Abstract Introduction Premenopausal women with intermediate-high risk HR+ breast cancer often receive near-complete estrogen deprivation with ovarian function suppression concurrent with an AI. Hypoestrogenemia is associated with cardiotoxicity, but the cardiovascular impact of this newer breast cancer treatment is largely unknown. With increases in survival rates and younger women being diagnosed, methods to monitor and predict cardiac damage due to OFS+AI therapy are needed. Left ventricle (LV) ejection fraction is currently used for monitoring cancer treatment-related cardiovascular degradation, and can detect major heart defects, but is insensitive to subclinical left ventricle function. Emerging methods include T1 mapping and estimation of myocardial strain to indicate fibrosis. We propose an extension of these methods by estimating cardiac tissue elasticity using LV wall deformation to drive a biomechanical model. Elasticity is a more functional and direct measurement of tissue response based on structural mechanics driven by patient-specific cardiac magnetic resonance imaging (CMR) data. Elasticity measurements may serve as a predictive biomarker of early AI-induced cardiac changes.MethodsThis study is a retrospective initial proof-of-concept correlative imaging study to an existing clinical study for the use of CMR to detect cardiovascular damage (ESPRIT). Two cohorts of premenopausal breast cancer patients either: (1) undergoing OFS+AI for HR+ breast cancer or (2) triple negative breast cancer (TNBC) patients that have already received chemotherapy, were imaged twice, 3-6 months apart using CINE CMR. TNBC patients serve as the control, with no expectation of further cancer treatment-related cardiac damage. Time steps during passive ventricular diastole were visually selected from CINE CMRs. Each slice was non-rigidly registered to estimate LV deformation during passive filling. Deformation was simulated on a finite element mesh of the LV based on linear elastic transverse isotropic mechanical equilibrium. Using an inverse problem formulation, simulated deformation was compared to model-calculated deformation to estimate the spatial tissue longitudinal and transverse elasticity. Elasticity maps of the LV at initial and final points are compared to determine regional stiffening of the LV wall, to be used as early biomarkers for LV fibrosis. ResultsIn this initial investigation, elasticity maps were analyzed for four patients (n=2 from each cohort). Passive LV tissue stiffening was observed in each AI patient, with 100% and 25% relative increases observed for longitudinal elasticity and 50% increases for transverse elasticity in the basal inferior region, and mid anterior region of the LV in each patient, respectively. No increases in stiffness of the LV were observed for TNBC patients. Ejection fraction remained consistent for all patients. ConclusionIn this proof-of-concept study, we demonstrate that elasticity maps indicate local stiffening of the LV using a biomechanical model-based elasticity imaging method that could be used to indicate cardiac dysfunction in breast cancer patients receiving AIs. Spatial elasticity mapping allows direct observation of structural mechanics to reveal specific areas of LV stiffening. Moving beyond traditional strain imaging, our method yields a functional measure of tissue stiffness to directly indicate cardiac fibrosis. This study demonstrates the use of biomechanical models to interpret CMR and provides potential for use of more advanced constitutive models. Our non-invasive biomechanical model-based elasticity imaging method shows significant promise to indicate early cardiac function deterioration, critical for premenopausal women undergoing extended cancer therapies. Citation Format: Caroline Elizabeth Miller, Jennifer Jordan, Alexandra Thomas, Jared Weis. Predicting cardiac dysfunction in breast cancer patients undergoing aromatase inhibitor treatment using biomechanical model-based elasticity imaging [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS3-21.