Cerebral venous sinus thrombosis after vaccination: the UK experience.

An important but rare complication of COVID-19 vaccination is vaccine-induced immune thrombotic thrombocytopenia (VITT) associated with the adenovirus vector vaccines, Ad26.COV2.S (Johnson & Johnson) and ChAdOx1 (Oxford–AstraZeneca).1, 2, 3, 4, 5 VITT occurs more commonly in women younger than 50 years who present within 5–24 days of vaccination with thrombosis in unusual sites—the majority with cerebral venous sinus thrombosis.1, 6 Thrombocytopenia, elevated D-dimer, decreased fibrinogen, and positive antibodies against platelet factor 4 (PF4) are commonly observed.1, 2, 3, 4, 5, 6 Recommended treatments for VITT, based on similarities with autoimmune heparin-induced thrombocytopenia (HIT), include non-heparin anticoagulation, intravenous immunoglobulin, and avoidance of platelet transfusions. Mortality associated with VITT is approximately 40%.

In The Lancet, Richard Perry and colleagues report on the largest series to date of patients with VITT-associated cerebral venous sinus thrombosis. In this multicentre cohort study, cerebral venous sinus thrombosis following COVID-19 vaccination was defined as VITT-associated if platelet count nadir was less than 150 × 109 per L and, if measured, D-dimer concentration was greater than 2000 μg/L. Between April 1 and May 20, 2021, the study enrolled 70 patients with VITT-associated cerebral venous sinus thrombosis and 25 patients with cerebral venous sinus thrombosis that did not meet criteria for VITT from 43 hospitals in the UK, as well as a large historical cohort of patients with cerebral venous sinus thrombosis.

All cases of VITT-associated cerebral venous sinus thrombosis occurred after a first dose of the ChAdOx1 vaccine. 56 (97%) of 58 patients with VITT for whom anti-PF4 antibody tests were available tested positive using an ELISA. Compared with those without VITT, patients with VITT were younger (median age 47 years [IQR 32–55] vs 57 years [41-62]; p=0·0045), were more likely to be female (39 [56%] of 70 vs 11 [44%] of 25), had more intracranial veins thrombosed (median 3 [IQR 2–4] vs 2 [2-3]; p=0·041), and had an increased likelihood of concurrent extracranial thrombosis (31 [44%] of 70 vs one [4%] of 25; p=0·0003). The primary outcome of death or dependency on others at the end of hospital admission occurred more frequently in patients with VITT-associated cerebral venous sinus thrombosis than in the non-VITT control group (33 [47%] of 70 vs four [16%] of 25; p=0·0061). The proportion of patients with VITT who were dead or dependent at discharge was lower in those who received a non-heparin anticoagulant (18 [36%] of 50 vs 15 [75%] of 20; p=0·0031) or intravenous immunoglobulin (22 [40%] of 55 vs 11 [73%] of 15; p=0·022) compared with those who did not receive these treatments.

Perry and colleagues' study proposes new diagnostic criteria for VITT based on patients whom the authors suspected of being misclassified according to existing criteria.1, 2, 9, 10 One patient in the non-VITT group had an elevated D-dimer (4985 μg/L) and positive anti-PF4 antibodies on two different ELISAs yet a platelet nadir of 158 × 109 per L. Two patients with clinical features highly suspicious for VITT were assigned to the non-VITT group on the basis of D-dimer concentrations less than 2000 μg/L, including one with positive HIT antibody testing. Perry and colleagues propose dividing cases of cerebral venous sinus thrombosis following COVID-19 vaccination into possible, probable, and definite VITT-associated cerebral venous sinus thrombosis, allowing for inclusion of atypical presentations with normal platelet counts, normal D-dimer, or negative HIT antibody testing.

The utility of the proposed criteria is yet to be determined. The patient with a platelet nadir of 158 × 109 per L would be, to our knowledge, the first reported instance of VITT with a normal platelet count, yet comparisons of presenting versus prevaccination platelet counts were not available in this study. Based on HIT paradigms, a relative decline in platelet count from baseline, rather than absolute thrombocytopenia, is likely to be a uniformly distinguishing feature of VITT. The exact rate and degree of platelet decline in VITT following COVID-19 vaccination are unknown1, 5, 6 and represent an area of active investigation. Although rare false negatives might occur, ELISA testing in VITT is generally very reliable11, 12 and it is unclear if patients with negative ELISA tests for anti-PF4 antibodies and functional HIT testing could still be classified as having VITT. A third of patients in Perry and colleagues’ study had anti-PF4 antibody testing using a chemiluminescent immunoassay; such immunoassays have poor sensitivity for VITT compared with ELISA testing11, 12 and could explain some of the negative test results.

An important consideration is that 19 (20%) of 95 study patients did not have anti-PF4 antibody testing available. Additional patients in the VITT group could have had negative anti-PF4 antibody testing, and additional patients in the non-VITT group could have had positive testing, and it is possible that a spectrum of VITT might exist, similar to HIT. Other limitations of the study include the small sample size, reflecting the rarity of cerebral venous sinus thrombosis, and a potential confounding bias due to age-based vaccine distribution policies, which might have contributed to the older age of the VITT and non-VITT groups compared with the historical cohort of patients with cerebral venous sinus thrombosis (median age 37 years).

The analysis by Perry and colleagues represents a landmark study, which is, to our knowledge, the largest thus far of VITT-associated cerebral venous sinus thrombosis, and the first to directly compare the clinical, laboratory, and radiographic features of VITT-associated and non-VITT-associated cerebral venous sinus thrombosis. The poor outcomes of VITT-associated cerebral venous sinus thrombosis highlight the need for accurate diagnostic tools to guide early recognition of this highly morbid condition. Additional studies are warranted to further guide treatment and management of VITT with the hope of improving outcomes for patients with this rare complication.