Abstract
Mechanical alignment (MA) techniques for total knee arthroplasty (TKA) may introduce significant anatomic modifications, as it is known that few patients have neutral femoral, tibial or overall lower limb mechanical axes.
A total of 1000 knee CT-Scans were analyzed from a database of patients undergoing TKA. MA tibial and femoral bone resections were simulated. Femoral rotation was aligned with either the trans-epicondylar axis (TEA) or with 3° of external rotation to the posterior condyles (PC). Medial-lateral (DML) and flexion-extension (DFE) gap differences were calculated.
Extension space ML imbalances (3mm) occurred in 25% of varus and 54% of valgus knees and significant imbalances (5mm) were present in up to 8% of varus and 19% of valgus knees. For the flexion space DML, higher imbalance rates were created by the TEA technique (p < 0 .001). In valgus knees, TEA resulted in a DML in flexion of 5 mm in 42%, compared to 7% for PC. In varus knees both techniques performed better. When all the differences between DML and DFE are considered together, using TEA there were 18% of valgus knees and 49% of varus knees with < 3 mm imbalances throughout, and using PC 32% of valgus knees and 64% of varus knees.
Significant anatomic modifications with related ML or FE gap imbalances are created using MA for TKA. Using MA techniques, PC creates less imbalances than TEA. Some of these imbalances may not be correctable by the surgeon and may explain post-operative TKA instability. Current imaging technology could predict preoperatively these intrinsic imitations of MA. Other alignment techniques that better reproduce knee anatomies should be explored.
