Cardiovascular Risk in Polycythemia Vera: Thrombotic Risk and Survival: Can Cytoreductive Therapy Be Useful in Patients with Low-Risk Polycythemia Vera with Cardiovascular Risk Factors?

The diagnosis of polycythemia vera (PV) often requires the presence of JAK2 mutation, in addition to the increase of hemoglobin/hematocrit ratio, whose threshold levels have been established by the 2016 World Health Organization (WHO) revised criteria (>16.5 g/dL or >49% for males and >16 g/dL or >48% for females) [1, 2]. Almost all patients with PV harbor a JAK2 mutation; approximately 96 and 3% of them display somatic activating mutations in exon 14 (JAK2-V617F) and JAK2 exon 12, respectively. In a recently published study conducted on 826 Mayo Clinic patients with essential thrombocythemia (ET), PV, or primary myelofibrosis, the respective median survivals were approximately 20 years for ET, 14 years for PV, and 6 years for primary myelofibrosis [3]. The increased risk of vascular complications over time is the main clinical feature of PV. With regard to PV, two risk categories are defined. In particular, the low-risk category applies for patients without a history of thrombosis and younger than 60 years, and the high-risk category for patients older than 60 years and/or with a history of thrombosis. All patients with PV should be managed with phlebotomy to maintain the hematocrit at <45% and take low-dose aspirin. Cytoreduction is indicated in high-risk patients. Current classifications of the thrombotic risk in patients with PV do not consider cardiovascular risk (CVR), and therefore the presence of CVR factors does not currently influence the choice of a cytoreductive therapy. In this study, we evaluated the frequency of CVR in a group of patients with PV and the possible impact of the CVR on survival.

From January 1997 to May 2019, 165 consecutive patients with PV were followed at our institution with a median follow-up of 58.18 months (0.3–289.3 months). We evaluated, with retrospective analysis, the main characteristics of the study population such as gender, age, and mutational status along with CVR frequency, considering cigarette smoking habits, hypertension, diabetes, obesity, and dyslipidemia. In particular, patients with only one of these conditions were distinguished by those with none or >1 cardiovascular risk factor. Moreover, the correlation of these cardiovascular risk conditions with the onset of thrombosis has been evaluated. The frequencies were calculated by using the χ² method and the medians comparison through the Kruskal-Wallis test. Furthermore, the Kaplan-Meyer method was used to evaluate survival, and the comparison between survival curves was studied with the log-rank test.

The main features including sex, median age, and the mutational status, along with the cardiovascular risk factors of the cohort under examination were, respectively, summarized in Tables 1 and 2. In particular, 37 PV patients (22.42) have no CVR, while 66 PV patients (49%) have only one cardiovascular risk factor, and 62 (37.57%) have >1 CVR. In patients with PV, we highlighted 49 (29.69%) cases of thrombosis at diagnosis or before diagnosis and 16 (9.69%) cases of thrombosis after diagnosis. Overall, PV patients show a thrombosis frequency of 65/165 (39.39%).

In the examined patients, cytoreduction was prescribed in patients with high-risk PV, according to ELN recommendations. Of the 119 patients with high-risk PV, 95 received hydroxyurea (HU) as first-line cytoreductive therapy, while 22 were prescribed with ruxolitinib after first-line HU because of HU intolerance (18 patients) or HU resistance (4 patients). All of the 165 patients received antiplatelet agents (117 patients) or anticoagulant drugs (65 patients). Between the antiplatelet agents prescribed, aminosalicylic acid was the most used (in 84 patients). The remaining 33 patients treated with antiplatelet agents received ticlopidine (27 cases) or clopidogrel (6 cases). Between the 65 patients treated with anticoagulant therapies, 44 received anti-vitamin K agents (warfarin or dicoumarol) and 21 patients received direct oral anticoagulants.

In our patients with PV, the frequency of thrombotic episodes is strictly correlated with cardiovascular risk factors. In fact, in PV patients without CVR, thrombosis is present in 8 of 37 cases, and 21 of 66 cases in patients with only one CVR, and 26 of 62 cases in patients with >1 CVR. Our data also show a significant correlation between cardiovascular risk factors and survival in the cohort of patients with PV, distinguished by number of cardiovascular risk factors (Fig. 1).

Patients with myeloproliferative neoplasms (MPNs) have an increased risk of thrombotic events if compared with the general population, with a greater risk of MPNs-associated morbidity and mortality [4, 5]. A recent single-center study conducted on 526 patients with MPNs with an overall study period of 3,497.4 years, reported an incidence rate of 1.7% of venous thrombosis per patient/year [6]. Overall, 38.4% of all venous thromboses occurred before or at diagnosis of MPNs, with 55.6% occurring at uncommon sites such as splanchnic or cerebral veins. PV is a MPN respectively characterized by erythrocytosis and thrombocytosis; other clinical features include leukocytosis, splenomegaly, thrombosis, bleeding, microcirculatory symptoms, pruritus, and risk of leukemic or fibrotic transformation. Moreover, thrombosis and cardiovascular disease are more prevalent in PV than in other myeloproliferative disorders. It has been estimated that 30–50% of PV patients have minor or major thrombotic complications, and vascular mortality accounts for 35–45% of all PV-related deaths [7]. Determining the thrombotic risk in PV has a pivotal importance to achieve a proper therapeutic choice. In PV, two risk categories are considered: high-risk (age >60 years or history of previous thrombosis) and low-risk (absence of both risk factors). In a previously proposed thrombotic risk classification model, an intermediate-risk category was added to the traditional high-risk and low-risk category, fitting patients aged <60 years, with no history of thrombosis, in presence of cardiovascular risk factors [8]. This thrombotic risk classification model did not achieve higher consideration, and it is not currently used in clinical practice. Cerquozzi et al. [9] explored the association of cardiovascular risk factors with the occurrence of arterial or venous events at or following diagnosis: they found that older age (≥60 years), hypertension, diabetes, hyperlipidemia, and normal karyotype were associated with arterial events, whereas younger age (<60 years), female sex, palpable splenomegaly, and history of major hemorrhage were associated with venous events. Today, for patients with PV aged <60 years and with one or more cardiovascular risk factors, there is no indication for cytoreductive therapy. In ET, the IPSET-thrombosis system that includes age, previous thrombosis, cardiovascular risk factors, and JAK2-V617F mutation, is the recommended prognostic system and it should be scored in all patients at diagnosis. This implies that general risk factors for thrombosis, including smoking habits, diabetes mellitus, arterial hypertension, and hypercholesterolemia, should also be considered, even in the absence of specific therapeutic indications [10]. Recently, a panel of experts is questioning the traditional classification of thrombotic risk in PV; a matter of discussion is whether these definitions predicting thrombotic risk are still valid and whether the newly proposed disease-related risk factors, such as cardiovascular risk factors, leukocytosis, JAK2V617F allele burden, may improve the grading system of the thrombotic risk [11].

According to our experience, it should be useful to design perspective studies to determine the real influence of CVR on the thrombotic risk in patients with PV and on survival in order to evaluate the opportunity to develop new specific therapeutic recommendations, such as early cytoreduction, for patients with PV aged <60 years and having cardiovascular risk factors.

This study was approved by our hospital’s Ethics Committee (“Palermo 1”). Patients have given informed consent to participate to this research.

S.M., F.P., G.A., A.P., and M.B.: none, M.S.: Honoraria from Novartis, Pfizer, and Incyte. V.A.: Honoraria from Novartis and Incyte. S.R.: Honoraria from Amgen and Sobi. A.R.: Honoraria for Advisory Board form Bristol-Myers Squibb, Pfizer, Bayer, Kyowa Kirin, and Ambrosetti, and speaker honorarium from Roche Diagnostics. S.S.: Honoraria from Sobi, Baxter, Amgen, NovoNordisk, and Novartis.

There are no funding sources to declare.

S.M.: study design, writing of the paper, and revision. M.S.: collecting data and paper revision. V.A.: data elaboration and writing of the paper. G.A. and S.R.: data collection. F.P., A.P., and M.B.: data generation and collection. A.R. and S.S.: paper revision.