Incidence of arterial thromboembolism in patients with myeloproliferative neoplasms treated with anticoagulation for venous thromboembolism

Abstract Background Patients with myeloproliferative neoplasms (MPN) have a high lifetime risk of both venous thromboembolism (VTE) and arterial thromboembolism (ATE), often leading to long-term anticoagulation. Rates of thromboembolism while on anticoagulation vary from 4.5% to 9.0%, but these data include patients with various anticoagulation indications and prior ATE. Aim To investigate the incidence of ATE in MPN patients on anticoagulation for VTE. Methods We conducted a retrospective study of patients at Memorial Sloan Kettering Cancer Center (MSKCC) and Beth Israel Deaconess Medical Center (BIDMC) from 2000 to 2025 with manual chart review. Patients were included if they had a history of polycythemia vera (PV), essential thrombocytosis (ET), or MPN-Not Otherwise Specified (MPN-NOS); and had a VTE within 6 months prior to MPN diagnosis, or any time thereafter. Patients were excluded if they were not on therapeutic dose anticoagulation after index VTE, had ATE prior to index, or had a different indication for antiplatelet therapy within the last year (e.g. vascular stent). Patients were indexed on date of first MPN-associated VTE and censored for death, transformation to leukemia/myelofibrosis, stopping anticoagulation, primary outcome of ATE, loss to follow up, and last date of record access. The primary outcome was ATE occurrence at any site (acute coronary syndrome, ischemic stroke, visceral/peripheral ATE). Secondary outcomes included non-surgical major bleeding as defined by the International Society on Thrombosis and Hemostasis. Baseline variables were documented at index. Antiplatelet and cytoreductive therapy was documented at 30 days post-index to allow for changes following the index VTE. We recognize that concomitant aspirin treatment (defined as aspirin documentation in the 30 days of index) may be a surrogate for patients at high cardiovascular risk and performed a subgroup analysis stratified for aspirin exposure. Outcomes were reported as cumulative incidences with 95% confidence intervals (CI) and death, stopping anticoagulation and transformation as competing risks. Results A total of 105 patients were included. Median age was 71 years (IQR 61-78); 63 (60%) were male. MPN types included ET (50, 48%), PV (47, 45%), ET/PV overlap (2, 2%), MPN-NOS (6, 6%). JAK2 617F mutation was positive in 84 (80%) patients. Median hemoglobin and platelets at index were 12.7 g/dl (IQR 10.4-14.3) and 416 x 109/L (IQR 317-614) respectively. Cardiovascular risk factors included hypertension (59, 56%) and hyperlipidemia (41, 39%). Index VTE site was pulmonary embolism and/or lower extremity deep vein thrombosis (DVT) in 84 (80%) patients, upper extremity DVT in 7, (7%), inferior vena cava or splanchnic vein thrombosis in 10 (10%) and cerebral vein thrombosis in 4 (4%). Therapeutic-dose anticoagulation for index VTE included direct oral anticoagulants (56, 53%), warfarin (36, 34%) and enoxaparin (13, 12%). At 30 days, 29 (28%) patients were also on antiplatelet therapy (all aspirin) and 56 (53%) were receiving cytoreductive therapy (53 hydroxyurea, 3 ruxolitinib). Aspirin treatment was associated with PV, use of hydroxyurea, aspirin treatment prior to VTE and numerically more hypertension. The cumulative incidence of ATE was 2.0% (95% CI 0.4-6.3) at 6 months, 4.0% (95% CI 1.3-9.3) at 1 year and at 2 years, and 6.4% (95% CI 2.6-13.0) at 3 years. Overall, 12 ATE events occurred on anticoagulation in 11 patients (one patient had ischemic stroke and splenic infarct concomitantly), including ischemic stroke (6), acute coronary syndrome (3), splenic infarct (1), limb arterial thrombus (1), and mesenteric ischemia (1). The ATE incidence in patients treated with aspirin was 6.9% (95% CI 1.2%-20.0%) at 6 months which was numerically higher than patients not treated with aspirin on Cox proportional hazard analysis (hazard ratio 3.14, 95% CI 0.92-10.76). Overall, the cumulative incidence of major bleeding was 1.9% (95% CI 0.4-6.2) at 6 months, 2.9% (95% CI 0.8-7.7%) at 1 year, and 4.0% (95% CI 1.3-9.2) at 2 years and at 3 years. Conclusions In this cohort, patients with MPN had a clinically significant ATE incidence of 6.4% at 3 years despite anticoagulation for VTE. ATE risk remained high in a subgroup of patients on aspirin treatment. The overall ATE risk was numerically higher than the risk of major bleeding. These results highlight the need for research on ATE risk reduction in patients with MPN on anticoagulation.