Now the SARS-CoV-2 variants that were first identified in the UK (B.1.1.7), South Africa (B.1.351), and Brazil (P.1) have been found in the US, a couple of thoughts (i.e., speculation...) on what happens next, as I been getting many questions about transmission and immunity.

Summary: I expect B.1.1.7 to rise rapidly in frequency in the U.S. over the next few weeks/months, but don't necessarily expect the same for B.1.351 / P.1 in most areas.

Despite this, all lineages require urgent and immediate action. Dense to follow

It is now firmly established B.1.1.7 is inherently more transmissible (~50%; doubling in relative frequency ~ weekly) - we have observed this in the UK, Denmark, Ireland, Portugal, Jordan, and other places. When a country finds this lineage - assume it'll become dominant quickly.

B.1.351 and P.1 are independent lineages, but they have several key mutations in common - L18F, K417N/T, E484K, and N501Y (the latter also shared with B.1.1.7). Seeing the same mutations pop up in multiple locations (convergent evolution) means there's selection.

While B.1.351 / P.1 have been growing rapidly in locations they were first identified, they have not yet been seen to spread rapidly outside those areas. While they are harder to identify (sequencing), it's possible this is because they are not inherently more transmissible.

Why are B.1.351 / P.1 then growing rapidly in S. Africa / Brazil? Likely because they are capable of causing a larger fraction of reinfections in previously infected individuals - because they can escape _some_ level of (B-cell mediated) immunity.

The Eastern Cape of S. Africa and Manaus in Brazil where B.1.351 / P.1 are believed to have emerged, were particularly hard hit in earlier COVID-19 waves - Manaus at an estimated 75% attack rate - yet cases are once again rapidly rising.

These findings - together with recent in vitro data and data from clinical trials of (very good, yay!) vaccines - suggest that high levels of standing immunity may help drive B.1.351 / P.1 rapidly to high frequency.

Importantly, the rapid increase in B.1.351 / P.1 may therefore be tightly linked to the level of immunity in a population - whether acquired naturally or via vaccines. If the level of population immunity is low, I am not convinced we'll see an immediate rise of these variants.

Which brings me back to B.1.1.7 - which we _know_ will rise rapidly in frequency in the coming weeks and months in the U.S. It is an extremely urgent situation that requires decisive action as we'll see effective reproduction numbers increase by ~50% as B.1.1.7 becomes dominant.

Should we expect B.1.351 / P.1 (and similar lineages) to rise rapidly in frequency in the U.S. immediately as well? Possibly, but I'm not convinced we will - since I'm not convinced (by current data - which, of course, will change) that they're inherently more transmissible.

However, I'm concerned this will change as we (necessarily) build more immunity in the population (via critically important vaccination) - B.1.351, P.1, and other lineages picking up the same / similar key mutations, will then have a fitness advantage and rise in frequency.

Most importantly - and this is key - vaccination is ABSOLUTELY CRITICAL to lower overall morbidity and mortality of ALL SARS-CoV-2 variants - including B.1.1.7, B.1.351, and P.1.

Despite B.1.351 / P.1 likely gaining a fitness advantage as more people become immune, the solution obviously isn't to stop or slow vaccination - quite the contrary, it's to accelerate vaccination. Why? Because the vaccines are still effective against them.

Likely not as effective as they are against non-B.1.351/P.1 lineages, but still effective. So the more vaccines we can get into the arms of people, the fewer numbers of overall infections (including those from B.1.351/P.1) we will have by the time they start to rise, the better.

However, immune evasion from B.1.351, P.1, and similar lineages represents an urgent and critical emergency - a massive wake up call! Why? Because while the vaccines will still be effective, they will not be effective enough and need to be updated.

We therefore need to figure out our whole strategy for how to update the vaccines - this is not a 6 week process as has been described many times. That's just the first part in the lab. We need to follow that by experiments, testing, trials, approvals, production, rollouts, etc.

The regulatory process also needs to be in place - what will be required for approval? How are we going to administer the updated booster? How many updated vaccine targets do we need? Are we going to combine multiple targets in one booster, or in multiple boosters?

More fundamentally, we need to know how B.1.351, P.1, and the yet-to-be-discovered SARS-CoV-2 variants like them can (partly) escape (B-cell mediated) immunity. Do they disrupt a key epitope? Or are they generally more resistant to Ab neutralization?

It's critically important we know the answers to these questions - and quickly. Updating a vaccine because a key epitope was disrupted might be relatively easy - but if these lineages are generally more resistant to neutralization (I suspect they are), that's a different story.

So to summarize - I expect we'll see B.1.1.7 rise rapidly in the coming weeks and months, but I don't necessarily expect the same to be true for B.1.351 and P.1 - at least not all across the U.S. (in places with high attack rates, we might / likely will).

However, all lineages require urgent and decisive action - now! Not when they have become prevalent. Those actions, however, will need to be tailored - one lineage will rise rapidly because it's more transmissible, the others as immunity is building up.

So let's get our genomic surveillance in place, better masks, more masks, MUCH more widespread testing and screening - and avoid gatherings and crowded settings. Combine that with accelerated vaccine distribution and an accelerated plan for updated boosters.

Tech speculation 1 - What drove the fixation of mutations in B.1.1.7?

I suspect intrahost selection in a (likely) chronically infected (likely) human that may or may not have been immunocompromised and may or may not have received treatment. Likely a very rare event.

Tech speculation 2 - What drove the fixation of mutations in B.1.351 / P.1?

I suspect intrahost selection with immune evasion being the primary driver. I think this is more likely than selection due to population immunity (creates a separate pressure leading to pop fixation).

Tech speculation 3 - How many co-circulating lineages should we expect for updated vaccines?
Are L18F, K417N/T, and E484K the tip of the iceberg, or do they represent most of what we'll see? Impossible to know, but we should prepare for the former.

You made it this far? Seriously? Here's a corgi.

One key caveat I should add - when I say I'm (currently) not convinced B.1.351 / P.1 are inherently more transmissible, it's because the data to address that question simply do not exist. We know they spread better in the countries first identified [...]

1/

[...] and we have data on immune evasion - which can explain the increased spread. However, we can't isolate the question of "inherent" transmissibility from immune evasion - until we have more data outside Brazil and S. Africa. I suspect we'll first get those from Denmark.

2/

So while we know B.1.351 / P.1 can evade _some_ immunity, whether they're inherently more transmissible as well, remains unknown for now. I _suspect_ they may not be (because immune evasion can explain what we observe in Brazil / S. Africa), but we'll have to wait for data.

fin/

A slightly updated version of my first slide with a few edits. Also, since the names come up often, all credit goes to @viralverity and @AineToole for coming up with such good ones. Yes, nerdy.