Risk of thrombosis in patients with essential thrombocythemia and polycythemia vera according to JAK2 V617F mutation status

Abstract

We compared the laboratory and clinical findings of 179 patients with essential thrombocythemia (ET) and 77 with polycythemia vera (PV) classified according to the presence of the JAK2 V617F mutation. A gradient between patients with JAK2 wild-type ET, JAK2 V617F ET and PV (all carrying the JAK2 mutation) was observed. The rate of thrombotic complications in JAK2-positive ET was significantly higher than in wild-type ET and not statistically different from that of PV patients.

The presence of the JAK2 V617F mutation divides essential thrombocythemia (ET) into two distinct subtypes. Patients with the mutation present higher hemoglobin levels, higher white cell counts and bone marrow hypercellularity. 1,4 An increased risk of thrombosis in mutated ET patients has been also reported by some investigators,1,2 but not by others.3,4 These observations have generated the hypothesis of a possible biological continuum between ET and polycythemia vera (PV).1 This interesting view would be further supported if the clinical vascular complications of JAK2 mutated ET were comparable to those of PV patients. At the best of out knowledge, a direct comparison between the thrombotic risk of patients with ET, stratified by JAK2 V617F status, and those with PV was not hitherto reported.

To tackle this issue, we evaluated the laboratory findings at presentation and the clinical complications at diagnosis and during follow-up in two parallel cohorts of 179 ET and 77 PV patients, regularly followed in our outpatient clinic, and classified according to the presence of the JAK2 V617F mutation. The following statistical analysis were carried out: (i) continuous variables were compared using the extension of the Wilcoxon rank-sum non-parametric test for trend across ordered groups; (ii) categorical variables were analyzed using the Pearson χ test; (iii) the risk of thrombosis (and the corresponding 95% confidence interval) was estimated using dummy variables coding in a logistic regression model; only the first thrombotic event occurred at presentation or during the follow-up for each patient was considered. The results are shown in Table 1. A highly significant trend across patients with JAK2 wild-type ET, JAK2 V617F ET and PV (all carrying the JAK2 mutation) was observed for all laboratory values. Hemoglobin and hematocrit levels as well as white cells number and activation parameters (polycythemia rubra Vera 1 gene expression and leukocyte alkaline phosphatase mean fluorescence intensity)5 progressively increase where-as platelet number decreases through the three disorders. In turn, a significant increase of thrombotic complications was registered in patients with JAK2 V617F ET (33%) and PV (43%) as compared with those with JAK2 wild-type ET (17%). Considering JAK2 wild-type ET as a reference group, the risk of thrombosis was 2.39 (95% CI 1.16–4.93) and 3.63 (95% CI 1.72–7.68) for JAK2 V617F and PV patients respectively. The maximum likelihood estimate of the odds ratio comparing JAK2 V617F ET and PV patients was 1.52, not statistically significant (p=0.18). The association between increased granulocyte PRV-1 and LAP expression and thrombosis is noteworthy. This finding is in line with our recent data showing that leukocytes of JAK2 mutated ET patients present a prothrombotic state indicated by a significant increased expression of surface tissue factor and fibrinogen and a tendency to form higher numbers of leukocyte-platelet mixed aggregates.6 These results may explain why hydroxyurea, a broad myelosuppressive drug, and not anagrelide, a megacaryocyte restricted inhibitory agent, is more effective in reducing the thrombotic events, particularly in JAK2 mutated ET patients.1,7

Table 1.Laboratory and clinical features of 256 patients with ET and PV according to their JAK2 mutation status.

Overall, the present analysis confirms that JAK2 mutation in ET identifies a distinct clinical entity with a biological phenotype intermediate between JAK2 wild-type ET and PV and for the first time presents a comparison between the thrombotic risk of ET and PV patients defined on the basis of their JAK2 mutational status.