8536 Background: t-AML is a syndrome occurring after exposure to chemo or radiotherapy. Since for similar treatments only some patients ends developing a secondary leukemia, it has been proposed a genetic predisposition associated to this syndrome. Methods: To analyse single nucleotide polymorphisms (SNPs) on genes that could be involved on risk of developing t-AML by means of RFLP and SNP genome screening using high density microarrays .Two groups of individuals were genotyped: Group A, composed by patients that develop t-AML after chemotherapy for breast cancer (BC) and Group B (control), formed by chemotherapy treated BC patients that after a period of more than 10 years have not developed t-AML. We have studied 12 polymorphisms on genes from drug detoxification pathways (NOQ1, GSTP1), DNA repair (XPC[3 ], XRCC1[2 ], NBS1, ERCC5 and XRCC3) and DNA synthesis (MTHFR[2 ]), in which the nucleotide change implies a change in the protein sequence (nA=16, nB=18) . Alternatively, for each patient, more than 10.000 SNPs were genotyped by means of of high density microarrays (Affymetrix) (nA=12, nB=18). The alele frequencies for each SNP between two groups were compared. Results: In RFLP study, we observe two SNPs on MTHFR gene displaying remarkably different allele frequencies between BC patients (Table). In microarray study, we found 12 SNPs with differences of allele frequency higher that 45% between A and B groups, located 6 on chromosome 8. Conclusions: The results suggest that the MHFTR gene is a candidate for being studied by its possible relation with the genetic predisposition to develop t-AML after BC treatment although its functional implication with the disease must still be elucidated. Moreover, data from SNP arrays suggest that the genome regions marked by those 12 SNPs, specially those on chromosome 8, are candidate for being studied by its possible relation with the genetic predisposition to develop t-AML after BC treatment. Financed by FIS G03/008. [Table: see text] No significant financial relationships to disclose.

INTRODUCTION: Therapy-related MDS/AML constitutes the most frequent secondary neoplasia and his incidence is increasing in the last years. This incidence is especially high when myeloablative chemotherapy is used before autologous stem cell transplantation (ASCT). Since for similar treatments only a percentage of the patients ends up by developing a secondary leukemia, it has been proposed the existence of some type of genetic predisposition associated to this syndrome. OBJETIVE: To search single nucleotide polymorphisms (SNPs) associated with the genetic predisposition to develop a t-MDS/AML by means of SNP genome screening using high density microarrays. MATERIAL AND METHODS: Two groups of individuals were genotyped: Group A, composed by 16 patients with t-MDS/AML post-ASCT (10 NHL, 4 HD, 2 multiple myeloma) and Group B (control), formed by 16 patients submitted to ASCT and that after a period of more than 10 years have not developed t-MDS/AML(14 NHL, 2 HD). For each individual more than 10.000 SNPs were genotyped by means of the use of microarrays of high density (Mapping 10K, Affymetrix®). Those SNPs are distributed along the whole genome and have a high average heterocigosity (0,37). The alele frequencies for each SNP between two groups were compared. RESULTS: We found a group of 4 SNPs located in the q23.3 band of the chromosome 5 in which the differences of alele frequency between the groups A and B were superior to 50 %. These SNPs are located in a region that comprises the only gene, CSS3 (chondroitin sulphate synthetase 3), being the difference of alele frequency of 57 % in the intronic SNP of this gene. CONCLUSIONS: The results suggest that the CSS3 gene is a candidate for being studied by its possible relation with the genetic predisposition to develop LMAt/SMDt, although its putative functional implication with the disease must still be elucidated. There is known that a counterpart of CSS3, CSS1, plays a key role in the proliferation and apoptosis of myeloma cells. Thus, it is possible that CSS3 could be somehow related with the hematopoyesis pathways and implied in the development of t-MDS/AML. Financed by FIS G03/008.