March 26, 2024, was a weird day for me because it was the only one in my life where I was actively trying to get bitten by mosquitos.
I had volunteered to be exposed to malaria as part of a study at the University of Maryland, Baltimore (UMB) evaluating MAM-01, an injectable drug meant to prevent infection. And by “exposed to malaria” I mean “bitten by mosquitos infected with malaria.”
The procedure takes place in a biosafety level 2 laboratory operated by the university’s Center for Vaccine Development, and while that sounds otherworldly and high-tech, the reality was more mundane. You sit in a foldable chair at a foldable table with a bunch of other volunteers, while nurses place a small cardboard cylinder containing several infected mosquitos against your arm, separated by a mesh that keeps the mosquitos from escaping but enables them to bite you. Which they do.
After that’s done, the researchers take away the mosquitos and dissect them, measuring how much blood they’ve taken from you. If the blood volume hits a certain threshold, you’re done. If you don’t, it’s time for more mosquitos to meet your arm.
When I was going into the lab, another volunteer who had just finished up gave me a tip: Use your socks. Take them off and rub the dirty sock soaked in your grimy foot sweat against your arm. He was right; Anopheles gambiae mosquitoes are apparently foot fetishists.
This mosquito feast in the middle of Baltimore was part of what’s called a “human challenge trial,” a class of medical experiments in which therapies and preventatives are tested by actually exposing human volunteers to the pathogen they’re meant to defend against. To know if vaccines and treatments work, you need to look at people who’ve been exposed to the pathogen in question. Most studies do that by studying places where the disease is spreading rapidly. But the easiest way to know people have been exposed is to expose them.
We’ve been doing human challenge trials to test drugs and vaccines for centuries. The UMB program I volunteered for dates back to the 1970s. But these trials are poorly understood by the general public, and I had heard from sources for years that given their clear benefits, they’re underutilized in efforts to develop new treatments and protective drugs.
So I decided to enroll in a challenge trial to find out what, exactly, it entails. The experience was long, lasting over a year with hours upon hours of appointments. But it was incredibly safe, and though (spoiler alert) I did wind up getting malaria, the symptoms were mild and over quickly.
My takeaway: This really isn’t that bad. And it does seem like a technique that could help save a lot of lives — if the medical research community and regulators are willing to use it, and ordinary people are willing to volunteer for it.
My challenge, explained
MAM-01, the subject of this malaria challenge trial, is a “monoclonal antibody.” While ordinary vaccines induce your immune system into producing antibodies that target particular illnesses, MAM-01 skips the immune system step: It’s just a lot of antibodies made in a lab and inserted into your bloodstream. Instead of teaching your immune system to fish, you’re giving your immune system some fish.
Malaria experts are very excited about monoclonals, especially as a prophylactic in parts of the world (like the Sahel region of Africa south of the Sahara) where malaria chiefly spreads in the rainy seasons that extend from roughly June to October. Take the monoclonal just before the season starts, and the protection could last until the coast is clear again. The hope is the shot would be more effective and offer more protection than the recently approved vaccines for the disease.
“In a place like Burkina Faso in the Sahel, where there’s near zero transmission outside a four- to five-month window, you could, with a single encounter, stop a full year of transmission,” Philip Welkhoff, who directs malaria efforts at the Bill & Melinda Gates Foundation, told me. But for that to work, the monoclonal has to be very highly effective, and to be that effective, the antibodies need to be specific enough to the strain of malaria that’s circulating. And unfortunately, Welkhoff added, “It looks as though the first two [candidates] are not going to be quite there yet.”
One of those first two candidates is MAM-01, and sure enough, about two weeks after my mosquito bites, I tested positive for malaria. It wasn’t a bad case: I was put immediately on Malarone, a combo pill of two anti-malarial drugs, and I only had one day of noticeably unpleasant symptoms: nausea, chills and fever, muscle aches. I binged season two of Will Trent and napped a lot, and then I was fine. But I couldn’t help but feel mildly bad: MAM-01 may yet be very effective, but it’s not 100 percent effective. My dumb immune system made sure of that. Sorry, Philip.
That was the most exciting week of the trial. The rest of it was incredibly routine. I got two injections of MAM-01, once in October 2023 and again in May 2024. In addition to wanting to see if this would prevent malaria infection, the researchers were interested in how long the antibodies stayed in my system. That meant returning to the lab (at first daily, then weekly, then fortnightly, then monthly) to have my blood drawn from an arm vein and/or squeezed from a prick in one of my fingers, to see if the antibodies were still circulating, both in large blood vessels and tiny capillaries. This also entailed driving an hour and 15 minutes each way between my house in Washington, DC, and the lab in Baltimore a total of 34 times, usually for a procedure that took maybe five minutes total.
Eighty-five hours of driving was, to be honest, pretty annoying — more annoying than actually getting malaria — but it had a silver lining: I got through most of the audiobooks of War and Peace and The Power Broker on those trips on I-95. Malaria’s major side effect was to make me more cultured, and also possibly the only person on Earth to associate a tropical disease with Moscow in winter.
This routine, mundane process means that malaria trials get a decent number of recruits compared to other challenge trials (especially from people who, unlike me, live in Baltimore). “We’re actually pretty lucky,” Kirsten Lyke, professor of medicine at the University of Maryland and the principal investigator in my trial, told me. “We’ve been pretty successful recruiting. Not to say that it’s easy, but we’ve developed social media and a lot of different techniques to reach out to people.” I should note that my study is still recruiting people, as are several others at the Center for Vaccine Development; I’m particularly excited for one testing a new mRNA vaccine for malaria. If you’re in DC or Baltimore, it’s worth considering.
That said, the kind of relaxed outpatient experience common for malaria studies isn’t necessarily the same for other illnesses. Keller Scholl, an AI policy researcher I know in DC, did a challenge trial in which he was exposed to and contracted Zika; his trial was inpatient, requiring him to stay in a special hospital dormitory for nine days, a burden against which my trial looks like child’s play.
Jake Eberts went viral after live-tweeting his experience as a volunteer in a challenge trial for shigella, one of the bacteria that causes dysentery. He got a fever of 103 degrees, plus diarrhea and blood in his stool. “I truly felt like I could not move,” Eberts later told a reporter. He also had to collect his watery, bloody waste for researchers to analyze. He earned slightly more money than I did for his trial, but personally I think he should have gotten maybe 10 times as much.
Unsurprisingly, Lyke told me that recruiting for shigella and cholera trials is a lot harder than for malaria.
A brief history of challenge trials
To join my challenge trial, I had to read through a 21-page document explaining in meticulous detail exactly what the trial entailed. I had to go through a screening appointment, pass a quiz about the contents of that 21-page document, and sign many forms. Every one of the dozens of visits I’ve made as part of the study has begun with a nurse asking me if I consent to continuing with the study.
For my troubles, I’ve been paid $4,740 so far. I’m donating all of that to the Against Malaria Foundation and the Malaria Consortium, since I didn’t want any readers to think my impression of the experience was inappropriately biased by being paid. After deducting the mileage costs and dividing by the 85 hours driving and maybe 30 hours in labs and tests, it works out to about $12.50 an hour, below Maryland’s minimum wage.
I say all of this to emphasize that UMD conducts these studies with the highest professional and ethical standards, and I emphasize that because I know “deliberately infecting people with diseases” is a genre with a long, and sometimes less than noble, history.
We’ve had challenge trials for as long as we’ve had vaccines. Edward Jenner, who pioneered the vaccination of patients against smallpox by injecting them with fluid from cowpox sores, repeatedly exposed his gardener’s son to smallpox after the inoculation, to prove it worked.
The modern history of challenge trials begins with the 19th-century Cuban doctor Cárlos Finlay, who for decades attempted to demonstrate that yellow fever was spread by mosquitos. The disease was a major scourge at the time; in a famous 1793 outbreak, about 1 in 10 Philadelphians died of the disease, and the federal government (then headquartered in the city) was forced to shut down.
Finlay conducted studies deliberately exposing volunteers to mosquitos, killing some subjects and infecting many, but the results were still inconclusive. His work inspired Jesse Lazear, a young Army doctor posted in Cuba after America conquered it from Spain, who sought in 1900 to prove the Finlay thesis. He did so by deliberately coaxing a mosquito to bite him. Lazear got yellow fever and died, evidence American officials took somewhat more seriously than Finlay’s work.
Lazear’s death helped inspire his boss, Walter Reed (after whom the famous military hospital would later be named), to adopt one of the first ethical codes ever for clinical trials. Participants in human trials overseen by Reed in 1900 signed a letter, in English or Spanish, confirming, “The undersigned understands perfectly well that in case of the development of yellow fever in him, that he endangers his life to a certain extent but it being entirely impossible for him to avoid the infection during his stay in this island, he prefers to take the chance of contracting it intentionally in the belief that he will receive … the greatest care and the most skillful medical service.” Every volunteer also received $100 (about $3,800 today), plus another $100 if he fell ill or died.
Reed’s trials would ultimately help prove Finlay right that mosquitos transmit disease, specifically yellow fever. That led to a mosquito eradication campaign that proved wildly successful: In the 1890s, an average of 462 people in Havana died of yellow fever every year, but not a single death was recorded after September 1901, after the eradication campaign was underway. Reed’s studies identified the main culprit behind the plague and enabled public health officials to eradicate it.
By the 1930s and ’40s, there were challenge units studying the flu in the Soviet Union, the US, Australia, and the UK. Perhaps the most storied effort was the Common Cold Unit located in a hospital in the small English city of Salisbury, which used human volunteers to study the virus (specifically a coronavirus) that causes the common cold. While the unit obviously did not produce a cure, it did give us a greater understanding of coronaviruses in general (which came in handy when a different coronavirus ignited a pandemic in 2020), as well as some delightfully British promotional videos:
But these challenge studies were also occurring at a time and in places where what we’d now consider absolute bare minimum ethical procedures were not followed. While Reed did pay his volunteers and get consent forms, he was still experimenting on incredibly poor Cubans, as a member of an occupying army. Some of the earliest challenge studies on malaria were conducted on prisoners at Stateville Penitentiary in Illinois. Anthony Allison, the researcher who determined that the same genes that cause sickle cell anemia also protect against malaria, demonstrated this through a challenge trial exposing Kenyans to malaria — conducted by Allison, a white South African, while Kenya was an English colony. These are conditions of domination where freely given consent is impossible.
So it feels important to clarify that today’s challenge trials, like the one I joined, are ethically scrupulous to a fault. The Center conducting the trial emphasized informed consent again and again until it became annoying; all participants knew the exact risks, and the exact procedures from the get-go, but they were still reminded constantly.
Moreover, those risks were very low. The reason why challenge studies have been going on longest for malaria out of any infectious disease is, in part, because today it is trivially easy to cure malaria in rich countries with strong medical infrastructure, especially when the strain of malaria is one that researchers have hand-selected, that we know is not drug-resistant, and in which the long-term effects of a single infection are minimal.
Studies evaluating the safety record of challenge trials confirm this. A recent comprehensive review of 308 studies from 1980 to 2021, which enrolled over 15,000 patients altogether, found that challenge studies during that period did not result in a single death. They led to 24 “serious adverse events,” or SAEs, a frequency of less than 0.2 percent. Even better, the authors clarify, “it is unlikely that any SAEs captured in this review were life-threatening,” not least because they occurred in hospitals where any problems could be caught very early.
Surveys of modern challenge trial participants have found that medical risks are among the least of their concerns. “Most participants said the burdens were greater than the risks, most commonly mentioning the time commitment and frequent blood draws,” the authors of one survey found. This fit my experience: It never felt dangerous, but it occasionally felt annoying.
Another survey of volunteers in the Netherlands found that 84 percent of participants reported being proud of their choice, 89 percent would encourage others to volunteer, and 85 percent would do a similar trial again. I agree with all three (though maybe not a trial that requires 85 hours of driving) — and generally, if such an overwhelming share of participants have the reaction “I want to do this again,” that’s a sign that you’re onto something.
Why challenge trials could be game changers for TB and hepatitis
I, like many people, first heard of challenge trials in the depths of the Covid-19 pandemic, when they appeared to be a way to test possible vaccines more rapidly, potentially saving huge numbers of lives by shortening the emergency. I wrote about them sympathetically and repeatedly — but ultimately, none of the three vaccines available in the US were tested through challenge trials. They were approved based on traditional “field trials,” based on how effectively they reduced infections among people just going about their business in a world full of Covid.
To some extent, this was only possible because the governments of the US, UK, and Brazil, where most of the trials occurred, failed so miserably at controlling the virus. “If the UK, US, and Brazil had been run by leaders interested in preventing viral transmission in the way that many other countries did, it’s really difficult to see how we would have developed vaccines,” Josh Osowicki, physician and senior researcher at the Murdoch Children’s Research Institute in Melbourne, told me.
The first randomized controlled trial evaluating effectiveness for the Pfizer vaccine, for instance, had 43,448 participants. Eight participants getting the vaccine wound up becoming infected with Covid; 162 of the placebo group were. The study needed to be that huge because a smaller study would have risked not having anyone in either group infected, not due to any feature of the vaccine but simply because at the time the virus was not spreading enough to be noticeable in a small-scale study. So even with a virus that was spreading quite rapidly, doing a field trial required massive investment.
That was fair enough in the Covid-19 case; the US government was willing to spend just about any amount of money to develop a vaccine as fast as possible. But what about neglected diseases, like malaria? “It’s one of the problems with malaria,” Lyke told me. “It’s not a problem that affects the US, so there’s never any sense of urgency.”
That means researchers operate under tougher budget constraints, while field trials can still prove to be hugely expensive. Very little information is publicly available on trial costs for vaccines, but a paper examining vaccines for rotavirus — which causes diarrhea in children and kills about 150,000 kids a year — reported a rough estimate of $2,000 to $3,000 per subject for phase 3 trials alone. (Phase 1 and 2 trials are tiny and cost only a few hundred dollars per subject).
That implies that, say, the recent 5,500-person trial of the R21 malaria vaccine could have cost some $11 million. For diseases with lower “attack rates” (that is, a smaller share of the population getting infected during the study period), the cost is even higher. At the extreme end, a phase 3 trial for a tuberculosis vaccine is budgeted to cost $550 million.
To be blunt: If we need to spend $550 million to test every TB vaccine candidate, we will never get a reliable TB vaccine. There just isn’t enough money, especially not for a disease that overwhelmingly kills poor people in poor countries.
One of the most important things challenge trials do is enable “down selection”: They let researchers evaluate candidate vaccines and drugs cheaply and quickly, with a few dozen volunteers, rather than over a long time with tens of thousands of participants. Because you are directly exposing people to the pathogen, you know the attack rate is very, very high. “Let’s say you test a vaccine here in the US [in a challenge trial]. It has zero efficacy,” Lyke says. “Then there’s no way, if you took it to the field, that’s going to work.” The challenge trial lets us eliminate that option and move on to other ones.
Osowicki and colleagues recently reviewed how well challenge trials predicted real-world efficacy. The correlation was pretty good. Sometimes, the challenge nailed it: Typhoid conjugate vaccine was 87 percent effective in a challenge study, and 95 percent effective in the field. But sometimes candidates that worked in challenge did worse in the field, as was the case with the RTS,S malaria vaccine or the intranasal flu vaccine. The reverse, though, didn’t happen: We don’t have any major cases of a vaccine that failed in a challenge trial but worked in real life (though, in fairness, few vaccines that fail at the challenge stage progress to field trials). Challenges thus serve as a weeding mechanism: You can cheaply rule out vaccines and treatments that seem unlikely to work in the field.
It’s also why researchers on diseases like TB are working hard on challenge models that would enable them to test candidate vaccines faster and more cheaply. This is tough for TB, which can take years after infection to show symptoms, requires a months-long course of antibiotics to treat, and even then can persist in patients. That means that field studies take agonizingly long and are extremely expensive, but also that we cannot ethically expose people to wild-type TB for a challenge study. It’s a horrible situation to be in with a disease that kills an estimated 1.3 million people a year, more than twice as many as malaria and more than any other infectious disease today.
To deal with that problem, Sarah Fortune, a professor of immunology at Harvard Medical School, has been developing a variant of TB to be used in challenges. “We’ve invented a TB strain that’s addicted to doxycycline and trimethylamine,” two different antibiotics, she told me. “In the absence of them, it blows up.”
This is exactly what you’d need for a challenge study: a version of TB that has a kind of “kill switch,” in that it will die off completely if participants stop taking those two drugs. That means a study participant could be infected, have TB for the duration of a challenge trial, and have it completely eradicated afterward if they simply stop taking the antibiotics.
“It’s taken us about seven years to figure out how to make that system meet our own safety standards,” Fortune says, but she thinks it’s ready. Their genetically modified TB variant is going through FDA review right now. And she and fellow professor Eric Rubin, a collaborator on the research, are putting their money where their mouths are in terms of safety: “Me and Eric decided we’d be volunteers one and two.”
Hepatitis C, which kills about 240,000 a year, is another illness where vaccine progress seems prohibitively expensive and difficult without a challenge model. The people most at risk for the illness are those using injection drugs like heroin (people who inject drugs or “PWIDs” in public health lingo), which is an inherently difficult population to do a field trial with.
“The last time they did a traditional trial for a vaccine candidate, it took over six years [and] was extremely expensive because it had to recruit from people who have high risk for hepatitis C, which tends to be both a very vulnerable population and a population that is difficult to retain in clinical trials,” Circe McDonald, who leads hepatitis C work at the pro-challenge trial nonprofit 1Day Sooner, told me.
McDonald notes that the one phase 1/2 trial of a hepatitis C vaccine candidate cost about $10 million to $11 million, with 548 participants; a phase 3 trial would require tens of thousands of participants and cost at least an order of magnitude more. By contrast, once we have a version of hepatitis C that’s safe to use in a challenge (which at least two teams are working on now), a challenge trial might cost $1 million to $2 million.
Osowicki and colleagues have put together a model for another neglected illness, although one you’ve probably heard of: strep, specifically “group A” streptococcus. While most associated with childhood sore throats and burnt plush rabbits in rich countries, group A strep bacteria impose a massive global burden. In extreme cases, the disease is much worse than strep throat or even scarlet fever. In “invasive group A streptococcal disease” (IGAS for short), “the presentation is severe, invasive, sudden, explosive infection,” Osowicki explains.
We don’t have as reliable data for strep as for TB, malaria, or hepatitis, but by “really really conservative estimates,” Osowicki says, 150-200,000 people a year die from iGAS. Another 300-400,000 people a year die from rheumatic heart disease, a chronic condition caused by damage from strep infection.
Despite this, we’ve had little progress toward a strep vaccine since the 1970s, when the FDA effectively banned further studies on the fear that vaccines themselves could lead to rheumatic heart disease or acute episodes (it lifted the ban in 2006). Part of the problem is it’s just incredibly hard to study in the field. IGAS only results in very rare cases; rheumatic heart disease develops after a long lag, making it difficult to detect in a time-limited study. It would be much easier to expose people to strep, and Osowicki has worked with experts around the world to develop a strep throat human challenge model.
Talking to researchers about projects like these, it almost started to feel like challenge volunteers, like me, were a kind of cheat code. Biomedical research is unbelievably difficult and expensive, especially if you’re working on diseases that primarily kill a lot of people in poor countries where they can’t pay much for vaccines. Anything that makes that easier seems like a worthy task. In my case, that meant some long car rides and a day of bad malaria. I think that was, easily, worth it. Maybe it’ll be worth it for you too.
