Deletions of the long arm of chromosome 5 are typical aberrations in AML and MDS. They occur either as the sole abnormality or within a complex aberrant karyotype. As the precise determination of the breakpoints and the size of the deletion is not possible with chromosome banding analysis alone we performed genomic arrays (Affymetrix 10K arrays) in 32 AML with 5q-deletion within a complex aberrant karyotype and in 17 cases with a 5q-deletion as the sole abnormality (AML: n=3, MDS as typical 5q- syndrome: n=6, other MDS subtypes: n=8). Gene expression analysis using Affymetrix U133A+B or 2.0 plus was performed in addition in all 32 AML cases with complex aberrant karyotype and in 3 cases with 5q deletion sole. The deleted region could be determined based on the genomic array data in 30 of 32 cases with AML and complex aberrant karyotype and in 9 of 17 cases with 5q deletion sole. All evaluable cases showed the 5q deletion in more than 40% of cells (interphase FISH with a 5q31-probe, range 40 to 98%), while less than 45% of interphase nuclei with 5q31-deletion in not evaluable cases (range 8% to 45%). Genomic array analysis mapped the variable proximal breakpoint in the 5q deletion sole group between 5q14.1 and 5q31.3 and the distal breakpoint between 5q32 and 5q35.1. The size of the deletion varied between 27.14 and 81.20 MB (median 73.13 MB). In cases with complex aberrant karyotype the proximal breakpoint was located between 5q11 and 5q23.1 while the distal was located between 5q32 and 5q35.3. The size of the deletion varied between 34.72 and 132.30 MB (median 94.77 MB). Two approaches were tried to determine the size of the deletion based on gene expression data. As a control groups 40 AML and 40 MDS both with normal karyotype were used. Each probe set expressed in at least 1 case of the control group with a precise localiasation on chromosome 5 available was included in the analysis. For each probe set a median expression within the control groups was calculated. First for each probe set a ratio between the individual patient with 5q deletion and the median expression of the respective control group was calculated. For each chromosomal band on chromosome 5 the proportion of genes showing a ratio <1 was determined. For chromosomal bands within the deleted region the proportion of genes showing a ratio <1 varied between 71–100% (mean 82%), while for chromosomal bands not involved in the deletion the proportion varied between 34–68% (mean 60%) (p<0.0001). In the second approach a median expression was calculated for all genes located within a certain chromosomal band. A ratio was calculated between each individual patient with 5q deletion and the control group. For chromosomal bands within the deleted region the ratios varied between 0.38 and 0.7 (mean 0.58), while for chromosomal bands not involved in the deletion the ratio varied between 0.71 and 1.4 (mean 0.97)(p<0.0001). In conclusion, 5q deletions are significantly larger in AML with complex aberrant karyotype compared to MDS/AML with a 5q deletion as the sole abnormality (mean: 95.65 MB vs 65.63 MB, p<0.001). 5q deletions can be identified in individual patients based either on the expression of genes located on chromosome 5 in comparison to the expression of 5q genes in a control group of AML or MDS with normal karyotype or by genomic arrays. These findings demonstrate that AML and MDS with 5q deletion show a loss of large genomic regions leading to a decreased expression of a large number of genes.

Disclosures: Torsten Haferlach is a consultant for F. Hoffmann-La Roche Ltd, Basel.; This study is supported in part by Roche Molecular Systems, Inc.

Author notes

*

Corresponding author

Sign in via your Institution