Unusual Site Thrombosis: Focus on Myeloproliferative Neoplasms with Splanchnic or Cerebral Venous Thrombosis

This study retrospectively evaluated 144 consecutive patients with unusual site thrombosis who referred to our Thrombosis Center between 2000 and 2016. All patients were classified as having either splanchnic venous thrombosis (SVT; n=127) or cerebral venous thrombosis (CVT; n=17). On the presence and type of provoking risk factors, then patients were categorized into three groups: unprovoked, those with possibly resolved provoking factors (PR), and those with persistent provoking factors (PP). Among the identified risk factors regarded as PP, we focused on Myeloproliferative Neoplasms (MPN) and performed a clinical comparison between MPN patients with SVT (MPN-SVT) and those with CVT (MPN-SVT). The characteristics of our cohort well reflects the clinical heterogeneity of clinical features commonly found in the routine clinical practice of diagnosing unusual site thrombosis. One hundred and twenty seven SVT patients were included: 7 unprovoked SVT, 10 SVT with PR, 110 SVT with PP; seventeen patients showed CVT, 5 unprovoked, 6 CVT with PR and 6 CVT with PP. Major risk factor for SVT with PP was liver cirrhosis (71.6%), whereas for CVT were MPN (5 patients, 29.4%). MPN was present in 8 patients (6.2%) of SVT (MPN-SVT vs MPN-CVT, p=0.009). For MPN-SVT, 3 patients showed the morphological features of polycytemia vera (PV), 2 of essential thrombocytemia (ET), 1 of primary myelofibrosis (PMF) and 2 fell into the MPN unclassified category (U-MPN); distribution of MPN subtypes for CVT was as follows: 1 PV, 2 ET and 2 U-MPN. Molecular analysis identified the JAK2V617F mutation respectively in 75% patients with MPN-SVT and in 60% patients with MPN-CVT; bone marrow histological features supported the diagnosis of MPN in all cases. Median age at MPN-SVT diagnosis was 46 years (range 17-78) vs 43 years (range 31-84) for MPN-CVT. Coexisting PR and thrombophilic abnormalities were identified in 75% and 20% of MPN-SVT and MPN-CVT respectively, so MPN are per se a strong risk for thrombosis and the leading systemic cause of CVT. In four of five cases (80%), CVT antedate the clinical phenotype of an overt MPN, whereas SVT occurred at MPN diagnosis in five patients and during MPN follow-up in three patients (median 164 months, range 88-215). At diagnosis MPN-SVT tended to have significantly lower platelet and white blood cell counts and haemoglobin concentration than MPN-CVT (respectively p<0.001, p<0.001 and p=0.002). Similar proportion of MPN patients in the two groups received cytoreductive treatment (hydroxyurea and alpha interferon) and appropriate anticoagulant therapy that consisted of low molecular weight heparin (LMWH) followed by oral vitamin K antagonists (VKA). MPN-SVT and MPN-CVT experienced similar median duration of anticoagulation (26 months for MPN-SVT, range 1-62, and 26 months for MPN-CVT, range 4-168) and a good quality anticoagulation control (median time within therapeutic range of 71 vs 87% respectively for MPN-SVT and MPN-CVT). Seventy-five percent MPN-SVT and 80% of MPN-CVT remain on indefinite anticoagulation. All patients were alive at last follow-up and the results of imaging techniques showed resolution of thrombosis in 87.5% and 80% of MPN-SVT and MPN-CVT respectively; the probability of recanalization of the occluded vessels was 18 months and 4 months for MPN-SVT and MPN-CVT, respectively. Only in one patient with MPN-SVT, a major bleed occurred on-treatment while the incidence of recurrent thrombosis was the same for both MPN-SVT and MPN-CVT (0.02 per 100 patients-year). In conclusion, our study evaluated unselected populations with unusual site thrombosis that were followed in our Thrombosis Center, an anticoagulation clinic well experienced on pathogenic mechanisms and anticoagulant treatment. Our findings have practical implications and point out the role of MPN as a major contributory factor for the pathogenesis of CVT with PP, even in absence of overt myeloproliferative features and additional prothombotic factors. Regarding management of vascular complications, patients with MPN-SVT and MPN-CVT responded equally well and efficacy of anticoagulation was not affected by the site of thrombosis.