Gene expression analysis of African-American and European-American breast tumors

PR-3 Observed differences in survival between African-American (AA) and European-American (EA) breast cancer patients are attributed to differences in socio-demographic and healthcare factors. However, recent studies have found that, after accounting for these differences, AA still exhibit lower breast cancer survival rates than EA, suggesting that differences in tumor biology may contribute to survival. AA breast cancer patients tend to have a greater prevalence of more aggressive, poorly differentiated, estrogen-receptor (ER) negative breast tumors and a higher rate of lymph node involvement than EA. More recent data indicate that AA ethnicity is an independent predictor of poor response to therapy. Despite this evidence, few studies have examined how differences in tumor biology between AA and EA breast cancer patients contribute to differences in survival. We examined gene expression profiles of micro-dissected breast tumors from 35 patients (18 AA, 17 EA) with invasive breast cancer using Affymetrix HG-U133A which contains 22,283 probe sets match to transcripts from approximately 13,000 human genes. The two groups of patients were well matched on clincopathological characteristics. We used laser capture microdissection to analyze the stromal gene signature separate from the tumor epithelium signature. The resulting datasets were analyzed on the gene and pathway level for expression differences overall, and by ER status. We identified several pathways and biological processes that were significantly enriched for genes with expression differences in the tumor epithelium and tumor stroma between AA and EA for both comparisons. The most significant identified processes included angiogenesis, blood vessel development in the tumor stroma and the regulation of chemotaxis and antigen presentation/processing in the tumor epithelium. These differences were independent of other prognostic factors that significantly affect the tumor gene signature such as p53 status. We used in-house software to find the previously recognized disease associations of those genes that were differently expressed in breast tumors by race/ethnicity. This analysis revealed a common association of these genes with diseases that arise from an inflammatory environment, indicating that differences in inflammation and immunobiology of breast tumors may play a role in the survival health disparity. We are currently corroborating the gene expression differences of those genes by RNA and protein analyses using a validation set of 60 breast cancer patients (AA and EA, both n=30). We are also examining non-cancer tissue samples to determine if these genes are generally differently expressed in the two race groups of our study. In conclusion, this study revealed that race/ethnicity is associated with differences in gene expression of breast tumors that may influence disease aggressiveness and response to therapy.