ABOVE: Volunteer Molly Perry receives an injection of malaria sporozoites from Jim Kublin at the Fred Hutch–based Seattle Malaria Clinical Trials Center as part of a challenge study.

As the 2009 pandemic flu virus infected millions around the world, Matthew Memoli began planning to expose healthy volunteers to the pathogen—a proposal that met with considerable opposition. The scientific community was divided on whether the approach was ethical or even warranted. 

As the director of the Clinical Studies Unit at the Laboratory of Infectious Diseases at the National Institute of Allergy and Infectious Diseases (NIAID), Memoli recognized the value of a so-called challenge trial for flu, and there was precedent for conducting one. “Our current vaccine and most of the antivirals that we use for flu were all developed in some part because of challenge studies,” says Memoli. But at the time he was applying for regulatory...

He made his case for reviving the approach to address the ongoing pandemic virus. He argued to numerous groups that a controlled exposure that mimicked what people could encounter in the real world would allow researchers to track subjects’ immune responses and test drugs and vaccines efficiently and with far fewer people than would be needed for traditional clinic trials. In the end, the US Food and Drug Administration (FDA) and the National Institutes of Health (NIH) ethics review board agreed with him, and he developed a challenge model using a wildtype strain of H1N1 that has since been sprayed up the noses of more than 400 volunteers.

While I’m a proponent of challenge models, I think there are some ethical requirements and those include having a significant knowledge of the long-term effects of the infection.

—Monica McArthur, University of Maryland School of Medicine 

“I personally take this very seriously. I’m making people sick on purpose. It’s not a joke; it’s a very serious thing that I’m doing. And I only want to be doing it if it really is truly something that is beneficial,” Memoli says. 

Ten years later, Memoli is once again working on developing a human challenge model for a pandemic virus—this time, SARS-CoV-2. It’s not yet clear whether it will be used to test out vaccines or even therapeutics for COVID-19, he says, but “I have always believed that we should be preparing to do it.” 

Since the early days of the COVID-19 pandemic, a debate has focused on whether a challenge study for COVID-19 vaccines should proceed. In June, an advisory group to the World Health Organization (WHO) issued a report providing guidelines for how such trials could be designed, but the committee was undecided over whether they could ethically proceed. Late last month, the Financial Times reported that the UK government announced it was considering a vaccine challenge trial for COVID-19 that could start early next year. (Imperial College London and the contract research organization hVivo, reportedly partnering on the trial, did not respond to requests for comment before deadline.)

For many pathogens other than SARS-CoV-2, challenge studies have already become a key part of research on vaccines and therapeutics. Besides influenza, scientists have developed challenge models for an array of other viruses as well as bacteria and parasites. “I think challenge studies have become more accepted, honestly,” says Memoli. “Since I started doing the flu challenge, I think there was a big resurgence.”

Determining the challenge

The first step in conducting a challenge study is to design the challenge—that is, what the researchers will expose study participants to. The strain, dose, and administration are considered, and the approach varies widely depending on the pathogen. The goal is to elicit a consistent response in a majority of volunteers while ensuring everyone’s safety. 

When clinical researcher Anna Durbin of Johns Hopkins Bloomberg School of Public Health and her collaborators at NIAID began developing a challenge model of dengue several years ago, they started with two naturally attenuated strains they hoped could serve as proxies for the more dangerous forms of the virus to test vaccines; the goal wasn’t to make people sick but simply to infect them. The researchers wanted to be able to see the virus in cell cultures grown from blood samples from at least 80 percent of volunteers before they began challenge trials to test an experimental vaccine. 

There is no effective rescue treatment and we have no way of predicting who may have serious complications due to infection.

—David Diemert, George Washington University

The researchers started with 10 participants who received 1,000 virus particles by subcutaneous injection, and all 10 developed an infection. They had an average of only about 300 virus particles per milliliter of blood, however—a far lower concentration than the 10,000 to 100,000 viruses per milliliter of blood typical of natural dengue infections. Several volunteers got a typical dengue rash over their whole body, but none got a fever. In 2016, Durbin and her colleagues published the results of a challenge trial of 41 volunteers showing a vaccine candidate to be 100 percent effective—none of the 21 participants who received the vaccine developed viremia when challenged with the virus whereas all 20 who didn’t get vaccinated did—prompting Brazil’s Butantan Institute to license the vaccine technology and launch an ongoing Phase 3 trial.

In other cases, researchers use the wildtype pathogen in an attempt to replicate natural disease. For challenge models of influenza, Memoli says, it’s important to actually induce illness. When it comes to respiratory viruses such as flu and SARS-CoV-2, he explains, a vaccine might not prevent infection, but rather reduce symptom severity, shorten the course of the disease, or limit the pathogen’s spread—effects researchers may miss if the challenge simply causes infection but not sickness. 

So far, Memoli says, developing a challenge model for SARS-CoV-2 has been “a bit simpler” than it was for flu, partly because the novel coronavirus has not yet evolved significantly different strains. “Really all the SARS-CoV-2 viruses that have been identified have pretty much the same properties,” he notes, and moreover, the coronavirus has proven relatively easy to grow in the lab. But there’s still a long way to go, Memoli adds, noting that his protocol must still be reviewed by the FDA—which must approve the proposed pathogen challenge as an investigational new drug before it is administered to people for an initial validation study—and an NIH institutional review board, as well as other committees. “We have lots to do over the next few months to even determine if everyone felt comfortable doing this before we could even make the decision to move forward,” he says. 

If a challenge trial were to proceed, the first study would simply serve to validate the model, enrolling only 5 to 10 people to receive a relatively low dose of pathogen. If the exposure proved safe enough, the researchers would expand the study and likely test higher doses until they are able to trigger illness in enough participants—typically at least 60–80 percent—to power future vaccine and drug trials.

With many labs now having characterized viral stocks of SARS-CoV-2, “that standardization and the qualification of the challenge material is one of the challenges that the coronavirus field is overcoming right now,” says Fred Hutchinson Cancer Research Center’s Jim Kublin, who conducts challenge studies of malaria. But, he adds, “given the pace at which we’re moving into these [COVID-19] vaccine studies, I’m skeptical” that a challenge trial would accelerate vaccine development.

Ensuring participants’ safety

A consistent and safe exposure is just the first requirement of a challenge study. After the challenge is delivered to volunteers, researchers must be prepared to care for the participants if they get sick, and make sure they don’t transmit the infection to others. 

Jim Kublin (left) watches Sanaria Inc.’s Jonathan Jackson and Meghan Marquette unpack malaria samples that will be injected into human test subjects during a clinical trial of a new malaria vaccine at the Fred Hutchinson Cancer Research Center in Seattle on April 21, 2016.

In some cases, that’s easy. In Kublin’s challenge studies of malaria, patients can simply go home after their exposure. Because Anopheles mosquitos that carry malaria parasites do not live in Washington State, there is no risk of them biting an infected person from the trial and spreading the disease, and the researchers closely monitor the participants’ blood for evidence of the parasite. Once it makes it into the person’s circulation, it must mature for another two weeks. By the time a volunteer might start to feel sick, they’ve been diagnosed and treated, says Kublin, who has used the model to test experimental vaccines and treatments

The situation is similarly benign in challenge studies of hookworm. While chronic infections with the parasite can cause gastrointestinal distress, anemia, and other complications in areas where it’s endemic, people who have volunteered to allow larvae of the pathogen to burrow through the skin on their arm don’t get sick, likely because of the relatively low dose and participants’ good nutrition, say Jeff Bethony and David Diemert, codirectors of George Washington University’s Vaccine Research Unit who are now conducting a hookworm vaccine challenge trial. (The Vaccine Research Unit is also a site of the Phase 3 trial of Moderna’s COVID-19 vaccine candidate.) The researchers confirm infection by looking for eggs in participants’ stool samples, and then treat them with a medication called albendazole that kills the worms. A few volunteers have agreed to remain infected and provide stool samples from which the team can harvest eggs for a new batch of larvae for the next challenge study. Diemert notes that it’s “almost impossible” for participants to spread the disease to others if they are using a toilet. 

In other cases, the pathogen presents enough of a risk to the exposed individual or to others that trial participants must stay for a period of time at an in-patient facility under quarantine. Such is the case when Johns Hopkins University infectious disease researcher David Sack has conducted vaccine challenge studies of cholera and other enteric bacteria. The team members use strains they know can be killed by an antibiotic, which volunteers are given, along with fluids, as soon as they meet the clinical criteria for having developed the illness. 

Similarly, participants who receive Memoli’s flu challenge stay in quarantine for one to two weeks, until they test negative for the flu on two consecutive days, although they don’t receive antiviral treatment so as to observe the infection take its course. Memoli notes that the need for an extended quarantine of perhaps three weeks for a SARS-CoV-2 challenge would be a hinderance, though not an insurmountable obstacle.

Considerations for a COVID-19 challenge trial

The more contentious part of the debate over whether a challenge study for COVID-19 should proceed is the risk to the individuals being deliberately exposed to the novel coronavirus, which is still very much an unknown—and one that can cause fatal disease. So far, SARS-CoV-2 has claimed more than 1 million lives around the globe, including more than 200,000 in the US. (In contrast, during the 2019–2020 flu season, the US Centers for Disease Control and Prevention estimates, there were between 24,000 and 62,000 flu-related deaths.) The Financial Times reports that the UK trial will treat participants who fall ill with remdesivir, a drug that appears to expedite recovery.

Researchers must be prepared to care for the participants if they get sick, and make sure they don’t transmit the infection to others.

“Personally, I do not think that it is ethical at this time to conduct COVID-19 challenge vaccine trials,” says Diemert. “There is no effective rescue treatment and we have no way of predicting who may have serious complications due to infection.” Sack agrees: “We say in general that young people don’t get as sick, but we also know that some young people die.” 

There’s also the risk that surviving participants could suffer long-term consequences, says Monica McArthur, a physician-scientist in the Center for Vaccine Development and Global Health at the University of Maryland School of Medicine. “While I’m a proponent of challenge models, I think there are some ethical requirements, and those include having a significant knowledge of the long-term effects of the infection.”

In making the evaluation of whether a COVID-19 challenge trial is warranted, the risks must be balanced with the need. As several researchers noted to The Scientist, there are many COVID-19 vaccine candidates in late-stage trials, with results expected over the next several months. If one or more of these studies yields results that support the vaccine’s approval, a challenge trial may not be necessary.

Challenge trials may be warranted when there isn’t enough disease transmission in the community to determine if a vaccine is protective, says Durbin, who adds that there is currently no shortage of natural COVID-19 spread in the population. “When you’ve got diseases that occur in sporadic outbreaks and you can’t necessarily predict when or where those outbreaks will occur to plan for a traditional Phase 3 efficacy study, challenge studies can be very useful,” says Durbin, who is developing a human challenge model for Zika for that reason. But back in 2016, when she and her colleagues first proposed a Zika human challenge model because they realized that the Zika outbreak was waning, the FDA told them that “it was unethical at the time to do a Zika challenge model,” Durbin says, in large part because there were still so many Zika cases. “But now we’re talking about a COVID-19 challenge trial. That kind of blows my mind.” 

Nevertheless, Durbin and others say it’s too soon to take the possibility of future challenge studies off the table. With so many moving parts, you never know what the future holds. “I count myself among the many who are still waiting for more information on the COVID virus that would be used as the challenge strain, the state of vaccines in the near future (i.e., How urgent is this by next year?), and any advances in COVID therapeutics,” Bethony tells The Scientist in an email.

Memoli is also a “maybe” on whether the SARS-CoV-2 challenge model he’s currently developing should ever be used to give people COVID-19. “As we look to see what happens here over the next few months, in terms of the vaccine, in terms of the disease itself and its spread in the world, we’ll have to make some decisions,” he says. For now, he reiterates, “we should definitely be preparing.”

Interested in reading more?

The Scientist ARCHIVES

Become a Member of

Receive full access to more than 35 years of archives, as well as TS Digest, digital editions of The Scientist, feature stories, and much more!
Already a member?