Advancing Our Understanding of the Etiologies and Mutational Landscapes of Basal-like Luminal A, and Luminal B Breast Cancers

Translational Research Program

Advancing Our Understanding of the Etiologies and Mutational Landscapes of Basal-like Luminal A, and Luminal B Breast Cancers

Principal Investigator: Christopher Li, MD, PhD

Background: Advances in our understanding of the molecular heterogeneity of breast cancer have led to the identification of several distinct molecular subtypes of breast cancer. While triple-negative (TN) breast cancers which primarily consist of basal-like tumors account for only ~15% of all invasive breast cancers, they are of considerable clinical and public health importance. This is because they have substantially poorer 5-year survival rates (35-80%) compared with luminal A breast cancer (the most common molecular subtype which has a ~90% 5-year survival). TN cancers also disproportionately impact young, African American, and Hispanic women. While the molecular signatures of these tumors have been well characterized, their risk factors and mutational spectrum are largely unknown since only a handful of relatively small studies have focused on advancing our understanding of the etiology of these tumors. Thus, there is a strong need for large scale studies of these poor prognosis tumors that evaluate both their epidemiology and underlying mutational landscapes.

Objective/Hypothesis: The objective of this proposal is to employ state of the art multidisciplinary approaches to advance our understanding of the unique etiologies of basal-like, luminal A, and luminal B tumors. Through this new synergistic collaboration we will elucidate both the risk factor profiles and mutational landscapes of the most common breast cancer molecular subtypes with a particular focus on poor-prognosis basal-like disease.

Specific Aims: With a sample size that is substantially larger than any other published study of basal-like breast cancer, this population-based case-control study will enable us to address the following specific aims:

1. Identify and quantify risk factors for each of the most common molecular subtypes of breast cancer, basal-like, luminal A, and luminal B tumors, in a large-scale population-based study. We will evaluate all of the established breast cancer risk factors and additional hypothesized risk factors. Preliminary data indicate that oral contraceptive use and parity are more strongly related to risk of basal-like tumors.

2. Discover and validate the mutational landscapes of basal-like, luminal A, and luminal B tumors. Using next generation sequencing, we will characterize a large majority of point mutations, indels, amplifications/deletions and gene fusions from RNA and DNA isolated from formalin fixed paraffin embedded (FFPE) tissue specimens in a subset of our cases and use the remaining cases for validation.

3. Characterize the relationships between subtype specific risk factors and mutational signatures. The biological underpinnings of the stronger relationships between certain risk factors and basal-like breast cancer are unknown. Through this aim we will identify how various exposures influence the tumor genome.

4. Develop and validate risk prediction models unique to each breast cancer subtype incorporating epidemiologic and clinical data. Basal-like tumors are more likely to be interval detected rather than screening detected given the rapidity of their growth. The identification of women at high risk for these tumors, who may benefit from more frequent screening or imaging with complementary modalities, has the potential to identify these tumors earlier, when they are more treatable, and improve their prognosis. 5. Identify and quantify the relationships between various exposures and mutational changes on risk of breast cancer recurrence among patients with basal-like, luminal A, and luminal B tumors. While patients with basal-like and to a lesser extent luminal B tumors have poorer outcomes compared to luminal A patients, little is known about factors that influence prognosis.

Study Design: We propose a population-based case-control study that will include 900 TN breast cancer patients, 1800 ER+ breast cancer patients, and 900 population-based controls 20-69 years of age living in the metropolitan Seattle-Puget Sound region. All participants will complete a telephone interview through which we will collect comprehensive information on established and suspected breast cancer risk factors. The medical records of cases will be reviewed to obtain data on breast cancer treatments and disease recurrences. Tumor tissue specimens from cases will be ascertained and centrally reviewed and tested so that basal-like cases can be identified among the TN cases, and so that luminal A and luminal B cases can be distinguished from each other. Tumor DNA/RNA and germline DNA will be extracted from FFPE tissue and a subset of the cases (n=400) will be subjected to target capture whole exome sequencing (70-100x coverage) of DNA, along with whole exome capture of cDNA followed by paired-end sequencing. This will enable us to detect a large majority of point mutations, indels, amplifications/deletions and gene fusions from RNA and DNA isolated from FFPE tissues. We will use the remaining cases in this study for validation of informative mutations we discover.

Innovation: Very little is known about the etiologies of basal-like and luminal B tumors. No prior study has been specifically designed to assess risk factors or genomic signatures for these poor prognosis tumors, to develop subtype specific risk prediction models analogous to the Gail Model, or to assess factors influencing prognosis. Furthermore, we have spent considerable effort in optimizing innovative approaches to next generation sequencing of RNA and DNA from FFPE tissues which has been a major challenge in the field.

Impact: Determining risk and prognostic factors specific to the major molecular subtypes of breast cancer could have broad impact in several respects: 1. Identifying modifiable risk factors for these tumors affords opportunities for prevention; 2. Developing subtype specific risk prediction models and identifying subtype specific factors influencing prognosis could impact clinical care and decision-making for both women in the general population and breast cancer patients, respectively; and 3. Determining etiologic pathways relevant to these molecular subtypes will help inform the development of novel prevention and treatment strategies.