Greek Registry Of Essential Thrombocythemia: Baseline Characteristics and Therapeutic Strategy

The Greek Registry of Essential Thrombocythemia (ET) is implemented under the auspices of the Acute Leukemias and Myeloproliferative Neoplasms Study Group of the Hellenic Society of Haematology. Hereby, we present results after four years of retrospective data collection. The total number of patients included is 1078, from 14 Greek sites; ET was diagnosed between 1982 and 2012.

The male to female ratio is 1:1.19. Median age at diagnosis is 63 years, median platelet counts (PLT) 826x10⁹/L, hemoglobin (Hb) 13.6 g/dL, white blood cell counts (WBC) 9.4x10⁹/L. The presenting symptoms were a thrombotic event in 6.8%, a hemorrhagic event in 1.5% of patients. In 79.8% of the patients the diagnosis was made after incidental finding of elevated platelet counts on routine laboratory investigation.

Molecular studies were performed in 677 patients and 42.8% of them were positive for the JAK2-V617F mutation. The presence of JAK2-V617F mutation (mutant vs wild type allele) was associated with baseline platelet counts (757.5 vs 882 x10⁹/L) and hemoglobin levels (14.4 vs 13.4 g/dL), p<0.001 (Mann-Whitney U-test). A history of thrombosis or hemorrhage was present in 18.6% and 6.6% of patients respectively. Chi-square test was performed to assess whether platelet counts at diagnosis (<600, 600-800, >800 x10⁹/L), Hb<13.8g/dL, WBC>9.5x10⁹/L, or splenomegaly are associated with thrombotic or hemorrhagic events in the past medical history or during the follow-up of ET patients. The only statistically significant difference was observed in the occurrence of thrombosis during the follow up: 10.1% of those with PLT between 600-800 x10⁹/L, 4.5% in PLT<600 and 5.6% in PLT>800 x10⁹/L. To assess for possible confounders the multivariable logistic regression model was used, with independent variables the PLT at diagnosis, age >60 years, history of thrombosis and first line therapy. The history of thrombosis was the only statistically significant risk factor with odds ratio (OR) 3.9 (p=0.0005), while PLT was not a statistically significant risk factor (OR=2.5, p=0.074).

Antiplatelet therapy was offered in 80% of patients (aspirin in 59.1%, clopidogrel in 4.7%, and combination therapy in 6.5%); anticoagulants (low molecular weight heparin or warfarin) were given in 2.3%, while the remaining 17.8% of patients did not receive any antithrombotic therapy. During the first six months post diagnosis, 31.6% of patients did not need any cytoreductive therapy. The rest 68.4% of the patients received first line therapy (hydroxyurea 80.6%, anagrelide 11.4% and interferon 5.4%). The response rates were 89.9%, 82.1% and 85.7%, respectively. Second-line therapy was received by 25.8% of the patients (hydroxyurea 23%, anagrelide 44.6%, interferon 9.5%), while the off-label combination of hydroxyurea and anagrelide was administered to 21.2% of the patients. Of the 852 patients treated with hydroxyurea as first line therapy, 12.1% switched to anagrelide and 1.2% to interferon. Of those initially treated with anagrelide, 27.6% switched to hydroxyurea and 8.2% to interferon. During the follow up phase, secondary solid tumor occurred in 4% and hematological malignancy in 2.7% of the patients.

The aim of the registry and the subsequent data analysis is to convey the practice of managing the disease. Moreover, useful conclusions can be reached regarding to the patients’ responsiveness to therapy and the minimization of thrombotic and hemorrhagic adverse events.