Stratifying DCIS Biopsies for Risk of Future Tumor Formation

Institution: University of California, San Francisco
Investigator(s): Thea Tlsty, Ph.D. -
Award Cycle: 2008 (Cycle 14) Grant #: 14OB-0165 Award: $750,000
Award Type: Translational Research Award
Research Priorities
Detection, Prognosis and Treatment>Imaging, Biomarkers, and Molecular Pathology: improving detection and diagnosis

Initial Award Abstract (2008)

While it is estimated that ~62,000 women were diagnosed with pre-cancers or pre-malignant lesions such as ductal carcinoma in situ (DCIS) in 2007, only 15% of these women will subsequently develop invasive breast cancer with potential to die of the disease. However, the vast majority of women diagnosed with DCIS undergo the same regimen of full breast removal (mastectomy), partial mastectomy or partial mastectomy followed by radio-, chemo-therapy and/or hormone drugs. Thus, many thousands of women are treated unnecessarily with additional radiation, hormonal therapy, and/or mastectomy in order to prevent the progression of an invasive cancer that will not, in fact, occur. This causes unnecessary suffering and side effects in a majority of otherwise-healthy women and places unnecessary demands on healthcare systems. Additionally, ~10% of DCIS patients do not receive adequate intervention, since they develop tumors even when the DCIS is removed.

Our goal is the development of a rapid, inexpensive prognostic clinical test that will provide individual risk information for all women diagnosed with DCIS.

The general methodology: Recently, we successfully found several biomarkers that, when expressed in DCIS, accurately predict formation of aggressive invasive tumors (called basal-like breast cancers) years before they form, as well as markers that identify women who will not develop future breast cancers after a pre-cancer diagnosis. Our general methodology was to use novel biological insights from basic studies of how cancer forms and grows to provide us with clues to identify markers that predict future invasive cancers. After the candidate biomarkers are hypothesized, they are tested in biopsy samples that were collected years ago in women that were diagnosed with DCIS and elected to have a subsequent lumpectomy. Standard practice at that time did not propose radiation or hormonal drugs. These women have been contacted through the years to monitor their health and determine if subsequent breast cancer developed. Since it is known which biopsies came from women who did or did not develop a subsequent invasive breast cancer, we can determine if our candidate markers are useful by testing them on the biopsies and evaluating their ability to accurately identify which women developed breast cancer.

For over four decades clinical investigators have tried to identify markers that would predict which women diagnosed with DCIS would be among the ~15% to develop a future invasive tumor. They looked for markers in the size, shape and pattern of the DCIS with no success. In the more recent molecular era, they have evaluated several molecular markers, again with no success. Our biological approach, using a unique and well-studied tissue culture model system, provided the first clues to this important clinical problem. Our analysis of how these human breast cells acquire cancer properties has provided insights for the prediction of future tumor formation. We propose to use this innovative model system to identify additional candidate markers that will provide information about the remaining fraction of biopsies that do not express the markers we have already identified. Validation of these additional markers would allow us to develop the first test that would tell a woman diagnosed with DCIS about her risk of future breast cancer.

Final Report (2011)

While only a minor fraction of premalignant lesions are associated with subsequent tumor events, most patients so diagnosed receive the same options for treatment. To date, studies have failed to identify characteristics of women who have been diagnosed with ductal carcinoma in situ (DCIS) and have a high or low risk of subsequent invasive cancer. The overall goal of this project is to identify which women diagnosed with DCIS will develop invasive cancers within the subsequent ten years following diagnosis and which will not. This would provide information that would allow clinicians to avoid under treatment of women who will develop aggressive tumors and over treatment of women that will have no subsequent tumor event.

Using information obtained from studies of human mammary epithelial cells, we investigated if biological clues would provide risk markers for women diagnosed with DCIS. We conducted a study in a cohort of 1162 women diagnosed with DCIS and treated by lumpectomy alone. Besides collecting clinical characteristics and information on subsequent tumors, we conducted standardized pathology reviews and immunohistochemical staining on the initial paraffin-embedded DCIS tissue for the estrogen receptor (ER), the progesterone receptor (PR), the proliferation marker Ki67, the tumor suppressor p53, the cell cycle inhibitor p 16, the epidermal growth factor receptor-2 (ERBB2, HER2/neu oncoprotein), and cyclooxygenase-2 (COX-2), a pre-malignant marker previously identified by us. Competing risk models were used to determine factors associated with risk of subsequent invasive cancer versus DCIS, and cumulative incidence survival functions were used to estimate 8-year risk.

Our study, described in detail in a recent issue of JNCI (Kerlikowske et al, 2010), shows that factors associated with subsequent invasive cancer differed from those associated with subsequent DCIS. In a multivariable model, DCIS lesions that were either detected by palpation or “triple-positive” for p16, COX-2 and Ki67, were statistically significantly associated with an increased eight year risk of subsequent invasive cancer compared to DCIS lesions that were detected by mammography and triple negative for p16, COX-2, and Ki67. Of particular note, nuclear grade was not statistically significantly associated with risk. Moreover, eight-year risk of subsequent DCIS was highest for women with DCIS lesions that had disease-free margins of 1 mm or greater combined with either ER-ERBB2+Ki67+ or pI6+COX-2-Ki67+ status.

We have therefore reached a stepping stone towards our goal to develop a rapid and inexpensive prognostic clinical test that will provide individual risk information for all women diagnosed with DCIS by showing that biomarkers can indeed identify which women who were initially diagnosed with DCIS are at high or low risk of subsequent invasive cancer, whereas histopathology information cannot. We are currently testing new DCIS biomarkers, which will be integrated with pertinent biological information into our current prognostic signature so we can further extend and complete our new found ability to predict future tumor formation. Such an optimized test would provide women and their clinicians with information to make informed choices for therapy after a diagnosis of a pre-malignancy. Another potential outcome may be identification of novel targets for early intervention.

Biomarker expression and risk of subsequent tumors after initial ductal carcinoma in situ diagnosis.
Periodical:Journal of the National Cancer Institute
Index Medicus: J Natl Cancer Inst
Authors: Kerlikowske K, Molinaro AM, Gauthier ML, and Tlsty, TD
Yr: 2010 Vol: 102 Nbr: 9 Abs: Pg:627-37

Premalignant breast neoplasia: a paradigm of interlesional and intralesional molecular heterogeneity and its biological and clinical ramifications.
Periodical:Cancer Prevention Research
Index Medicus: Cancer Prev Res
Authors: Berman HK, Gauthier ML, Tlsty TD
Yr: 2010 Vol: 3 Nbr: 5 Abs: Pg:579-87

DNA damage drives an activin a-dependent induction of cyclooxygenase-2 in premalignant cells and lesions.
Periodical:Cancer Prevention Research
Index Medicus: Cancer Prev Res
Authors: Fordyce C, Fessenden T, Pickering C, Jung J, Singla V, Berman H, Tlsty T
Yr: 2010 Vol: 3 Nbr: 2 Abs: Pg:190-201

Human mammary cancer progression model recapitulates methylation events associated with breast premalignancy.
Periodical:Breast Cancer Research
Index Medicus: Breast Cancer Res
Authors: Dumont N, Crawford YG, Sigaroudinia M, Nagrani SS, Wilson MB, Buehring GC and Tlsty TD
Yr: 2009 Vol: 11 Nbr: 6 Abs: Pg:R87