High Early Failure Rate of the Columbus® Posterior Stabilized High-flexion Knee Prosthesis

Background Most studies report high survivorship rates for TKAs, however, we observed higher than anticipated rates of dislocation and femoral component loosening after implanting a Columbus (R) posterior stabilized prosthesis. Questions/purpose We therefore determined (1) the incidence of dislocation and aseptic loosening that occurred after implantation of posterior stabilized high-flexion prostheses in TKAs, (2) the causative factors of dislocation and aseptic femoral component loosening when comparing two designs of prostheses, and (3) the mechanisms of dislocation. Methods We retrospectively reviewed 319 patients who underwent 384 TKAs from May 2007 to July 2008. These patients had been assigned alternately to receive a Scorpio (R) posterior stabilized knee prosthesis (Group I, 158 patients, 187 knees) or a Columbus (R) posterior stabilized knee prosthesis (Group II, 161 patients, 197 knees). We followed the patients with clinical and radiographic evaluations for a minimum of 24 months (mean, 26 months; range, 2438 months). Ten retrieved prostheses were examined visually. Results Ten dislocations (5.1%; 10 of 197 knees) and seven aseptic loosenings of femoral components (3.6%; seven of 197 knees) occurred in Group II at a mean of 10.9 months postoperatively. However, no dislocation or loosening occurred in patients in Group I. Most dislocations were associated with varus flexion or flexion rotation movements during normal daily activities. The cam jump distance at 908 flexion for the Columbus (R) prosthesis was lower than for the Scorpio (R) prosthesis. Conclusions We observed a high rate of early failure during short-term followup after implantation of a Columbus (R) posterior stabilized prosthesis. It appears that early failures of the Columbus (R) design were related to a different cam-post design attributable to a low jump distance during knee flexion. We no longer recommend using this device.