Accuracy of anatomical references used for rotational alignment of tibial component in total knee arthroplasty

Purpose This study aimed to research which was the most reliable of the four techniques based on local anatomic markers used to determine tibial component rotation in total knee arthroplasty, and whether the markers varied in knees with varus deformity. Methods The study included 33 knees with a normal anatomic axis and 32 knees with a varus deformity and osteoarthritis. On the MR images, the femoral transepicondylar axis (TEA) was determined and transposed to the standard tibial resection level. At this level, four axes were drawn on the axial sections: tibial posterior condylar line (PC), tibial plateau anterior line (AC), a vertical line (AA) drawn to Akagi’s line, and the maximal mediolateral distance (MMLD). The relationships of these lines and the transposed TEA were compared between two groups. Results In all the knees, the mean values of the PC, AA, and MMLD axes compared to TEA reference were 5.5° ± 5.7 (mean ± SD), 7° ± 3.2, and 6.7° ± 8.1 internal rotation, respectively, and the AC axis was 8.9° ± 6.7 external rotation. In the AC, AA, and MMLD axes, the change occured because of varus deformity was statistically meaningful. For all the observers, the axis with the least SD and the most accuracy was the AA axis. Conclusions Of the four axes used to determine tibial component rotation, only the PC axis is not affected by varus deformity, and the least affected axis according to the observers was the AA axis, and thus the AA and PC axes can be used for guidance in determining the rotation of the tibial component. Level of evidence Prognostic studies—investigating natural history and evaluating the effect of a patient characteristic: High-quality prospective cohort study with >80% follow-up, and all patients enrolled at same time point in disease, Level I.