Background:In addition to cartilage degeneration, knee osteoarthritis (OA) causes bone changes, including cortical bone thickening, subchondral bone density decrease, and bone shape changes as a result of widening and flattening condyles and osteophyte formation. Knee joint distraction (KJD) is a joint-preserving treatment for younger (<65 years) knee OA patients that has been shown to reverse OA cartilage degradation. On radiographs, KJD showed a decrease in subchondral bone density and an increase in osteophyte formation. However, these bone changes have never been evaluated with a 3D imaging technique.Objectives:To evaluate cortical bone thickness, subchondral trabecular bone density, and bone shape on CT scans before and one year after KJD treatment.Methods:19 KJD patients were included in an extended imaging protocol, undergoing a CT scan before and one year after treatment. Stradview v6.0 was used for semi-automatic tibia and femur segmentation from axial thin-slice (0.45mm) CT scans. Cortical bone thickness (mm) and trabecular bone density (Hounsfield units, HU) were measured with an automated algorithm. Osteophytes were excluded. Afterwards, wxRegSurf v18 was used for surface registration. Registration data was used for bone shape measurements. MATLAB R2020a and the SurfStat MATLAB package were used for data analysis and visualization. Two-tailed F-tests were used to calculate changes over time. Two separate linear regression models were used to show the influence of baseline Kellgren-Lawrence grade and sex on the changes over time. Statistical significance was calculated with statistical parametric mapping; a p-value <0.05 was considered statistically significant. Bone shape changes were explored visually using vertex by vertex displacements between baseline and follow-up. Patients were separated into two groups based on whether their most affected compartment (MAC) was medial or lateral. Only patients with axial CT scans at both time points available for analysis were included for evaluation.Results:3 Patients did not have complete CTs and in 1 patient the imaged femur was too short, leaving 16 patients for tibial analyses and 15 patients for femoral analyses. The MAC was predominantly the medial side (medial MAC n=14; lateral n=2). Before treatment, the MAC cortical bone was compared to the rest of the joint (Figure 1). One year after treatment, MAC cortical thickness decreased, although this decrease of up to approximately 0.25 mm was not statistically significant. The trabecular bone density was also higher before treatment in the MAC, and a decrease was seen throughout the entire joint, although statistically significant only for small areas on mostly the MAC where this decrease was up to approximately 80 HU (Figure 1). Female patients and patients with a higher Kellgren-Lawrence grade showed a somewhat larger decrease in cortical bone thickness. Trabecular density decreased less for patients with a higher Kellgren-Lawrence grade, and female patients showed a higher density decrease interiorly while male patients showed a higher decrease exteriorly. None of this was statistically significant. The central areas of both compartments showed an outward shape change, while the outer ring showed inward changes.Conclusion:MAC cortical bone thickness shows a partial decrease after KJD. Trabecular bone density decreased on both sides of the joint, likely as a direct result of the bicompartmental unloading. For both subchondral bone parameters, MAC values became more similar to the LAC, indicating (partial) subchondral bone normalization in the most affected parts of the joint. The bone shape changes may indicate a reversal of typical OA changes, although the inward difference that was seen on the outer edges may be a result of osteophyte-related changes that might have affected the bone segmentation. In conclusion, KJD treatment shows subchondral bone normalization in the first year after treatment, and longer follow-up might show whether these changes are a temporary result of joint unloading or indicate more prolonged bone changes.Disclosure of Interests:None declared.
3D image registration is critical to ensure accurate detection of longitudinal changes in trabecular bone density, microstructure, and stiffness measurements in rat tibiae by in vivo microcomputed tomography (μCT)