Early cement damage around a femoral stem is concentrated at the cement/bone interface

This study aimed to improve understanding of the mechanical aspects of cemented implant loosening. After aggressive fatigue loading of stem/cement/femur constructs, micro-cracks and stem/bone micro-motions were quantified to answer three research questions: Are cracks preferentially associated with the stem/cement interface, the cement/bone interface or voids? Is cement damage dependent on axial position? Does cement damage correlate with micro-motion between the stem and the bone? Eight Charnley Cobra stems were implanted in cadaveric femora. Six stem/cement/femur constructs were subjected to "stair-climbing" loads for 300 kcycles at 2Hz. Loads were normalized by construct stiffness to avoid fracture. Two additional constructs were not loaded. Transverse sections were cut at 10mm intervals, stained with a fluorescent dye penetrant and examined using epi-fluorescence stereomicroscopy. Crack lengths and cement areas were recorded for 9 sections per specimen. Crack length-density was calculated by dividing summed crack length by cement mantle area. To isolate the effect of loading, length-density data were offset by the baseline length-density measured in the non-loaded specimens. Significantly more cracks were associated with the interdigitated area (35.1%+/-11.6%) and the cement/bone interface (31.0%+/-6.2%) than with the stem/cement interface (11.0%+/-5.2%) or voids (6.1%+/-4.8%) (p<0.05). Load-induced micro-crack length-density was significantly dependent on axial position, increasing proximally (p<0.001). Micro-motions were small, all stems rotated internally. Cement damage did not correlate with micro-motion.