In the wake of COVID-19’s first surge, we did our best to shelter at home, while doctors and nurses faced the mysterious virus head-on. In emergency rooms and intensive care units across the country, they not only witnessed first-hand how deadly their new adversary could be, but the many unpredictable ways that it could kill. Emergency medicine doctor David Lee recalls being struck by the gamut of different trajectories. Some patients were “just fine,” experiencing nothing more than a temporary loss of taste and smell. Others had mild-to-moderate upper respiratory symptoms that resembled the common cold. In the worst cases, though, people would start getting better, and then, “bam, suddenly, something went wrong.” And that something was frighteningly variable: asphyxiation, heart attacks, blood clots, and in rare cases, even outbreaks of psychosis. It almost seemed like they were dealing with multiple diseases in one.

Lee and his colleagues eventually deduced that it wasn’t just the virus wreaking havoc on patients. SARS-CoV-2 was making some people’s immune systems go haywire, even young people without known pre-existing conditions. As 2020 wore on, clues started to pile up that some of the sickest COVID-19 patients were experiencing autoimmune reactions—their immune systems were attacking their healthy tissues, including cells in the lungs and blood. This, then, was another way that the pandemic hit our society in one of its soft spots. Autoimmune diseases like Multiple Sclerosis and Lupus have long been taking a massive toll on public health (with around 7% of Americans affected). And, beyond the fact that they probably arise from a combination of genetic predisposition and environmental triggers, we just don’t know very much about them. How and why exactly the body turns on itself remains an open question.

This knowledge gap is consequential. Research on how to prevent autoimmune disorders before they can be triggered, and how to treat them once they have started, is still in its early stages. Existing interventions can’t cure most of these conditions, and even with treatment, most of them come with a decreased life expectancy. There are very particular historical reasons for the underdeveloped state of this field. Our knowledge of autoimmune disorders was significantly impeded by the work of one person—quintessential great man of science, Paul Ehrlich—and by other great men of science who abided by his findings uncritically. A hundred years later, we’re still running into hurdles he left behind. Arguably, now more than ever.

A Complicated Legacy

Ehrlich’s life-saving contributions to science are well-known. In 1908, he shared the Nobel Prize in Medicine for his work on immunity. He was instrumental in theorizing the existence of  antibodies—proteins with shapes that bind to proteins that stick out of pathogens, like a lock to a key. His work contributed to the first treatment for diphtheria, to Gram staining methods used to categorize bacteria, and to the development of chemotherapy as we know it. His one great disservice to humanity, meanwhile, has been largely overlooked.

At the turn of the 20th century, Ehrlich coined the term horror autotoxicus, or the horror of self-toxicity, which reflected Ehrlich’s staunch rejection of the then-new hypothesis that autoimmunity was possible. With the help of his assistant, Ehrlich conducted experiments in which rabbits, goats, and other animals were injected with the blood of another species or the blood of the same species. They found that individuals injected with a different species’ blood formed antibodies against the foreign cells, while those injected with the same species’ blood didn’t form disease-causing autoantibodies. Ehrlich thus purported to prove that the immune system simply wouldn’t do something as counterproductive as damaging healthy cells.

This kind of self-destruction “would be dysteleological to the highest degree,” he wrote in his widely-acclaimed 1906 book Collected Studies on Immunity. “[T]he organism possesses certain contrivances by means of which the immunity reaction, so easily produced by all kinds of cells, is prevented from acting against the organism's own elements. Further investigations made by us have confirmed this view, so that one might be justified in speaking of a ‘horror autotoxicus’ of the organism. These contrivances are naturally of the highest importance for the existence of the individual.”

That same year, physicians Julius Donath and Karl Landsteiner conclusively (and correctly) demonstrated that paroxysmal cold hemoglobinuria (PKH), which damages blood cells and causes anemia, was an autoimmune disease. More proven examples of autoimmune diseases followed. Ehrlich, a much bigger name than any of the careful scientists disproving his theory, swatted all evidence aside. By then, horror autotoxicus was an established part of immunology. In the two years leading up to Ehrlich’s Nobel prize, immunologists and physicians all over the world devoured his Collected Studies on Immunity wholesale, flushing the good and the bad bits through the system for generations.

Up until his death in 1915, even when presented with evidence of autoimmunity—like the formation of antibodies against an organism’s own sperm in vitro (outside a living system in a laboratory)—Ehrlich insisted that horror autotoxicus would prevent self-toxicity in a living organism. When a pioneer stands at the forefront of a scientific discipline, it's easy to mistake their ideology and biases for objectivity. Mistake or not, this kind of insistence has consequences. For decades thereafter, scientists regurgitated that dictum, even after repeated evidence surfaced that autoantibodies harm the tissues of organisms.

If the existence of autoimmune disorders was still treated as an open question well into the 1950s, this was largely thanks to Ehrlich and his influential disciples. Noel Rose, the father of autoimmune disease research, made this connection explicit in a story about his mentor Ernest Witebsky, who was a “second generation” follower of Ehrlich’s. At the celebration of the Paul Ehrlich centennial in 1954, Witebsky lauded the “ingenious” concept of horror autotoxicus, saying, “[n]o living organism would be capable of producing—or would dare to produce, if you wish—an antibody against constituents of its own body, for this would be incompatible with life.” He believed so deeply in the dictum that, when his experiments with Rose showed the existence of thyroid autoantibodies, he suppressed the results and spent three years trying to find experimental errors in their work. He finally recanted, and the results were published in 1956.

When such an idea takes hold in a discipline, it can be difficult to purge. And when it is purged, you have to make up for lost time. In this case, fifty years of subsumed research amounted to lives that could have been spared—several generations of doctors who didn’t take autoimmune disorders seriously—and a key driver of COVID-19’s mass casualties that could have long been attenuated.

Why It Matters Now

With many viral illnesses, the severity of a case will increase with the amount of virus particles in the body. For instance, when a patient’s HIV viral load is high, they get full-blown AIDS, whereas a low viral load is typical for patients in remission. In a significant subset of patients, COVID-19 follows the opposite trajectory. This was one of the things that made the novel virus such a mystery at first, says Emergency medicine doctor David Lee. “[Patients] expect doctors to know every single answer,” he says, creating a cognitive dissonance as we encountered this new disease that “doesn’t fit the mold.” Experts had a tendency to revert to their tried-and-true answers and treat the brand new virus as if it were like familiar older diseases to, in Lee’s words, help “resolve the cognitive dissonance." He adds that “most of the scientific world in medicine is still treating this as a viral pneumonia with viral complications,” and is “basically looking at this as just like some viral-mediated disease.” But with SARS-CoV-2, often, “your viral load is highest [when] you have the sniffles...or maybe a sore throat or maybe you lost your sense of smell. Why is it that by the time I see the patients dying on a ventilator, their viral loads are much lower?”

This led Lee and other doctors to an obvious hypothesis: if the viral load is low in many people with severe, prolonged diseases, then autoantibodies, or some other autoimmune phenomenon like autoinflammation, could be at work. This hypothesis has since been more and more widely substantiated. “[I]t is pretty clear that a subset of severely ill COVID patients suffer from unchecked inflammation, cytokine storm and autoimmunity,” says PJ Utz, a physician scientist in immunology and rheumatology at Stanford University.

Utz points to the work of Jason Knight’s Lab at the University of Michigan, which studies antiphospholipid syndrome (APS), a condition in which antibodies mistakenly target blood cells and the lining of blood vessels. It’s thought to be underdiagnosed in patients, and the inner workings of the disease, which can cause blood clots, heart attacks, permanent disability, shortened lifespan, and maternal and fetal complications, have yet to be elucidated. As the pandemic’s death toll began to mount in the early spring of 2020, autopsies revealed that microvascular clotting, or blood clots in the rich network of tiny vessels in the lungs, were causing the respiratory distress. Knight and his colleagues took note, and that November, they published results in Science Translational Medicine showing that, among 172 patients hospitalized with COVID-19, half had autoantibodies in the blood and vessels like those found in patients with the worst cases of APS.

It’s not entirely clear whether infection with SARS-CoV-2 activates existing autoantibodies in seemingly healthy people or if it somehow triggers the formation of new ones (healthy people can also have a resting level of autoantibodies). The question is how and why these dispositions go awry. In a preprint published in January 2021, Utz and his colleagues report findings that suggest that a significant subset of COVID-19 patients develop new-onset autoantibodies. In another paper published this month, Lee and his co-authors report finding higher levels of autoantibodies against the protein annexin A2 — which helps maintain cell membrane structure — in patients who died of COVID-19 than those who survived. “Many peer reviewed publications strongly suggest that SARS-CoV-2 markedly alters the immune system,” Utz notes. “And it is unclear how long it will take before immune homeostasis reverts back to normal.” There’s also evidence that the virus may trigger autoimmunity against interferons, which are protein molecules that govern early immune response, allowing the virus to infiltrate under the body’s radar. There’s even evidence of COVID-19 triggering the formation of autoantibodies in the brain, which may be causing neurological symptoms as well as an autoimmune reaction against the pancreas, leading to new-onset type 1 diabetes.

The goal of the COVID-19 vaccines is to induce the formation of antibodies against SARS-CoV-2 without having to go through serious illness. That way, when vaccinated people do inhale COVID-19-laden droplets, the vaccine-induced antibodies oust the virus before it can cause severe symptoms. But by the time COVID-19 patients get severely ill, many of them aren’t dealing with viral damage.

“COVID-19 will forever change how we think about the interplay between infection and autoimmunity, including horror autotoxicus,” says Utz. “I predict that in the future, patients with severe COVID or other severe viral infections will be screened for anti-cytokine antibodies, particularly anti-interferon alpha, and if they have such antibodies their treatment could include trying to remove the pathogenic antibodies or to alter treatment in other ways.”

Lee says that the link between autoimmunity and not only COVID-19 but potentially hundreds of viral infections isn’t receiving enough attention. “The hallmark of science is reproducibility,” he says. But if not enough researchers believe that it’s important to investigate, then they’re not “going to venture to attempt a study they don’t believe in.” As the pandemic spread its way through summer of 2020, Lee laments that doctors "were siloed into our little different fields” as usual. He recalls that when he spoke with infectious disease doctors, they would say, “we need more antibodies, we need more antibodies [for hospitalized patients].” But when he spoke with rheumatologists, they would say that “actually some antibodies [may have] caused some of the worst patient outcomes.” This fatal separation between medical specializations is nothing new. It goes back to Ehrlich, his flawed ideology, and the spurious findings that resulted from his biased experiments.

Would we be better off in a world in which Ehrlich and his adherents didn’t find self-toxicity so horrifying as to be impossible? That’s not a question with a clear answer, nor one that can be disentangled from all of the medical advances he inspired. Nevertheless, with our incomplete understanding of immunity; with millions of people suffering from autoimmune diseases without correct diagnoses or hopes of a cure; and with one in three COVID-19 survivors dealing with long-term illness, horror autotoxicus still casts its long shadow. It stands as a painful example of what can happen when scientific acclaim and authority trump scientific rigor—when we forget about the inherent, limiting human-ness of science. ♦