Chronic myelomonocytic leukemia (CMML) represents a diagnostic and therapeutic challenge characterized by highly heterogeneous clinical and laboratory aspects, contrasting from mainly dysplastic (MD) to predominantly proliferative (MP) in different patients. Although no specific cytogenetic or molecular aberration has been associated to CMML, next generation sequencing (NGS) has recently led to the discovery of at least one lesion in up to 90% of patients. Nonetheless, the role of the identified genetic aberrations in CMML onset and progression remains to be clarified.

In a series of 40 consecutive patients we previously reported a higher frequency of RAS and JAK2 mutations and a shorter survival in those with MP- than in those with MD-disease. Furthermore, paired samples analysis showed RAS mutations acquisition in concomitance with progression from MD- to MP-CMML, suggesting these lesions as second hits that confer a proliferative advantage to the malignant clone, leading to poor outcome. In addition to these findings, a highly significant shorter life expectation in the MP-variant of CMML was more recently confirmed in an extended population of 74 patients (p=0.0005), further supporting the association of molecular acquisition of gene aberrations with disease progression.

By comprehensive next generation sequencing (NGS) of selected genes, here we aimed to further investigate the spectrum of aberrations contributing to CMML development and progression and to examine whether MD- and MP-CMML may be also discriminated at the molecular level.

We designed a NGS study (Oxford Gene Technology, Oxford UK) of 44 genes in DNA prepared from MNCs from 12 CMML patients after obtaining informed consent. Of the 21 samples analyzed, 17 were consecutively collected from 9 patients at the time of MD-CMML and later on during the disease course, showing either long lasting stable MD-CMML disease (median follow-up of 102 month), or progression to MP-CMML or AML, and 4 more were obtained from patients with MP-CMML (2 with previous MD-phase). In some patients, DNA prepared from purified CD3+ cells selected by FACS cell sorting was also analyzed.

Candidate mutations were validated by Sanger sequencing. Deep sequencing analysis confirmed TET2 mutations as the most frequent (10/12 patients, 83%) and, the earliest known event in CMML, being present since time of referral in 100% of our cases with sequential samples, supporting their possible role of initiating lesions in CMML. Overall, 9 patients harbored frameshift/nonsense mutations and 1 had an essential splice site substitution. Non-synonymous variations of yet unknown origin were detected in 3 cases while in 1 case the substitution found in MNCs DNA was identified by direct sequencing also in DNA from buccal swab and thus annotated as a SNP. Other documented mutations in variable proportions involved ASXL1, SRSF2, SF3B1, EZH2, CBL, DNMT3A, MPL, NOTCH1, NOTCH2, N- and K-RAS.

Among patients who were investigated with sequential samples collected at different time points and/or different disease phases, TET2, SRSF2 and ASXL1 mutations were documented from the first presentation in all cases, suggesting their acquisition as early events possibly driving molecular mechanisms of disease onset. In contrast, besides RAS mutations, which were detected at the time of disease progression from the MD- to the MP-variant in 2 patients, other aberrations possibly associated with disease evolution included EZH2 and CBL mutations, both detected in a small fraction of cells at diagnosis but significantly expanding after progression to MP-CMML. Of note, in one case harboring TET2, ASXL1, EZH2 and CBL concomitant mutations the sequencing of DNA from purified CD3+ cells unveiled the presence of TET2, ASLX1 and CBL mutations also in a significant fraction of T-lymphocytes, suggesting the aberration to possibly arise in a multipotent progenitor, whereas the EZH2 mutation appeared restricted to the myeloid lineage.

A combined analysis of sequential samples and single-cell-derived colonies is currently ongoing to better elucidate clonal evolution in CMML, which in turn could help the improvement of disease classification as well as the early identification of patients at risk of disease evolution.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution