We propose that collagen features, such as alignment, fiber straightness and width, and other features of extracellular matrix composition can be used as an early biomarker of breast cancer. It is the goal of this proposal to determine the range of collagen structural features that characterize normal, benign and early disease, to determine how these features relate to mammographic density, and to identify the features that best predict risk, recurrence, and progression. This proposal leverages a diverse collaborative team that includes experts in imaging and understanding the tumor microenvironment, breast surgery, experts in population health and biostatistics, and an expert in proteomic analysis of the ECM. This team will analyze a set of patient cohorts to develop imaging and analysis of collagen structure and ECM composition as a potential early biomarker for breast cancer.
Aims : 1) Characterize collagen features that define heterogeneity of the normal state. 2) Characterize collagen features in pre- and early breast cancer lesions. 3) Characterize the proteomics of biopsies representing benign, at risk, and invasive breast cancer. Significance: This project meets the goals of the RFA to develop and validate combined imaging and biomarker approaches to improve early cancer detection and the diagnosis of early-stage cancers. Our findings have the potential to reduce overdiagnosis and false positives and discern lethal from non-lethal disease.

Public Health Relevance

The widespread adoption of screening mammography throughout the 1980s and 1990s led to a reduction in breast cancer mortality as more women were diagnosed at an earlier stage of disease and could be more successfully treated. Apart from their use in visualizing the tumor, mammograms are quantifiable based on the density of the signal in the image, which is based on the absorbance of x-rays by cells and the filamentous protein collagen. Collagen is the most abundant protein in the body and in the breast is present as a scaffolding network that supports the milk-secreting epithelial cells of this gland. In non-diseased breasts, the mammographic density is a function of the amount of collagen present, and is highly correlated to a 4-6 fold increased risk of developing breast cancer in the future. Mammographic density has been deemed so important that over a dozen states have recently passed legislation requiring physicians to inform patients of this value. However, the problem is that mammographic density does not predict survival once a tumor has been detected. Thus, although we have improved our detection of breast cancers, we do not have the ability to sort patients into groups where disease progresses from those that do not. Our current paradigm of treating everyone aggressively needs to change to a targeted plan of treating only those women at risk of future life-threatening disease. It is our hypothesis that it is not the overall macroscpic amount of collagen present, but rather the structure and organization of collagen at the local level which predicts cancer progression. We will examine tissue samples from a variety of non-diseased human patient cohorts, with the goal of discovering a biomarker for disease progression based on collagen structure. This will be done non-invasively by imaging slides that are already routinely prepared using a technique called second harmonic generation (SHG) microscopy which selectively images collagen. We have advanced software which will measure the length, width, straightness, density and orientation of fibers that we can use to analyze tens of thousands of collagen fibers from dozens of images per sample. This high level of quantitation improves our statistical ability to discover the underlying biomarker from the noise. By understanding the normal non-diseased state, we can apply this knowledge to very early cancer stages in order to sort patients into aggressive and non-aggressive therapy groups. As a compliment to these studies, we will also determine the protein composition of the collagen matrix, as there are dozens of other proteins present. Preliminary data has shown changes in the makeup of proteins found in normal tissue compared to breast cancer, implying that the change in collagen structure is related to the underlying expression of proteins there. Through our proposed experiments, we will establish what a truly 'normal' collagen matrix looks like in terms of structure and composition.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Research Project (R01)
Project #
1R01CA199996-01
Application #
8985889
Study Section
Special Emphasis Panel (ZRG1-SBIB-F (59))
Program Officer
Mazurchuk, Richard V
Project Start
2015-09-01
Project End
2020-08-31
Budget Start
2015-09-01
Budget End
2016-08-31
Support Year
1
Fiscal Year
2015
Total Cost
$505,303
Indirect Cost
$130,419
Name
University of Wisconsin Madison
Department
Anatomy/Cell Biology
Type
Schools of Medicine
DUNS #
161202122
City
Madison
State
WI
Country
United States
Zip Code
53715
Tomko, Lucas A; Hill, Ryan C; Barrett, Alexander et al. (2018) Targeted matrisome analysis identifies thrombospondin-2 and tenascin-C in aligned collagen stroma from invasive breast carcinoma. Sci Rep 8:12941
Conklin, Matthew W; Gangnon, Ronald E; Sprague, Brian L et al. (2018) Collagen Alignment as a Predictor of Recurrence after Ductal Carcinoma In Situ. Cancer Epidemiol Biomarkers Prev 27:138-145