Imbalances of Chromosome 21 in MDS/AML Include Cryptic Deletions but NOT RUNX1 Amplifications.

Abstract 2561

We present the results of a study demonstrating that the genome profile of RUNX1 in MDS/AML is characterised by hitherto unreported partial deletions and absence of amplifications. This is in stark contrast to reports of chromosome 21 amplifications in ALL. We speculate that the absence of RUNX1 deletions results from them being well below a size detectable by commercial FISH probes.

Extra chromosome 21 is the second most common acquired trisomy after (+) 8 in adult myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). It is rarely observed as sole abnormality but seen as part of complex karyotype in some 3–7% of the AML (Atlas of Genetics and Cytogenetics in Oncology and Haematology, http://atlasgeneticsoncology.org). Although the gene(s) in trisomy 21 associated with leukemia are unknown, the 21q22 region appears to be critical since it houses the RUNX1 gene. Multiple amplified copies of the RUNX1 carried by marker chromosomes, such as iAML21, are described in both acute lymphoblastic leukemia (ALL) and AML. A common 5.1 Mb amplification containing the RUNX1, miR-802 and genes mapping to the Down syndrome critical region identified in 91 children with iAML21, was shown to be the likely initiating event in this rare form of childhood B-cell ALL (Rand et al., Blood, 2011). In contrast, recent studies of AML in a Down syndrome and a constitutionally normal individual showed lack of RUNX1, ETS2 and ERG involvement (Canzonetta et al., BJH, 2012). Here we present 16 MDS/AML cases with imbalances of chromosome 21 identified by genomic array screening from a cohort of 83 cases. Whole genome screening (aCGH) was performed on presentation samples of MDS /AML and de novo AML cases using an oligonucleotide array platform (Agilent) at 60K, 244K, 400K and 1M density. G banding and FISH analysis were also successfully performed.

Gain of an extra copy (trisomy) of chromosome 21 (+21) was found in 9 patients, all but one with complex karyotypes. In 2 AMLs high level amplifications were detected at 21q22, which involved the ETS2 and ERG but not the RUNX1 sequences. While several commercially available RUNX1 FISH probes showed gene multiple signals, custom FISH probes covering the relevant regions confirmed that the amplifications excluded the RUNX1 but affected both EST2 and ERG thus rendering the commercial probes unfit to assess CNA in this genome area.

In another two cases with trisomy 12, cryptic loss of 43Kb and 98Kb resp. within the RUNX1 sequences was detected and confirmed by FISH. Furthermore, similar deletions within the 21q22.12 were also found in another 7 cases all of which had diploid set of chromosome 21 but had multiple changes at G banding level and high TGA score. These RUNX1 deletions were variable in size, ranging from 98Kb to 2.7Mb. Although our observations excluded clinical correlations it is note worthy that most of the patients with RUNX1 loss have not achieved complete cytogenetic remission.

These findings suggest role for the RUNX1 loss as indicator of progressive disease and provide a novel insight into pathogenesis of MDS/AML.