Single– versus two–femoral socket anterior cruciate ligament reconstruction technique

Single– Versus Two–Femoral Socket Anterior Cruciate Ligament Reconstruction Technique: Biomechanical Analysis Using a Robotic Simulator Tatsuo Mae, M.D., Konsei Shino, M.D., Takahide Miyama, M.D., Hirotaka Shinjo, M.D., Takahiro Ochi, M.D., Hideki Yoshikawa, M.D., and Hiromichi Fujie, Ph.D.

Purpose: Although anterior cruciate ligament (ACL) reconstruction with multistrand autogenous hamstring tendons has been widely performed using a single femoral socket (SS), it is currently advocated to individually reconstruct 2 bundles of the ACL using 2 femoral sockets (TS). However, the difference in biomechanical characteristics between them is unknown. The objective of this study was to clarify their biomechanical differences. Type of Study: This is a cross-over trial using cadaveric knees. Methods: Seven intact human cadaveric knees were mounted in a robotic simulator developed in our laboratory. By applying anterior and posterior tibial load up to ⫾ 100 N at 0°, 15°, 30°, 60°, and 90° of flexion, tibial displacement and load were recorded. After cutting the ACL, the knees underwent ACL reconstruction using TS, followed by that using SS, with 44 or 88 N of initial grafts tension at 20° of flexion. The above-mentioned tests were performed on each reconstructed knee. Results: The tibial displacement in the TS technique was significantly smaller than that in the SS at smaller flexion angles in response to anterior and posterior tibial load of ⫾ 100 N, and the in situ force in the former was significantly greater than that in the latter at smaller flexion angles. Furthermore, in the TS technique, the posterolateral graft acted dominantly in extension, while the anteromedial graft mainly resisted against anterior tibial load in flexion. However, in the SS technique, the anteriorly located graft functioned more predominantly than the posteriorly located graft at all flexion angles. Conclusions: The ACL reconstruction via TS using quadrupled hamstring tendons provides better anterior-posterior stability compared with the conventional reconstruction using a single socket. Key Words: Anterior cruciate ligament—Hamstring tendon grafts—Two femoral sockets—Single femoral socket—Laxity—Robotic simulator.

A

nterior cruciate ligament (ACL) reconstruction using autogenous graft materials is a common procedure for restoring knee stability after the rupture of the ACL.1-3 The patellar tendon as well as the hamstring tendon is often used as the graft. Previous

From the Department of Orthopaedic Surgery, Osaka University Medical School, Suita; and the Department of Systems and Human Science, Graduate School of Engineering Science, Osaka University, Toyonaka (H.F.), Osaka, Japan. Address correspondence and reprint requests to Tatsuo Mae, M.D., Department of Orthopaedic Surgery, Osaka University Medical School, 2-2, Yamada-oka, Suita, Osaka 565-0871, Japan. E-mail: [email protected] © 2001 by the Arthroscopy Association of North America 0749-8063/01/1707-2849$35.00/0 doi:10.1053/jars.2001.25250

708

reports on the clinical outcome of the procedure indicated that ACL-reconstructed knees could be restored to nearly normal stability 2 years or more after the surgery.4-6 In addition, in vitro studies were also performed to evaluate the biomechanical characteristics of the knee that experimentally underwent the ACL reconstruction surgery. Markolf et al.7 reported that the laxity-matched initial tension of patellar tendon grafts in ACL reconstruction was 28.2 N, and that a greater force was generated in grafts than in the normal ACL under several external loads. Woo et al.8 found that the traditional ACL reconstruction using either a patellar tendon or hamstring tendon graft via a single femoral socket showed 104% to 129% of intact anterior laxity under 110 N of anterior load to the knee. They also indicated that forces carried by the

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 17, No 7 (September), 2001: pp 708 –716

SINGLE– VERSUS TWO–SOCKET ACL RECONSTRUCTION ACL grafts in response to 110 N of anterior load to the knee were 50% to 90% of those carried by the intact ACL, while the initial tension applied to the grafts was not controlled in their study. Ishibashi et al.9 found that knee stability after ACL reconstruction was sensitive to the graft fixation site at the tibia. Woo et al. and Ishibashi et al. used a robotic system to perform biomechanical tests on the knee that had been originally developed for determining the biomechanical characteristics of the natural ACL and posterior cruciate ligament.10-14 The system was capable of controlling the displacement of and the force/moment applied to the knee at 6 degrees of freedom (6-DOF) based on a mathematical description of knee kinematics and kinetics.15 Therefore, it was possible to apply single or combined loads to the knee while still allowing natural joint motion using the system. More importantly, it was possible to reproduce any 3-dimensional path of the knee motion, which allowed the determination of the forces in the ACL, its bundles, and the ACL grafts under the principle of superposition.16 Generally, the normal ACL can be divided into 2 main bundles: the anteromedial bundle (AMB) and the posterolateral bundle (PLB).17-20 The traditional ACL reconstruction has usually been performed using a single femoral socket (SS), which was aimed to mainly reconstruct the AMB, but recent innovations in surgical techniques and in arthroscopic instruments have allowed surgeons to separately reconstruct the AMB and the PLB using 2 femoral sockets (TS) using autogenous quadrupled hamstring tendon graft.21 It might be hypothesized that the TS technique is advantageous over the SS technique in terms of biomechanics, as shown by the clinical study in which the TS technique achieved a better outcome than the SS in restored anterior stability at follow-up of over 2 years after surgery.22 The objective of the present study was to determine their biomechanical differences using our newly developed robotic simulator.23-25

709

were separately harvested from the limb as graft materials and wrapped in gauze soaked with saline solution for later use. The ends of the femur and tibia were potted in cylindrical molds of acrylic resin (Ostron II; GC Corp, Tokyo, Japan). The fibula was cut to 5 cm in length from the tibiofibular joint junction and was fixed to the tibia using acrylic resin to maintain its anatomic position. The femoral and tibial cylinders were then secured to the clamps of the manipulator of the robotic simulator developed in our laboratory (Fig 1).23,25 This simulator consisted of a 6-DOF manipulator, servo-motor controllers, and a control computer. The 6-DOF manipulator consisted of a lower mechanism with a translational axis, and an upper mechanism with 2 translational and 3 rotational axes. The femoral clamp was fixed to the lower mechanism, while the tibial clamp was fixed to the upper mechanism by a universal force/moment sensor (UFS) (UFS-45A50-U760/N; JR3, Inc, Woodland, CA). Position and force controls allowed a hybrid 6-DOF

METHODS Seven intact human cadaveric knee joints were used. The mean age of the specimens was 75 years (range, 67 to 86 years). No ligamentous injury or significant degenerative joint disease was found. The tibia and femur were cut to 15 cm in length from the joint line, and the muscles including the quadriceps muscle–patella–patellar tendon and soft tissues were carefully removed from the knee, leaving the joint capsule intact. Semitendinosus and gracilis tendons

FIGURE 1. The robotic simulator consists of a 6-DOF manipulator, servo-motor controllers, and a control computer.

710

T. MAE ET AL.

control of the displacement and the force/moment at the knee to be achieved. Calculating the motion of the manipulator was repeated at a rate of 300 to 500 milliseconds. Control and data acquisitions were performed using a personal computer in the C-language programming environment with a multi-task operating system (CP/Q; IBM Japan, Kanagawa, Japan). Preliminary studies indicated that the manipulator had a positional accuracy of less than 120 nm under the application of 500 N to the clamp and had a clampto-clamp stiffness of more than 312 N/mm. And the force determined by the robotic system was also found to be the same as the force by direct measurement for the pilot study. Normal Knees (Group N) After the specimen was fixed to the simulator using locations of the femoral insertion sites of the medial and lateral collateral ligaments as positioning datum points,25 the simulator flexed the knee from 0° to 90° of flexion 5 times for preconditioning. The simulator then applied cyclic anterior and posterior (A-P) drawer load up to 100 N to the tibia at a rate of 0.7 mm/sec at flexion angles of 0°, 15°, 30°, 60°, and 90°, while the initial 5 cycles of the 3-dimensional path of knee motion and the 6-DOF forces/moments to the knee at the fifth cycle were recorded. After the ACL was transected near its tibial insertion site, the simulator reproduced the initial 5 cycles of the identical path of the intact knee motion at each flexion angle, while measuring the 6-DOF forces and moments to the knee at the fifth cycle. The force in the ACL in response to 100 N of anterior load was calculated from the difference of forces and moments between the intact and the ACL-transected knee under the principle of superposition. The A-P laxity, defined as anterior and posterior translation between ⫾ 100 N of anterior and posterior loads to the knee, was also determined from the fifth cycle of the 3-dimensional path of the intact knee motion.

eter according to the maximum diameter of the graft. A femoral socket 30 mm deep for the anteromedial graft (AMG) was created in an inside-out manner at the superior margin of the anteroproximal half of the ACL femoral footprint (11 o’clock in the right knee or 1 o’clock in the left knee) on the lateral wall of the intercondylar notch using a guide pin and an endoscopic cannulated reamer 6 mm in diameter. The pin was then further overdrilled to the anterolateral femoral cortex with a 4.5-mm diameter drill bit. The second femoral socket/tunnel of 4.5 mm diameter for the posterolateral graft (PLG), was created in the same fashion at the superoposterior margin of the posterodistal half of the ACL femoral footprint (9:30 o’clock in the right knee or 2:30 o’clock in the left). Extreme care was taken to leave 3 mm distance between the 2 femoral sockets to avoid their overlapping (Fig 2). Previously harvested semitendinosus and gracilis tendons were cut down to 14 cm in length and doubled over to form a quadrupled tendon graft. A No. 5 braided polyester suture on a tapered needle was whip-stitched in the free ends of the folded grafts, while a 6-mm polyester tape was passed through the loop and the EndoButtons (Acufex), and tied at the narrower part of the femoral tunnels. These doublelooped semitendinosus and gracilis tendons were separately prepared as the AMG and the PLG, respec-

ACL Reconstructed Using the TS Technique (Group TS) Reconstruction was performed in the same knees as in group N according to the technique described by Rosenberg et al.21,26 A Kirschner wire 2.4 mm in diameter was inserted as a guide pin from the anteromedial aspect of the tibia to the center of the tibial ACL attachment site using a tibial drill guide system (Acufex, Mansfield, MA). The pin was then overdrilled with a cannulated reamer 8 to 9 mm in diam-

FIGURE 2. Diagram showing the location of the femoral sockets. The oval with solid line on the intercondylar notch indicates the sockets in the TS technique. The oval with the dotted line indicates the socket in the SS technique. Note that the over-drilled socket in the SS technique never overlaps the lower one for PLG in TS technique.

SINGLE– VERSUS TWO–SOCKET ACL RECONSTRUCTION tively. The grafts were passed from the tibial tunnel to the femoral sockets, and the femoral side was fixed with 2 EndoButtons on the anterolateral femoral cortex. At this stage, the excursion of the graft throughout the entire range of motion was checked using a tension isometer (Isometric Positioner; Acufex), and it was found that all of the grafts showed over-the-top pattern or became elongated for 2 to 3 mm approaching full extension. Finally, the tibial sides of both bundled grafts were fixed to custom-made force gauges set at the anterior cortex of the tibia distal to the tibial tubercle (Fig 3). Meticulous care was taken to locate the PLG in the posterolateral portion of the intraarticular opening of the tibial tunnel. After a pretension of 22 or 44 N was applied to each graft (44 or 88 N in total) at 20° of flexion for 300 seconds in each test, the initial tension for each graft was set at 22 or 44 N (44 or 88 N in total) at 20° of flexion. During pretension and fixation, the robotic arm with the tibia was able to move under a control of UFS, allowing 5-DOF except for the flexion angle. The above-mentioned A-P drawer tests were performed, and the 5 cycles of the 3-dimensional path of the ACL-reconstructed knee and the 6-DOF forces/moments at the fifth cycle were recorded. After the A-P drawer tests of the ACL-reconstructed knee, the grafts were removed. The simulator then reproduced the 5 cycles of the identical path of the ACL-reconstructed knee while measuring the 6-DOF forces/moments to the graft-removed knee at the fifth cycle. The total force in the grafts in response to 100 N of anterior load was determined from the difference of forces and moments between the ACL-reconstructed and the graftremoved states under the principle of superposition.

711

Moreover, force sharing between grafts under 100 N to anterior load was determined by the output from the force gauges set at the tibial side. The A-P laxity was also determined from the fifth cycle of the 3-dimensional path of the ACL-reconstructed knee. ACL Reconstructed Using the SS Technique (Group SS) The femoral socket for the AMG used to analyze the TS technique was dilated to 8 to 9 mm in diameter for the SS technique.26 This socket did not overlap with the previously created femoral socket for the PLG because extreme care was taken to leave 3-mm distance between the 2 femoral sockets in the TS technique (Fig 2). The grafts were also prepared as separate bundles; the larger bundle was placed anterior and the smaller bundle was situated posterior in the bone tunnels or in the joint. The proximal ends of the grafts were fixed with 1 EndoButton at the femur, and the distal ends were secured to the force gauge with the initial tension of 22 or 44 N for each graft (44 or 88 N in total) at 20° of flexion at the tibia. The same test for determination was then performed. The reconstructions were performed by one of the senior authors (K.S.) who has had more than 20 years experience in knee surgery. Calculation of forces in the ACL and the grafts was based on a hypothesis that the force could be represented by a single resultant force and that the femur and tibia were much stiffer than soft tissues and grafts.27 A possible interaction between ACL grafts was also assumed to be negligible. Statistical analyses were performed using a repeated-measures analysis of variance to detect significant differences (P ⬍ .05). RESULTS Three-Dimensional Motion Under A-P Tibial Load

FIGURE 3. The custom-made force gauge has 2 arms, which enables separate measurement of the force of each graft.

The A-P laxity in group N was 8.2 ⫾ 1.8 mm at 0° of flexion, and increased to 10.5 ⫾ 1.1 mm at 15° of flexion, then decreased to 6.3 ⫾ 1.0 mm at 90° of flexion. When the initial tension was set at 44 N, the A-P laxity in group TS was significantly smaller than that in the other groups at 0° of flexion (TS/SS/N mean ⫾ SD; 6.8 ⫾ 2.5 mm / 7.9 ⫾ 2.5 mm / 8.2 ⫾ 1.8 mm). Similarly, the A-P laxity in group TS was significantly smaller at 15° and 30° of flexion (TS/SS; 15°: 8.5 ⫾ 2.0 mm / 9.9 ⫾ 2.1 mm, 30°: 8.7 ⫾ 1.5 mm / 9.4 ⫾ 1.7 mm), while no significant differences were found

712

T. MAE ET AL. 6.2 ⫾ 2.6 mm, 15°: 6.1 ⫾ 2.3 mm / 6.9 ⫾ 2.3 mm, 30°: 6.1 ⫾ 1.7 mm / 6.9 ⫾ 1.9 mm). At 90° of flexion, no significant difference was found in A-P laxity between group TS and group SS (Fig 4B). Simultaneously, the tibia in the reconstructed groups of TS or SS rotated more externally than that in group N under anterior tibial load of 100 N except at 0° of flexion. With the initial tension of 44 N, the coupled external tibial rotation was ⫺ 0.1° ⫾ 3.3° at 0°, 6.1° ⫾ 7.9° at 30°, or 3.2° ⫾ 8.0° at 90° of flexion in response to 100 N of anterior load in group TS, whereas that in group SS was ⫺0.2° ⫾ 3.2° at 0°, 4.9° ⫾ 9.0° at 30°, or 1.9° ⫾ 8.6° at 90° of flexion (Table 1). There was no statistically significant difference in coupled tibial rotation between the 2 groups. Similarly, no significant difference was found in coupled tibial rotation between group TS and SS when the graft initial tension was set at 88 N. The tibia in the reconstructed groups also showed more varus rotational motion than that in group N in combination with anterior tibial movement (Table 2). No significant difference was found in coupled varus/ valgus rotation between group TS and SS at each initial tension. In Situ Force

FIGURE 4. A-P laxity at initial tension of (A) 44 N or (B) 88 N under ⫾ 100 N of anterior-posterior tibial force.

at 90° (Fig 4A). When the initial tension of the graft was set at 88 N, the A-P laxity in group TS was also significantly smaller than those in the other groups at 0°, 15°, and 30° of flexion (TS/SS; 0°: 5.6 ⫾ 2.6 mm /

The force carried by the normal ACL (group N) in response to 100 N of anterior load was 107 ⫾ 13 N at 0° of flexion, and increased to 112 ⫾ 8 N at 15° of flexion, then decreased to 95 ⫾ 5 N at 90° of flexion. The graft force with the graft’s initial tension of 44 N was significantly greater in group TS than in the other groups at 0° (TS/SS/N; 132 ⫾ 13 N / 122 ⫾ 6 N / 107 ⫾ 13 N), while there were no differences at or beyond 30° of flexion. At 15° of flexion, there was no significant difference in the graft force between group TS (131 ⫾ 9 N) and group SS (125 ⫾ 4 N), whereas the force in group N (112 ⫾ 8 N) was significantly smaller than that in the other groups (Fig 5A). Simi-

TABLE 1. Coupled External Tibial Rotation (Reconstructed-Normal) (mean ⫾ SD) 44 N Flexion Angle

Group TS

0° 15° 30° 60° 90°

⫺0.1° ⫾ 3.3° 4.1° ⫾ 7.3° 6.1° ⫾ 7.9° 5.3° ⫾ 9.9° 3.2° ⫾ 8.0°

88 N Group SS ⫺0.2° ⫾ 3.2° 3.5° ⫾ 7.5° 4.9° ⫾ 9.0° 4.9° ⫾ 11.9° 1.9° ⫾ 8.6°

Group TS ⫺1.3° ⫾ 3.7° 1.1° ⫾ 8.0° 3.0° ⫾ 9.8° 1.4° ⫾ 11.1° 1.5° ⫾ 8.6°

Group SS 1.1° ⫾ 3.4° 2.1° ⫾ 8.1° 3.9° ⫾ 9.8° 1.4° ⫾ 12.6° 3.4° ⫾ 9.2°

NOTE. No significant differences were found between the 2 groups at each initial tension.

SINGLE– VERSUS TWO–SOCKET ACL RECONSTRUCTION

713

TABLE 2. Coupled Varus Rotation (Reconstructed-Normal) (mean ⫾ SD) 44 N

88 N

Flexion Angle

Group TS

Group SS

Group TS

Group SS

0° 15° 30° 60° 90°

0.9° ⫾ 0.7° 1.1° ⫾ 1.3° 2.1° ⫾ 1.6° 3.5° ⫾ 2.2° 3.1° ⫾ 2.5°

0.4° ⫾ 0.4° 0.9° ⫾ 1.0° 1.9° ⫾ 1.6° 3.5° ⫾ 2.5° 3.2° ⫾ 2.8°

0.8° ⫾ 0.5° 1.1° ⫾ 1.3° 1.8° ⫾ 1.9° 3.3° ⫾ 2.7° 2.9° ⫾ 2.7°

0.5° ⫾ 0.2° 0.9° ⫾ 1.0° 1.7° ⫾ 1.6° 3.2° ⫾ 2.7° 2.7° ⫾ 2.7°

NOTE. No significant differences were found between the 2 groups at each initial tension.

larly, when the initial tension of the graft was set at 88 N, the graft force in group TS was significantly greater than that in the other groups at 0° of flexion (TS/SS/N; 178 ⫾ 9 N / 163 ⫾ 5 N / 107 ⫾ 13 N) (Fig 5B). At

flexion angles of 15° or more, no significant difference in the graft force was found between group TS and SS, while the force in group N was significantly smaller than that in the other groups. Force Sharing in the Grafts With regard to the force sharing between the 2 grafts: in group TS under the anterior tibial load of 100 N with the initial tension of 44 N, the percentage of force sharing by AMG increased with flexion angle from 48.4% ⫾ 4.5% at 0° to 56.4% ⫾ 6.8% at 90° of flexion, and that by PLG decreased with flexion angle from 51.6% ⫾ 4.5% at 0° to 43.6% ⫾ 6.8% at 90° of flexion (Fig 6A). The pattern of force sharing at initial tension of 88 N was the same as that in case of 44 N, and significant differences between the 2 bundles were found at 60° and 90° of flexion (AMG/PLG: 60°, 54.3% ⫾ 2.3% / 45.7% ⫾ 2.3%; 90°, 59.2% ⫾ 5.9% / 40.8% ⫾ 5.9%) (Fig 6B). In group SS, the force in the anteriorly located graft (AG) was significantly larger than that in the posteriorly located graft (PG) at all flexion angles when the initial tension was set at 44 N (AG/PG: 0°, 62.4% ⫾ 1.8% / 37.6% ⫾ 1.8%; 30°, 61.3% ⫾ 2.1% / 38.7% ⫾ 2.1%; 90°, 56.3% ⫾ 4.2% / 43.7% ⫾ 4.2%) (Fig 6C), and at 88 N (AG/PG: 0°, 63.5% ⫾ 2.5% / 36.5% ⫾ 2.5%; 30°, 61.4% ⫾ 2.5% / 38.6% ⫾ 2.5%; 90°, 57.9% ⫾ 3.4% / 42.1% ⫾ 3.4%) (Fig 6D). DISCUSSION

FIGURE 5. In situ force in response to 100 N of anterior load with initial tension of (A) 44 N or (B) 88 N.

The TS reconstruction could be assumed to be biologically advantageous over the SS technique because of its greater graft-bone contact area leading to potentially earlier graft-bone healing. Biomechanically, this is the first study to show that the TS technique provides better A-P stability in response to the A-P load than the SS technique, not at larger but smaller flexion angles. This may be attributable to the

714

T. MAE ET AL.

FIGURE 6. Graphs showing force sharing between grafts under 100 N of anterior load with initial tension of (A) 44 N or (B) 88 N in group TS, and with initial tension of (C) 44 N or (D) 88 N in group SS.

following differences. First, the anterior walls of the femoral sockets were located slightly more posteriorly in the TS technique than in the SS, because of smaller diameter of the femoral sockets in the former. Because a thicker graft is subjected to greater deformation due to creep phenomenon, the exact center of the grafts in the SS technique is prone to shift more anteriorly. Second, both the AMG and PLG were directly in contact with the anterior edge of the femoral sockets in the TS technique, whereas only the anteriorly located part of the graft was closely in touch with the anterior wall of the femoral socket in the SS technique. This suggests that the former is advantageous over the latter in terms of resisting anterior drawer force. However, it should be noted that this advantage becomes negligible at greater flexion angles, because

the alignment of the grafts becomes similar between both techniques with increased flexion angle. In this study, no significant difference in A-P laxity was found between group N and group SS with the initial tension at 44 N. Woo et al.8 also found no significant difference in A-P laxity between normal knees and SS ACL-reconstructed knees at the initial tension of 45 N applied to the ACL grafts. Therefore, the laxity-matched initial tension in the SS technique using quadrupled hamstring tendons could be assumed to be around 44 N. Considering that A-P laxity with the initial tension of 44 N in group TS was smaller than that in group N at all flexion angles including 0° and 15° of flexion, an initial tension smaller than 44 N might be enough to restore normal stability in the TS technique. Because the smaller initial tension could be

SINGLE– VERSUS TWO–SOCKET ACL RECONSTRUCTION assumed to impose less stress on the graft or its fixation sites, the TS reconstruction could be regarded as a more efficient way to stabilize the knee. It was also found that the in situ force with the initial tension of 44 N in group TS, in response to the anterior tibial load, was greater than that in the other groups, including group N at smaller flexion angles. This suggests that, in group TS, the anterior load applied to the tibia was more efficiently transmitted to the grafts, and that initial tension smaller than 44 N might be sufficient to restore normal stability. There was no previous study about the force sharing between the 2 bundles in the TS technique. The present study has first shown that the force in the PLG was greater than that in the AMG at small flexion angles, and that the force in the AMG was larger than that in the PLG at larger flexion angles under the anterior load. This pattern of force sharing between the 2 grafts in the TS technique was similar to that between AMB and PLB in the normal ACLs (Fig 7).24,27-29 On the contrary, the force in the AG was always larger than that in the PG at all flexion angles in the SS technique. This suggests that the TS reconstruction is superior to the SS in terms of force sharing as well. In this experiment, the same knees were used for both techniques to minimize variability in socket position or specimen quality, including soft tissues and bone, as in previously published studies.30,31 It might be a concern that the TS technique was always tested before the SS technique, which required the socket of larger diameter. The data obtained in the later exper-

FIGURE 7. Graph showing force sharing between the normal ACL bundles under 100 N of anterior load.

715

iment could be almost equally evaluated as those in the earlier one for the following reasons: (1) the load shared by the anteromedial socket in the TS technique was much smaller than that shared by the expanded one in the SS technique, (2) the second posterolateral socket was carefully created in the exact position leaving 3 mm distance to the first one to avoid overlapping of the 2 sockets for the later experiment, and (3) the specimens were kept moist to minimize biomechanical attenuation during the experiments (Fig 2). It is well known that the function of the ACL is to resist against not only anterior tibial load but also rotational loads such as varus-valgus or internal-external moments. Therefore, further studies are needed to determine exactly the biomechanical behavior of ACL-reconstructed knee in response to various external loading conditions. Moreover, the other factors, including stress relaxation and biologic remodeling following graft implantation, still remain major issues to be studied. In conclusion, ACL reconstruction via 2 femoral sockets using quadrupled hamstring tendons more efficiently stabilizes the knee than the conventional technique using a single socket. REFERENCES 1. Johnson RJ, Beynnon BD, Nichols CE, Renstrom PA. The treatment of injuries of the anterior cruciate ligament. J Bone Joint Surg Am 1992;74:140-151. 2. Sgaglione NA, Del PW, Fox JM, Friedman MJ. Arthroscopically assisted anterior cruciate ligament reconstruction with the pes anserine tendons. Comparison of results in acute and chronic ligament deficiency. Am J Sports Med 1993;21:249256. 3. Karlson JA, Steiner ME, Brown CH, Johnston J. Anterior cruciate ligament reconstruction using gracilis and semitendinosus tendons. Comparison of through-the-condyle and overthe-top graft placements. Am J Sports Med 1994;22:659-666. 4. Bach BR, Jones GT, Sweet FA, Hager CA. Arthroscopyassisted anterior cruciate ligament reconstruction using patellar tendon substitution. Two- to four-year follow-up results. Am J Sports Med 1994;22:758-767. 5. Otero AL, Hutcheson L. A comparison of the double semitendinosus/gracilis and central third of the patellar tendon autografts in arthroscopic anterior cruciate ligament reconstruction. Arthroscopy 1993;9:143-148. 6. Maeda A, Shino K, Horibe S, Nakata K, Buccafusca G. Anterior cruciate ligament reconstruction with multistranded autogenous semitendinosus tendon. Am J Sports Med 1996;24: 504-509. 7. Markolf KL, Burchfield DM, Shapiro MM, Davis BR, Finerman GAM, Slauterbeck JL. Biomechanical consequences of replacement of the anterior cruciate ligament with a patellar ligament allograft. Part 1,2. J Bone Joint Surg Am 1996;78: 1720-1734. 8. Woo SL-Y, Fox RJ, Sakane M, Livesay GA, Rudy TW, Runco TJ, Li G, Allen CR, Fu FH. Force and force distribution in the

716

9.

10.

11.

12.

13.

14.

15. 16.

17. 18. 19. 20.

T. MAE ET AL.

anterior cruciate ligament and its clinical implications. First place winner of the inaugural GOTS Beiersdorf Research Award for Sports Medicine. Sportorthopadic-Sporrtaumatologie 1997;13:37-48. Ishibashi Y, Kim HS, Rudy TW, Livesay GA, Xerogeanes JW, Fu FH, Woo SL-Y. The effect of the ACL graft fixation site at the tibia on knee stability: A robotic evaluation. Arthroscopy 1997;13:177-182. Fujie H, Mabuchi K, Woo SL-Y, Livesay GA, Arai S, Tsukamoto Y. The use of robotics technology to study human joint kinematics: A new methodology. J Biomech Eng 1993;115: 211-217. Livesay GA, Rudy TW, Woo SL-Y, Runco TJ, Sakane M, Li G, Fu FH. Evaluation of the effect of joint constraints on the in-situ force distribution in the anterior cruciate ligament. J Orthop Res 1997;15:278-284. Xerogeanes JW, Fox RJ, Takeda Y, Kim HS, Ishibashi Y, Carlin GJ, Woo SL-Y. A functional comparison of animal anterior cruciate ligament models to the human anterior cruciate ligament. Ann Biomed Eng 1998;26:345-352. Fox RJ, Harnar CD, Sakane M, Carlin GJ, Woo SL-Y. Determination of the in situ forces in the human posterior cruciate ligament using robotic technology. A cadaveric study. Am J Sports Med 1998;26:395-401. Sakane M, Livesay GA, Fox RJ, Rudy TW, Runco TJ, Woo SL-Y. Relative contribution of the ACL, MCL, and bony contact to the anterior stability of the knee. Knee Surg Sports Traumatol Arthrosc 1999;7:93-97. Fujie H, Livesay GA, Fujita M, Woo SL-Y. Forces and moments in six-DOF at the human joint: Mathematical description for control. J Biomech 1996;29:1577-1585. Fujie H, Livesay GA, Woo SL-Y, Kashiwaguchi S, Blomstrom G. The use of a universal force-moment sensor to determine in-situ forces in ligaments: A new methodology. J Biomech Eng 1995;117:1-7. Arnoczky SP. Anatomy of the anterior cruciate ligament. Clin Orthop 1983;172:19-25. Girgis FG, Marshall JL, Monajem A. The cruciate ligaments of the knee joint. Anatomical, functional and experimental analysis. Clin Orthop 1975;106:216-231. Odensten M, Gillquist J. Functional anatomy of the anterior cruciate ligament and a rationale for reconstruction. J Bone Joint Surg Am 1985;67:257-262. Harner CD, Baek GH, Vogrin TM, Carlin GJ, Kashiwaguchi

21. 22.

23.

24.

25. 26. 27.

28.

29. 30. 31.

S, Woo SL. Quantitative analysis of human cruciate ligament insertions. Arthroscopy 1999;15:741-749. Rosenberg TD, Graf B. Techniques for ACL reconstruction with Multi-Trac drill guide. Mansfield, MA: Acufex Microsurgical, 1994. Hamada M, Shino K, Horibe H, Mitsuoka T, Miyama T, Shiozaki Y. Single-socket versus bi-socket ACL reconstruction using autogenous multiplied hamstring tendons and EndoButton femoral fixation: a prospective randomized study. Presented at the 8th Congress of the European Society of Sports Traumatology, Knee Surgery, and Arthroscopy, Nice, France, April 29-May 2, 1998. Sekito T, Fujie H, Ota Y, Hayashi K. Development of a novel robotic simulator for the analysis of knee mechanical function. Proceedings of the 1997 Bioengineering Conference of the American Society of Mechanical Engineers, 1997;35:393-394. Fujie H, Mae T, Sekito T, Shino K. Mechanical functions of human ACL bundles: Development and application of a robotic knee simulator. Proceedings of the Tokyo International Symposium on Human Life Support Biomechanics, 2000;255260. Fujie H, Sekito T. Development of a novel robotic simulator for the human knee joint. J Biomech Eng (in press). Rosenberg TD, Brown GC, Deffner KT. Anterior cruciate ligament reconstruction with a quadrupled semitendinosus autograft. Sports Med Arthrosc Rev 1997;5:51-58. Amis AA, Dawkins GPC. Functional anatomy of the anterior cruciate ligament. Fiber bundle actions related to ligament replacements and injuries. J Bone Joint Surg Br 1991;73:261267. Sakane M, Fox RJ, Woo SL-Y, Livesay GA, Li G, Fu FH. In situ forces in the anterior cruciate ligament and its bundles in response to anterior tibial loads. J Orthop Res 1997;15:285293. Takai S, Woo SL-Y, Livesay GA, Adams DJ, Fu FH. Determination of the in situ load on the human anterior cruciate ligament. J Orthop Res 1993;11:686-695. Race A, Amis AA. PCL reconstruction. In vitro biomechanical comparison of ‘isometric’ versus single and double-bundled ‘anatomic’ grafts. J Bone Joint Surg Br 1998;80:173-179. Harner CD, Janaushek MA, Kanamori A, Yagi M, Vogrin TM, Woo SL-Y. Biomechanical analysis of a double-bundle posterior cruciate ligament reconstruction. Am J Sports Med 2000; 28:144-151.