Biomechanical comparison of different graft positions for single-bundle anterior cruciate ligament reconstruction

Sixteen fresh-frozen cadaveric knees were used in the study. Each specimen was screened by a CT scan and a manual examination, and the existence of an intact ACL was confirmed by arthroscopy. Knees were excluded if evidence of any of the following was present: ligament injuries, previous knee surgery, osteoarthritis and bony abnormalities. All specimens were stored at −20 °C and thawed at room temperature 24 h before the test. All soft tissues until a distance of approximately 10 cm away from both sides of the joint sides were removed. The exposed femur and tibia were secured in an epoxy compound (Bondo, Atlanta, Georgia, USA) for mounting in custom-made aluminium clamps. The femoral side was rigidly fixed to a base, while the tibial side was mounted to the end-effector of the robot through a universal force-moment sensor (UFS), as shown in Fig. 1. The specimens were kept moist throughout the testing.

Four ACL reconstruction techniques were evaluated: (1) anatomic mid-bundle single-bundle reconstruction (MID–MID; n = 8), with the graft being placed from the midpoint between the tibial AM and PL footprints to the midpoint between the femoral AM and PL footprints; (2) anatomic AM bundle single-bundle reconstruction (AM–AM; n = 8), with the graft being placed from the tibial AM footprint to the femoral AM footprint; (3) anatomic PL bundle single-bundle reconstruction (PL–PL; n = 8), with the graft being placed from the tibial PL footprint to the femoral PL footprint; and (4) conventional single-bundle reconstruction (PL-high AM; n = 8), with the graft being placed from the tibial PL footprint to the femoral high AM position in the notch. Of the sixteen knees used, eight were used for MID–MID reconstructions, and the other eight knees were used for the remaining reconstructions (AM–AM, PL–PL and PL-high AM) with the three reconstructions being done in each knee (Fig. 2). There was no significant difference in the age or sex between the MID–MID reconstructed knee group and the multiple-reconstructed knee group (AM–AM, PL–PL and PL-high AM). The MID–MID reconstruction was performed in one group because it was difficult to place femoral tunnels at all positions (MID, PL, AM high AM) simultaneously; in particular, MID femoral tunnel creation between AM and PL femoral tunnels posed a risk of tunnel overlap.

A 3-portal technique was used with anterolateral, anteromedial and accessory medial portals with a 30° scope [13]. The two functional ACL bundles (AM and PL bundles) were differentiated by their tension patterns during knee motion. The ACL was removed with an electrothermal arthroscopy system (Vulcan, Smith and Nephew, Endoscopy, Andover, MA). After identifying the ACL foot print and removing the ACL, ACL reconstruction was performed. A guide wire (2.4 mm diameter) was inserted into the centre of each tibial ACL bundle footprint (AM and PL) from the anteromedial aspect of the tibia using a tibial drill guide system (Smith and Nephew Endoscopy, Andover, MA). The wire was then over drilled with a cannulated reamer (6 mm diameter). In 8 knees, three femoral tunnels (PL, AM and high AM) were created using a trans-portal technique [13]. A guide pin was inserted into the centre of the AM footprint, PL footprint and high AM position, which was located at the 11 or 1 o’clock position of the superoanterior portion of the ACL femoral footprint on the lateral wall of the intercondylar notch (Fig. 3a). The pin was then over drilled to the anterolateral femoral cortex by using a cannulated reamer (6 mm diameter). For the MID–MID reconstruction, the tibial and femoral tunnels were drilled between the centres of the AM and PL footprints of the tibia and femur in eight knees. Although one of the principles of an anatomic reconstruction is individualized surgery [28], 6 mm diameter bone tunnels were made for every reconstruction in this study to avoid AM and PL tunnel overlap (Fig. 3b). The knees were checked for the occurrence of tunnel overlap by CT scans after each test.

Previously harvested semitendinosus and gracilis tendons from human cadaver knees were used as grafts and were trimmed to a diameter of 6 mm. A No.5 braised polyester suture was whip stitched with a tapered needle into the free ends of the folded grafts. The graft was passed through the extra-cortical button loop (EndoButton, Smith and Nephew, Endoscopy, Andover, MA) to make it double stranded. At graft fixation, an initial tension of 60 N load was applied (ligament tension meter; Meira Corp, Nagoya, Aichi, Japan) [9] at 0° of knee flexion for the PL–PL reconstruction and at 30° of knee flexion for the other reconstructions (MID–MID, AM–AM and PL-high AM) [11, 16, 35]. Staples were used to fix the graft on the tibia.

The robotic/UFS testing system was used to determine knee joint kinematics and in situ forces of the ligament and the reconstruction graft [27, 32, 33, 34]. The robotic manipulator (CASPAR, OrthoMaquet, Rastatt, Germany) is a six-joint, serial articulation device that allows 6-degree-of-freedom motion of the knee with repeatability of 0.02 mm at each joint according to the manufacturer. The UFS (model 4015; JR3 Inc, Woodland, California, USA) is capable of measuring 3 orthogonal forces and moments with repeatability of 0.2 N and 0.1 Nm, respectively, according to the manufacturer. A custom MATLAB program with a multitask operating system (Math Works Inc., Natick, Massachusetts, USA) was utilized to control the testing system, to monitor knee kinematics and calculate the in situ forces of the ACL and the reconstructed grafts. During the experiment, this testing system was operated in both the force- and displacement-control modes.

The specimen was initially mounted to the testing system at full extension of the knee (measured with a goniometer). The path of passive flexion–extension of the intact knee was determined with the robotic/UFS testing system by moving the tibia from full extension to 90° of flexion by 0.5° increments. At each incremental flexion of knee flexion, the forces and moments generated by the specimen in the remaining 5-degree-of-freedom were minimized by the iterative algorithm of the robot control software. The positions at full extension and 15°, 30° and 60° of flexion were used as the starting positions for applying external tibial loads throughout the test. The following external loading conditions were applied to the tibia: (1) 89 N of anterior tibial load (simulated KT 1000 test) [36] at full extension and 15°, 30° and 60° of flexion and (2) a combined 7 Nm valgus torque and 5 Nm internal tibial rotation torque (simulated pivot-shift test) [20] at 0°, 15°, 30° and 45° of flexion. The anterior tibial load was as applied since the ACL is a major restraint to anterior tibial translation, and its application corresponds to the Lachman test and anterior drawer test. The force of 89 N is equivalent to that used in the KT-1000 arthrometer [10]. The combined valgus and internal tibial torque was chosen to simulate the pivot-shift test. While the external tibial loads were applied at each flexion angle, the 5-degree-of-freedom kinematics, forces, and moments of the intact knee were monitored. The ACL was transected arthroscopically, and the knee was tested with both modes of the testing system. Initially, the same loading conditions were employed in force control mode of the testing system to obtain kinematics of the ACL-deficient knee at each testing angle. Subsequently, in displacement-control mode, the identical path of motion of the intact knee was repeated in the ACL-deficient knee while new forces and moments of the specimen were recorded. By the principle of superposition, the vectorial difference in the measured forces between the intact and the ACL-deficient knee during the same path of motion gives the in situ force of the ACL [27, 33, 34].

All ACL reconstructions were performed sequentially. For the multiply reconstructed knees, the three ACL reconstructions, AM–AM, PL–PL and PL-high AM, were performed in random order. For the single reconstruction knees, only anatomic mid-bundle single-bundle reconstruction (MID–MID) was performed. The same graft (hamstring) and the same tibial fixation technique were used in all the reconstructions. External tibial loads were applied, and the kinematics of each reconstructed knee was analysed. After graft removal, the path of motion of the reconstructed knee at each testing angle was replayed while new forces and moments were monitored. The in situ force of the graft with each reconstruction technique was the vector difference in the measured forces between the reconstructed knee and the ACL-deficient knee with an identical path of motion.

Differences in ATT and in situ force at the different flexion angles were analysed using the Kruskal–Wallis test for comparison of all groups and the Mann–Whitney U test between all the pairwise comparisons. A Bonferroni approach was used to adjust the alpha level for the pairwise post hoc comparisons, and statistical significance was assumed when p < 0.05 for the Kruskal–Wallis test and p < 0.01 for the Mann–Whitney U test. Statistical analysis was performed using the software package SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). Pairwise comparisons were made between the data of all reconstruction methods and the intact ACL and between the AM–AM, PL–PL and MID–MID reconstructions and the PL-high AM reconstruction.