Implant malalignment in primary TKA has been reported to increase stresses placed on the bearing surfaces of implant components. We used a longitudinally maintained registry coupled with an implant retrieval program to consider whether preoperative, postoperative, or prerevision malalignment was associated with increased risk of revision surgery after TKA. (1) What is the relative polyethylene damage on medial and lateral compartments of the tibial plateaus from revised TKAs? (2) Does coronal TKA alignment affect implant performance, such that TKAs aligned in varus are predisposed to experience increased polyethylene damage? (3) Does TKA alignment differ between postoperative and prerevision radiographs, and if so, what does this difference suggest about the mechanical contact load placed on a knee with a TKA? Between 2007 and 2012, we performed 18,065 primary TKAs at our institution. By March 2016, 178 of those TKAs (1%) were revised at our center at least 2 years after primary surgery at our institution. Eighteen of those TKAs were excluded from this analysis because the tibial insert was not explanted during revision surgery, and four more were excluded because the inserts were lost or returned to the patient before the study was initiated, leaving 156 retrieved polyethylene tibial inserts (in 153 patients) revised at greater than 2 years after the primary TKA for this retrospective study. Patients who underwent revision surgery elsewhere were not considered here, since this study depended on having retrieved components. Polyethylene damage modes of burnishing, pitting, scratching, delamination, surface deformation, abrasion, and third-body debris were subjectively graded on a scale of 0 to 3 to reflect the extent and severity of each damage mode. On preoperative, postoperative, and prerevision radiographs, overall alignment, femoral alignment, and tibial alignment in the coronal plane were measured according to the protocol recommended by the Knee Society. Knees with more overall varus alignment after TKA had increased total damage on the retrieved tibial inserts (Spearman's rank correlation coefficients of -0.3 [95% CI, -0.4 to -0.1; p = 0.001]). We also found revised TKAs tended to drift back into greater varus before revision surgery, with a mean (SD) of 3.6A degrees +/- 4.0A degrees valgus for postoperative alignment compared with 1.7A degrees +/- 6.4A degrees prerevision (p = 0.04). Despite surgical efforts to achieve neutral mechanical alignment, remaining varus alignment places an increased contact load on the polyethylene articular surfaces. The drift toward further varus alignment postoperatively is consistent with the knee adduction moment remaining high after surgery. While we found a predisposition toward recurrence of the preoperative varus deformity, we did not find increased medial as opposed to lateral polyethylene damage, which may be explained by the curve-on-curve toroidal design of the articulating surfaces of the TKA implants in this study.
