Thread by @NathanGrubaugh

Some further musings on our work with @carolilucas, @VogelsChantal, @InciYildirim11, @SaadOmer3,
& @VirusesImmunity. I'll start by nerding 🤓 out on some of the genetics and end with some thoughts 🤔 about vaccination strategies (/gasp). (1/22)

https://twitter.com/VirusesImmunity/status/1416928709030907904?s=20

One of the 😎 parts of our study is that we used virus isolates (not pseudovirsues) that represent much of the genetic diversity in our region. This allowed us to examine local effects and to dive into the genetic components of 💉neutralization (2/22)

https://twitter.com/VirusesImmunity/status/1416928742664986634?s=20

These are the results that I want to spend some time with as there is a lot to unpack here. I know that I am a bit biased, but this is such a 😎 figure! (3/22)

https://twitter.com/VirusesImmunity/status/1416928756602654726?s=20

The spike L452R and E484K mutations have received the most attention as they are associated with many VOC/VOIs and are known to impact neutralization from some antibodies (4/22)

From our panel of viruses, we estimate that L452R and E484K have the greatest negative impact on neutralization following 2x mRNA vaccination 💉. To my surprise, when considering the mutations independently, L452R > E484K. But of course its more complicated than that. (5/22)

Here the lineages are stacked by least to most reductions on neutralization titers. Even though we estimate that L452R has the single most impact on neutralization, lineages with this mutation are intermediate. This includes Delta (B.1.617.2). (6/22)

Lineages with the E484K mutation are found at the bottom, but also in the middle. This includes one (B.1.526+E484K) that didn't significantly reduce neutralization. So just having a L452R or E484K mutation is not enough to push a virus towards the lower end. (7/22)

Quick aside: I'd like to discuss B.1.526 (Iota) as this lineage is complex. 2 different clades contain SN77N (a+b in neut fig 👆, orange 👇), 1 clade with E484K (c; yellow), & 1 clade with L452R (d; blue). Lineages + E484K were the most closely tracked by public health. (8/22)

With the power of hindsight, I don't really think that B.1.526+E484K was particularly important compared to the other 526s. From the freq plot 👆, the other versions were stably transmitted along side it. Also, we found that B.1.526+L452R (d) had greater impact on neut. (9/22)

The astute observers probably already saw me slow-play what the 5 isolates with greatest impact on neutralization have in common: E484K + N501Y/T. While I just spent several tweets explaining that we can't over-simply things, this combination is hard to deny. (10/22)

Adding E484K to B.1.1.7 (already with N501Y) moves this lineage from one extreme to the other in neutralization. While this isn't a particularly new finding, it's interesting to see this jump in the context of our large panel of isolates. (11/22)

Finally, we even included a B.1 lineage with E484K + N501T that has yet to be classified into a specific sub-lineage, let alone as a VOC/VOI, and it too is at the lower end of neutralization. The combination of these two sites appear to be quite important in this context. (12/22)

Thus my team will start tracking all viruses with E484K + N501Y/T in Connecticut and report them to the @CDPH, regardless of lineage. Locally, this combination never exceeded 19%, and is now being pushed out by Delta, but it's still important to watch. (13/22)

Another aside: What about L452R + E484K? A search on @DiseaseOutbreak shows that this combination has only been found in 63 of the >2.4 million sequenced viruses. So my guess is that they are antagonistic in terms of virus fitness, and isn't a major worry at this point. (14/22)

So what does this all mean for levels of protection and vaccination strategies? Our preprint, and subsequent twitter threads, have initiated a lot of discussion/speculation on this front. (15/22)

https://twitter.com/EricTopol/status/1416917815932579841?s=20

First, our work does NOT imply that natural infection is better than vaccination in terms of neutralization, as some commenters have suggested. And we are certainly not suggesting that people try to get infected as a way to boost their NAbs. (16/22)

What we are saying is that IF NAb titers are directly correlated with protection (which there is probably more to it than that), than a boosted immune response could help increase protection against the variants on the far end of the spectrum (e.g. Beta, Gamma). (17/22)

But real-world data still shows high efficacy to all variants from 2 doses of the mRNA vaccines. This is because, as suggested below, some reduction in neutralization is probably ok. (18/22)

https://twitter.com/gadboit/status/1417021962057043968?s=20

In my mind the problem is that we don't yet know what NAb titers are still effective at (1) preventing infection and (2) preventing disease.

Beta and Gamma may be already pushing those limits for some vaccines. (19/22)

Also, the mRNA vaccines (and most others) are very effective against Delta, which is or will be dominant in most locations. We find that it leads to a low/mid reduction in neut. If that is the most important correlate, then immediate boosters are not likely needed for it. (20/22)

Finally, the **host has a far greater impact on neutralization than the variant**. Individuals with high neutralization to one typically have high neutralization to others, and vice versa. We have saw 2 that didn't produce NAbs to any of the tested viruses. (21/22)

Recap:
- E484K + N501Y/T 📉📉 neutralization
- Delta low/moderate 📉 neutralization
- Boosting *may* help with Beta, Gamma
- Host >> variant
- Get VACCINATED!

Our study does not say if the CDC/FDA or Pfizer are correct regarding boosters. Way more goes into those decisions. /fin

**Caveat** There are others FAR more knowledgable about vaccines than I am. Please follow our partners @InciYildirim11, @SaadOmer3, & @VirusesImmunity for more coherent thoughts.