INTRODUCTION
Patients with posterosuperior irreparable rotator cuff tears (PSIRCTs) complain of unbearable shoulder pain and great disability, including loss of active range of motion (ROM) [1]. The optimal treatment of PSIRCTs can be challenging, especially for those who do not qualify as ideal candidates for reverse total shoulder arthroplasty, such as young and active patients with high physical demands or elderly patients [1]. Although there are several surgical options for PSIRCTs, tendon transfer is a viable treatment alternative for this population [2]. Several tendons can be used for tendon transfer to improve shoulder function by reconstructing the PSIRCT. Among these, lower trapezius tendon transfer (LTT) is an effective option to restore native glenohumeral biomechanics and the rotator cuff force couple in PSIRCTs [2-4].
Recently, arthroscopically assisted LTT transfer (aLTT) has received attention as a more effective treatment option than latissimus dorsi transfer due to the biomechanical advantages of aLTT [5]. Several studies have demonstrated favorable clinical and radiological results of aLTT to treat PSIRCT in short-term follow-up [5-10]. Moreover, aLTT showed better results than superior capsule reconstruction and latissimus dorsi tendon transfer [5,6]. Furthermore, favorable mid-term outcomes of aLTT were reported recently [11].
As it is anatomically impossible to attach the released lower trapezius tendon to the supraspinatus (SSP) footprint on its own, an interposition graft is necessary to complete the transfer. Initially, the Achilles tendon allograft (ATA) was introduced as an interpositional graft and showed good clinical outcomes in short-term and mid-term clinical studies [3,5-7,11]. Although an ATA has advantages, such as no donor site morbidity, reduced operative time, and broad accessibility, there are concerns regarding high cost, infection-related problems, graft rejection, and/or delayed incorporation [12,13]. To minimize allograft-related complications, some studies have reported the use of autografts in aLTT. Use of semitendinosus tendon and hamstring tendon autografts as interpositional grafts is associated with good clinical outcomes in aLTT for PSIRCT [14,15]. Recently, a fascia lata autograft (FLA) was introduced as a reliable interpositional graft during aLTT [16]. However, no studies have compared allografts versus autografts as interpositional grafts in aLTT for PSIRCT.
The purpose of this study was to compare the clinical and radiologic results of an ATA versus FLA as the interpositional graft in aLTT for PSIRCT. We hypothesized that although aLTT may lead to significant improvement in radiologic and clinical results in patients with PSIRCT, use of an ATA may be a significant independent risk factor for graft retear.
METHODS
The Institutional Review Board of the Ministry of Health and Welfare approved this study (No. 2023-040-001) and waived the necessity for informed consent due to the retrospective nature of the study. However, the informed consent for the use of photographs was obtained from the patients with a full explanation.
Patient Selection
We retrospectively reviewed patients treated with aLTT for PSIRCT between January 2017 and December 2021. PSIRCT was diagnosed based on physical examination, plain radiography, magnetic resonance imaging (MRI) scans, and arthroscopic results (Fig. 1). PSIRCT was diagnosed based on the follows criteria: (1) more than two rotator cuff tendon tears including SSP and infraspinatus (ISP) tears, (2) 3 or 4 fatty infiltration grade using the Goutallier classification for rotator cuff muscles, (3) medial retraction or shortening of the torn tendons to the glenoid level corresponding to Patte stage 3, (4) rotator cuff tendons retracted to the glenoid not reduced to the footprint despite interval sliding techniques during diagnostic arthroscopy.
Indications for aLTT to treat PSIRCT were as follows: (1) unbearable pain and severe loss of functional ability interfering with activities of daily living, (2) symptomatic PSIRCT with failed conservative treatment, (3) intact subscapularis (SSC) or a reparable SSC tendon, (4) no arthritic changes or Hamada grade 2 arthritic changes. Patients who lacked preoperative or at least a 2-year postoperative clinical/MRI evaluation were excluded from the study.
Finally, 90 patients who underwent aLTT for PSIRCT with a minimum clinical and MRI follow-up of 2 years were enrolled in the study. Among them, ATA was used as an interposition graft in 64 patients (“group ATA”), and FLA was used in 26 patients (“group FLA”) (Fig. 2). Among them, 12 patients (18.8%) in the ATA group and seven patients (26.9%) in the FLA group underwent aLTT as a revision procedure after rotator cuff retear. Because this study was performed retrospectively, patients were not grouped randomly according to the type of interpositional graft used. The decision to use an ATA or FLA was based on each patient’s overall condition. Advantages and disadvantages of using an ATA or FLA were described to the patients prior to surgery. In most cases, choice of an ATA or FLA was based on the patient’s decision.
Surgical Techniques
All surgeries were carried out by a single experienced surgeon (CHB). Patients were placed in the lateral decubitus position on the operating table with a traction device under general anesthesia with an interscalene nerve block. The reparability and mobility of the SSP and ISP tendons were evaluated during the diagnostic arthroscopic examination. Non-viable or residual fibrous rotator cuff tendon tissue was removed. If the SSC tendon was torn and reparable, the SSC was repaired. After implementation of the interval sliding technique, PSIRCTs was identified if the SSP and ISP were not reparable. The patient then made the decision whether to undergo aLTT or not (Fig. 3A). The subacromial spur was removed to prevent postoperative irritation of the interpositional bridging graft of the aLTT during acromioplasty. Tenotomy, tenodesis, or superior capsular reconstruction of the long head biceps was performed based on clinical symptoms including tenderness at the biceps groove or findings of the biceps-associated physical examination, and biceps pathology such as tear, subluxation, or degeneration. Irreparable and non-viable parts of the rotator cuff tendons were removed and the SSP was prepared. In more detail, the footprint of the SSP was prepared medially to the cartilage of the humeral head and laterally to the greater tuberosity (GT) to extend the contact area between the footprint and the interpositional bridging graft. The interval between the remnant ISP and deltoid muscle was identified and developed for the LTT.
To harvest the lower trapezius tendon, the scapular spine, scapular body, and lower trapezius tendon were identified as surface land markers. A 5-cm oblique skin incision was made from the mid-portion of the scapular spine inferior border to the medial border of the scapular body. Fat tissue over the lower trapezius tendon was removed. The insertion point of the lower trapezius tendon was identified around the mid-portion of the scapular spine inferior border. The triangular-shaped lower trapezius tendon was detached and mobilized from the underlying ISP fascia. The end of the lower trapezius tendon was prepared with the Krakow suture method with nonabsorbable suture material (Fig. 3B). The infraspinatus fascia below the lower trapezius tendon was incised to make sufficient space for the transferred lower trapezius tendon.
ATA preparation
After preparation of the LTT, the Achilles tendon was prepared as an interpositional allograft. After removal of the calcaneus portion, the ATA was folded two or three times to obtain adequate graft thickness (at least 6 mm), width (at least 2 cm), and length (at least 15 cm). ATA was prepared using the Krakow method with #2 Fiberwire. These traction stiches were created for insertion into the SSP footprint (Fig. 4).
FLA harvest
The FLA was harvested from the lateral aspect of the thigh. A 15-cm-long straight skin incision was made and the FLA was identified. After dissecting the fascia lata, at least 15 cm × 4 cm of fascia lata was harvested. The harvested FLA was folded two or three times to obtain adequate graft thickness (at least 6 mm) and length (at least 15 cm). One side of the FLA was prepared with Krakow sutures methods. These traction stiches were created for insertion into the SSP footprint (Fig. 5).
After harvesting the interpositional graft, two medial-row anchors were inserted into the anterior and posterior area of the SSP footprint, respectively. These two anchors were located just lateral to the articular cartilage of the GT. Three suture threads of the posterior anchors were passed into the posterior remnant ISP with a suture passer. These three threads were used for side-to-side sutures with the interpositional graft. The remaining three threads were taken out through the lateral portal and passed into an interpositional graft for attachment to the SSP footprint. A long curved clamp was inserted from the lateral portal to the infraspinatus fascia incision of the previous incision. The traction stiches of the interpositional graft were grasped and drawn out to the lateral portal. The unprepared site of the interpositional graft was pulled through the scapular incision to maintain the appropriate tension. After placing the interpositional graft on the SSP footprint, it was fixed using a double-row suture bridge technique with the two previously inserted anchors. Then, a side-to-side suture between the interpositional graft and posterior remnant rotator cuff was performed. Three lateral-row anchors were inserted at the posterolateral, middle–lateral, and anterolateral aspects of the GT. The interpositional graft was fixed more tightly with these lateral-row anchors using the suture bridge technique (Fig. 6).
After fixation of the interpositional graft on the SSP footprint, the interpositional graft was sutured to the inferior border of the lower trapezius tendon using the Krakow method with nonabsorbable suture material. The inferior border of the lower trapezius tendon is similar to the native line of pull of the ISP. Anastomosis of the interpositional graft to the lower trapezius tendon was carried out with maximal external rotation (ER) and 60° abduction of shoulder for physiological tension.
Postoperative Management
Postoperatively, a shoulder ER brace was applied to all patients at 0° ER. Patients wore the brace for 6 weeks. During this 6 week period, intermittent movement of elbows, wrists, and fingers was permitted. After 6 weeks of wearing the shoulder ER brace, patients began active ROM exercises in forward elevation, abduction, internal rotation (IR), and ER for 4 weeks. Then, patients were allowed to perform full ROM and underwent gentle strengthening exercises. Six months postoperatively, patients were encouraged to return to high level activities.
Clinical Evaluation
Patients visited the out-patient department every 3 months for 1 year postoperative, and thereafter every year. Demographic characteristics, including medical comorbidities, symptom duration, smoking status, body mass index (BMI), age, and sex were collected. Preoperative and postoperative clinical results were evaluated using the visual analog scale (VAS) score for pain, Constant score, American Shoulder and Elbow Surgeons (ASES) score, University of California-Los Angeles (UCLA) score, and activities of daily living that require active ER (ADLER) score. Active ROM (aROM) of the shoulder was evaluated using a standard goniometer in forward elevation, abduction, ER at 0° abduction, ER at 90° abduction, and IR. IR was assessed as the level reached by the thumb when the patient internally rotated their arm backward (10, T7; 8, T12; 6, L3; 4, lumbosacral junction; 2, buttock; 0, greater trochanter). An experienced research coordinator recorded all clinical scores and assessed aROM at each follow-up visit. If the patient was unable to perform active forward elevation more than 90°, the patient was diagnosed with pseudoparalysis. Complications, including nerve palsy, hematoma formation, adhesions, and surgical infection, were recorded.
Radiologic Evaluation
To evaluate radiologic outcomes, true anteroposterior plane radiographs (Grashey [17]) and shoulder MRI scans were obtained. On the true anteroposterior (Grashey) view, the acromiohumeral distance (AHD), which is the shortest distance from the upper portion of the humeral head to the lower margin of the acromion undersurface, was evaluated preoperatively and at 2 years postoperative. Hamada grade was used to evaluate arthritic changes of the glenohumeral joint. Hamada grades 3 or more (3, 4, or 5) at 2 years postoperative indicated marked progression of arthritis.
MRI was performed at 6 weeks, 1 year, and 2 years postoperatively to evaluate graft integrity. The integrity of the interpositional graft was assessed using the five criteria of Sugaya et al. [18]. If the interpositional graft was classified as type IV or V, retear of the graft was considered. An independent musculoskeletal radiologist who was unaware of the clinical details of this study evaluated the MRI scans.
Statistical Analyses
Intra-observer reliability and inter-observer reliability of AHD and the Hamada grade were evaluated using the intraclass correlation coefficient. One author measured preoperative and postoperative 2-year AHD and Hamada grade at 1-week intervals to evaluate intra-observer reliability. Two other authors measured preoperative and postoperative 2-year AHD and Hamada grade to evaluate inter-observer reliability.
Preoperative and postoperative continuous data were analyzed using the nonparametric Wilcoxon signed-rank test. Preoperative and postoperative categorical data were analyzed using McNemar’s test. Continuous clinical and radiological data were compared between the two groups using the nonparametric Mann-Whitney test. Categorical data were compared between groups using Fisher’s exact test. SPSS version 21.0 (IBM Corp.) was used for analyses and the level of significance was set to 95%.
RESULTS
The patient population comprised 51 men (35 in the ATA group and 16 in the FLA group) and 39 women (29 in the ATA group and 10 in the FLA group). Average age at surgery was 63.3±6.4 years in the ATA group and 63.0±5.0 years in the FLA group. There were no significant differences in demographic data including BMI, medical comorbidities, preoperative fatty infiltration of the rotator cuff, management of SSC, or management of the biceps between the two groups (Table 1).
Significant improvement in clinical parameters after surgery were confirmed in both groups. There were no significant differences in the VAS score for pain, Constant score, ASES score, UCLA score, ADLER score, or aROM between the two groups (Table 2). AHD and Hamada grade showed good-to-excellent intra-observer reliability and inter-observer reliability. Significant improvements in radiological outcomes were confirmed in both groups. There were no significant differences in the 2-year follow-up AHD or Hamada grade between the two groups. Graft retear of the interpositional graft was observed in eight patients in the ATA group and two patients in the FLA group in the postoperative 2-year MRI study (Fig. 7). The rate of graft retear was higher in the ATA group than that in FLA group, but without statistical significance (P=0.718). All retears occurred at the humeral fixation site, medial to the footprint; none occurred at the tendon-graft interface (Table 3). Among the FLA group, one patient reported consistent discomfort at the donor site, and one patient reported consistent swelling due to a hematoma at the donor site. However, these patients were successfully treated conservatively.
DISCUSSION
aLTT is an effective treatment option for recovery of shoulder ER and abduction with restoration of glenohumeral kinematics in PSIRCT [7,11]. Although interpositional grafts should be used due to the limited length of the lower trapezius tendon, no study has compared if the type interpositional graft used in aLTT affects outcomes. Therefore, this study is meaningful as it is the first to compare clinical and radiologic results of using an ATA versus FLA as the interpositional graft in aLTT for PSIRCT. We confirmed that aLTT leads to significant improvement in clinical and radiologic results in patients with PSIRCT, regardless of whether an ATA or FLA was used as the interpositional graft. The retear rate of the interpositional graft was not associated with graft status. Measures to promote graft healing should be considered.
We believe that the good clinical and radiological results of aLTT were due to adherence to the tendon transfer principle and the biomechanical advantages of aLTT. The transferred lower trapezius tendon has a similar line of pull to the physiologic line of pull of the ISP [2]. Certainly, the excursion and tension of the transferred lower trapezius tendon are adequate to replace that of the native ISP [18]. Moreover, during shoulder ER and abduction, in-phase contraction of the transferred lower trapezius tendon promotes retention of the lower trapezius tendon in patients because the lower trapezius tendon plays a role in shoulder ER and abduction [19]. These advantages of the lower trapezius tendon may facilitate better recovery of shoulder kinematics and good long-term outcomes. A biomechanical study measured the aLTT-reconstructed joint reaction force and glenohumeral kinematics in both compressive-distractive and anterior-posterior planes [20]. In a three-dimensional computed biomechanical study, the LTT produced abduction moment arms during shoulder ROM and imitated the native SSP [4]. Moreover, transfer of the lower trapezius tendon through the subacromial space and its attachment to the footprint of the SSP provided static stability and a spacing effect [7]. This biomechanical advantage of the aLTT may improve shoulder function and prevent the progression of arthritic changes.
Since the initial introduction of an ATA as an interpositional graft by Elhassan et al. [5], several studies have reported favorable short-term clinical results for aLTT using an ATA as the interpositional graft. Elhassan et al. [5] reported a significant improvement in clinical outcomes in 41 patients followed-up for an average of 13 months. Rodríguez-Vaquero et al. [8] showed improved ROM in six patients followed-up for an average of 14 months. Stone et al. [10] demonstrated satisfactory relief of pain and significant improvement in ROM in 15 patients who were followed-up for 1 year. Similarly, Baek et al. [6,7] reported improved clinical and radiological outcomes in 36 patients and 42 patients with a minimum follow-up period of 2 years. Recently, Baek et al. [11] confirmed that favorable clinical outcomes and aLTT effectiveness were maintained for a mean of 58.2±5.3 months of follow-up after PSIRCT. However, they reported seven (19.4%) patients with retear at the final follow-up.
To overcome the disadvantages of ATA, such as delayed incorporation or graft retear, several studies have reported favorable clinical outcomes after using autografts as interpositional grafts in aLTT. Valenti and Werthel [14] used a semitendinosus tendon autograft instead of an ATA to prevent allograft-associated complications. Ye et al. [15] proposed the hamstring autograft as an interpositional graft with good clinical outcomes without graft retear. Recently, Baek et al. [16] proposed an FLA graft as a viable interpositional graft during aLTT for PSIRCT. They reported favorable clinical and radiological outcomes in 22 patients with two (9.1%) cases of graft retear over a mean follow-up period of 35.9±15.9 months [16]. In this study, we confirmed that the retear incidence of the interpositional graft was higher in the ATA group (12.5%) than the FLA group (7.7%). However, this result was not statistically significant and our hypothesis that that use of an ATA would be a significant independent negative predictor of graft retear was not supported by the data. Nevertheless, measures to promote graft healing and prevent retear of the graft should be considered, regardless of the graft type.
This study had several limitations. First, because of its retrospective design, patients were not grouped randomly according to the type of interpositional graft used. The decision to use an ATA or FLA was based on the patients’ overall condition and decision. This limitation may have resulted in selection or performance bias. Second, the small sample size may have limited the validity of this study in clinical practice. Third, patients were only followed-up for a short period of time. Nonetheless, this study is meaningful as it is the first study to compare the clinical and radiological outcomes of the use of an ATA versus FLA as the interpositional graft in aLTT for PSIRCT. Fourth, the method of fixation between the interpositional graft and the lower trapezius tendon evolved from a Pulvertaft-weave to a side-by-side suture in a Krakow configuration during the course of this study. This was due to complaints of a moving "lump" on the back during shoulder use. However, there were no cases of tendon-graft interface failure during the study, regardless of the fixation method used. Finally, a formal cost comparison could not be performed as the actual fee the patient had to pay out-of-pocket varied greatly depending on the patient's insurance coverage. In general, the cost of an Achilles allograft is approximately 1,000 USD while autograft harvest costs 200 USD.
CONCLUSIONS
aLTT may lead to significant improvement in clinical and radiological outcomes in patients with PSIRCT, regardless of whether an ATA or FLA is used as the interpositional graft. The retear rate of the interpositional graft is not associated with graft status. Measures to promote graft healing should be considered.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
