Abstract
Introduction
Component position and overall limb alignment following total knee arthroplasty (TKA) have been shown to influence prosthetic survivorship and clinical outcomes1. The objective of this study was to compare the accuracy to plan of three-dimensional modeled (3D) TKA with manual TKA for component alignment and position.
Methods
An open-label prospective clinical study was conducted to compare 3D modeling with manual TKA (non-randomized) at 4 U.S. centers between July 2016 and August 2018. Men and women aged > 18 with body mass index < 40kg/m2 scheduled for unilateral primary TKA were recruited for the study. 144 3DTKA and 86 manual TKA (230 patients) were included in the analysis of accuracy outcomes. Seven high-volume, arthroplasty fellowship-trained surgeons performed the surgeries. The surgeon targeted a neutral (0°) mechanical axis for all except 9 patients (4%) for whom the target was within 0°±3°. Computed tomography (CT) scans obtained approximately 6 weeks post-operatively were analyzed using anatomical landmarks to determine femoral and tibial component varus/valgus position, femoral component internal/external rotation, and tibial component posterior slope. Absolute deviation from surgical plan was defined as the absolute value of the difference between the CT measurement and the surgeon's operative plan. Smaller absolute deviation from plan indicated greater accuracy. Mean component positions for manual and 3DTKA groups were compared using two-sample t tests for unequal variances. Differences of absolute deviations from plan were compared using stratified Wilcoxon tests, which controlled for study center and accounted for skewed distributions of the absolute values. Alpha was 0.05 two-sided. At the time of this report, CT measurements of femoral component rotation position referenced from the posterior condylar axis were not yet completed; therefore, the current analysis of femoral component rotation accuracy to plan reflects one center that exclusively used manual instruments referencing the transepicondylar axis (TEA).
Results
Coronal positions of the femoral components measured via CT for manual and 3D TKA, respectively, were (mean ± standard deviation) 0.1°±1.6° varus and 0.0°±1.4° varus (p=0.533); positions of the tibial components were 1.9°±2.4° varus and 0.9°±2.0° varus (p=0.002). Positions of external femoral component rotation relative to the TEA were 1.1°±2.3° and 0.5°±2.3°, respectively (p=0.036). Tibial slopes were 3.7°±3.0° and 3.2°±1.8°, respectively (p=0.193). Comparing absolute deviation from plan between groups, 3DTKA demonstrated greater accuracy for tibial component alignment [median (25th, 75th percentiles) absolute deviation from plan, 1.7° (0.9°, 2.9°) vs. 0.9°(0.4°, 1.9°), p<.001], femoral component rotation [1.4° (0.9°, 2.5°) vs. 0.9° (0.7°, 1.5°), p=0.015], and tibial slope [2.9° (1.5°, 5.0°) vs. 1.1° (0.6°, 2.0°), p<.001] (Table 1). Accuracy for femoral component alignment was comparable [1.0° (0.4°, 1.7°) vs. 0.9° (0.4°, 1.5°), p=0.159] (Table 1).
Discussion and Conclusions
Our findings support improved accuracy to the surgical plan utilizing 3DTKA compared with manual TKA. Compared to manual TKA, 3DTKA cases were typically 47% more accurate for tibial component alignment, 62% more accurate for tibial slope, and 36% more accurate for femoral component rotation (calculated as percent reduction of median absolute deviation). The evaluation of femoral component coronal alignment reflected already very good baseline accuracy of the surgeons utilizing the intramedullary femoral guide system (Table 1). As optimal component position in TKA affects joint kinematics and may positively influence implant longevity, it is important for surgeons to maximize the opportunity to direct component positioning. Further clinical data is needed to study potential longer-term benefits of robotic technologies.
For figures, tables, or references, please contact authors directly.
