Work in my lab is focused on identifying and understanding the molecular events associated with the initiation and progression of human cancer. The clinical significance of these events in malignant and pre-malignant human breast tumors is of primary interest. Our involvement in collaborative research provides a unique opportunity for translation of basic science discoveries to questions that can be posed in large, clinical and population-based studies.

The current model of carcinogenesis is that of a multi-step accumulation of genetic changes within cells that supplant normal controls on cell division and lead to increased cell proliferation. Abnormalities of the cell cycle in a highly proliferative cell population may then lead to additional molecular alterations. In fact, derangements in the cell cycle may not only contribute to uncontrolled cell growth but may be causal factors in the development of cancer. In collaboration with epidemiologists and basic scientists, we were one of the first groups to identify the loss of cell cycle inhibitor p27kip1 as an important indicator of poor prognosis in breast cancer. Follow up in a phase III Southwest Oncology Group (SWOG) clinical trial showed that, in women with steroid receptor positive tumors, but not in women with steroid receptor negative tumors, loss of nuclear p27 protein expression is a marker of poor prognosis for women treated with Adriamycin. Recent data evaluating mislocalizatin of p27 to the cytoplasm, we found that cytoplasmic p27 contributed to lapatinib resistance in Her2+ breast cancer cells by suppressing apoptosis. Our results suggest that p27 localization may be useful as a predictive biomarker of therapeutic response in patients with Her2+ breast cancer.

Steroid receptor status if one of the main differentiating characteristics of breast cancer and lack of ER, PR, and HER2 expression in tumor cells (‘triple negative’ breast cancer (TNBC)) is associated with a specific constellation of risk factors such as inherited BRCA1 mutations, high rates in African American women, aggressive tumor characteristics, poor outcome, and lack of response to hormonal treatment. In breast cancers of young AA and Caucasian women in a population-based case-control Atlanta study, we found high frequency gene copy number alterations (CNAs) in genomic locations that were subtype-specific, and of the highest frequency in TNBC compared to the other subtypes of breast tumors, and at specific genomic sites in tumors of AA women. We are using this data to guide the evaluation of gene copy number, methylation and LOH alterations of specific genes in the genomic regions identified.

Although others and we have successfully identified molecular markers for tumorigenesis using conventional approaches, single gene and protein changes are also unlikely to reflect the complexity of the molecular changes present in tumor cell populations. Using global approaches, we are now in a position to elucidate the molecular components, and the connections between the components, that coincide with the acquisition of malignant traits. We are conducting discovery projects in diverse populations including women exposed to radiation from the Chernobyl nuclear accident, and Latina women from the U.S. and Latin America evaluating global tumor mutations, gene copy number changes, gene expression and DNA methylation, as well as inherited genetic contributions in studies of precursor and malignant lesions that will help to better understand, and respond appropriately to, the clinical heterogeneity of breast cancer.