Abstract
Disparities in breast cancer stage of presentation and survival rates exist in patients of different ethnicities. These differences are undoubtedly a result of a combination of factors, including socio-economic, lifestyle, tumor characteristics and inherent factors, such as genetic composition. Our group remains focused on analyzing genetic/genomic contributions to these disparities, with the ultimate goal of increased biological understanding, leading ultimately to individualized, ethnic-specific diagnostic and therapeutic approaches. We have previously reported ethnicspecific expression patterns in matched tumor and adjacent samples from a cohort of triple negative breast cancer (BC) samples. Here we report results from a parallel study focusing on gene expression profiling in a multi-ethnic collection of normal breast tissues. Study samples were cut from FFPE (formalin fixed paraffin-embedded tissue) blocks saved from local reduction mammoplasty cases [5 Caucasians (CAU); 7 African-American (AA); and 4 Hispanics (HIS) women with no personal or family BC history]. These were sent to Almac Diagnostics for RNA isolation, cDNA preparation, and hybridization of cDNAs to a cancer focused gene expression array (Xcel) containing 110,961 probes, representing 19,905 unique known genes. Arrays were quantile normalized and log transformed to the median of all samples. The probes were filtered to remove variation within each ethnic group to ≤ 0.5 SD, while the variation between the three groups was maintained at ≥ 0.2 SD. Samples which had < 90 % similarity based on ‘Pearson correlation’ were removed from subsequent analyses. This removed three AA, 2 CAU and 1 HIS samples. Finally, the filtered subset of 68,145 probes common to the three ethnic groups was then compared on GeneSpring® analytical software. A mean centered Principal Component Analysis (PCA) was performed on the filtered subset of normalized data. In addition, a one way ANOVA between the three ethnic groups identified 1884 significantly differentially expressed probes representing 237 unknown and 1647 known genes (P ≤ 0.05; Benjamin-Hochberg multiple testing corrected). Post-hoc analysis of the 1884 significant probes identified 207 probes significantly differentially expressed between CAU and AA; 1863 probes were found to be differentially expressed between HIS and CAU, and 1873 probes were significantly differentially expressed between HIS and AA. These results suggests that the HIS group/samples have a unique gene signature in comparison to the other ethnic groups analyzed. When the differentially expressed gene list across ethnicities is reduced to those genes with 1.5 fold change, the number of statistically-significant differentially-expressed genes across ethnicities decreases dramatically (7 AA vs. Cau; approx 600 AA vs HIS; 600 CAU vs His), suggesting a much more select group of differentially-expressed genes in normal breast tissue across these ethnic groups. Since the overall sample size is small, we are continuing to validate these data on a larger set of samples. In addition, we are performing laser capture micro-disssection of these samples, to compare gene expression patterns in stroma vs. epithelial. We will present our latest study results, including pathway analysis of statistically-significant differentially-expressed genes across the three ethnic groups.
Citation Information: Cancer Epidemiol Biomarkers Prev 2011;20(10 Suppl):B39.