Abstract
Objectives:
Histological studies have shown that the ACL has a direct and indirect insertion on the femur [1]. The direct insertion is located along the lateral intercondylar ridge and the indirect insertion is located ‘lower’ on the lateral wall of the notch. The trend towards anatomic ACL reconstruction using the anteromedial (AM) portal technique has resulted in ‘lower’ non-isometric femoral tunnel positions and increased graft failures [2]. To our knowledge, the load transfer properties of the direct and indirect ACL insertions have not been studied. This information may help in understanding the increased failures reported with AM portal drilling. The purpose of this study was, 1) to compare the load transferred across the native ACL at the direct and indirect femoral insertions and, 2) to determine the strain behavior of ACL grafts placed at different tunnel locations within the direct and indirect insertions.
Methods:
Ten fresh-frozen cadaveric knees (mean age, 52.5 years; range, 29-65) were mounted to a six degree of freedom robot. A 134N anterior load at 30 and 90° flexion and a combined valgus (8Nm) and internal (4Nm) rotational moment at 15° flexion were applied. The ACL was subsequently sectioned at the femoral footprint by detaching either the direct or indirect insertion (partially sectioned state), followed by the remainder of the ACL (completely sectioned state) (Figure 1). The kinematics of the intact knee were replayed after each stage of sectioning to determine the loads transferred across the direct and indirect ACL fibers. Loads were expressed as a percentage of the total load borne by the ACL. Strain behaviour was tested by generating 3D models of the femur and tibia from CT scans of each knee. Three tunnel locations (anteromedial bundle [AM], center [C], posterolateral bundle [PL]) each were selected for the direct and indirect insertions and a virtual ACL graft was inserted. The isometry of the virtual graft was calculated through a flexion path of 0 to 90°.
Results:
Under an anterior tibial load at 30° flexion, the direct insertion carried 83.9% of the total ACL load compared to 16.1% in the indirect insertion (p<0.001). The direct insertion also carried more load at 90° flexion (95.2% vs 4.8%; p<0.001). Under a combined rotatory load at 15° flexion, the direct insertion carried 84.2% of the total ACL load compared to 15.8% in the indirect insertion (p<0.001). A virtual ACL graft placed at the AM position in the direct insertion demonstrated the best strain behaviour with a mean 10.9% change in length. This value was significantly lower (p<0.001) than the isometry at all 3 tunnel positions in the indirect insertion (AM = 18.5%; C = 24.9%; PL = 30.9%).
Conclusion:
Fibers in the direct insertion of the ACL carry more load than fibers in the indirect insertion. Virtual ACL grafts placed in the ‘higher’ direct location are more isometric than in the ‘lower’ indirect location during range of motion testing.
Clinical Relevance:
‘Low’ ACL grafts in the indirect ACL insertion, resulting from AM portal drilling techniques, may experience higher loads in-vivo due to unfavorable biomechanics. With the current shift towards anatomic ACL reconstruction, it may be beneficial to create a ‘higher’ femoral tunnel within the direct insertion at the lateral intercondylar ridge. This position remains anatomical but may also be biomechanically favorable.