Frequent Molecular Aberrations in Normal Karyotype (NC) Precursor B-Cell Acute Lymphoblastic Leukemia (pB-ALL) Detected by Comparative Genomic Hybridization.

The prognosis of adult NC pB-ALL has not improved over the last decade mainly because separation into distinct molecular subsets has been lacking. We conducted a proof-of-principle experiment to screen the genome of blasts from adult NC pB-ALL patients for novel genomic alterations using 19,000 bacterial artificial chromosomes (BAC) by array CGH and verified our results with gene expression profiling (GEP, Agilent oligonucleotide expression arrays) and fluorescent in situ hybridization (FISH) using the same BAC probes or commercially available probes if available. The median age at diagnosis of the 6 males and 4 females was 55 years; all were treated with a 5-drug induction regimen followed by intensive consolidation and maintenance regimens. Five achieved CR of whom 3 relapsed with complex karyotype; samples were available on 2 of the patients at relapse. Samples with <60% blasts (2 at diagnosis, 1 at relapse) were enriched for >95% purity by sorting according to their immunophenotype pattern. Only one case had a normal CGH pattern. A decreased copy number of p16 was found in 4 of the 10 diagnostic samples; 3 on whom mRNA was available demonstrated a decreased p16 mRNA expression by GEP. In addition, a loss at 9p21.2 and a loss and a gain at 8q21.2 were detected in one sample each; those were verified by FISH. Copy number losses were also observed across multiple samples at 1p34–1p36, 4p16, 5p15.33, 8q24, 9q34, 10q26, 11q12–11q13, 12q24, 14q32, 17p, 19p13.3, 22q11, and 22q13. Copy number gains across multiple samples were observed at 15q11.2 and 18q21. Those are currently being verified by FISH. Comparison of samples at diagnosis and at relapse revealed a disparate molecular pattern. In summary, enriching samples by sorting from NC pB-ALL to be studied by CGH is feasible; normal chromosome pattern does not correlate with normal molecular pattern; and diagnostic and relapsed samples have a different molecular pattern. Our work on 10 patient samples at diagnosis and 2 at relapse demonstrates the feasibility to detect specific molecular aberrations in NC pB-ALL. Since attaining good quality metaphase cells in ALL is often suboptimal, this work suggests that array CGH can complement conventional cytogenetic analysis to better characterize NC pB-ALL by identifying cryptic copy number changes. Work on a larger number of samples is warranted to determine the prognostic significance of the aberrations detected herein.