Introduction

Recently, rotatory knee instabilities have gained increasing interest. While the lateral side of the knee has extensively been investigated, little is known about anteromedial rotatory knee instabilities (AMRI). Thus, the purpose of this study was to biomechanically investigate knee kinematics in response to external loads in knees with simulated high-grade anteromedial injuries and reconstructions. It was hypothesized, that in cases of a high-grade anteromedial injury an isolated reconstruction of the superficial medial collateral ligament (sMCL) might not fully restore knee kinematics.

Methods

12 fresh-frozen human cadaveric knees were used for this study. Knees were mounted onto a 6 degrees of freedom robotic testing system(KUKA KR 60-3). Knee kinematics were recorded with a force/moment sensor(ATI Theta FT-sensor; Schunk). The following loads were applied: 1) 134N anterior tibial translation (ATT) performed at 5Nm external tibial rotation (ER) to simulate AMRI, 2) 5Nm ER, 3) 5Nm internal tibial rotation (IR), and 4) 10Nm valgus rotation. These loading conditions were tested in the native knee and in the following knee states: 1) sMCL and deep MCL (dMCL) resection, 2) sMCL reconstruction, 3) posterior oblique ligament (POL) resection, 4) POL reconstruction, 5) anteromedial capsule and retinaculum (AM) resection, and 6) AM reconstruction. The order of antero- and posteromedial cutting and reconstructions was randomized. A repeated-measures ANOVA with a post-hoc Bonferroni correction was performed (p < 0.05).

Results

ATT performed in ER significantly increased after removal of the sMCL and the underlying dMCL at 0°, 30°, 60°, and 90° (p<0.015). Even though sMCL reconstruction was able to reduce ATT in ER, there was a persistent difference to native knee kinematics at all flexion angles (p<0.05). Despite a lack of statistical significance, removal of the AM resulted in increased ATT, which could not be restored by an AM reconstruction (p<0.05).
Similarly, isolated ER was significantly increased by removing the sMCL and dMCL at all flexion angles (p<0.001). sMCL reconstruction decreased isolated ER to values not significantly different to the native knee at 0° and 30° (NS). However, at higher flexion angles sMCL reconstruction was not able to restore ER (p<0.001). With AM resection, ER was not statistically significant but clinically relevant increased by >5° compared to the native knee. This could not be restored to normal when performing an AM reconstruction.
IR was restored with a POL reconstruction at 0°. At 30° and higher flexion angles, IR was significantly higher compared to the native knee (p<0.05).
Valgus rotation was significantly increased with MCL deficiency (p<0.001) and could be restored at 0° and 30° with sMCL reconstruction.

Conclusion

Data of this study indicate, that with an isolated sMCL reconstruction, AMRI cannot be fully restored. Even the anteromedial reconstruction with a Semitendinosus-tendon graft did not restore knee kinematics entirely. Thus, when facing high-grade AMRI, surgeons should be aware, that current techniques might not be sufficient to treat these entities.
From an anatomical perspective, one could speculate, that one explanation of these findings is the fact, that current techniques do not mimic the broad shape of the sMCL.