NIH researchers infect volunteers with the flu virus in an ongoing effort to improve vaccines.


Studies that intentionally infect people with disease-causing bugs are on the rise

Vibrio cholerae, a comma-shaped bacterium that contaminates water and food, can kill fast. Acids in the stomach can wipe out billions of the bacteria, but if a person swallows as few as 1000 with food, some may survive the swim to the small intestine. There, the invaders will release enzymes to penetrate a thick mucous layer that lines the epithelium. Once through, the bacteria will attach to epithelial cells and begin dividing, establishing microcolonies that secrete toxins. Then the death clock begins ticking.

Irritated by the main cholera toxin, the intestine will gush fluid, and a person will develop cramping, vomiting, and, most notoriously, diarrhea that rapidly becomes a staggering volume of “rice-water” stool: a watery liquid filled with mucous flakes and epithelial cells. In severe cases, people lose a liter of rice-water stool per hour and, without rehydration to replace lost body water and electrolytes, can die within half a day.

In 1976, at the behest of a U.S. government panel, Myron “Mike” Levine of the University of Maryland School of Medicine in Baltimore began intentionally giving humans V. cholerae. He is still doing so today.

Myron “Mike” Levine

Myron “Mike” Levine

Paul Fetters

Forty years ago Levine was one of a tiny cadre of researchers doing so-called human challenge studies—intentionally infecting people with V. cholerae and other pathogens to test drugs and vaccines. But in the past few decades, this practice, which has a long and checkered past, “has become much more mainstream,” Levine says. Stricter safety procedures and new ways to weaken pathogens to reduce their risks are leading investigators in industry, universities, and government to take a new look at human challenge trials, which offer a powerful tool for studying diseases and potential therapies. There’s even a commercial company, hVIVO in London, that specializes in human challenges. Today, people are being deliberately infected with malaria, influenza, shigella, dengue, norovirus, tuberculosis, rhinovirus, Escherichia coli, typhoid, giardia, and campylobacter.

The risks are obvious: Otherwise healthy people can suffer harm and, if the disease is contagious, potentially sicken others. But if done right, the benefits are compelling, a growing number of researchers say. The standard pharmaceutical development path for products that target pathogens moves slowly from studying safety, dosing, and biological responses in hundreds of people to an expensive efficacy trial with thousands of participants at high risk of becoming naturally infected. Human challenge studies, which only involve a few dozen volunteers, speed the process of deciding whether to scrap or pursue a promising lead, saving time and money. And tests that intentionally infect people can quickly and efficiently flag potential side effects, advocates say. “You certainly can’t do a $100 million study for every candidate vaccine that appears safe and immunogenic,” says Mark Mulligan, a molecular virologist who heads the vaccine center at Emory University in Atlanta and does human challenges with norovirus and tuberculosis.

Insights from human challenges reach far beyond drugs and vaccine development. Christine Moe, another Emory University researcher, has shown that norovirus more readily transmits via vomit than diarrhea, and that this “Ferrari of viruses,” famous for the speed at which it races through vacationers on cruise ships, is impervious to alcohol-based hand sanitizers and to power-washing the oysters that carry it. She notes that “sometimes human challenge studies are the only way to answer critical questions.”

Checkered past

Human challenges date back to the 18th century and the first vaccine, when English physician Edward Jenner attempted to persuade the world that infecting a person with harmless cowpox could prevent disease from its dreaded cousin, smallpox. Jenner scraped “matter” taken from a cowpox sore on a dairymaid’s hand into the skin of 8-year-old James Phipps, the son of his occasional gardener, and then repeatedly tried to infect him with smallpox. “Poor Phipps,” as Jenner later referred to the boy, never came down with smallpox. Jenner reported that some 6000 other people were vaccinated and then the “far greater part of them” were challenged with smallpox. Two centuries later, the vaccine Jenner pioneered eradicated the virus from the human population.

Vaccine pioneer Edward Jenner put cowpox fluid into James Phipps’s arm in 1796, then challenged him with smallpox.

Vaccine pioneer Edward Jenner put cowpox fluid into James Phipps’s arm in 1796, then challenged him with smallpox.

University of Michigan Health System, Gift of Pfizer Inc. UMHS.23

Intentionally infecting a human—let alone a child—with a disfiguring and even deadly disease would never pass ethical muster today. But as recently as the early 20th century, intentional infection was seen as cutting-edge: Austrian psychiatrist Julius Wagner-Jauregg won the 1927 Nobel Prize in Physiology or Medicine for injecting blood from people with malaria into patients with neurosyphilis, which putatively cured them of insanity and paralysis. As journalist Lawrence Altman documented in his book Who Goes First?, many investigators have challenged themselves with pathogens to prove the worth of their own experimental medicines or theories. Some died.

In the 1940s, the University of Chicago in Illinois and the U.S. Army collaborated on challenge experiments that tested malaria drugs in 400 Illinois prisoners. Nazi doctors, who horrified the world with their own medical experiments, including malaria tests that killed several hundred people, cited the U.S. studies in their defense when they were put on trial in Nuremberg, Germany, in 1947. This led to the Nuremberg Code, which spells out what are now the standard research principles of informed consent, voluntary participation, and the freedom to quit a study.

Yet U.S. experiments on prisoners continued, leading to investigative journalist Jessica Mitford’s 1973 exposé in The Atlantic Monthly, “Experiments Behind Bars.” Levine, who was just starting his career, was then challenging humans with shigella and typhoid at the Maryland House of Correction in Jessup—experiments he insists were conducted ethically. “The studies done at Jessup were 2 decades ahead of their time in terms of the methods of informed consent,” he says. But in 1976, the U.S. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research—the country’s first bioethics policy effort—issued a report that effectively brought human challenge experiments in prisons to a halt.

The first question I ask is, ‘Would I want my kids … to participate?'

Myron “Mike” Levine, University of Maryland School of Medicine

Outside of prisons, however, research continued. In 1974 the U.S. National Institute of Allergy and Infectious Diseases (NIAID) in Bethesda, Maryland, awarded the University of Maryland a half-million dollars to create a new vaccine testing center, headed by Levine, that would recruit volunteers from colleges and church groups. The center began with influenza challenges, which were conducted in refurbished rooms at the University of Maryland Hospital that had bunk beds for 22 people and an isolated air system. The researchers had little trouble recruiting volunteers, who received the same fee as jurors ($20 a day), which the researchers deemed fair but not coercive. Volunteers had to take a written test to prove that they understood the risks.

Two years later, at the request of NIAID’s Cholera Panel, Levine’s group added challenges with V. cholerae to test cholera vaccines. “One of the really big questions was, ‘Would anyone be willing to participate?’” Levine recalls. “It’s one thing doing flu, which most people experience every other year or so, and it’s another thing to take this exotic tropical infection and set that up.”

Maryland required that the hospital fly a yellow flag to warn of a cholera quarantine area. Again, finding volunteers presented few obstacles. “These were the same young people who would go down the hairiest parts of rivers on rafts,” Levine says.

In a 1970s University of Maryland cholera study, this man needed 26 liters of intravenous electrolytes to replace lost fluids.

In a 1970s University of Maryland cholera study, this man needed 26 liters of intravenous electrolytes to replace lost fluids.

Courtesy of Myron M. Levine

The cholera studies led to the scuttling of a leading vaccine candidate, a finer understanding of effective immune responses, and, ultimately, compelling evidence that a different cholera vaccine worked. In June, the U.S. Food and Drug Administration (FDA) will consider licensing a cholera vaccine for travelers based largely on Levine’s work. This is the most influential role the human challenge model has ever played in the FDA approval process.

Modern volunteers

Over the next decade, Levine’s group expanded to challenges with Salmonella typhi, E. coli, and rotavirus. The only other substantial human challenge operation was the Common Cold Unit established by the Medical Research Council in Salisbury, U.K. Then in 1985, a team led by Ripley Ballou began human challenges with malaria at the Walter Reed Army Institute of Research (WRAIR) in Silver Spring, Maryland. That program pioneered advances that have lowered the risks of human malaria challenges and increased the benefits, opening the way to the trials flourishing today in several places.

Ballou, who now heads vaccine R&D in the United States at GlaxoSmithKline (GSK), and his team bred mosquitoes in an insectary, and then fed them on human blood infected with the malaria parasite Plasmodium falciparum. He and five other Army colleagues each took a candidate malaria vaccine and then let five infected mosquitoes—which their group had determined was the number needed to reliably transmit the parasite—lunch on their arms. “I got a full-blown case of malaria and was never so sick in my life,” Ballou says, even though he was promptly treated. “It made a huge impression on me and I was committed to finding a way to stop this disease.”

In WRAIR’s first vaccine trial, all the participants were “my friends in the laboratory or from down the hall,” says Ballou, and they went home after being infected. Now, WRAIR recruits civilians—special policies govern participation of people serving in the military—who for up to 10 days stay in a hotel together, where they receive regular checkups. Technology has also made the experiment safer than when Ballou infected himself: The polymerase chain reaction test can detect minute amounts of parasite DNA and identify an infection 2 days earlier than traditional microscopy, and if volunteers receive immediate treatment, they rarely suffer any symptoms.

Ripley Ballou led malaria challenge trials to test vaccines at the Walter Reed Army Institute of Research starting in 1985. Volunteers—including Ballou—willingly exposed their arms to malarial mosquitoes.

Ripley Ballou led malaria challenge trials to test vaccines at the Walter Reed Army Institute of Research starting in 1985. Volunteers—including Ballou—willingly exposed their arms to malarial mosquitoes.

Walter Reed Army Institute of Research

The Army’s malaria challenge studies have yielded impressive dividends. “We trashed a whole bunch of vaccines,” Ballou says. They also contributed to the development of GSK’s RTS,S, the only malaria vaccine that has so far demonstrated efficacy, albeit modest, in a large-scale field trial.

The resurgence

Trials launched more recently face greater regulatory scrutiny than Levine’s and Ballou’s did. Since the mid-1990s, FDA has deemed that organisms used in challenge studies are experimental medicines, and the agency has required researchers to submit Investigational New Drug applications before conducting trials. Institutional reviews have intensified, too.

Human challenge studies with influenza provide a glimpse of the new landscape. In the 1980s and 1990s, for example, Frederick Hayden of the University of Virginia School of Medicine in Charlottesville conducted challenge studies with influenza that helped speed the development of Tamiflu and Relenza, drugs that have become the mainstays of treatment. But the work ground to a halt in 2000 after a volunteer in one of Hayden’s studies experienced what FDA calls an “adverse event.” A 21-year-old man testing a flu drug developed heart abnormalities after being challenged with the virus. “I still don’t know what caused that episode,” Hayden says. “There were a lot of sleepless nights.” No long-term harm occurred, but the incident has led to a thorough review of cardiac events in other influenza challenge studies.

So there was considerable concern when NIAID’s Matthew Memoli proposed new human challenge studies with influenza in 2011, which ultimately aimed to test novel treatments and vaccines. Some of his colleagues were so wary that the ethics department at the National Institutes of Health (NIH), NIAID’s parent, was asked to conduct a formal review of the protocol. “We went through a lot of steps,” Memoli says. The ethicists were particularly concerned about the proposed “high levels of payment”—up to $4000—but deemed this was not an “undue influence” because no one had an obligation to accept the offer.

Volunteers were “meticulously” screened, Memoli says: They had to be under 45 and undergo a battery of tests, including electrocardiograms. Memoli and colleagues also worked with FDA to grow a strain of the virus that met the agency’s good manufacturing practices, and they precisely calculated the minimum dose needed to cause disease in most volunteers.

In ongoing studies, the researchers spray influenza virus into the noses of volunteers via a mist, created by an atomizer that only produces particles larger than 10 microns. These relatively fat particles can cause infections in the upper respiratory pathway but do not reach the lungs, where influenza virus can cause life-threatening pneumonia. To avoid infecting others, participants remain in hospital isolation rooms for 9 days. “I’ve challenged nearly 200 people and have had no serious complications,” Memoli says. “The worst thing that happened is a guy slipped in a shower.”

Memoli stresses that intentionally infecting people is an odd pursuit for a doctor. “We’re purposefully making people sick,” Memoli says. “It’s a different idea than what you originally go to medical school for. But over the course of the next few years I think we’re going to get information that’s going to be tremendously helpful.” In a study published online in mBio on 19 April, Memoli and his co-workers reported that their challenge studies indicated that a widely ignored antibody response to influenza vaccines might be a better predictor of effectiveness than the antibody routinely analyzed today.

Five years ago, the small community that studies dengue began discussing challenge trials, which made some people nervous, says Anna Durbin of Johns Hopkins Bloomberg School of Public Health in Baltimore. The mosquito-borne infection can trigger a high fever, serious joint pain, and intense rashes; in rare cases, it can lead to hemorrhaging and death. No drugs specifically target dengue virus. “We heard, ‘You can’t treat dengue so you can’t do a human challenge model,’” Durbin says. Human challenges with dengue date back a century, but the last intentional infections of volunteers took place at WRAIR in 2001, and a few of the participants developed dengue fever.

In 2011, WRAIR and NIH sponsored a workshop to discuss “reintroducing” the human challenge model for dengue. Several attendees, including Durbin, argued that the trials could be done safely and would speed development of a badly needed vaccine for this disease.

With the blessing of FDA, Durbin in June 2013 began challenging volunteers who had received a dengue vaccine made by NIAID. Instead of using wild-type dengue virus, Durbin and her Hopkins team infected people with a naturally weak isolate of the virus that had been further attenuated in the lab. “I don’t think you need to make people sick” to see whether they develop an infection, Durbin says.

As she and her group reported online on 16 March in Science Translational Medicine, none of the 21 people who received the vaccine became infected after the challenge, but all 20 controls had the virus in their blood, and 16 developed a rash. Based in part on these results, the Butantan Institute in São Paulo, Brazil, this year launched an efficacy trial of the vaccine that plans to enroll 17,000 people. 

Regulations of human challenge studies differ from place to place. In the United Kingdom, for example, challenge agents are not considered drugs, and experiments with them thus don’t require regulatory approval. A group at the University of Oxford led by pediatrician Andrew Pollard has conducted a challenge study of experimental vaccines against typhoid and paratyphoid. Although both diseases are contagious, the researchers allow volunteers to go home instead of staying in isolation. “They’re potentially shedding organisms that are going into a flush toilet,” says Levine, who collaborates with the Oxford group. “That’s something that’s not amenable to being carried out in the USA.”

The United Kingdom is “much more permissive,” agrees Pollard, but he says the trials go through extensive ethical reviews, and the risk of transmission is “near zero” if people have good hygiene.

The human challenge model has its limits, Levine stresses, noting that his group declined to participate in an experiment done elsewhere that put Neisseria gonorrhoeae in a penile catheter to study gonorrhea transmission. “The first question I ask is, ‘Would I want my kids, siblings, or spouse to participate?’” Levine says. “If the answer is ‘no,’ we don’t do it.” And he worries that even though researchers today address risks more carefully than ever before, someone could push too far and undo the gains the field has made. “This should not be a Wild West show,” he says. “Some newcomers may not be totally aware of the burden the pioneers went through. It has taken a lot of time to get buy-in from everyone imaginable.”