Cytoreductive therapy reduces recurrent thrombosis in patients with myeloproliferative neoplasm-associated splanchnic vein thrombosis: A global abdominal/splanchnic thrombosis retrospective observational study (GASTRO-MPN) study of 757 patients

Abstract Background Patients with myeloproliferative neoplasms (MPNs) are at risk of splanchnic vein thrombosis (SVT). Prior studies failed to demonstrate a benefit of hydroxyurea in preventing SVT extension/recurrence, but these were limited by small sample sizes and inability to account for starting, stopping, or switching of cytoreductive therapies. We therefore investigated the role of cytoreduction as a time-dependent covariate on thrombosis recurrence in a large, international cohort of MPN-SVT patients. Methods We performed a retrospective cohort study of MPN-SVT patients included in the international GASTRO-MPN consortium. Patients diagnosed with MPN at any time before, concurrently with, or up to 60 days after initial SVT were included. Primary outcome was time from initial SVT to first thrombosis recurrence, defined as the composite of SVT extension/recurrence, transjugular intrahepatic portosystemic shunt (TIPS) thrombosis, or venous/arterial thrombosis outside the splanchnic veins. Secondary outcomes included SVT extension/recurrence, TIPS thrombosis or extra-SVT thrombosis individually, clinically relevant bleeding (composite of major and clinically relevant non-major bleeding), MPN progression to myelofibrosis or blast phase, and death. Cytoreductive therapy was treated as a time-dependent covariate to account for discontinuation and therapy changes. Multivariable Cox proportional hazard models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CI). Results A total of 757 patients were included for this analysis. Median age was 51 years (IQR 38-62), 59% were female. SVT involved the portomesenteric veins in 558 (77%) patients, hepatic veins in 95 (13%), and both in 54 (7.4%). MPN was known prior to initial SVT in 316 patients (42%) and diagnosed concurrently or up to 60 days after SVT in 441 (58%). Polycythemia vera (PV) was the most common MPN subtype (43%), followed by essential thrombocythemia (ET) in 24%, and primary myelofibrosis in 14%. JAK2V617F mutation was present in 86% and CALR in 2.9%. 13% had prior non-SVT thrombosis. 607 patients (83%) were treated with cytoreduction. At time of SVT, 131 patients (17%) were already on cytoreduction; in those who started after SVT, median time to initiation was 54 days after SVT, and median duration of cytoreduction per patient was 63 months. Cytoreductive therapies used at any point in treatment included hydroxyurea in 504 (69%), ruxolitinib in 182 (25%), and interferon-based in 85 (12%) patients. Anticoagulation (AC) was started within 30 days after SVT in 504 (68%) patients. Median duration of any AC per patient was 24.6 months. During a median follow up of 7.4 years, 245 patients (33%) had thrombosis recurrence; 143 (20%) had SVT extension/recurrence/TIPS thrombosis, 129 (18%) extra-SVT thrombosis (of which 74% venous), 237 patients (32%) experienced clinically relevant bleeding, and 163 (22%) died. After adjusting for age, sex, timing of MPN diagnosis relative to SVT diagnosis, AC within 30 days of SVT, PV/ET vs other MPN, JAK2 mutation, previous thrombosis, non-MPN cancer, and cirrhosis, cytoreductive therapy was associated with a significantly decreased risk of recurrent thrombosis (HR 0.67 95%CI 0.50-0.89, p=0.01). Age was associated with an increased risk of recurrent thrombosis (p=0.01). When considering types of recurrent thrombosis, cytoreduction was associated with a decreased risk of SVT extension/recurrence/TIPS thrombosis (HR 0.62 95%CI 0.42-0.90, p=0.01), but not thrombosis outside the splanchnic veins (HR 0.97 95%CI 0.72-1.30, p=0.82). There was no association between cytoreduction and bleeding (HR 1.10 95%CI 0.82-1.47, p=0.52). Similarly, there was no association between cytoreduction and time to MPN progression (p=0.24) or overall survival (p=0.30). In a subgroup of patients without erythrocytosis/thrombocytosis at initial SVT (n=271), cytoreductive therapy was associated with a HR 0.65 (95%CI 0.40-1.10, p=0.09) for recurrent thrombosis. Limited sample size and low event-rate precluded evaluation of the impact of individual cytoreductive therapies. Conclusions In the largest cohort of MPN-SVT patients to date, we show for the first time that cytoreductive therapy is associated with a decreased risk of thrombosis recurrence, driven largely by decreased SVT extension/recurrence/TIPS thrombosis. Collectively, these findings support use of cytoreductive therapy in individuals with MPN-SVT to prevent recurrent thrombosis.