staggered: a very low dose to five people, a higher dose to five more, a yet-higher dose to five more, and so on. The weakest dose to pass the threshold then gets ramped up to another 10 to 15 people until you get the sample of 20 necessary to proceed.
Once that’s all done and a dose is determined, researchers need enough of the virus to expose a full study (40 to 50 people per “arm” of the study: 40 to 50 for the control, and 40 to 50 for each candidate vaccine). They’ll also need enough medical staff, with PPE and other necessary gear, to care for that many people.
And they’ll need a secured biosafety facility that can house all those participants for about a month, with sufficient distance between them that unnecessary, unplanned infections don’t occur; indeed, there might not be enough existing facilities at biosafety level-3 or above to support a trial, necessitating new construction. The facility should ideally be close to or part of a hospital complex so that participants whose condition deteriorates get the care they need.
That all requires significant funding, be it from the federal government or private business or philanthropy, and a lot of manpower and facilities. In practice, Memoli says, NIH or foreign equivalent agencies are probably the only ones with the resources and money necessary to get it done. “I don’t know what Fauci or Collins think about this or what they want to do, but I absolutely think we’re capable of it,” Memoli says. “We can do it ourselves, or with universities or contract research organizations.”
The ethical and regulatory barriers to a human challenge trial
Memoli notes that these resource and monetary barriers are only half the battle. Researchers would also need to satisfy regulators and bioethicists who are monitoring to make sure the trial is done safely.
The main legal barrier in the United States is the approval of the FDA, to which vaccine testers/developers must apply before any trial, including human challenge trials (the European Drug Administration does not require the same). Specifically, the FDA usually has said that proposals to expose participants to a virus for challenge trials have to go through the Investigational New Drug (IND) application process. That can be time-consuming, but Memoli notes there are ways around it. And the IND process has the advantage that once producing the virus has been approved, approval can come much quicker for each subsequent challenge trial because the FDA already has background information on the virus.
But the FDA’s approval isn’t the only requirement. Basically all biomedical research institutions use institutional review boards (IRBs) to evaluate research proposals involving human subjects to make sure they’re methodologically and ethically sound. The NIH’s IRB, and those of drug companies or universities involved in a challenge trial, would have to approve moving forward.
The NIH also has a dedicated bioethics faculty, and many hospitals have staff bioethicists or academics they consult on tricky bioethical issues. “When I first started doing challenge studies over 10 years ago now, before we started to do any of them we did a full bioethics review with the whole NIH bioethics faculty,” Memoli recalls.
And the bioethics community, and biomedical research community more broadly, hardly have a consensus on the acceptability of challenge trials for SARS-CoV-2. Some people have expressed the view that without a reliable treatment for Covid-19, a challenge study of the underlying virus is inappropriate. Myron Levine, a professor and veteran conductor of challenge studies at the University of Maryland, told Stat’s Helen Branswell, “I think one needs to have [a treatment] before one would undertake a challenge. That’s my personal view.”
Others disagree. “It is a common misconception that to do a challenge study you must have a specific treatment for the disease,” says Seema Shah, professor of medical ethics at Northwestern University Medical School and a practicing medical ethicist at Lurie Children’s Hospital of Chicago who has written about the ethics of human challenge trials. “Challenge studies are conducted with dengue and influenza, where enough is known about the disease or the strain that was selected that the risks to participants are lower and can be managed.”
Josh Morrison, founder of the volunteer group 1 Day Sooner, argues that the decision ultimately has to be up to volunteers. Bioethicists and researchers should be clear with volunteers about what the risks are, and that those risks are serious. “Deciding to participate in a study like this is not something anyone should take lightly,” Memoli notes.
But Morrison argues, “If challenge trials are likely to benefit society, and well-informed people want to participate, we should respect their freedom to do so. Ethicists have an important role in laying out the process and the criteria volunteers should consider, but ultimately the decision should be up to those volunteers. It would be paternalistic to stifle such acts of public service and perverse to do so when the benefits are so potentially enormous.”