System Behavior of Commonly Used Cranial Cruciate Ligament Reconstruction Autografts

Veterinary Surgery, 18, 6,459-465, 1989

System Behavior of Commonly Used Cranial Cruciate Ligament Reconstruction Autografts

STUART G. JOHNSON, DVM, DONALD A. HULSE, DVM, DiplomateACVS, HARRY A. HOGAN, PhD, JAMES K. NELSON, PhD, and HARRY W. BOOTHE, DVM, MS, DiplomateACVS

Biomechanical analysis was performed on the cranial cruciate ligament (CCL) and three autogenous tissues used for CCL reconstruction in the canine stifle. The autogenous tissues were patellar ligament-based autografts described for over-the-top CCL replacement and included the central one third of the patellar ligament, the medial one third of the patellar ligament, and the lateral one third of the patellar ligament with fascia lata. Tension testing produced abrupt failure of the central and medial autografts but sequential failure of the lateral autograft. Structural properties were determined for the overload condition and within the load range of normal activity for the CCL (physiologic range). None of the autograft systems approached the stiffness, maximum load, and energy absorbed to maximum load of the CCL. The central and lateral autografts were stiffer, had greater maximum loads, and absorbed more energy to maximum load than the medial autograft. The central and lateral autografts had an elastic range, as defined by proportional limit, which corresponded to the physiologic range of loading for the CCL. Loads that corresponded to physiologic displacement of the lateral and central autografts were near the maximum load of the fixation site, which under- scored the need for postoperative support of the repaired stifle.

R A N I A L CRUC~ATE LIGAMENT (CCL) rupture occurs C frequently in dogs. Intra-articular repair has become a popular method of reconstruction as veterinary surgeons strive to develop physiologic and biomechanically sound methods.'-3 In several procedures, autogenous tissues are placed in the over-the-top position'.' or through bone tunnels'-7 to mimic the course of the normal CC'L. Autogenous tissues that have been tested clinically in- clude the medial one third of the patellar ligament.'-' the central one third of the patellar and the fascia lata alone'-'." or in combination with the lateral one third of the patellar ligament.'.""'" Results of clinical studies of these replacement tissues have been good in 85% to 92% of the cases.i-2.5.y.in However. graft selection for canine CCL reconstruction is still based on limited biomechanical information.

An ideal autograft reconstructive procedure should have the following qualities: ease of collection. adequate initial structural properties, proper placement within the joint, secure fixation to bone, and proper tension applied during tightening. ' ' - I4 The starting point for any discus-

sion of a ligament replacement tissue is the structural properties of that tissue. '' The biomechanical properties of the transplanted tissue determine to a large extent the initial strength of the repair and early return to function. 13-1 5 The initial structural biomechanical properties of the autografts used for reconstruction of the canine CCL are unknown.

The objectives of this study arc as follows: ( I ) to determine the structural biomechanical properties of three autograft systems used for CCL reconstruction and to compare these properties to the normal CCL and (2) to determine the behavior of each system within the physiologic range for normal activity ofthe cranial cruciate ligament.

Methods

Seven pairs of grossly normal stifle joints were collected immediately after euthanasia of healthy. young, adult dogs weighing I4 to 35 kg. The tissues included the stifle joint with surrounding ligamentous support, fascia

From the College of Veterinary Medicine, Department of Small Animal Medicine 8 Surgery (Johnson Hulse, Boothe), and the College of Engineering, Department of Mechanical Engineering (Hogan), and Department of Civil Engineering (Nelson) Texas A 8 M University, College Station, Texas

Presented at the 24th Annual Meeting of the American College of Veterinary Surgeons, Reno, Nevada, February 8, 1989 Correspondence address Stuart G Johnson, DVM, Houston Veterinary Referral Surgical Service. 1551 Campbell Rd Houston, TX 77055

459

460 STRENGTH OF CCL AUTOGRAFTS

TABLE 1. Number and Dimensions of Cranial Cruciate Ligament and Patellar Ligament-Based Autograft Preparations

Mean k SD No. in

Preparation Group Length (mm) Width (mm)

Cranial cruciate ligament 7 20.62 f 2.1 3 4.96 f 0.78 Central autograft 7 74.80 k 7.71 5.05 f 0.38 Lateral autograft 7 67.80 +- 5.16 15.94 f 2.05 Medial autograft 6 74.00 k 4.74 4.72 f 0.31

SD-standard deviation.

lata, quadriceps tendon-patella-patellar ligament mech- anism, and joint capsule. Each stifle specimen was stored in saline-moistened gauze inside a plastic bag at -80°C until the day of testing. Further dissection of the specimens produced the following four preparations: ( 1 ) the intact cranial cruciate ligament as a bone-ligament- bone preparation with the femur and tibia, (2) the central one third of the patellar ligament attached distally to the tibia and proximally to a central wedge of cortical bone from the patella, quadriceps tendon and fascia lata,' (3) the medial one third of the patellar ligament attached distally to the tibia and proximally to a medial wedge of patella, quadriceps tendon, and fascia lata,' and (4) the lateral one third ofthe patellar ligament and adjacent fascia lata attached distally to the tibia and proximally to the patellar retinaculum and fascia lata.'.4 The CCL and central autograft preparations were taken from one stifle of each dog, and the medial and lateral autograft preparations were taken from the contralateral stifle. Collection of the autograft preparations followed the techniques described in the literature and was randomized between right and left limbs.

The length and width of each preparation were measured with digital calipers (Table l). The length of the CCL was determined from the average of four measure- ments taken around the ligament. The loaded length of each autograft was measured from its tibial attachment to the point of fixation at the femoral condyle when the over-the-top position was simulated. The width of each preparation was determined as the average of four mea- surements along its length.

Each bone was held by two 4-mm stainless steel pins passed transversely through the bone and into a specially designed clevis (Fig. 1). The bones of the CCL preparations were held in 40" ofjoint flexion. All tibias were held at 35" to the axis of loading. For the autograft preparations, the proximal soft tissues were held within wedge action grippers* that had corrugated inserts to prevent crushing. The autograft and the CCL preparations were

positioned parallel to the axis of loading with care taken not to twist the preparations. Physiologic saline solution was sprayed on the tissues to prevent desiccation. The preparations were loaded to failure in the materials testing machine? at a constant rate of 25 mm/sec with load and displacement data recorded simultaneously by the machine. A computer within the testing machine stored the data until it could be transferred to a personal computer for reduction and manipulation. All tests were vid- eotaped to failure. One medial autograft preparation was excluded from the study because it was damaged during mounting and failed prematurely.

From the load-displacement data of each test, structural properties were calculated under two conditions- traumatic overload and physiologic range. The overload parameters examined were failure mode, maximum load, energy absorbed up to maximum load, displacement to maximum load, and stiffness. Failure mode was determined by examination of the failed specimen and the videotape of each test. Energy absorbed up to maximum load was the area under the load-displacement curve up to maximum load. Stiffness was determined by regression analysis of the upper part of the linear region of each curve. For the physiologic range, stiffness and displacement up to a load of 169 newtons (N), load up to 4 mm of displacement, and the proportional limit (load achieved at the end of the linear region) were determined. The structural properties determined in each test were averaged by preparation and all preparation aver- ages were pooled and an analysis of variance among preparation groups was performed. Initially, all variables were tested for normality. For variables that showed a skewed distribution, a square root transformation proved sufficient for normalization. Orthogonal and linear contrasts were used to identify significant differences at the p < .05 level.

Results

Faillire Mode

The CCL preparations failed by femoral avulsion (3 dogs), tibial avulsion (2), or interstitial failure (2). The central autograft preparations failed either through the fascia lata proximal to the quadriceps tendon (4 dogs) or through the patellar ligament (3) (Fig. 2). Maximum load was reached at similar times for the central and lateral autograft preparations; however, the lateral preparation sustained load longer than the central preparation (Fig. 3). The lateral preparations failed sequentially (Fig. 3). with separation of the patellar retinacula from lateral to medial (Fig. 4). All six medial autograft preparations

~

* Instron Corp, Canton, MA. t MTS Systems Corp.. Minneapolis, MN.

JOHNSON, HULSE, HOGAN, NELSON, AND BOOTHE 461

central and lateral autograft preparations were stiffer than the medial preparation.

Physiologic R N nge

The values for the physiologic range are given in Table 3 . Stiffness of the CCL preparations in the physiologic range was less than i n the overload range but was still significantly greater than that of the autograft preparations. For the autograft preparations, the values for stiffness in the physiologic range were similar to those in the overload range. The stiffness of the central preparation

Fig. 1 . Central autograft preparation mounted in the materials testing machine.

failed abruptly through the fascia lata proximal to the quadriceps tendon (Fig. 4) early in the loading period (Fig. 3 ) .

Overload Properties

The overload results are listed in Table 2. The CCL preparations achieved a greater maximum load than the autograft preparations. The central and lateral autograft preparations reached similar maximum loads, which were approximately 29% of the CCL value. The medial preparation had a significantly lower maximum load, 10% of the CCL value. The CCL absorbed significantly greater energy up to maximum load than the autografts. The central and lateral autograft preparations absorbed similar amounts of energy up to maximum load, significantly more than the medial preparation. All values for displacement to load were significantly different from one another. The CCL Preparations were significantly stiffer than the autograft preparations. The

Fig. 2. Failed central autograft preparations. The modes of failure shown are proximal through the fascia lata (right) and distal through the patellar ligament (left).

462 STRENGTH OF CCL AUTOGRAFTS

- LATERAL CENTRAL

plied in the 1aborat0ry.l~~~"'~ In vivo loads have been calculated to be from 10% to 25% of the tissue's ultimate strength under normal condition^.'^.'^ Therefore, in ad- dition to maximum values for structural properties, we determined the function of each autograft in consider- ation of expected physiologic conditions.

Biomechanical data are available for human anterior cruciate ligament reconstruction tissues.I4 In that study, two of the specimens tested were an isolated strip of fascia lata and one third of the patellar ligament as a bone- ligament-bone preparation. The latter tissue was found to be one of the strongest with a strength of nearly 160%

TIME

Fig. 3. Representative load behavior curves of three autograft preparations. The curves have been offset on the ordinate to differentiate them. The medial preparation failed abruptly, early during loading. The central preparation also failed abruptly, but the lateral preparation failed sequentially, sustaining load longer.

was greater than the stiffness of the medial but not the lateral preparation. The displacement response of the central preparations up to a load of 169 N was significantly less than the displacement response of the lateral preparation. Only one medial preparation reached 169 N of load: its 5.58 rnm displacement was significantly greater than the central but not the lateral preparation. The load response of the CCL preparations to 4 mm of displacement was significantly greater than the autograft preparations. The lateral and central autograft preparations had greater loads to 4 mm of displacement than the medial preparation. The values for proportional limit were greatest for the CCL preparations, and the value for the medial preparation was significantly less than the other autograft preparations.

Discussion

There is little published information about the biomechanical behavior of CCL reconstruction grafts. In one study, failure during the early stages of healing occurred at the fixation site." Until sufficient time has elapsed for fibrosis to the femoral condyle (2-12 week~), '~ . ' ' only the fixation site is being tested and not the replacement tissue. With a more ideal fixation system, slippage of the tissue would be prevented, and the initial strength of the repair would more closely reflect the mechanical properties of the graft itself. The maximum, or near-maximum, structural properties of the replacement tissue would become more t' a discussion Of graft biomechanical behavior. in vivO forces On ka- mentous structures are considerably less than those ap-

Fig. 4 . Failed medial and lateral autograft preparations. The medial autograft (/eft) failed through the fascia lata. The lateral autograft (right) failed from lateral to medial through the peripatellar tissues.

JOHNSON, HULSE, HOGAN, NELSON, AND BOOTHE

TABLE 2. Structural Mechanrcal Properties for the Overload Condition of the Cranial Cruciate Ligament and Patellar Ligament -Based Autograft Preparations

Mean 2 S D

Preparation

Maximum Energy Absorbed to Displacement to Load Maximum Load Maximum Load Stiffness (N) (Nm) (mm) "m)

~ ~~ ~ ~~~~~~~ ~~~~ ~ ~~ ~~~~~ ~ ~

Cranial cruciate ligament 1469.25 * 377' 5.17 f 1.89' 7.17 2 1.39' 238.22 2 47.71' Central autograft 422.09 t 124 2.16 t 0.97 9.29? 1 88$ 53.762 11.34 Lateral autograft 427.10 t 120 2.71 f 0.83 12.46 t 2 29$ 44.82 t 13.60 Medial autograft 150.02 * 1007 0.49 f 0.50 t 4 90 * 0 77$ 27.07 2 8.917

SD-standard deviation. * Significantly different from the autograft values (p < .05). t Significantly different from lateral and central autograft values (p < -05). * Significantly different from other autograft values (p < .05)

of the anterior cruciate ligament. We found that the patellar ligament-based grafts, placed in the over-the-top position in the canine stifle, relied on the fascia lata for strength. Therefore, the autograft used for canine CCL reconstruction is different from the human anterior cruciate ligament reconstruction and extrapolation of the human biomechanical data to the canine is erroneous. To portray the mechanical properties ofthe CCL reconstruction autografts accurately, we chose instead to test the entire autograft system.

To simplify the procedure, two preparations from each stifle were tested. This method of testing risked hav- ing the collection and testing of one preparation interfere with the collection and testing of the other preparation from the same stifle. Using the techniques described in the literature, we found that the lateral autograft could be collected from one stifle specimen without lessening the proportions of the medial autograft. Therefore, failure of the medial autograft through the fascia lata was

considered to be caused by relative weakness of the fascia lata proximal to the quadnceps tendon. Conversely, the location of the central autograft allowed a much stronger proximal tissue to be collected so that failure occurred in the patellar ligament in some of the preparations. The lateral autograft failed consistently through the lateral patellar retinaculum, suggesting an area of the graft that could be modified or reinforced. Also, the mechanical response at failure indicated by the load behavior curves (Fig. 3 ) showed that sequential failure of the lateral autograft may permit recoverability of function after a traumatic overload.

Maximum load i s believed by some to be an important factor in the resistance of a structure to traumatic over-

Similarly. the energy up to maximum load reflects the ability of the structure to absorb the shock of an acute overload. The average maximum load for our CCLs ( I469 N) was similar to that reported by others for the canine CCL.'h-''.''l The maximum loads and energy

load, 12.14.1?

TABLE 3 Structural Mechanrcal Properties for the Physiologic Loadrng Condrtion of the Cranial Crucrate Ligament and Patellar Ligament Based Autograft Preparatrons

Mean _+ S D

Stiffness to 169 Displacement to Load to 4 mrn Proportional Newtons 169 Newtons Displacement Limit

Preparation (Nimm) (mm) (N) (N)

Cranial cruciate ligament 161.04 2 46.02t 1 1 6 2 0.49t 800.26f 173.01t 1425.72 f 393t Central autograft 50.35 t 4.78$ 3.38 -t 0.349 201.42f 1921 352.56 -t 105 Lateral autograft 38.86 t 8.32 4 55 2 0 98 154.69f 33 74 373.12 t 100 Medial autograft 27.29 t 10.21 5.58' 94.75 t 23.427 125.22i 951

SD-standard deviation. Only one medial preparation reached 169 N before failure.

t Significantly different from autograft values (p i ,051). * Significantly different from medial autograft value (p < .05) 5 Significantly different from lateral and medial autograft values (p < .05). ll Significantly different from central and lateral autograft values (p .05).

STRENGTH OF CCL AUTOGRAFTS

absorbed up to maximum load were considerably less for the autograft preparations than for the CCLs, indicating that the ability of the autografts to withstand acute over- loads was marginal in the initial reconstruction. The difference in the average values among the autograft preparations would favor the selection of those grafts with the highest maximum load and greatest energy absorbed, the lateral and central autografts.

Reporting only the structural properties at maximum load neglects a tissue’s performance within the load range of normal activity (physiologic range).” The response of a ligamentous structure to in vivo forces de- pends on the stiffness of the s t r~cture . ’~.~’ Since these forces represent the lower end of the load-displacement curve,” we recalculated stiffness for each preparation. A load of 169 N was chosen because it was reported to be the maximum load achieved before failure at the fixation site in the early stages of healing.“ Also, 169 N was 12% of the average maximum load for the CCLs and corresponded to the lower end of the physiologic range for the CCL. For the CCL, the physiologic range was in a re- duced stiffness region of the load displacement curve. This was in agreement with the results of a previous study in which cyclic testing was performed on the canine CCL.2’ Furthermore, in that study, loads up to 200 N did not produce permanent deformation of the CCLs tested. For the autografts, similar stiffness values for the overload and physiologic loading conditions reinforced the contention that the overall structural stiffness could be used to evaluate graft stiffness under physiologic conditions.” Likewise, the load and displacement responses within the physiologic range reflected the relative stiffness ofeach autograft system. The stiffest autograft (central) had the least displacement up to 169 N and greatest load to 4 mm of displacement. Given the importance of stiffness to the reconstruction, the central and lateral autografts compare favorably within the physiologic range and offer an advantage to the medial autograft in overall system stiffness.

The secondary joint restraints of the stifle would prevent the large displacements to maximum load of the autografts.” We chose to use 4 mm of displacement, because it was below the displacement to maximum load of the CCL and more accurately depicted a physiologic displacement for the graft tissue. Four millimeters of displacement produced a load response in the central and lateral autografts near the maximum load for the fixation site and in the medial autograft near its maximum load. Therefore, the central and lateral autografts would be expected to undergo physiologic displacement from loads that could exceed the maximum load of the fixation site. A similar physiologic displacement could produce permanent deformation of the medial autograft. This rein- forces the need for early postoperative s u ~ p o r t , ~ ’’ not

only to protect the graft from overload but also to protect the fixation site.

The proportional limit of a structure is a conservative estimate of the load beyond which permanent deformation can occur and is the upper end of the elastic range. The proportional limits of the lateral and central autografts were 25% of the average maximum load of the CCLs. This corresponded to the upper limit of the physiologic range for normal activity in the CCL. If the fixation site could withstand loads at or beyond the proportional limit of the autograft, the reconstruction would have an elastic range within the physiologic range for the CCL. This may explain generally good results in clinical studies and a lack of graft failures in research studies.

Many factors are involved in the success ofa particular CCL replacement procedure. The structural characteris- tics of the central and lateral autografts would favor their use, but the accessibility of the lateral autograft may make it preferable.’.’’ The lateral autograft is a wider graft than the central autograft, but the similar structural properties suggest that the lateral autograft is made of a weaker material. However, the size of the medial and central autografts is restricted by location, whereas the lateral autograft can be modified by the surgeon to in- clude more periarticular tissue to strengthen the structure. Other factors, such as enhanced collagen remodel- ing, isometric intra-articular placement, and improved fixation methods, must be addressed systematically before a completely effective reconstruction can be devel- oped for the cranial cruciate-deficient canine stifle.

References

I . Arnoczky SP, Tarvin GB, Marshall JL, Saltzman B. The over-the- top procedure: a technique for anterior cruciate ligament substi- tution in the dog. J Am Anim Hosp Assoc 1979; I5:283-290.

2. Shires PK, Hulse DA, Liu W. The under-and-over fascia1 replacement technique for anterior cruciate ligament rupture in dogs: a retrospective study. J Am Anim Hosp Assoc 1984;20:69-77.

3. Arnoczky SP. Torzilli PA, Marshall JL. Biomechanical evaluation of anterior cruciate ligament repair in the dog: an analysis ofthe instant center of motion. J Am Anim Hosp Assoc i 977; I3:553- 558.

4. Hulse DA, Michaelson F, Johnson C, Abdelbaki YZ. A technique for reconstruction of the anterior cruciate ligament in the dog: preliminary report. Vet Surg 1980;9:135-140.

5. Dickinson CR. Nunamaker DM. Repair of ruptured anterior cruciate ligament in the dog: experience of 101 cases, using a modi- tied fascia strip technique. J Am Vet Med Assoc 1977; 170:827- 830.

6. Paatsama S. Ligament injuries of the canine stifle joint: A clinical and experimental study. Thesis, Helsinki, 1952.

7. Pichler ME, Bacon JP. Evans JA. The fascia lata as a replacement for the cranial cruciate ligament: two new surgical techniques. J Am Anim Hosp Assoc 1982: 18:779-784.

8. Arnoczky SP. Tarvin GB, Marshall JL. Anterior cruciate ligament replacement using patellar tendon. J Bone Joint Surg 1982;64- A12 I 7-224.

JOHNSON, HULSE, HOGAN, NELSON, AND BOOTHE 465

9. Denny HR, Barr AR. An evaluation of two "over the top" techniques for anterior cruciate ligament replacement in the dog. J Small Anim Pract 1984:25:759-769.

10. Denny HR. Barr AR. A further evaluation of the "over-thc-top" technique for anterior cruciate ligament replacement i n the dog. J Small Anim Pract 1987:28:681-686.

1 1, Smith G. Fibular head transposition for repair of thc cruciate-deli- cient stifle in the dog. J Am Vet Med Assoc 1985; I87:375-383.

12. Butler DL, Grood ES, Noyes FR, Sodd AN. On the interpretation of our anterior cruciate ligament data. Clin Orthop 1985: 106: 26-34.

13. Noyes FR. Butler DL, Paulos LE. Grood ES. Intra-articular cruciate reconstruction. I: Perspectives on graft strength. vasculariia- tion. and immediate motion after replacement, Clin Orthop 1983: 172:71-77.

14. Noyes FR. Butler DL, Grood ES. et al. Biomechanical analysis of human ligament grafts used in knee-ligament repairs and recon- structions. J Bone Joint Surg 1984;66-A:344-352.

15. James SL. Biomechanics of knee ligament reconstruction C'lin Or- thop 1980: 146:90-101.

16. Butler DL. Hulsc DA. Kay MD. et al. Biomechanics of cranial cruciate ligament reconstruction in the dog. 11: Mechanical properties. Vet Surg 19x3; 12- I 13-1 18.

17. Butler DL., Grood ES, Noyes FR. ct al. Mechanical properties of primate vasculariLcd \s. nonvasculanzed patellar tendon grafts: changes over time. J Orthop Res 1989:7:68-79.

18. Torzilli PA. Arnocrhy SP. Mechanical properties of the lateral col- lateral ligament: etTect olcruciatc instability in thc rabbit. J Bio- mech Eng 1988: I I0:1OX-2 17.

19. Arms SW. Beynnon B. Fischer RA. r t al. The biomechanics of ACL. reconstruction in the canine model. Trans 33rd Ann Meet Orth ResSoc I9X7:12:101

20. Vasscur PB. Pool RR, Arnocrk? SP. 1,au RE. Correlative hiorne- chanical and histologic stud) of the cranial cruciate ligament in dogs.Am J Vet Res 1985:lh:I847-I854.

7 I . Dorlot JM, Ait Ba Sidi M. Tremhlav GM, Drovin G. L.oad elonga- tion behavior of the canine antenor cruciate ligament. J Bio- mech Eng 1980; 102:19O-193.

22. Butlcr DI . Noyes FR. Cirood ES. Ligamentous restraints to anteri- or-posterior drawcr i n the human knee. J Bone Joint Surg 19x0: 62-A:259-270

Abstract of Current Literature

RECOVERY OF THE DOG QUADRICEPS AFTER 10 WEEKS OF IMMOBILIZA- TION FOLLOWED BY 4 WEEKS OF REMOBILIZATION Lieber RL, McKee-Woodburn T. Gershuni DH Joitrnul ofOrthopuedic Rtwurch 1989: 7:408-4 I2

Skeletal muscle fiber areas were measured in three heads of the canine quadriceps femons after 10 weeks ofimmobilization followed by 4 weeks of remobilization. By two-way analysis of variance, there was a significant decrease in both type I ( p < ,005) and type 2 (p < .001) fiber areas. However, there was no significant difference between the three heads of the quadriceps ( p > .2). Although muscle fiber areas had not returned to control levels after remobilization, the fraction that was perimysial and epimysial connective tissue was not significantly different from control values ( p > .15). These data suggest that although the degree of muscle atrophy after 10 weeks of immobilization is severe and muscle specific, following 4 weeks of remobilization, muscles uniformly recover to approximately 70% of control values.

Documents

System Behavior of Commonly Used Cranial Cruciate Ligament Reconstruction Autografts