The first tuberculosis vaccine was developed in 1921, by two French scientists, Albert Calmette and Camille Guérin. It was called Bacillus Calmette-Guérin, or B.C.G., and has long been one of the world’s most widely administered shots. From the beginning, its power was surprising. B.C.G. contains a bacterium similar to the one that causes TB, and engenders an immune defense specific to that disease. But, as Calmette noted in a paper in 1931, those vaccinated with B.C.G. at birth were around seventy-five per cent less likely to die in their early years of any cause. The effect seemed out of scale with the incidence of TB. There were, Calmette thought, two possibilities. TB might be more widespread than was commonly supposed. Alternatively, the vaccine might somehow confer a broader benefit—“a special aptitude to resist those other infections which are so frequent in young children.”
No one knew how to explain the phenomenon. But researchers observed a similar effect when polio vaccines were introduced, in the nineteen-fifties. Two married virologists, Marina Voroshilova and Mikhail Chumakov, conducted the Soviet Union’s clinical trials for Albert Sabin’s oral polio vaccine, known as O.P.V.; they noticed that it reduced not just the incidence of polio but of many other viral infections. Voroshilova started giving O.P.V. to her children every flu season, as a kind of prophylactic. Meanwhile, she looked at the medical records of more than three hundred thousand people, most of whom had received vaccines for polio and related viruses, across three winters. She found that those who’d received the vaccines were about seventy per cent less likely to have suffered acute flu or respiratory infections. Certain vaccines, she wrote, seemed to offer “a potential means of overcoming the diversity of pathogenic viruses.” (Her son Konstantin Chumakov is now the associate director for research at the F.D.A.’s Office of Vaccines Research and Review. “Knock on wood, but I don’t remember ever having flu,” he told me.)
In 1978, a Swedish organization sent Peter Aaby, an anthropologist with a doctoral degree in medical research, to Guinea-Bissau, in West Africa, to study the country’s high rate of child mortality. The next year, a measles outbreak hit a district in the capital city of Bissau, and Aaby began vaccinating children there. He, too, found that the vaccine reduced child mortality over all—by around half, according to his measurements, which was higher than what one would expect if the vaccine were preventing deaths from measles alone. “That was a very stunning experience,” he told me, on a video call from Bissau. “I guess that is what has kept me here, to try to understand what actually happened.”
Aaby published his initial findings in 1984, then followed up with a fuller report in 1995, describing similar results across ten studies in Bangladesh, Benin, Burundi, Guinea-Bissau, Haiti, Senegal, and Zaire (now the Democratic Republic of the Congo). Since then, he and others have reported that B.C.G., O.P.V., and a measles vaccine called M.V. have significantly reduced mortality from non-targeted diseases in low-income countries. Aaby has run randomized trials on infants in Guinea-Bissau and found that, according to some measures, B.C.G., O.P.V., and M.V.—all of which contain “live” bacteria or viruses, rather than chopped-up bits of pathogens—have reduced child mortality by at least thirty per cent. For this work, he won the Novo Nordisk Prize, Denmark’s most prestigious medical-research award. Surveying data collected by Aaby and others, researchers convened by the World Health Organization concluded that B.C.G. and measles vaccines reduce all-cause child mortality more than would be expected from specific protection. In a report published in the British Medical Journal, in 2016, they wrote that they “strongly recommend further studies.”
There’s a story we typically tell when we explain how vaccines work. We say that they prepare our immune systems to target specific intruders by programming antibodies and T cells. But for decades we’ve also had evidence of another phenomenon. Some vaccines appear to build a degree of defenses against nearly anything that comes our way. In this second mode of action, they work as a general immune-system booster. It’s like taking a cross-training class—for some period afterward, your whole body is extra fit.
The extent of these effects isn’t fully understood, in part because they’re understudied. “It’s almost counterintuitive to use one live vaccine to help protect against diseases caused by one or more other infectious organisms,” the medical researcher David Naylor, the co-chair of Canada’s COVID-19 Immunity Task Force and a former president of the University of Toronto, told me. In general, he said, “the scientific compass for many years has been swinging toward precision medicine.” Still, if we knew more about how some vaccines manage to provide a degree of broad protection, we could use that benefit to our advantage. In theory, we could use a preëxisting vaccine to protect against a new virus while more specific vaccines are still being developed. This is the strategy that Naylor adopted for himself during the pandemic. In the fall of 2020, he got a shingles shot.
Mihai Netea, a fifty-three-year-old immunologist at Radboud University, in the Netherlands, coördinates a research group of about twenty scientists, studying how organisms respond to severe infections. In 2010, his team was running an experiment to assess the impact of B.C.G. vaccination on Toll-like receptors—proteins that our cells use to respond to a broad class of microbial structures. The idea was simple. An initial encounter with B.C.G. should increase the production of immune-system molecules called cytokines in response to mycobacteria, the genus of bacteria that causes TB, and it should have no effect on the body’s response to a control stimulus—in this case, a fungus. But the cells showed increased reactivity to fungi, too. Netea thought that one of his students had made a mistake.
After they repeated the experiment and arrived at the same result, Netea searched the literature to see whether anyone else had documented the effects of B.C.G. on diseases besides TB. He discovered epidemiological studies, including Aaby’s, and also lab work done in mice. In those animals, “in the sixties and seventies, people were showing that B.C.G. protects against influenza, Listeria, malaria—everything,” Netea said. “And I thought, Oh, my God.” He developed a hunch about the mechanism at work, and, in a 2011 paper, gave it a name: trained immunity.
Most vaccines target what’s called the adaptive immune system: they work by aiming antibodies and T cells at specific pathogens. But we also have an innate immune system, a more indiscriminate first line of defense, which includes our skin, mucous membranes, and generalist proteins throughout the body that inhibit viral replication. Scavenger cells in this system attack foreign intruders—even ones the body has never seen before—and killer cells destroy any infected cell. All this happens no matter which pathogen is attacking us. Inflammation and fever, mediated in part by cytokines, are tools of innate immunity. Netea compared the adaptive and innate immune systems with specialists and hard laborers: one takes weeks to prepare, while the other goes to work in hours or less.
One proposed explanation for B.C.G.’s broad effectiveness was focussed on the adaptive immune system. Perhaps B.C.G. and the control fungus looked so much alike that adaptive immune cells aimed at the former also reacted against the latter—a phenomenon called cross-reactivity. But Netea suspected something else. Even in the absence of an adaptive immune system, one infection can bolster responses against future infections. In 1933, a similar effect, now called systemic acquired resistance, was identified in plants. Adaptive immunity evolved just half a billion years ago, roughly three billion years after life first appeared on Earth; many living things, including plants and all invertebrates, have only innate immune systems. And yet immunity in those organisms has a kind of memory, too—it can be sharpened by experience.
What could be the mechanism behind such effects? Netea thought that infections might be altering innate-immune cells through a process called epigenetic reprogramming. When cells make proteins—including the ones involved in innate immunity—they do so using instructions that are hard-coded into our DNA. But experience can affect which instructions a cell executes, and how often. In 2012, Netea confirmed that epigenetic changes were behind his lab’s earlier B.C.G. results. Increased production of certain signalling proteins had insured that cells launched an innate-immune response to the TB-causing bacterium while also doing the same for both a very different bacterium and a fungus. In a commentary published alongside the paper, Aaby and his longtime collaborator Christine Stabell Benn wrote, “It is rare that epidemiological and immunological data support each other to such an extent, telling a completely coherent and plausible story.” And yet the story was unfinished: the cells Netea had studied survive for only a few days, but epidemiological evidence showed that trained immunity could last for months or even years. Netea suspected that epigenetic changes took place elsewhere, too—possibly in the cells in our bone marrow that divide and differentiate into innate-immunity cells. Such changes would persist through time even as innate-immunity cells died and reproduced.
In 2020, a multinational team led by Sandrine Sarrazin and Michael H. Sieweke, of the Centre d’Immunologie de Marseille-Luminy, deepened the story, suggesting how the reprogramming might take place. If you were to stretch out all the DNA strands in your body and place them end to end, they would encircle Earth’s equator roughly two million times; they fit inside cellular nuclei by wrapping around proteins called histones, forming a tightly coiled complex called chromatin. A chromatin fibre, Netea explained, is like a book telling the body how to operate. Normally, most of it is closed—but its pages must open so that cells can read the instructions and make proteins. Inflammation, he explained, can alter the chemical structure of histones, essentially placing bookmarks. “You close the book, but the bookmark says, ‘This is where the chapter on fighting infection is,’ ” he said. “So, the next time you fight an infection, you can open it much easier.” Training makes innate-immunity cells faster and better at finding the plans for proteins that kill infected cells, produce signalling molecules, and trigger adaptive immunity.
It’s not just vaccines that train our immune systems. We get sick all the time, experiencing colds, flus, or fatigue; when such illnesses aren’t severe, they usually leave us stronger, better prepared to battle the next infection. Netea recalled how, in his children’s first years of schooling, they continually got sick, with fevers and runny noses. Slowly, illness became less frequent. “Well, part of it is producing antibodies, and so on,” he said. “But also the innate immune system matures. You put in these bookmarks. You catalogue the information.” Some vaccines, it turns out, can help us build those capacities while avoiding sickness.
In 1984, Robert Gallo co-discovered H.I.V. as the cause of AIDS; in 2011, he co-founded the Global Virus Network, an international coalition of virologists aiming to prevent and control viral epidemics. Early in 2020, when Chinese researchers published the genetic sequence of SARS-CoV-2, Gallo grew interested in the possibility of using trained immunity to slow the spread of the virus. A few years earlier, he’d attended a lecture given by Konstantin Chumakov about protecting against the flu with O.P.V. He’d also read about bats, which harbor several coronaviruses simultaneously without growing ill, and without antibodies. “How do they do fine?” Gallo said. “It’s not the classical adaptive-immune response. It’s on all the time. They keep a balance, so that the coronavirus is there but not harmful. This really caused me to become deeply interested in innate immunity.” In 2020 and 2021, Gallo co-authored two high-profile articles—with Chumakov, Aaby, Benn, and Netea on one or both—advocating an attempt to fight the coronavirus by eliciting trained immunity through the use of existing vaccines.
To some extent, the idea had already been put into practice by accident, through ordinary vaccination. Researchers at Virginia Tech and the National Institutes of Health have found that, among twenty-two socially similar countries, those with greater B.C.G. coverage had lower COVID-19 mortality rates. (This finding is remarkable considering that the last common ancestor between the bacterium that causes TB and the virus that causes COVID-19 existed more than three billion years ago.) Studying O.P.V.’s capacity for COVID-19 protection is more difficult, because supply is reserved for polio eradication, but an analysis by Chumakov, Gallo, and others found that Iranian mothers who’d been indirectly exposed to O.P.V. through their children’s vaccinations—the vaccine is transmissible—were better protected against the coronavirus. A study in Brazil, conducted by Swiss and Brazilian researchers, found that a flu shot reduced the odds of death from COVID-19 by sixteen per cent.
Other studies seem to tell similar stories. Netea and his collaborators conducted a small study at a Dutch hospital and found that those who’d been immunized against the flu were roughly forty per cent less likely to contract COVID-19, thanks, it appears, to changes in their innate immune systems. In a GlaxoSmithKline study of nearly half a million adults in California aged fifty and above, those who’d received G.S.K.’s shingles vaccine were sixteen per cent less likely to contract the coronavirus, and thirty-two per cent less likely to be hospitalized because of it. And, among more than a hundred and thirty-seven thousand Mayo Clinic patients, those who’d received any of several vaccines within the past one to five years—including shots for chicken pox, flu, hepatitis, measles, pneumonia, and polio—had lower chances of COVID-19 infection. Polio vaccines reduced the probability of infection by forty-three per cent—even when controlling for comorbidities, other vaccinations, demographics (age, gender, race, ethnicity, county of residence), and regional COVID-19 incidence and testing rates. Numerous studies by researchers around the world have presented harmonious findings.