Pre-resection ligament balance more frequently reproduces joint line orientation compared to inverse-kinematic workflow in robotic total knee arthroplasty.

Abstract

Robotic total knee arthroplasty (TKA) enables the acquisition of ligament balance data at various stages in the surgical workflow to inform implant planning. This study aims to compare pre-resection joint balancing versus inverse-kinematic workflows on implant positioning, joint line obliquity and coronal plane alignment of the knee (CPAK) phenotype. A single-surgeon single-institution series of 422 primary posterior-cruciate sacrificing TKAs using a digital ligament tensioner was reviewed. The first 213 cases assessed gaps after tibial and before femoral resections (tibia-first group) in an inverse-kinematic alignment (iKA) workflow. The subsequent 174 cases assessed gaps before any bone resections (pre-resection group) in a femur-first kinematic alignment (KA) workflow. Native and postoperative medial proximal tibial angle (MPTA) and lateral distal femoral angle (LDFA) were compared using Wilcoxon rank-sum tests, and CPAK phenotype distributions were compared using chi-square tests. Mean preoperative MPTA and LDFA were within 0.5° between groups. Postoperative measurements differed between tibia-first and pre-resection for both MPTA (88.3° ± 1.5° vs. 86.7° ± 1.7°, p < 0.001) and LDFA (90.7° ± 2.1° vs. 89.4° ± 2.4°, p < 0.001). The tibia-first group had greater preoperative to postoperative change in MPTA (0.7° ± 1.3° vs. -0.5° ± 1.7°, p < 0.001), LDFA (2.5° ± 1.7° vs. 1.4° ± 1.5°, p < 0.001) and joint line obliquity (3.2° ± 2.5° vs. 0.9° ± 2.6°, p < 0.001). The pre-resection group had more patients whose CPAK phenotype did not change postoperatively (35% vs. 25%, p < 0.05), and fewer patients with CPAK changes > 1 (9% vs. 29%, p < 0.001). Final mediolateral balance was within 1 mm of neutral for both groups throughout flexion. While both techniques achieved a similar balance, a pre-resection ligament balancing workflow better restored native joint line obliquity and CPAK phenotype compared to an iKA workflow. Level III.