Comparison of anatomic axes with a navigated functional rotation axis determined by ligament tension for rotational femoral component alignment in cadaver knee arthroplasty

TL;DR

This study compares different methods for aligning knee implants and finds that all show significant deviations from a gold standard reference axis.

Contribution

A new navigated functional rotation axis is compared with traditional anatomic axes for femoral component alignment in knee arthroplasty.

Findings

All tested axes showed significant angular deviation from the flexion-extension axis (FEA).

The anatomical transepicondylar axis (aTEA) had the highest mean angular deviation (5.2°).

The surgical transepicondylar axis (sTEA) and posterior condylar axis (PCA) showed smaller deviations.

Abstract

The malimplantation of the total knee arthroplasty (TKA) components is one of the main reasons for revision surgery. For determining the correct intraoperative femoral rotation several anatomic rotational axes were described in order to achieve a parallel, balanced flexion gap. In this cadaveric study prevalent used rotational femoral axes and a navigated functional rotational axis were compared to the flexion-extension axis defined as the gold standard in rotation for femoral TKA component rotation.

Thirteen body donors with knee osteoarthritis (mean age: 78.85 ± 6.09; eight females and five males) were examined. Rotational computer tomography was performed on their lower extremities pre- and postoperatively. Knee joint arthroplasties were implanted and CT diagnostics were used to compare the preoperatively determined flexion-extension axis (FEA). The FEA is the axis determined by our surgical technique and serves as an internal reference. It was compared to other axes such as (i) the anatomical transepicondylar axis (aTEA), (ii) the surgical transepicondylar axis (sTEA), (iii) the posterior condylar axis (PCA) and (iv) the functional rotation axis (fRA).

Examination of 26 knee joint arthroplasties revealed a significant angular deviation (p*** < 0.0001) for all axes when the individual axes and FEA were compared. aTEA show mean angular deviation of 5.2° (± 4.5), sTEA was 2.7° (± 2.2), PCA 2.9° (± 2.3) and the deviation of fRA was 4.3° (± 2.7). A tendency towards external rotation was observed for the relative and maximum axis deviations of the aTEA to the FEA, for the sTEA and the fRA. However, the rotation of the posterior condylar axis was towards inwards.

All axes showed a significant angular deviation from the FEA. We conclude that the presented technique achieves comparable results in terms of FEA reconstruction when compared with the use of the known surrogate axes, with certain deviations in terms of outliers in the internal or external rotation.

The online version contains supplementary material available at 10.1007/s00402-024-05394-3.

Figures