LAUSR

The use of ChAdOx1 and Ad26.COV2.S anti-COVID-19 adenovirus vector vaccines has been associated with unusual cases of cerebral venous sinus thrombosis (CVST) with thrombocytopenia and anti-platelet factor 4 (anti-PF4) antibodies. In this syndrome, which has been named vaccine-induced immune thrombotic thrombocytopenia (VITT) [1], the binding between anti-PF4 antibodies and platelet FcγRIIA receptor causes platelets activation with the release of procoagulant microparticles leading to CVST [2,3]. A similar mechanism has been observed in cases of heparin-induced thrombocitopenia [4–7]. For both vaccines, on April 7 and 20, 2021, the European Medicines Agency (EMA) stated ‘[. . .] COVID-19 is associated with a risk of hospitalization and death. The reported combination of blood clots and low blood platelets is very rare, and the overall benefits of the vaccine in preventing COVID-19 outweigh the risks of side effects.’ This statement stemmed from the analysis of individual case reports (169 for ChAdOx1 and 12 for Ad26.COV2.S) [8,9] of adverse events reported for these adenovirus vaccines and the demonstrated efficacy of vaccination against COVID-19 transmission and clinical severity [10–12]. Indeed, the burden of thromboembolic events reported after vaccination (469 reports, 191 of them from the European Economic Area [EEA]) was lower than that expected in the general population [13]. Although the EMA did not impose age-related limits for the use these vaccines, several EU countries opted for preferential use among subjects older than 60 years, given that CVST case reports mainly concerned younger women (median age: 42 years) (female/male ratio: 49/12). Nevertheless, these decisions did not consider the individual risk of venous thromboembolism (VTE) among younger individuals. In this respect, the EMA also provided a visual risk contextualization in which ChAdOx1 showed a favorable benefit/risk profile for almost all age categories [14], and a recent statement of CHMP (EMA) further confirmed this position [15]. The discussion regarding VITT was also featured by misinformation and miscommunication, which clearly contributed to vaccine resistance and hesitancy [16,17], although the pandemic in the UK was initially controlled using this type of vaccine [18]. Thus, to fuel the discussion on the individual benefit/risk evaluation for anti-COVID-19 vaccines and to support the public health and ethical role of vaccination against COVID19, we used the Health Search Database (HSD) [19], a general practice research database that covers the electronic healthcare records of about 1 million patients, to quantify the individual risk of VTE among those subjects who developed CVST before vaccine availability, according to age and a validated score employed to predict such a risk [20]. Including 21 clinical risk factors, our VTE score showed good accuracy with an AUC of 0.82 (95% CI: 0.82–0.83), explaining 27.9% of the variation for VTE occurrence, and a margin of error between the predicted and observed risk of less than 10% (under or overestimation) in 70% of the validation cohort [20]. We identified all subjects who were at least 18 years old between 1 January 2000and 31 December 2019. They were followed from the start date (1 January 2000) or their 18 birthday until the occurrence of whichever of these events came first: diagnosis of CVST (ICD-9-CM: 437.6x, 325x), death, transfer out of general practice or end of the study period (31 December 2019). The individual VTE risk among CVST cases was assessed using the aforementioned score, whose determinants were operationally defined on 31 December 2019. We identified 92 cases of CVST (mean age: F: 57.4 ± 18.2 vs. M: 58.5 ± 17.9) with an incidence rate equal to 3.44 (95% CI 2.77–4.22) per 1,000,000 person-years. Figure 1 displays those subjects diagnosed with CVST (squares and triangles according to age) within the background risk of VTE for the general population. Of the 92 cases, the mean score of VTE was equal to 1.67 ± 0.41, thus predicting 30-day risk of VTE equal to 0.7 per 1000 (vs. 0.5 per 1000 in the overall population). Sixteen were younger than 60 years and were staged as very high (n = 10) or high risk (n = 6), and most were female (n = 11) (Figure 2). Also, the total number of patients younger than 60 years who were at high or very high risk in the overall population was not negligible. Our results show that the risk of CVST in the Italian general population is in line with the most recent estimates.