Next Generation MRI for Virtual Mechanical Testing of Bone Fracture Healing

Bone fractures are common injuries that impact an estimated 9.6 million Americans each year, resulting in 1.6 million years of living with disability. There is currently a lack of reliable biomarkers that can detect problems with bone healing early, which represents a major barrier to improved care. This proposal describes a translational validation study of next generation methods for fracture diagnostics. In the study, we will use ultrashort echo time magnetic resonance (UTE-MRI) sequences and finite element (FE) based virtual mechanical testing to quantify and predict bone healing. Tibial fractures will be employed as a test bed for more global applications. Our central hypothesis is that the structural quality of bone healing can be measured using MR-based virtual mechanical testing. The study will consist of 3 Aims. In Aim 1 we will establish MR assessment of callus mechanical properties. We will use an ovine tibial osteotomy model, fixed by intramedullary (IM) nailing. UTE-MRI will be used to construct FE models and in situ inverse optimization will define the mechanical properties of callus by matching the virtual mechanical tests to the postmortem physical biomechanical tests. We hypothesize that UTE-MRI measures of bone density will be negatively correlated with callus elastic modulus, and that image-based modeling from UTE-MRI can reliably replicate bench tests of healing long bones. In Aim 2, we will predict late fracture healing using early MR scanning in the same ovine model. Here, we will use MR to measure the earliest stages of repair and predict the later formation of mineralized callus in normal and delayed healing animals. We hypothesize that the sigmoid functional recovery curve of mechanical rigidity in delayed-healing fractures is significantly different from the recovery curve that characterizes normal healers. In Aim 3 we will deploy MR-based virtual mechanical testing to measure fracture healing in a clinical setting. In this observational pilot study, we will recruit 50 patients with tibial shaft fractures treated by IM nailing. Results from 6-week UTE-based virtual mechanical testing will be correlated with 12- week CT and MR measures and time to clinical union. Receiver-operator characteristic analysis will be used to identify the diagnostic cutpoint for early image-based virtual mechanical testing that best discriminates between normal healers and patients who experience delayed union (>20 weeks). We hypothesize that MR scanning can replicate the results of CT-based virtual mechanical testing for assessment of fracture healing in a clinical setting and that soft callus formed by 6 weeks predicts clinical outcomes (binary classification: union or delay). This project will demonstrate how objective measurements of bone healing using radiation-free MR imaging will fundamentally shift fracture research and clinical care paradigms by enabling an early and proactive approach to the management of high-risk fracture patients. Our findings will lead to future applications focused on femoral shaft fractures, fragility fractures, pediatric fractures, and limb salvage.