Cerebral Vein Thrombosis In Patients With Myeloproliferative Neoplasms

Patients with Philadelphia-negative myeloproliferative neoplasms (MPN) can develop venous thrombosis and MPN are the leading cause of splanchnic vein thrombosis. Cerebral vein thrombosis (CVT) is a rare life-threatening disease that in approximately 3% of cases encounters MPN among risk factors and tends to recur in 2-4% of patients as CVT and in 4-7% as venous thrombosis at other sites. Little or no information is available on patients with MPN who develop CVT.

To investigate the characteristics and clinical course of CVT in patients with MPN we carried out a multicenter (n=11), observational, retrospective cohort study.

Centers were asked to provide information on index patients with MPN who developed CVT (group MPN-CVT). For each of them, 2 patients with MPN and venous thrombosis other than CVT (group MPN-VT) and 4 patients with MPN and no venous thrombosis (group MPN-NoVT) were provided, matched by sex, age at diagnosis of MPN (+/-5 years) and type of MPN (polycytemia vera, essential thrombocytemia, myelofibrosis) with index patients.

From January 1982 to June 2013, 48 MPN-CVT, 87 MPN-VT and 178 MPN-NoVT patients were identified in a population of 5,500 patients with MPN. Diagnosis of MPN and thrombosis coincided in 46% of MPN-CVT and 29% of MPN-VT patients (p=0.046). Compared to MPN-NoVT, MPN-CVT and MPN-VT patients had a higher prevalence of thrombophilia abnormalities (40% and 35% vs 21%, p=0.015) and, among those with essential thrombocytemia, of the JAK2 V617F mutation (76% and 78% vs 55%, p=0.059). Compared to MPN-VT, MPN-CVT patients had a higher rate of recurrent thrombosis (42% vs 25%, p=0.049) that in two-third of patients in both groups was venous, with a similar site distribution. This difference occurred despite a shorter median follow-up period (6.1 vs 10.3 years, p=0.019), a higher proportion of patients on long-term antithrombotic treatment (94% vs 84%, p=0.099) and a similar proportion of patients on cytoreductive treatment (75% vs 72%, p=0.745) among MPN-CVT than MPN-VT patients. The incidence of recurrent thrombosis was 8.8% patients/year in MPN-CVT and 4.2% patients/year in MPN-VT patients (log-rank test, p=0.022) and CVT was the only variable in a multivariate model including blood counts, thrombophilia, cytoreductive and antithrombotic treatment, that was predictive of recurrent thrombosis (HR 1.86, 95%CI 1.00-3.58).

Patients with MPN develop recurrent thrombosis in a much higher proportion than those without, particularly if they had a CVT. Patients with MPN and CVT have an approximately 2-fold increased probability to develop recurrent thrombosis than those with MPN and venous thrombosis at other sites, independently of other risk factors.

No relevant conflicts of interest to declare.