Thread by @udarnik

There are a few people who follow me on here at least in part because of COVID, and for those I want to riff on this finding to explain some of my current attitudes, and in the true spirit of a Cargo Cult Scientist, also explain what data will *make me change my mind*.

https://twitter.com/sailorrooscout/status/1379044284477898760

So @sailorrooscout is an expert, so definitely listen to what they have to say about T cells and other elements of the immune system.

But you will have also, almost definitely, heard right wing pundits, especially a few months ago, espousing the idea that T cell immunity was sufficient for protection from COVID.

This is untrue.

T cells and B cells work in concert in the immune system, and antibodies amplify the effects of T cells, and T cell-derived lineages such as NK cells.

So, am I happy to see that COVID-recognizing T cells are upregulated in casual contacts of COVID patients who did not have symptomatic disease?

Hell yes.

However, I also want to point out that this is one piece of circumstantial evidence for the "threshold model" of infection.

If you get hit with just one viron, you're likely to fight the infection off, even if it manages to infect a cell.

Two virons, even.

But at some point, depending on your immune status and other factors, you reach a quantitative point of exposure at which you *will* get sick.

It's really, really hard to conduct an experiment to prove this, but all we understand about the interaction of viruses with the immune system points to the idea that there's this fuzzy cutoff point, above which exposure leads to disease.

Is it 1000 virons? 10,000? 100,000?

At the moment, it beats our pair of jacks, but in principle, you can "vaccinate" with *just the right amount of virus to elicit an immune response, without causing illness.

The cutoff point is fuzzy. It depends on recent illness, native immune strength, the amount of times the host has seen similar bugs where antibodies might be "cross-reactive" (attach to both types of bug), even on hormonal influences like menstrual status.

In other words, dose selection's a bitch.

This is why we don't use this method in practice, but rather rely on vaccines.

However, we used to rely on this method.

Look up variolation and the *real* story of how Jenner's cowpox discoveries led to vaccination - it's quite a bit more complex and nuanced than the "before Jenner, nothing, after Jenner, smallpox utopia" myth that's taught in general history.

ncbi.nlm.nih.gov/pmc/articles/P…

People used to die from variolation, by the way. That puts "vaccine safety" into a whole new perspective.

Anywho, back to T cells, B cells, and trying to quantify immunity.

So, this Nature Comms study shows a significant "dose response" in the T cell populations they found.

That leaves open a lot of questions. The first, for those right wingers who think evidence of T cell response is enough to protect against COVID, is "when is it not enough"?

Or do they think this is a magic binary, and any T cells at all are fully protective?

If there's a threshold of virons below which you get an immune response but no illness, is there a threshold of Tcells that won't protect you if you get shwacked with enough virons? And how do those thresholds interact in a quantitative model of infectious risk to an individual?

Before I go on to quantitative risk and mask-wearing post vaccination, I want to add a little aside on antibodies / T cell interactions so you can see why experts in this field look askance at certain interpretations of data.

I've been a T cell guy most of my career, only recently having switched to dealing with B cells, so my scientific biases as far as emphasizing the importance of one over the other are a bit blunted. But I will say I know a *lot* more about T cells, even after 4 years in B cells.

There are 2 main groups of T cells, with multiple subgroups, but let's just concentrate on the 2 major groups. There are T helper cells that activate other parts of the immune system. These are command and control cells that overlap with the signal corps of the immune system.

Then there are T killer cells that are the hitmen of the T cell lineage. The killing process is modulated and enhanced by interactions with the stem of antibodies. Antibodies are Y-shaped. The arms bind to foreign invaders, and the stem interacts with receptors on our own cells.

Without antibodies, killer T cells are nowhere near as effective at killing.

nature.com/articles/nri22…

Now, just because you don't have many *detectable* antibodies doesn't mean you don't have any antibodies. Plasma cells can hang out for a long, long time, secreting barely detectable antibodies until they're called upon.

journals.plos.org/ploscompbiol/a…

But whenever someone, especially a non-specialist, comes along touting one part of the immune system over another part, ask some skeptical questions, please. It's a complex system, perhaps the most complex one, after the nervous system.

So back to immunity, masks, and vaccinations. I got invited to a few twitter disputes yesterday, but I was in a mellow mood, and I really practiced my mantra that you don't have to attend every fight you're invited to.

But the spherical cow dipshittery was in full force.

So, let's posit a (rhetorical from my point of view) question, should you wear a mask after you're vaccinated, and for how long?

The answers from me are "yes" and "I don't know, but that doesn't mean forever".

You can transmit the virus after vaccination. Someone I know just tested positive for COVID a month after being vaccinated (a HCW). She was completely asymptomatic.

This is what the vaccine is supposed to accomplish.

BUT.

She was contagious for at least some period of time.

And we can't rely on vaccines to deliver that kind of protection all the time. The current studies, even the Phase III / IV ones ongoing, are in the face of the NPI precautions we're mostly taking.

Remove those NPI interventions, and vaccine efficacy will drop.

How much will it drop?

Now we're back to the quantitative questions of "how many virons?" & "how many T/B cells?" I posed above, complicated by the existence and evolution of variants.

Now throw in the fact that 20 - 30% of the population won't get vaccinated in a timely manner

Getting back to spherical cow nonsense, when I made the assertion that 25 - 30% of the population won't get vaccinated, someone tweeted this at me in rebuttal:

cnn.com/2021/03/06/pol…

My lower bound on that figure fluctuates a bit, but looking at that story, I just had to shake my head. While the number of people changing their mind and getting a vaccine is increasing, the fraction of hard "no" is around 20%.

It's also a *really* good bet that many of those still "waiting for more data" are just hard nos who don't want to admit that to a pollster.

I'm really at a loss for why someone would think that the poll indicates anything but skepticism around the proposition that things will get back to normal, soon.

First, there's the threshold: do you, especially if you're over 60 and staring immune senescence in the face, really feel like gambling on the extrapolation of the vaccine efficacy results from a trial run during heavy NPI use to *your* particular protection in a given situation?

Did the virus fire 6 rounds, or only 5? Do ya feel lucky, punk? Well, do ya?

Hell, I'm in my 50s and I don't feel that lucky.

Then there's the variant issue. With somewhere between 1/5 and 1/3 of the population not getting vaccinated, that's an immense viral evolutionary playground. And, as I explained above, the vaccine isn't a bulletproof vest, it's a sprinkler system.

The virus doesn't bounce off a vaccinated person, they get infected.

If they're not wearing a mask, it's possible that if they're in a superspreader event, they'll get sick. Maybe very sick - especially if variants heading in the direction of escape mutations are present.

But even if they don't get sick, if, to take a rhetorical middle ground, 1 out of every 4 people they meet hasn't been vaccinated, even with a rational estimate that half of refuseniks are naive (meaning the other half have had COVID and have at least some immunity)...

... they are going to be playing a helluva lot of viral ping pong if they're not wearing masks.

And every time they're receiving a serve from that viral ping pong, the vaccinated person's body is selecting for escape mutations.

That poses a threat to both the individual, and to the collective.

Now this is somewhat tempered in my mind by the fact that this is a relatively simple virus, and many of the Variants Of Concern are swirling around a few major deletions / mutations.

There is only so much the spike can mutate and still be effective at latching onto ACE2, which is the cellular crowbar it uses to get inside human cells.

But there are other parts of this virus, and the IFN-evasion story that @VirusesImmunity talks about could certainly hold some nasty surprises, especially in non-spike proteins.

So I'm not chicken-littleing on the variants, but I *am* giving them the side eye.

So, I think I'm finally ready to get to my own risk / benefit calculations around masks.

Right now, there are too many unknowns around vaccine efficacy. If I'm out, unmasked, and there are zillions of virons flying around, am I going to do my own, personal corrective experiment on the reproducibility of the vaccine efficacy trials?

I. Don't. Know.

I'd like to see a little more data on the exact dynamics of infection and the limitations of vaccine protection. I think that's coming in the next 6 months or so.

If it looks good, that's one point in favor of ditching masks.

But the other question is around public health - what's the evolutionary future of the virus, and how quickly will enough people get vaccinated / get the disease to reduce its evolutionary degrees of freedom?

That's going to take more than a few months to work out. We need to keep a close eye on the variants, and if any other mutations / deletions start consistently showing up.

We also need to see how quickly this fourth surge burns through the refuseniks.

My best guess, is that all of this is going to take about 6 - 8 months.

Yes, we'll have to wait that fucking long.

I don't understand why people get pissed off at scientists bearing this message.

Why aren't they angry at the fucking refuseniks who could cut that time in half if they got a fucking shot?

I hoping we can have a maskless Christmas and New Year's, but I think it's iffy. I'm pretty sure we'll have a maskless Valentine's next year.

There is one significant caveat, here. The dynamics of immune response of those infected worry me a bit.

First, we are pretty sure that the worse you get COVID, the better your immune memory. We don't know how long immune memory lasts from the vaccine OR infection.

But this paper made me sit up and take notice:

medrxiv.org/content/10.110…

That paper is one piece of evidence among several, that immunity from catching the disease - even if it's a severe case - isn't as good as vaccinated immunity.

COVID response can be so wild you actually make antibodies to elements of your *own immune response*.

With COVID, the immune system often looks less like a Roman column, locking shields and striking out with discipline, and more like a bunch of berserkers wildly flailing about and doing damage to their own lines as well as the enemy.

So how does that translate into the longevity of immunity? In other words, if the refusenik has already contracted COVID, how long will their immunity last?

We have no clue. And if humanity drew the short straw on that one, we may see reinfections in 6 months or so.

That would turn my calculations above completely on their head.

I've got as many questions as I do even guesses at answers, but I hope that helps the laymen on my TL who are trying to parse the significant noise in the scientific signal as its reaching the lay press.

As always, feel free to ask me questions, and if I don't know the answers, it's more than likely I do know someone who does.