Leukocytosis Can Predict Thrombotic Events in Myelofibrosis

Thrombotic events are the main complication of Philadelphia-negative chronic myeloproliferative neoplasms (MPN). In polycythemia vera (PV) and essential thrombocythemia (ET), risk factors for thrombosis are well established, such as age greater than 60 years and previous thrombosis. However, the role of JAK2 V617F mutation and leukocytosis at diagnosis as risk factor for thrombosis is still controversial. Our aim was to identify factors related to the risk for thrombotic events in the studied population.This study is a retrospective non-interventional cohort. All of the analyses were performed using the database of 142 patients with MPN regularly followed at the Hematology Division (at UNIFESP-SP) from 1992 to 2014. Diagnosis was established according to WHO criteria. We analyzed the JAK2 V617F mutation, hemoglobin (g/dL), hematocrit (%), white blood cells (x10⁹/L) and platelets (x10⁹/L) at the diagnosis and DIPSS-Plus risk score (International Working Group for Myelofibrosis Research and Treatment, 2009). These variables were associated with thrombotic event at any time.Of the 142 patients, 54 had diagnosis of PMF, 28 of PV, 33 of ET and 27 of post-essential thrombocythaemic myelofibrosis (post ET MF) or post-polycythaemic myelofibrosis (post-PV MF). This last group was included in myelofibrosis group for statistical purposes.

Thrombotic events were more frequent in PV patients (39.2%), followed by ET (33.3%), and PMF (20.9%).

From those which JAK2 mutation was obtained, it was positive in 92.4% of PV patients, 62% of PMF and 50% of ET. In none of the three groups, the presence of JAK2 V617F mutation was related to increased risk of thrombosis.

In myelofibrosis, leukocytosis was higher among thrombotic patients (median of 13.7 in thrombotic group versus 9.7x 10⁹/L; p 0.0379). None of the other parameters, hemoglobin, hematocrit, platelets and DIPSS-Plus were statistically significant. In ET, the hemoglobin level at diagnosis was significantly higher in the presence of thrombosis (mean of 14.57 in thrombotic group against 13.03 g/dL in the non-thrombotic one, p 0.0428). The other parameters, hematocrit, white blood cells and platelets were not relevant. The median WBC in the thrombotic group was 9.4 and in the non-thrombotic one 9.3 x10⁹/L. Finally, in polycythemia vera, none of the variables were related to thrombosis. Among the studied population, leukocytosis was increased in patients with thrombotic event in MF. Thus, monitoring leukocyte count in MF is essential to predict thrombosis risk and should be further studied in order to define therapeutic goals in these patients.