What to Make of Research Suggesting Immunity to Covid-19 Is Short-Lived

Be careful not to jump to conclusions

Yesterday, a preprint that was recently posted to the server medRxiv (meaning the study has not undergone peer review, an important step for accurate science research) generated a flurry of excitement — and criticism — for its implication that immunity to the coronavirus might only last a few months. The paper suggested that the levels of antibodies people produce in response to Covid-19 infection rapidly decline after they hit their peak.

The paper’s findings, which were covered by large news outlets like The Guardian, CNBC, and CNN, suggest that people’s immunity after a Covid-19 infection wanes over time. And that doesn’t seem to bode well for hopes of herd immunity or a vaccine. A couple days earlier, Vox published a story written by a doctor suggesting that a patient who recovered from Covid-19 got reinfected, raising similar concerns about immunity to the coronavirus.

But as critics pointed out on Twitter, these case studies need to be a part of a larger understanding and study of Covid-19 and immunity. The study has not yet been reviewed by scientists, and the Vox story is a single opinion piece about one case. Importantly, this framing also assumes that antibodies are the only way to achieve immunity, which is simply not the case.

In the medRxiv study, King’s College London researchers repeatedly tested the antibody levels of 96 people who tested positive for Covid-19 between March and June. They observed that while 60% of people had a “potent” antibody response around 23 days after their symptoms first appeared, only 16.7% of them still had the same level of potency an average of 65 days later. They also found that people who had more severe Covid-19 infections tended to have more and longer-lasting antibodies than people with milder cases.

Altogether, the findings raised concerns that antibodies can’t be counted on to provide long-lasting protection to the virus — which is troubling, given that a successful vaccine is generally thought of as one that produces long-lasting antibodies.

But antibodies aren’t the only way in which the body can fight Covid-19, thanks to the existence of the immune system’s multiple branches. It’s understandable if you didn’t know that they exist: Over the course of the pandemic, many science experts and communicators (myself included) have written about immunity solely as it relates to antibodies, but the reality is actually much more complex than that.

The antibody-mediated branch of the immune system has been widely discussed during the pandemic — specifically, its ability to respond to infection by making neutralizing antibodies, the kind that can fight off the virus upon future encounters. Neutralizing antibodies are the ones that the preprint refers to, that vaccines are meant to produce, that antibody tests try to detect, and that are sought in convalescent plasma. But there are many more types of antibodies that also work toward immunity — they just haven’t been talked about as much.

The other main branch of the immune system is known as cellular immunity or cell-mediated immunity. It doesn’t have much to do with antibodies at all. Instead, it relies on T-cells, which have different methods of fighting off pathogens they have previously encountered. There is ongoing research on cellular immunity to Covid-19, but it too hasn’t been talked about as much. If antibody-mediated immunity turns out to have a limited role in Covid-19 infection, cell-mediated immunity could be a much-needed alternative or complement.

I keep returning to a Twitter thread by Eric Topol, MD, chair of Innovative Medicine at Scripps Research, on how to think about new research on immunity and the coronavirus. In it, he emphasizes the importance of acknowledging the immune system’s other branches. “So when you interpret seroprevalence, serology, antibody only data, please keep the incomplete nature of that assessment in terms of the broader cellular immune response to #SARSCoV2,” he writes. “There’s a lot more to learn about the stuff we don’t measure.” He also points to research showing that cells that aren’t normally considered to be part of the immune system, like those in the thymus or liver, have their own immune-like response to the coronavirus.

There remain several ongoing questions about reports of people testing positive for Covid-19 after seemingly recovered from the virus. Are they having a relapse of symptoms, or did they indeed catch Covid-19 a second time? Scientists are still working through the possibilities.

A newsletter from the Johns Hopkins Center for Health Security sent out today similarly cautions readers about interpreting the medRxiv paper and the Vox story. The “duration of immunity is unclear,” it says, and it’s “unknown whether or not recurrent infections are tied to waning immunity after initial infection.” Furthermore, it notes, just because a person tests positive for antibodies doesn’t mean they’re protected from reinfection, since antibody tests don’t necessarily pick up on the neutralizing antibodies required for immunity.

In a blog post published today, Francis Collins, MD, director of the National Institutes of Health, acknowledges “rare reports” of people who recover from Covid-19 only to test positive again. (Though he doesn’t specify which reports.) He says that “these results might be explained by reports that the virus can linger in our systems,” but reiterates that many questions remain about immunity. In a recent study, Collins points out, macaques that had been deliberately infected with Covid-19 mounted a protective antibody response when they were reinfected 28 days later.

None of this is to say that the King’s College London study or the case study described in Vox can’t be meaningful — just that their findings (which need to be studied and replicated before they are considered strong evidence) represent only one part of the potential immune response to Covid-19. This is helpful to keep in mind when thinking about the World Health Organization’s recent acknowledgment that antibodies may wane after several months and the recent study by the Spanish government that found antibodies could disappear after a few weeks. It isn’t known whether antibodies last a long time, whether they actually protect us from reinfection, and whether cell-mediated immunity has a role to play, but at this point, being uncertain how to answer these questions leaves us in a better state than definitively answering them with a “no.”