Bony landmarks with tibial cutting surface are useful to avoid rotational mismatch in total knee arthroplasty.

PURPOSE: The purpose of this study was to define various anteroposterior axes of the tibial component as references and to evaluate their accuracy and variability using virtual surgery. It was hypothesized that (1) Akagi's Line could result in high accuracy and low variability in varus osteoarthritic knees; (2) anteroposterior axes defined by using the tibial bony cutting surface as a landmark might be good substitutes for Akagi's Line; and (3) extra-articular bony landmarks might influence the variability of the anteroposterior axis.

METHODS: Three-dimensional bone models were reconstructed from the preoperative computed tomography data of 111 osteoarthritic knees with varus deformities. Seven different anteroposterior axes of the tibial component were defined: Akagi's Line, Axis MED, Axis 1/6MED, Axis 1/3MED, Axis of Oval Shape, Axis of Anterior Crest, and Axis Second Metatarsus. The rotational mismatch angle was measured between the tibial anteroposterior axis and the line perpendicular to the transepicondylar axis projected on the cutting surface (positive value: external rotation of the tibial anteroposterior axis).

RESULTS: The average rotational mismatch angles (referring to the projected anatomical/surgical epicondylar axes) were - 2.7° ± 5.8°/1.0° ± 6.0° (Akagi's Line), - 4.2° ± 7.7°/- 0.5° ± 7.8°, 2.9° ± 7.2°/6.6° ± 7.2°, 9.8° ± 7.0°/13.5° ± 6.8° (Axis MED, Axis 1/6MED, Axis 1/3MED), - 5.1° ± 7.9°/- 1.4° ± 7.8° (Axis of Oval Shape), and 19.3 ± 9.5°/23.0° ± 9.6°, - 2.0° ± 11.3°/1.7° ± 11.4° (Axis Anterior Crest, Axis Second Metatarsus), respectively.

CONCLUSIONS: Akagi's Line provided the best accuracy and least variability in varus osteoarthritic knees. Axis 1/6MED and Axis MED are good substitutes for Akagi's Line due to the difficulty of identifying the attachment site of the posterior cruciate ligament after the proximal tibia has been cut. Extra-articular bony landmarks should not be used for alignment due to their high variability. This study will aid surgeons in choosing the proper anteroposterior axis of the tibial component to reduce rotational mismatch and thus achieve good clinical knee outcomes.

LEVELS OF EVIDENCE: III.