A biomechanical analysis of tibial ACL reconstruction with graft length mismatch

Abstract

Introduction:

The incidence of graft length mismatch (GLM) during anterior cruciate ligament (ACL) reconstruction is reported to be up to 13%, with a rate of 20% when using bone-patellar tendon-bone (BPTB) allografts. Multiple techniques have been described to accommodate for the longer BPTB graft. As no study has compared the biomechanical properties of these methods (with cyclic loading), we evaluated the strength of four different surgical techniques used to accommodate for GLM during ACL reconstruction.

Methods:

A total of 32 fresh-frozen bovine tibiae and patellar tendons were divided into four groups based on the method of tibial graft fixation: (1) sutures tied over a post, (2) bone staples, (3) screws and washers, and (4) soft-tissue conversion with interference screw. Biomechanical testing was performed comparing the tensile properties of graft fixation techniques under cyclic loading. Ability to withstand 1500 cycles of load, the maximum tensile strength at load-to-failure, and the mode-of-failure were evaluated.

Results:

Only group 4 had all grafts intact after 1500 loading cycles, while the other groups had one graft failure at 338 (group 1), 240 (group 2), and 309 (group 3) cycles. The highest mean load-to-failure was observed in group 3 at 762 ± 173 N, which was found to be significantly higher than the other groups. The mean loads to failure in groups 1–4 were 453 ± 86 N, 485 ± 246 N, 762 ± 173 N, and 458 ± 128 N.

Conclusion:

While there are multiple viable techniques for fixation of a BPTB graft in the case of GLM, this study demonstrated that direct screw fixation offers the strongest construct.

Introduction

Rupture of the anterior cruciate ligament (ACL) is one of the most common injuries of the knee, with an annual incidence of up to 200,000 cases in the United States alone. ¹ Accordingly, estimates of 60,000–175,000 ACL reconstruction surgeries are performed each year. ² Although commonly performed, the debate over the best type of graft to use continues to exist; physician survey data shows regionally based differences among surgeons using bone-patellar tendon-bone (BPTB) grafts or soft-tissue constructs as well as autografts versus allografts. ^{3 –6}

Graft length mismatch (GLM) is a specific scenario that can occur when utilizing a BPTB graft. The fixed patellar tendon length is longer than the intra-articular space available; hence, the tibial bone block protrudes from the extra-articular aperture of the tibial tunnel after appropriate tensioning. ⁷ In recent studies, the incidence of GLM is reported to be approximately 13% of BPTB autografts, with an incidence of up to 20% when using BPTB allografts. ^8,9 Several solutions for graft fixation have been described in the literature, including adjusting the angle of the tibial tunnel, twisting the patellar tendon graft to shorten it, tying sutures inserted through the tibial bone plug over a post, seating the bone block in a trough and securing it with screws or staples, and excising the bone block such that a soft-tissue interference screw can be used. While multiple biomechanical studies have been performed to evaluate the individual stability of such fixation methods, we know of no study that has directly compared these salvage techniques to each other. Also, no biomechanical study of this particular scenario used a cyclic loading protocol, which can better mimic in vivo conditions of the knee during ambulation.

The objective of our study was thus twofold: (1) to compare the biomechanical strength of four different surgical techniques used in the case of GLM during ACL reconstruction in a bovine model and (2) to perform a study using a cyclic loading protocol to demonstrate whether such methods are viable options under in vivo conditions. The specific techniques we compared included suture fixation over a post, screw fixation of the tibial bone block, staple fixation of the tibial bone block, and soft-tissue interference screw fixation of the patellar tendon after excision of the protruding tibial bone block.

Materials and methods

A total of 32 fresh-frozen bovine tibiae and BPTB grafts were obtained for testing from the anatomy laboratory at our institution. These were randomly divided into eight samples for each of the four fixation techniques tested: group 1—sutures tied over a post, group 2—bone staple fixation, group 3—screw and washer fixation, and group 4—soft-tissue interference screw fixation. The bovine tibiae were defrosted for 24 h and cleared of all musculature and bone–ligament interfaces prior to preparation and testing. In all cohorts, using a standard 10 mm ACL drilling guide, tunnels were drilled into all tibiae at an angle of 50°. An additional trough was created with a bone burr at the distal edge of the anterior tibial aperture for samples in groups 1–3 (a total of 24 tibiae) to accommodate a 10 × 25 mm bone plug. For testing, the base of the tibia was potted in a custom-milled jig (Instron® 8874 Load Frame; Instron Corporation, Norwood, Massachusetts, USA) that oriented the specimen such that load from the dual-column load frame would be applied in line with the tibial tunnel (see Figure 1(a)).

Figure 1.

Setup of fresh-frozen bovine tibia and BPTB allografts in Instron loading frame. (a) Base of tibia potted to custom jig. (b) BPTB allograft fixed in tibial tunnel. Jig aligns patellar tendon to be loaded in line with the tibial tunnel, thus pulling longitudinally across tendon fibers. Free patellar tendon end is secured in pneumatic clamp attached to the actuator of loading frame. BPTB: bone-patellar tendon-bone.

Preparation of the BPTB grafts for all biomechanical testing included excision of the femoral bone plug to allow for measurement of patellar tendon strength during testing. These free patellar tendon ends were secured proximally in a pneumatic clamp that was attached to the actuator of the load frame (see Figure 1(b)). The tibial ends were standardized with a rongeur and bone burr to 10 × 25 mm size bone blocks as per the technique described by Kurosaka et al. ¹⁰ For group 1 samples, the tibial bone plug was drilled transversely with a 3.5 mm drill at 3 mm away from the proximal and distal edges. A nonabsorbable, braided, polyester suture (Ethibond Excel Number 2; Ethicon, Somerville, New Jersey, USA) was then passed through each drill hole. The graft was inserted through the tibial tunnel such that the bone plug laid in the trough. A 4 × 30 mm screw was inserted with a washer 2 mm distal to the trough. Prior to full seating of the screw and washer, the suture ends from the tibial bone plug were firmly tied beneath the washer. For group 2 specimens, the graft was again pulled through the tibial tunnel such that the bone plug laid in the trough. Two bone staples were then applied in the anterior to posterior direction, transverse to the axis of the graft, to secure the bone block to the tibia. In group 3, once the bone plug was seated in the trough, two 2.7 mm drill holes were made in the anterior to posterior direction at 3 mm away from the proximal and distal ends. Two 3.5 × 40 mm screws were then inserted with washers for fixation. In group 4, the tibial bone block was excised and a whipstitch was placed through the distal end of the patellar tendon with a Number 2 Ethibond Excel suture. With the femoral end clamped proximally in the Instron® machine, the tendon graft was passed through the tibial tunnel, hand tensioned, and secured in place with a 9 × 20 mm soft-tissue interference screw. Excess tendon distally was removed from the tibial aperture.

Biomechanical testing

Loading tests were performed with the Instron 8874 hydraulic mechanical tester in a similar fashion as previously published studies comparing the tensile properties of graft fixation techniques with a cyclic program. ^{11 –16} Three phases of loading were utilized: (1) preloading involved gradually tensioning the grafts to 50 Newtons (N) to account for stress relaxation of the patellar tendon; (2) cyclic loading applied 50–200 N of load across the grafts at 2 Hz for 1500 cycles to mimic the gait cycle of an in vivo graft; and (3) loading to failure included tensioning the graft at 50 mm/min until failure occurred, which measured the remaining strength of the graft following 1500 cycles of simulated ambulation. All specimens were kept moist with a saline spray throughout preparation and biomechanical testing.

Data analysis

Data collected included the number of cycles each specimen survived under cyclic loading, the ultimate load-to-failure if the graft completed 1500 cycles, and the mode-of-failure. The Kruskal–Wallis test was used to test for differences between nonparametric group data. Dunn’s multiple comparison test was used to evaluate for significance in fixation strength between the various fixation methods. Level of significance was set at a p value of less than 0.05.

Results

In all groups that retained the distal bone plug (groups 1–3), seven of the eight samples (87.5%) withstood 1500 cycles of loading without failure. The number of cycles prior to failure were similar, with one sample in group 1 failing at 338 cycles due to suture cutout, one sample in group 2 failing at 309 cycles due to intratendinous rupture, and one sample in group 3 failing at 240 cycles due to bone block pullout. For specimens that completed 1500 cycles of loading, the mean ultimate loads to failure were 453 ± 86 N (range, 383–620 N) for group 1, 762 ± 173 N (range, 572–934 N) for group 2, 485 ± 246 N (range, 249–847 N) for group 3, and 458 ± 128 N (range, 265–672 N) for group 4. Statistical analysis demonstrated a significant difference between the fixation methods with the nonparametric Kruskal–Wallis test (p = 0.1222), as well as a significant difference in fixation strength for group 2 (screw and washer fixation) over the other fixation techniques after cyclic loading with Dunn’s multiple comparison test (p < 0.05).

The modes of failure varied according to the graft fixation method. From group 1 (sutures tied over a post), five of the eight specimens (62.5%) failed secondary to suture cutout, two of the eight (25%) failed secondary to suture rupture, while one of the eight (12.5%) failed secondary to intratendinous rupture. In group 2 (screw and washer), five of the eight samples (62.5%) failed due to bone block fracture, while the remaining three (37.5%) failed due to intratendinous rupture. For group 3 (bone staples), six of the eight specimens (75%) failed from bone block pullout and the remaining two (25%) failed from bone block fracture. All samples in group 4 (soft-tissue interference screw) failed secondary to tendon pullout.

Discussion

BPTB grafts remain as a reliable choice for ACL reconstruction procedures because of its high load-to-failure rate as well as faster graft integration from bone-to-bone healing. In the uncommon case of GLM, particularly when utilizing allograft specimen, tibial fixation can become an issue. The incidence of GLM has been reported to be as high as 13%, and up to 20% when using BPTB allografts. ^8,9

Several techniques have been developed to accommodate for the longer graft, with an emphasis on preserving the tibial bone block so as to retain the quicker bone-to-bone healing relationship. One such method is to adjust the tibial tunnel angle according to the length of the patellar tendon graft (the “N + 7” rule). ¹⁷ However, the accuracy and effectiveness of this calculation has been shown to be variable. Another method is to rotate the patellar tendon graft to shorten its overall length. Berkson et al. evaluated this method by externally rotating 35 BPTB composite porcine grafts by 0, 180, and 540° and subjecting them to a cyclic loading protocol. ¹⁸ The grafts were shortened by 1.7 ± 0.8 mm at 180° of rotation and by 7.6 ± 2.0 mm at 540° of rotation (p < 0.01). Their results demonstrated a statistically significant increase in strain in BPTB grafts rotated 540° when compared to nonrotated grafts, but no significant differences in maximum load, yield stress, yield strain, or modulus of elasticity.

Other salvation techniques for GLM consist of various methods to fixate the tibial end of the BPTB graft. These include suture fixation over a post, direct screw fixation, staple fixation, and conversion to soft-tissue interference screw. Several studies have been performed to investigate the biomechanical strength of these methods. Kurosaka et al. reported a significant mechanical advantage for direct screw fixation compared to staple fixation and suture tying over a button. ¹⁰ A similar study by Novak et al. compared suture tying over a screw post to free bone block interface fixation for BPTB grafts in 28 bovine knees. ¹⁹ This study showed that interference fixation failed at significantly higher force (669 N; range, 511–819 N) when compared to sutures tied over posts (374 N; range, 266–491 N; p < 0.001). There was also an observed increase in stiffness with free bone block interface fixation compared to sutures tied over posts. Some studies are in disagreement, such as Matthews et al. who found no statistical difference in maximum tensile strengths between interference screw fixation and tying sutures over a post. ²⁰ Gerich et al. examined interference screw fixation versus staple fixation in 55 fresh-frozen human cadaver specimens and showed increased stiffness with staple fixation but no statistical difference in load to failure. ²¹ However, there was no study, to the best of our knowledge, that utilized cyclic loading in their testing protocols, which is described as more accurately reflecting the stresses a knee joint undergoes during ambulation. In addition, no study has been performed to evaluate the specific circumstance of tibial fixation for GLM. The objective of this study therefore was to directly compare the biomechanical strength of four different surgical techniques to accommodate for GLM using a cyclic loading test protocol.

Our study demonstrated that direct screw and washer fixation of the protruding tibial bone plug produced a significant advantage by withstanding higher load forces before failure compared to suture tying over a post, staple fixation, and conversion to soft-tissue interference screw with excision of the tibial bone plug. This result suggests that direct screw and washer fixation may be a favorable technique when GLM is encountered intraoperatively. It retains the bone-to-bone healing relationship while providing a stronger and stiffer construct until integration with the tibia has occurred. This construct also provides a higher ultimate load to failure, which could mean that patients could potentially progress in rehabilitation without delay.

There were several limitations to our study. While bovine tibiae and BPTB grafts are an established and acceptable model for biomechanical studies of ACL reconstruction, there are marked differences from in vivo studies of human ACL reconstructions. The healing process is not accounted for with in vitro animal studies and neither are different patient demographic nor rehabilitation practices. We acknowledge that the results of our study therefore may not indicate clinical significance in terms of patient function and satisfaction. However, the bovine model used was able to provide some insight on an uncommon scenario that would otherwise be difficult to gather a sample size large enough to produce data of significant difference, if any. Furthermore, our study did not have a control group as there is no established standard fixation method in the case of GLM. Another limitation is that the actual graft length in excess will vary per patient and may not be so long as to allow creation of a bone trough to seat the bone plug at the extra-articular aperture of the tibial tunnel. However, we believe that the results of our study reliably evaluated the biomechanical strength of different BPTB constructs, which may potentially correlate with patient outcomes.

In conclusion, our biomechanical comparison of four possible salvage techniques for GLM during ACL reconstruction showed that suture tying over a post, screw fixation, staple fixation, and conversion to soft-tissue interference screw fixation following excision of the tibial bone plug are all viable methods. However, a significantly higher load to failure was found when seating the tibial bone plug in a trough and securing it with screws. Despite encountering GLM intraoperatively, patients may benefit from reduced failure rates with a stronger and stiffer construct.