Project 3: Molecular Pathways to Breast Cancer Mortality Among African American and White Women

Partnership

This project will transfer experience, knowledge, and methodology so that a team of North Carolina Central University (NCCU) Partnership Faculty (Drs. Williams-DeVane and Chen) and their students can be full, long-term participants in the conduct of large scale population research on minority disparities. NCCU faculty will be fully integrated into the technology effort of extracting high quality nucleic acids from FFPE tissue and its analysis by various RNA expression modalities up to and including RNAseq and the handling, data basing, bioinformatics analysis and statistical integration with health services, treatment and outcomes data.

Dr. Melissa Troester is Associate Professor of Epidemiology, Co-Leader of the Cancer Epidemiology program in the Lineberger Comprehensive Cancer Center, and Co-Principal Investigator of the Carolina Breast Cancer Study. She is leading molecular epidemiologic research on disparities in breast cancer survivorship. In the current proposal, Dr. Troester will lead the selection of genes for profiling in collaboration with Dr. Williams-DeVane, will coordinate the analysis of samples in collaboration with Dr. Chen, and will lead epidemiologic analyses.

Dr. ClarLynda Williams-DeVane is the director of the Bioinformatics, Genomics, and Computational Chemistry Core and an Assistant Professor of Bioinformatics and Biostatistics at NCCU. Dr. Williams-DeVane is an expert in data integration methods. In the current proposal, Dr. Williams-DeVane will develop and implement bioinformatics methods to select genes for analysis and will collaborate with UNC investigators in classifying samples after molecular analyses are complete.

Dr. Xiaoxin Luke Chen will support Aim 1 of this Full Project, working closely with Dr. Troester’s team in extracting high-quality nucleic acids from FFPE tissues, analyzing these samples for RNA expression with Nanostring platform, and developing datasets of complex research data. This partnership will aid NCCU in building infrastructure for the analysis and integration of molecular epidemiologic data types. Dr. Williams-DeVane will leverage the CBCS data resource to develop data integration pipelines. Dr. Chen and his staff will develop expertise in RNA expression analysis (Nanostring platform, RNAseq, etc). Such expertise will benefit cancer researchers at NCCU in their future work on human tissue samples. By Year 3, we expect the partnership to produce high impact methodological manuscripts and new findings on biological pathways in breast cancer disparities. NCCU partners will also submit an R21 grant application focused on application of newly developed data integration methods to the CBCS data set.

Cancer Disparity

For more than two decades, the Carolina Breast Cancer Study has been researching epidemiologic determinants of mortality and incidence disparities among African American women. Under age 40, African American women have higher breast cancer incidence, and at all ages African American women have higher breast cancer mortality. Beginning in 1993, the CBCS Phase 1 and Phase 2 recruited cases and controls from 24 counties in North Carolina, with oversampling of African Americans to enhance power to detect effect modification by race. Significant recent discoveries from CBCS include discovery that African American and young women have higher rates of the Basal-like subtype and lower rates of ER-positive Luminal A breast cancer than Caucasian women [1], demonstration of unique risk factor profiles for basallike vs. luminal breast cancer (CITE), and discovery that even among ‘good prognosis’ Luminal A breast cancers, African American patients have poorer outcomes [5], suggesting that additional biological and health care access-related determinants of mortality can be elucidated.

To better understand the complex interaction of biology and access that drive breast cancer disparities, we undertook the Carolina Breast Cancer Study Phase 3 between 2008 and 2014. CBCS3 is a case-only study of 3000 breast cancers (1500 each African American and Caucasian). With funding from a Specialized Program of Research Excellence to UNC, we are classifying all CBCS3 tumors by intrinsic subtype and p53 expression subtype. However, there are many other important biological factors in tumors that are unstudied. Among these are: distinct estrogen response profiles that distinguish Luminal A and Luminal B breast cancers and may predict response to endocrine therapy (CITE), differences in hepatocyte growth factor signaling that are expected to vary by race and predict survival among basal-like breast cancers (CITE), and differences in immune response to tumors that are prognostic and may vary by race (CITE). Thus, the current project will advance understanding these factors that are critical to our
understanding of AA breast cancer incidence and mortality disparities, while also enhancing research collaborations between North Carolina Central University (NCCU) and UNC. We have selected the specific gene expression signatures to be studied in this project both for biological significance and for mutual interest by investigators at both institutions, to facilitate translation of basic research findings from previous NCCU-UNC collaborations. This project will refine our knowledge of how breast cancer biology varies between African American and Caucasian women, will identify relationships between risk factors and tumor biology, will identify the prognostic and predictive value of these signatures in a population based cohort, and will allow advances in our understanding of interactions between these specific biological pathways, intrinsic subtype and treatment. Ultimately, these studies will support novel clinical and public health interventions to prevent breast cancer incidence and reduce mortality.

Abstract

Relative to white women, African American women have higher incidence of breast cancer before age 40 and suffer higher mortality at all ages. The Carolina Breast Cancer Study has shown that African American women are more likely to get estrogen receptor negative breast cancer and triple negative or basal-like breast cancers. Furthermore, when African American women get estrogen receptor positive breast cancers, their survivorship is lower than white women with similar disease. To better understand the biological pathways that lead to incidence mortality disparities, we will collect RNA expression data from Carolina Breast Cancer Study tumors. The Carolina Breast Cancer Study Phase 3 is a cohort study of 3000 women with breast cancer, half of which are African American women. The study was conducted in 44 counties and used population-based sampling, therefore representing catchment for the state. Detailed treatment and clinical data are available for survivorship analyses. RNA expression levels will be assayed with a focus on the following pathways: estrogen responsiveness among luminal breast cancers, HGF-signaling among basal-like breast cancers and IGF-signaling among all cancers. Data on the expression of these pathways will be combined with existing data on other important breast cancer pathways (e.g. intrinsic subtype, p53 expression subtype, EGFR signaling, hypoxia signaling, etc.) to develop a complete picture of the biology of breast cancer disparities. The current project will extend successful methods developed in the UNC Breast Cancer Specialized Program of Research Excellence, and leverage these methods to support our Lineberger-NCCU partnership. UNC Lineberger has a state-wide mission, a top ranked school of public health, and an outstanding medical school. In this partnership, UNC will engage NCCU faculty and students in projects that would not typically be conducted at a smaller campus without these health sciences strengths. Thus, while the research addresses a health disparity, the implementation of the project will address a gap faced in conducting high impact public health and clinical research. The RNA expression analysis, conducted primarily at UNC and therefore maximally leveraging UNC resources, will support bioinformatics collaborations with NCCU and will allow for training of basic researchers from NCCU in novel methods. Training will therefore support career development of partnering NCCU investigators, while also advancing research on important pathways leading to breast cancer mortality disparities.

Co-Principal Investigators

ClarLynda Williams-DeVane, PhD NCCU	Melissa Troester, PhD UNC Lineberger