That’s a question that’s on a lot of people’s minds, because frankly, we have answers that seem to disagree with each other. Let’s have a look:
For one, there’s this letter in the NEJM from researchers at Beth Israel Deaconess hospital and from Janssen (J&J’s pharma branch). That’s a look at immunity in 20 patients who had the single dose (and in a few cases, two doses), versus controls, and it compares serum from these patients against the original strain along with the Alpha, Beta, Gamma, Delta, Epsilon, and Kappa variants, with many of these tested over time to see how the neutralizing antibody levels might have changed over a period of months.
In general, the peak responses were at the Day 71 blood draw, but these had not gone down that much by Day 239 – and still not far off of the values on Day 29, which is good to see, considering that you can already see protection showing up in the clinical trial data even earlier than that. As in other studies with other vaccines, the Beta variant (B.1.351) showed the biggest effects in lower neutralizing antibody titers (most possibility of evading vaccine protection), but that showed an interesting patter. At Day 29, the neutralizing antibody levels were 13 times lower against Beta than they were for the parent strain, but by Day 239, they were only 3-fold lower. This strongly suggests that the antibodies being produced are in fact expanding and evolving in “germinal centers” in the lymph nodes over time (which lead to longer-lived plasma cells in the bone marrow) which is very good to see. That link goes into more details, but basically a good vaccination keeps on going, stimulating your immune system to keep on producing new antibody variants and expanding production of the ones that seem to be doing the best job. We’ve recently had direct evidence that the mRNA vaccines produce long-lived germinal centers, which is excellent news in a landscape that features new variants coming on.
Comparing the variants head-to-head, the single-shot patients at day 239 had a median pseudovirus neutralizing antibody titer of 184 against the parent strain (two-shot patients were at 192). Against Alpha, this titer was median 147, Beta 62 (there’s that threefold drop), Gamma 129, Delta 107, Epsilon 87, and Kappa 171. So as has been said before, while the more infectious delta form is somewhat less neutralized than the original coronavirus is, the spread of the Delta strain (B.1.l617.2) is at least crowding out some variants (like Beta) that are even better at getting past the antibodies. Whether it’s possible to have a more infectious strain that’s as vaccine-antibody-evading as Beta (or worse) is as yet unknown. There might be no amino acid variations possible for the virus to hit on that trick, or it might have done so this morning somewhere. We simply don’t know.
So this work shows that the J&J vaccine still seems to be protective against Delta, albeit at lower levels (but not so low as to break through constantly). We need to contrast those results, though, with this new preprint that came out two days ago from a team at NYU. It’s very similar in its methods: pseudoviruses containing the business end (the Spike protein) of the different coronavirus variants are evaluated versus serum from recovered coronavirus patients as well as vaccinated patients to see what the neutralizing antibody titers are. And this one compares the J&J values to those from people vaccinated with the Moderna and Pfizer/BioNTech mRNA vaccines as well. The actual J&J titer numbers I’m about to show should not be directly compared with the ones just mentioned above, but the trends can be.
I’ve assembled the NYU data into one table below. The convalescent numbers are the mean values from 8 different patients, the Pfizer/BioNTech ones are a mean of 9 patients, the Moderna ones are a mean of 8 patients, and the J&J ones are a mean of 10 patients. The standard deviations on these numbers are fairly large, since different patients responded differently. But even this sort of spread is definitely not large enough to wipe out the larger differences that we’re seeing between vaccines (and versus convalescent patients who went through the immunity process the all-natural way). The convalescent data were collected somewhat earlier than the vaccine sera, as you’ll note, and we saw from that NEJM paper that the J&J antibody titers (in their samples, anyway) peaked at their Day 71 sample. So it’s possible that the convalescent data might look a bit better at a later time point, but they’re surely still not going to hit the levels seen in the mRNA-vaccinated patients.
But these numbers do not make the J&J antibody levels look good at all. They’re worse than convalescent patients, and as mentioned, these samples were presumably collected around the peak of the J&J levels if we can draw that conclusion from the earlier NEJM paper. I am sure that the appearance of this preprint is causing consternation at J&J, and I look forward to any statements that they might make on it. So far, their comment seems to be that these new numbers “do not speak to the full nature of immune protection“. That same New York Times article linked also quotes an author of the NEJM paper (from Beth Israel Deaconess) saying that he doesn’t feel that the numbers really look that different, but I can’t really see that. He also notes that the new preprint does not touch on T-cell numbers (the NEJM work showed what appear to be durable CD8+ cell numbers). That’s a fair point, and T-cell data always lag antibody data because the numbers are so much harder to obtain (which means that we have only a fuzzy idea of the importance of T-cells as a part of the total immune response to the coronavirus, although they must surely be important.
But even taking all this into account, these NYU numbers, if they are indeed what they seem, would indicate that the single-dose J&J vaccination might well have a greater chance of losing efficacy against variant strains. The huge majority of people being hospitalized with Delta infections here in the US are not vaccinated at all, as we’ve all been hearing, and that just makes you want to beat your head against the wall in general, because we are, horribly, well stocked with vaccine doses that are going unused while the people who are passing them up are getting sick. But we also should be collecting data on the far smaller cohort who have shown serious disease after having been vaccinated, and see if there’s a trend for single-shot J&J patients to stand out in that group.
I would certainly think that these data are being collected – if they’re not, we have still more problems – and I hope that we can get a read on this. It’s the actual patient numbers that are the most important here, and that’s what we’re currently lacking. . .