OSTEO: Engineering a predictive digital twin of bone metastasis

Bone metastases (BMs) occur when cancer spreads to bone and are a life-threatening problem for about 40% of patients with renal cancer and 80% of patients with prostate cancer. BMs cause pain, weak bones, and fractures, and often require major surgery. Understanding how cancers affect bone health and fractures is challenging. While animal studies can help, they raise ethical concerns and are limited for quickly and efficiently addressing complex problems. Using advanced “digital twin” (DT) models of BM, scientists can learn more about how cancer weakens bones. This project will build a computational model - Osseous System Twin for Experimental Optimization (OSTEO) - that will bring together 3D images of bone, experimental cancer data, and bone-strength measurements on one platform. OSTEO will help better understand how BMs grow, how bones weaken, and when they are at risk for fracture. OSTEO will first be tested using data from experiments in mice to ensure accuracy. The model will then run simulations to help decide which lab studies can be most effective. OSTEO will be freely available with a point-and-click app for researchers, doctors, students, and teachers. Videos, handouts, and example cases will be used to explain how the model works. Workshops and hands-on demos for students and the public will show how fields like engineering, cancer research, and advanced machine learning (ML) models can work together to improve research of complex diseases. OSTEO will be a modular, image-driven, 3D DT of BM that couples an in vivo-inspired agent-based model of tumor–bone–vascular dynamics. A finite element method (FEM)-based mechanical module will be used to estimate bone weakening and fracture susceptibility. Biological simulations will include tumor cells, osteoclasts (OCs), osteoblasts (OBs), vasculature and bone matrices within micro-computed tomography (μCT)-derived geometries. The platform will be initially calibrated using osteolytic prostate and renal cancer models supported by local imaging, histology, therapy-response, and mechanical datasets. The mechanical module will convert evolving μCT-derived architecture into specimen-specific FEM models using calibrated density-modulus relationships and post-yield constitutive laws. Calibration and validation will combine longitudinal μCT, intravital imaging, therapy-response data, and ex vivo mechanical testing. Sensitivity analysis and uncertainty quantification will identify drivers of tumor growth, osteolysis, OC/OB imbalance, vascular remodeling, and stiffness or strength loss, while propagating uncertainty into predictive intervals. OSTEO will help support preclinical virtual experiments across tumor type, bone site, sex, age, and therapeutic scenarios, including cabozantinib and zoledronic acid. Outcomes include a validated, transparent, user-refinable DT that advances mechanistic understanding of tumor-bone ecosystems, reduces exploratory animal use by in silico triage, and provides fracture-susceptibility surrogates to advance health outcomes. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.