Derek Lowe (In the Pipeline blog in Science Translational Medicine) updated his blog from 2 weeks ago on antibodies, T-cells, immunity and more - and what it all might mean, as well as detailing how little we really know.
The fundamental problem is we have snapshots of antibody levels in state and even country-level serology surveys of close to random populations (i.e., those who had a viral infection and those who didn't know they had one) and we have selected data on T-cell levels active against the virus in small populations (dozens to a few hundred, since measuring these is far more complex than measuring antibodies) of infected and never infected people (from pre-COVID blood donations), but we don't know how these antibody levels might change over time, or how someone's T-cell response might change over time, and (importantly) what a protective profile looks like for both of those.
We've presumed that infected and recovered patients will have immunity from months to a few years, but we don't know that for sure and we've assumed unexposed people have no immunity and that might also be wrong if "cross-reactivity" from exposure to other coronaviruses somehow has given people some or total immunity to the virus from the T-cell part of the immune system. A link and excerpt from his blog is below, The 2nd link below is to a fantastic paper in Cell, which he talks about, which also includes the great graphic, below, showing the two main parts of the immune response (antibodies, which typically work to prevent the virus from latching onto and infecting cells and T-cells, which are special white blood cells which target and destroy virus infected cells) and even this is greatly simplified.
https://blogs.sciencemag.org/pipeli...-on-t-cells-antibody-levels-and-our-ignorance
https://www.cell.com/trends/pharmacological-sciences/pdf/S0165-6147(20)30130-9.pdf?_returnURL=https://linkinghub.elsevier.com/retrieve/pii/S0165614720301309?showall=true
We have similar data here in the US: several surveys of IgG antibodies show single-digit seroconversion. You could conclude that we have large numbers of people who have never been exposed – and indeed, the recent upswing in infections in many regions argues that there are plenty of such people out there. But we need to know more. We could have people who look vulnerable but aren’t – perhaps they show no antibodies, but still have a protective T-cell response. Or we could have people who look like they might be protected, but aren’t – perhaps they showed an antibody response many weeks ago that has now declined, and they don’t have protective levels of T-cells to back them up. Across the population, you can use the limited data we have and our limited understanding of it to argue for a uselessly broad range of outcomes. Things could be better than we thought, or worse, getting better or deteriorating in front of our eyes. We just don’t know, and we have to do better at figuring it out.