EXPERT REACTION: Brain blood clots may be more likely after COVID-19 infection than following vaccines

The new pre-print paper (not yet peer reviewed), provides valuable information on the incidence of cerebral venous thrombosis (CVT) (also known as cerebral venous sinus thrombosis) and portal vein thrombosis (PVT) in patients with COVID-19, in patients after vaccination with mRNA vaccines and the general population.

This research fills a gap in knowledge by providing a comparison (using data from about half a million subjects) of the risks of CVT and PVT in people infected with COVID-19 and those vaccinated with Pfizer’s and Moderna’s vaccines (data for AstraZeneca’s vaccine is also included in the study by those data come from a different dataset).

The importance of this research is that it does not look at increases in the risk of thrombosis in general, but it concentrates on specific types of thrombosis associated with COVID-19 infection and COVID-19 vaccination.

Overall, they found that CVT events are rare: 4.1 per million following vaccination compared to 39 per million in COVID-19 infection. The rate in the general population is 0.4 per million.

The rates for PVT were higher: 45 per million following vaccination and 436 per million in COVID-19 infection.

The data should be interpreted conservatively given that the study was based on an electronic database (hence the accuracy of diagnoses of CVT and PVT is unknown).

The current concern about AstraZeneca’s (and Johnson & Johnson vaccines) is that, in rare cases, they produced a specific immune reaction the leads to thrombosis. This immune reaction can be detected by a particular laboratory test.

An important point is that in the new study no specific haematological tests were reported in patients diagnosed with thrombosis, therefore we don’t know if the thromboses developed in these patients were due to an immune reaction or to other underlying conditions.

In general, this study should not be a reason for undue concern regarding vaccination with mRNA-based vaccines. The authors state “we cannot conclude that the mRNA vaccines studied here are associated with an increased risk of CVT; far larger samples are needed to address this question”.

This paper also provides useful context regarding the benefits of vaccination: the risk of CVT is about 100 times higher if people are infected with COVID-19 than in the general population.

This observational study uses administrative coding data from electronic health records in a US population to compare the rates of CVT, portal vein thrombosis following COVID-19 infection against the risk following administration of either the Pfizer BioNTech or Moderna vaccines. The study clearly shows that the risk of these specific thrombotic events following COVID-19 infection is higher than either the mRNA or the AZ vaccine. The absence of this signal in >172K people with influenza is interesting and underscores the effect of COVID on the vascular system where it causes dysfunction and dysregulation accounting for its multisystem effects which are distinct to the flu.

The concern with this study is that it is conflating clotting in unusual sites (CVST, PVT) from COVID to a recently discovered syndrome (VITT) that has been described with the viral vectored (AZ, JJ) but not mRNA vaccines.

VITT has a higher published mortality of 55-60% from the German and Norwegian series compared to the 18-20% described in this report and is a syndrome with a constellation of findings which includes clotting in unusual sites such as the brain, portal vein but other sites also (lung, kidney, heart, abdominal vessels) as well as a bleeding predisposition-patients are clotting and bleeding at the same time.

Unlike COVID, it is challenging to diagnose (VITT) requiring a high index of suspicion (because it is rare) and specialised tests that are not always available like scans (CT, MRI) and antibody tests for confirmation. This delay to diagnosis may well be contributing to poor outcomes.

The study from Oxford University aims to estimate how common thromboses were following COVID-19 disease, in particular cerebral venous thromboses (CVT) and portal vein thromboses (PVT). They then compared these rates with the incidence of CVT and PVT in populations receiving mRNA (Pfizer and Moderna) or COVID-19 AstraZeneca vaccines.
 
The study finds a significant increase in the rate of thromboses after COVID-19 disease, which corresponds with existing data showing that COVID-19 has a procoagulant or pro-clotting impact in both adults and children. According to the study, there seems to be a lower rate of these same clots in people who have had either a mRNA or viral vector COVID-19 AstraZeneca vaccine.
 
However, a number of methodological and applicability challenges emerge when trying to compare this data between COVID-19 disease and COVID-19 vaccines. Firstly, the data sources were from different populations and countries and are not directly comparable. They were not matched for age or other confounding factors. Secondly, it is unclear what the background rates of CVT are in each of these populations. This means that it is not clear whether the reported higher rate of clots represents a true increase above an expected baseline rate of CVT or PVT in those specific populations.
 
Furthermore, comparing isolated CVT/PVT with thrombocytopaenia syndrome (TTS), the syndrome of interest following COVID-19 AstraZeneca vaccine, means we actually may be comparing two separate entities and drawing conclusions that are not entirely accurate. There is not enough clarity within the existing study data to indicate whether there were other clinical signs and laboratory signals like low platelets to indicate that there was a TTS-like syndrome after COVID-19 disease.
 
Finally, this study reinforces that effective strategies such as vaccinations to reduce the health burden of COVID-19 is important. However, with a low incidence of COVID-19 disease in Australia, and hence an absence of this increased risk of thrombosis, this study does not directly impact or should not change our approach to a COVID-19 vaccine rollout. Individual benefit-risk assessments based on known and associated benefits of preventing severe COVID-19 disease and risks such as TTS remains paramount as part of an individual’s decision to receive a COVID-19 vaccine.

This is an interesting study comparing the incidence of cerebral venous thrombosis (CVT) in the period immediately following COVID-19 diagnosis with that following influenza diagnosis and receipt of the BNT162b2 (Pfizer) or mRNA-1273 (Moderna) vaccines.

CVT is part of the syndrome that is seemingly linked to the viral vectored AstraZeneca vaccine and perhaps the Johnson & Johnson vaccine (which is also a viral vectored vaccine) although the particular syndrome being discussed as potentially induced by the vaccines also includes other features particularly thrombocytopenia (or low platelet counts), hence is referred to by names including vaccine induced prothrombotic immune thrombocytopenia or VIPIT amongst others.

It is also noteworthy that this comparison has been done with mRNA vaccines that are not to date associated with this syndrome.

This is a large study utilising electronic health records that includes over 500 000 COVID-19 diagnoses, nearly 200 000 influenza diagnoses and nearly 500 000 recipients of mRNA vaccines.

The reported incidence of CVT following COVID-19 diagnosis was 30.0 per million people, which is significantly higher than the rates seen following influenza of 0.0 per million people and following receipt of mRNA vaccine at 4.1 per million people.

In the preprint they quote the estimate from the European Medicines Agency for the incidence associated with the ChAdOx1 (AstraZeneca) vaccine as 5.0 per million people which if we compare to the figures of CVT outlined in this study is less than that associated with COVID-19 infection yet greater than that following influenza diagnosis and slightly higher than that reported following the receipt of either mRNA vaccine.

While interesting, it is still a pre-print so it is yet to be peer reviewed, and the study itself compares CVT following the mRNA vaccines so the same methodology has not been applied to the actual vaccines thought to cause the syndrome in question (i.e. the viral vectored vaccines).

Despite some limitations this is an interesting study in so far as that it shows there are potentially other causes of CVT including COVID-19 itself and highlights that when we are making assessments of the risks versus benefits of vaccines in the context of adverse events we also have to consider other causes of these events including the infection itself as well as the background risks of such events.

Blood clotting leading to thrombosis following COVID-19 infection is 8-10 times higher than seen in cases after vaccination.  In the UK, 39 in a million COVID-19 patients presented with thrombosis, while post-COVID-19 mRNA vaccine (Pfizer or Moderna), the rate was 4 in a million and post AstraZeneca vaccine, 5 in a million people were impacted.

There is some risk of blood clotting post any COVID-19 vaccination but that is much lower than the rate after getting COVID infection. It has been seen that the risk of thrombosis from COVID-19 is about 10 times greater than after mRNA vaccines and 8 times greater than after DNA vaccine AstraZeneca vaccines.  

This data and similar data from Europe show balance between risks and benefits of vaccination should be considered.

Currently, there is insufficient evidence suggesting vaccines are directly responsible for thrombosis. In rare patients, an unusual immune response may be triggered by vaccination leading to blood clots. It is possible people impacted were predisposed to clotting issue and the vaccine triggered it.

I believe vaccinations should continue in regions with community transmission of the virus. More evidence is required, and quickly, to make a conclusive statement but in meantime some protection is better than nothing.