Project Summary The main goal of our project is to investigate the added value of multi-modality breast imaging in prognostic assessment for breast cancer. Accurate prognostic assessment is a key component of personalized treatment. Breast cancer prognosis has historically been determined based on tumor histopathology (i.e., size, grade, stage, etc) and immunohistochemistry (i.e., estrogen, progesterone, human epidermal growth factor receptors). Recently, molecular assays have also become available (i.e., OncotypeDX, MammaPrint, etc) that measure tumor gene expression as related to prognosis. Although a lot of progress has been made, there is still a need for substantial improvements in identifying women who are at risk for morbidity due to overly or insufficiently aggressive therapy. Currently, histopathology and the molecular characteristics of tumors are mainly analyzed based on selective tissue sampling. As it is increasingly recognized that intra-tumoral heterogeneity plays an important role in tumor progression and resistance to treatment, selective tissue sampling may be inadequate for fully capturing such heterogeneity, potentially resulting in incomplete information for guiding treatment. Imaging is increasingly used in routine care for screening, diagnosis, and treatment of breast cancer, with different modalities offering complementary information. This ability, coupled by a potential for high-resolution 3D visualization, has provided a new means for capturing vital aspects of tumor heterogeneity in-vivo, and therefore potentially complementary prognostic information. The overarching goal of our study is to address this fundamental question: Can tumor imaging phenotypes provide additional information to established histopathologic and emerging molecular markers for predicting breast cancer recurrence? We propose four aims: AIM1) Develop a multi-modality imaging phenotype vector that captures structural (e.g., shape, morphology, texture) and functional heterogeneity (e.g., contrast uptake) of primary tumors. AIM2) Determine the prognostic value of the imaging features in predicting breast cancer recurrence; predictive value of features will also be explored. AIM3) Develop an augmented recurrence risk assessment model that incorporates tumor imaging features with standard tumor histopathology and emerging molecular markers, and AIM4) Perform independent validation of our model with prospectively collected data. In our study, we will investigate the prognostic value of multi-modality imaging for women diagnosed with primary invasive breast cancer. We will utilize a cohort of women with imaging and tumor tissue biomarker data from an NIH trial completed at our institution, from which 10-year follow-up from initial diagnosis and treatment is currently available. Ultimately, by integrating imaging with tumor histopathology and molecular markers in an augmented recurrence risk assessment tool we will be able to help better guide treatment dec...