The exploration of musculoskeletal systems is limited by capture of only a single or few measurements to describe complex conditions. Observing mechanostructural interactions in arthritis, and developing specific and sensitive measures of change would generate new advances in understanding human health. Micro-computed tomography, microCT, provides 3D image datasets that can be employed to access 3D metrics that define OA in clincial and preclinical models. Despite the wealth of information available in these datasets, it is rarely exploited. In this work, a methodology is demonstrated in a preclinical rabbit model and clinical data.

Eight New Zealand white rabbits underwent anterior cruciate ligament (ACL) desmotomy on one knee, and the non-operated contralateral joint served as a control, and sacrificed 8 weeks post-op. Humeral head samples were harvested from 21 patients undergoing replacement for end stage OA. Consent from patients and approval of the local ethics committee was received. All samples were scanned using microCT (SCANCO Medical). Humeral heads were also scanned by magnetic resonance imaging, MRI (Philipps). A range of new metrics were defined and applied to the data. A mixed models analysis, adjusted for multiple comparisons, was performed to test for significant differences (p < 0.05).

The metrics for the rabbit model are sensitive to changes due to traumatic OA. The lateral compartment of operated joints had larger joint space width, thicker femoral cartilage and reduced bone volume, Figure 1a [1]. The humeral head results indicate a spectrum of arthritis disease [2]. Samples were classified as OA-like, OP-like or OA/OP-like, with positive associations between bone and cartilage morphometric parameters, Figure 1b.

The ability to correlate whole joint, bone and cartilage changes is valuable, as these cues allow for a targeted diagnostic approach and a better classification of the disease.