A point that is often made is that widespread transmission and rapid viral evolution should make the virus become milder over time. We and others have pointed out that this is not the case for this virus (h/t @ariskatzourakis, @tryangregory). (2)
There's no reason to expect the virus to "settle in" due to rapid evolution- we showed in a recent paper that the virus could kill almost everyone it infects with minimal impact on its fitness (ability to succeed evolutionarily): mdpi.com/2673-8112/2/12… (3)
Our paper above suggested that virulence may be under drift, at a minimum (meaning that it may increase or decrease over time, based on luck). But what if there were biological advantages to increased virulence?(4)
That's not something that can be addressed in a modeling paper, but fortunately for us, there is an enormous body of literature on this virus at present (193k papers), making it (already) one of the most studied pathogens of all time! (5)
So, we dug into that literature a bit for clues to this question, using a simple lens- what changes in virus biology are likely to lead to both an increase in virulence and viral fitness (transmission potential)? This is a simple question, but a powerful one.(6)
Over 2yrs ago we predicted rapid immune evasion of the spike based on a similar observation- the virus had essentially unlimited mutational space to increase its immune evasion without impacting fitness.
In this review, we looked at the biological processes involved in viral pathogenesis (infection, replication) & immune evasion/suppression. We found many examples where the virus could increase fitness by increasing its ability to suppress/evade the immune system, e.g. (8)
Both loss-of-function immune evasion & gain-of- immune suppression have been reported extensively in the literature. There is a whole panoply of viral genes whose function it is to suppress the immune system directly. (9)
Notably, mutations in these genes are not expected to the efficiency of infection or transmission, meaning that the virus is free to fine-tune this machinery without impacting fitness. That's a lot of evolutionary whitespace for the virus to explore!(10)
Another potential way in which the virus can access increased virulence is through antibody-dependent enhancement (ADE), where non-neutralizing antibodies help the virus enter host cells- in effect, playing for the wrong team!(11)
ADE occurs in dengue, for example- when Abs from a 1st infection bind to an infecting viral particle during a 2nd infection with a different serotype, they can't neutralize the virus, but instead, help the virus infect immune cells more efficiently. (12)
We've pointed out in a separate preprint that the risk of serotype formation (and hence ADE) is very much on the table for this virus, given enough time for viral evolution). It hasn't happened yet, but the risk remains real. (13)
Several aspects of SARS-CoV-2 biology also factor into the intrinsic virulence of viral variants. Many of these aspects are likely to have a neutral impact on transmissibility (and hence, from an evolutionary standpoint, be under drift rather than natural selection).(14)
But some changes may yield a benefit to the virus for fitness, while also increasing virulence. For example, tighter binding of the viral spike protein to ACE2, or an increase in the viral load, or the ability to enter host cells through other pathways (altered tropism).(15)
The thing is- none of these mechanisms is theoretical. They've all been demonstrated at one point or the other, just not at the same time in an evolutionarily successful SARS-CoV-2 variant. Lucky for us- so far, but luck is not a good basis for a public health strategy.(16)
The rapid pace of viral evolution means deadlier variants can- by dumb luck alone- be outcompeted by less deadly variants. If virulence and transmissibility are uncorrelated, or weakly correlated, you could expect viral interference to do some of the heavy lifting for us. (17)
But viral evolution is not our friend (🧵👇). And viral recombination is afoot already, letting the virus access big changes in properties in a relatively short span of time. Recombination makes things even more volatile- it's not hard to see why! (18)
Our review suggests that one of these days, along will come a variant with properties that *we've seen before*. And these properties may well lead to a higher infection fatality rate for a highly transmissible variant. There are many ways in which this could all go sideways. (19)
A point we've made before is that under 'let it rip', even relatively small increases in IFR could be devastating. So, such a situation, while entirely predictable, will also be something that we are caught reacting to. (20)
Because increases in fatality rates are a lagging indicator, and only become apparent after a ton of people have been infected.
Unfortunately, a reactive public health strategy in the face of rapid viral evolution is just not good enough. It’s a disaster waiting to happen.(21)
We are in year 4 of what looks like a long-term pandemic & only a few months into the scenario where countries worldwide have opted for unlimited viral transmission. We know that relying on vaccines alone will not slow the evolution of this virus. (22)
The virus still has explored only a vanishingly small portion of its evolutionary space. Our review suggests that the quadrillion monkeys on typewriters working tirelessly to improve the virus are not done with us yet. We could still be in for some really nasty surprises! (23)
This review is not aimed at predicting what will happen next. No one knows that.
It’s simply pointing out that the biological basis (direct gain or loss of function resulting from mutations) for accessing higher virulence has been documented for this virus already.(24)
The bottom line (again!) is that we need to take viral evolution seriously. Our work suggests the virus has showed behaviors that should be cause for concern already. Rather than explaining why we shouldn't panic, public health should focus on mitigating worst-case scenarios.(25)
Check out this article from Jan 2021, by the always-entertaining @marc_veld about our post-pandemic future. An endless litany of opinions masquerading as facts, mixed in with statements that were known to be wrong at the time. (1/): nature.com/articles/s4157…
“COVID-19 vaccinations have started. They will stop the pandemic... maintenance of population immunity will not depend on continued vaccinations but on the endemic presence of SARS-CoV-2.” (2/)
For reference, this is what we predicted (nov 2020)- a high endemic disease burden and repeat reinfections. Vaccinal suppression was within reach at the time and remains so (3/): journals.plos.org/plosone/articl…
Although the specifics of viral evolution are not predictable, its role in the trajectory of the pandemic has been very predictable, unfortunately. In the fall of '20, we posted a preprint predicting the virus would rapidly evolve to evade Abs: (2) journals.plos.org/plosone/articl…
In the spring of '21, we posted a preprint predicting that reopening after having partially vaccinated the population would lead to a massive variant-fueled rebound in cases (this was in the same week as the CDC said you can vax & relax): journals.plos.org/plosone/articl… (3)
TL;DR dosing existing vaccines more often could lead to ⬇️ in both disease severity & infection risk. If that pans out, you could use the vaccines to better protect both populations & individuals . (2/)
The point is not to say that the @CDCgov should immediately recommend 4 boosters a year for everyone. The point is that we haven't explored our options fully yet. We know more about vax efficacy than we did 2yrs ago (neutralizing antibodies & not T-cells are the key) (3/)
Flipping the "personal freedom" argument on its head by allowing people the freedom to avoid covid will not end the pandemic overnight. But it provides a start, because it will reduce the cost to individuals for avoiding covid. (26/)
There are many other things that can and should be done (improved indoor air quality, access to antiviral prophylactics, cheap testing, subsidies for vaccinations) to actively manage the burden of covid. (27/)
These things can be done without trying to change the minds of those who don't think covid is a risk. (28/)
Last week, the FDA advisory panel met to discuss a simplified COVID vaccine strategy going forward. The plan is to offer updated boosters to the general population 1x/yr.
Was this the right choice? A preprint posted by us today explores that question. A (long) 🧵follows (1/):
We were interested in understanding the impact of vaccination frequency on protection against severe disease and infection. To examine this, we used an agent-based modeling framework coupled with the outputs of a population pharmacokinetic model (2/)
In simple terms, we fitted a model to a date set consisting the rate of neutralizing antibody waning following vaccination, and coupled that model to a simulation where agents (think Sims) were free to interact and be infected. We accounted for both natural & vax immunity (3/)
We hear a lot about public health being a personal thing these days. “If you’re vulnerable, shield and stay up to date on your vaccines” is what PH advocates for now. So, how well will this work? Our latest preprint explores this question ( 🧵 follows, details 👇👇) (1/:)
TL; DR is that shielding is *very* challenging to pull off under current conditions. We have all “volunteered” for a mass infection experiment that is difficult if not impossible to opt out of (2/)
And for most people, this will come with very real costs: “At an individual level, people who are vaccinated & not taking measures to reduce their contact rate can expect to spend an average of 6 days/yr acutely sick with COVID-19 and also incur a 12% risk of long covid “ (3/)