Cerebral venous thrombosis (CVT), a rare thrombotic event that can cause serious neurologic deficits, has been reported after some ChAdOx1 nCoV-19 vaccinations against coronavirus disease 2019 (COVID-19). However, there are few reports of associations between COVID-19 mRNA vaccination and CVT. We retrospectively analyzed CVT occurrence, time of onset after vaccination, outcomes (recovered/not recovered), and death after COVID-19 vaccination from adverse drug reactions (ADR) reports in VigiBase. A disproportionality analysis was performed regarding COVID-19 mRNA vaccines (BNT162b2 and mRNA-1273) and the ChAdOx1 nCoV-19 vaccine. We identified 756 (0.07%) CVT cases (620 (0.05%) after BNT162b2 and 136 (0.01%) after mRNA-1273) of 1,154,023 mRNA vaccine-related ADRs. Significant positive safety signals were noted for COVID-19 mRNA vaccines (95% lower end of information component = 1.56; reporting odds ratio with 95% confidence interval (CI) = 3.27). The median days to CVT onset differed significantly between the BNT162b2 and ChAdOx1 nCoV-19 vaccines (12 (interquartile range, 3–22) and 11 (interquartile range, 7–16), respectively; p = 0.02). Fewer CVT patients died after receiving mRNA vaccines than after receiving the ChAdOx1 nCoV-19 vaccine (odds ratio, 0.32; 95% CI, 0.22–0.45; p < 0.001). We noted a potential safety signal for CVT occurrence after COVID-19 mRNA vaccination. Therefore, awareness about the risk of CVT, even after COVID-19 mRNA vaccination, is necessary.
The key findings of our study of CVT cases from VigiBase reported by 130 countries are that the potential safety signal for the development of CVT was noted in mRNA-based COVID-19 vaccines as well as the ChAdOx1 nCoV-19 vaccine compared with the entire dataset.
There are few reports on CVT after mRNA-based COVID-19 vaccination [17,28]. These studies suggested that CVT occurrences related to mRNA-based COVID-19 vaccines may be due to endothelial dysfunction caused by spike glycoprotein interactions with endothelial cells resulting in immunothrombosis. If the spike glycoprotein of mRNA-based COVID-19 vaccines binds to the angiotensin-converting enzyme 2 receptor, several inflammatory and thrombogenic molecules, such as leukocyte chemotactic factors, cell adhesion molecules (vascular cell adhesion molecule 1 and intercellular adhesion molecule 1), and procoagulant cytokines, can be activated. This mechanism may cause endothelial dysfunction, particularly in brain endothelial cells , which could contribute to a significant disruption of brain endothelial barrier integrity, ultimately promoting thrombus formation. Moreover, a previous study suggested that the spike glycoprotein may induce platelet aggregation and activation and eventually result in thrombus formation . Although the period of time in which the spike glycoprotein persists has not been clearly established, several studies have suggested that it may last for weeks. Thus, spike glycoprotein-related platelet activation triggered by mRNA-based COVID-19 vaccines could explain the trend of CVT occurrences after mRNA-based COVID-19 vaccinations [30,31]. Furthermore, in line with these previous case reports, our results showed that CVT occurred mainly within a few weeks of mRNA-based COVID-19 vaccinations.
In agreement with the well-known relationship between the ChAdOx1 nCoV-19 vaccine and CVT [32,33], our results showed a potential safety signal of the ChAdOx1 nCoV-19 vaccine for CVT. Additionally, in one recent study, about 90% of CVT cases following COVID-19 vaccination occurred after the administration of the ChAdOx1 nCoV-19 vaccine . The ChAdOx1 nCoV-19 vaccine can cause systemic thromboembolism due to thrombosis with thrombocytopenia syndrome . The Food and Drug Administration has found a causal link between the adenovirus vector COVID-19 vaccine and thrombosis with thrombocytopenia syndrome and has provided updates on rare clotting or thrombotic events following adenovirus vector COVID-19 vaccination, primarily in young women . Clinical courses and laboratory test results suggest that the pathogenesis of thrombosis with thrombocytopenia syndrome is similar to that of autoimmune heparin-induced thrombocytopenia. Autoimmune heparin-induced thrombocytopenia is caused by the formation of antibodies to PF4, a component of platelet alpha granules released during platelet activation. Unlike classical heparin-induced thrombocytopenia, endogenous polyanions, such as chondroitin sulfate or polyphosphate, may trigger PF4 antibody formation in autoimmune heparin-induced thrombocytopenia [19,35]. Other previous studies and reports by the European Medicines Agency have demonstrated that an adverse immune reaction called immunosenescence may occur in young people, leading to disseminated intravascular coagulation-like blood changes after ChAdOx1 nCoV-19 vaccination [36,37,38,39]. Our research supports the evidence for the risk of CVT after ChAdOx1 nCoV-19 vaccination with real-world data.
Interestingly, our results showed that there was a difference in the onset of CVT after exposure to the mRNA-based COVID-19 vaccines and ChAdOx1 nCoV-19 vaccine. The median values of 13 and 11 days for the time to onset of CVT for the mRNA-based COVID-19 vaccines and the ChAdOx1 nCoV-19 vaccine, respectively, were similar. However, the mRNA-based COVID-19 vaccines had the highest number of CVT cases in the first week after vaccination, after which the incidence decreased gradually. In contrast, the ChAdOx1 nCoV-19 vaccine showed the highest incidence of CVT in the second week after vaccination and a sharp decrease thereafter. These timelines suggest that the thrombosis mechanisms of these vaccines differ. The mRNA-based COVID-19 vaccines, which target the spike glycoprotein of SARS-CoV-2, directly induce intracellular production of the spike protein. This spike protein plays a key role in the initiation of the immune response, which may last for up to a few weeks. CVT occurred from 1 to 9 days after vaccination [17,28], and the neutralizing antibody titer after vaccination was maintained for 35 to 119 days [40,41]. This evidence supports a wide temporal distribution of CVT occurrence related to mRNA-based COVID-19 vaccination in our study. In contrast, the formation of PF4 antibodies and PF4–polyanion complexes in vaccine-induced immune thrombotic thrombocytopenia, the most well-known thrombotic complication after ChAdOx1 nCoV-19 vaccination, takes time . A type-II heparin-induced thrombocytopenia that develops via a similar mechanism occurs 5 to 14 days after exposure due to the time required for the formation of antibodies . In another study of the ChAdOx1 nCoV-19 vaccine, CVT was reported 5 to 30 days (median 14 days) after vaccination . Regardless, it is possible that other mechanisms of thrombosis have not yet been elucidated.
In this study, the time until CVT occurrence after vaccination was significantly different between the BNT162b2 and ChAdOx1 nCOV-19 vaccines but not between the mRNA-1273 and ChAdOx1 nCOV-19 vaccines. Various factors, such as vaccine components or immune responses, might cause this difference between mRNA vaccines. However, our study could not provide acceptable evidence related to these differences. Recently, several studies that compared the BNT162b2 and mRNA-1273 vaccines reported the possibility of differences in clinical responses or outcomes in addition to SARS-CoV-2 antibody responses [44,45]. The rates of breakthrough infections and 60-day hospitalizations were significantly lower in those vaccinated with mRNA-1273 compared with BNT162b2; this result suggests that these vaccines may act via different mechanisms . Further studies on the specific mechanism of CVT occurrence after mRNA vaccination and prospective studies on clinical outcomes are needed.
In our study, the number of deaths in CVT patients was lower after mRNA vaccination than after vaccination with ChAdOx1 nCoV-19. A previous study showed that significant risk factors for mortality due to thrombosis with thrombocytopenia syndrome after ChAdOx1 nCoV-19 vaccination were intracerebral hemorrhage and CVT . However, there have been few studies on CVT occurrence after mRNA vaccination and CVT-related mortality. Furthermore, it is difficult to present accurate evidence because VigiBase does not provide information on various parameters related to mortality, such as laboratory data, brain-imaging findings, or the occurrence of systemic thromboembolism.
Our study has limitations. First, if the national drug-monitoring center of a country does not report ADRs, these cases will not be present in VigiBase. However, VigiBase includes rare ADRs and generalized ADR information from more than 130 countries. Second, VigiBase does not provide any validation of laboratory findings, radiologic information, or accuracy of diagnosis. Information on whether CVT occurred after the first or the second vaccine dose was also not included. Third, vaccine-induced immune thrombotic thrombocytopenia received major public attention after April 2021 , and this may have affected the increased reports of CVT cases after COVID-19 vaccination. Lastly, as mentioned above, it is difficult to directly compare outcome parameters, including death, between the different types of COVID-19 vaccines in VigiBase.
Our study demonstrated a potential safety signal for occurrence of CVT for COVID-19 mRNA vaccination. It is necessary to be aware of the risk of CVT occurrence, even after COVID-19 mRNA vaccination.