Preoperative age, BMI and KL grade predict long-term functional outcome of knee arthroplasty better than CPAK-classification

Purpose: To assess whether coronal plane alignment of the knee (CPAK) is associated with functional outcome of total knee arthroplasty (TKA) and identify factors predicting functional outcomes during 10 years of follow-up in a prospective cohort. Methods: In total, 250 patients aged 65 or younger were enrolled. Primary knee arthroplasty was performed using mechanical alignment (MA). Baseline variables, including joint line obliquity (JLO) and arithmetic hip-knee angle (aHKA), were used to predict Oxford knee score (OKS) at 1, 2, 5, 8 and 10 years postoperatively using ordinary least squares regression analysis. Variable importance and overall predictive ability were assessed. Results: In total, 203 patients were included in the analyses. At 1 year, the mean for OKS was 39.7 (SD = 7.3), and (Formula presented.) was 0.247, showing moderate predictive performance. Higher age and higher Kellgren–Lawrence (KL) grade predicted higher OKS at 1 year. At 2 years, the mean for OKS was 40.5 (SD = 7.1), and (Formula presented.) was 0.298. Higher age, higher preoperative OKS, lower BMI and higher KL grade predicted better OKS at 2 years. At 5 years, the mean for OKS was 41.2 (SD = 6.7), and (Formula presented.) was 0.219. None of the variables were statistically significant at 5 years. At 8 years, the mean OKS was 41.5 (SD = 6.9), and (Formula presented.) decreased to 0.074. Higher BMI predicted lower, though not significant, OKS. At 10 years, the mean for OKS was 36.1 (SD = 6.6), and (Formula presented.) was 0.233. Variables that predicted higher OKS at 10 years were valgus mechanical axis, valgus aHKA and higher KL grade. JLO and aHKA had a negligible effect on OKS at 1, 2 and 8 years. At 5 and 10 years, aHKA and JLO influenced the OKS, but clinical significance is uncertain. Conclusion: Predictive ability of preoperative variables was moderate. CPAK variables were uncertain predictors at every follow-up in relation to age, BMI and KL grade. Level of Evidence: Level II.