Ethically advancing bone biomechanics in preclinical research: A micro-CT-based finite element framework to assess mouse-specific tumor-induced and age-related bone deterioration

Pathophysiological processes, such as aging and tumor colonization, disrupt bone architecture and reduce load-bearing capacity, increasing fracture risk. In preclinical research, bone mechanical properties are traditionally assessed destructively, precluding longitudinal within-subject assessment and requiring large animal cohorts. Recent advances in imaging and computational modeling now enable longitudinal, specimen-specific, non-destructive simulations of bone loading, reducing cost, time, and animal use; but their application to metastases and aging remains limited. We developed a micro-computed tomography-based finite element model for specimen-specific virtual mechanical testing of long bones from three mouse strains (C57BL/6, BALB/c, and nu/nu nude strains) that varied in age and were colonized by osteolytic tumors. Model calibration used a healthy dataset; validation used tumor-bearing and age-varied cohorts to quantify effects of tumor-induced bone loss and senescence on skeletal integrity. We incorporated a post-yield (elastic-plastic) constitutive material law and propagated simulation and experimental uncertainties. Across cohorts, the model reproduced sample-specific force-displacement curves with accuracy above 80% and, stiffness and yield load with R 2 =0.985 and R 2 =0.646, respectively, while also providing predictions of stress-strain distributions. Three-point bending tests were sensitive to age-related changes at the midshaft, but not to metastatic lesions located proximally away from the midshaft. Axial compression tests were therefore simulated as alternative strategy to investigate metastasis-induced weakening, revealing a significant decline in mechanical properties over time (p< 0.05). Overall, this platform provides a non-invasive, data-driven tool to link bone geometry with mechanical competence, reducing animal use, and supporting longitudinal monitoring and virtual “what-if” analyses in murine bone disease models.