In a bit of a surprise move, Pfizer and their partner BioNTech announced yesterday that they were moving their BNT162b2 mRNA vaccine candidate forward into Phase II/III trials. The surprise was because all the publications from this effort so far had been on another one of their four candidates, BNT162b1.
Fans keeping score at home will know that there were originally four candidates: two with modified RNA bases, one with extra uridines, and one self-amplifying RNA. From that press release linked above, it appears that the latter two fell behind during the preclinical and early clinical studies – we haven’t seen that data, but apparently BNT162b1 and BNT162b2 have been the front-runners for some time as far as the two companies were concerned. (In retrospect, maybe it’s a good thing that I never got around to doing a detailed post on the self-amplifying RNA mechanism! Update: OK, with the Imperial College vaccine doing human trials, I’ll go ahead and do that post anyway. . .) So what’s the difference between the two remaining ones?
It comes down to the antigen(s) being coded for. The b1 candidate, the one we’ve been hearing about, codes for the coronavirus Spike protein’s receptor-binding domain (RBD), and this was constructed as a trimer, three RBDs attached to a “foldon” protein core. Meanwhile, the b2 candidate codes for what they say is an “optimized full-length Spike” protein instead, not just the receptor-binding domain. Pfizer’s press release says that both the b1 and b2 candidates “induced favorable viral antigen specific CD4+ and CD8+T cell responses, high levels of neutralizing antibody in various animal species, and beneficial protective effects in a primate SARS-CoV-2 challenge model“. But they made the choice for the b2 variety partly because it seemed to be better tolerated on injection, and also because it led to a wider variety of T-cell responses. These include both CD4+ and CD8+ T-cells, and these were raised not only to recognize the RBD region, but also other regions of the Spike protein that weren’t contained at all in the b1 candidate. And they’re quite right – that could well be beneficial, and the better tolerability is a bonus. The release says that the neutralizing antibody response was similar between the two candidates.
I look forward to seeing the numbers on that – in fact, I look forward to seeing the numbers on everything. That’s going to be really important as we move forward, because any hint or appearance of secrecy will be toxic here. We already have enough people who are (1) suspicious of vaccines, (2) suspicious of what various authorities say about coronavirus treatments, and (3) suspicious of the motives of the pharmaceutical industry. I might add a fourth group, who are suspicious of the motives of the Trump administration, particular as regards getting a vaccine announcement out in a timely manner before the November election. It’s a mess, an awful mess, and we do not need to make it any worse by giving one or more of these groups more ammunition.
The only solution (as far as I can see) is full disclosure. Let’s have all the clinical data for every candidate and have the decision-making process be as open as it can be. We have a lot of decisions to make (it’s going to take a whole other post to get into that), and it’s going to be way too easy to screw that process up and make things worse than they are already. That actually seems to have been the hot happenin’ trend here in the US for 2020; I’d be very happy if we could manage not to do it for the vaccine selection and rollout. . .