A long-lived antibody response requires that B-cells become activated in lymph nodes, develop first into plasmablasts, and then finally into plasma cells, which live in the bone marrow rather than the bloodstream or lymph nodes. Plasma cells can live for many years, and are specialized to churn out a stream of antibodies. Fun fact of the day: a single plasma cell can secrete up to 20,000 antibodies per second. All of the antibodies that come from a single plasma cell are identical (monoclonal), but normally you have multiple plasma cells in your bone marrow that make antibodies that target different parts of a given pathogen, so your total antibodies to that pathogen are said to be polyclonal.
Their location in the bone marrow makes plasma cells difficult to study. But a small number of generous people volunteered to have a needle biopsy of their bone marrow at 7 and 11 months post-covid so that their plasma cells could be studied.
This study also involved getting a larger number of people to give blood at intervals over the first 11 months after infection, so that their antibodies can be traced.
As controls, the researchers also looked for antibodies and plasma cells to influenza and plasma cells to a mix of tetanus and diptheria. Most people get five doses of tetanus and diptheria vaccine during childhood, and some get boosters in adulthood, because the immunity to these tends to wear off faster than immunity to other things. Influenza plasma cells tend to live longer than those to tetanus and diptheria.
For the blood samples, the greatest antibody decline was in the first three or four months. After that, levels were still decreasing, but at a slower rate, and it looks like some people are leveling out between 6-11 months. That's something that's been seen before, and it's a good sign that there are plasma cells in the bone marrow making antibodies.
18 people agreed to donate bone marrow to the study at 7 months, and 5 agreed to donate again at the 11-month mark. As expected, all of the bone marrow had lots of anti-flu plasma cells. There were fewer plasma cells for tetanus/diptheria than for flu, and a few people had none.
At 7 months post-symptom-onset, 12 out of 18 people had some anti-spike plasma cells in their bone marrow. The levels varied a lot, but on average they were lower than those for tetanus+diptheria.
At 11 months post-symptom-onset, 5 out of 6 people had some anti-spike plasma cells in their bone marrow. Five of the six had bone marrow taken at both 7 and 11 months. Out of those five, one had none at both time points (so probably no lasting antibody response). One completely lost IgA-secreting plasma cells between 7 and 11 months. The other four had increasing levels of plasma cells between 7 and 11 months. So although this data is limited, it appears that a little more than half are making a long-lived antibody response after covid. We don't know for sure that the levels of antibodies produced will be enough, especially for the variants. But this is in general good news, because some people have been worried that because cold-type coronaviruses tend not to produce lasting antibodies, we wouldn't make lasting antibodies to SARS CoV-2.
They also found that people with more plasma cells tended to have more antibodies, which makes sense, and it means that antibody levels long-term give a rough idea of one's plasma cell status.
Link to study: https://www.nature.com/articles/s41586-021-03647-4_reference.pdf