Abstract

An acquired V617F mutation in JAK2 occurs in most patients with polycythemia vera (PV) but only half of those with essential thrombocythemia (ET) and idiopathic myelofibrosis. It is not known whether mutation-bearing patients are biologically distinct from those lacking the mutation, or why the same mutation is associated with different phenotypes. Two sensitive PCR-based methods were used to assess the JAK2 mutation status of 806 patients with ET, including 776 from the MRC PT-1 trial and two other prospective studies. The combined cohort provides a unique resource for studying ET, particularly in view of its large size, centralized review of end-points and comprehensive follow-up. The involvement of a large number of secondary and tertiary centers, together with the inclusion of patients in all risk categories, suggest the results are of general relevance.

JAK2 mutation status divided the cohort into two distinct subgroups. Mutation-positive patients (53.4%) displayed multiple features resembling PV, with significantly increased hemoglobin levels (p<0.0001), neutrophil counts (p<0.0001), bone marrow erythropoiesis (p=0.001) and granulopoiesis (p=0.005). They suffered more venous thromboses in the year before diagnosis (p=0.04) and during follow-up (p=0.06), together with a higher incidence of polycythemic transformation (p=0.01). To explore the resemblance between V617F-positive ET and PV further, we analysed factors that might constrain erythropoiesis. Compared to mutation-negative patients with ET, mutation-positive patients had lower serum epo (p<0.0001), lower ferritin (p=0.01), and were more likely to be microcytic (p<0.0001). V617F-positive patients were more sensitive to hydroxyurea, requiring lower doses to control platelet count than V617F-negative patients (p<0.0001), a pattern not seen with anagrelide. Despite lower doses of hydroxyurea, V617F-positive patients had greater reductions in hemoglobin, platelet and white cell counts than V617F-negative patients, with no analogous differences noted between V617F-positive and negative patients randomized to anagrelide (p=0.004, p=0.04, p=0.0003 for platelet count, Hb and WCC). Mutation-negative patients did exhibit many clinical and laboratory features characteristic of a myeloproliferative disorder, including cytogenetic abnormalities, hypercellular bone marrow, abnormal megakaryocyte morphology, PRV1 over-expression, growth of epo-independent erythroid colonies, and a risk of myelofibrotic or leukemic transformation.

JAK2 mutation status defines two biologically distinct subgroups of ET with differences in presentation, outcome and response to therapy. Our results suggest a model in which V617F-positive ET and PV form a continuum, with the degree of erythrocytosis determined by physiological or genetic modifiers, including iron stores, epo homeostasis, gender and V617F-homozygosity. This concept has major implications for the classification, diagnosis and management of MPDs.

Author notes

Corresponding author