Quantification of Implant Micromotion, Strain Shielding, and Bone Resorption With Porous-Coated Anatomic Medullary Locking Femoral Prostheses

Fourteen femora containing porous-coated anatomic medullary locking (AML) femoral prostheses were retrieved from 12 patients at autopsy. Clinical roentgenograms in 13 femora showed bone remodeling changes, indicating that the implants were fixed by osseointegration. Under simulated physiologic loading, micromotion between the implant and the bone was measured using electrical displacement transducers connected to the implant and to the adjacent cortex. The micromotion between the implants at the areas of porous coating and the adjacent cortex in the one case of failed bone ingrowth measured 150 microns. Maximum relative motion between the cortex and the implant in the areas of porous coating for the 13 cases showing signs of bone ingrowth was 40 microns, and this was completely elastic relative displacement. With all implants, the micromotion between the cortex and the stem was always greatest over the uncoated portion of the stem. Four of the implants were proximally porous coated. With these, the micromotion was greater over the uncoated areas than with more extensively coated stems and was always greatest at the uncoated tip of the prosthesis. The amount of micromotion was directly related to the extent of porous coating on the implant. Maximum tip motion for the proximally coated implants was 210 micra, whereas for the fully porous-coated implants, it was 40 microns. In nine of the autopsies, the contralateral normal femur was obtained in addition to the femur containing the AML (the in vivo remodeled femur). These were used for comparative studies of strain shielding and femoral remodeling. Cortical strains were measured in the in vivo remodeled femora and were compared with measurements made in the contralateral normal femora before and following implantation of a stem identical to that present on the clinically treated side. The data showed major strain reductions in all the postmortem implanted normal femora. Comparison of the strain data from the postmortem implanted normal femora with those from the in vivo remodeled femora clearly indicated that extensive bone remodeling did not result in restoration of cortical strain levels anywhere near normal. Strain shielding continued to exist in all of the remodeled specimens, even up to 7.5 years after surgery. This strain shielding was associated with bone remodeling changes that resulted in regional reductions in bone mineral content that ranged from 7% to 78%. These observations are unique, important, and valuable in defining the in vivo function and clinical behavior of this type of porous-coated femoral component.