Dissecting the Hierarchical Level of Hematopoietic Progenitors' Involvement in AML with NPM1 Gene Mutation and Their Engraftment Potential in Immunocompromised Mice.

Abstract 480

Acute myeloid leukemia with mutated NPM1 gene and cytoplasmic nucleophosmin (NPMc+ AML) [Falini B et al, NEJM 2005;352:254-266] is a new entity of WHO classification that shows distinctive biological and clinical features [Falini B et al, Blood 2007;109:874-885] which include negativity for CD34 antigen expression at both immunohistochemistry and gene expression profiling. Flow cytometric analysis shows that, in most NPM1-mutated AML, percentages of CD34+ cells are in the low range (< 5-10%). Detection of NPM1 mutations by molecular techniques and/or immunohistochemistry and Western Blot analysis with specific antibodies provides an important tool for tracking the genetic lesion in leukemic cells at different hierarchical stage. We previously reported involvement by NPM1 gene mutation of the CD34+ cell fraction isolated from patients with NPM1-mutated AML, and, in one case, the involvement, in particular, of the early progenitor CD34+/CD38- [Martelli MP et al, Blood (ASH Annual Meeting Abstracts) 2008;112:307]. Here we expand and confirm our previous observation in 5 cases of CD34-negative NPM1-mutated AML. CD34+/CD38- cells were isolated by either FACS (3 cases, purity >98%) or MACS-sorting (2 cases, purity >92%) and analyzed by molecular analysis and Western Blot with a specific anti-NPM1 mutant antibody, respectively. The presence of either NPM1 gene mutation or mutant protein was demonstrated in all samples analyzed proving the CD34+/CD38- cells belong to the leukemic clone. This cell subpopulation displayed also immunophenotypic features classically associated to leukemic stem cells (LSCs) (CD123+/CD33+/CD90-) in all (16/16) samples analyzed, suggesting they might actually represent the LSCs in NPM1-mutated AML. Indeed, CD34+ cell fraction isolated from NPM1-mutated AML was able to generate leukemia in immunocompromised mice resembling the original patient's disease. However, there is experimental evidence that, at least in some CD34-negative AML, also the CD34- population may contain LSCs. Whether the CD34- cell compartment in NPM1-mutated AML is also able to engraft and outgrow into leukemia in mice remains to be clarified. For this purpose, we assessed the engraftment ability of CD34- cells from 5 NPM1-mutated AML patients. No engraftment was observed in one case. Interestingly, in three patients with myelomonocytic (M4, 2 cases) and myelocytic (M2, 1 case) AML, the CD34- fraction resulted into marrow engraftment by human CD45+/CD33+ myeloid cells that, at morphological and immunohistological grounds, consisted of a mixed population of macrophage cells expressing the CD68 (PG-M1) antigen and mature looking myeloperoxidase (MPO)-positive cells. This pattern possibly reflects short-term engraftment by leukemic cells devoid of self-renewal potential that differentiated into mature elements. However, the neoplastic nature of engrafted cells could be established with certainty only in one case by western blotting detection of NPM1 mutant protein. Immunohistochemistry could not help in these cases to establish the leukemic nature of human cells since terminally differentiated leukemic cells in NPM1-mutated AML show nucleus-restricted NPM1 positivity. In contrast, the pure CD34+ fraction (availabel for comparison in one of these three cases) engrafted as AML with clear blastic morphology and cytoplasmic dislocation of nucleophosmin. In a fourth patient, the highly purified CD34- fraction from relapsed NPM1-mutated AML engrafted in mice with a typical AML picture. These preliminary findings suggest that in general the CD34- fraction from NPM1-mutated AML may have more limited engraftment potential than the CD34+ fraction. Further studies are ongoing to address this issue.