UKH

This article explores the intriguing suggestion that the virus burns out after around 20% of a population have been infected:

https://theconversation.com/coronavirus-could-it-be-burning-out-after-20-of...

Possible explanations for this include the existence of ‘immunological dark matter’, which turns out to be immunity conferred from exposure to other corona viruses. 
 

Going back to swine flu, every one else in the house was knocked out for a week or more with fever and I had a hot flush and a bit of a headache for an afternoon.  This makes me think that there could well be something to what you are saying.  There is a lot of unknown stuff out there that we have not been monitoring and who knows whether there have been CVs in the past that either only infected some people or had such mild or diverse symptoms that they were not picked up in the past decades.  I believe in the possibility of this, not that it is fact as we simply don't yet have enough knowledge.

We exist in an ocean of viruses, always have and always will.

I read something yesterday that added to the weight of evidence that T cells deal with the virus in many cases and confer immunity. The real rate of cases might be 3x higher than currently estimated. It’s certainly weird that if only 5 or 10 Percent of the population has been infected, all these demonstrations etc haven’t led to a big jump in cases. 

I'll guess we'll find out once the pubs re-open! 

I agree there may be something in the arguments but we don't know from the science and I disagree at that level... as NY had around that level infected with a large proportion of their population isolating and social distancing. It's the same problem as the Oxford study you can't claim 0.1% mortality when more than that died in NY and most there will have not been exposed.

https://news.ki.se/immunity-to-covid-19-is-probably-higher-than-tests-have-...

If this is the case, then why do 57% of the population of Bergamo have antibody evidence of infection?

I think we have to be very careful of premature optimism, and simply be very patient to see how it all pans out. I'm still in total lockdown, in a 'bubble' with my partner.

The article mentions that infections on the Diamond Princess plateaued at 20%.

Infections plateaued on the Diamond Princess at 20% on the 24th February.

The last passengers disembarked on 24th February.

https://en.wikipedia.org/wiki/COVID-19_pandemic_on_Diamond_Princess#/media/...

The article mentions places that have gone much higher than 20% so the article itself might contain enough information to undermine the 20% idea.

NZ has plateaued at 0.024% (1200 cases, 5 million people) , I doubt they are 99.976% "immunological dark matter" or have 800 times less immunological vulnerability (0.024% rather than 20%). I think public policy of "go early and go hard" and suppression seems more probable difference between countries 

There is much unknown. Anything is possible. Not everything is probable.

It looks as if it fits very well with the evidence.

Providing you cherry-pick the evidence and make excuses for all the exceptions, that is...

"If it seems to good to be true..."

That said, who knows? Almost everything about this pandemic is clouded by uncertainty; it'd be great if it was true.

Off to buy a lottery ticket now.

Total nonsense... 

True, Bergamo doesn't fit. These kinds of exception are addressed in the article.

They do present counter-examples in the article so they can hardly be accused of cherry-picking

That's one thing we can say with certainty!

Which bits are nonsense? There is hard data in here that seems to be sound, so not total nonsense...

I read an abstract to a paper that proposed 43% based on the accepted model but refining the variables based on recorded data.

Probably for exposure up to a certain viral load and not beyond.

They were outdoors and viral load is generally down so I don’t find this weird.

I do find it a reason to fear the winter if England keeps at a plateau in active cases and doesn’t hammer it to zero soon.

60% infections, 25% pensioners. 

At least 35/75 of the non-pensioners must be infected if we make the implausible* assumption all pensioners infected. 

35/75 or 47% of non-pensioners infected - not good support for 20% idea.

*Covid infection rates are pretty uniform across all age groups but Covid symptoms obviously get massively worse with age

"There is no evidence of differences in the proportions of individuals testing positive for COVID-19 between different age categories", see URL below.

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/con...

But they make excuses for all of them, some of which don't look very likely.

It looks a lot like cherry-picking the evidence to fit the desired conclusion.

That doesn't necessarily mean the conclusion is wrong, but it isn't a very compelling case for it - especially given the clusters where it has spread rapidly to a large number of those present - e.g. German meat factories, various places of religious worship.

Alternative hypothesis, plucked out of the air: lockdown measures work quite well, but about 20% of the population can't or won't stick to lockdown, or are closely associated with people who can't.

"That said, the fact that more than 20% of people have been infected in other places means that the T-cell hypothesis is unlikely to be the sole explanation either. Indeed, if a 20% threshold does exist, it applies to only some communities, depending on interactions between many genetic, immunological, behavioural and environmental factors, as well as the prevalence of pre-existing diseases."

(My italics)

They're not really cherry-picking, just highlighting an anomaly. They hedge the hypothesis about with caveats all the way, and appear well aware of the many confounding variables. 

There's more on the 'dark matter' hypothesis etc. here:

https://www.theguardian.com/world/2020/jun/07/immunological-dark-matter-doe...

Yes, I haven't read it all properly yet, but Grifoni Cell paper the Guardian cites looks interesting and sensible.  Not sure why that should be a surprise, given the journal it's in, but boy, have I read some crap about this recently! 

Let's hope so.

There still seems to be a lot of what I would consider fundamental information, we have no idea about.

Maybe you are right to be surprised... 😁

https://www.theguardian.com/commentisfree/2013/dec/09/how-journals-nature-s...

The day that I lose my optimism will be a very grim day!

I would love to be a fly on the wall at a few SAGE meetings. I do wonder if the content would be like a more detailed version of this thread...

This is indeed likely - coronaviruses are intracellular (live inside cells).  Antibodies aren't very good at killing pathogens that live inside cells, they are much better at killing pathogens that live outside your cells (e.g, in your gut, in your blood, or in the spaces between cells).  Certain types of T cells are capable of finding and killing infected host cells and are much more effective against pathogens that live inside cells.

However, in most cases you will still see an antibody response even if the antibodies are not responsible for killing the pathogen.  The reason for this is to do with how the adaptive immune response is activated. Naive B cells that have never seen pathogens before have a type of antibody called IgM stuck on their surface.  When this encouters an antigen (e.g., from a virus), these surface antibodies bind the antigen and internalise it.  This causes two things to happen.  Firstly, the B cells switch to making different kinds of antibody that are secreted into the environment. These may or may not be capable of killing the pathogen. The other thing that happens is that the internalised antigen is processed by the B cells and then put back out on the surface associated with another protein called MHC.  When the B cell comes into contact with a naive T cell, this MHC/antigen combination can activate the T cell. This two-pronged approach means that generally antibodies are produced during an infection if they aren't ultimately responsible for clearing the infection.

This is an over-simplification but it isn't the case that if a T cell response has cleared an infection you won't be able to detect antibodies so by itself this doesn't account for the apparent underestimation of infection rates based on antibody tests.

They don't claim it to be anything else. What's the problem?

"This has led to speculation about whether a population can achieve some sort of immunity to the virus ..."

I think you're expecting a scientific paper but that's not really what they're trying to do is it? Think of it as news journalism and it's a useful piece. 

Thanks for this authoritative response, even though it poses even more questions! 

There's another theory that it takes less % of the population than simple models would suggest because it is naive to think everyone has the same risk of spreading.

Some groups come into contact with far more people than others - commuters on public transport, health workers, shop workers etc.   Those groups catch it first and once they are immune it is harder for the virus to spread.  

On the other hand if we go back to business as usual and people start going to church, cinema, pubs and so on then we could be out of the stable state and it could all kick off again with new opportunities for spreading.

Makes no sense as 42% of. Ischgl have antibodies and New York City  is probably over 35% now. It was 25% end of April and since then cases have risen by nearly 50 percent.

When it comes to measles etc.. most vaccination schemes aim for 70-90% of the population. 

Well if true it means that London is the best place to be. Apparently 20% of Londons poulation has had covid whilst the rest of the country is at about 8%.

As per Prof Ferguson this morning.

Great post; thanks.

Can the mechanisms that target double stranded RNA be primed (through in-cell learning or by a vaccine) to target a specific virus, or are those candidate vaccines likely just boosting the general response to double stranded RNA?

If I went back and did it all over again I’d either study to be an archeologist or to do immune studies in one capacity or another.  As it is I’m spending my days on nematodes which tend to be on the receiving end of immune responses more often than the giving end...

That's true, but I wonder how much of the active humoral response is mucosal.  It may just be that there isn't a high titre of circulating IgG and IgM.  That said, I read that it was possible to find antibodies against other coronaviruses in blood at least 6 years post-infection.

Also, I'm getting a bit rusty on the role of B cell antigen presentation vs 'professional' APCs, for new antigens, particularly in mucosal immunity.  I should probably go back and do some homework!

There's a lot we still don't understand about this and I'm far from confident that the testing is robust.

Yes, my post above was a vast oversimplification! There are several reasons why antibody tests could be under-reporting.  As you suggest, if the humoral response is mucosal there could be limited circulating antibody. I haven't looked into this test specifically, but it could be that in some individuals class switching is behaving differently (e.g., if the humoral response is mucosal, they could be produing IgA - I don't know if the test responds to this), or it could be that re-activating a memory response in individuals who have previously seen other coronaviruses is bypassing some of these systems. Or, other things I haven't thought of.  Still too much we don't know I think.

Yes, I am rusty here too - a lot has moved on since I last worked in this field! If I remember though, many professional APCs activate B cells as well as T cells (DCs certainly can), although this may depend on what other signals are received.

Absolutely, I am reasonably confident that testing is not robust. The point of my original post was just to say that if T cell responses are important, that by itself is not sufficient to explain why the tests are not robust.

Responses to dsRNA are innate - they are intrinsic and not "learned" during an infection. The RNA vaccines currently being tested actually prime an immune response against one of the viral surface proteins, although the protein itself is not a component of the vaccine. The way this works is that the RNA encodes the protein.  Your own cells can use the RNA to make copies of the protein (without the rest of the virus).  As this protein wasn't in your body when your immune system was learning about what is "self" and what is not, it will be recognised as foreign even though it has been made by your own cells. Your immune system then raises a response to this protein.

Innate immune responses, such as responses to dsRNA are great - they happen a lot faster than adaptive responses so they can clear an infection before it becomes established or keep a pathogen in check while the adaptive response kicks in, and they can also guide which arms of the adaptive response are activated, but in many cases they are not strong or specific enough to clear an established infection without help from the adaptive system.

Thansk.  I thought one of the “RNA” vaccine candidates was targeting the double stranded immune response as opposed to using RNA as a vector for training antibodies.  I may be wrong - homework time!  I didn’t know if that response could be “trained” on certain RNA or not, know I know - thanks.  If could still presumably somehow be primed in a generic way?

This is why I’ve wondered if it takes a stronger infection to leave an antibody response.  If so we’ll never know how many people have had infections that were seen off by innate immunity.  Well unless we get everyone to take and store weekly blood samples...    I’m wishing we had some for Ms Wintertree from January but I’m not crazy enough to keep a weekly blood log...

Possibly - I may have missed that one! I was talking specifically about the trial currently underway headed by Robin Shattock at Imperial.

It could be related to which tissues are infected.  The 'got away with it lightly' scenario might be; relatively small viral load arriving predominantly in upper respiratory tract and thence local draining lymph nodes, predominantly mucosal immune response (and maybe heavily skewed to CD8+ T-cells) , virus quickly eliminated (especially if some pre-existing partial immunity from other coranaviruses). 

For people who get properly poorly it might be more like; large viral load also infecting lower respiratory tract, local viral replication causing a lot of inflammation and viral spread to other tissues, larger systemic immune response with high (or at least measurable) circulating IgM/G.  Plus, of course, these people are likely to be older and/or have significant comorbidities.

Or it might just correlate with previous exposure or genetics (immune repertoire or relative innate resistance to some sneaky new immune evasion strategy this particular virus uses).        

My warped brain prefers an interpretation that "not everyone has the same risk of catching" to the idea that "not everyone has the same risk of spreading". Thus in practice if, say, a large % of the fraction of a population that were actually at risk were immune then that might be equivalent to the 70% (or whatever) needed for herd immunity in some models if the entire population was equally at risk.