Abstract B35: Cumulative results from an investigation of transcriptome differences in breast cancer samples from African American and East African triple-negative breast cancer patients

Abstract

Abstract We report our cumulative results of a multi-year study focusing on the genomic similarities and differences between Triple-negative breast cancer (TNBC) in native African and U.S. African patients. We used a cohort of African-American (AA) TNBC patients from South Florida and an indigenous cohort of Kenyan TNBC patients. Strong U.S. epidemiological data supports breast cancer (BC) as the second leading cause of cancer death among AA women, with a 20% greater mortality rate than that in Caucasian-American women (CA). AA women often present with tumors that are of higher grade and later stage and their relapse-free and overall survival period is significantly lower than CA. Additionally, TNBC has a higher prevalence in AA women. All of these observations lead to a much poorer prognosis for AA BC patients- thus defining the all-encompassing term “disparities in BC in women of African ancestry”. In our study, we used archived formalin fixed paraffin embedded (FFPE) blocks, collected across a stage-matched multi-ethnic U.S. cohort of node-negative TNBC from the University of Miami and similar TNBC samples from the Kijabe Hospital (Kenya). Using 10 μm scrolls from each block, total RNA was isolated, cDNA prepared, and hybridized to a breast cancer enriched gene expression array (Affymetrix Platform BC DSA Research Tool) in collaboration with Almac Diagnostics. Gene expression analysis was conducted using GeneSpring 12.1® analytical software. The raw data, a total of 60,856 gene/probes, was pre-processed using Robust MultiArray Average (RMA) algorithm; background corrected, log2 transformed to the baseline median of all samples and then quintile normalized. The samples that passed final QC were then filtered to remove low-intensity signals (< 20% signal). The final analysis cohort consisted of 10-AA, 13-CA and 21-Kenyan. PCA analysis revealed that the samples clustered well with respect to ethnicity. Unsupervised cluster analysis, based on ethnicity and genes (p value <.05, fold change >2.5), was performed. The resulting dendogram clearly segregated into distinct subgroups based on ethnicity, revealing a pattern of differential gene expression between the cohorts. A list of differentially expressed genes from each cohort (DEG) were selected using ANOVA analysis (fold change > 3.0, p value <.05) followed by the Benjamin and Hochberg method for multiple-testing correction. Finally, the lists of DEG were uploaded into GeneGo MetaCore to identify functionally enriched pathways. These analyses revealed significantly deregulated genes associated with the Wnt/β-catenin pathway in the AA cohort, as compared to the CA, suggesting that this pathway may contribute to the more aggressive phenotype in AA women diagnosed with TNBC. Additionally, significantly deregulated genes associated with the Oncostatin M pathway were discovered in the Kenyan cohort, as compared to the AA and CA tumors. In particular, STAT1 was significantly downregulated in the Kenyan cohort compared to the CA and AA. Thus, our results indicate gene expression differences within several key pathways across these ethnic groups. In summary, this study represents the first direct comparison of gene expression in TNBC specimens across U.S. CA and AA BC patients and Kenyan East Africans. These studies have important implications for further understanding BC ethnic disparities, as well as tailored approaches to prediction, prevention and therapeutic measures. Citation Format: Julie E. Getz, Mary E. Ahearn, Carmen Gomez, Biju Issac, Jennifer Clarke, Mark D. Pegram, Peter Bird, John D. Carpten, Lisa L. Baumbach. Cumulative results from an investigation of transcriptome differences in breast cancer samples from African American and East African triple-negative breast cancer patients. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B35. doi:10.1158/1538-7755.DISP13-B35