Using metaphase cytogenetics (MC), chromosomal abnormalities can be found in approximately 50% of patients with MDS (in our cohort (N=356), 46% of patients had normal karyotype by MC), in whom they have important prognostic implications. However, patients with identical lesions, including normal karyotype may show variable clinical behavior. We hypothesized that if a more precise method is used, previously cryptic karyotypic lesions can be found in patients with known aberrations as well as in those with normal MC. High-density SNP arrays (SNP-A) can be used for precise LOH and gene copy number analysis. We have applied this new platform (Affymetrix 250K SNP arrays) to study chromosomal lesions in bone marrow samples from 112 MDS patients, 6 hematologic and 36 healthy controls. Our MDS cohort included patients with RA/RCMD (N=30), RARS/RCMD-RS (N=18), RAEB1/2 and sAML (N=45), and CMML1/2 (N=19); by traditional MC, aberrations of chromosomes 5, 8, 7, and complex karyotypes were present in 13%, 9%, 6% and 8% of patients, respectively. A normal MC exam was obtained in 44% of this sample; in 4% of cases the results were non-informative due to lack of growth. We first applied whole genome scan by SNP-A to establish parameters for minimal pathogenic lesions in healthy controls in whom copy number polymorphisms were easily detectable, but only a limited number of small random defects was found (O’Keefe, ASH 2006). Hematologic controls all showed a normal whole genome scan. However, when this method was applied to MDS patients, chromosomal aberrations were detected in 79% (vs. 52% by MC, p<0.001). Previously unrecognized lesions were detected in both patients with a normal MC test, as well as in those with known lesions. Consequently, a higher proportion of patients showed >1 genomic lesion (e.g. for MC vs. SNP-A, 2 defects in 10/112 vs. 27/112, and ≥3 in 9/112 vs. 31/112, respectively). Newly identified lesions were confirmed by microsatellite and TaqMan PCR copy number analysis in clonal and wt hematopoietic cells. Most significantly, in a proportion of patients, we have identified segmental uniparental disomy (UPD), a lesion resulting in LOH that cannot be detected by MC; it was found in 24% of patients. Most often, UPD involved chromosomes/regions that are frequently affected by loss of genetic material, including chromosome 7q (N=5), 11q (N=5) and 6p (N=3), but also in chromosomes 1 (N=5) and 17 (N=3). As a result, shared areas of LOH were identified in a higher proportion of patients. For example, in addition to known 7/7q deletions (N=7), we have detected 2 new losses involving 7q34 (N=3) and 7q22.1 (N=2) as well as UPD in 7q (N=5), increasing the proportion of patients with aberrant chromosome 7 from 6% by MC to 15% by SNP-A (p<0.03). Clinical analysis of the impact of previously cryptic lesions analogous to those with established adverse prognostic impact (new del7/upd7 or complex) suggests that that SNP-A karyotyping will have clinical utility above and beyond the value of MC. In sum, SNP-A-based karyotyping allows for precise detection of chromosomal lesions in MDS. Previously cryptic defects, including UPD may have clinical and prognostic relevance and help identify genes responsible for the phenotype of the dysplastic clone.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author