Array-Based Comparative Genomic Hybridization (Array-CGH) in Multiple Myeloma (MM) and Plasma Cell Leukemia (PCL): Recurrent Imbalances within a Highly Heterogeneous Background of Genomic Disruption.

Introduction: array-CGH is significantly impacting on cancer cytogenetic. We used this technique to perform pan-genomic screening in 15 patients with MM and 4 with de novo PCL. Patients and methods. Bone marrow samples were employed. If tumor contamination was below 20% plasma cells were purified with Myltenyi columns. Array-CGH was performed as follows: genomic DNAs, from both the tumor and normal reference cells, labeled with different fluorescent dyes were cohybridized to 1 Mb resolution arrays containing 2600 Bacteria Artificial Chromosome (BAC) clones (Spectral Genomics Inc, Houston, TX, USA) according to manufacturer’s recommendations. Variations in DNA sequence copy number for each BAC clone was assessed by relative fluorescence signal intensities, in a single hybridization, providing a locus-by-locus measure of DNA copy-number changes.

Results: The assay was validated as follows: three normal DNAs were tested revealing no genetic imbalances. One normal male was tested against one normal female and sex chromosome (ch) imbalances were effectively detected. Array-CGH results for ch 13q14.3 deletions were matched with FISH results and concordance was seen in 87% of cases. The median number of lesions/patient observed in our panel was 17 (4–135) (fig 1a). Also the amount of the total genome affected by chromosomal imbalances was highly variable (median 3.9% range: 0.14%–27%) (fig 1a). The amount of involved genome did not correlate with the actual number of lesions (fig 1a). A good correlation was noticed between the amount of losses and gains in each patient (fig1b). Notably PCL do not have a more disrupted genome compared to MM patients as one might expects based on the highly malignant behavior (fig 1a). Interestingly two patients with a prolonged clinical history of MGUS prior to MM diagnosis had massive presence of losses ad gains. Of 2600 BACs 934 were never affected, 864 were targeted only in one patient (pt), 401 in two pts, 296 in 3–5 pts and only 105 were targeted in six pts or more (fig 1c). These 105 recurrent imbalances could be attributed to 9 different abnormalities. Among these we have identified five recurrent lesions (occurring in at least six patients) that have not been previously described. These are 19p13.2 (gain 9 pts, loss 1 pt), 14q12 (loss 3 pts, gain 4 pts), 16q12.1 (loss 6 pts) 11q24 (gain 6 pts), 9q23 (gain 6 pts).

Conclusions:

1. array-CGH allow effective pan-genomic screening of MM patients;

2. the pattern of genetic disruption is highly heterogeneous with a majority of non-recurrent or uncommonly recurrent lesions;

3. a number of highly recurrent lesions have been identified that will require assesment for prognostic impact;

4. the overall amount of perturbed genome does not seem to correlate with more aggressive disease, and might be the reflection of alternative biologic features (f.e. a prolonged history of clonal disease).

View largeDownload slide

Figure

View largeDownload slide

Figure

Close modal