Quantitative characterization of tumor heterogeneity using habitat imaging for the prediction of patient outcome in triple negative breast cancer

PROJECT SUMMARY The overall goal of this training proposal is to employ quantitative imaging to noninvasively characterize tumor microenvironmental heterogeneity in triple negative breast cancer (TNBC) for the prediction of treatment response and outcome. TNBC is an aggressive breast cancer subtype with notable diversity in disease biology

and clinical presentation. Recently-approved immunotherapies introduce exciting new avenues for neoadjuvant treatment of TNBC; however, response to immunotherapy is variable and treatment can entail significant adverse effects and financial cost. Variable response to therapy can be attributed in-part to heterogeneity of the tumor

microenvironment, affecting therapeutic delivery and efficacy. Through an innovative approach known as “habitat imaging”, multiparametric magnetic resonance imaging (mpMRI) can be used to spatially resolve local microenvironments within a lesion into distinct tumor subregions, or habitats. For the research component of this

proposal, we propose to use quantitative breast imaging and informatics techniques to identify tumor habitats and whole-lesion habitat signatures for TNBC patient stratification. We hypothesize that imaging-derived habitat signatures identified prior to treatment can aid in stratifying TNBC patients with increased probability of achieving

pCR and/or decreased risk of recurrence. We will test this hypothesis through the following specific aims: Aim 1 (K99) To retrospectively identify tumor habitat signatures from pretreatment mpMRI to stratify TNBC patients and predict treatment outcome; Aim 2 (R00) To prospectively employ habitat imaging using hybrid positron

emission tomography (PET)/MRI for improved TNBC patient stratification. Successful completion of my research aims will provide a clinically-translatable methodology for improved understanding of an individual’s lesion physiology that could guide personalized treatment strategies for optimal patient outcome. To provide me with

the necessary training to successfully carry out these research aims, this proposal outlines a mentored-training plan with three areas of focus: 1) strengthen my expertise in clinical breast cancer research, 2) receive additional training in computational pathology, and 3) obtain educational training and hands-on experience with PET/MRI

to prepare for research in the independent R00 phase and beyond. This training program will be executed under the direct mentorship of NIH-funded researchers and clinicians specializing in medical oncology, nuclear medicine, pathology and radiology, and take place within the well-equipped and established cancer research

environment at the University of Washington and Fred Hutch Cancer Center. As outlined in my career development plan, funding from this proposal will be used to dedicate time for educational workshops and training seminars, along with regular meetings with my mentorship team. Together, this training and research proposal

will ensure that I am well-prepared to achieve my career goal to become an independent investigator focused on the translation of innovative imaging techniques to clinical practice to further personalize breast cancer care.