Abstract In a previous report (López et al, ASH 2007, abstract 1916) we have demonstrated the presence of cytogenetic aberrations on mesenchymal stem cells (MSCs), the bone marrow (BM) stroma progenitors, from patients with myelodysplastic syndromes (MDS). Since there is a chance that the observed changes could be the result of the expansion in the culture process the aim of the present study was to analyze whether genomic changes could be present in non-expanded BM MSC and also to assess if the genomic changes detected could be correlated with specific subtypes of MDS. To address our first issue, in four MDS cases, MSCs were enriched by sorting mononuclear cells from BM with the following phenotype: CD45−/CD73++/CD34−/CD271+++, purity being >99%. Array based comparative genomic hybridization (array-CGH) was performed with this cell population. DNA was extracted, and after it was amplificated using the GenomiPhi kit. The postamplification cleanup was achived by ethanol precipitation using the sodiumacetate/EDTA. In order to know whether the amplified products contained actual genomics a PCR, using specific primers was performed; both test and reference DNA were amplified using the same amount of starting DNA. With this approach genomic changes were observed in this sorted cell population, with gains being more frequent than loses. The more frequently affected genomic regions were: 1q31, 10q26 and 20q13. In order to be sure that these features were also present in expanded MSCs, FISH using the same bacterial artificial chromosome (BAC) probe than used in the array-CGH studies was performed on expanded MSCs from the same patients and the same genomic alteration was observed, showing that not only expanded but also freshly isolated MSCs have genomic aberrations. Once demostrated that. uncultured MSCs were genomically abnormal we wanted to know whether these aberrations could be related to any clinical parameter. For this purpose expanded MSCs from 13 MDS patients were studied using array-CGH and an unsupervised hierarchical cluster analysis performed. Two different clusters were identified: one of them included all the 5q- syndrome patients, while the other incorporated the remaining MDS subtypes. The differences between these two clusters occurred just in 26 BAC. When looking at these clones and taking into account only regions involving at least two consecutive BACS, chromosomal regions 11q13, 17q25, 19q13, 20q13, and 22q13.2 were gained in 100% or 80% of the 5q-cases; by contrast a normal copy or a deletion of these regions were observed within the rest of patients. Our results show, for the first time, that both cultured-expanded and freshly-isolated MSC from MDS display genomic aberrations, as assessed by array-CGH and FISH, some of them specially linked to a specific clinical entity, the 5q- syndrome.