Thread by @michaelzlin

Going to wrap up 2021 with an optimistic thread.

If we can avoid creating worse variants with molnupiravir, COVID19 can finally become like the flu in 2022.

1/n

The SA experience, mirrored in Western countries so far, suggests Omicron IFR is an order of magnitude lower than previous variants. This is due to preexisting immunity and lower virulence of Omicron.

2/n
nytimes.com/2021/12/30/wor…

Specifically, deaths per counted case in SA are 16% of the Delta wave so far, but cases are almost certainly undercounted several-fold.

3/n

https://twitter.com/michaelzlin/status/1476247343297146890

Many cases were found in SA hospitals by routine screening of asymptomatic people, and the same is true in the US, and we're see ~4x higher viral RNA in wastewater than the Delta peak, compared to 1.3x by official testing.

4/n

https://twitter.com/michaelzlin/status/1476762634040922112

`

This suggests actual infection fatality rate (IFR) of Omicron is probably 5-10% as a conservative guesstimate of Delta. If Delta has a IFR of 1.5% in the US, then we're at a IFR of 0.075-0.15.

An average seasonal flu has an IFR of 0.05-0.1 in the US.

5/n

Flu has an IFR of <0.1% because it doesn't usually cause viral pneumonia in health younger adults, and most have partial immunity. Flu deaths are in older and immunocompromised, like COVID19.

See slide 8 in the "coronadeck" (rest of deck outdated)

6/n

https://twitter.com/michaelzlin/status/1239675364172914688

Yes SA is younger than Western countries, but the preexisting immunity levels were similar, and the ~10x IFR reduction was seen across ages in SA, as far as we can see, so the same 10x should apply to any population regardless of age distribution.

7/n

The evidence for lower virulence of Omicron continues to build, with recent studies finding less lung pathology in multiple animal models. People aren't small animals, but these particular models were chosen for being very susceptible to disease.
nytimes.com/2021/12/31/hea…

8/n

So what are the conditions necessary to get us to a flu-like state with COVID19? We need >90% with immunity to protect them from death from seasonal SARSCoV2 waves. We will have that, if not 100%, by early 2022.

9/n

A high level of immunity also keeps infection rates low so that the chance of getting COVID19 in each year may drop to<50% even without masks, like the flu.

10/n

We then need annual boosters, both so that we can maintain this state of low endemicity, and so that when we do catch SARSCoV2 eventually, our immune system will be ready to clear SARSCoV2 rather than give it a chance to cause pneumonia.

11/n

Finally, we need the virus not to make big jumps in evolution. Naturally coronaviruses mutate in people at a rate of 2 nucleotides per transmission event, and those events happen 4 days apart on average, and then people clear the virus.

12/n

This is a low enough rate that we are unlikely to get major jumps in immunoevasiveness each year (would be more akin to flu antigenic drift than the large antigenic shifts we fear). This allows existing immunity to remain effective and any breakthroughs to be mild.

13/n

Omicron was an exception in having enough mutations in the spike receptor binding domain (16) to evade most prior immunity. Why it happened is a mystery; could have been a long infection in an immunocompromised person.

14/n

We got lucky Omicron happened to be less virulent too. To prevent future Omicron-like antigenic shifts that are also more virulent and can outpace a humoral reactivation and T cell organ protection, we need to do two things in the future.

15/n

Obviously, we need to vaccinate the world to reduce the ability of the virus to pick up a large number of mutations in immunocompromised hosts, or even just a smaller number of mutations in healthy hosts which will still be bothersome.

16/n

This will make antigenic "shifts" rare enough, and the spread of any shifted virus slow enough, to hopefully buy us time to adapt our vaccines. Companies can then have a schedule of forecasting and adapting their vaccines ahead of time for each fall, exactly as with flu

17/n

The 2nd thing we need to do is severely limit molnupiravir (MOV) use. MOV accelerates viral mutagenesis by 5x (Merck's data) and doesn't reduce viral levels for the first 2 days. So in that time patients on MOV are breeding grounds for potential antigenic shifts

18/n

If the potential for molnupiravir to create worse variants is a surprise, it's because the risk has been overlooked in the rush to have something against Omicron. Ironically MOV is not very effective at all, and Omicron is milder. More info here

19/n

https://twitter.com/michaelzlin/status/1475518939413184518

So does it mean that Omicron is a good thing to get and nothing to worry about? No. Just as you wouldn't go out and catch the flu, I can't imagine anyone wanting to go out and get Omicron. It will still set you back a week, and then do the same for your close contacts.

20/n

And just as you wouldn't go visit grandma with a bad flu, you should avoid seeing people with weaker immune systems if you have Omicron or may have it after being exposed. You should thus avoid those New Year parties unless you like getting sick or getting others sick.

21/n

Now the next variant to create a wave may not have the inherently milder severity of Omicron. It could be Omicron regains some lung infection ability, or the next variant is not derived from Omicron. But as long as it's not both antigenically shifted and infectious...

22/n

... then in high-immunity countries our preexisting immunity will limit severe outcomes. Hence it could be like a regular seasonal flu most years, and how we think about the risk of antigenic shift and a particularly bad epidemic could be similar to the risk of shifted flu.

23/n

So that's it. I'm hoping next year we can start making COVID19 like flu: get an annual booster, tell your friends and family to do the same, stay home if you're sick.

But we need to vaccinate the world, and we need to avoid molnupiravir.

Wishing everyone a healthier 2022.

n/n