Antibodies after Vaccination and Disease

Last time, we built a mental model of the SARS-CoV-2 virus and used that model to make predictions of antibody test results.  This time we will see how that model squares against real-world observations.

Here’s the table we created last time:

	Infected	Uninfected
Vaccinated	S+ N+	S+ N-
Unvaccinated	S+ N+	S- N-

With their permission, I tested more than 40 individuals who fit into one of the four categories.  Here’s what I found:

	Infected	Uninfected
Vaccinated	4 S+ N+ 7 S+ N-	17 S+ N- 3 S+ N+
Unvaccinated	4 S+ N+ 2 S+ N-	4 S- N- 2 S+ N+

Previously Infected and Vaccinated.  There were eleven people in this category, but only 4 had both the S- and N-antibodies that our model predicted.  Surprisingly, nearly two thirds of the people in this group lacked N-antibodies.  This is not what our model predicted.  It seems some people may not form N-antibodies.  Let’s keep looking.    

Previously Infected but Unvaccinated.  There were six people in this category.  Four had both S-antibodies and N-antibodies, but two had only S-antibodies.  Again, we’re missing some of the N-antibodies predicted by our model.  What’s going on here?  I’ll offer some speculations later.  

Vaccinated without Known Infection.  There were twenty people in this category, and all but three of these individuals had the expected S+ N- antibody pattern.  All outliers were S+ N+, suggesting they had asymptomatic infections sometime during the pandemic.  Is this suggestion reasonable?  I think so.   During the pandemic we tested healthy patients before elective surgeries and found a significant number of asymptomatic infections, so we know this can happen.

Unvaccinated without Known Infection.  There were initially five people in this category, and they had neither S nor N antibodies detected in their blood.  Except for one person.  She was surprised to learn she of a silent previous infection based on the finding of both S and N antibodies in her blood.  Subsequent testing of her husband, who also is unvaccinated without previously known infection, found S and N antibodies in his blood too, bringing the total number of people in this group to six with two outliers.  

We can summarize what we’ve learned as follows:

• Both SARS-CoV-2 infection and vaccine stimulate the production of S-antibodies, 100% of the time in this study.
• A significant number of people, about 20% in this group, had silent SARS-CoV-2 infections during the pandemic.
• SARS-CoV-2 infection does not seem to stimulate the production of N-antibodies consistently.  This is a pesky observation that does not fit our model.

Could it be that N-antibody production relates to the severity of disease?  Probably not since quite a few of the S+ N+ individuals in this study had asymptomatic infections.  Could it be that variant viruses cause N-antibody negative infections?  Or is the N-antibody test not very good?  All are possible, but, as we’ve said so many times since the outbreak of the pandemic, we really don’t know for sure.   What we can say is that tests for “COVID antibodies” are more complicated than they seem at first glance.  Laboratories should clearly label the antibodies measured when reporting SARS-CoV-2 antibody results.

How do the antibody levels caused by disease compare with the antibody levels caused by vaccine?  And which vaccine provides the best immune response?  We will examine these questions next time.