A Physis-Based Three-Dimensional CT Reference for Precise Evaluation of Varus Deformity Origin and Joint Line Reconstruction in Patients with Severe Knee Osteoarthritis: A Retrospective Observational Study

Abstract Background To evaluate the efficacy of a physis-based three-dimensional CT measurement method for assessing varus deformity origin and reconstructing the native joint line in severe knee osteoarthritis (KOA), and to validate its clinical utility in total knee arthroplasty (TKA) and unicompartmental knee arthroplasty (UKA). Methods Three parallel substudies were conducted. (1) A cross-sectional CT study (258 patients, 516 knees) compared the lateral distal femoral angle (LDFA) and medial proximal tibial angle (MPTA) on the conventional bony surface (b-LDFA, b-MPTA) versus the physis reference plane (e-LDFA, e-MPTA). (2) A retrospective longitudinal study (53 bilateral KOA patients, imaging ~ 1 year apart) examined dynamic changes of both measurements during KOA progression. (3) A retrospective cohort study (96 TKA patients: 49 mechanical alignment [MA] vs. 47 kinematic alignment [KA] based on physis-derived planning) evaluated 1-year outcomes, including hip-knee-ankle angle (HKA), range of motion (ROM), Knee Society Score (KSS), and Forgotten Joint Score (FJS). Results (1) In patients without bone loss, e-LDFA and e-MPTA correlated strongly with b-LDFA and b-MPTA (r = 0.624 and r = 0.674, both p < 0.001), with the relationship unaffected by age or sex. In severe OA, ΔLDFA and ΔMPTA increased significantly (p < 0.05), indicating disruption of the stable physis-to-joint geometry by bone attrition. (2) Longitudinally, b-LDFA increased by 1.17°(p < 0.001) versus 0.64°for e-LDFA (p < 0.001); b-MPTA decreased by 2.54°(p < 0.001) versus 0.97°for e-MPTA (p < 0.001). Changes in bony-surface measurements were significantly larger than those in physis-reference measurements (both p < 0.001), confirming varus progression arises from both intra-articular wear and extra-articular remodeling. (3) At 1 year post-TKA, the KA group retained physiological varus (HKA 176.74°±0.52°) versus the MA group (179.02°±0.57°, p < 0.001). ROM and KSS showed no clinically meaningful differences, but KA group had significantly higher FJS (85.24 ± 3.31 vs. 80.52 ± 2.77, p < 0.001), indicating better joint awareness. Conclusions The physis-based CT method accurately reflects pre-arthritic alignment, enables quantitative differentiation of intra-articular versus extra-articular deformity, and facilitates individualized TKA alignment reconstruction. This approach provides a reliable basis for personalized knee arthroplasty planning and may aid UKA patient selection.