Myeloid Blasts in Transformed JAK2-V617F Positive Myeloproliferative Disorders Are Frequently Negative for the JAK2-V617F Mutation.

Acute myeloid leukemia (AML) is a common complication of myeloproliferative disorders (MPDs). The role of the JAK2-V617F mutation in this process is unknown. We performed a retrospective analysis of DNA samples from MPD patients with secondary AML. We analysed DNA samples taken at the time of transformation to AML from 54 MPD patients (24 PV, 21 ET, 9 IMF). In addition, DNA samples taken at diagnosis of MPD were obtained in 21 of these patients. DNA was extracted from bone marrow or peripheral blood films, purified granulocytes or frozen cells. FACS sorting of blast cells, T cells and neutrophils was performed in some of the samples. The allelic ratio of JAK2-V617F was determined by allele-specific quantitative PCR (AS-PCR). We obtained AS-PCR data on 52/54 samples taken at the time of transformation (96%), whereas 2 samples did not yield PCR products: 24/52 samples were negative for JAK2-V617F (46%) and 28/52 were positive (54%). For 14/24 negative patients (58%) we had additional DNA samples taken at the time of MPD diagnosis and interestingly, 5 of these 14 patients (36%) were positive for JAK2-V617F at this earlier time point before AML transformation. This suggests that in these patients the JAK2-V617F positive clone was lost during the evolution to AML. Furthermore, comparison of the JAK2-V617F allelic ratios with the percentage of blast cells in patient samples positive at transformation revealed 8/28 cases where the JAK2-V617F allelic ratio was markedly lower than the percentage of blasts, e.g. 8%T-allele and 52% myeloid blast cells. In these patients a JAK2-V617F negative AML clone most likely co-exists with a JAK2-V617F positive MPD clone. To address the question whether the AML clone arose independently from the JAK2-V617F clone, we analyzed loss of heterozygosity on chromosome 9p (9pLOH) in one informative patient who displayed a high allelic ratio of mutant JAK2 at diagnosis (94%T). The CD15+ cells from this patient showed 9pLOH at diagnosis, as demonstrated with two independent microsatellite markers. In contrast, the FACS sorted blast cells at the time of transformation contained both parental alleles in the 9p region and were JAK2-V617F negative by AS-PCR. This excludes the possibility that the AML clone lost the JAK2V617F in the process of undergoing mitotic recombination at a stage heterozygous for JAK2-V617F. Analysis of additional patients is under way. In summary, we found in a cohort of 54 MPD patients, 13 patients initially positive for JAK2-V617F that transformed into JAK2-V617F negative AML. Although not confirmed in the one patient analyzed, we cannot exclude that other patients the JAK2-V617F positive MPD clone lost the JAK2 mutation during the process of transformation. Alternatively, the AML clone could have developed de novo from a JAK2-V617F negative progenitor or stem cell. The latter model has difficulties explaining the high incidence of de novo AML (8/54 patients), unless the JAK2-V617F negative progenitor already carried an as yet unknown mutation and was part of the MPD clone.