Abstract
Background The Tape Locking Screw (TLS®) system for anterior cruciate ligament reconstruction (ACLR) is an emerging short graft with tape suspension (SGTS) technique which uses a single hamstring tendon with tape suspension and dedicated interference screw fixation. Few studies have examined the SGTS technique versus other techniques. The objective of this study was to compare LOE in patients undergoing ACLR with the SGTS technique with other ACLR techniques. We hypothesised that the SGTS technique would be superior to others in terms of minimising LOE.
Materials and methods We retrospectively reviewed 138 patients who received primary ACLR between January 2015 and December 2017 and had elected to participate in a rehabilitation program with the hospital physiotherapy department had postoperative knee extension assessed. Patients were followed up by a department physiotherapist until baseline function was achieved. Patients were grouped as having either SGTS ACLR or non-SGTS ACLR during chart review and compared for LOE at initial assessment and when the maximum extension had been achieved.
Results There were 44 patients in the SGTS group and 94 patients in the non-SGTS group. The SGTS group had a significantly larger median graft diameter (8.5mm vs. 8.0mm, P <0.001) and less incidence of notchplasty (17.8% vs. 44.7%, P <0.001). LOE incidence was significantly associated with the use of notchplasty and graft type at initial assessment. At the time of maximum extension, LOE incidence was significantly associated with LOE at initial assessment, graft size, graft type and lateral meniscus injury. Time to maximum extension was significantly associated with the presence of LOE at initial assessment.
Conclusion In this retrospective analysis, the SGTS technique was not inferior to other ACLR techniques with respect to postoperative loss of extension. Longer time to achieve full extension in the SGTS group was not considered clinically significant. Further studies are needed to assess longer term success and functional outcomes.
1. Introduction
Anterior cruciate ligament (ACL) rupture is a common knee injury affecting both the active, athletic population as well as the general population [1]. While the condition can be managed nonoperatively in lower demand patient groups, for those wishing to return to recreational or elite level sport, a reconstruction procedure (ACLR) is recommended [2]. There are numerous methodologies in use for ACLR which vary in graft type, preparation and fixation. Despite extensive investigation, outcomes vary widely and consequently there is still debate on the optimal ACLR technique [3]. Several common complaints have been reported post-ACLR, including subjective instability, anterior knee pain and failure to return to pre-injury level of function [4]. One other common post-operative complaint after ACLR is reduced range of knee motion (ROM), in particular loss of extension (LOE), which may present a greater functional limitation than pre-operative instability [5]. Previous studies have investigated factors affecting LOE, including surgical technique, however few have been able to definitively identify significant contributors to maximising knee ROM [5–7].
The Tape Locking Screw (TLS®) system for ACLR is an emerging technique which uses a four or five strand single hamstring tendon with polyethylene terephthalate (PET) strips at both ends, fixed with dedicated TLS® interference screws [8]. We describe this technique as a short graft with tape suspension (SGTS). The proposed benefits of this technique include requiring only a single hamstring tendon which can be passed through smaller bone tunnels [9]. It has also been suggested that the dedicated interference screw gains better traction against the PET strip compared to soft tissue graft, which may reduce slippage of the graft-tape construct [10]. Finally, the technique involves fixing the graft in near-full knee extension which theoretically reduces risk of over-constraint [11].
Since the earliest reports of the TLS® technique in 2003, there have been few publications studying outcomes after surgery. Comparison of knee extension deficit after ACLR using a SGTS technique versus other traditional techniques has not been reported at this time. The objective of this study was to compare LOE in patients undergoing ACLR with the SGTS technique with other ACLR techniques at follow-up during the rehabilitation period. We hypothesised that the SGTS technique would be superior to other techniques in the incidence of LOE.
2. Materials and methods
2.1 Study setting
This was a retrospective dual-centre study of ACLR performed on patients between January 2015 and December 2017 at two metropolitan hospitals offering Government/Public healthcare (Queen Elizabeth II Jubilee Hospital [QEII] and Mater Private Hospital Springfield [MPHS]). All patients had presented to QEII initially. Patient lists were sourced from a previous dataset (ACL Model of Care) with patients confirmed to have participated in rehabilitation at the QEII physiotherapy department, which coincided with the period that the SGTS technique had been introduced. A second patient list was sourced from the department clinical registry (SHARKS, ACTRN12617001161314) to include patients not in the Model of Care dataset. Chart review was used to confirm the status of the second patient list against the inclusion/exclusion criteria.
Inclusion criteria:
Electing to undergo primary ACL reconstruction at either hospital
Participation in post-reconstruction rehabilitation at the QEII hospital outpatient physiotherapy department.
Isolated ACL injury +/- meniscal injury only
Exclusion criteria:
Revision ACLR
Concurrent surgical treatment of other ligamentous injury (PCL, LCL, MCL, PLC)
Receiving physiotherapy external to the hospital outpatient department.
As per the hospital’s usual practice, new patient referrals were triaged by a consultant orthopaedic surgeon and assigned to a lower limb surgeon. The decision to proceed to surgery was made between the surgeon and patient and was not protocolised. From July 2017, patients booking for surgery were also captured in the department clinical registry (SHARKS) and consented using a written-informed opt-in approach. Ethical approval was granted for the registry by the Metro South Health Human Research Ethics Committee (HREC/16/QPAH/732).
2.2 Surgical Technique − Tape Locking Screw
During the period of interest, 14 surgeons performed ACLR of which two used a SGTS technique. There were 44 patients in this group. All patients were examined under anaesthesia. Tendon(s) were harvested prior to commencing arthroscopy. The standard TLS® surgical technique was used and has been described in detail by Collette and Cassard [8]. The Semitendinous tendon was harvested in isolation and Gracilis was used only in the event of graft failure or insufficient graft material. The graft length was set 50 to 55mm. Prior to implanting the graft, arthroscopic examination of the knee was undertaken. Any meniscal injuries were addressed using an all-inside technique repair or resected to stable margins. The prepared graft was inserted into the knee via the anterior-medial portal. Once the graft was adequately positioned and fixed into the femur, a second interference screw was inserted with the knee in full extension. The knee was then examined arthroscopically in full extension to evaluate the intercondylar notch and notchplasty was performed if required.
2.3 Surgical technique – Others
The remaining 94 patients had ACLR by 12 surgeons using non-SGTS techniques. These comprised of 69 femoral closed-loop button/tibial interference screw, 18 femoral cross-pin/tibial interference screw and 7 femoral and tibial interference screws. All patients had hamstring (either Semitendinosus or Semitendinosus and Gracilis) autografts from the ipsilateral leg. Grafts were fixed with the knee in 20-30° flexion. Meniscal pathology was addressed intraoperatively either by partial meniscectomy or primary repair. Notchplasty was performed as required. Surgical approach and closure were as per surgeon preference.
2.4 Rehabilitation
Following ACLR, patients at both hospitals were reviewed by the ward physiotherapist with identical post-operative instructions and referred for outpatient physiotherapy at QEII hospital. The aim was for all patients to have an initial assessment within 10 days of surgery. Frequency of appointments was based on patient progress. We used the four-phase, goal-directed program described in the Randall Cooper ACL Rehabilitation Guide [12]. Patients were discharged once they achieved baseline function and were provided with written information regarding ongoing rehabilitation, risk of re-injury and osteoarthritis. Knee ROM was collected and documented at each physiotherapy session. ROM was measured using a standard universal goniometer with the patient seated. Patients were deemed to have LOE if the deficit was more than five degrees compared to the non-operative leg.
2.5 Data collection
A list of all patients who had a primary ACLR at QEII and MPHS between January 2015 and December 2017 was sourced from hospital administrators. Patients who had participated in physiotherapy at QEII following surgery were included in the review. Operative notes from those patient charts were used for ACLR technique and arthroscopic findings. Physiotherapy examination and progress were also recorded in the patient charts.
2.6 Data and Statistical Analysis
Data were retrieved from a sample of convenience, based on clinical volume before and after the introduction of SGTS technique, changes to electronic medical records processes and data availability. At the conclusion of the chart review, the spreadsheet was locked to further editing and prepared for analysis. Responses were recorded for graft type to standardize into response groups (ST, STG). Extension angle (deg) was dichotomized into loss of extension (yes; no) at ≥5° at both initial and maximum extension time points. The technique groups (SGTS; non-SGTS) were compared with univariate analyses, Mann-Whitney U tests for continuous variables and Fisher’s exact test for categorical data. Binary logistic and multivariable linear regression mixed-models were constructed to assess the relationship between patient (random factor), pathology and treatment factors on knee extension outcomes during the rehabilitation period (Table 1).
Model characteristics to assess factors associated with extension outcomes following ACL reconstruction
3. Results
3.1 Patient characteristics and group comparison
A sample of 138 patients who met the inclusion criteria were included in the final analysis (Figure 1). All patients had a maximum extension recorded (no loss to follow up) prior to termination of the rehabilitation program (patient self-select, therapist discharge, failed to attend). The sample were majority male, aged in mid-20s with a normal BMI (Table 2). Missing data rates ranged from 0 − 10%. There were significant (P<0.05) differences between SGTS and non-SGTS cases for gender, surgery delay, graft type, graft size, notchplasty and surgery to first physio appointment period (Table 2).
Summary characteristics of patients sampled in present analysis (median with interquartile range)
STROBE [13] flow chart of inclusion, exclusion and analyses.
The probability of a patient having no LOE recorded at initial assessment (32.6% of groups combined) was explained by a model with adjusted R2 of 13.7%, with positive associations to an absence of a notchplasty, ST grafts, higher BMI and a recorded injury/reconstruction to the contralateral limb (Table 3). There was a significant difference in the incidence of ST-only grafts in the SGTS group (71.1% of SGTS vs. 4.4% of non-SGTS, P < 0.001). The probability of LOE at the time of achieving maximum extension (18.4% of groups combined) was explained by a model with adjusted R2 of 18.5%, with STG grafts, receiving SGTS technique, notchplasty, lower BMI, no injury to the lateral meniscus and LOE at initial measurement all risk factors identified in the final model (Table 3). The variance in time between initial appointment and maximum extension achieved (median 3 weeks, IQR 0 − 8.1) was explained by (adjusted R2 23.7%) a mixed linear model with graft type (ST) (4 [95%CI 0.3 − 7.6], P = 0.033) and LOE at initial assessment (4.9 [2.7 − 7.1], P = <0.001) as significant factors (Figure 2).
Summary of binary logistic regression results for LOE at Initial and Maximum Extension assessments
Main effects on initial appointment to maximum extension (weeks) following ACLR. Fixed Flexion − loss of extension >5° at initial appointment.
4. Discussion
4.1 Summary
The primary objective of this study was to identify if the emerging SGTS technique would yield improved post-operative outcomes with respect to LOE after ACLR compared to other common techniques. Our hypothesis, that the SGTS technique would be superior to others in terms of LOE, with an overall incidence of 18.4% at final followup, was rejected. Receiving the SGTS technique was not a significant risk factor for LOE at initial assessment, but was at time of maximum extension, when controlled for other patient and management factors. Our results are consistent with previously reported LOE after ACLR (Table 4). Previous studies have attributed LOE to, among others, graft placement and fixation, impingement within the intercondylar notch, immobilisation prior to rehabilitation and early ACLR post-injury [5–7]. There is notable variability in LOE within all timepoints in Table 4, reinforcing the concept that LOE is multifactorial. Despite the varied incidences of LOE, however, there is a general trend of ongoing reduction in LOE incidence with longer term follow up.
Summary of literature reporting incidence of loss of extension after ACLR
4.2 Knee position during graft fixation
We had anticipated that fixing the graft with the knee in full extension would prevent over-constraint and allow greater ROM post-operatively. Austin et al reported in a cadaveric study that there was significantly greater LOE when the graft was fixed with the knee in 30° flexion compared 0° flexion [11]. They concluded that the ACL is under most tension in full extension and therefore if fixed in this position, the risk of over-constraint would be reduced. Asahina et al. performed an in vivo study investigating knee flexion angle at ACL graft fixation and found that when fixed at 0° flexion, there were significantly lower extension deficits compared to 30° knee flexion [18]. These findings were not reproduced in our study.
One notable difference between the two groups was the time to achieve full extension. The SGTS group took an average 6.4 weeks (range 1 – 14.6 weeks) from initial physiotherapy evaluation to achieve maximum knee extension, compared to only 2.5 weeks (range 0 – 6.7 weeks) in the non-SGTS group. The SGTS graft usually consists of only the Semitendinosus tendon which is then looped to form four to five strands. When using a combined Semitendinosus and Gracilis graft, each tendon is usually double-looped to form a longer 4-strand graft. The additional strands of single tendon could reduce the tensile load per strand and subsequently reduce stretch after implantation. We hypothesise that the shorter graft used in the SGTS group is stiffer than other hamstring grafts. Regarding graft size, we found that the average diameter was 8.5mm in the SGTS group versus 8mm in the non-SGTS group, which was statistically significant. Biomechanical studies have found a direct correlation between graft diameter and resistance to stretch [19]. It is possible then that the shorter, thicker SGTS graft may be more resistant to stretch in the immediate post-operative setting, requiring more time for the patient to achieve maximum knee extension. Longer term follow-up is required to determine if this contributes to reducing rates of graft re-rupture or failure. While the difference in time to maximum extension was statistically significant, it may not be clinically significant as is reflected by the similarity in overall incidence of LOE. Sheldon investigated graft cross-sectional area and incidence of LOE two years post-ACLR and found no apparent relationship despite a 20% difference in graft size between groups [20]. We conclude that if graft size does contribute to early post-operative LOE, it is likely to resolve over time, however this needs to be formally evaluated.
There was a significant difference in the use of notchplasty between the SGTS group (17.8%) and the non-SGTS group (44.7%). None of the observed techniques mandate use of notchplasty and this discrepancy may reflect surgeon preference. We found a positive association between patients who did not receive notchplasty and patients who had no extension deficit at initial physiotherapy session (32.6% of combined-group patients, p = 0.001), however this association was unable to be assessed in each group alone due to an insufficient number of subjects in the SGTS group. It is therefore difficult to draw correlations about the role of notchplasty in the SGTS technique versus others. Though there was an initial association with notchplasty and LOE, there was no association at final review, which suggests that any initial LOE may not be clinically important. Koga et al. demonstrated that in patients having undergone double-bundle ACLR, those with notchplasty had significantly greater extension deficits at all time points beyond 6 months post-op (max. follow-up 2 years) [21]. They concluded that bleeding from the notchplasty site caused fibrosis of the infrapatellar fat pad and resulted in LOE in those patients. Longer term follow-up would be required to assess if notchplasty does indeed result in delayed LOE as reported by Koga et al.
One of the proposed benefits of the SGTS technique is most grafts can be fashioned using only ST tendon. In this study, ST tendon alone was used in 71.1% of cases in the SGTS group compared to 4.4% in the non-SGTS group, which was statistically significant. Previous studies have shown that, when comparing STG tendon graft to ST-only tendon graft, there was reduction in both isokinetic and isometric knee flexion strength in the STG groups [22]. Conversely, Ardern et al retrospectively compared patients with STG autograft and ST-only autograft at a mean of 32.5 months after surgery and found no significant difference in isokinetic or isometric hamstring strength [23]. We did not assess knee flexion strength in this study. While the significance of preserving Gracilis tendon remains controversial in current literature, ST tendon alone for ACL graft appears to be a non-inferior choice and a reliable option for the SGTS technique.
4.3 Limitations
There were some limitations noted in this retrospective study. The primary objective was to evaluate the SGTS technique with respect to LOE compared to other techniques. The patients in the non-SGTS group were operated on by 12 different surgeons using 3 different fixation methods. While this may replicate the breadth of ACLR techniques, it becomes difficult to draw conclusions about the superiority of the SGTS method over any single method. What can be concluded is that the SGTS was not inferior to other techniques combined with respect to persistent LOE in this study. Future studies could investigate the SGTS technique versus individual non-SGTS techniques to identify its strengths and weaknesses. The main limitation of this study was that it was a retrospective case-control study. As has been identified, cause of LOE is multifactorial and some of these variables could be better controlled with a prospective study. A prospective randomised control trial is needed to better evaluate the strengths of the SGTS method over other techniques for ACLR.
5. Conclusion
This study investigated if ACLR using the SGTS technique had superior results in terms of LOE compared to other techniques employed at QEII Hospital. We found no significant difference in incidence of LOE between the two groups. Patients in the SGTS group took longer to achieve maximum knee extension however, and this may be contributed to by the shorter, thicker graft used in the SGTS technique. This finding is of questionable clinical significance, given that difference in LOE incidence at final follow up was not significant. Previous studies have shown that notchplasty can contribute to extension deficit after ACLR, which was observed initially in this study as well. We conclude that the SGTS technique is comparable to other ACLR techniques in terms of incidence of LOE and that notchplasty, independently, could contribute to early, but not persistent, extension deficits after ACLR.
Data Availability
Blinded data may be available from the authors at reasonable request.
Conflicts of interest
Corey Scholes holds shares in EBM Analytics, which is contracted to perform registry activities for QEII Jubilee Orthopaedics. No other authors have conflicts of interest to disclose.
Acknowledgements
The authors thank Meredith Harrison-Brown for her assistance with editing and formatting the manuscript.
References
