Anterior Cruciate Ligament Reconstruction: A Comparative Clinical Study Between Adjustable and Fixed Length Suspension Devices

Abstract

Background:

Adjustable-length cortical suspension devices provide technical advantages over fixed-length devices for femoral graft fixation during anterior cruciate ligament (ACL) reconstruction but have shown increased lengthening during cyclic loading in biomechanical studies. The purpose of this study was to prospectively measure graft elongation in vivo along with patient reported outcomes.

Methods:

Thirty-seven skeletally mature patients diagnosed with anterior cruciate insufficiency who underwent ACL reconstruction using autogenous hamstring graft were included in this study. Thirteen patients received an ACL reconstruction using a fixed loop device (FL) and twenty-four patients were treated with an adjustable-length device (AL) based on surgeon preference. Bilateral knee laxity was measured with a KT1000 Arthrometer before surgery and immediately after surgery with the patient under anesthesia, and at the 6-week, 3-month, and 6-month clinical follow-up appointments. All measurements were made by the same operator with maximum force testing. Differences between the affected knee and the contralateral knee were measured. Patient reported outcomes were collected at 6 and 24 months post-operatively.

Results:

No difference was found between the FL and AL groups in either knee laxity or patient reported outcomes. Average side-to-side difference at 6 months was 1.8 ± 2.6 mm for the FL group and 1.7 ± 2.4 mm for the AL group (p=.874). One patient in the FL group (7.7%) and two in the AL group (9.5%) had a side to side difference in laxity greater 5 mm. Patient reported outcomes did not differ between groups and no patients underwent revision surgery.

Conclusions:

The adjustable-length cortical suspension device (AL) did not demonstrate increased laxity as compared to fixed-length devices. There was no difference in patient reported outcomes between the groups.

Level of Evidence: IV

Introduction

The outcome of anterior cruciate ligament (ACL) reconstruction depends on many surgical factors including tunnel position, graft tension, and fixation techniques.¹.

Anatomic positioning of femoral and tibial bone tunnels attempts to approximate native knee kinematics, allowing appropriate graft tension throughout range of motion.² Along with maintaining appropriate tunnel positioning, graft-fixation devices must provide sufficient fixation to ensure that graft tension is maintained until incorporation of the graft to native bone.^3,4 Hamstring tendons are one of the most popular grafts used for ACL reconstruction; however, controversy exists about the best graft fixation.

Common soft tissue ACL reconstruction femoral fixation implants include interference screws, cortical suspension devices, and cross pins. Previous work has shown trends in mechanical behavior for most of the fixation mechanisms, with positive clinical outcomes for all three types of devices.⁵ Cortical-cancellous suspension fixation with transcondylar devices seems to offer the best results in terms of graft elongation, fixation strength, and stiffness⁴, but tends to show high rates of intra- and postoperative complications.^5-7 Interference devices may allow for relatively higher graft slippage and failure at lower ultimate loads when used to secure hamstring tendon grafts.^4,5,8

Fixed-length cortical suspension devices have been shown to be a good option for soft tissue graft fixation in biomechanical studies in terms of limiting graft slippage and providing sufficient fixation strength.^3,9,10 However, there are technical challenges in measuring and inserting the device that may result in insufficient graft length in the femoral tunnel for incorporation. This is aggravated by shorter femoral tunnels with anatomic femoral tunnel placement, increasing the risk for insufficient graft length in the femoral tunnel for incorporation.⁵

Adjustable-length suspension can provide greater ease of insertion by obviating the need to calculate the loop length, allows complete graft fill of the femoral tunnel, and allows the same implant to be used regardless of tunnel placement or depth.^9,11 The potential disadvantage of the adjustable-length design is loop lengthening after fixation, which can lead to graft loosening and consequently surgical failure.⁵

The objective of this study was to compare side-to-side knee laxity and outcomes after primary ACL reconstruction with hamstring autograft between adjustable-length or fixed-length cortical suspension devices. We hypothesized that there would be increased knee laxity in the adjustable-length suspension as compared to fixed-length suspension, but no difference in patient reported outcomes or failure rates.

Methods

Institutional review board approval was obtained at the University of Iowa for this prospective non-randomized case control study comparing groups of patients undergoing ACL reconstruction surgery with either fixed-length (FL) or adjustable-length femoral cortical suspension devices. The fixed-length device in this study was the EndoButton (Smith&Nephew, Andover, Massachusetts), and the adjustable-length device was the TightRope (Arthrex, Naples, Florida).

The study included all skeletally mature patients who were diagnosed with anterior cruciate insufficiency and elected to undergo ACL reconstruction using an autogenous hamstring graft between July 2014 and March 2015. Exclusion criteria were previous knee surgery, multi-ligament injuries, contralateral ACL deficiency or reconstruction, or connective tissue disorders.

Patients consenting to participate in the study had their knee laxity measured in both their reconstructed knee as well as with side-to-side comparison measured with a KT1000™ Arthrometer (MEDmetric, San Diego, California) at maximum force by a single trained orthopaedic surgeon who was present for each surgical case. Measurements were done immediately before and after surgery with the patient under anesthesia (either general or spinal), and then postoperatively at six weeks, three months and six months after surgery. All measurements were done by the same surgeon to reduce bias caused by the low inter-rater reliability of the KT 1000 arthrometer.^12,13 These timepoints were all measured as an attempt to cover the initial rehabilitation period through tunnel incorporation and healing. Differences in laxity were compared using t-tests.

The study included the patients of four fellowship trained surgeons in our sports medicine group (AA, MJB, CMH, BRW). The patients were divided into two groups, Fixed-Length (FL) or Adjustable-Length (AL), based on surgeon preference. One surgeon used only FL, two only used AL, and one used both. The surgeon preferring the fixed-length fixation device used a medial portal technique and the other three surgeons utilized an outside-in technique, using a Flipcutter (Arthrex Inc., Naples, Florida). Femoral and tibial tunnels were created within the center of the native femoral and tibial ACL footprints. Both adjustable-loop or fixed-loop cortical suspension devices were used according to the manufacturer’s guidelines. The grafts were manually tensioned. Tibial fixation consisted of an interference screw, except for three patients, in which an All-Inside® technique (Arthrex Inc., Naples, Florida) was used with double femoral and tibial adjustable loop fixation.

Patient reported outcomes (PROs) were collected preoperatively as well as at six and twenty-four months post-operatively. PRO’s included the Marx Activity Scale, Knee Injury and Osteoarthritis Outcome Scale (KOOS), Short Form-36 (SF-36), and EQ5D. PROs were analyzed using t-tests to compare baseline to 6 month, and 24 month scores between FL and AL groups. Change scores (baseline to 6-month and baseline to 24-month timepoints) were also calculated independently for each of the KOOS subscales and compared with t-tests, which allowed the highest sensitivity for detecting any differences between groups. KOOS subscales were also compared with Fisher exact tests to analyze the percentage of patients who achieved the threshold patient acceptable symptomatic state (PASS) for ACL reconstruction as described by Muller et al.¹⁴

Results

A total of thirty-seven patients were enrolled. Thirteen patients (35.1%) were in the FL group: seven females, eight right knees. AL group consisted of twenty-four patients (64.9%): thirteen females, eleven right knees. Of all patients, seventeen had concomitant surgery (thirteen in the AL group and four in the FL group) for a total of twenty-one procedures, as detailed further in Table 1. Three patients of the AL group were lost to follow-up after their initial pre and postoperative testing, resulting in follow-up rates for the 24 month PROs of 100% in the FL group and 87.5% in the AL group.

Table 1.

Demographics, Tunnel Diameter, and Concomitant Procedures

	AL Group	FL Group	P valuea
N	24	13
Age	31.3 ± 11.7	33.9 ± 11.2	0.510
Gender Female	13 (54%)	7 (54%)	0.985
Laterality Right	11 (46%)	8 (62%)	0.570
Femoral Tunnel Diameter	9.3 ± 0.8 mm	8.4 ± 0.7 mm	0.002
Patients with Concomitant	13 (54%)	4 (31%)	0.173
Procedures
Medial Partial Meniscectomy	5 (21%)	2 (16%)
Medial Meniscus Repair	2 (8%)	2 (15%)
Lateral Partial Meniscectomy	5 (21%)	1 (8%)
Lateral Meniscus Repair	0 (0%)	1 (8%)
Plicectomy	1 (4%)	0 (0%)
Loose body removal	1 (4%)	0 (0%)
Patellar Chondroplasty	1 (4%)	0 (0%)

^aP values calculated using t-tests and chi-squared tests

There was no significant difference between maximum side-to-side KT-1000 arthrometer testing immediately after surgery and at six months postoperatively between groups (Figure 1). The FL group showed an increase of 2.5 ± 2.3 mm (median 2.0 mm) from postoperative testing to testing at six months, and the AL group showed an increase in 2.8 ± 2.8 mm (median 2.0 mm) with a p-value of 0.759 between the two groups. One patient in the FL group (7.7%) and two in the AL group (9.5%) showed a 6-month side-to-side difference in laxity over 5 mm (p=.855). No patients underwent revision surgery.

Figure 1.

Side-to-side differences in KT-1000 arthrometer measurements (brackets represent 95% confidence intervals; p values from independent samples t test).

Preoperatively, the FL group had significantly higher side-to-side laxity (6.5 mm vs 4.7 mm, p=.034). Both groups had less laxity than the contralateral side immediately after surgery, with laxity progressively increasing at six weeks, three months and six months, although no post-operative timepoint had significant side-to-side differences between FL and AL (Figure 1). Average side-to-side difference at six months was 1.8 ± 2.6 mm for the FL group and 1.7 ± 2.4 mm for the AL group (p=.874).

Six and 24-month questionnaires were completed by 81.1% and 75.0% of patients respectively. Table 2 summarizes differences in PROs at preoperative baseline, six months, and 24 months. No significant differences were noted between FL and AL groups at baseline. The only significant difference noted at six months was a higher KOOS Quality of Life score in the AL group (67.4 vs 50.3, p=.032). However, both groups were again found to be similar at 24-month follow-up with no significant differences identified for any PROs. Marx activity decreased at both time points but was similar for both groups. KOOS scores were calculated for FL and AL groups and are summarized in Table 3. Analysis showed both similar baseline scores and improvement. Additionally, similar percentages of each group achieved the patient acceptable symptomatic state for each of the KOOS subscales, as shown in Table 4.

Table 2.

Patient Reported Outcomes at Baseline, 6 Months, and 24 Months^a

	Baseline			6 Month			24 Month
	FL	AL	P value	FL	AL	P value	FL	AL	P value
KOOS Symptoms	51.5 ± 15.5	54.5 ± 15.9	.595	71.1 ± 16.6	71.4 ± 15.1	.560	79.1 ± 18.5	80.6 ± 13.9	0.808
KOOS Pain	62.0 ± 11.9	61.3 ± 21.3	.919	77.5 ± 15.3	83.8 ± 11.1	.239	91.0 ± 5.3	89.3 ± 13.0	0.738
KOOS ADL	70.5 ± 12.0	71.7 ± 18.9	.858	85.9 ± 14.5	92.2 ± 9.3	.165	97.8 ± 2.1	95.6 ± 5.7	0.149
KOOS Sport	21.7 ± 17.8	26.3 ± 22.0	.533	61.7 ± 24.1	63.2 ± 19.3	.847	79.4 ± 22.3	80.3 ± 20.4	0.921
KOOS QOL	31.6 ± 16.8	35.4 ± 18.2	.549	50.3 ± 26.0	67.4 ± 14.5	.032	75.8 ± 18.7	74.0 ± 18.4	0.823
Marx Activity	11.3 ± 5.0	12.7 ± 3.7	.379	7.7 ± 5.3	9.3 ± 5.6	.485	6.0 ± 5.5	6.5 ± 4.1	0.785
EQ5D	80.1 ± 14.0	82.3 ± 13.1	.644	77.1 ± 14.8	85.2 ± 9.1	.082	77.8 ± 9.1	84.8 ± 9.6	0.165
SF-36 PCS	38.5 ± 13.4	42.1 ± 11.7	.450	47.4 ± 7.8	50.5 ± 6.1	.237	55.7 ± 4.8	56.9 ± 4.6	0.569
SF-36 MCS	57.7 ± 6.0	56.1 ± 7.1	.547	54.1 ± 6.3	54.7 ± 6.3	.818	50.9 ± 6.3	52.5 ± 5.6	0.506

^aReported as mean ± SD; P values calculated using independent samples t-tests. ADL = activities of daily living. QOL = quality of life. PCS = physical component summary. MCS = mental component summary

Table 3.

KOOS Change Scores at 6 and 24 Months from Preoperative Baseline^a

	FL	AL	P value
KOOS Symptoms
Baseline	51.5 ± 15.5	54.5 ± 15.9	0.595
6-month change	+21.4 ± 18.4	+15.8 ± 22.5	0.492
24-month change	+23.3 ± 9.4	+23.7 ± 11.1	0.924
KOOS Pain
Baseline	62.0 ± 11.9	61.3 ± 21.3	0.904
6-month change	+18.7 ± 13.6	+20.6 ± 20.1	0.786
24-month change	+25.7 ± 14.9	+25.7 ± 14.3	0.993
KOOS ADL
Baseline	70.6 ± 12.0	71.7 ± 18.9	0.858
6-month change	+19.8 ± 17.1	+17.7 ± 18.9	0.748
24-month change	+29.8 ± 13.6	+21.4 12.8	0.137
KOOS Sport
Baseline	21.7 ± 17.8	26.3 ± 22.0	0.533
6-month change	+41.8 ± 25.8	+34.7 ± 26.3	0.488
24-month change	+59.4 ± 23.8	+53.4 ± 24.6	0.568
KOOS QOL
Baseline	31.6 ± 16.8	35.4 ± 18.2	0.549
6-month change	+20.5 ± 27.4	+32.2 ± 24.5	0.246
24-month change	+46.6 ± 13.0	+37.3 ± 22.8	0.291

^aReported as mean ± SD; P values calculated using independent samples t-tests. ADL = activities of daily living. QOL = quality of life

Table 4.

Percentage of Patients Who Achieved the Patient Acceptable Symptomatic State (PASS) by 24 Months for Each of the KOOS Subscales

	FL	AL	P valuea
KOOS Symptoms (57.1)	87.5%	100%	0.296
KOOS Pain (88.9)	75%	63.2%	0.676
KOOS ADL (100)	37.5%	31.6%	0.999
KOOS Sport (75)	87.5%	73.7%	0.633
KOOS QOL (62.5)	75%	74.1%	0.999

^aP values calculated using Fisher exact tests. PASS for each subscale listed in parentheses

Discussion

The aim of our study was to prospectively investigate the in vivo effects of different types of cortical suspension in patients undergoing ACL reconstruction using hamstring autograft. Our study found that there were no significant differences in side-to-side knee laxity in patients using fixed-length versus adjustable-length cortical fixation devices at any timepoint. Additionally, we observed no differences in PRO scores at six months and twenty-four months following surgery.

Previous in vitro studies have shown increased laxity with adjustable-length fixation devices, and this effect of cyclic loading on these devices is an area of clinical concern, due to potential elongation during the acute postoperative period leading to decreased graft tension as well as graft slippage within the bone tunnel. These can both negatively affect postoperative healing and clinical outcomes because of increased knee laxity.^3,15

Petre et al.¹⁰ showed a significantly higher displacement for adjustable-length devices (2.74 ± 0.39 mm for TightRope and 3.34 ± 1.28 mm for Biomet ToggleLoc) than for fixed-length devices (1.82 ± 0.23 mm for XO Button (Linvatec), 1.88 ± 0.25 mm for Endobutton).¹⁰ The clinically relevant amount of construct lengthening is not known; however, ACL failure has been described as greater than 3 mm of difference between side-to-side anterior tibial translation; therefore, 3 mm of loop lengthening would cause clinical failure.¹⁶ Although their study was not designed to directly compare fixed- and adjustable-length fixation devices, it raised awareness about possible differences between implant types.

A subsequent study directly addressing fixed versus adjustable-length fixation was performed by Barrow et al.⁵ and showed an increase in loop lengthening during cyclic testing in the adjustable-length devices. This was partially caused by suture slippage in the adjustable-length loop. The Arthrex TightRope reached clinical failure of 3 mm lengthening after fewer cycles than the Biomet ToggleLoc (2576 ± 73), while the Smith & Nephew EndoButton did not reach clinical failure during cyclic testing. With the free suture ends tied, after 4500 cycles, the Arthrex TightRope had a significant decrease in lengthening but not sufficiently to prevent failure.⁵ The elongation of the devices is particularly concerning given the relatively small loads under which it began to occur. At a normal pace, the native ACL is subjected to forces of 303 N during level-ground ambulation and up to 445 N during walking on a downhill slope,^17,18 which are higher than the forces at which elongation occurred during these studies.

A possible explanation for these results is the unloading of the suture during cyclic preconditioning at 10- 50N. The literature shows that there is complete unloading of the ACL during certain phases of normal walking or rehabilitation. Biomechanical cadaveric studies have shown that the in situ loads on the whole ACL during passive flexion between 0° and 90° is less than 10 N,¹⁹ approaching 0 N.^20,21 Measurements in cadavers and in vivo have shown that the strain levels for both bands of the ACL were at or below zero between 10° and 110° of passive flexion.^21-23 A biomechanical study using live quadruped models showed that ACL forces dropped to zero during the swing phase in all trials,²⁴ and studies based on 3D calculated models also show the ACL strain dropped to zero during the first portion of the swing phase.¹⁶ Thus, if the unloading of the sutures produces suboptimal fixation strength and increases slippage, this could yield negative clinical outcomes.

One potential criticism of the biomechanical studies is the direction of the applied load. Parallel loads to the tunnel axis are likely to produce higher forces than the actual forces that the implant would see in vivo. This is in accordance with most previous biomechanical studies, but differs from clinical conditions.⁹ These testing protocols subjected the implants and specimens to more strain than they would see in vivo. The results can show differences in the performance of the implants, but the clinical differences may be less substantial.¹¹ Additionally, increased laxity in the FL group could be the result of in vitro studies representing only a portion of the overall ACL graft displacement. In vivo, the soft tissue grafts are also subject to graft elongation, graft slippage, or cutout of the femoral bone, all of which probably contribute to further displacement of the ACL construct, resulting in an increase in laxity greater than the one reported for FL in biomechanical studies.

More recent literature supports our findings, with adjustable-length devices achieving comparable results to fixed-length devices.²² The effects of graft lengthening in a patient population was investigated in a retrospective study by Boyle et al,²³ comparing adjustable-loop and fixed-loop cortical suspension devices. The study included a consecutive series of 188 patients (73 AL and 115 FL). Their study showed no significant differences in maximum side-to-side difference in KT-1000 testing at six months of 1.51 mm (AL) vs. 1.79 mm (FL). They also observed no significant difference between the two groups in the rate of graft failure (10% vs. 11%, p = 0.71). Similarly, our study showed no significant difference in laxity between the two groups at any time point following surgery. Both groups had less laxity than the contralateral knee immediately after surgery, with the grafts progressively loosening over the post-operative course.

A study by Choi et al. retrospectively compared clinical and radiographic outcomes between the Arthrex TightRope and the Smith & Nephew EndoButton.²⁵ The study found no difference in radiographic femoral or tibial tunnel widening and no difference in Lysholm scores or Tegner activity scale.

Similarly, our study found no difference in patient reported outcomes between TightRope and EndoButton. Baseline, 6-month, and 24-month scores showed no difference in outcomes based on KOOS, Marx activity, SF-36, and EQ5D scores. Change in KOOS and the 24-month PASS rates were also similar between the groups, further suggesting that implant choice has no effect on the rate of improvement and clinical outcomes.

Limitations of our study include only gathering KT data until the 6-month time point. We did not collect KT data past this point because soft tissue incorporation should occur between 8-12 weeks after surgery,^11,26 with progressive soft tissue to bone healing up to six months postoperatively.²⁷ Any further increase in laxity after this point is not likely to result from femoral cortical suspension slippage. We also do not know the causes for the three patients that had laxity of greater than 5 mm. These did not present until their 6-month clinical follow up, and the patients did not report any increased laxity or known injury. The relatively small size of the study may also be a limitation, although we felt that collection of KT 1000 data by a single surgeon was a more important factor than enrolling a larger cohort with multiple surgeons measuring, due to the low inter-rater reliability of KT 1000 measurements. Therefore, these findings should be interpreted within the context of limitations with the KT 1000 arthrometer. Further, the study was non-randomized, and implant choice along with technique was dictated by surgeon preference, leading to a larger number of patients treated with adjustable-length implants. No patients in this study required revision surgery.

Conclusion

There was no difference in side-to-side knee laxity between adjustable and fixed-length suspension devices in vivo, indicating that the adjustable-length devices are safe and effective, withstanding loads from the early rehabilitation period. Patient reported outcomes improved in both groups compared to pre-operatively and were equivalent at 6 and 24 months post-operatively. Therefore, implant type should be dictated by the operating surgeon based on personal preference and expertise.