Monoclonal myelopoiesis and subnormal erythropoietin concentration are independent risk factors for thromboembolic complications in essential thrombocythemia

Essential thrombocythemia (ET), one of the chronic myeloproliferative disorders, was recently shown to consist of a heterogeneous cohort of patients. Indeed, when clonality analyses were performed using X-chromosome inactivation patterns (XCIP) in samples of ET patients, about 50% showed constitutive skewing, about 25% had monoclonal myelopoiesis, and 25% had polyclonal myelopoiesis.1,2 Further, about 50% of newly diagnosed ET patients presented with subnormal plasma erythropoietin (EPO) concentrations.3,4 ET patients with monoclonal myelopoiesis were, compared with those with polyclonal disease, shown to have a significantly higher risk for developing thrombotic complications by Harrison et al2; this finding was confirmed by Chiusolo et al.5 Also, ET patients with subnormal EPO concentration at the time of diagnosis were shown to have a significantly higher requirement for myelosuppressive treatment.4 Further, in the group of ET patients with subnormal EPO, 10 out of 13 developed cerebrovascular or other macrovascular events compared with 7 of 18 patients with normal plasma EPO concentration at diagnosis. The difference between the groups was statistically significant (P < .02).4 The aim of the present study was to investigate whether the group of ET patients with monoclonal myelopoiesis was identical to the group of ET patients with subnormal EPO concentration at diagnosis, or whether these 2 variables are independent risk factors for the development of thromboembolic complications in ET.

Female patients with ET, all fulfilling the Polycythemia Vera Study Group (PVSG) criteria,6 were recruited from the Hematology Outpatients Clinic at Sahlgrenska University Hospital, Göteborg, Sweden, and from the University College London Medical School, United Kingdom. XCIP analyses on neutrophils and T cells were performed using the human androgen receptor (HUMARA) assay as described by Gale et al.7 At the time of sampling, some patients had a platelet count of less than 600 × 10⁹/L, but all patients had platelet counts greater than this limit at the time of diagnosis. None of the patients had been treated with myelosuppressive agents. All patients were on low-dose aspirin prophylaxis.

Eight English patients were selected among earlier XCIP-tested patients who had frozen serum stored and venous blood counts taken. Twenty-two Swedish patients were investigated, using the HUMARA assay, at the Department of Clinical Genetics at Lund University Hospital Lund, Sweden; of these, 9 were not interpretable due to skewing, 2 were excluded due to anemia (ie, hemoglobin concentration < 120 g/L, which possibly could affect the EPO concentration), and 11 patients were found to be evaluable. All serum or plasma EPO concentrations were analyzed at the Clinical Chemistry Laboratory at Sahlgrenska University Hospital, using an immunoenzymometric assay (Quantikine IVD Human Erythropoietin DEP 00; R&D Systems, Minneapolis, MN). The reference range for healthy individuals, given by the manufacturer, is 3.1-14.9 IU/L.

No significant differences between means (Student ttest) for the groups of monoclonal and polyclonal ET patients were found when hemoglobin concentration, EPO, concentration, or disease duration were compared. Two of the 7 ET patients with monoclonal myelopoiesis had a subnormal EPO, compared with 2 of the 12 ET patients with polyclonal myelopoiesis; this difference did not reach statistical significance (Fisher exact test). However, the mean platelet count was significantly higher in the monoclonal group of patients compared with the mean for the ET patients with polyclonal myelopoiesis (P = .02; Table 1).

The present results consequently do not support the hypothesis that the group of ET patients with monoclonal myelopoiesis is identical to the group of ET patients in whom the EPO concentrations were subnormal at diagnosis. Also, not all ET patients exhibiting monoclonal myelopoiesis present with subnormal EPO levels at diagnosis. The XCIP analyzes were done with DNA, which restricts the evaluable cells to myelopoiesis. The possibility that diverging results could have been obtained if an RNA-based technique had been employed by using platelets or erythrocytes cannot be excluded, however, but appears unlikely.2 The present results therefore indicate that subnormal EPO concentrations in ET patients at diagnosis and monoclonal myelopoiesis should be considered as independent risk factors for thromboembolic complications.

The authors wish to thank Bertil Johansson MD, PhD, Department of Clinical Genetics, Lund, for his valuable help with the HUMARA analyzes on the Swedish patients. The present study was supported by grants from FOU Västra Götaland, Stiftelsen Jubileumsklinikens Forskningsfond mot Cancer, and Stiftelsen Assar Gabrielssons Forskningsfond.