A Biomechanical Analysis of Tibial Fixation Methods in Hamstring-Graft Anterior Cruciate Ligament Reconstruction

Abstract

Background:

There are conflicting results on the biomechanical properties of tibial fixation devices in anterior cruciate ligament reconstruction. The objective of this study is to compare the initial biomechanical properties of tibial fixation in hamstring-graft ACL reconstruction with interference screw, suspension button, and Tape Locking Screw^TM devices. We hypothesized there are no differences in the initial biomechanical properties of these three tibial fixation techniques.

Methods:

Twenty-one fresh-frozen porcine tibiae were equally divided into three groups of seven tibiae to evaluate the fixation of human hamstring tendon grafts with interference screw, suspension button, or Tape Locking Screw fixation. Using a servohydraulic materials testing system, each graft was subjected to 500 cycles of loading followed by a monotonic failure test.

Results:

Interference screw fixation demonstrated significantly lower cyclic displacement (1.28 ± 0.73 mm) than the other groups fixated with either a suspension button device (2.54 ± 0.27 mm, p = 0.003) or a Tape Locking Screw (2.32 ± 0.42 mm, p = .009), and a significantly greater cyclic stiffness (212.19 ± 40.30 N/mm) than the Tape Locking Screw (137.64 ± 26.17 N/mm, p = 0.002). The interference screw also demonstrated significantly higher pullout stiffness (166.83 ± 23.22 N/mm) than the suspension button (112.78 ± 24.14 N/mm, P = 0.002) and Tape Locking Screw (109.11 ± 12.91 N/mm, P = 0.0002).

Conclusions:

Tibial fixation with an interference screw demonstrated superior biomechanical properties for cyclic testing compared to the suspension button and Tape Locking Screw. Load to failure did not differ between groups, and there were no significant biomechanical differences between the suspension button and Tape Locking Screw fixation devices.

Clinical Relevance:

Despite the initial biomechanical differences, all three fixation devices exhibited mean loads to failure and cyclic displacements below clinically relevant thresholds of failure. These data suggest all three fixation methods are viable options for achieving a functional ACL reconstruction.

Level of Evidence: V

Introduction

Anterior cruciate ligament (ACL) reconstruction remains one of the most common orthopaedic procedures performed, with more than 100,000 ACL reconstructions performed annually in the United States.^1,2 The primary goals of ACL reconstruction are to restore knee stability and pre-injury level of function and prevent injury and degeneration to the knee joint. Surgeons utilize both autologous hamstring tendon and bone-patellar tendon-bone (BPTB) grafts in ACL reconstruction. While use of hamstring autograft has been shown to decrease knee flexion strength at high flexion angles,³ many prefer this option due to reports of decreased incidence of anterior knee numbness,⁴ kneeling pain,⁵ and anterior knee pain.^6,7

Stable graft fixation is necessary during biological incorporation to avoid graft elongation and failure.⁸ This is especially true with the trend of early range of motion, weight-bearing and return to sport following ACL reconstruction.⁹ While studies have demonstrated higher ultimate load to failure of hamstring graft compared to BPTB graft,^10,11 the fixation of hamstring graft to tibial bone is often the site of failure due to weaker tibial metaphyseal bone compared to that of the femur.¹²

The tibial fixation construct of choice is difficult to determine because clinical studies are limited by variability in outcomes reporting and various surgical techniques. Those that have compared various methods of tibial fixation have demonstrated no differences in clinical outcomes.^13,14,15,16 Many studies have compared the biomechanical properties of tibial fixation systems with varying results. Depending on the study, interference screws,¹⁷ screw and washer fixation,^18,19 stirrup,²⁰ suspension button^8,21 or a combination of these²² have been linked to superior results.

As the gold-standard tibial fixation device in hamstring-graft ACL reconstruction remains unclear, and a variety of devices are commonly implanted for this procedure, further research is warranted on the subject. This is especially true as the Tape Locking Screw (TLS,^® FH Ortho, Chicago, USA), to our knowledge, has only been evaluated biomechanically twice previously,²³ one of which did not compare this device to other fixation methods.²⁴ The objective of this study is to compare the initial biomechanical properties of tibial fixation in hamstring-graft ACL reconstruction with three devices: Delta interference screw (Arthrex, Naples, USA), Tape Locking Screw, and TightRope RT (Arthrex, Naples, USA) suspension button. We hypothesized there are no differences in the initial biomechanical properties of these three techniques.

Methods

Tibia Preparation:

A total of 21 skeletally mature (same age at slaughter, 11-12 months of age) fresh porcine tibiae were obtained from a local abattoir. The tibia was dissected of all soft tissue and stored in a freezer until the day of testing, at which time they were thawed for a minimum of 12 hours. The 21 porcine tibias were then divided into three groups based on ACL reconstruction and fixation technique. The first group of seven tibiae underwent ACL reconstruction with tibial fixation with a Delta interference screw. Seven tibiae underwent ACL reconstruction with the CoLS Classic System (FH Ortho, Chicago, USA), which utilizes a TLS. The remaining seven tibiae underwent reconstruction with the Graft Link All-Inside ACL (Arthrex, Naples, USA), utilizing TightRope RT suspension button fixation.

Tibial tunnels were prepared in accordance with the reconstruction method to which they were randomized. For the interference screw group, a tibial tunnel guide was set to 55° and the entry site placed midway between the tibial tubercle and posteromedial cortex. A tunnel matching the graft diameter was reamed in antegrade fashion with its intra-articular exit site located at the footprint of the native porcine ACL. The two remaining groups had identical landmarks to insert the guide pin so that it also exited at the footprint of the native porcine ACL. The sockets were created in a retrograde fashion to a depth and width as determined by their respective system.

Graft Preparation and Fixation:

Human hamstring tendon grafts were wrapped in saline-soaked gauze and stored frozen. On the day of testing, the grafts were thawed with normal saline and kept at room temperature for at least one hour. They were prepared depending on their respective product guides. For the interference screw technique, the original allograft was divided into two equal length segments and then each doubled over to create a four-strand graft with a total length of 60 mm—30 mm of tibial tunnel length and 30 mm outside the bone (extra-articular length). Number 5 Ethibond (Ethicon, Somerville, USA) sutures were utilized to whipstitch the free ends of the graft.

The GraftLink All-Inside ACL advocates a combined socket length (end of femoral socket to end of tibial socket) at least 10 mm longer than the graft to ensure that the graft does not bottom out in either socket. Thus, a single graft was quadruple-folded to a total length of 10 mm less than the pre-determined tibial socket length; an extra-articular length of 30 mm was established. Lastly, the CoLS Classic System determines total graft length by the patient’s height. On the femoral side, the graft length is always 10 mm and the tibial side is always 15mm; the intraarticular length varies based on patient’s height. Accordingly, for this testing model, a quadruple-folded graft with a total length of 45 mm was used: 15 mm of tibial tunnel length and 30 mm of extra-articular length.

Grafts for all reconstruction methods were sized to an 8 mm diameter determined by a sizing block. In addition, as is performed during surgery, the allografts underwent preconditioning prior to testing. The interference screw group was preconditioned at 67 N for 10 minutes on a pretensioning board while the TLS and suspension button groups were pre-tensioned according to their respective product manuals.

For Delta interference screw testing, all allografts were fixated using nine mm screws, one mm larger than the tunnel diameter. Using a guide wire, each screw was inserted co-linear to the tibial tunnels and inserted until the distal end of the screw abutted the cortical border of the tibia. A graft tensioner tensioned each strand of the graft to 15 N. For the remaining groups, grafts were fixated according to product manuals for the CoLS Classic System and GraftLink All-Inside ACL. The testing sequence was randomized to minimize potential surgical bias. All surgical fixations were performed by the same orthopaedic surgeon [HF].

Biomechanical Testing:

Mechanical testing was performed using a servohydraulic materials testing system (Figure 1). Each tibia was secured in a custom-designed vise fixture, which allows unconstrained positioning of tibial specimens prior to testing. The base of the vise was mounted to the testing platform and the tibia positioned so the tunnel was parallel to the test actuator axis. This allowed for testing in a “worst-case” scenario with direct, in-line force on the graft. The proximal end of the graft was looped around the cross-pin of a “trapeze” fixture, which was connected directly to the load cell.

Figure 1.

Mechanical testing of tibial fixation with human hamstring tendon in porcine tibia using a servohydraulic materials testing system.

All tests were conducted at room temperature and the allografts regularly moistened with saline during testing. An initial preload of 5 N was applied to each graft before testing and the length of tendon outside of the tibial tunnel A (approximately 30 mm) was reassessed for any displacement. Each specimen was then pre-conditioned at a displacement rate of 0.75 mm/sec between 10 and 50 N for 10 cycles. Next, cyclic tensile testing (havertriangle waveform) was performed at the same displacement rate using 500 cycles between 50 and 250 N. Finally, a load-to-failure test was conducted at 20 mm/min.

For each specimen, four parameters were computed for analysis. Cyclic displacement (net change in peak cyclic displacement over 500 cycles) and cyclic stiffness (the slope of the secant line joining minimum and maximum points of the loading phase of the load-deformation curve reported from the 500^th cycle) were quantified from the cyclic protocol. Maximum failure load and pullout stiffness (the steepest slope of the load-deformation curve spanning 20% of the data points up to maximum load) were calculated from the load-to-failure test. To qualify for statistical analysis, the specimens must have completed the 500-cycle protocol and failed via the load-to-failure test. Any specimen that failed during cyclic loading was removed from statistical analysis. Student’s t-tests were used to compare the three groups.

Results

One specimen from the suspension button group failed during cyclic testing and was removed from statistical analysis. Of the remaining specimens that completed testing, the suspension button failed via tibial button loop rupture in five cases and button migration into the tunnel in one case. All constructs in the interference screw and TLS groups failed by graft pullout from the tunnel.

Interference screw fixation demonstrated significantly lower cyclic displacement (1.28 ± 0.73mm) than the other groups fixated with either a suspension button device (2.54 ± 0.27 mm, p = 0.003) or a TLS (2.32 ± 0.42 mm, p = 0.009), and significantly greater cyclic stiffness (212.19 ± 40.30 N/mm) than the TLS (137.64 ± 26.17 N/mm, p = .002) (Table 1). During load-to-failure testing, the interference screw trended toward a lower mean ultimate failure load (674.91 ± 183.85 N) than both the suspension button (880.05 ± 176.64 N) and TLS (843.93 ± 193.17 N), although there were no significant differences. The interference screw had significantly higher pullout stiffness (166.83 ± 23.22 N/mm) than the suspension button (112.78 ± 24.14 N/mm, p = 0.002) and TLS (109.11 ± 12.91 N/mm, p = 0.0002). There were no significant biomechanical differences between the suspension button and TLS.

Table 1.

Biomechanical Results of Tape Locking Screw, Suspension Button, and Interference Screw Tibial Fixation Devices

Fixation Device	Cyclic	Displacement (mm)	Cyclic Stiffness (N/mm)	Load to Failure (N)	Pullout Stiffness (N/mm)
		(mean ± SD)	(mean ± SD)	(mean ± SD)	(mean ± SD)
TLS		2.32 ± 0.42	137.64 ± 26.17	843.93 ± 193.17	109.11 ± 12.91
SB		2.54 ± 0.27	171.75 ± 32.16	880.05 ± 176.64	112.78 ± 24.14
IS		1.28 ± 0.73a,b	212.19 ± 40.30c	674.91 ± 183.85	166.83 ± 23.22d,e

^aSignificantly different from TLS (p = 0.009).

^bSignificantly different from SB (p = 0.003).

^CSignificantly different from TLS (p = 0.002).

^dSignificantly different from TLS (p = 0.0002).

^eSignificantly different from SB (p = 0.002).

TLS = Tape Locking Screws, SB = Suspension Button, IS = Interference Screw

Discussion

Hamstring autograft is commonly used in ACL reconstruction due to decreased incidence of anterior knee numbness,⁴ kneeling pain,⁵ and anterior knee pain.^6,7 Stable graft fixation is necessary during biological incorporation to avoid graft elongation and failure,⁸ and the fixation to tibial bone is often the site of failure due to weaker metaphyseal bone compared to that of the femur.¹² The ideal fixation technique of ACL reconstruction with hamstring graft remains controversial, and multiple studies have investigated the initial biomechanical properties of tibial fixation devices with differing results.^8,17-23 The objective of this study was to compare the initial biomechanical properties of tibial fixation with three devices: Delta interference screw, TLS, and TightRope RT suspension button. We hypothesized there are no differences in the initial biomechanical properties of these three techniques.

An interference screw is considered a type of aperture fixation, which compresses the graft against the cortical or cancellous bone in the wall of the tunnel. Robert et al.’s study also used porcine tibiae and their Delta interference screw fixation exhibited cyclic displacement 3.81 ± 11.25 mAm, cyclic stiffness 309.7 ± 75 N/mm,B load to failure 844 ± 394 N, and pullout stiffness 195.7 ± 59 N/mm.²³ Suspension button devices utilize proprietary loop configurations to seat the tensioned graft in the tunnel and fix the button against cortical bone. Potential advantages of this method include the necessity of only one tendon graft, implant fixation on cortical instead of weaker cancellous bone, and a larger area of graft to cancellous bone surface area for healing.²¹

The TLS achieves fixation via an interference screw with a polyethylene terephthalate strip, which attaches to the tendon graft loop. Potential advantages of this technique include the need to only harvest a single tendon, better interference fit between the strip and screw compared to the graft,²⁵ and a shorter working length of the suspension construct resulting in a lower likelihood of tunnel widening often seen with suspension button fixation.²⁴ To our knowledge, the initial biomechanical results of TLS fixation has only been studied twice previously. Ayzenberg et al. found load to failure of 523 ± 269 N, but they did not conduct cyclic tensile testing.²⁴ Only one study compared it to other modes of tibial fixation: Robert et al.’s TLS analysis exhibited cyclic displacement 1.23 ± 0.36 mm, cyclic stiffness 295.9 ± 78 N/mm, load to failure 1015 ± 129, and pullout stiffness 138.2 ± 35 N/mm.²³

Many studies have compared the biomechanical properties of tibial fixation systems with varying results. Interference screws have been studied extensively and compared to multiple other devices. Kousa et al. compared a titanium interference screw, a bioabsorbable interference screw, a sheathed interference screw, and washer and screw fixation, and demonstrated the sheathed interference screw had the lowest residual displacement and was the strongest in load-to-failure.¹⁷ Chivot et al. compared an interference screw, a sheathed interference screw, an absorbable cross pin, and a nonabsorbable cross pin, and determined the interference screw exhibited higher load to failure and less slippage²⁶ Giurea et al. compared a stirrup, a clawed washer with cancellous bone screw, a soft titanium interference screw, and a rounded-head interference screw, demonstrating significantly higher load to failure with the stirrup.²⁰

Specific to the devices used in this study, Mayr et al. compared the TightRope RT suspension button and the BioComposite (Arthrex, Naples, USA) biodegradable interference screw, demonstrating that grafts with interference screw fixation showed less cyclic displacement, higher pull-out stiffness, and a lower ultimate failure load compared to suspension button fixation.⁸ In contrast, the TightRope ABS (Arthrex, Naples, USA) suspension button has also been compared to BioComposite interference screw fixation, demonstrating a higher load to failure but comparable cyclic displacement.²¹ In our study, the Delta interference screw demonstrated lower cyclic displacement, no difference in load to failure, and higher pullout stiffness compared to the TightRope RT.

To our knowledge, the TLS’s biomechanical properties have only been investigated in one other comparative study: Robert et al. compared the TightRope RT, WasherLoc (Zimmer Biomet, Warsaw, USA) screw and washer fixation, Delta interference screw, and TLS using porcine bone and human tendons.²³ In their study, TLS demonstrated lower cyclic displacement compared to the other three devices. This contrasts with our study, in which the interference screw group had lower cyclic displacement than the other groups. The difference may be attributed to the fact that Robert et al. included trials with lacerated tendons, which they say accounted for high cyclic displacement in the interference screw group, whereas tendons that could not complete cyclic testing were excluded from our study. Robert et al.’s study demonstrated no difference between groups for cyclic stiffness, but our study showed higher cyclic stiffness for the interference screw than the TLS.

For load-to-failure in Robert et al.’s study, both the TLS and Delta interference screw had the higher yield load compared to the two remaining systems; the TLS and interference screw yield loads were not significantly different. During load-to-failure testing in our study, the interference screw group trended toward a lower mean ultimate failure load than the other groups, although there were no significant differences. The interference screw group also exhibited higher pullout stiffness than the other groups in both studies. There were no significant biomechanical differences between the suspension button and TLS fixation devices in our study. The comparison of these two studies is summarized in Table 2.

Table 2.

Fixation Superiority for Initial Biomechanical Results

	Cyclic Displacement	Cyclic Stiffness	Load to Failure	Pullout Stiffness
Robert et al.	TLS > IS TLS > SB TLS > W	-	TLS > W TLS > SB	IS > TLS IS > SB IS > W
Current Study	IS > TLS IS > SB	IS > TLS	-	IS > TLS IS > SB

IS = interference screw, TLS = Tape Locking Screw,

SB = suspension button, W = screw/washer

This study has a few limitations. First, it describes the initial biomechanical properties of hamstring-graft ACL reconstruction tibial fixation, but it does not evaluate the femoral fixation, the properties of both constructs in tandem, nor the change in biomechanics of these constructs over time. Also, porcine bone was used instead of human bone and direct extrapolation to tibial fixation in human ACL reconstruction cannot be made. However, young human cadaveric bone is difficult to obtain, and a previous study has demonstrated the average density of porcine bone is similar to young human bone and significantly higher than elderly human cadaveric bone.²⁷ Further, there is low variability in porcine tibiae bone mineral density.²⁸ For these reasons, combined with the fact that porcine bone has been used for similar studies in the past,^17,18,23 we believe it is an acceptable substitute.

In sum, Delta interference screw fixation demonstrated less cyclic displacement and greater pullout stiffness than the TLS and Arthrex TightRope RT suspension button and greater cyclic stiffness than the TLS. Load to failure did not differ between groups, and there were no significant biomechanical differences between the suspension button and TLS fixation devices.

Adequate tibial fixation in ACL reconstruction is necessary to allow early rehabilitation and a stable environment for ligamentization and tendon-bone integration. A 500 N load is the estimated tension of an ACL graft during intensive rehabilitation.¹⁸ The post-operative stability of the knee after ACL reconstruction can also be measured using a KT-1000 arthrometer, with a side-to-side difference greater than three mm considered reconstruction failure.⁸ While cyclic displacement is not synonymous with change in anterior laxity, it has been used as an approximation for the purpose of clinical correlation in previous studies.^8,23 Therefore, despite the initial biomechanical differences, all three fixation devices exhibited mean loads to failure and cyclic displacements below these clinically relevant thresholds of failure. These data suggest all fixation methods investigated in this study are viable options for achieving a functional ACL reconstruction.