Page 144 - Haematologica March 2020
P. 144

J. Woo et al.
mutations and disease phenotype which differed for HCT outcomes. In particular, the total number of mutations and more mutations in epigenetic regulators were significantly associated with relapse and with inferior survival. Reminiscent of observations in patients with MPN or MDS,15,38 patients with more mutations experienced a higher incidence of relapse, especially when the mutations occurred in genes involved in epigenetic regulation. Those
mutations occurred predominantly in patients with dys- plastic CMML, as also observed by others.15 Of note, how- ever, our unsupervised clustering analysis uncovered a previously unrecognized group of high-risk patients with a higher mutation burden involving epigenetic regulators. This group of patients was not closely associated with high-risk disease as determined by cytogenetics, suggest- ing independent mechanisms and heterogeneity in the dis-
AB
D
Figure 2. Molecular profiling and risk factors associated with hematopoietic cell transplantation outcomes in patients with chronic myelomonocytic leukemia. (A) Unsupervised hierarchical clustering of a correlation matrix between mutations and clinical parameters. Group 1 consists of mutations in mitotic signaling pathways, myeloproliferative type of chronic myelomonocytic leukemia (CMML) (Prolifer), mutations in TP53, and high-risk disease by cytogenetics and prognostic scoring sys- tems. Group 2 consists of mutations in epigenetic pathways, methylation, and splicing regulation. (B) Increased relapse following hematopoietic cell transplantation in patients with high mutation burden, stratified by total number of mutations (≥10 vs. <10), and (C) stratified by number of mutations in epigenetic regulators (≥4 vs. <4; classified in Online Supplementary Table S1). (D) Unsupervised clustering of association (odds ratio, corresponding to the scaling bar) between groups of mutations (signaling pathways, epigenetic regulators, and tumor suppressors), high-risk cytogenetics (cyto), risk stratification systems (CPSS, MDAPS) and relapse. CMML with a high mutation burden in epigenetic regulators (epigene) was distinct from previously recognized high-risk diseases by cytogenetic abnormalities, blast count, and mutations in tumor suppressors (red boxes). (E) Time from diagnosis to transplant. High mutation burden (≥10), in particular, in epigenetic processes (≥4), was associated with a longer interval from diagnosis to transplant, consistent with the concept of disease evolution and acquisition of more mutations (in epigenetic but less so in signaling pathways) with disease duration *P<0.05 (individual values in Online Supplementary Table S3). **P=0.01, ***P=0.03.
C
E
658
haematologica | 2020; 105(3)


































































































   142   143   144   145   146