Biclonal expansion and heterogeneous lineage involvement in a case of chronic myeloproliferative disease with concurrent MPL^W515L/JAK2^V617F mutation

To the editor:

Chronic myeloproliferative diseases (CMPDs)/myeloproliferative neoplasms (MPNs) are caused by clonal mutations of tyrosine kinase or receptor genes such as BCR-ABL, JAK2^V617F, and MPL^W515L.^1,2  Recently, we and others showed that these mutations may be combined in hematopoietic cells of some MPN cases.^3,4  It is not yet clear whether, in cases with a double mutation, subclonal evolution has taken place or whether 2 unrelated clones proliferate independently. To address this issue, bone marrow and peripheral blood cells in a case of unclassifiable MPN (Figure S1, available on the Blood website; see the Supplemental Materials link at the top of the online article) with combined JAK2^V617F and MPL^W515L mutation were separated by laser microdissection as well as fluorescence-activated cell sorting, followed by quantitative analysis of mutated allele burden by pyrosequencing.⁵ 

Hematopoietic lineages demonstrated a heterogeneous involvement of MPL^W515L and JAK2^V617F (Table 1). Besides double-mutated cell types (CD34⁺, CD14⁺, CD14⁻/CD15⁺/CD16⁺), lineage-specific isolates with only one of the 2 mutations were observed affecting either JAK2^V617F (megakaryocytes, erythroid cells, and eosinophils; Figure S2) or MPL^W515L (T and B lymphocytes).

CD34⁺ precursor cells exhibited 94% MPL^W515L and 44% JAK2^V617F alleles. This is consistent with the existence of double-mutated cells. Even if trisomy 9 with 3 mutated JAK2^V617F alleles is taken into account, at least 20% of the cells should be affected by both mutations. Single-cell analysis showed MPL^wild-type/JAK2^wild-type in one granulocyte (data not shown). These findings indicate that, besides mutated stem cell clones, nonmutated clones were still proliferating, and that the wild-type alleles detectable in BM cells and CD34⁺ cells did not belong completely to hemizygously mutated cells. Heterozygosity for JAK2^V617F was found in megakaryocytes whereby the value below 50% might be due to polyploidy and cutting artefacts.⁵ 

Different patterns of lineage involvement may be explained by varying localization of the mutations on divergent levels of the hematopoietic maturation hierarchy. When colony formation by CD34⁺ precursor cells from MPL^W515K/L-mutated CMPD was compared with JAK2^V617F-mutated cases, it was recently found that MPL^W515K/L might affect more primitive cells than the JAK2^V617F mutation,⁶  which is in accordance with the results of mutation analysis in lymphocytes in this case which were positive only for MPL^W515L but not for JAK2^V617F. Heterogeneous lineage involvement in this case, amounting to various combinations of mutations of both MPL^W515L and JAK2^V617F with each other or wild-type, is consistent with the simultaneous proliferation of at least 2 unrelated stem cell clones with either mutation. The findings argue strongly against a concept of clonal evolution with emerging subclones which have subsequently acquired MPL^W515L and JAK2^V617F. Interestingly, the independently coexisting JAK2^V617F and MPL^W515L-mutated precursor cell clones in this case demonstrated divergent preferences to differentiate into the main hematopoietic lineages. Restriction of the JAK2^V617F mutation to 1 lineage or sparing of a single lineage has been reported previously.^7,8 

It has been hypothesized that different gene dosages⁹  or a varying combination of different molecular defects in one pathologic stem cell might be responsible for the phenotypic heterogeneity of CMPD.¹⁰  This study lends support to the concept that, at least in a subfraction of CMPD, heterogeneity might also be caused by parallel proliferation of unrelated abnormal hematopoietic stem cell clones.