If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The failure load of anterior cruciate ligament repair with Internal Brace Augmentation is high.
Higher load to failure for anterior cruciate ligament repair with augmentation than without.
There is higher stiffness for the anterior cruciate ligament repair with Internal Brace.
There is more energy until failure for anterior cruciate ligament repair with Internal Brace.
Different mode of failure for anterior cruciate ligament repair augmented with Internal Brace.
Newer repair techniques of anterior cruciate ligament tears, including augmentation with internal brace, have shown promising clinical results. Few biomechanical studies exist comparing anterior cruciate ligament repair only versus repair with internal brace. The purpose of this study was to compare the load to failure and stiffness of anterior cruciate ligament repair with internal brace augmentation versus repair-only.
Proximal femoral avulsion type anterior cruciate ligament injuries were created in 20 cadaver knees. Anterior cruciate ligament repair-only or repair with internal brace was performed using arthroscopic tools. Load to failure and failure modes were collected, with calculations of stiffness and energy to failure performed.
The average load to failure for the internal brace group was higher than the repair-only group: 693 N (SD 248) versus 279 N (SD 91), P = .002. The stiffness and energy to failure values were higher for the internal brace group than the repair-only group: 83 N/mm versus 58 N/mm, P = .02 and 16.88 J (SD 12.44) versus 6.91 J (SD 2.49), P = .04, respectively. Failure modes differed between groups (P = .00097) with 80% failure in the repair-only due to suture pull through the anterior cruciate ligament and 90% failure in the internal brace group due to suture button pull through the femur.
There was higher load to failure, stiffness, and energy to failure for the internal brace group compared to the repair-only group, and a high positive correlation between bone density and load to failure for the internal brace group.
Anterior cruciate ligament repair with internal brace augmentation demonstrates significantly higher load to failure. It may be a useful adjunct to protect the anterior cruciate ligament repair from failure during the early stages of healing.
). Feagin and Curl reported ACL repair data which demonstrated deterioration in mid- to long-term follow-up with 53% re-injury rate at five years, in conjunction with high rates of pain, instability, and stiffness (
More recently, there has been renewed interest in ACL repair. In order to protect the ACL repair and improve outcomes, augmented repair with internal brace (IB) has been developed. MacKay et al. reported supplementation of primary ACL repair with heavy-braided augmented suture (2.5 mm polyethylene tape). This was made possible due to the use of an alternative means of fixation using suspensory button techniques rather than suture anchors (
). In their study, only 1 of 27 patients who underwent primary repair with an internal brace construct went on to failure and was subsequently converted to formal reconstruction. The theory behind the internal brace is that it protects the repair and increases the strength during early healing, allowing early mobilization (
). With this suspensory type of fixation, it may be that bone with lower bone density has a lower load to failure as the button pulls through the softer bone.
While there is increasing clinical data on the ACL repair with IB augmentation, there is a paucity of studies evaluating the stiffness and failure loads of this new technique. The purpose of this study was to compare the load to failure and stiffness of ACL repair with IB augmentation versus ACL repair without augmentation. The hypothesis was that ACL repair augmented with IB would have a higher load to failure and stiffness than ACL repair alone. A secondary aim was to evaluate the modes of failure between each group and to determine if there was a correlation between bone density and load to failure in both groups.
As this study was a cadaver biomechanical study, it did not require Institutional Review Board approval under current guidelines. Twenty fresh frozen cadaver knees were used in this study with 10 in each group: ACL repair-only (RO) group or the ACL repair with IB group. Out of the 20 cadaver knees, there were 10 right knees and 10 left knees. Fourteen knees were paired from the same cadavers, whereas the remaining 6 knees came from 6 different donors. The 6 knees represented with 3 right and 3 left knees and were paired according to demographic data. Within each pair of knees, the assignment of the left vs right knee to each group was randomly selected to the RO vs IB group. For the final groups, there were 5 left knee IB versus 5 right knee RO pairs and 5 right knee IB compared to 5 left knee RO pairs. The average age of the cadaver knees was 62.2 years (SD 4.0) with 7 male and 13 female knees.
The knees were thawed at room temperature for 24 h. The soft tissue was carefully dissected, including removal of the collateral and posterior cruciate ligaments, so that the ACL remained the only attachment remaining between the femur and the tibia. The tibia and femur were cut 6 in. from the joint line to standardize the moment arm. The center of the proximal ACL footprint was marked on the femur, and the ACL was elevated from the femur using a scalpel (to simulate an ACL avulsion off the femur). Then the ACL repair or repair with IB was performed. In order to mitigate variability, all of the repairs in both groups were performed by the primary author.
2.2 Repair-only (RO) technique
In order to simulate arthroscopic conditions, the tibia and femur were placed in separate vice-clamps with the knee at 120 degrees of flexion. A #2 FiberWire (Arthrex, Naples, FL USA) was passed in a common Krackow fashion with 3 locking stitched passed down one side of the ACL to the mid-point of the ligament longitudinally, and then passed to the other side of the ligament and run back up proximally with three more locking stiches. A 2.4 mm guide-pin was passed through the footprint of the ACL with a 7 mm offset guide placed (to simulate an anteromedial approach). The two ends of the repair sutures were passed with the guide-pin, hand tensioned, and tied over a four-hole 12 mm button (Arthrex, Naples, FL USA) on the lateral femur (see Fig. 1). All knots were tied with a surgeon's knot with 5 additional half-hitches.
2.3 Repair with internal brace (IB) technique
The technique for the knees in the ACL repair with suture-tape augmentation group was performed in a similar manner to the RO group. The femur and tibia were secured in 120 degrees of flexion while a #2 FiberWire® (Arthrex, Naples, FL USA) was passed through the proximal ACL as stated above. The same 2.4 mm guide-pin was passed through the ACL footprint on the femur with the same 7 mm offset guide. In contrast to the RO technique, a 4.5 mm reamer was passed over the guide-pin. Next, a 2 mm wide polyethylene tape (InternalBrace, Arthrex, Naples, Fl USA) was passed through the loop of an ACL TightRope (Arthrex, Naples, Fl USA). The TightRope button was passed with the FiberWire from the ACL repair through the femoral tunnel and flipped on the lateral femur. The ACL TightRope was cinched so that the loop of the tape was juxtaposed to the femoral tunnel. The #2 FiberWire which had also passed through the tightrope button was tensioned and tied. For the tibial side of the InternalBrace ® (Arthrex, Naples, FL USA), a 2.4 mm tibia tunnel was drilled using an ACL guide set to 55 degrees with the aiming point set on the center of the tibial ACL footprint. The two limbs of the suture tape were then passed through the tibia tunnel (see Fig. 2). In order to set a standardized length of the internal brace and simulate a posterior drawer, the femur was held flat on a one-inch-thick polyethylene block with the tibia held flat on a table. The two limbs of the polyethylene tape were hand tensioned, then tied over a four-hole button on the tibia. After preparation of the repair, the tibia and femur were instrumented along the medial-lateral direction with a three-inch long 3/16″ stainless steel rod and potted in 3.5x3x2 inch metal boxes using a polyester resin (3 M Bondo, Maplewood, MN).
During mechanical testing, the femur was constrained to the actuator of a servohydraulic testing system (Model 8874, Instron, Canton, MA) in a flexed position relative to the stationary tibia to simulate anterior drawer testing (see Fig. 3). The specimens were preconditioned with 100 Cycles at 0.5 Hz in load ranging from 50 to 200 N, and then loaded in tension at a rate of 20 mm/min until failure. The condition of failure was defined as reduction in load equivalent to 80% of the maximal measured load. The mechanical performance of each specimen was evaluated via failure load, stiffness, and toughness.
The failure load (Fmax) was defined as the peak load measured during testing. The toughness was calculated as the energy spent to peak load (Epeak), and up to failure (Efail). The graft stiffness was characterized by values of stiffness measured as the slopes of the linear regressions built for intervals of 50 N surrounding the reference loads of 50 (S50), 150 (S150), and 300 N (S300). These three load values were used in consideration of the loads at which the ACL is reported to be subject to during gait (
). The stiffness referenced was the S150 value, as most repairs survived past 150 N.
During testing, it was noted that the primary mode of failure for the IB group occurred at the bone implant interface; therefore the decision was made to evaluate the bone density of the specimens to determine if there was a relationship between the failure and the bone density as a secondary aim. The density of the distal femur has been previously estimated through the values of CT attenuation coefficient expressed in Hounsfield Units (HU) (
). In accordance, each isolated femur was CT scanned with a LightSpeed VCT (GE Healthcare, Little Chalfont, United Kingdom) at 120 kV/100 mA with a pixel size of 0.516 mm and a slice thickness of 0.625 mm. The reconstructed, 3D images of each femur were segmented in 3D slicer (
). The density was measured for each segment with a computer model in the plane perpendicular to the anatomical axis and passing through the most proximal point of the trochlea considering 0 HU (Hounsfield Unit) as lower limit for the threshold (
). The average HU value was then computed as the average HU of the voxel included in each of the obtained segments. The bone density values were used to evaluate any differences between groups or correlation between load to failure and bone density.
An a-priori power analysis was done to determine the number needed in each group for a power equal or 1 - β equal to 0.8. Based off of the load to failure from previous studies comparing failure load of ATFL (Anterior-talofibular ligament) repair (mean 68.2 N SD 27.8) and ATFL repair with IB augmentation (mean 250.8 N SD 122.7), it was determined that the number needed for each group was 3 (
). These were the only available studies reporting failure loads of ligament repair versus ligament repair with IB augmentation.
The failure loads, stiffness and bone density were evaluated and compared using a matched student's two sample t-test with SPSS (IBM SPSS Statistics, Armonk, NY) to compare both groups. A 95% confidence interval was calculated for both the RO and IB groups. A regression analysis was performed to evaluate if there was any correlation between bone density based on HU and load to failure. A chi square was used to compare categorical differences between the two groups (
Twenty human cadaver knees were tested, 10 in the RO group and 10 in the IB group. The average age of the cadaver knees was 62.2 years (SD 4.0). The average age of the RO group was 62.3 years (SD 3.3) versus 62.1 years (SD 4.8) for the IB group with no significant difference (P = .91). There was no statistical difference in sex distribution between the RO and IB groups (4 males, 6 females in RO group, 3 males and 7 females in IB group, P = .64). The average load to failure for the IB group was 693 N (SD 248), range 430 to 1128, which was higher than the RO group load to failure of 279 N (SD 91), range 100 to 392, P = .002 (see Fig. 4). The IB load to failure 95% confidence interval was 539 N to 846 N, while the RO group was 223 N to 335 N. The stiffness was also higher for the IB group than the RO group (83 N/mm2 versus 58 N/mm2, P = .02). The energy to failure was higher in the IB group than the RO group (P = .04) (see Table 1).
Table 1Biomechanical data of ACL repair only group versus ACL repair with Internal Brace group.
The modes of failure in the RO group were 8 suture cut-out of the ligament (see Fig. 5) and 2 ACL mid-substance ruptures below or distal to the repair. In the IB group, the modes of failure were 8 due to pull through of the button through the lateral femoral cortex (see Fig. 5b), 1 pull through of both the distal button through the tibia and the proximal button through the lateral cortex of the femur, and 1 mid-substance ACL tear. There was a statistically significant difference in the modes of failure between the RO and IB group (P = .00097).
The average overall bone density of the femurs used was 238 HU (SD 65). There was no statically significant difference in the bone density based on HU between both groups, with an average for the RO group versus IB group of 245HU (SD 65) versus 233HU (SD 54), respectively, P = .68. There was a small correlation between bone mineral density based on HU versus load to failure overall, r2 = 0.26. However, only the IB group showed a strong correlation with bone density versus load to failure, r2 = 0.82 (P = .003). The repair only group had a weak negative correlation with bone density vs load to failure, r2 = −0.16 (P = .61).
While ACL repair techniques have demonstrated high failure rates historically, newer arthroscopic techniques have yielded more successful results (
). Several authors have postulated that ACL repair may offer many clinical benefits including preservation of blood supply, preservation of proprioception, potential for faster healing times, and decreased morbidity. These authors also proposed that the reason for lower failure rates may be attributable to newer instrumentation and the ability to protect the repair with suture-tape augmentation. While early clinical outcomes for ACL repair are promising, many questions remain. The current study demonstrated that the ACL repair with IB had a higher load to failure, stiffness and energy to failure compared to RO technique. Additionally, the mode of failure was different between groups, and there was a correlation between bone density and failure load for the IB group.
The forces placed on the ACL during daily physiologic activity vary. Some studies report peak forces with ground level walking to be as high as 303 N to 355 N (
). Our study showed the ACL repair alone had an average load to failure of 279 N which is lower than the peak forces placed on the ACL with walking. A recent study by van der List and DiFelice evaluated repair only of ACL proximal stump with a suture button versus a suture anchor. They showed the failure load was 310 N for the suture button repair group versus 176 N in the suture anchor group, with no difference between the two techniques (P = .144) (
). The failure loads of the suture repair with button performed by van der List and DiFelice were comparable to the RO group in our study, which was also below the peak forces during level ground walking. This may explain the high failure rates of ACL RO, as the repair sutures are likely not able to sustain the peak loads of normal walking.
An additional study performed by Hoogeslag et al. evaluated anterior tibial translation for ACL intact knees, ACL deficient knees, ACL repair with suture-tape augmentation, and ACL repair with dynamic interligamentary augmentation. While no load to failure was determined, the authors did compare an internal-brace-like construct (static augmentation) with ACL intact and deficient knees. They showed that ACL repair with static tape augmentation had anterior tibial translation which was statistically no different than the ACL intact knees and less than the ACL deficient knees (
Providing a baseline for the native ACL biomechanical properties, there are several studies on the native ACL and various suspensory constructs. A previous study comparing the Endobutton (Smith & Nephew, London, UK) to the ACL TightRope showed ultimate loads (N) of 656 N versus 749 N respectively (
). Previous studies show failure loads of native ACL and suspensory devices range from 622 N to 849 N, which is comparable to the failure load of the IB group in the current study. The mode of failure of the suspensory device study is also similar to our study where the button pulled through the cortex (
). The current study showed stiffness values of the ACL repair with IB to be 83 N which are comparable to these values for the native ACL. Higher stiffness is not necessarily better; but the ACL repair with IB may have a stiffness more similar to the native ACL than ACL repair alone.
While IB ACL biomechanical data is limited, IB has been more extensively tested on anterior talo-fibular (ATFL) repair techniques. Biomechanical evaluation of failure load and torque for IB has been evaluated by Viens et al., on different ATFL repair techniques. In their study, no significant difference was found in load to failure or stiffness between suture-tape augmented Brostom repair and the native ATFL (
). Stiffness of ATFL repair only was calculated to be 6.8 N/mm, while another study showed ATFL repair with IB stiffness was 21.1 N/mm. An additional study by Schuh et al. demonstrated ATFL repair with IB had a higher torque at failure than ATFL repair alone (11.2 Nm versus 8.0 Nm respectively, P = .04) (
). These results are comparable to the current study, where both the load to failure and stiffness were higher with IB than RO.
In the current study, the predominant mode of failure in the repair only group was pull-through of the suture through the ACL ligament. This is consistent with previous studies on ATFL repair, where all RO constructs failed at the ligament-suture interface (
). In the current study, the main failure mode of IB was pull of the button through the bone of the femur and, in one case, both the tibia and femur. A previous bovine femur study showed similar failure mechanisms for suspensory fixation where 6/6 of the cortical buttons pulled through the bone with a suture-button and 3/6 of the cortical buttons pulled through the bone with tight rope (
). In all these studies utilizing tape augmentation, the main mode of failure is at the screw or button interface with bone which is the main mode of failure of our IB group.
An additional goal of the current study was to evaluate if there was correlation between bone density and failure load of ACL repair. The average bone density in our study was 238 HU. A previous study demonstrated that normal bone density T-score of −1.0 or greater correlated to a CT measured density of 133 HU.(
) Compared to this previous study, our specimens were not osteoporotic. There was, however, a correlation between lower bone density (HU) and lower load to failure with the IB group. The IB fails at higher loads than the RO, thus, mainly fails via button pull through the bone. This may explain why the IB group showed a correlation between load to failure and bone density. Due to the fact that the failure load of the IB was much higher, it failed via the button pulled through the bone instead of suture rupture.
Additionally, a recent biomechanical study for the medial collateral ligament (MCL) compared the native intact MCL, MCL repair, MCL repair with IB and allograft reconstruction. The MCL repair with IB had 29.4% higher moment to failure than the repair alone: 95 Nm (SD 31.9) and 73.4 Nm (SD 27.6) respectively, p = .05) (
). Of note, the problem of over-tightening and stress-shielding the ATFL suture-tape augment was addressed by using a hemostat deep to the construct while securing the anchors. This process is not easily performed arthroscopically for ACL repair, and presents a technical challenge when performing IB. The ideal length or tension for the internal brace has not been described. For the current study, the internal brace was secured using a standard protocol in order to set a consistent length. Proper tensioning maneuvers for suture-tape augmented ACL repair warrants further evaluation to determine the effects on healing, and over-tensioning. The early clinical data from recent reports have shown successful early outcomes. A retrospective study of 56 patients showed a 14.3% rate of failure with repair alone versus 7.1% with repair and suture augmentation (
). An additional case series on 68 patients who received ACL repair with IB showed successful outcomes at one year with only one failure. While early clinical data is promising, ACL repair with IB augmentation should be done selectively. Van der list and DiFelice recommend reserving repair for proximal avulsions in the acute setting, when good ACL tissue is present. Additionally, when performed, the surgeon should be ready to perform a reconstruction if tissue quality or location prohibits direct repair. While protecting the repair with the IB may prevent the repair from failing during physiologic loads, this does not necessarily translate into better healing. There are still concerns that, clinically, the internal brace may overconstrain the knee. While our study showed stiffness values for the internal brace similar to previous reported values on the native ACL, further studies should be done to evaluate the effect of overtightening the internal brace. Future analysis should evaluate how to properly tension the IB to recreate the normal function of the knee. Additionally, future studies should be done to evaluate the effects of IB on the healing of ACL tissue.
The following limitations regarding this study are described. First, this study was not performed arthroscopically. It was necessary, during the preparation of a proximal femur ACL avulsion to open the knee and carefully resect it from the footprint. Although arthroscopic instruments and positioning of the specimens provided simulated arthroscopic conditions, it remains unclear how arthroscopy would affect these repairs biomechanically. Secondly, the average age of our sample was 62 years and consisted of a 13:7 female-to-male ratio. The concern with using an older female model for ACL repair testing is the potential impact of osteopenia on the testing data. We were able to address this with bone mineral density correlation to failure for the IB group. However, this does not truly reflect the younger population that ACL repair procedures would typically be performed in. Variations in the data due to the bone interface with the IB group may also be further compounded by issues of tunnel trajectory. Depending on specimen anatomy, femoral tunnel trajectory may result in obliquity as it pierces the extra-articular cortex. The result is an ovoid tunnel opening upon which the button seats. Additional studies are warranted to determine the best method and position of femoral tunnel placement for maximal pull-out strength. Also, the RO technique was tied over a simple button; while the IB augmentation technique was performed with a tightrope button. This was done in order to re-create previously described techniques (
). Nevertheless, despite these limitations in button placement and usage, the IB group still had higher failure loads than the RO group. Finally, the aim of this cadaveric study was solely a determination of failure load and stiffness. Further studies should be done to determine clinical relevance when assessing differences between ACL repair and ACL repair augmented with IB.
There was higher load to failure for the ACL repair with IB compared to ACL repair alone. There was a higher stiffness and energy to failure for the ACL repair with IB compared to ACL repair alone. There was also a difference in the mode of failure between both groups with a high positive correlation between bone density and load to failure for the ACL repair with IB group.
Author contributions statement
The following six authors were instrumental in the development of this manuscript in the following manner:
Patrick Massey, MD: Wrote the introduction, methods, results, and conclusion, developed the figures and table, and edited the manuscript.
David Parker, MD: Contributed to the discussion, modified the figures and edited the manuscript.
Kaylan McClary, MD: Wrote the abstract, contributed to the discussion, edited the manuscript.
James Robinson: Created the reference section, applied referencing throughout the paper, and edited the manuscript.
R Shane Barton, MD: Contributed to the discussion and editing of the manuscript.
Giovanni Solitro, PhD: Contributed to the methods and results section of the manuscript.
All authors have read and approved the final submitted manuscript.
Declaration of Competing Interest
This study was funded by a grant from Arthrex for cadaver specimens, supplies, and testing costs. Patrick Massey, David Parker, Kaylan McClary, James Robinson, R. Shane Barton, and Giovanni Solitro declare that they have no conflict of interest other than the grant described above.
This article does not contain any studies with human participants performed by any of the authors.
We would like to acknowledge that this study was funded by a grant from Arthrex for cadaver specimens, supplies, and testing costs.
Arthrex ACL TightRope and Biomet ZipLoop with ToggleLoc: Mechanical Testing.