Loss of heterozygosity (LOH) is detectable in many forms of cancer, including leukaemia. LOH contributes to tumorigenesis through the loss of one of the two alleles of one tumor suppressor gene at a given locus, caused by deletion or uniparental disomy (UPD). In this study, we analysed 146 patients with primary AML, using high-density single nucleotide polymorphisms (SNPs) arrays. Leukemic cells and healthy T-cells from each patient were obtained using FACS-Vantage cell sorting. The genome wide SNP analysis was carried out using Human Mapping 10K Arrays (Affymetrix). So far LOH was identified in 27 cases with normal karyotype (18.5%). Potential LOH regions were confirmed by analysis of short tandem repeat (STR) markers using DNA samples from T-cells and blasts. In 21 of these cases, STR-analysis of T-cells representing the corresponding tumor-free material confirmed the regions of UPD. This aberration affected different chromosomes, but most frequently chromosome 2, 6, 11, 13, 21, and covered between 11.5 and 88 Mb. In the remaining 6 LOH cases, long stretches of homozygosity occurring at identical positions in the T-cells and in the leukemic blasts were found. These cases were detected and separated from those with UPD.

So far, correlation of clinical variables and LOH results in this cohort did not indicate significant differences based on age, WBC-counts, sex and FAB-subtype. Additionally, we did not detect any proven correlation between FLT3-ITD, MLL-PTD, NPM1 mutations and UPD. Interestingly, 80% of the patients with UPD that had achieved CR after chemotherapy relapsed within 6 months. In contrast, the rate of relapse in patients without UPD was only 54% . Furthermore, we were able to confirm the relationship between UPD and gene aberrations. We demonstrated that the UPD cases found in chromosomes 21, 19, and 11 were correlated with point mutations in the RUNX1, CEBPA and WT1 genes, respectively. Furthermore, AML cases with and without UPD showed different gene expression profiles, revealing different expression levels for some genes involved in double strand brake repair.

In conclusion, the use of the SNP array technology allowed the identification and mapping of LOH in 18.5% of AML patients with normal karyotype. Our data also underline the necessity of analysing tumour-free material in order to confirm the somatic origin of the alteration. Furthermore, our results indicate a higher relapse rate in patients with UPD.

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author

Sign in via your Institution