First, do harm

An unusual statue stands in the courtyard of the San Fernando
Medical School in Lima, Peru. It commemorates a student named
Daniel Alcides Carrión, who died in 1885 after
he deliberately exposed himself to bartonellosis, a bacterial
disease spread by the bites of infected sand flies. Carrión was
trying to prove that the disease was also responsible for
distinctive lesions that erupted from the skin of people in the
region. He did so — but only after dying when he injected himself
with material taken from one such lesion on a patient.

Carrión’s risky self-experiment was very much of its time. But a
new generation of scientists have revived and updated the
principle, and are once again purposely exposing, if not
themselves, then other human volunteers to harmful diseases. The
good people of Southampton, UK, are
being asked to inhale a bad bug
that is driving a lethal,
international revival of infant whooping cough. Fifteen
strong-willed men and women in 2010 endured a week of stomach cramps and diarrhea at the
University of Vermont after they knowingly swallowed a
food-poisoning agent found in raw milk. And in the last few decades
hundreds of men in the southern United States have come forward to be inoculated with gonorrhea,
well, you know where.

Supporters of these types of experiments — called human
challenge trials or controlled human infection models — argue that
they are the quickest and cheapest way to develop new vaccines,
test medicines and study the basic progression of some of
humanity’s most enduring infectious foes, as well as some new ones.
The strategy is popular with organizations such as the British
biomedical charity the Wellcome Trust and the Bill & Melinda
Gates Foundation, as they spend their billions trying to tackle
neglected ailments of the developing world. Malaria, dengue fever,
cholera, influenza, typhoid and tuberculosis — which among them
kill some three million people each year — are all being
investigated by scientists who make healthy people purposely

A colored scanning electron micrograph shows Vibrio cholerae, a rod-shaped bacterium with a single long hair-like flagellum that it uses to propel itself through the water. Challenge trials with this bacterium, which cases cholera, led to the development of Vaxchora, the first FDA-approved vaccine for the disease.

The Vibrio cholerae bacterium (shown here in a colored
scanning electron micrograph) is the cause of cholera, an
infectious and sometimes fatal disease. Human challenge trials with
the microbe led to the development of Vaxchora, the first
FDA-approved cholera vaccine.


“It’s not without risk, but we work to make it as safe as we
possibly can,” says Robert Read, an infectious disease expert at
the University of Southampton who is leading the whooping cough
trial. The conventional way to develop vaccines is expensive and
inefficient, he says. “It wastes the lives of thousands of animals
that are not physiologically relevant — and then, when we do
pre-clinical work in people, we find it doesn’t work.”

Still, ethicists say it’s time for a more detailed assessment of
the rights and wrongs of human challenge trials. That’s because,
although the vast majority have taken place in rich nations,
scientists increasingly seek volunteers in places where the
diseases being studied are endemic. The shift is important if
treatments and vaccines are to work, proponents of these studies
say, because people who live under constant threat of a disease
often develop a different immunological response.

Known risks, uncertain benefits

Deliberate infections can help researchers develop new vaccines,
as well as test the effectiveness of some that have been in use for
decades and try to improve them. Take Read’s Southampton project.
Like most countries (both developed and developing), the UK
routinely vaccinates against whooping cough in childhood, but
protection by the vaccine is patchy, and serious epidemics crop up
every few years. The Southampton study aims to understand why that
is, by analyzing physiological responses,
especially of people who silently harbor and pass on the bug
without getting ill themselves.

Photograph shows a young man sitting in a doctor’s office holding a vial as two observers wearing gloves and aprons give him the thumbs-up sign.

A volunteer prepares to drink from a vial containing the
bacterium that causes typhoid fever, which kills roughly 200,000
people each year. The study, conducted in Oxford, England,
recruited local volunteers to test the effectiveness of a new
typhoid vaccine.


Read’s team has squirted a solution containing up to 100,000 of
the Bordetella pertussis bacteria that cause whooping
cough into the noses of 52 people. Thirty-two were colonized by the
microbe, and all were confined and closely watched in hospital for
two weeks, at which point the infection was ended with antibiotics.
The scientists now want to launch a second study that would allow
volunteers to live at home and spend time in the community,
enabling the number of enrollees to be much larger. The researchers
have been working with public health experts to make sure it would
be safe, for example by showing that infected volunteers do not
shed infectious bacteria from their noses.

Such work poses a unique ethical conundrum. The volunteers — and
sometimes those around them — are required to take on a known risk
without the promise of any direct benefit to themselves in return.
That’s very different, say, from people who already have cancer or
are infected with HIV signing up for an experimental medicine. It’s
why human challenge trials tend to focus on conditions with acute
(that is, immediate) symptoms that can be treated and cleared up
with no lasting damage. Ethical review committees charged with
assessing and approving these projects have given most the green

Most, but not all. In February 2017, bioethicists put the kibosh
on a human challenge trial that a group of scientists had proposed
for Zika — a virus that flared across Latin America in 2016 and is
linked to severe birth defects. The trial was going to give
volunteers a potential vaccine and then see how well it worked
against injections of small doses of the virus.

The ethicists, led by Seema Shah, then at the University of
Washington School of Medicine in Seattle, raised two sticking points. The first is what’s
known as the bystander risk. Because Zika can be sexually
transmitted, participants could pass it on to partners who did not
know about the trial and had not given consent. The second is that
in rare cases, Zika infections can lead to chronic paralysis and
even death, and the panel judged that a risk too far.

A table lists the risks to community members when participants take part in voluntary studies, including human challenge trails for malaria and dengue, and standard early safety studies for vaccine-testing. The table includes factors such as who might be at risk (for example, fetuses or members of the volunteer’s community), and the type of risk, such as joint pain or Guillain Barré Syndrome.

Participants in human challenge trials may expose their
communities to infection and the risks that come with it. In 2017,
a bioethics committee voted against human challenge trials for the
Zika virus after concluding that the potential risks to third
parties were too high. The chart compares these risks with risks
from human challenge trials for malaria and dengue fever, and from
standard early safety studies done for drug or vaccine

Shah didn’t see enough justification when the virus was widely
circulating: Why not inoculate and try to protect people in natural
populations? Since then, though, natural cases of the disease have
dramatically declined, so there could now be no realistic
alternative. And recent discoveries about the virus have also
tipped the ethical scales in favor of such a human challenge trial
— for example, new research shows that infected men are contagious
to sexual partners for just 30 days, which is less time than
previously thought. “It would be more ethically acceptable to do it
today,” says Shah, who is now at Lurie Children’s Hospital and
Northwestern Medical School.

The Zika case highlights a conundrum in vaccine testing, says
Michael Diamond, a virologist and immunologist at Washington
University School of Medicine in St Louis, Missouri. “A waning
epidemic is bad news for a vaccinologist,” he says, because to show
that vaccines work, “after they have been vaccinated you need
people to get infected.” The severe drop in Zika cases — while good
news for everybody else — means it’s become much harder to do that,
and much more expensive. A major benefit of human challenge trials
is that, because they deliberately cause infection, vaccines can be
verified using far fewer people — in the Zika case, perhaps just
200, says Diamond, who cowrote an update on Zika vaccine development in the
2019 Annual Review of Medicine.

A bar graph shows a spike in the number of cases of Zika virus in early 2016, predominantly in southern South America, Central America and the Caribbean. The number of cases drops sharply by mid-2016, just as trials to test vaccines begin.

After the 2016 explosion in Zika virus cases, the numbers
subsided, just as vaccine testing began (red bars). Since the best
way to test a vaccine’s effectiveness is in the middle of an
epidemic, when people will be exposed to the pathogen, a drop in
cases is a problem for vaccine development. Human challenge trials
offer an alternative testing route.

More money, more challenges

Such a focused approach is of great interest to funders. In
November 2018, the Wellcome Trust invited proposals for further
human challenge trials to run against more diseases endemic in low-
and middle-income countries. With grants of up to £5 million (about
$6.3 million) available for each project, the trust asked for plans
on how deliberate infection could accelerate vaccine development
for conditions from dengue and shigellosis to cholera and

Until now, much of the focus in such countries has been on
malaria. Since 2012, at least seven trials have infected healthy
people with the malaria parasite in Tanzania, Kenya and Gabon; a
further four are underway in Kenya, Equatorial Guinea and Mali.
These aim to test vaccines and antimalarial drugs, and to probe
basic science around infection, such as studying how natural
immunity develops. The push is backed by investment in lab
facilities (volunteers must typically be quarantined and watched
carefully) and advances in technology (a technique to extract,
purify and inject consistent doses of the malaria parasite has replaced the need to rely on mosquito

Now, as the Wellcome initiative and others gear up, the use of
challenge trials for other diseases in these parts of the world is
expected to accelerate, presenting procedural challenges. “We
should be especially vigilant when working in low- and
middle-income countries,” says Michael Selgelid, a bioethics expert
at Monash University in Melbourne, Australia. A central principle
of bioethics is to avoid working with vulnerable populations, or to
be especially careful when doing so. And volunteers in developing
countries, he says, are more likely to be vulnerable. Rules there
to protect human rights and govern research may not be as strict.
And participants may be less well-educated than in developed
countries and so be less informed about what they are consenting

Maureen Njue, a researcher with the KEMRI-Wellcome Trust
Research Programme in Kilifi, Kenya, has collected responses from
participants in one of the ongoing Kenya trials for malaria, and
has raised some concerns. In a paper published last year in the
journal Wellcome Open Research, she and colleagues warn
that current ethical guidance for more than minimal-risk
studies is inadequate
. One of the thorniest issues is how much
to pay volunteers — or even whether to call it a payment. Njue’s
team was careful not to do so, viewing the cash as compensation for
lost time and work. But cash is also a strong motivator for people
who participate in clinical studies.

The Kenyan volunteers received $480 for the three-week study.
That’s about five times greater than the typical unskilled wage
rate in the country, but still ten times less than what’s paid to
people on similar challenge trials in the US and UK. Several Kenyan
participants said that being offered more money would have made
them suspicious that the risk was higher than study organizers were
letting on.

Photograph shows four individuals sitting around a table. Each person has one arm resting on a large white paper cup containing mosquitoes.

In this re-creation of a challenge trial to investigate
treatments for malaria, volunteers place their arms over cups
containing infected mosquitoes. In this phase of the trial,
conducted at the Walter Reed Army Institute of Research, each
volunteer had to be bitten by five mosquitoes.


Shah says a useful way for scientists to assess informally the
bioethics of a project is to ask if they would be happy for
themselves or their loved ones to take part. Indeed, she says, some
researchers of human challenge trials go further and do
try it themselves — sometimes making themselves ill in the process.
(In doing so, these scientists are following a long tradition of
self-experimentation. Perhaps most famously in recent times, the
Australian Nobel Prize winner Barry Marshall drank a solution of the bacterium Helicobacter pylori to show it causes
stomach ulcers.)

But however well-meaning the participants and however
well-motivated they may be (by the death of a loved one, for
example), it’s vital that the proper checks are in place when
people join research projects, especially human challenge trials.
“We can’t let people consent to whatever they want,” Shah says.
“Informed consent is not a perfect protection and there has to be
additional scrutiny.”

Transparency and better communication are the simplest ways to
ensure that trials are done properly and to make improvements where
necessary, ethicists say. For example, although the Kenyan
volunteers consented to have 412 ml of blood removed during the
study, some told Njue they were still taken aback by how much that
turned out to be. Visuals to better demonstrate the quantity that
will be withdrawn are now included in the educational materials
shown to participants as part of the process. One participant, a
32-year-old man with 12 years of education, told the survey how a
clinician taking blood “brought a cup of water, a spoon and a
syringe” to show how much blood would be drawn, “so we were able to
verify that it was OK. So, I no longer have any doubts.”

There are other issues, too. Researchers in India point out in
an article in press at the International Journal of Infectious
that the very different hygiene practices and
sanitation infrastructure in that country mean that study protocols used in other places must be
. A typhoid challenge study in the UK, for example, can
safely send infected volunteers home between tests. But in India,
with unreliable sanitation, they would need to be quarantined to
ensure that the pathogen did not spread to the community.

Done properly, challenge trials are a powerful tool. The US, for
example, has used them over the last four decades in its
investigations of cholera. That research, performed with the help
of almost 200 healthy volunteers who were given the disease-causing
bacteria, produced a new single-dose oral vaccine, Vaxchora,
which was approved by the FDA in 2016.

Maryland required hospitals taking part to fly the traditional
yellow flag of quarantine. That’s an example of the openness that
experts say is the best way to avoid misinformation and nasty
outcomes in this growing and important field. As visitors to the
San Fernando Medical School in Lima can see for themselves, there
have been enough of those already.