Methods: Two cohorts of 50 patients each were preoperatively matched to receive the same TKA, having a J-curve femoral design with an adapted "medially congruent" polyethylene insert; the second cohort (group B) underwent the intraoperative sensor-check. Intraoperative sensor data were recorded as tibiofemoral load at 10°, 45°, and 90°. We considered stable knees those with a pressure <50 lbs on the medial compartment, <35 lbs on the lateral, and a mediolateral inter-compartmental difference <15 lbs. Clinical outcomes were evaluated according to the Oxford Knee Score (OKS) and Knee Society Score (KSS).
Results: All patients (group A: no sensor; group B: sensor) were available at 2-year minimum follow-up (FU; min. 24 months, max. 34 months); no preoperative statistical differences existed between groups in the average range of motion (ROM), OKS, KSS, and body mass index. There were no statistical differences at final FU between groups in the average OKS (group A: 41.1; group B: 41.5), in the average KSS (group A: 165.7; group B: 166.3), or in final ROM (group A: 123°; group B: 124°). One patient in each group required a manipulation under anesthesia. In the sensor group, an accessory soft tissue release/bone recut was necessary after sensor testing with trial components in 24% to obtain the desired loads; in the same group, the level of constraint in the final components was increased to posterior-stabilized in 12% because of an inter-compartmental difference >40 lbs. Surgical time was 8 min longer in the sensor group.
Conclusion: The use of this sensing technology did not improve the clinical outcome but supported multiple intraoperative decisions aimed to better reproduce the medial pivot kinematic of the normal knee.
Keywords: TKA; knee; medial pivot; navigation; sensor; soft tissue balance; total knee arthroplasty.