Cerebral Venous Sinus Thrombosis Associated with Vaccine-Induced Thrombotic Thrombocytopenia—A Narrative Review

: In March 2021, cerebral venous sinus thrombosis and thrombocytopenia after vaccination with adenovirus-based vaccine against SARS-CoV-2 were ﬁrst reported. The underlining condition has been termed vaccine-induced immune thrombocytopenia (VITT). Anti-platelet factor 4 antibodies have been proposed as a central component of the pathomechanism. Treatment recommendations en-tailed immunomodulation with intravenous immunoglobulins, avoidance of heparins and avoidance of platelet transfusions. Although mortality from VITT-associated cerebral venous sinus thrombosis has decreased over time, it remains high. The aim of this narrative review is to describe different aspects of this disease according to the current state of knowledge.

The aim of this narrative review was to describe different aspects of CVST-VITT according to the current state of knowledge.

Pathophysiological Aspects
VITT, in which anti-platelet factor (PF) 4 antibodies are produced after adenovirusbased SARS-CoV-2 vaccination as part of the inflammatory response and immune stimulation, is a condition very similar to HIT, and even more similar to autoimmune HIT (aHIT). Venous thrombosis and/or arterial vessel occlusions and thrombocytopenia are hallmarks of both VITT and aHIT. Heparin-induced thrombocytopenia (HIT) is a systemic hypercoagulable state that occurs when heparin-dependent IgG antibodies bind to PF4/heparin complexes. aHIT is a condition in which anti-PF4 antibodies are produced without recent exposure to heparin [18].
Anti-PF4 antibodies were discovered in VITT, as in aHIT, by an enzyme-linked immunosorbent assay (ELISA) [3,9]. In addition to the similarities in clinical presentation, the thrombus structure in VITT is also similar to aHIT, with platelet-rich 'white' venous thrombi as opposed to 'red' stasis thrombi seen in classical CVST [4].
It remains to be determined which components of the adenovirus-based SARS-CoV2 vaccines trigger the production of anti-PF4 antibodies [3,19,20]. PF4 is a naturally occurring, positively charged chemokine that binds to negatively charged molecules, such as heparin, bacterial cell walls and polyanionic compounds used in prosthetic surgery [19]. PF4-anionic complexes trigger the production of anti-PF4 antibodies. Greinacher et al. 2021 demonstrated the formation of PF4-vaccine complexes, which stimulated a pro-inflammatory response of B-lymphocytes, a phenomenon that was induced by the addition of DNA [20]. Other mechanisms possibly involved in the pathophysiology of VITT are suspected, such as the downregulation of ACE-receptors, molecular mimicry between the SARS-CoV-2 spike protein and PF4 molecules, RNAemia, and the involvement of vaccine constituents, such as ethylenediaminetetraacetic acid (EDTA) [20].
The occurrence of CVST and other venous thromboses and arterial vessel occlusions after mRNA-based SARS-CoV-2 vaccines have also been described [7,[21][22][23][24][25]. However, so far, there is no report on VITT in these patients. The mechanism of venous thrombosis and arterial vessel occlusion after mRNA-based SARS-CoV-2 vaccination is, so far, less understood than after adenovirus-based SARS-CoV-2 vaccination [26]. In the following sections, we focus exclusively on VITT after adenovirus-based SARS-CoV-2 vaccination.

Epidemiology
The phase 3 trial of ChAdOx1 nCoV-1 (AstraZeneca) randomized 23,848 participants [27], and the trial of Ad26.COV2.S (Janssen) vaccine 39,321 participants [28]. However, these clinical trials, which led to the clinical approval of the adenovirus-based SARS-CoV-2 vaccines, were underpowered to detect rare adverse events such as VITT.
So far, more than 11 billion SARS-CoV-2 vaccinations have been administered worldwide [29]. The largest vaccination campaign in history is still ongoing.
An analysis of patients vaccinated in Europe using the Eudravigilance database from 1 January to 30 July 2021 showed that 38,664,988 people had been vaccinated with ChAdOx1 nCoV-1 (AstraZeneca/Oxford) and 10,972,234 with Ad26.COV2.S (Janssen/Johnson & Johnson) [30]. In this analysis, the rate of CVST with thrombocytopenia during the study period of 6 months was 21.6 (95%CI 20.16-23.11) per 1 million vaccination days for ChA-dOx1 nCoV-1 and 11.48 (95%CI 9.57-13.67) per 1 million vaccination days for Ad26.COV2.S vaccination [30]. Another study by the UK Medicines and Healthcare Products Regulatory Agency and the US Centers of Disease Control and Prevention came to very similar conclusions [31].
Overall, VITT and CVST-VITT are rare after SARS-CoV-2 vaccination; CVST after COVID-19 is more common, and ischaemic strokes or disabling or deadly events show even higher incidence rates [32].
Several conventional CVST risk factors are known, but they seem to be less present in patients with CVST-VITT. Conventional CVST risk factors may be temporary (prothrombogenic drugs, pregnancy and puerperium, infections) or permanent (common prothrombotic disorders such as thrombophilia, antiphospholipid syndrome, myeloproliferative disorders and malignancies) [8,10,11,17,34].
The occurrence of thrombosis with thrombocytopenia caused by idiopathic thrombocytopenic purpura (ITP) has been described after the administration of the measlesmumps-rubella vaccine [35]. Given the rarity (1 case for every 40,000 doses) and the lower total number of doses administered compared with COVID-19 vaccines, the evidence on ITP following measles-mumps-rubella vaccination in the context of post SARS-CoV-2 VITT is of limited relevance.

Definition and Laboratory Features of VITT
Different societies and research groups have proposed criteria for VITT (Table 1) [8,[36][37][38][39][40][41]. Some criteria require only the detection of thrombosis and thrombocytopenia, while others require additional laboratory parameters, such as D-dimer and fibrinogen levels and positive anti-PF4 antibodies, for the diagnosis of VITT. In addition, the time window after vaccination varies considerably depending on the definition (Table 1) [8,[36][37][38][39][40][41].  The symptoms at the time of onset of VITT have been described as varying [42,43]. For example, headache has been suggested as an early symptom of CVST-VITT without radiologic evidence of early stage CVST [43][44][45]. Furthermore, anti-PF4 antibodies are potentially negative at the beginning of the clinical course and only become positive later [42]. In some patients, D-dimer and fibrinogen levels were described to be within the normal range throughout the clinical course [46].
In summary, some patients diagnosed with VITT may not fully meet the proposed VITT definition criteria.
Thrombocytopenia and positive anti-PF4 antibodies have been repeatedly described in CVST-VITT. A descriptive analysis of baseline thrombocytopenia in patients with CVST from the pre-COVID-19 era, conducted by the International Cerebral Venous Thrombosis Consortium (ICVST), found baseline thrombocytopenia-vastly non-severe-in 8% of patients, with obvious explanations for the thrombocytopenia in the majority of cases, such as cancer, inflammation, alcohol dependence or intake of drugs. HIT with anti-PF4/heparin antibodies was diagnosed in one single patient. In the pre-COVID-19 era, baseline thrombocytopenia was uncommon in patients with CVST, and HIT and anti-PF4/heparin antibodies were rare. The results suggest a causal association between adenovirus-based SARS-CoV-2 vaccination and CVST with thrombocytopenia [47].
Around 7% of healthy individuals develop antibodies to PF4/polyanion complexes at lower titres than in VITT. This urges caution when screening healthy individuals for VITT, as in a significant proportion of individuals, antibodies against PF4/polyanion complexes have no clinical significance and do not require treatment [48].
Other laboratory features that have been described as typical for VITT include increased D-dimer levels, a shortened pro-thrombin time and an activated partial thromboplastin time (aPTT) ( Table 1).

Location of VITT
Thromboses in VITT have been shown to occur in the cerebral venous system, but also in other, sometimes unusual, locations. According to an analysis of 220 patients with definite or probable VITT, about half of all patients suffered from CVST, one third from pulmonary and deep venous thrombosis, one fifth from splanchnic venous thrombosis, one fifth from arterial vessel occlusion, and 29% had more than two affected vascular beds [8]. Moreover, according to a meta-analysis, about half of the patients with VITT suffered from CVST [19].
A role of PF4 in the interaction of the drainage of the splanchnic and cerebral venous systems with bacteria and viruses from the intestinal and nasal tissues has been proposed as an explanation for the unusual location of thrombosis in VITT. Endothelial activation facilitates the activation of the innate immune system through an interaction between the pathogens and PF4. A higher titre of tissue factor and PF4 in these environments may lead to a higher titre of immunogenic complexes at these sites with stronger local immune responses and consecutive thrombosis [49].
Other symptoms depend on the extent and location of the thromboses and arterial vessel occlusions. It is noteworthy that a considerable number of CVST-VITT patients have severe disease with coma and intracerebral haemorrhage [9,10,35].

Treatment
The first reports on VITT also included treatment recommendations largely based on the similarities with aHIT and the in vitro inhibition of platelet activation by the VITT antibodies after the addition of intravenous immunoglobulin (IVIG) [3,8,9]. Treatment recommendations consist of IVIG and immunosuppression, administration of non-heparin anticoagulants (anti-thrombin and anti-Xa) and strict avoidance of heparin and platelet transfusions.
There is limited evidence on the efficacy and safety of these recommendations [35,54]. Their scientific validity is also changing due to new insights into the pathophysiology of this disease.
The effectiveness of IVIG is limited to case reports only [55]. Platelet transfusions are not recommended, as they can trigger immune reactions. However, no study has yet investigated how platelet transfusions affect outcome.

Conclusions
CVST-VITT is rare and differs from CVST in the pre-pandemic era. The benefits of SARS-CoV-2 vaccination far outweigh the small risk of VITT. The risk of CVST after COVID-19 is higher than the risk of CVST-VITT.
CVST-VITT has a heterogeneous and severe clinical picture, and although mortality has decreased over time, it remains considerable.
Due to its heterogeneous presentation, the diagnosis of VITT is challenging. The currently proposed case definitions for VITT are heterogeneous, and the diagnosis of VITT should be made on a case-by-case basis.
The type of SARS-CoV-2 vaccine could be adjusted according to patient characteristics, such as age. For example, adenovirus-based SARS-CoV-2 vaccines may be safer in middleaged and elderly people than in younger people.
Currently, there is limited evidence to support treatment recommendations for VITT, and data on the long-term prognosis, including mortality, disability and risk of recurrence, and long-term secondary prevention are lacking. Further research is needed.

Conflicts of Interest:
The authors declare no conflict of interest.