Incidence and Localization of Vascular Events in a Well Characterized Cohort of 477 Patients with Myeloproliferative Neoplasms

Myeloproliferative neoplasms (MPNs) are hematopoietic disorders characterized by bone marrow and blood myeloid cells proliferation. MPNs are secondary to the acquisition of clonal mutation. Polycythemia vera (PV) cases are mutated in JAK2V617F in 97% and JAK2 exon12 in 3%. Essential Thrombocythemia/ Primary Myelofibrosis (ET/PMF) cases are JAK2V617F mutated in 50-60%, CALR mutated in 20-30% and MPL mutated in 5-10%. Thromboses are the most frequent mid term complications in MPNs, and in most series the ratio arterial/venous thrombosis is 2/3 vs 1/3. The patients carrying CALR mutations have been recently reported as presenting a lower incidence of thrombosis, perhaps due to a lower leukocytes count and hemoglobin level.

We report a cohort of 628 MPNs (ET and PV) patients followed in a single institution. JAK2, CALR and MPL mutational status is available to date in 477 cases. PV cases were divided into PV and masked PV (mPV; normal Hb/Ht level but increased red cell mass and high platelets count). Among the 477 patients, we identified 158 pts (33.1%) with PV, 57 pts (11.9%) with mPV, and 262 pts (54.9%) with ET. All PV and mPV cases were JAK2 mutated. In the ET group, 184 pts (70.2%) carried the JAK2 V617F mutation, 40 pts (15.3%) a CALR mutation, 5 pts (1.9%) a MPL mutation, and 33/262 pts (12.6%) were triple negative (JAK2V617F, CALR, MPL negative). All vascular events have been collected and the subtype (arterial or venous) and sites of thrombosis have been registered. Only the thrombotic events which occurred in the 2 years preceding the diagnosis of MPN were considered to be related to the MPN.

We observed a similar incidence of thrombosis before and after diagnosis of MPN (22%, n=105 vs 21%, n=100). Among the 105 patients with prior thrombosis, 119 thrombotic events were recorded; during the follow-up, the 100 pts have presented 128 events of thrombosis. About one third of these ‘post-diagnosis’ episodes were thrombotic recurrences. There were no significant differences in term of incidence of thrombosis before and after diagnosis according to the different phenotypes (PV 37 pts / 22% both; ET 19.1%, n=50 vs 19.8%, n=52) except for mPV: 18 pts (31.6%) with prior thrombosis vs 11 patients (19.3%) who presented a thrombosis after. This remained true among ET patients even when deciphering by the mutational status: JAK2V617F positive cases (24.5%, n=45 vs 23.9%, n=44), CALR positive (7.5%, n=3 vs 10%, n=4), MPL positive or triple negative cases (6.1%, n=2 vs 9.1%, n=3). We observed only 9 episodes of thrombosis in CALR mutated ET patients: 7 arterial events (3 cases of transient ischemic attack, 2 cases of ischemic stroke, 2 cases of acute myocardial infarction and 1 mesenteric ischemia) and 2 cases of venous thrombosis (1 deep venous thrombosis and 1 cerebral venous thrombosis).

The frequency of CALR mutations was lower in our cohort than described in the literature, but allowed us to confirm the overall lower incidence of thrombosis in that subgroup of ET. In addition to previously published data we showed that venous thrombosis were extremely rare and that the majority of vascular events were arterial in CALR mutated patients who presented a thrombosis. Cerebral and myocardial infarctions were predominant. In parallel, we have performed a systematic CALR mutational analysis on 394 pts presenting a first unprovoked venous thrombosis without identify any mutant.

Here we confirmed the very high incidence of thrombosis in MPNs, especially in JAK2 mutated cases, and showed that the frequency of vascular events remains very high even after the diagnosis of the MPN, demonstrating the lack of effective prevention of first episode or recurrence in this high-risk population. Future therapeutic strategies should better take into account all risk factors including the mutational status and the type of thrombotic events.