We’ve had some good news on coronavirus immunity recently – good because it gives us some more clarity on the whole situation, and because it suggests that both people who have already recovered from the infection and people who will be getting vaccinated can have good protection.
We have this preprint from several of the Boston/Cambridge area institutions, comparing antibody levels in 259 infected patients (out to 75 days post-symptoms) with 1548 pre-pandemic samples. They’re looking specifically at IgM, IgG, and IgA comparisons. IgM is the first antibody type to appear in response to an infection – they’re the ones I mentioned in this post as being five of the Y-shaped units put together. IgG are the ones that most people are talking about when they talk about antibodies in the blood, and IgA are secreted mucosal antibodies, found in the saliva and nasal/lung tissues. That mucosal immunity is surely an important factor in a disease that appears to be spread largely by inhaled airborne droplets.
They estimate that it takes about 11 days to “seroconvert” after infection, that is, to show evidence that your immune system has raised these new antibodies to the coronavirus. Looking at hospitalized individuals versus milder cases, the former actually seroconverted a couple of days earlier, and their IgM response tended to drop off a bit more quickly. One limitation of this paper is that the coronavirus patient cohort was skewed towards the hospitalized patients, rather than mild infections. But overall they found that IgG antibodies were still detectable in serum 75 days out (the last time point measured) and were a very useful marker for infection, although IgM and IgA had mostly gone back down by then. The paper concludes that although we don’t yet know the optimum levels of antibodies for protection (and we’re still collecting later time points!), “the association between RBD-IgG with neutralizing titers and the persistence of these antibodies at late time points is encouraging“.
Here’s a report on the same topic from a multicenter team in Canada that’s comparing antibody level both in the blood and in saliva. In contrast to the report just mentioned, the authors here found IgA against the coronavirus persisting for at least three months after infection, and correlating well with IgG in the blood. Not every study has shown that sort of persistence, but the authors believe that this might be due to the techniques used for detection. They also found that several of their negative controls – samples banked from people pre-pandemic, who had definitely not been exposed to coronavirus – also showed IgA titers in saliva, presumably cross-reactive antibodies that were raised from some other infection. “It is tempting to speculate“, they write, “that these preexisting IgA antibodies may provide some stop-gap protection against SARS-CoV-2 in the oral cavity, and if so, it is essential to ascertain their original antigenic specificity.”
Then there’s this preprint from a team in Arizona, which confirms these results by finding that antibodies against the RBD part of the coronavirus Spike protein persisted for at least three months. “In contrast to other reports“, they say, “we conclude that immunity is durable for at least several months after SARS-CoV-2 infection“. They also checked antibodies to the nucleocapsid protein (N) as well as to the spike and found that the N response was more variable. One possibility they raise is that there are cross-reactive antibodies to other coronavirus N proteins (which is a more conserved domain across the various types than the Spike), and that these were raised by previous infections with different viruses.
Moving past antibodies, we have this paper on T-cell responses. The authors, a multinational team led out of the Karolinska Institute in Sweden, have put in a lot of work looking at the T-cell situation in unexposed individuals, people with acute coronavirus infections and those who have recovered, and family members of those patients as well. The acute phase subjects had just the sort of cellular profile you’d expect: highly activated and cytotoxic, out there killing virus-infected cells as T-cells were born to do. In the convalescent patients this had calmed down, as it’s supposed to, and they detected stem-memory-type cells, which is just what you’d want to see. Importantly, these were also found in people who had recovered from much milder infections and in the asymptomatic family members tested as well. (The paper provides a great deal of detail on the exact sorts of responses in the various T-cell types that I’m not going into, but it’s valuable information).
They also detected potentially cross-reactive T cells in 28% of people who had donated blood before the pandemic even hit, which is consistent with several other reports. 41% of the overall patients who were seronegative in antibody tests were still positive for T-cells (CD4+ and CD8+ alike) against coronavirus proteins (Spike, nucleocapsid, and membrane). They conclude that the T-cell response is indeed non-redundant and apparently an important part of immunity to this virus, and that using seroprevalence (antibody levels) as a marker for exposure in a population will almost certainly underestimate the real situation. That’s good news, since it would mean that more people have already been exposed (and are to some good degree immune) than we would think. But that doesn’t mean that we can blow the all-clear whistle, either, as the various surges in infection around the world have shown: the situation may be better than feared, but we don’t seem to be anywhere near “herd immunity” levels yet. And we would be killing off an awful lot more people to get there without a vaccine.
And finally, we have this report from the University of Washington, which is about as direct a measure of immune protection as we’re likely to get before the vaccine efficacy trials read out. The authors studied the crew of a fishing vessel, before and after its voyage. Three crew members showed a positive antibody response beforehand, indicating that they had been infected earlier in the epidemic – they were not positive in RT-PCR testing, though, indicating that they did not have active infections. In fact, none of the other 120 crew members tested (out of 122 total) had such a positive reading on departure.
But an outbreak occurred on the ship anyway – someone was early enough in the infection that they hadn’t shown positive yet. This crewmember became sick and the vessel returned to port on Day 18 of the voyage. (Sequencing of viral samples from a number of crew members confirmed that the outbreak seemed to originate from a single source). Testing then and over the next few days showed that 104 of the 120 crew members were now positive for the coronavirus – but not the three who had antibodies beforehand. This high attack rate (!) and apparent protection makes a strong case for protective immunity, because God knows everyone on board had plenty of chances to catch the disease.
How long this protection lasts, what part of it is due to antibody response and what part to T cells, and what the exact cutoffs are for those – these are the important things we don’t quite know yet. But the picture is becoming clearer. And what we’re seeing is that this virus, which it has definitely has some unusual features, is also something that our immune systems are dealing with in the just the way that you would hope to see. That gives us hope that the vaccines are in turn going to raise protective, lasting responses. We just have to see the details!