Abstract
Myelodysplastic syndromes (MDS) and chronic myelomonocytic leukemia (CMML) are hematological disorders at high risk of progression to acute myeloid leukemia (sAML). Previous high-throughput sequencing studies have provided insight into the mutational dynamics and clonal evolution underlying disease progression. However, large serial sequencing studies are still required to define which type of mutations alone or in combination contribute to leukemic transformation.
To assess the mutational profiles and mutational dynamics underlying progression from MDS to sAML, a targeted-deep sequencing (TDS) of 117 MDS/AML related-genes was performed in 110 bone marrow serial samples from 50 MDS/CMML patients who evolved to sAML and 5 patients who did not evolved (controls), at two different time-points: at the time of diagnosis and at sAML progression or after a median of 3 year follow-up, respectively.
A total of 269 mutations in 57 different genes were identified at second sampling. At diagnosis, all patients, progressing and not progressing (controls), presented similar number of mutations (p=0.15). Moreover, patients evolving to sAML were then divided by FAB/WHO subtypes at diagnosis (CMML, low-risk and high-risk MDS subgroups) and no differences were observed in the number of mutations (p=0.71) and variant allele frequency (VAF) between each group (p=0.63). It should be noted that mutations in the splicing pathway were significantly more frequent in low-risk MDS patients (89% low-risk MDS vs. 56% high risk MDS, p=0.038). However, after progression, those patients who evolved to sAML displayed a statistically significant increase of mutations (p=0.001) at the leukemic phase, while controls did not at the follow-up sample (p=0.88). This higher number of mutations at second sampling in patients who evolved to sAML, independently of their diagnostic subtype, may be indicative of a higher genomic instability during disease evolution.
To study the mutational dynamics and what mutations could be important during disease evolution, the VAFs of mutations detected at both time-points in each patient of transformation cohort were compared. We observed that some mutations identified at the sAML stage (163 mutations) were already present at the MDS stage, at clonal or subclonal levels, and were retained during evolution, for example in genes such as SRSF2 and DNMT3A. However, 106 mutations increased in clonal size or were newly acquired. Interestingly, most of mutations in Ras signaling pathway showed a same pattern: they were not present at time of diagnosis and appeared at sAML. In fact, mutations in this pathway were detected in 25 of 50 patients (50%) included in this cohort and in 22 of them (88%) mutations displayed this dynamic. Therefore, in this study, Ras signaling was the most common pathway involved in the progression from MDS to sAML. Of note, 9 of these patients (18% of the whole cohort) presented, independently of diagnosis, a co-occurring cohesin mutation, that was already present at diagnosis and, in most cases, markedly increased in clonal size at sAML. Thus, the combination of mutations in these two pathways could play an important role during disease evolution.
In addition, 22 of 50 patients were treated with a disease-modifying agent (18 azacytidine and 4 lenalidomide) before they progressed to sAML, while the remaining 28 patients received no treatment or supportive care and were considered as non-treated. Thus, we studied the effect of disease-modifying therapy on mutational dynamics in this cohort of patients progressing to sAML. In the treated patients, a higher proportion of newly acquired or increasing mutations at sAML in chromatin modifiers was observed, while in non-treated patients most mutations remained stable (61% vs. 28.6%, p=0.013). By contrast, regarding treatment, no differences were detected in the mutational dynamics of cohesin (p=0.56) or Ras pathway (p=1.00).
MDS progression to sAML was characterized by a higher genomic instability, independently of MDS subtypes of patients at diagnosis. Ras signaling was the most frequent affected pathway during disease evolution in this cohort and, interestingly, the co-occurrence of Ras signaling and cohesin mutations could play an important role in the progression. Moreover, mutations in chromatin modifiers genes could be related to the evolution of patients who received disease-modifying treatment before progression to sAML.
Olivier:Celgene: Honoraria; Jassen: Honoraria. Díez-Campelo:Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau.
Author notes
Asterisk with author names denotes non-ASH members.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal