Like everything in this pandemic the science around the rare clotting disorder seen in AstraZeneca vaccinees has moved at an incredible pace. As the link has become clear, hints for a cause have emerged
Story with @GretchenVogel1 is here, thread to come: sciencemag.org/news/2021/04/h…
@GretchenVogel1 As we explained in an earlier story, the combination of thromboses in unusual places and a low platelet count, quickly led researchers to think of HIT (heparin-induced thrombocytopenia), a rare side effect in people given the blood thinner heparin.
@GretchenVogel1 On Friday, reports from a group in Germany and one in Norway appeared in @NEJM. Both show that the patients have some of the hallmarks of HIT, like antibodies against platelet factor 4 and platelet activation. nejm.org/doi/full/10.10… nejm.org/doi/full/10.10…
@GretchenVogel1@NEJM Quick note on the name:
Greinacher earlier suggested the name "vaccine-induced prothrombotic immune thrombocytopenia” or VIPIT for the rare disorder.
But the two papers now both call it VITT: "vaccine-induced immune thrombotic thrombocytopenia"
@GretchenVogel1@NEJM So what’s the mechanism behind VITT and is it likely to be caused by other vaccines than the AstraZeneca one as well?
Let’s go through what scientists are thinking:
@GretchenVogel1@NEJM One early thought was that this might be a rare reaction in people who had already had #covid19 before being vaccinated. But that’s been largely dismissed. All 5 Norwegian cases for instance appear not to have had a prior #SARSCoV2 infection.
@GretchenVogel1@NEJM Another hypothesis: That the antibodies against spike that the vaccine induces cross-react with PF4. That would have been particularly worrying because in that case the effect we want from the vaccine (the anti-spike antibodies) would also be the problem.
@GretchenVogel1@NEJM Greinacher and colleagues investigated this in a new paper (out as a preprint) and their results suggests this hypothesis is false too. For one, the platelet-activating antibodies isolated from #VITT patients did not react to the coronavirus spike protein. researchsquare.com/article/rs-404…
@GretchenVogel1@NEJM That suggests that producing antibodies against spike does not automatically run the risk of producing anti-PF4 antibodies as well.
That’s why Greinacher said on Friday, that the finding was “fantastic news for the vaccination program.”
@GretchenVogel1@NEJM So what mechanism does Greinacher think IS at play?
Well, HIT is caused when PF4, which is small and positively charged, binds to heparin, a large, negatively charged molecule. That then makes PF4 “more visible” to our immune system.
@GretchenVogel1@NEJM Greinacher thinks something similar could be happening with free DNA (also a negatively charged molecule) in the vaccine. There are about 50 billion viral particles in a dose of the vaccine and some of them may break apart spilling DNA, he suggests.
@GretchenVogel1@NEJM Again, this is just a hypothesis. What is interesting though is that there is some research suggesting that extracellular DNA triggers thromboses as part of an evolved response by our body to deal with injured cells.
(See here for instance: frontiersin.org/articles/10.33…)
@GretchenVogel1@NEJM Another possibility:
PF4 antibodies are already present in vaccinees.
We all harbor some auto-antibodies that could create problems, but they are usually kept in check through an immune mechanism called “peripheral tolerance”. Could the vaccination trigger a breakdown of that?
@GretchenVogel1@NEJM As Gowthami Arepally told me: “When you get vaccinated, sometimes the mechanisms of peripheral tolerance get disrupted. When that happens, does that unleash any autoimmune syndromes that you are predisposed to, like HIT?”
@GretchenVogel1@NEJM We know the AZ vaccine produces a lot of inflammation. So one hypothesis is that this strong inflammatory response breaks down peripheral tolerance in some people with PF4 antibodies leading to VITT.
Again: This is another hypothesis that scientists are testing.
@GretchenVogel1@NEJM There are other hypotheses, too, of course. Pinpointing the mechanism at work here is important to figure out what can be done to reduce the risk and what other vaccines may carry the same risk.
@GretchenVogel1@NEJM If inflammation is crucial, an easy fix may be to halve the dose of the vaccine, Greinacher told us.
People that accidentally got a half-dose of the vaccine did experience fewer side effects. “Part of the problem might be that they just overdose” the vaccine, says Greinacher.
@GretchenVogel1@NEJM Of course, we don’t know any of this for sure.
As @Cox_A_R told me, having fewer common side effects does not necessarily mean rare side effects are fewer too: “We can’t automatically assume effects on more common side effects map to these extremely rare reactions."
@GretchenVogel1@NEJM@Cox_A_R One piece of the puzzle that will become clearer soon is whether we will see the same side effect in other vaccines, particularly the other adenovirus vector ones, like those from J&J and CanSino and Russia’s Sputnik V vaccine.
That will give us important information too.
@GretchenVogel1@NEJM@Cox_A_R Already, EMA is investigating four cases of similar clotting seen in U.S. patients who received the J&J vaccine.
We should not jump to conclusions, this may be a coincidence, but as Greinacher says: “It’s at least very suspicious.”
@GretchenVogel1@NEJM@Cox_A_R I’ll write a separate thread about the risk-benefit calculations here later today or tomorrow.
For now just a very general reminder that these clotting disorders are extremely rare events and in most places for most people the benefits far outweigh the risks.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
I'm seeing a lot of confusion already out there about #mpox and the differences between clades and lineages. I will get into this in more detail later, but for now:
We really don't know for sure whether there is any material difference between clade Ia, Ib, IIa and IIb.
The differences we see might have very little to do with the virus and everything to do with it affecting different populations in different places and spreading different ways once it gets into certain contact networks. Real world data is not comparing apples and apples here...
We will learn a lot in the coming weeks and months and things will become much clearer. But for now there is a lot of uncertainty. My advice as always: Don’t trust anyone who pretends that things are clear and obvious.
In May I wrote about researchers' plans to infect cows in high-security labs with avian influenza #H5N1 to better understand the infections and how easily the virus is transmitted. The results from two of these experiments are now out here in a preprint: biorxiv.org/content/10.110…
WHAT DID THEY DO?
In one experiment (at Kansas State University) 6 calves were infected with an #H5N1 isolate from the current outbreak oronasally and then housed together with three uninfected animals ("sentinels") two days later.
In the other experiment (at Friedrich Loeffler Institut) 3 lactating cows were infected through the udder with an #H5N1 isolate from the US outbreak and 3 other lactating cows the same way with a different #H5N1 isolate from a wild bird in Europe.
One question at the heart of the #h5n1 outbreak in US cows has been: Is there something special about this virus? Or is H5N1 generally able to do this and this particular version was just "in the right place at the right time"?
Quick thread, because it seems we have an answer
Researchers in Germany have done an experiment in a high-security lab infecting cows directly with the strain of #H5N1 circulating in cows in the US (B3.13) and infecting others with an #h5n1 strain from a wild bird in Germany.
(I wrote about the plans here: )science.org/content/articl…
In both cases they infected the udders directly through the teats and in both cases the animals got sick. They "showed clear signs of disease such as a sharp drop in milk production, changes in milk consistency and fever." That suggests there is nothing special about B3.13.
The thing that I find most frustrating about the entire mpox/gain-of-function debate is how the uncertainties that lie at the base of it all just become cemented as certainties that are then carried forward.
(If you know anything about me you know I love me some uncertainty...)
Most importantly: The interim report on the investigation into these experiments released on Tuesday numerous times calls clade II "more transmissible" or even "much more transmissible".
But that is a claim that has very little evidence at all.
In fact you can find plenty of literature that argue the exact opposite, that in fact clade I is more transmissible.
Just, as an example, here is Texas HHS:
"Clade I MPXV, which may be more transmissible and cause more severe infection than Clade II..." dshs.texas.gov/news-alerts/he…
Some more details on the latest (3rd) human case of #H5N1 #avianflu linked to the current outbreak in dairy cows:
- second case in Michigan but not linked to the other case (different farm)
- reported cough and eye discomfort with watery discharge
- given oseltamivir, isolating at home
"As with the previous two cases (one in Texas, one in Michigan), the person is a dairy farm worker with exposure to infected cows, making this another instance of probable cow-to-person spread."
"The patient reported upper respiratory tract symptoms including cough without fever, and eye discomfort with watery discharge. The patient was given antiviral treatment with oseltamivir, is isolating at home, and their symptoms are resolving."
A thought on communication:
In today’s presser's opening statement @USDA presented work on killing H5N1 in ground beef through cooking. They mentioned no virus being present at 160°F and 145°F. Only later when someone asked, they mentioned that at 120°F there was some virus left.
@USDA I have thought a lot about trust and transparency in the wake of the #covid19 pandemic and to me this seems exactly the kind of communication style that does not build trust.
Yet another lesson not learnt in my book...