Cytogenetic testing improves diagnosis in myeloid disorders; chromosomal (chr) aberrations have important clinical implications. SNP arrays (SNP-A) can be applied for karyotyping with a superb resolution of unbalanced defects and detection of uniparental disomy (UPD). We stipulated that SNP-A will enhance diagnostic value of metaphase cytogenetics (MC) and uncover new random/recurrent lesions. We applied 250K SNP-A to analysis of 76 controls and 318 patients, including 95 MDS, 64 AA, 20 PNH, 48 MDS/MPD, and AML both as primary (N=32) and secondary (N=59). Multiple samples were obtained in 13 patients. Minimal clonal size detectable by SNP-A was 25–50% by dilution studies. Repetitive testing resulted in congruent results; analysis of chr X in males showed >99% fidelity. To obtain reference, deletions and duplications seen in controls were analyzed. These abnormalities correspond to germ line encoded copy number variants (CNV). In patients such CNV were not deemed pathogenic. SNP-A confirmed 82% of unbalanced chr lesions detected by MC; discordant cases included defects involving smaller clones (<8/20 metaphases) and aberrations of Y. SNP-A allowed for detection of defects in 63% vs. 37% by MC, including 77% vs. 58% in MDS, 75% vs. 37% in MDS/MPD, 33% vs. 0% in AA, 30% vs. 0% in PNH, 59% vs. 31% in AML and 76% vs. 53% in sAML. New lesions were confirmed by paired SNP-A and microsatellite analysis. Concurrent analysis of blood and marrow showed concordant results suggesting utility of SNP-A performed on blood. Serially followed patients N=6, showed occurrence of new lesions (del(4)(q) and del(7)(q)) and earlier detection of the chr aberrations. In sAML, differential analysis of blasts and granulatocytes revealed occurrence of new lesions e.g., UPD6 or 7. In both MDS and AML, UPD of various chrs was present in 20% of patients and found in up to 35% of MDS/MPD (in addition to 9p involving also chrs 6,7,11 & 14). Other newly detected lesions included isolated/recurrent microdeletions and duplications involving genes such as AML1 or Ftl3 among others. Clinical utility of SNP-A depends on whether SNP-A karyotypig will have impact on disease parameters. In all groups tested the newly detected lesions showed impact on overall survival. While the detailed results will be a subject of our presentation, survival analysis in AML can illustrate our point; cases with a normal karyotype showed superior OS to those with newly detected defects (21 vs. 6 mo, p=.05). Similarly, new additional lesions worsen the survival as compared to those with confirmed MC (3 vs. 10 mo, p=.004). The impact on OS was also established for some of the new recurrent lesions such as UPD7q (3 vs. 39 mo, p=.002). Clinical relevance of SNP-A karyotyping is also demonstrated in AA;
it may help to distinguish AA from hypocellular MDS (clonal chr. defects, including UPD, occur in 33% of AA patients),
AA with normal SNP-A testing showed superior response to immunosuppression as compared to patients with a totally normal karyotype.
Aside of the clinical relevance, new overlapping/recurrent lesions point towards genes involved in the disease process. We conclude that SNP-A karyotyping may enhance MC in diagnosis of chr. defects and allow for a better clinical correlations of the defects with the phnenotypic and clinical features.
Disclosure: No relevant conflicts of interest to declare.