Abstract
PURPOSE: Accurate femoral component (FC) rotation is critical in total knee arthroplasty (TKA). Residual posterior condylar cartilage is difficult to assess intraoperatively and may bias the posterior condylar axis. Quantitative data remain scarce, particularly for lateral compartment osteoarthritis, and the assumption of a uniform 2-mm cartilage thickness has rarely been tested. METHODS: Twenty-five valgus knees in 22 Asian women undergoing primary TKA were analysed; 31 previously measured female varus knees, assessed with the same protocol, served as controls. After posterior femoral resection, resected condyles were radiographed perpendicular to the cut. Medial (m) and lateral (l) cartilage thicknesses were measured with magnification adjustment. The thickness difference (C (diff)) was regressed against theoretical rotational deviation (θ). The slope and intercept were compared between valgus and varus cohorts. The proportion of knees within a prespecified '2 mm range' (1.5-2.5 mm) was calculated. RESULTS: FC rotation shifted 1.31° per 1-mm cartilage asymmetry (95% confidence interval [CI]: 1.28-1.35; R (2) = 0.99). Slopes were similar in varus and valgus cohorts. The mean θ was 1.6° ± 1.3°, with values up to 4°. Thin cartilage (<1.5 mm) was more frequent in varus than valgus (42% vs. 24%), not significant. Overall, ≈52% of knees fell within the '2 mm range' (1.5-2.5 mm). CONCLUSION: Residual posterior condylar cartilage directly affects FC rotation in TKA. In Asian women, each 1-mm medial-lateral asymmetry produced ≈1.3° of rotational error irrespective of deformity pattern. Only half of the knees approximated 2 mm, indicating that a fixed 2-mm correction is not universally applicable. LEVEL OF EVIDENCE: Level III.