Integrating preoperative gait kinematics into total knee arthroplasty planning: Associations with intraoperative robotic laxities and prediction of postoperative function
Jean Baltzer, Alix Cagnin, Julien Erard, Clément Favroul, Laurence Chèze, Nicola Hagemeister, Alex Fuentes, Elvire Servien, Cécile Batailler, Sébastien LustigAbstract
Purpose
Total knee arthroplasty (TKA) is evolving from static measures towards function‐focused, robotic‐assisted personalisation. Integrating knee kinematics captured under weight‐bearing conditions may add clinically relevant information, but remains underexplored. This study aimed to explore the relevance of integrating preoperative knee dynamic coronal alignment during gait (mean varus‐valgus [V‐V] alignment during stance/swing) to better inform TKA planning.
Methods
This prospective observational study enrolled 30 consecutive patients scheduled for robotic‐assisted TKA; 26 completed a preoperative 3D gait assessment (knee Kinesiography exam). Of these, 17 completed a 6‐month postoperative gait assessment. The primary analysis evaluated associations between preoperative dynamic V‐V alignment and intraoperative laxities (in extension and flexion) using Pearson correlations. The secondary analysis explored whether preoperative dynamic V‐V and intraoperative planned laxities jointly predicted postoperative dynamic V‐V alignment using multiple linear regressions in a varus morphotype subgroup ( n = 13). Pearson correlations, Chi‐square, T ‐tests were also used to evaluate associations between postoperative dynamic V‐V alignment and PROMs (satisfaction and knee osteoarthritis outcome score joint‐replacement [KOOS‐JR]).
Results
The primary analysis ( n = 26) reported that patients exhibiting more varus alignment during gait before TKA had significantly greater initial lateral laxity in extension, and smaller medial laxity in flexion (0.592 ≥ │ r │≥0.475, p ≤ 0.03). Secondary analyses showed significant prediction of postoperative dynamic V‐V alignment with preoperative dynamic V‐V and robotic planned laxities in flexion as predictors ( R 2 > 0.90, all p ≤ 0.04). Postoperative dynamic V‐V during stance was associated with KOOS‐JR scores (postoperative dynamic varus >2°, 94.4 vs. 73.2, p < 0.001).
Conclusion
Results suggest that preoperative knee coronal kinematics during gait is linked to intraoperative laxities in TKA. Exploratory results further indicate associations between dynamic behaviour and outcomes, while suggesting that complementing robotic planned laxities with preoperative dynamic coronal alignment may enhance the ability to predict post‐TKA function in varus knees. Although additional studies will be needed, gait kinematics promise to help personalise TKA.
Level of Evidence
Level IV, observational study.