Abstract
INTRODUCTION
The purpose of TKA is to restore normal kinematics and functioning to diseased knees. The purpose of this study was to determine whether intraoperative kinematic data are correlated with minimum one-year outcomes following primary TKA.
METHODS
We reviewed data on 185 consecutive primary TKAs in which sensor-embedded tibial trials were used to evaluate kinematic patterns following traditional ligament balancing. Procedures were performed by two board-certified arthroplasty surgeons. The same implant design and surgical approach was used for all knees.
Contact locations on the medial and lateral condyles were recorded for each patient at 0°, 45° and 90° of flexion, and full flexion. Vector equations were created by contact locations on the medial and lateral sides and the vector intersections determined the center of rotation between each measurement position. Center of rotation was calculated as the average of vector intersections at 0 to 45°, 45 to 90°, and 90° to full flexion. If the average center of rotation was between 16 and 1000 mm of the contact location on the medial side it was considered a medial pivot knee. Knees were also classified as medial (16 to 200 mm on medial side), lateral (16 to 200 mm on lateral side), translating (> 200 mm medially or laterally), and other (< 16 mm on both medial and lateral sides). The new Knee Society Scoring System (KSSO objective score, KSSS satisfaction score, KSSF function score), the EQ-5D™ Health Status Index, and the University of California Los Angeles (UCLA) Activity Level Score were measured preoperatively and at minimum one-year follow-up (average 20.4 months).
RESULTS
Thirty-three TKAs were excluded to eliminate potential bias due to sensor device malfunction, atypical hardware, unresurfaced patella, surgery at a non-study hospital, or early postoperative infection, revision due to aseptic loosening, ipsilateral hip disease, and subsequent neurologic disease or death unrelated to the index TKA, resulting in a final sample size of 152 knees. Twelve (7.9%) patients were lost to follow-up, and two were excluded due to outlier values for average center of rotation. Seventy-five percent of the final sample was female. Mean age and BMI were 63.6 years 33.9, respectively.
Average center of rotation ranged from −1017 to 1562 mm with negative signifying the lateral side. Medial pivot knees comprised 40% (55) of the total sample. Sex, age, height, surgeon, implant side, and implant type were unrelated to pivot classification. Patient weight (100.2 vs. 90.9 kg; p = 0.012) and BMI (35.5 vs. 32.8; p = 0.044) were greater in medial pivot knees.
Controlling for BMI, KSSO, KSSF, KSSS, EQ5D, UCLA, and pain scores at latest follow-up did not differ in medial and non-medial pivot knees (p ≥ 0.151). The amount of improvement in outcomes from preoperative baseline also did not differ in medial and non-medial pivot knees (p ≥ 0.161). Outcomes did not vary among knees with translating medial, lateral, and other pivots (p ≥ 0.065).
DISCUSSION
Our results suggest that a medial pivot kinematic pattern may not be a substantial governor of clinical success.