One area of concern for patients who have a reconstructed anterior cruciate ligament is the potentially large forces that develop in the graft as the knee is loaded during daily activities, rehabilitation, and participation in sports. Large forces in the graft during the early postoperative period may jeopardize the initial fixation of the graft or have biological consequences related to remodeling and maturation of the graft. To date, most studies related to forces in substitutes for the anterior cruciate ligament have involved synthetic ligaments implanted in specimens from cadavera1,2,5-7. In those studies, the forces in the graft were measured during straight extension of the knee or application of a straight anterior tibial force. We found no studies in which the force in a graft at laxity-matched pre-tension (the level of pre-tension that produces normal anterior-posterior laxity at 30 degrees of flexion) was directly compared with the force in the intact anterior cruciate ligament. The cruciate load-cell technique that we developed allows a comparison of the force in the graft with that in the intact anterior cruciate ligament for each loading experiment, with each knee used as its own control. In the present series, we subjected knee specimens to constant anterior tibial force, varus or valgus tibial bending moment, and internal or external tibial torque as the femur was extended from 90 degrees of flexion to 5 degrees of hyperextension. The load-cell measured the resultant force in the intact anterior cruciate ligament or in the graft during these loading tests.
The objectives of this study were to compare the magnitudes of forces that were generated in correctly pre-tensioned and over-tensioned patellar ligament allografts with those in intact anterior cruciate ligaments and to measure the effects of a graft on tibial rotation during passive extension of the knee (the so-called screw home mechanism).
Materials and Methods
Thirteen specimens were prepared for testing, and the anterior cruciate ligament load-cell was installed, as described in Part I of this study4. Four additional specimens that had unacceptable placement of the graft were analyzed separately. Not all of the constant loading tests were completed for all of the specimens. In some, the cap of bone containing the proximal attachment of the anterior cruciate ligament fractured, while in others, a collateral ligament tore after testing of the specimen had been partially completed.
The femur was placed in the same fixture used for the anterior-posterior tests described in Part I of this study4, with one important difference: the distal end of the tibia was supported by a roller mounted on the tibial extension shaft. The top and bottom surfaces of the roller contacted parallel, horizontal flat plates. This arrangement maintained the tibia in a level position as the femur was extended. Care was taken to mount the femur on the flexion bar so that the tibia demonstrated minimum varus-valgus angulation and had a minimum tendency to piston in the proximal-distal direction when the knee was moved through a range of motion. During extension of the knee, the tibia was allowed to rotate about its long axis and free varus-valgus angulation of the tibia was allowed in the horizontal plane.
For the passive knee-extension tests with a constant anterior tibial force, the cylinder of polymethylmethacrylate in which the tibia was potted passed through a large bearing mounted in a housing. A 100-newton weight that acted at the joint line was suspended from the horizontal undercarriage bar integral with this housing (Fig. 1).
For the flexion tests with a constant internal or external tibial torque, a weight was suspended from a long torque bar clamped in a horizontal position on the tibial extension shaft (Fig. 2). This produced a constant tibial torque (of ten newton-meters) about the axis of the polymethylmethacrylate cylinder in which the tibia was potted. Varus-valgus angulation of the tibia was permitted in the horizontal plane during the constant tibial torque test.
For the flexion tests with a constant varus or valgus moment, a cord looped around the roller passed over a pulley and suspended a weight that produced a medial or lateral force on the tibial extension shaft (Fig. 2). This force, acting at a right angle to the tibia, produced a constant varus or valgus bending moment of ten newton-meters in the horizontal plane while tibial rotation was unconstrained.
Each test began with the femur flexed to 90 degrees (with the tibia level). The femur was manually extended at approximately 10 degrees per second until the knee was hyperextended 5 degrees. The resultant force in the anterior cruciate ligament or in the graft, the tibial rotation, and the angle of flexion of the knee were recorded continuously.
The specimen was mounted in the test apparatus, and the constant tibial loading tests just described were performed for the intact anterior cruciate ligament. The ligament was replaced with a patellar ligament graft, and the laxity-matched pre-tension was determined. The knee-extension tests under constant load were repeated at laxity-matched pre-tension. The graft was then pre-tensioned to forty-five newtons more than the laxity-matched pre-tension (the over-tensioned condition), and the test series was repeated a third time.
Two-way analysis of variance with repeated measures was used to determine the significance of differences between the mean force in the intact anterior cruciate ligament and that in the graft for a given loading test. Pairwise comparisons between the means were made at specific angles of knee flexion with the Student-Newman-Keuls test. A similar model was used to analyze the differences between the means for tibial rotation before and after insertion of the graft. The level of significance was p ≤ 0.05.
Forces in the Graft
Passive Extension of the Knee
The level of force in the intact anterior cruciate ligament during passive extension of the knee remained low until the last 10 degrees of extension (Fig. 3). The force in the graft was minimum near 40 degrees of flexion and increased as the knee was flexed or extended from this position. The mean force in the graft at laxity-matched pre-tension was significantly greater (p ≤ 0.05) than the corresponding mean force in the intact anterior cruciate ligament at all positions of flexion (Fig. 3). The mean force in the over-tensioned graft was significantly greater (p ≤ 0.05) than the corresponding mean force in the graft at laxity-matched pre-tension for all angles of flexion. At full extension, the force in the intact anterior cruciate ligament ranged from eighteen to 110 newtons. In contrast, the force in the graft at full extension ranged from fifty to 297 newtons at laxity-matched pre-tension and from 145 to 467 newtons in the over-tensioned condition.
Elevated forces in the graft were recorded between 30 and 90 degrees of flexion during the passive knee-extension test in the first four specimens tested (Fig. 4). At 90 degrees of flexion, the forces in the graft ranged from 100 to 190 newtons. These levels of force were considered unacceptable, and the four specimens were not included in the main analysis. This force profile indicated tightening of the graft with flexion of the knee and most likely resulted from placement of the femoral tunnel too far anteriorly. Beyond 20 degrees of flexion, the mean forces in the four poorly placed grafts were significantly greater (p ≤ 0.05) than the corresponding mean forces in the remaining thirteen grafts (Fig. 4). The mean laxity-matched pre-tension (and standard deviation) for the poorly placed grafts (102.3 ± 37.2 newtons) was significantly greater (p ≤ 0.05) than that for the so-called acceptably placed grafts (28.2 ± 16.8 newtons).
Constant Loading Tests
When a constant anterior force of 100 newtons was applied to the tibia as the knee was extended, the mean forces in the graft were significantly greater (p ≤ 0.05) than the corresponding mean forces in the intact anterior cruciate ligament (Fig. 5). Over-tensioning of the graft by forty-five newtons significantly increased the mean force in the graft at all angles of flexion (p ≤ 0.05).
The mean force in the graft was significantly greater than that in the intact anterior cruciate ligament (p ≤ 0.05) at all angles of flexion when constant internal or external torque was applied to the tibia during extension of the knee (Figs. 6 and 7). Over-tensioning of the graft did not significantly increase the mean force generated in the graft by the tibial torque at any angle of flexion.
The mean force in the graft was significantly greater (p ≤ 0.05) than that in the intact anterior cruciate ligament at all angles of flexion when constant valgus or varus moment was applied to the tibia during extension of the knee (Figs. 8 and 9). Over-tensioning of the graft significantly increased the mean force in the graft at all angles of flexion (p ≤ 0.05) for applied varus and valgus moment. With the exception of full extension and hyperextension, the increase in the mean force in the graft for both loading directions was approximately equal to the increase in the pre-tensioning of the graft (forty-five newtons).
Tibial Rotation during Passive Extension of the Knee
As the intact knee was extended passively from 90 degrees to full extension, the tibia rotated externally a mean of 9 degrees; there was no significant change in tibial rotation as a function of the angle of flexion after insertion of the graft.
Limitations of the Study
Our experimental apparatus was designed to apply a constant load to the tibia in an unconstrained fashion during passive extension of the knee. The advantage of this approach was uniform loading of the knee. As muscle forces were not simulated during these tests, our data represent knee-loading that is resisted by only ligamentous restraints and the contact surfaces of the joint. The varus-valgus moment and the levels of tibial torque (ten newton-meters) applied to the knees represent substantial in vitro loading. In vivo, the stabilizing effects of the muscles about the knee act to control medial-lateral opening of the joint and tibial rotation. Therefore, one would expect less force to develop in the graft from these loading modes when the muscles of the knee are active. However, high quadriceps activity would increase loading of the graft generated by the anterior component of the patellar tension force because forces in the graft from this loading mechanism could be much higher than the 100-newton level used for these experiments.
The maximum force that we were able to record in the intact anterior cruciate ligament was limited by the fixation strength of the cap of bone, which in turn was limited by the quality of the cancellous bone on its undersurface. For this reason, only specimens from men were used in this study; osteoporotic changes in the bone of specimens from women made them unsuitable. It was somewhat remarkable that, in certain instances, forces in the anterior cruciate ligament of approximately 350 newtons were recorded without failure of the cap of bone. The maximum sustainable forces in the graft were somewhat larger since the grafts were obtained from younger individuals. A maximum force of 425 newtons was recorded in one graft without failure of the fixation.
Even though the edge of the tunnel hole was well rounded, abraded areas of the graft were observed in this region. It was our impression that if the sharp edge of the hole had not been rounded, the graft would have eventually worn through at this point.
Because of the occasional mechanical failure in the cap of bone on the grafts and the failure of a collateral ligament in some knees, not all tests were completed for all specimens. Thus, the repeated-measures analysis of variance for the intact anterior cruciate ligament compared with the graft at laxity-matched pre-tension sometimes included a different number of specimens than did the analysis comparing the graft at laxity-matched pre-tension with the over-tensioned graft. Strictly speaking, we should have presented two curves for laxity-matched pre-tension, one for each number of specimens (Figs. 3 and 5, 6, 7, 8 and9). However, these two mean curves were virtually indistinguishable from one another, and we elected to present the curve with the larger number of specimens (in other words, the mean curve used for the analysis of the intact anterior cruciate ligament compared with the graft at laxity-matched pre-tension).
Forces in the Graft
Passive Extension of the Knee
The straight, unloaded passive knee-extension test is of particular clinical interest because it is performed routinely while the patient is on the table at the time of an operation. No appreciable force was generated in the intact anterior cruciate ligament at more than 10 degrees of flexion. In contrast, loading of the graft occurred at less than 30 degrees of flexion. Although we took care when determining the site for the femoral tunnel, the force in the graft was still elevated slightly between 40 and 90 degrees of flexion, the position at which the force in the intact anterior cruciate ligament was nominally zero. Thus, even with our best efforts, a perfect force in the graft profile was not achieved.
As expected, over-tensioning of the graft increased the force in the graft during passive extension of the knee; between 30 and 90 degrees of flexion, the increase in force was approximately equal to the amount of over-tensioning of the graft (forty-five newtons). Increasing the initial tension in the graft by forty-five newtons at 30 degrees of flexion produced a mean increase of 120 newtons in the force at full extension. This apparent magnification effect could be the result of the more highly pre-tensioned graft being loaded on a stiffer region of its stress-strain curve.
Constant Loading Tests
Anterior tibial force is possibly the most direct loading condition for the generation of force in the anterior cruciate ligament, as at 30 degrees of flexion the anterior cruciate ligament is the primary restraint to anterior translation of the tibia. It is thought that in many sports-related injuries the anterior cruciate ligament is ruptured by inertial forces related to sudden changes in forward acceleration of the lower limb. It is important to bear in mind that forces in a graft under the same loading conditions could be greater than those in the intact anterior cruciate ligament in the critical 0-to-30-degree range of flexion. The forces in an over-tensioned graft increase even further (by approximately the amount of increased pre-tension in the graft). Clearly, sudden accelerations or decelerations may pose a substantial risk to a graft that has not matured fully.
As we reported previously3, varus moment generated a greater force in the anterior cruciate ligament than did valgus moment. This was also true after insertion of the graft. Varus moment is the more important of the two loading directions, especially in the 40-to-90-degree range of flexion. In this region, large forces were generated in the graft whereas forces were normally negligible in the intact anterior cruciate ligament, and forces were even greater in the over-tensioned graft.
We also reported previously that internal tibial torque generates greater force in the intact anterior cruciate ligament than does external torque3. Again, this pattern was the same for the graft. The standard deviations for forces in the intact anterior cruciate ligament and in the graft were greater for internal torque than for any other loading state. The reasons for this increased scatter are unknown. In the present study, the greatest forces in the intact anterior cruciate ligament and in the graft were recorded for knees that were subjected to internal torque at full extension. Therefore, this loading condition should be avoided after insertion of a graft. One could argue that a loading state of external tibial torque is also of concern after insertion of a graft because forces in the graft are elevated in the 20-to-90-degree range of flexion, whereas forces in the intact anterior cruciate ligament are essentially zero. Although it is reasonable to expect that over-tensioning of the graft would increase the force generated in the graft by tibial torque, excessive scatter in the data prevented demonstration of any significant effect of over-tensioning for either direction of applied torque.
Tibial Rotation during Passive Extension of the Knee
An interesting finding in this study was the elimination of the so-called passive screw home mechanism of the knee after insertion of the graft. This result could be attributed to geometric differences between the cord-like graft tethered within a tunnel and the complex multi-banded intact ligament, which is anchored securely to bone over a broad femoral insertion site. The ability of the graft to slide freely around the perimeter of the hole for the tunnel during tibial rotation may affect its ability to control this motion.
NOTE: The authors thank Mike Shepard and Steve Jackson for their assistance in the testing and analysis of the data, and Fred Dorey, Ph.D., for his consultation and advice regarding the statistical analysis. The patellar ligament grafts used for this study were donated by the Musculoskeletal Transplant Foundation.
*Although none of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article, benefits have been or will be received but are directed solely to a research fund, foundation, educational institution, or other non-profit organization with which one or more of the authors are associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding sources were Grant RO1 AR40330 from the National Institutes of Health and grants from the Orthopaedic Research and Education Fund and the Dorothy and Leonard Straus Fund.
Investigation performed at the Department of Orthopaedic Surgery, Biomechanics Research Section, University of California at Los Angeles, Los Angeles
1. Fleming, B.; Beynnon, B. D.; Johnson, R. J.; McLeod, W. D.; and |and |Pope, M. H.: Isometric versus tension measurements. A comparison for the reconstruction of the anterior cruciate ligament. Am. J. Sports Med.
, 21: 82-88, 1993.
2. Lewis, J. L.; Lew, W. D.; Engebretsen, L.; Hunter, R. E.; and |and |Kowalczyk, C.: Factors affecting graft force in surgical reconstruction of the anterior cruciate ligament. J. Orthop. Res.
, 8: 514-521, 1990.
3. Markolf, K. L.; Gorek, J. F.; Kabo, J. M.; and |and |Shapiro, M. S.: Direct measurement of resultant forces in the anterior cruciate ligament. An in vitro study performed with a new experimental technique. J. Bone and Joint Surg.
, 72-A: 557-567, April 1990.
4. Markolf, K. L.; Burchfield, D. M.; Shapiro, M. M.; Davis, B. R.; Finerman, G. A. M.; and |and |Slauterbeck, J. L.: Biomechanical consequences of replacement of the anterior cruciate ligament with a patellar ligament allograft. Part I: Insertion of the graft and anterior-posterior testing. J. Bone and Joint Surg.
, 78-A-A: 1720-1727, Nov. 1996.
5. More, R. C., and |and |Markolf, K. L.: Measurement of stability of the knee and ligament force after implantation of a synthetic anterior cruciate ligament. In vitro measurement. J. Bone and Joint Surg.
, 70-A: 1020-1031, Aug. 1988.
6. Muneta, T.; Yamamoto, H.; Sakai, H.; Ishibashi, T.; and |and |Furuya, K.: Relationship between changes in length and force in in vitro reconstructed anterior cruciate ligament. Am. J. Sports Med.
, 21: 299-304, 1993.
Copyright 1996 by The Journal of Bone and Joint Surgery, Incorporated
7. Penner, D. A.; Daniel, D. M.; Wood, P.; and |and |Mishra, D.: An in vitro study of anterior cruciate ligament graft placement and isometry. Am. J. Sports Med.
, 16: 238-243, 1988.