The JAK2 V617F mutation is present in almost all Polycythemia Vera (PV) and in 60-70% of Essential Thrombocythemia (ET) cases. There are experimental evidences that this acquired clonal driving mutation is involved in disease phenotype in murine models and in humans. However, it is still unclear if the %V617F changes over time and if it is necessary to systematically quantify %V617F at diagnosis and during follow-up.

Patients and Methods

We are conducted a prospective multicentre (n=5) study cohort in newly diagnosed PV and ET patients, harvested before any cytotoxic therapy and at 3 years, with the aim to correlate the %V617F at diagnosis, the %V617F 3 years later, and the variation of the %V617F between y0 and y3 with clinical evolution. Primary or secondary myelofibrosis, and secondary leukemia were not included.

The primary objective is to determine whether an increase of %V617F between diagnosis and after 3 years could be related to a poor evolution of the disease. A secondary objective is to determine whether the %V617F at diagnosis is associated with the evolution at 3y. The primary endpoint is the worsening (Y / N) at y3 defined by the presence of at least one of the following criteria: leukocytosis >12G / L or immature granulocytes >2% or erythroblasts >1%; anemia or thrombocytopenia not related to treatment toxicity; development or progressive splenomegaly; thrombocytosis on cytoreductive therapy; patients inadequately controlled with treatment (defined by at least one treatment change for reason other than an adverse event).


To date, 201 patients have been included. Clinical datas are currently available for the 176 first included (91 PV and 85 ET; 90 females and 86 males) and %V617F measurement for 167 cases.

At diagnosis, overall median age was 65, identical for PV and ET patients, but curiously women were a little older than men (72 vs 65, p<0.001). Family history of hematological malignancies was identified in 11% of the cohort. Thrombotic events (arterial or venous) were noted in 35% before or at the time of MPN diagnosis. A cytotoxic therapy has been introduced during the first 3 months after diagnosis in 73.9% of cases. After 3 years 10/91 PV patients still need phlebotomies and only 20 patients remains without any myelosuppressive therapy. The median %V617F at diagnosis was 27.5%. As expected the median %V617F was significantly higher in PV than in ET pts (42.7% and 16.8% respectively, p <0.0001) and a high %V617F is correlated with the presence of a splenomegaly, asthenia, pruritus, a higher hemoglobin level, and a more frequently need for cytoreductive therapy (29% versus 17%, p=0.01).

After 3y of evolution, clinical evaluation reveals 36 major events:

- 3 patients experienced an hematologic transformation: 2 post-PV myelofibrosis and one ET which has evolved in PV and none secondary acute leukemia. These 2 transformed-PV displayed a very high initial %V617F (over 90%) that remained high during evolution.

- 10 patients (5,6%) experienced a poor evolution (7 PV and 3 ET) as defined in the method section. Interestingly, these 10 patients who worsened have an higher initial %V617F than others (44.17% versus 25.73% p=0.03). This seems to be true when we look at PV and ET separately, but the results did not reach statistical significance probably due to the low number of patients. When %V617F at y3 is used the difference increased (40.9% vs 17.4%, p=0.002).

- 23 new vascular events occurred: thrombosis in 17 cases and hemorrhages in 6 cases.

Finally, we identified three distinct molecular evolutive profiles that represent each one third of the cases: stable %V617F, increasing %V617F and decreasing %V617F. It is too early to draw definitive conclusions about clinical evolution and these groups, but first analysis suggest that patients who have an unfavorable evolution are more common among those with an increased %V61F during follow-up.

Discussion and conclusion

Despite the short 3-years follow-up, we already recorded several major events and found that a high %V617F at diagnosis is correlated with a poorer evolution of the disease, suggesting that these patients could have a more advanced disease despite a classical hematological phenotype of ET or PV. Furthermore, the increase of %V617F seemed higher in evoluted PV in comparison to stable PV. In contrast, we didn't found any correlation between %V617F and thrombotic or bleeding events.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution