Quantification of bone microarchitecture damage in human femoral press-fit implantation using HR-pQCT and digital volume correlation

INTRODUCTION: During prostheses press-fit implantation, high compressive and shear stresses at the implant-bone interface are generated.Permanent bone damage occurs[1], but the extent remains unknown.The aim of this study isto quantify, using high-resolution peripheral quantitative computed tomography (HR-pQCT) and Digital Volume Correlation (DVC), permanent bone changes and deformation due to press-fit femoral knee implantation.METHODS: Pre-implantation scan: Six human cadaveric distal femora (85±3 years) were resected and scanned with HR-pQCT (XtremeCT II, Scanco) at 60.7μm/voxel, isotropic.Implant fitting: An experienced surgeon fitted the femurs with a cementless cruciate retaining femoral knee implant Sigma® (DePuy Synthes, UK)with porous surface coating (Porocoat). Post-implantation scan: Implants were then carefully removed and femurs rescanned with HR-pQCT. Image Registration:3D image-registration of pre- and post-implantation cross-section images was performed (DataViewer, Skyscan-Bruker). Image Analysis: For each femur, volumes of interest(VOI) were selected (CT-Analyzer, Skyscan-Bruker) in the posterior-medial and posterior-lateral condyles and anterior flange; VOIs included 10 mm of depth in the coronal plane atthe bone-implant surface. The bone volume fraction (BV/TV) for the VOIs in pre- and post-implantation images and their ratio (BV/TVpost/BV/TVpre) was calculated, slice by slice, at increasing depth. Digital Volume Correlation:DVC direct correlation (DaVis, LaVision) [2](3-step multipass, final subvolumes 20 voxels, 1.2 mm)was applied on the same VOIs pre- and post-implantation, to assess trabecular bone displacements and plastically accumulated strains. RESULTS AND DISCUSSION: The “BV/TVpost/BV/TVpre ratio vs. depth” graphs (Fig. 1a) showed,consistently among the six femurs, three characteristic points, differing significantly in BV/TVpost/BV/TVpre ratios and depth among them(p<0.05, Friedman and Wilcoxon signed rank tests), indicating: bone removal (A, ratio<100%),compaction (B, ratio>100%) and correspondence between pre- and post-implantation (C, ratio=100%)(Fig.1a, b). Correspondingly, trabecular bone displacement computed by DVC for the 6 femurs, suggested bone compaction up to 2.3±0.5mm in depth(Fig.1c), with peak third principal strains of -150,500±39,500με(mean absolute error: 1,000-2,000με, SD: 200-500με; Fig.1d), well above the yield strain of bone (7,000-10,000με). CONCLUSIONS: Combining high-resolution 3D-imaging and DVC enables to quantify permanent bone damage occurring after press-fit knee implantation. The apparent damage occurs up to 2 mm in depth, first with bone removal, followed by compaction and by no changes (correspondence). This data can be used to inform surgeons and manufacturers, advancing implant development.