Infect volunteers with Covid-19 in the name of research? A proposal lays bare a minefield of issues

In 2016, a team of scientists asked the National Institutes of Health for permission to develop a “human challenge model” for Zika virus infections — a sometimes controversial tool to fast-track research on vaccines and drugs. Under the model, healthy volunteers would be infected in a highly controlled setting.

An ethics panel denied the application in January 2017, saying that deliberately infecting volunteers with Zika — which is innocuous to most people but can profoundly damage the brains of fetuses infected in the womb — would pose too much risk to the participants and their sexual partners.

More than three years later, other scientists are advocating for human challenge trials to test vaccines against the coronavirus that causes Covid-19 — an infection that is far more dangerous than the one caused by the Zika virus.

“A lot has changed in 3 1/2 years,” Anna Durbin, a vaccine researcher at Johns Hopkins University who pushed for the establishment of a Zika challenge model, noted dryly in a recent interview with STAT.

The idea of controlled human infection trials, as they are also called, for Covid-19 research was first raised in late March in an article in the Journal of Infectious Diseases. Since then, the proposal has gained momentum, with other experts making a similar proposal. Last month 35 members of the House of Representatives wrote to the Food and Drug Administration and the Department of Health and Human Services, arguing such trials should be allowed. There is even a website, 1daysooner.org, set up to encourage people to volunteer to be in human challenge trials or to advocate on their behalf.

Durbin and a number of other researchers who have done these studies and who believe fervently in their value find themselves in the paradoxical position of arguing for caution and restraint. They note that developing the challenge model itself is a painstaking endeavor that would take over a year, perhaps even two. A scientist who works for a contract research organization that conducts challenge studies estimated the timeline at 12 to 18 months.

In either scenario, some of the earliest vaccines may already be in Phase 3 studies — the large trials meant to show if a vaccine is protective — by the time the model is ready to be put to use. Still, proponents of human challenge studies say they could help sidestep Phase 3 trials for some potential vaccines.

Skeptics, meanwhile, note that there is still no treatment for the disease that has been shown to prevent deaths, arguing that without a way to cure any volunteers who might go on to develop severe illness in a trial, a challenge study cannot be conducted.

“I think one needs to have therapy before one would undertake a challenge. That’s my personal view,” said Myron “Mike” Levine, associate dean for global health, vaccinology, and infectious diseases at the University of Maryland. Levine has conducted human challenge trials for decades, using this approach on research into cholera, shigella, and malaria, among other diseases.

Levine said this week’s news that the drug remdesivir seems to speed recovery in some patients, while welcome, does not assuage his concerns. It remains to be seen, he said, whether the drug “can reliably halt the progression to severe and fatal disease.” The clinical trial results reported Wednesday did not show a lower death rate among people who were treated with remdesivir.

The debate over challenge trials is growing heated, even if both proponents and skeptics have the same goal: to develop safe and effective Covid-19 vaccines as quickly as they can be made.

One of the proponents, Stanley Plotkin, a professor emeritus of pediatrics at the University of Pennsylvania and an éminence grise in the world of vaccine development, acknowledged in an interview that there will be ethical questions around subjecting healthy volunteers to a virus that occasionally kills even young, seemingly healthy people. But he said there are ethical quandaries in running a traditional Phase 3 trial in which some people receive a placebo as well. Given that a Phase 3 trial will have to involve a large number of people — thousands, maybe more — he noted there could be deaths in the placebo arms of Covid-19 vaccine trials.

“So we’re talking about two different methods, both of which have faults,” Plotkin told STAT. “But the issue from my point of view is to speed up emergency use of vaccines that are shown to be protective.”

Plotkin is pulling together a proposal aimed at getting major players in the search for Covid-19 vaccines on board, including the World Health Organization. One of his targets includes the Coalition for Epidemic Preparedness Innovations — CEPI — which he hopes will fund development of the challenge model, work that could in itself cost millions of dollars.

Katherine Littler, co-lead of the global health ethics and governance unit at the WHO, said her group is drawing together guidance for governments on how to make ethical decisions on whether to allow human challenge studies, but the document — which she hopes will be published on the agency’s website in the next week or two — will not advocate for or against the studies.

Among the advantages of a challenge study is that, by using small numbers of people in a very controlled setting, scientists can rule in or rule out experimental vaccines or drugs faster than if they had to vaccinate, say, 10,000 people — half with a vaccine, half with a placebo — and wait until there were enough illnesses among the trial volunteers to see if infection rates were lower in the vaccinated group.

But in order to conduct human challenges, scientists need a challenge model for Covid-19, and there currently isn’t one. Just developing the model could take between 18 months and two years, said Beth Kirkpatrick, who runs a human challenge trials unit at the University of Vermont, where she is chair of the department of microbiology and molecular genetics.

“The question of whether this is going to speed up vaccine development, I think, is maybe forgetting how much time it takes to develop the model itself before you can use it to test anything,” Kirkpatrick said.

Because this type of work involves deliberately infecting people, the Food and Drug Administration closely regulates it at every step, she said. Researchers have to identify and grow supplies of a strain of the virus to be used in the challenges (work that would need to be done in a biosecurity level or BSL-3 laboratory), prove that it contains no other pathogens or “adventitious agents,” as they are known, test it in animals and get it certified for use, Kirkpatrick said. Those steps alone can take six months.

Once the challenge strain is approved, dosing studies would need to be conducted to determine how much virus to expose volunteers to in the actual vaccine trials. Because of the danger, the researchers would have to start with low doses and work their way up, a process that would take months.

Marc Lipsitch, an epidemiologist at the Harvard T.H. Chan School of Public Health and an author of the March paper proposing the studies, doesn’t think establishing the challenge model would take “quite so long.” Plotkin believes a challenge model could be established in between three to six months.

But the fact that many people infected with Covid-19 have no or few symptoms could add to the complexity. Volunteers in the trials would need to be exposed to enough of the virus to develop detectable illness, said Durbin. But how many symptoms, and for how long would they need to be symptomatic?

“What are your endpoints going to be? You know, do you want to give people pneumonia? I hope not,” Durbin said. “You want people who feel crappy to some extent, who definitely know they’re sick. And what that balance is or what that line is between just feeling crappy and then developing pneumonia, I think one of the scary things is we really don’t know what that is.”

Christine Grady, chief of the department of bioethics at the NIH clinical center, agreed that the absence of drugs to cure Covid-19 and the fact that the disease can cause fatal illness in some young, healthy adults makes the idea of challenge studies difficult at present.

“We don’t yet know why some people get sick and others don’t or why some people get certain manifestations of … Covid that others don’t get,” Grady said. “There’s so much emerging information about this sort of clinical course of infection and also susceptibility to infection that it makes an assessment that it’s OK to subject a certain age group to risk a little bit too fast for me.”

There are also concerns about whether one could conclude — based on the findings of challenge studies in young, health adults — that a vaccine would be protective in other age groups.

“You would only know that that works in healthy young adults, probably a very narrow age,” said Levine. “And that may or may not be extrapolatable information to the population with the mortality, which is, you know, 60s and 70s, and 80s.”

Levine and Mark Feinberg, head of the International AIDS Vaccine Initiative, favor compressing the timelines for studying vaccines in the way Merck’s Ebola vaccine during the West African Ebola outbreak of 2014-2016 was handled.

The Ebola vaccine was tested with unprecedented speed, going from a “first-in-humans” Phase 1 to completion of a Phase 3 efficacy trial in 10 months.

Feinberg, who was chief scientific officer for Merck at the time, believes timelines for Covid-19 vaccines will be similarly or even more aggressively compressed. Already, some of the vaccines in development are taking similar steps. The vaccine being developed at the University of Oxford is being tested in a blended Phase 1 and 2 trial; if all goes well, that group says, the vaccine might be ready for emergency use as early as September.

“I think it may be possible that, you know, this the advocacy for these controlled human infection studies may wane as the pace of other developments [increases],” he said. “I think we are seeing candidates entering the clinic with a much shorter time lag than is typically the case and actually has ever before been seen.”

Skeptics of challenge studies worry about whether they will produce enough safety data if they replace Phase 3 studies.

Research more than a decade ago on SARS vaccines — work that was shelved when it became apparent there was no more SARS, and no more SARS vaccine market — showed that some vaccines appeared to make mice more vulnerable to severe infection. Watching Covid-19 vaccines to ensure they don’t behave in similar fashion is critical.

But that kind of phenomenon might not be detected in challenge studies because they would involve so few people, Durbin said. She estimated between 50 to 100 people would be needed to show if a Covid-19 vaccine was protective.

Lipsitch countered that if a candidate Covid-19 vaccine increases the risk of infection or severe disease in a small number of people who get vaccinated — a one in 100,000 or a one in a million event, say — that won’t be detected by a Phase 3 trial either.

He said trying to determine where to conduct standard large Phase 3 trials could be difficult. If one area drives down its Covid-19 transmission rate, a trial might take a long time to gauge if a vaccine works because there is too little spread of the virus. In places where the risk of getting infected remains high, people may be ordered to avoid as much contact with others as possible, or may adopt physical distancing of their own accord. Either reality could drag out the time to an answer.

“Given all the uncertainty about the trajectory of the virus … picking a site for a large scale and rapid Phase 3 trial is not at all trivial,” Lipsitch said.

The other concern is that there are nearly 100 vaccines in development. Most probably won’t progress, but if enough reach the point where efficacy data are needed — given how much vaccine is required, worldwide, this is to be hoped for — challenge studies done on the first vaccines could help fast-track those that come later.

The idea is that these studies might help establish what are known as the “correlates of protection” for Covid-19 — mapping out which parts of the immune system need to be activated and to what degree if someone is to be protected against this infection.

If challenge studies show that this level of these antibodies or a particular type of T cells needs to be activated, vaccines that can hit those targets could be deployed more rapidly.

Grady is among those who feel it’s too soon to try human challenge studies, at least until drugs to prevent illness from becoming life-threatening are in place, or until there’s a way to determine who cannot safely take part in such studies. But that doesn’t mean the idea should be entirely abandoned, she said.

“I wouldn’t take it off the table, but I certainly wouldn’t say we’re ready for it now,” Grady said. “And I certainly wouldn’t let it divert activity from other ways of testing vaccines.”

Kirkpatrick agreed. “I don’t think I would always be a ‘no’ for a human challenge model. But I think right now we do not have the data we need to proceed in a safe and ethical way.”