Femoral Bowing in Total Knee Arthroplasty: Angular Deviation as a Primary Mediator of Poor Five-Year Function and Evidence-Based Thresholds for Intervention.

Abstract

Femoral bowing compromises intramedullary-guided distal femoral resection accuracy in total knee arthroplasty (TKA), yet its impact on five-year functional outcomes remains incompletely characterized. This study investigated whether bowing affects TKA outcomes through angular deviation during resection and established clinically actionable thresholds for risk stratification. This study analyzed 422 patients who underwent primary TKA with intramedullary-guided resection with five-year follow-up. Femoral bowing angle (FBA), postoperative angular deviation of resection (postoperative-Δθ), alignment, and Knee injury and osteoarthritis outcome scores (KOOS) were assessed. Pearson's correlation, multivariate linear regression, and mediation analyses evaluated mechanistic pathways, while receiver operating characteristic analyses established thresholds. Patients were stratified by thresholds for comparison. The mean FBA was 1.4 ± 2.9° (range, -5.9 to 11.5), with 6.9% exhibiting severe bowing. The FBA correlated with postoperative-Δθ (r = 0.716) and negatively with five-year KOOS (r = -0.363) (both P-values < 0.001). Mediation analysis revealed that bowing's impact on improvement in functional outcomes is primarily through postoperative-Δθ (β = -0.356, P < 0.001), rather than mechanical alignment. The thresholds were established: |FBA| = 5.9° and |postoperative-Δθ| = 4.6° (both P-values < 0.001). Patients who had severe bowing had worse pain and sports outcomes, while those who had unacceptable alignment deviation had worse daily living and quality of life outcomes (all P-values < 0.05). Patients who had both severe bowing and unacceptable deviation fared worse across all scales compared to patients who had neither condition (all P-values < 0.05). Bowing adversely affects TKA functional outcomes primarily through angular deviation during resection rather than through direct effects or alignment alone. The established thresholds provide |FBA| = 5.9° for preoperative risk stratification and |postoperative-Δθ| = 4.6° for intraoperative accuracy objective. For patients who have severe bowing, advanced surgical technologies to enhance resection accuracy may be warranted to optimize functional outcomes.