Abstract
Arthrofibrosis remains a dominant post-operative complication and reason for returning to the OR following total knee arthroplasty. Trauma induced by ligament releases during TKA soft tissue balancing and soft tissue imbalance are thought to be contributing factors to arthrofibrosis, which is commonly treated by manipulation under anesthesia (MUA). We hypothesized that a robotic-assisted ligament balancing technique where the femoral component position is planned in 3D based on ligament gap data would result in lower MUA rates than a measured resection technique where the implants are planned based solely on boney alignment data and ligaments are released afterwards to achieve balance. We also aimed to determine the degree of mechanical axis deviation from neutral that resulted from the ligament balancing technique.
Methods
We retrospectively reviewed 301 consecutive primary TKA cases performed by a single surgeon. The first 102 consecutive cases were performed with a femur-first measured resection technique using computer navigation. The femoral component was positioned in neutral mechanical alignment and at 3° of external rotation relative to the posterior condylar axis. The tibia was resected perpendicular to the mechanical axis and ligaments were released as required until the soft tissues were sufficiently balanced. The subsequent 199 consecutive cases were performed with a tibia-first ligament balancing technique using a robotic-assisted TKA system. The tibia was resected perpendicular to the mechanical axis, and the relative positions of the femur and tibia were recorded in extension and flexion by inserting a spacer block of appropriate height in the medial and lateral compartments. The position, rotation, and size of the femoral component was then planned in all planes such that the ligament gaps were symmetric and balanced to within 1mm (Figure 1). Bone resection values were used to define acceptable limits of implant rotation: Femoral component alignment was adjusted to within 2° of varus or valgus, and within 0–3° of external rotation relative to the posterior condyles. Component flexion, anteroposterior and proximal-distal positioning were also adjusted to achieve balance in the sagittal plane. A robotic-assisted femoral cutting guide was then used to resect the femur according to the plan (Figure 2).
CPT billing codes were reviewed to determine how many patients in each group underwent post-operative MUA. Post-operative mechanical alignment was measured in a subset of 50 consecutive patients in the ligament balancing group on standing long-leg radiographs by an independent observer.
Results
Post-operative MUA rates were significantly lower in the ligament balancing group (0.5%; 1/199) than in the measured resection group (3.9%; 4/102), p=0.051. 91.3% (42/46) of knees were within 3° and 100% (46/46) were within 4° of neutral alignment to the mechanical axis post-operatively in the ligament balancing group.
Conclusions
Gap driven femoral based planning in TKA resulted in a significantly lower post-operative manipulation rate than in the measured resection approach, while maintaining acceptable overall alignment to the mechanical axis.
