Investigating Clonal Dynamics of Different Mutations in Myelodysplastic Syndromes

In a whole-exome sequencing (WES) and targeted deep sequencing study published in Nature Genetics, researchers identified several biomarkers that could improve diagnosis and therapy selection for patients with myelodysplastic syndromes (MDS). The study also highlighted the complex clonal dynamics during MDS progression.

“Our molecular analysis parallels the risk classification of MDS, showing that progression steps previously defined by pathologic criteria are accompanied or mediated by distinct molecular changes,” wrote Hideki Makishima, MD, PhD, from the Department of Translational Hematology and Oncology Research at the Taussig Cancer Institute at the Cleveland Clinic in Ohio, and co-authors.

The authors used previously published genotyping data and a new cohort of patients to assess mutational enrichment patterns for 2,250 patients (lower-risk MDS = 1,207; higher-risk MDS = 683; secondary acute myeloid leukemia [sAML] = 360). They identified 2,322 mutations in 1,383 genes, of which 49 genes were significantly mutated when compared with the background mutation rate (p<0.01). These driver gene mutations (those thought to be directly responsible for causing the disease) were present in 77 percent of patients analyzed by WES.

Univariate comparison between lower- and higher-risk MDS revealed that mutated genes occurred more commonly in higher-risk MDS, with the exception of SF3B1, which was more frequently mutated in low-risk MDS.

The mean number of mutations was 11.4 per patient; patients with sAML had a significantly higher mutation rate compared with patients with lower-risk MDS (p=0.002), and the number of mutations increased significantly over time (p=0.032). Although most of these driver mutations had been reported previously, the authors identified several possible new drivers, including C1QTNF3, IRF2, NEURL1, GNL2, PCDHA1, and PDGFRB.

“The driver genes can be classified into molecular subtypes differentially associated with lower-risk MDS, higher-risk MDS, or sAML,” the authors wrote. Multivariate analyses identified seven genes (FLT3, PTPN11, WT1, IDH1, NPM1, IDH2, and NRAS) that could be designated as “type-1” mutations, which were significantly enriched in sAML compared with higher-risk MDS. They also identified 10 “type-2” mutations, which were significantly enriched in higher-risk MDS: GATA2, NRAS, KRAS, IDH2, TP53, RUNX1, STAG2, ASXL1, ZRSR2, and TET2. “This new categorization provides insights into clonal dynamics and allows for the use of sub-clonal events as MDS progression biomarkers,” they explained.

The researchers also conducted a separate analysis to characterize the chronologic behavior of type-1 and type-2 mutations in the 122 patients who were followed longitudinally – 90 of whom progressed to sAML. Most of those patients (n=109; 82%) had one of the 401 mutations in known driver genes detected at one or more timepoints. In most cases (n=70), the number of mutations was higher at the second timepoint (average = 2.7 mutations) than at diagnosis (average = 1.9 mutations).

“Overall, driver mutations tended to increase their clone sizes and were more likely to be newly acquired than lost between two timepoints,” Dr. Makishima and co-authors reported.

Type-1 mutations in sAML were more likely to be newly acquired than present before progression, compared with type-2 and other mutations (p=0.0001), whereas type-2 mutations were more frequently persistent than newly acquired (p=0.002).

The time from the acquisition of type-1 mutations to sAML progression was, on average, shorter than the time from the acquisition of other mutations to sAML (hazard ratio [HR] = 1.82; 95% CI 1.08-3.05; p=0.025). Overall survival (OS) also was significantly shorter for patients with type-1 mutations (HR=1.50; 95% CI 1.20-1.86; p=0.001).

“When present in MDS, [type-1] mutations were associated with a higher risk of progression to sAML and shorter OS, compared with other mutations,” the authors concluded. “Close monitoring of the emergence of type-1 mutations might allow for early diagnosis of progression to sAML.”

The study’s findings are limited by the potential shortcomings of the previously published data that were included in the current trial. The role of SF3B1 mutations (which was strongly enriched in lower-risk MDS but not higher-risk MDS) also confounded the results. “The two novel sets of gene mutations might be simply associated with sAML evolution from MDS, because the examined cohort included only MDS cases, even though sAML is sometimes derived from other types of myeloid neoplasms, including MDS/ myeloproliferative neoplasms,” the authors noted.


Makishima H, Yoshizato T, Yoshida K, et al. Dynamics of clonal evolution in myelodysplastic syndromes. Nat Genet. 2016 December 19. [Epub ahead of print]