Stemmed femoral knee prostheses: effects of prosthetic design and fixation on bone loss

Although the revision rates for modern knee prostheses have decreased drastically, the total number of revisions a year is increasing because many more primary knee replacements are being done. At the time of revision, bone loss is common, which compromises prosthetic stability. To improve stability, intramedullary stems are often used. The aim of this study was to estimate the effects of a stem, its diameter and the interface bonding conditions on patterns of the bone remodeling in the distal femur. We created finite element models of the distal half of a femur in which 4 types of knee prostheses were placed. The bone remodeling process was simulated using a strain-adaptive bone remodeling theory. The amount of such remodeling was determined by calculating the changes in bone mineral density in 9 regions of interest from simulated DEXA scans. The computer simulation model showed that revision prostheses tend to cause more bone resorption than primary ones, especially in the most distal regions. Predicted long-term bone loss due to a revision prosthesis with a thin stem equalled that around a prosthesis with an intercondylar box. However, strong regional differences were found--the stemmed prostheses having more bone loss in the most distal areas and some bone gain in the more proximal ones. A prosthesis with a thick stem led to an increase in bone loss. When the prosthesis-cement interface was bonded, more bone loss was predicted than with an unbonded interface. These results suggest that a stem which increases stability initially may reduce stability in the long term. This is due to an increase in stress shielding and bone resorption.