The best outcome is that you get even stronger immunity. That seems to be what happens when people who received the oral (Sabin) polio vaccine were then given the injectable (Salk) form. The first is an attenuated live virus, and the second is a completely inactivated one. The Salk vaccine is better at producing humoral immunity (antibody and T-cell response), and the Sabin vaccine needs multiple doses to be effective. But it is better at producing mucosal immunity in the gut, which has a better chance of interrupting the spread of the disease in children. The choice about which one to use has always been a matter of argument. But the study linked above showed that in children who had already had the oral Sabin vaccine, that an injection of the Salk vaccine boosted their intestinal immunity better than another round of the oral vaccine. Again, you wouldn’t necessarily have predicted that – if it had come out that the injected dose didn’t seem to do much for mucosal immunity, it would have been easy to rationalize that as well.
There are other cases where multiple vaccines are available for the same pathogen, and where a mix-and-match approach doesn’t seem to make a difference either way. An example is hepatitis A, where there are several inactivated-virus options. In that case, it appears that the vaccines are basically interchangeable: the booster-shot schedule can be completed any way you like. The same goes for the two monovalent vaccines for hepatitis B, and for the three vaccines that target meningococcus group A, C, W, and Y. (Here’s an overview of vaccine interchangeability).
That said, all of those vaccines in each of those cases are rather similar to each other, and we now have the unusual – very, very unusual – situation of several different vaccine platforms coming into potential use against the same virus at almost the same time. By the spring we may well have two mRNA vaccines (Pfizer/BioNTech and Moderna), two different adenovirus vaccines (Oxford/AZ and J&J), and a recombinant protein vaccine (Novavax). We don’t have efficacy data on the J&J and Novavax candidates yet (numbers are on the way), and we can argue about the data for Oxford/AZ, but it’s certainly possible that all of them will be out there simultaneously. Putting one of these on top of the other is a step into the unknown.
And there are examples of vaccines for the same pathogen having some interference. Several vaccines for bacterial diseases are in the “conjugate vaccine” category: they have a bacterial polysaccharide fused to a carrier protein, which can give a more useful immune response than just dosing the polysaccharide by itself. But for pneumococcal vaccines, both types are given (with a different range of immune response to cover a variety of bacterial serotypes). It’s been found that if you give a pneumococcal polysaccharide vaccine (PPSV) followed by a pneumococcal polysaccharide conjugate vaccine (PCV), there’s a lower antibody responses for some serotypes targeted by the conjugate vaccine than there is if you give them in the opposite order. So the rule in this area is to give both for maximum protection, but to always give the conjugate vaccine first. Another tricky part is that the use of the same sorts of carrier proteins in different vaccines – you could imagine a situation where an immune response against the carrier protein causes a later vaccine to be less effective.
That last problem is similar to what we’re talking about with the immune response to adenovirus vectors and booster-shot dosing regimens with the same vaccine. But the Oxford/AZ vaccine is a chimpanzee adenovirus and the J&J one is Ad26, so that’s a different situation, and I have no idea of what would happen if you mixed those two. (The Russian vaccine is, in fact, a mixture of two different adenovirus vectors, one in the original shot and one in the booster). I also don’t know what happens if you take both an mRNA vaccine and one of the other types.
Overall, though, I would tend to think that it would work out. All of the coronavirus vaccines we’re talking about target the Spike protein, after all, and they are, by different means, presumably raising a pretty similar suite of antibodies (with perhaps more differences in T-cell response, which remains to be seen in detail). So the chances are that the immune response will be similar (as with the hepatitis vaccines) or perhaps even a bit better (as with mixing the polio vaccines), rather than worse. But we haven’t proven anything like that in the clinic yet, and educated guesses will only take you so far. I would assume that there will be people who end up taking both types, for all sorts of reasons, and I hope that we collect as much data from those cases as we can.
