Sometime in the late 1980s, in a town in southwest England called Salisbury, 15 volunteers agreed to have a cold-causing coronavirus known as 229E squirted into their noses in a saline solution. Ten of the volunteers were successfully infected, as determined by viruses recovered from their noses in the days following, although only eight displayed symptoms. Researchers monitored the levels of antibodies and immune cells in their blood over the ensuing weeks.
A year later, 14 of the same volunteers came back for another round. Of the nine people who’d become infected with the first exposure, six became infected again, but none developed colds. Moreover, they only shed virus from their noses for a couple of days, compared with an average of five and a half days the first time around. As for the five people who’d resisted infection the first time around, all became...
The researchers struggled to explain the results. “These data do not fit any simple model,” they wrote in their report. “It may be that the small amounts of antibody remaining in the original infected group contributed to resistance to reinfection in some volunteers. It may also have prevented colds and shortened the duration of virus shedding.”
Three decades on, as a pandemic caused by a different coronavirus rages, researchers are still scratching their heads over whether coronaviruses—in particular, SARS-CoV-2—provoke lasting immunity in people they infect. With no long-term data yet on the immune and other effects of SARS-CoV-2 infection, some scientists are returning to 229E and three other coronaviruses that have been with humans much longer—OC43, NL63, and HKU1—in hopes of finding clues to this question.
One study hoping to shed light on immunity after a coronavirus infection, posted on medRxiv in June and not yet peer-reviewed, drew on blood samples from healthy control subjects in an ongoing HIV project that began in 1985. [Update: This study has now been peer-reviewed and was published in Nature Medicine on September 14.] Researchers based at the Amsterdam University Medical Center (UMC) and their colleagues at other institutions analyzed stored samples from 10 subjects who had their blood collected every three to six months for at least 10 years, looking for antibodies to proteins from the four known cold-causing coronaviruses that would indicate a recent viral infection.
The research team knew of the earlier 229E reinfection study, so they weren’t surprised to see multiple 229E infections in the same subjects crop up in their own data, as revealed by increases in antibody levels, says Arthur Edridge, a physician and Amsterdam UMC graduate student who is the paper’s first author. “What was surprising for us is that [reinfection] actually seemed to be a common feature for all the seasonal coronaviruses that we studied,” he says. All but one study subject had been infected with a particular coronavirus multiple times over the period of the study, and in some cases the time between infections with the same virus was as little as six months to a year, indicating an “alarmingly short duration of protective immunity,” the authors write in their paper.
Edridge cautions that it’s not clear whether SARS-CoV-2 will follow the same pattern as these more familiar coronaviruses—but if it does, then the idea that allowing the virus to spread in order to achieve herd immunity wouldn’t be a successful strategy, he adds.
Another recent study to find evidence of coronavirus reinfection was an analysis of data from a respiratory virus monitoring program conducted between 2016 and 2018. That study, which included 214 children and adults in New York City and relied on self-reports of symptoms and viral RNA swabbed from the back of the throat, found 12 instances of reinfection by the same coronavirus, although nine of these were in children, whose immune systems are less developed than those of adults. Reinfections were found for three of the four cold-causing coronaviruses (OC43, HKU1, and 229E).
It’s not clear whether SARS-CoV-2 will follow the same pattern as these more familiar coronaviruses—but if it does, then the idea that allowing the virus to spread in order to achieve herd immunity wouldn’t be a successful strategy.
Marta Galanti, a postdoc at Columbia University and the study’s first author, notes that the reinfections fell into two clusters in terms of timeline: at four to eight weeks after the initial infection, and at 8 to 10 months after the initial infection. She and her coauthor weren’t able to rule out the possibility that the earlier reinfections were in fact persistent first infections, she says, although they’re working on this in a follow-up study.
Like Edridge, Galanti is clear that the reinfection results don’t necessarily apply to SARS-CoV-2. But, she says, they indicate that “we have to be prepared [for] the possibility that the multiple subsequent infections can happen” with the novel coronavirus.
“Maybe it’s possible that if you only have these mild respiratory symptoms [with SARS-CoV-2 infection], you don’t develop a really strong immune response, and you could get reinfected,” says Rachel Roper, an immunologist at East Carolina University who was not involved in either of the studies. But she still thinks there’s uncertainty about whether reinfections occur with endemic coronaviruses, and she adds that infection with murine hepatitis virus, a coronavirus that causes serious disease in mice, confers lifelong immunity, as she suspects more severe cases of COVID-19 would. “If you had a serious infection the first time, all indications are you’ve got a stronger immune response,” and would either be immune to a second infection or experience only mild symptoms the second time around.
Prospects for long-term immunity
In general, whether a viral infection elicits long-term immunity depends on characteristics of both the virus and the host, says Akiko Iwasaki, an immunologist at Yale University. She adds that studies have found that the number of antibodies produced by COVID-19 patients varies widely, which may indicate that their protection against future infection also varies. Other studies have shown that levels of antibodies to SARS-CoV-2 tend to wane after a few months. But Iwasaki adds that antibodies don’t reveal the full picture of immunity, as even if antibodies to a virus are at undetectable levels, people who’ve been exposed to the pathogen before may be able to mount a quick response thanks to T cells and B cells that “remember” the previous infection.
Even if immunity to SARS-CoV-2 does turn out to be short-lived after infection, it might not be bad news for vaccine development, says Roper. “A lot of viruses, when they infect, also do things to turn off the mammalian immune response,” she says. “Hopefully, when we design vaccines, the vaccines won’t have those proteins that shut off the host immune response. So it should be possible to get a better immune response to the vaccine than you do to the virus.”