From the start of the COVID-19 pandemic, many of researchers’ nagging questions involved trying to understand what constitutes immunity to future infections. People who had been infected by the virus produced varying amounts of antibodies, and it wasn’t clear what levels were needed to provide protection. Similar issues applied to figuring out how long protection lasted, given that antibody levels appeared to decline over time. Those questions have implications for whether we will eventually need booster shots to maintain our immunity.
The most common way of looking at immunity at the beginning of the pandemic was to check for neutralizing antibodies, which could block the virus’s ability to infect new cells. But we’ve gone through much of the pandemic without knowing exactly how levels of these antibodies relate to protection.
Evidence has been building that neutralizing antibodies directly correlate with protection, and a new paper provides some of the most decisive evidence yet. The authors also provide some hints about the sort of decline in immunity we might expect.
There are many easy ways to check for the presence of antibodies that target SARS-CoV-2. Most attention, however, has focused on neutralizing antibodies, which stick to the virus in such a way that it is prevented from infecting cells. There are a number of methods to assess neutralization, too, from using the actual SARS-CoV-2 virus as antibody bait to using a harmless virus carrying the gene that encodes the SARS-CoV-2 spike protein.
Many of the vaccine clinical trials have included information on the levels of neutralizing antibodies people make in response to immunization; there is also a lot of data on levels of neutralizing antibodies people make after natural infections. Researchers have followed many of those vaccinated and recovered people to determine whether they get infected afterward. It would seem that we’re well-positioned to see how neutralizing antibodies correlate with immunity, and data suggests that it does.
The problem is that most of these studies often ran for different periods of time and used different means of assaying for neutralizing antibody levels, so the data didn’t add up to a coherent, consistent body of evidence. And individually, most of the studies were too small to provide a clear answer.
The new study, performed by a large group of Australian researchers, normalizes all the results from clinical trials of vaccines (seven in all) back to a comparison with the levels of neutralizing antibodies generated following an infection, all while accounting for the time elapsed since vaccination. The researchers also included a single study that tracked the outcomes of infected individuals, with the data adjusted for time since infection.
The results were very clear, showing “a remarkably strong, non-linear relationship between mean neutralization level and the reported protection across different vaccines.” In other words, the higher the levels of neutralizing antibody that someone produces, the more likely they are to be protected.
This doesn’t necessarily mean that neutralizing antibodies are solely providing the protection; it could be that they correlate with some other aspect of protective immune function. But it does mean we can measure levels of neutralizing antibodies and have some confidence that we’re getting a measure of overall immunity.
To demonstrate this, the researchers used vaccine trial data that was released while they were performing their study. Based on the levels of neutralizing antibody generated by the vaccine, the researchers’ data would suggest that the vaccine should provide about 79.6 percent efficacy. The actual reported result was 80.6 percent. So it seems like we now have a reasonable measure of protection from COVID-19.
A slow decline
The other big issue addressed in the study is how quickly immunity declines, which should tell us a lot about how long heavily vaccinated populations can go without needing vaccine boosters or worrying about having to restart preventative measures.
Antibodies produced by vaccines and infections declined at roughly similar rates, with half-lives of 58 days and 65 days, respectively. But given the strength of the response to vaccines, this result isn’t a huge problem. A vaccine with 95 percent efficacy after the second dose would still have an estimated 77 percent efficacy 250 days out. And that’s for protection against a symptomatic infection. The protection against severe COVID-19 is much stronger and would likely take far longer to decline.
Vaccines with a lower initial efficacy present more of a potential problem. A starting efficacy of 70 percent would be down to 33 percent efficacy at 250 days.
There are a few significant caveats to this data, mostly identified by the researchers themselves. One is that it assumes that protection from COVID-19 remains linked to the levels of neutralizing antibodies in the blood stream over time. For long-term protection from a pathogen, however, the immune system switches over from a bunch of active cells to a smaller number of memory B cells. While this process results in lower levels of antibodies, the memory cells can rapidly remobilize in response to a new exposure. There are clear indications that SARS-CoV-2 infection results in the production of memory immune cells, suggesting that the same thing is likely to be happening among vaccinated people.
Other aspects of immunity, such as the T cell response, will also play a role in controlling the virus. Even if antibody production fades, it’s possible that other aspects of the immune response can partially compensate.
So we’re still left with a significant question. If the memory cells and other immune mechanisms can respond quickly enough to keep infections from becoming severe, we may end up with longer-term immunity than the decline of antibodies would suggest. We don’t know that’s the case, however, so we can’t dismiss the need for a booster at this point.
None of these results is a surprise, and a lot of the data has been hinted at by prior work (the paper itself also sat on a public pre-print server prior to going through peer review). But getting validation that things are working as expected provides an important level of confidence and a clear indication that we have the tools to address some critical questions going forward.