Ten Years of Cerebral Venous Thrombosis (CVT) in Melbourne, Australia: Male Gender and Presence of Myeloproliferative Neoplasm Is Associated with Thrombotic Recurrence in Unprovoked Events

Aim

Cerebral venous thrombosis (CVT) accounts for 0.5-1.0% of all strokes and is a common cause of stroke in young people. The presentations are often heterogeneous and can be associated with significant morbidity and mortality. This review aims to evaluate our local experience in CVT compared to other venous thromboembolism (VTE) with a focus on risk factors for thrombotic recurrence.

Methods

Retrospective evaluation of consecutive CVT presentations from January 2005 to June 2015, at two major tertiary hospitals in Melbourne, Australia. Data collected included demographics, risk factors, management, complications, modified Rankin score (mRS) and mortality.

Results

52 patients (31 female, 21 male) with median age 9.5 (18-83) years, including 4 with cancer, presented with 53 episodes of CVT. Females were younger (32 vs 41 years, p=0.06). Typical presenting symptoms were headache (87%), nausea/vomiting (43%), visual disturbances (38%), focal neurological deficits (28%) and seizures (17%). All but one case was symptomatic, with 53% reporting symptoms in the preceding week. 18 (34%) failed to be diagnosed on initial presentation while 35% (13/37) of CT brain yielded false negative for thrombosis; all of which were subsequently diagnosed on magnetic resonance imaging (MRI) or CT angiography/venography. Commonly thrombosed sinuses included transverse/sigmoid (40%), superior sagittal (11%) or both (43%), with no location-dependent outcome differences. Nine (17%) had CVT-related haemorrhagic transformation and was associated with CVT-related death (2/9 vs 0/44; p=0.04).

28 episodes were provoked - twice more common in female (p=0.02) with 45% attributed to oral contraceptive pill(OCP). 44 patients (85%) had thrombophilia screen performed with 21% positivity. Median duration of anticoagulation was 6.5 months (8 remained on long-term); 78% treated with warfarin. Eight (15%) required intensive care support, while 2 patients required decompressive surgery. 12 (23%) were not followed up in our institutions. At last follow-up of the remaining 40, 2 (5%) had worsening mRS of ³ 2 compared to premorbid, 2 had CVT-related deaths and 2 succumbed to malignancy. 30% reported ongoing symptoms such as headaches, residual neurological deficits, seizures and memory impairment. There were three clot recurrences (1 CVT, 2 portal vein thrombosis) - all male with initial unprovoked events and were subsequently diagnosed with myeloproliferative neoplasm (MPN). Of the 3, one was positive for JAK2V617F mutation. Men with unprovoked CVT had a 20% risk of recurrence, significantly higher compared to women with unprovoked events (3/15 vs 0/10; p=0.02). Clot progression, defined as increased clot burden on repeat imaging, occurred in 2 patients - one was associated with MPN while another progressed in the setting of subtherapeutic anticoagulation post partum. There was one episode of Grade III bleeding (following a procedure) in addition to the 2 (4%) clot-related deaths discussed prior.

Table 1.compares the characteristics of CVT and other VTE previously audited by us.

Conclusions

CVT is rare and may be missed on initial presentation (34%)_with a high degree of clinical suspicion required to improve detection rate. Given there was 35% of CT brain had false negative, MRI or CT angiography is the preferred modality of investigation. It is more common in young people, particularly females on OCP. The presence of haemorrhagic transformation was associated with higher mortality. All thrombotic recurrences in this audit occurred in men with unprovoked events, who were subsequently diagnosed with MPN. This suggests the need for further evaluation, particularly for MPN in males with unprovoked events.