CNAs are the most abundant form of genetic alteration in cancer; however, due to the difficulties associated with their identification, only a limited number of CNAs can be used in a clinical setting. Targeted cancer-related gene panel NGS has become a routine clinical tool in recent times and can be used for the detection of CNAs, in addition to the detection of somatic mutations. One typical example is the Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) study, which enables precision oncology in patients with solid tumors [19]. Prospective clinical sequencing of 10,000 patients with advanced solid cancer using the MSK-IMPACT cancer panel revealed the presence of highly frequent genes with CNAs, such as PIK3CA, EGFR, and KIT, in addition to frequently mutated genes such as TP53 and KRAS [20]. It has also been reported that the concurrent detection of CNAs and mutations using targeted cancer-related gene panel is possible in myeloid malignancy [21]. The authors analyzed the mutations and CNAs in 270 patients with myeloid malignancy (MDS, MDS/MPN, MPN, and AML) using a targeted cancer-related gene panel and identified the presence of targeted CNAs in 68 patients [21]. However, the clinical and prognostic implications of CNAs from the target gene panel analysis in MDS have not been well-studied. In this study, our focus was to investigate whether the detection of CNAs defined from the targeted NGS data provided an additional prognostic value over somatic mutations, and also to explore whether a synergistic effect is observed on combining the somatic mutation and CNA data. Indeed, on taking into account both somatic mutations and CNAs, more genetic aberrations were found to be significantly associated with the prognosis of MDS, and the significance level became more reliable than it was on considering somatic mutations alone.
The identification of prognostic genetic factors and their integration into the clinical data is an essential process for the prediction of survival and disease progression of MDS. Karyotype has been the gold standard of MDS genetic data, and is a crucial component of the IPSS and the revised IPSS [22, 23]. Conventional karyotype analysis can detect chromosome abnormalities in only half of the de novo MDS patients. Thus, several studies have attempted to utilize molecular genetic data such as somatic mutations and CNAs to improve the precision of MDS prognostication and to aid the identification of therapeutic targets. In this study, we simultaneously detected somatic mutations and CNAs using the same targeted NGS data, which led to the identification of further genetic aberrations in 17 genes, compared to those detected upon using somatic mutations alone (Table III). Somatic mutations of two genes, TP53 and NRAS, had significant prognostic impacts on the survival of our MDS patients, as previously reported [1, 6, 7, 24, 25]. All five patients with TP53 CNAs also harboured TP53 mutations, and only two patients had NRAS CNAs without NRAS mutations. In the case of these two genes, the detection of CNAs did not prove to be of additional clinical significance. In comparison, genetic aberrations of four genes, including TET2, LAMB4, U2AF1, and CBL, showed a significant correlation with survival only after a combined analysis of both somatic mutations and CNAs, but not by the analysis of somatic mutations alone. We found eight TET2 CNAs (loss and gain in four each). The adverse prognostic impacts of copy number loss [11] or somatic mutations [24, 26] of TET2 have been previously reported, but none of reports have demonstrated the clinical significance of copy number gain of TET2. When we analysed the prognostic impacts of each combination of copy number loss or copy number gain with somatic mutations, each of these was associated with poor OS (P = 0.039 for loss and P = 0.037 for gain, data not shown). However, this kind of analysis does not seem to be appropriate given the small size. CBL mutations are known to contribute to myeloid leukemogenesis through a gain-of-function phenotype of CBL mutant [27]. The mutations recurrently occur in various myeloid neoplasms, including MDS and secondary AML [28]. In this study, copy number gain of CBL was found in four patients without CBL mutations, and the analysis of combined CNAs and somatic mutations of CBL showed a significant adverse impact of CBL aberrations on OS and AFS. U2AF1 mutations result in altered pre-mRNA splicing and may contribute to MDS pathogenesis [29]. Some studies have shown inferior survival in MDS patients with U2AF1 mutations [30, 31]. In this study, seven patients harboured U2AF1 CNAs (loss in 3 and gain in 4), and U2AF1 aberrations were associated with inferior OS. LAMB4 mutations are recurrently found in MDS patients, but little is known about their prognostic implications and contribution to MDS pathogenesis [2]. In this study, 17 patients harbored LAMB4 mutations, and detection of CNAs resulted in 9 more patients being identified with LAMB4 aberrations. The results suggest that concurrent detection of targeted CNAs and mutations is feasible [21] and may improve the prognostic implications of genetic aberrations in several MDS-related genes through detection of more aberrations.
There are several issues which should be considered while interpreting our results. Firstly, we suggest that the prognostic implications of CNAs identified in this study on survival should be confined to MDS patients receiving active treatment, as most of the patients in this study received chemotherapy and/or allogeneic hematopoietic stem cell transplantation. The timing of collection of bone marrow samples should also be taken into consideration, with some of them being acquired after hypomethylating treatment, though none after transplantation. This should be done in order to account for the clonal dynamics of genetic aberrations during the course of MDS [32–34]. Our targeted cancer-related gene panel was designed to target 28 genes based on PCR amplification, hence, technically it is possible to detect only the gene level CNAs. An NGS panel with more cancer-related genes and backbone targets may identify the large structural alterations and chromosomal rearrangement.