Abstract
Introduction
Medial unicompartmental knee arthroplasty (UKA) restores mechanical alignment and reduces lateral subluxation of the tibia. However, medial compartment translation remains abnormal compared to the native knee in mid-flexion Intra-operative adjustment of implant thickness can modulate ligament tension and may improve knee kinematics. However, the relationship between insert thickness, ligament loads, and knee kinematics is not well understood. Therefore, we used a computational model to assess the sensitivity of knee kinematics, and cruciate and collateral ligament forces to tibial component thickness with fixed bearing medial UKA.
Methods
A computational model of the knee with subject-specific bone geometries, articular cartilage, and menisci was developed using multibody dynamics software (Fig 1a). The ligaments were represented with multiple non-linear, tension-only force elements, and incorporated mean structural properties. The 3D geometries of the femoral and tibial components of the Stryker Triathlon fixed-bearing UKA were captured using a laser scanner. An arthroplasty surgeon aligned the femoral and tibial components to the articular surfaces within the model (Fig 1b). The intact and UKA models were passively flexed from 0 to 90° under a 10 N compressive load. The tibial polyethylene insert was modeled by the orthopaedic surgeon to create a “balanced” knee. The modeled polyethylene insert thickness was then increased by 2 mm and decreased 2mm (in increments of 1mm) to simulate over- and under-stuffing, respectively. Outcomes were anterior-posterior (AP) translation of the femur on the tibia in the medial compartment, and forces seen by the ACL and MCL during mid-flexion (from 30 to 60° flexion). The mean differences between the intact knee model and all other experimental conditions for each outcome were calculated across mid-flexion.
Results
All outcomes are presented relative to the intact knee model. A balanced medial UKA caused an average of 1.7 mm anterior translation of the medial compartment through mid-flexion (Table 1). Understuffing increased anterior translation of the medial compartment up to 3.7 mm on average. Overstuffing altered the anterior position of the medial compartment to 1.2 mm on average. The predicted ACL and MCL average loads in the balanced medial UKA differed from the intact model through mid-flexion by 2.7 N and 6.0 N, respectively (Table. 1). Overstuffing increased ACL and MCL load by 15.5 N and by 27.2 N, respectively. Understuffing by 2 mm decreased ACL and MCL load by 1.5 N at most.
Discussion
A computational model incorporating a fixed-bearing medial UKA revealed that understuffing increased AP translation of the medial compartment in mid-flexion. In contrast, overstuffing increased ACL and MCL loads in mid-flexion). Thus, it is critical to achieve a compromise between insert thickness and ligament balance to obtain the most normal kinematics and ligament loads. Interestingly, no amount of over- or under-stuffing restored the AP position of the medial compartment to that of the intact model. This could be due to the inability of the flat polyethylene surface to provide “normal” AP constraint. Altogether, sensitivity analyses using computational modeling provide a valuable tool to assess the effect of implant position on knee mechanics.
