Using Stromal Markers to Improve Risk Stratification of Ductal Carcinoma in Situ

Investigator: Thea D. Tlsty, PhD
Sponsor: NIH National Cancer Institute

Location(s): United States


The widespread adoption of screening mammography has led to an increase in the diagnosis of ductal carcinoma in situ (DCIS) of the breast. While it is estimated that 55,000 women in the US will be diagnosed with DCIS in 2013, only a small fraction of these women (~15%) will subsequently develop invasive breast cancer. However, most women with DCIS are treated as if they will develop invasive cancer. Almost a third of these women opt for a full mastectomy. In a majority of remaining women, the DCIS lesion is surgically excised with lumpectomy and, in over half of these women, additional radiation or tamoxifen treatment is provided. Thus, many women are receiving unnecessary adjuvant therapy to prevent invasive cancers that will not occur. Additionally, ~ 15% of women are not receiving adequate intervention because they will subsequently develop a subsequent invasive tumor even after receiving lumpectomy and adjuvant therapy. Identifying molecular markers that can accurately predict subsequent invasive cancer could aid in stratifying an individual's risk for subsequent tumors and their need for adjuvant therapy. Using a large group of women diagnosed with DCIS and treated by lumpectomy alone and with long-term follow-up, we identified signatures (e.g. p16/COX-2/KI67-positive) within epithelial cells of DCIS lesions that, when combined with clinical factors, can identify women at high and low risk for subsequent invasive cancers. These markers can identify 63% of women that will develop invasive breast cancer subsequent to a diagnosis of DCIS. Following this work, we have identified interactions between epithelial cells and their surrounding stromal cells that influence
subsequent tumor formation. We found a novel activin A-dependent DNA damage response pathway that originates in epithelial cells and that elicits cancer-promoting behaviors in neighboring stromal cells, in particular fibroblasts and potentially in immune cells. We also characterized a pro-tumorigenic multicellular stromal program in both cancerous and pre-cancerous breast tissues at high risk of progressing to cancer. This program, characterized by repression of the cell surface protein CD36, constitutes a cornerstone of our proposal. Using biopsy material from the group of women diagnosed with DCIS described above, we will evaluate the potential predictive value of the novel stromal markers identified by our recent findings as well as stromal immune biomarkers and stromal biomarkers and histological features recently reported to predict ipsilateral recurrence. Our goal is to identify every woman that will develop an invasive cancer subsequent to a diagnosis of DCIS so that they can take appropriate action.