Use of SNP-array-based karyotyping for cytogenetic prognostication in unclassified cases of myelodysplasia and associated overlap disorders

Abstract

7016 Background: Myeloproliferative disorders (MPD) and myelodysplastic syndromes (MDS) often have overlapping features resulting in unclassifiable cases (MDS-U and MDS/MPD-U). Chromosomal abnormalities impact prognosis, but 50% of cases show normal karyotype by metaphase cytogenetics (MC). Single nucleotide polymorphism arrays (SNP-A) are novel karyotyping tools with superior resolution and ability to detect copy neutral loss of heterozygosity, a defect not detected by MC. Methods: MDS-U (N = 17) and MDS/MPD-U (N = 61) patients were selected from an MDS database (N = 720, median age = 76, median follow-up = 42 mos). SNP-A was performed on 67 patients and 751 controls. An algorithm for identification of somatic lesions was designed: 1) Lesions detected by MC and SNP-A required no further analysis; 2) Micro-duplications/ deletions overlapping with copy number variants (CNV) were excluded. Lesions not in CNV databases were confirmed by CD3 lymphocytes; 3) UPD <25 Mb were unlikely somatic and excluded. Telomeric and interstitial (≥ 25 Mb) UPD were considered somatic. International Prognostic Scoring System (IPSS) was used to assess routine risk. Fisher's exact test was used for categorical variables. Overall (OS) and event-free (EFS) survival defined by the MDS working group criteria were analyzed by Kaplan Meier analysis (log-rank or Wilcoxon and 2-sided significance). Results: SNP-A yielded superior detection rate for chromosomal defects compared to MC (71% vs 47%, p = 0.008). UPD was seen in 17 patients and frequently involved chromosomes 1, 3, 6, 8, 11, 17. MDS/MPD-U and MDS-U patients had similar OS and EFS (OS = 42 vs. 45 mos, p = 0.13; EFS = 42 vs. 45 mos p = 0.63). SNP-A revealed a more complex karyotype in patients with advanced MDS. Furthermore, SNP-A karyotyping resulted in prognostic refinement of previously assigned IPSS: Unclassified cases = 6% versus 0%, int-1 = 45% versus 53%, int-2 = 6% versus 19%, high = 5% versus 8%. Overall, patients with new SNP-A lesions had worse OS and EFS (OS = 41 mos vs NR, p = 0.07; EFS = 32 vs 112 mos, p = 0.07). Conclusions: SNP-A karyotyping complements MC in detecting chromosomal defects in MDS-U and MDS/MPD-U. This technology will be helpful in refining diagnosis based on characteristic recurrent chromosomal lesions including UPD. No significant financial relationships to disclose.