DUBLIN – A newly published retrospective analysis of the electronic health care records of more than 500,000 COVID-19 patients has found that infection with SARS-CoV-2 carries “a significantly and substantially” greater risk of cerebral venous thrombosis (CVT) than does either one of the two approved mRNA COVID-19 vaccines or infection with influenza virus.

The absolute numbers are small. The study investigators estimate the risk of CVT developing within two weeks of a COVID-19 diagnosis is 39 per million (with a range of 25.2 – 60.2 per million within a 95% confidence interval). A direct comparison with the current estimates for CVT developing in individuals taking Astrazeneca plc’s COVID-19 vaccine was not possible, as the dataset was obtained from the U.S., which has not approved the vaccine.

According to the latest estimates from the EMA, Astrazeneca’s adenoviral vaccine carries a five-in-a-million risk of CVT with thrombocytopenia developing in vaccinees. “The risk of having a CVT after COVID-19 appears to be substantially and significantly higher than it is after receiving the Oxford/Astrazeneca vaccine,” lead author Maxime Taquet, of Oxford University’s Department of Psychiatry, told journalists during a media briefing. In comparison, the background rate of CVT, calculated by examining the health records of all 81 million in the federated TriNetX Analytics network, was 0.41 per million in any two-week period. That is not an absolute measure, however. “It’s actually surprisingly difficult to work out the baseline incidence of a very rare condition,” said corresponding author Paul Harrison, professor of psychiatry and head of the Translational Neurobiology Group at Oxford University.

The study has not yet been peer-reviewed, but was posted April 15 on a preprint server, given its potential significance. “We think that’s quite important in helping both regulators and the public understand better the risk-benefit question, which of course is at the heart of what the policy makers and regulators are trying to do, to get it right for the vaccines,” Harrison said.

CVT is, said study co-author Masud Husain, of Oxford University’s Nuffield Department of Clinical Neuroscience, interchangeable with the term “cerebral venous sinus thrombosis,” which has been employed by the FDA, the CDC and the EMA to describe the very rare but life-threatening clotting events in blood vessels that drain blood from the brain. The unusual coincidence of CVST with thrombocytopenia – or low blood platelet counts – has aroused concerns on both sides of the Atlantic, as the problem has also occurred at a low frequency in patients who received Johnson & Johnson Co.’s adenoviral COVID-19 vaccine.

The present analysis was inconclusive on the issue of thrombocytopenia. “We can’t really speak to the thrombocytopenia angle, because the lab data weren’t complete, but there’s a hint that we’re seeing the association with that as well,” Harrison said. At the same time, the 20% CVT-associated mortality rate observed in the study was, said Taquet, similar to what the EMA reported for those who developed CVST and thrombocytopenia.

The age range of those COVID-19 patients affected by CVT in the Oxford study was wide. It extended from 16 to 78 years. “Thirty percent of all the events that we’ve observed in the patients with COVID-19 were among patients who were under 30 years old,” Taquet said.

Still very, very rare’

CVT is, in itself, not a major problem in COVID-19, given its rarity and given the greater frequency of other clotting events, such as ischemic stroke. “It’s only as the concerns with the vaccine have grown that it becomes important in COVID, even though it is still very, very rare in COVID, compared to other thrombotic and serious [clotting] events,” Harrison said.

The Oxford team also found an association between COVID-19 and an increased incidence of portal vein thrombosis (PVT), a rare but life-threatening clotting event involving the portal vein, which drains blood from the gastrointestinal tract, gallbladder, spleen and pancreas into the liver. The incidence in those who had COVID-19 was an estimated 436.4 (382.9-497.4) per million, as compared with 98.4 (61.4-157.6) per million in those with influenza or 44.9 (29.7-68) per million in those who had received an mRNA COVID-19 vaccine.

The study appeared just as Denmark became the first European country to drop the Astrazeneca vaccine from its national COVID-19 immunization campaign. The Danish Health Authority made the decision on April 14, because, it said, the epidemic in Denmark “is under control” and because the country has made good progress in vaccinating older age groups, who are most vulnerable to developing severe COVID-19. Denmark is now exploring how to make available to other countries the doses of Astrazeneca vaccine it has acquired, and it will rely on the mRNA vaccines from New-York-based Pfizer Inc. and Cambridge, Mass.-based Moderna Inc. to continue its vaccination campaign.