High-Resolution Genomic Profiles Identify Novel Genes and/or Chromosomal Regions with Prognostic and Oncogenic Significance in Myeloma Patients.

Despite major improvements in the treatment of myeloma over the last decade, disease course is variable due to genetic heterogeneity. Even though cytogenetics is a difficult art in myeloma, chromosomal abnormalities are present in 100% of the patients. In order to define the spectrum of unbalanced chromosomal abnormalities in myeloma, we studied a cohort of 200 patients ≤ 65 years of age, homogeneously treated with high dose chemotherapy in the IFM 99 trials, using the Affymetrix 500K SNP arrays. Bone marrow was obtained at diagnosis, and plasma cells were sorted using anti-CD138-coated magnetic beads. In all cases, a plasma cell suspension with a purity > 90% was obtained, and DNA was extracted using standard methods. After labeling, DNA was hybridized on the two Nsp and Sty chips, containing 260,000 SNP each. As expected, copy number variations were observed in all the cases. The most frequently involved chromosomes were chromosomes 13, 16, 1, 8 and 6. We then performed a prognostic analysis to correlate survival and chromosomal abnormalities. Many previously unreported chromosomal regions were identified; for example, 89 deleted chromosomal regions as well as 53 regions present in > 2.5 copies were associated with a survival < 3 years. Among the deleted regions with prognostic impact, the most frequent were 1p, 12p, and 4q. The most frequently gained regions associated with overall survival were 1q, 8q, and chromosomes 7 and 21. A similar analysis was performed for association with event free survival (EFS). Losses at 11q, 4q, and 12p were highly predictive of a short EFS, whereas gains of chromosome 15 predicted a longer EFS. Global analysis revealed a marked heterogeneity in the gains observed in hyperdiploid karyotypes, explaining the inconsistency in the literature regarding the prognostic value of hyperdiploidy. A more detailed analysis identified more than 80 genes presenting a double deletion, suggesting their role as putative tumor suppressor genes. This study represents the largest cohort of patients analyzed with such high-density arrays, enabling the description of a highly confident landscape of chromosomal abnormalities in myeloma. Moreover, since all the 200 patients were treated with a similar high-dose regimen, it allows powerful statistical analysis of the impact of chromosomal gains and losses on response to therapy, event free survival and overall survival. Finally, since all the patients were also annotated for other prognostic parameters such as t(4;14) or beta2-microglobulin, a multivariate prognostic model will be presented.