Background: Patients (pts) with myeloproliferative neoplasms (MPN) have a significantly increased risk of arterial and venous thrombotic events (TEs), while some pts may paradoxically suffer from bleeding manifestations. Thrombosis risk evaluation in polycythemia vera (PV) and essential thrombocythemia (ET) has important implications for patient management, however, the currently used risk stratification models are extremely limited.
Methods: We designed a pilot study to assess the utility of an expanded coagulation laboratory profile in addition to other clinical variables for thrombosis and bleeding risk evaluation in a heterogenous MPN patient population. Our laboratory profile included PT, aPTT, fibrinogen, D-dimer, von-Willebrand Factor (vWF) antigen and activity testing, platelet (PLT) aggregation studies and thrombin generation assay (TGA, ST-Genesia, Stago, France), along with CBC and comprehensive biochemistry studies. Blood was drawn from pts visiting the hematology clinics or day care at Hadassah Medical Center. Clinical and demographic data were collected retrospectively from electronic medical records. The study was approved by the local IRB committee.
Results:74 pts were included in the study; the mean age was 59.7 years and 51% were women. Sixty percent had a diagnosis of PV, 20% had ET, 17.6% had myelofibrosis (MF) and 2.7% had MPN unclassifiable. 78% of the pts harbored a JAK2V617F mutation, 8% and 2.7% harbored CALR and MPL mutations, respectively, and 11% were negative to all three MPN driver mutations. 63% of the pts had at least one cardiovascular risk factor, most commonly hypertension (HTN) and dyslipidemia. Most of the pts were treated with antiplatelet agents (62%), anticoagulants (16%) and/or cytoreductive agents (58%). Of 74 pts, 31% had a history of TEs (20% arterial, 11% venous), 22% reported suffering from microvascular symptoms, and 20% had experienced bleeding. The high rate of bleeding manifestations in our study population suggests an underestimation of the bleeding burden in MPN, as proposed by our group previously [Rottenstreich, Eur J Intern Med 2017]. Other patient characteristics were overall comparable to MPN patient populations in other studies (Table 1).
In multivariate analyses, history of TEs was significantly associated with older age, type 2 diabetes mellitus (DM2), HTN and dyslipidemia, but not with gender, MPN subtype, driver mutation status or smoking. Pts with history of TEs had significantly lower hematocrit and prolonged PT/INR, associated with a higher rate of treatment with cytoreduction and anticoagulation, respectively, in these pts. Interestingly, pts with TEs had significantly higher vWF antigen levels (p=0.007), which were independent of age, blood counts and treatments, a finding not previously reported in MPN. No association was observed between TEs and PLT or WBC counts, D-dimer levels and vWF activity (Table 2), however, vWF activity was decreased in pts with higher PLT counts.
TGA was of key interest in this study. In 62 pts not receiving anticoagulation, we observed an association between prolonged lag time (suggesting decreased coagulability) and older age, male gender and (near significant trend) JAK2V617 negative status, but did not find a correlation to TEs, blood counts and other variables. PLT aggregation indices were not associated with TEs, bleeding history or blood counts.
In multivariate logistic regression analyses, history of DM2 and HTN, and vWF antigen levels, showed the strongest association to TEs, which was independent of age, blood counts and all other variables. A risk-model incorporating these variables showed a sensitivity of 77% and specificity of 92% for history of TEs in our study population.
Conclusions: Despite being limited by a small and heterogenous study population and retrospective data collection; these findings support the utility of a coagulation laboratory profile in the evaluation of thrombosis and bleeding risk in MPN. The potential of TGA in risk evaluation in MPN is intriguing but merits further research. The multivariate risk-model incorporating history of DM2 and HTN along with vWF antigen levels merits validation in larger studies. Based on this pilot, a prospective study following over 250 pts with PV and ET (excluding overt MF) was initiated. This study will include NGS mutational profiling and JAK2V617F allele burden testing and omit the use of PLT aggregation studies.
Disclosures
No relevant conflicts of interest to declare.
