That makes it difficult to know if vaccine developers are really on the right track. Their hunch is based, primarily, on how the immune system responds to other pathogens. But some viruses evade the typical patterns. They short-circuit the immune response. The most infamous example of that is HIV, Wherry says, which attacks the very T cells that would coordinate the immune response to the virus. SARS-CoV-2 has already offered its own twists and turns, like its propensity to prompt runaway immune responses. For Covid-19, “there’s no prototypical immune response, especially in severe cases,” Wherry says.

Lately, though, systematic studies of T and B cell responses to SARS-CoV-2 have begun to elicit some patterns. Recently, researchers at the La Jolla Institute for Immunology looked at T cell responses in what they considered “average” cases of the disease—people who got sick but didn’t need to be hospitalized. In a study published in Cell in May, they found that all of their subjects developed helper T cells, and 70 percent had killer T cells. The level of the T cell response, they found, roughly corresponded with levels of neutralizing antibodies. Other studies, including a recent preprint from a team at Oxford, have come to similar conclusions.

Simply having T cells—or even antibodies—that recognize the virus doesn’t mean you’re protected. There’s much more to learn about that. But on the vaccine front, the findings looked like good news. “It’s confirming that [helper] T cells are going to be an important factor in generating a robust antibody response,” Permar says. That’s a relief for pharma companies trying to replicate that process with a vaccine.

But immunology is never straightforward, remember? The La Jolla group’s findings came with a wrinkle: In a control group of blood donors who had never been exposed to SARS-CoV-2, the researchers also found T cells that recognized the virus. They speculate that those T cells might be “cross-reactive” with other viruses. Say you’ve had a lot of colds—especially colds caused by other coronaviruses. Then perhaps your immune system is primed to recognize this new virus based on its experience with other viral proteins, the researchers suggested.

It’s too soon to tell if those particular T cells offer any useful protection against SARS-CoV-2, the researchers caution. (In fact, let’s just make that the blanket caveat, for now.) But cross-reactive cells could have implications for vaccines too, Permar says. Those T cells could be a good thing, if they give their bearers a head start in producing antibodies after vaccination. Or they could backfire if a vaccine stimulates them to generate the wrong kind of antibodies, elbowing out a more fine-tuned response to SARS-CoV-2.

And what about when those antibodies wane in exposed people, or if they don’t show up at all? To address that question, Buggert’s team in Sweden took a slightly different approach. In addition to Covid-19 patients and two control groups—blood donors who gave samples before and after the pandemic began—they added members from the households of people with known cases. Those people, they reasoned, were at a greater likelihood of having been exposed to the virus than the general population, even if they had never shown symptoms or got tested.

Read all of our coronavirus coverage here.

Like other researchers, Buggert’s team saw cross-reactive T cells in unexposed people. But they also looked for responses that were specific to this virus by identifying a unique array of viral proteins those T cells recognized—a response not seen in the pre-pandemic blood donors. “People who have really been infected [with SARS-CoV-2] tend to respond against multiple different regions of the virus,” Buggert explains. “They have a broader response.” In a few blood donations taken during the pandemic, and in the household members of Covid-19 patients, they found that unique T cell response, but no antibodies.