• Home
  • News
  • Calendar
  • About DF/HCC
  • Membership
  • Visitor Center

Kornelia Polyak, MD, PhD

Professor, Department of Medicine, Harvard Medical School

Professor of Medicine, Medical Oncology, Dana-Farber Cancer Institute

Contact Info

Kornelia Polyak
Dana-Farber Cancer Institute
450 Brookline Avenue
Boston, MA, 02215
Mailstop: Dana 740C
Phone: 617-632-2106
Fax: 617-582-8490
Kornelia_Polyak@dfci.harvard.edu

Assistant

Claudia Steele
Dana-Farber Cancer Institute
Phone: 617-582-7646
claudia_steele@dfci.harvard.edu
Claudia Steele
Dana-Farber Cancer Institute
Phone: 617-582-7646
claudia_steele@dfci.harvard.edu

DF/HCC Program Affiliation

Breast Cancer
Cancer Genetics

Lab Website

Polyak Lab

Research Abstract

Our goal is to identify differences between normal and cancerous breast tissue, determine their consequences, and use this information to improve the clinical management of breast cancer patients. The three main areas of our interests are: (1) how to accurately predict breast cancer risk and prevent breast cancer initiation or progression from in situ to invasive disease, (2) better understand drivers of tumor evolution with special emphasis on metastatic progression and therapeutic resistance, and (3) novel therapeutic targets in breast cancer with particular focus on “bad” cancers such as triple negative breast cancer and inflammatory breast cancer. All of our studies start with analyzing samples from breast cancer patients (or normal healthy women for the risk studies), formulate hypotheses based on our observations, use experimental models to test these, and then translate back our findings into clinical care.
Highlights from our cancer heterogeneity studies: With rare exceptions tumors are thought to originate from a single cell. Yet, at the time of diagnosis the majority of human tumors display startling heterogeneity in many structural and physiological features, such as cell size, shape, metastatic proclivity, and sensitivity to therapy. This intratumor diversity complicates the study and treatment of cancer because small tumor samples may not be representative of the whole tumor and because a treatment that targets one tumor cell population, may not affect another, leading to a poor clinical response. On the positive side, intratumor diversity is a type of “looking glass” for a particular cancer from which we can both learn its past and predict its future.
We have developed a model of intratumor clonal heterogeneity in breast cancer and utilized this to assess the functional relevance of clonal interactions in metastatic progression. We found that polyclonal tumors were commonly metastatic, even though none of the individual clones present in them showed this behavior in monoclonal tumors. We have also analyzed breast tumor samples before and after pre-operative chemotherapy, or at different stages of disease progression (i.e., primary and metastatic lesions) for the degree of intratumor genetic and phenotypic heterogeneity at the single cell level. We found that tumors with the lowest pretreatment genetic diversity responded the best to treatment and that distant metastatic lesions had higher genetic diversity compared to primary tumors and lymph node metastases. Lastly, we have developed mathematical models based on these experimental data that can infer the evolution of tumors during treatment. Based on these preliminary data, we hypothesize that intratumor heterogeneity per se is a driver of metastatic spread and therapeutic resistance. Thus, measures of intratumor heterogeneity can be used to predict the risk of metastasis and to personalize therapy based on this. At the same time, understanding of how heterogeneity within tumors promotes disease progression may reveal new therapeutic targets and would allow us to design more effective and individualized treatment strategies.
Highlights from our breast cancer risk and prevention study: The highest impact on breast cancer-associated morbidity and mortality will be achieved with two tools. The first tool is a test that accurately predicts an individual’s risk of being developing breast cancer. This will allow us to identify who needs the preventive action and who does not. Second, is to discover the best agent for prevention that will be universally effective. We know that inheriting mutated BRCA1 and BRCA2 genes confer a high risk of breast cancer, and the most effective prevention strategy currently available is prophylactic oophorectomy and mastectomy. Other significant determinants of breast cancer risk are reproductive history and mammographic density. Epidemiological data suggest that pregnancy induces long-lasting effects in the normal breast, except in BRCA1 and BRCA2 mutation carriers, where pregnancy does not decrease breast cancer risk.
What cells must in the breast need to be eliminated to reduce risk? A number of studies have shown that breast epithelial progenitor cells are the likely the “cell-of-origin” of breast cancer. It stands to reason then, that eliminating them will abolish tumor development. In recent work we analyzed and characterized multiple cell types from normal breast tissues of nulliparous and parous women, including BRCA1 and BRCA2 mutation carriers. We detected the most significant differences in breast epithelial progenitors and found that the frequency of these cells is higher in women with higher risk of breast cancer. We have also identified key signaling pathways important for their proliferation and showed that by modulating the activity of these pathways we can decrease the frequency of the progenitor cells, thus, potentially reducing breast cancer risk. We propose that the progenitor markers we identified can be used for breast cancer risk prediction and that depleting these progenitors will decrease the risk of breast cancer. We are pursuing these studies in large cohorts in women and in rodent models of breast cancer (prevention) with immediate plans to translate our findings to high risk women as the drugs used to deplete these progenitors are already in clinical trials for cancer treatment.

Publications

View All Publications