4. Myelodysplastic Syndrome
Myelodysplastic syndrome (MDS) is a group of clonal hematopoietic neoplasm characterized by ineffective hematopoiesis, cytopenia, morphologic dysplasia, and potential progression to acute myeloid leukemia [50]. Most patients with MDS die of bone marrow failure rather than transformation to acute myeloid leukemia. MDS is diagnosed based on the World Health Organization (WHO) classification, but the prognosis for survival is stratified based on the international prognostic scoring system (IPSS) and revised IPSS (IPSS-R) that incorporate cytogenetic abnormalities, percentage of bone marrow myeloblasts, and number of cytopenias [51,52]. As part of the IPSS-R system, cytogenetic abnormalities are stratified from very good with single del(11q) and −Y to very poor with complex karyotype (>3 abnormalities). Approximately half of patients with MDS have normal karyotype that is associated with good prognosis in the IPSS-R system, but MDS patients with normal karyotype are still heterogeneous genetically. SNP arrays were therefore performed on MDS to identify additional occult genetic abnormalities, especially in patients with normal karyotype.
Gondek et al. [53,54] was the first group to study MDS with SNP array. They applied an Affymetrix 50K SNP Assay to 66 and 72 patients with MDS, and found chromosomal defects in 82% of MDS patients, including 68% patients with normal karyotype, 81% patients with abnormal karyotypes, with chromosomes 8, 7, 5, and 11 most frequently involved. Segmental uniparental disomy (sUPD) was found in 33% of patients with MDS, and usually in regions frequently affected by deletions detected by metaphase cytogenetic analysis, including 7q and 11q. In a similar study, Mohamedali et al. [55] studied 119 patients with low-risk MDS with 50 K, 250 K and 500 K SNP arrays and identified deletions in 10%, amplifications in 8%, and UPD in 46% of cases. Nowak et al. [56] showed that CNAs and LOH can be identified in the CD34 positive blasts. These early studies suffered from absence of paired normal tissue for each case, which made distinction of inherited CNAs and LOH from somatic acquired ones difficult. Heinrichs et al. [57] performed a prospective study of matched pairs of bone marrow and buccal cell (normal) DNA from 51 patients with MDS by 250K SNP array, and identified somatically acquired genomic abnormalities in 41% patients, including 15% in MDS with normal karyotypes. UPDs affecting chromosome 7q was associated with rapidly progressive clinical course despite a low-risk IPSS score [57]. Similarly, Tiu et al. [58] analyzed 250 cases of MDS by 250K and Affymetrix SNP Array 6.0 with paired bone marrow and CD3+ lymphocytes to distinguish germline lesions. In this study, they showed that combined metaphase cytogenetics and SNP array had a higher diagnostic yield of chromosomal defects (74% vs. 44%), compared with conventional karyotyping. The genetic abnormalities detected by SNP arrays were deletions and aUPD involving chromosomes 1, 5, 7, 11, 17, and 21. While Mohamedali’s study failed to show independent prognostic significance of the genetic lesions identified by SNP array on multivariate analysis, Tiu showed that the presence of new genetic lesions detected by SNP array was predictive of poor prognosis in MDS by univariate and multivariate analyses [58]. These studies proved the utility of SNP array in detecting submicroscopic genetic lesions in MDS as a complement to metaphase cytogenetics, and the lesions identified by SNP arrays can further help prognostic stratification of MDS patients.
MDS can be difficult to diagnose clinically due to many mimickers of the disease and the lack of significant morphologic dysplasia in a small subset of cases. For example, severe aplastic anemia (AA) may be difficult to distinguish from hypoplastic MDS morphologically and cytogenetically. Afable et al. [59] demonstrated the utility of SNP analysis in AA to complement metaphase cytogenetics for the detection of clonal chromosomal lesions. Combined metaphase cytogenetics and SNP array identified chromosomal lesions in 19% of AA and 54% of hypoplastic MDS. Persistent detection of chromosomal lesions by SNP array would be highly suspicious for hypoplastic MDS and less response to immunotherapy (ATG/cyclosporine) for AA. Therefore, in diagnostically-challenging and equivocal cases of MDS, SNP array can be used to establish the presence of clonal hematopoiesis in patients with normal karyotype and allow appropriate management of the patients.
With the advent of next generation sequencing, SNP arrays are not likely to be used for the identification of genes involved in disease. In the last decade, SNP arrays have played an important role in the identification of individual genes important for the pathogenesis of MDS. TET2 gene was identified by SNP array genomic profiling and genomic sequencing in 102 patients with MDS, and acquired deletions, missense and nonsense mutations in the TET2 gene were found in 26% cases of MDS [60]. Recurrent aUPD and microdeletion of chromosome 7q led to identification of EZH2 gene and mutations in MDS [61,62]. The TET2 gene mutations have been found to be associated with better response to hypomethylating agents [63], while EZH2 mutations are poor prognostic marker for MDS. By combining SNP-array and gene expression profiling, Merkerova et al. [64] identified BMP2 and TRIB3 genes located in 20p UPD as potential candidate genes for the pathogenesis of MDS.