The chronic myeloproliferative disorders (MPD), polycythemia vera (PV), essential thrombocytosis (ET) and idiopathic myelofibrosis (IMF) differ phenotypically but share the same mutation, JAK2V617F. To determine how the same mutation yields three different phenotypes, we examined the relationship between disease phenotype and the quantitative JAK2V617F allele burden in both CD34+ cells and neutrophils. The study population consisted of 126 PV, 41 ET and 18 IMF JAK2V617F-positive patients with a median disease duration of 8 years. The JAK2V617F allele burden was determined by allele-specific quantitative PCR using genomic DNA from neutrophils and circulating purified CD34+ cells; in 6 ET, 6 IMF and 13 PV patients, erythroid colonies cloned from CD34+ cells were genotyped for JAK2V617F. As we previously reported, the mean neutrophil JAK2V617F allele burdens were lower in ET compared to PV and IMF (38% vs 66%; p<.001). In PV but not in ET, the neutrophil JAK2V617F allele burden correlated directly with the reticulocyte count (R=.452, p<.001), the white cell count (R=.519; p<.001) and splenomegaly (R=0.378; p<.001) and inversely with the platelet count (R=0.323; p=0.004). In both ET and PV, the median CD34+ cell JAK2V617F allele burdens were lower than the corresponding neutrophil allele burdens (25% vs 39% in ET, p=.04, and 56% vs 64% in PV, p=.01), but these were equivalent in IMF (77% and 77%). The median CD34+ cell allele burden was lower in ET patients compared to PV or IMF patients (p<.001). In contrast to neutrophils in which there was no correlation between allele burden and disease duration (R = 0.067; p = ns), in PV the CD34+ cell allele burden correlated with disease duration (R=0.358; p = 0.035), and more strongly correlated with the white cell count (R=.725; p<0.001) and spleen size (R=0.635; p<0.001) than the neutrophil allele burden. BFU-e clonogenic assays identified JAK2 wild-type progenitors in ET and PV patients in whom the CD34+ JAK2V617F allele burden was lower than the neutrophil allele burden, and did not identify wild-type progenitors in PV and IMF subjects in whom the CD34+ cell and neutrophil allele burdens were similar. Based on this observation, we defined clonal dominance at the progenitor level as a difference of 10 or less between the neutrophil and CD34+ allele burdens. Clonal dominance was present in 22% of ET, 53% of PV and 90% of IMF patients, and was independent of the clonal genotype as clonal dominance was present with heterozgous only (ET, PV and IMF), homozygous only (PV or IMF) or mixtures of heterozygous and homozygous clones (PV, IMF). Within PV, clonally dominant subjects had significantly longer disease durations (11 vs 5.5 years, p=.01), higher white cell counts (18.2 vs 11.1, p=.02), larger spleens (4 vs 0 cm below the costal margin, p=.003), and were more commonly receiving chemotherapy than the nonclonally dominant PV patients (46% vs 22%, p=.02). Importantly, the neutrophil allele burden did not differ between clonally dominant and nondominant PV patients (mean 72.8 vs 68.7%, p=0.506). We conclude that the extent of JAK2V617F CD34+ cell clonal dominance is a major determinant of disease class within the MPD, and within PV, is a major determinant of extramedullary disease and leukocytosis. The extent of clonal dominance at the progenitor level cannot be determined from the neutrophil allele burden alone, and thus limits the utility of the neutrophil allele burden as a disease marker.

Author notes

Disclosure: No relevant conflicts of interest to declare.

Sign in via your Institution