Bridging the Gap Between Species: Translating a Fracture Callus Mechanical Properties Model From Ovine to Human Use

New approach methodologies (NAMs) can reduce reliance on animal testing and enable outcomes assessments in humans that were previously possible only in animal studies. Translating NAMs from animals to clinical use requires consideration of differences between the preclinical and clinical settings. The objective of this study was to translate and assess the performance of virtual mechanical testing of tibial fracture healing from a large animal model to clinical use. We translated a dual-zone material model for soft and hard callus, which we previously validated in sheep, to clinical use. Image-based models, also known as digital twins, were generated from computed tomography (CT) scans of healing human tibiae at 12 weeks post-op. Scaling adjustments were applied to correct for scanner-specific variability in X-ray attenuation values. The threshold for differentiation between soft and hard callus was inferred from sheep using comparative densiometric analysis. The selected hard/soft callus cutoff value was 998 Hounsfield units (HU), corresponding to 0.5372 of the expected cortical bone density mode of 1858 HU. The human-scaled dual-zone model reduced virtual torsional rigidity (VTR) by 41% compared to a single-zone material model developed based on cortical bone mechanics. With the dual-zone model, half the cohort achieved torsional rigidities in the range of intact tibiae, which corresponded well with modified radiographic union score for tibial fractures (mRUST) scores showing that half the patients achieved union (mRUST >= 11) at this timepoint. These results demonstrate the potential for translation of a validated preclinical virtual mechanical test to clinical use.