The surgical management of concomitant rotator cuff and SLAP (superior labral anterior-posterior) lesions remains controversial, especially in middle-aged patients, who are thought to be prone to stiffness postoperatively. Options for treatment of the SLAP lesion include biceps anchor tenotomy, tenodesis, and repair. The former two procedures have been advocated in the setting of rotator cuff repair, with the belief that the third results in stiffness and decreased shoulder motion1.
Débridement alone for the treatment of type-II SLAP lesions has resulted in poor outcomes in younger age groups2,3. Additionally, Pagnani et al.4 reported a trend toward poor results in several patients who had undergone a SLAP lesion repair with concurrent subacromial decompression when they were eighteen to forty years of age. The authors suggested a staged approach, with initial stabilization of the SLAP lesion followed by subacromial decompression for persistent symptoms of impingement. They did not advocate repair of superior labral lesions in patients over the age of forty.
However, in a study of SLAP lesion stabilization that included several patients treated concurrently for a rotator cuff tear and impingement syndrome, Samani et al.5 observed no adverse trends in the patients who had undergone the simultaneous procedures. This was partly attributed to an intensive postoperative rehabilitative program. Voos et al.6 also reported promising results of combined arthroscopic rotator cuff and labral repair in a case series with more than two years of follow-up.
The purpose of the present study was to investigate whether concomitant arthroscopic repairs of a rotator cuff tear and a SLAP lesion can yield satisfactory outcomes comparable with those seen with rotator cuff repair alone. To our knowledge, there have been no previous studies comparing patients with a simultaneous repair with a cohort of patients with an intact biceps anchor who underwent an isolated arthroscopic rotator cuff repair.
After obtaining institutional review board approval, we retrospectively reviewed the cases of all patients who had undergone combined arthroscopic SLAP lesion and rotator cuff repair between 2003 and 2005. Patients were assigned to two different cohorts on the basis of ICD-9 (International Classification of Diseases, Ninth Revision) and CPT (Current Procedural Terminology) codes, and operative notes were reviewed to extract data and to confirm that the respective procedure(s) had been performed. Thirty-four patients underwent rotator cuff repair and SLAP lesion repair (the RCR + SLAP cohort), and twenty-eight patients underwent rotator cuff repair alone (the RCR cohort). The RCR + SLAP cohort included all patients who met the inclusion and exclusion criteria. The patients in the RCR cohort were a consecutive subset. The selection of patients was not based on successful contact and callback.
Patients who had a mini-open or open procedure, a history of ipsilateral shoulder surgery, a subscapularis tendon tear, Outerbridge7 grade-III or IV glenohumeral osteoarthritis, a massive rotator cuff tear and pseudoparalysis, or adhesive capsulitis preoperatively were excluded from the study. Patients with advanced fatty infiltration (Goutallier grade 2, 3, or 48,9) and/or advanced muscle atrophy (Warner grade 3 or 410) were excluded from the study as well. All patients in both cohorts had grade-0 or 1 fatty infiltration and grade-1 or 2 muscle atrophy. In addition, an eighty-four-year-old patient with a history of septic arthritis was excluded from the RCR cohort.
Fourteen of the twenty-eight patients in the RCR cohort and seventeen of the thirty-four patients in the RCR + SLAP cohort received one or more subacromial steroid injections preoperatively (p = 1.000) (Table I). The number of injections per patient ranged from zero to four, for a total of twenty-seven injections, in the RCR cohort, whereas the patients in the RCR + SLAP cohort received zero, one, or two injections, for a total of twenty-three injections in that cohort.
All patients were treated by the same sports-fellowship-trained orthopaedic surgeon (S.D.M.). Outcomes were assessed independently by one of us (B.F.), who was not involved in any of the surgical procedures.
All thirty-four patients in the RCR + SLAP cohort had magnetic resonance imaging arthrograms confirming the presence of a full-thickness rotator cuff tear and a SLAP lesion, and all of them had a positive O'Brien active compression test. All of these patients were thus diagnosed with a SLAP tear both clinically and by magnetic resonance imaging before the operation. All patients had a symptomatic rotator cuff tear with at least one grade of weakness on manual strength testing. The operative findings in all of the patients in this cohort involved exposed bone or an intrasubstance superior labral tear with separation and instability of the biceps anchor, with a positive peel-back sign. All patients had minimal or no fraying of the biceps tendon. Any visualization of a structural biceps tendon lesion was a contraindication to SLAP lesion repair. All patients presented with a rotator cuff tear combined with a SLAP lesion, which was type II in thirty-two patients, type IV in one patient, and type V in one patient. All underwent combined rotator cuff and biceps anchor repair; none underwent biceps tenotomy or tenodesis.
All twenty-eight patients in the RCR cohort had magnetic resonance imaging arthrograms confirming the presence of a full-thickness rotator cuff tear as well as an intact biceps anchor. All patients had a symptomatic rotator cuff tear with at least one grade of weakness on manual strength testing. All had a negative O'Brien active compression test.
Three patients in the RCR cohort underwent a biceps tenodesis to treat a tear of the biceps tendon exceeding 30% of the tendon's diameter. All tenodeses were performed arthroscopically within the bicipital groove, with a single double-loaded anchor utilized to place a horizontal mattress knot. These patients were included in the RCR cohort to preserve the integrity of the consecutive series. Five patients in the RCR cohort had a concomitant type-I SLAP lesion (defined as minimal superior labral fraying and biceps tendon fraying of <10% of the tendon diameter with a stable biceps anchor), which was simply débrided.
All rotator cuff tears in both cohorts were repaired with the goal of reapproximating the rotator cuff to the anatomic footprint; all patients in both cohorts also underwent subacromial decompression.
Operative Technique
All of the surgical procedures were conducted with the patient in the beach-chair position. After a standard posterior arthroscopy portal was established, the glenohumeral joint was examined and all intra-articular pathological involvement was identified. Once an unstable SLAP lesion was noted, two 7-mm cannulas were placed anteriorly. One cannula was placed through the rotator interval or through the supraspinatus tear, and the second was placed above the subscapularis. The biceps anchor footprint and the superior aspect of the glenoid were prepared with a 4.5-mm shaver to remove any fibrous debris. A single 3-mm Bio-SutureTak anchor (Arthrex, Naples, Florida) was placed at the midpoint of the biceps anchor on the superior aspect of the glenoid rim after the superior aspect of the glenoid had been punched with an awl. All anchors were double-loaded with number-2 FiberWire suture (Arthrex). The sutures were shuttled underneath the biceps anchor and the labrum with a 90° Suture Lasso (Arthrex). One suture limb was shuttled anterior to the biceps anchor and the labrum, and the other was shuttled posterior to the biceps anchor and the labrum. All sutures were tied with use of an arthroscopic sliding, locking Weston knot followed by three alternating half hitches.
A subacromial decompression was performed in all patients, with particular attention paid to preserving the coracoacromial ligament. The bursectomy was extended into the lateral gutter. The size of the rotator cuff tear and the extent of tendon retraction were measured as accurately as possible with a graduated probe, calibrated in millimeters, in both the anteroposterior and the mediolateral plane. Double-row repairs were performed for all rotator cuff tears. Preparation of the footprint included clearance of fibrous debris, with preservation of the corticocancellous bone. The medial row of anchors was placed at the articular margin with use of horizontal mattress sutures. The lateral anchors were placed at the lateral edge of the greater tuberosity, and simple sutures were passed through the rotator cuff. All anchors were inserted at a "dead man's angle" of 45°. We utilized 5.0 or 6.5-mm bioabsorbable Bio-Corkscrew Suture Anchors (Arthrex), depending on the quality of the bone. Sutures were placed with use of a combination of Mitek Clever Hooks (Raynham, Massachusetts), a Suture Lasso (Arthrex), and a Scorpion suture passer (Arthrex). The goal of rotator cuff repair was reconstitution of the total footprint contact area, with a tension-free repair.
Postoperative Management
All patients underwent operative repair as day surgery. Postoperatively, the shoulder was immobilized with a sling and an abduction pillow providing 15° of abduction. The patients in both cohorts underwent the same intensive rehabilitation program. Pendulum exercises were started on postoperative day 1, and the patient was allowed an active range of motion of only the elbow, wrist, and hand. All patients were permitted to remove the sling and abduction pillow starting on postoperative day 1. This enabled use of the arm in a controlled setting with the elbow at the side—i.e., for typing and eating. Use of the sling and abduction pillow was continued for sleeping at night and for outdoor activities for six weeks. Passive range-of-motion and shoulder isometric exercises were started at two weeks postoperatively. Motion was increased as tolerated with the goal of achieving a full range of motion by eight weeks. Strengthening began at ten weeks postoperatively. Light activities below shoulder level were allowed at six weeks, and above-shoulder-level activities were allowed at three months.
Evaluation
All sixty-two patients were successfully contacted, and fifty-five returned for a follow-up clinical examination. The seven remaining patients submitted updates on their clinical status by telephone interview. These seven patients had been evaluated clinically at one year after the surgery. The primary outcome measures included the range of motion, Constant scores11, normalized Constant scores12, and American Shoulder and Elbow Surgeons (ASES) scores13. Dynamometer strength testing was performed on all patients as an adjunct to qualitative assessments14. We calculated normalized scores, as described by Katolik et al.12, to perform sex and age-matched functional assessments.
Statistical Methods
A power analysis (p = 0.80) determined that, with twenty-five patients in each cohort, we could detect a clinically relevant difference of 7 points with the ASES standardized shoulder assessment. In a study by Michener et al., the minimal clinically important difference in ASES scores was found to be 6.4 points15. Power analyses were not performed for the secondary variables.
The chi-square test was used to detect differences in variables expressed as proportions, including the patient's sex, number treated with steroid injections, operatively treated side, hand dominance, Workers’ Compensation status, return to work, return to participation in sports, fatty infiltration, and muscle atrophy. The Student t test was used to compare differences in variables expressed as means. These included age; number of months between the onset of symptoms and the repair; duration of follow-up; tear size; tear retraction; preoperative and postoperative Constant, normalized Constant, and ASES scores; and range of motion. Means are reported with 95% confidence intervals. A p value of <0.05 was considered significant.
Source of Funding
No external funding was received for this investigation.
Of the thirty-four patients in the RCR + SLAP cohort, thirty returned for clinical evaluation and dynamometer testing and four completed telephone interviews. In the RCR cohort, twenty-five of the twenty-eight patients returned for clinical evaluation and dynamometer testing and three completed telephone interviews. There were fifteen men and thirteen women with an average age of 59.6 ± 7.9 years (range, forty-three to seventy-five years) in the RCR cohort, and there were twenty-one men and thirteen women with an average age of 56.9 ± 7.9 years (range, forty-three to seventy-five years) in the RCR + SLAP cohort (Table I). In the RCR cohort, sixteen of the tears were traumatic in etiology, with an average of 10.8 months (median, 10.5 months; range, one to twenty-four months) from the onset of symptoms to the operative repair, and twelve of the tears were atraumatic (insidious in onset), with an average of 54.6 months (median, 20.5 months; range, six to 192 months) from the onset of symptoms to the operative repair. In the RCR + SLAP cohort, eighteen of the tears were traumatic in etiology, with an average of 9.6 months (median, 8.0 months; range, two to eighteen months) from the onset of symptoms to the operative repair, and sixteen of the tears were atraumatic, with an average of 60.1 months (median, twenty-two months; range, two to 360 months) from the onset of symptoms to the operative repair.
In the RCR cohort, the mechanism of injury was sports-related in seven patients. Other traumatic mechanisms included a fall from a height (three), a fall from a standing position (three), lifting a heavy item (two), and an accident during construction work (one). In the SLAP + RCR cohort, the mechanism of injury was sports-related in six patients. Other traumatic mechanisms included a motor-vehicle accident (two), a fall from a height (one), a fall from a standing position (five), and lifting a heavy item (four). Patients who reported a remote history of trauma (more than two years prior to the first clinical presentation) were classified as having an atraumatic tear, given the difficulty in reliably correlating the onset of symptoms with a singular traumatic event.
The right shoulder was involved in forty-three patients and the left shoulder was involved in nineteen patients in the study. The right and left sides were involved, respectively, in twenty-one and seven patients in the RCR cohort and in twenty-two and twelve patients in the SLAP + RCR cohort. The dominant extremity was involved in twenty patients in the RCR cohort and twenty-three patients in the SLAP + RCR cohort.
Of the twenty-eight patients in the RCR cohort, two had a work-related injury and had made a Workers’ Compensation claim. Of the thirty-four patients in the SLAP + RCR cohort, four had a work-related injury and had made a Workers’ Compensation claim and two were injured in a motor-vehicle accident with pending litigation (see Appendix).
Intraoperative Findings
Three patients in the RCR cohort underwent a simple biceps tenodesis, as described above, because of a biceps tendon tear that was >30% of the tendon diameter. In the RCR + SLAP cohort, thirty-two patients had a type-II SLAP lesion, one patient had a type-IV SLAP lesion, and one patient had a type-V SLAP lesion; all were repaired.
Postoperative Outcomes
The ASES scores increased significantly in both cohorts after the repairs; the mean score increased from 34.3 ± 24.7 points to 92.3 ± 12.1 points (p < 0.001) in the RCR cohort and from 22.6 ± 15.8 points to 96.4 ± 9.2 points in the RCR + SLAP cohort (p < 0.001). Although, preoperatively, the mean ASES score in the RCR + SLAP cohort was significantly lower than that in the RCR cohort (p = 0.027), no difference between the groups was noted postoperatively (Table II).
The Constant scores similarly increased significantly after the repairs in both cohorts; the mean score increased from 53.8 ± 18.9 points to 85.0 ± 6.5 points in the RCR cohort (p < 0.001) and from 49.9 ± 16.0 points to 91.0 ± 8.0 points in the SLAP + RCR cohort (p < 0.001.) While the Constant scores did not differ significantly between the cohorts preoperatively, the mean Constant score in the SLAP + RCR cohort was significantly higher than the mean score in the RCR cohort postoperatively (p = 0.002). Notably, the postoperative 6-point difference between groups is probably not clinically relevant.
The normalized Constant scores also increased significantly in both cohorts after the repairs; the mean score increased from 60.7 ± 20.4 to 95.8 ± 6.7 points in the RCR cohort (p < 0.001) and from 55.1 ± 17.0 to 101.0 ± 7.3 points in the RCR + SLAP cohort (p < 0.001.) Again, despite a lack of a significant difference between the groups preoperatively, the mean normalized Constant score in the RCR + SLAP cohort was higher than that in the RCR cohort postoperatively (p = 0.006). The postoperative difference of 5.2 points between groups is probably not clinically relevant (see Appendix).
The active range of motion increased significantly in both cohorts postoperatively (Table III). In the RCR cohort, the overall mean preoperative active range of motion was 145.2° of forward flexion (range, 30° to 170°), 138.6° of abduction (range, 30° to 170°), 48.6° of external rotation (range, 20° to 80°), and internal rotation to T11 (range, T8 to the sacrum). Postoperatively, the mean range of motion improved to 162.5° of forward flexion (range, 110° to 180°), 158.2° of abduction (range, 100° to 170°), 68.9° of external rotation (range, 40° to 85°), and internal rotation to T8 (range, T6 to L5). The improvements in forward flexion (p = 0.005), abduction (p = 0.003), and external rotation (p = 0.001) were statistically significant and clinically relevant. In the RCR + SLAP cohort, the overall mean preoperative active range of motion was 147.1° of forward flexion (range, 40° to 170°), 139.0° of abduction (range, 40° to 175°), 53.5° of external rotation (range, 25° to 80°), and internal rotation to T10 (range, T7 to the sacrum). Postoperatively, the mean range of motion improved to 164.6° of forward flexion (range, 140° to 175°), 161.6° of abduction (range, 120° to 175°), 68.1° of external rotation (range, 45° to 80°), and internal rotation to T8 (range, T7 to L5). The improvements in forward flexion (p = 0.001), abduction (p = 0.006), and external rotation (p = 0.001) were statistically significant and clinically relevant. There was no clinically relevant difference in the active ranges of motion between the two cohorts either preoperatively or postoperatively (see Appendix). Furthermore, all patients in both cohorts had a negative result of the O'Brien active compression test on follow-up physical examination.
Of the twenty-five patients in the RCR cohort who had been employed before the surgery, twenty-two returned to work. Of the thirty patients in the RCR + SLAP cohort employed before the surgery, twenty-nine returned to work. Of the eighteen patients in the RCR cohort who regularly participated in sports activities, sixteen returned to full participation. Of the thirty-one patients in the RCR + SLAP cohort who regularly participated in sports activities, twenty-nine returned to full participation.
Complications included an acute 2-cm rerupture of the rotator cuff five months postoperatively in a patient in the RCR cohort; the patient subsequently presented to another surgeon for a revision rotator cuff repair and had a good clinical outcome. In addition, one patient in the RCR + SLAP cohort developed a deep venous thrombosis in the perioperative period, necessitating a course of therapeutic-dose pharmacologic anticoagulation.
Of note, all patients in both cohorts stated that they were satisfied with the result of the surgery and that they would undergo the procedure again under similar circumstances.
Some surgeons might be reluctant to combine a SLAP repair with a rotator cuff repair because of the possibility of a poor clinical outcome and shoulder stiffness4,16. Our study supports the notion that concomitant arthroscopic rotator cuff and SLAP lesion repair is not contraindicated.
The patients in this study were treated during the same three-year time period and were well matched demographically; the two cohorts were comparable with respect to patient age and sex, the size of the rotator cuff tear, the extent of tendon retraction, the degree of muscle atrophy and fatty infiltration, handedness, Workers’ Compensation status, the mechanism of injury (traumatic or atraumatic), the time from the onset of symptoms to the surgery, the number of injections, and the duration of follow-up. The return-to-work rates were similar and the cohorts were comparable with respect to the rate of return to sports, although the SLAP + RCR cohort had a higher initial sports participation rate (thirty-one of thirty-four patients compared with eighteen of twenty-eight patients in the RCR cohort; p = 0.03). Of note is the fact that the overall average patient age in our study was 58.1 years, which is older than that reported in most studies of SLAP repairs2-5,17-19. The average age in a comparable study by Voos et al. was forty-eight years6.
The patients in the RCR + SLAP cohort presented with lower initial ASES scores (mean, 22.6 points compared with 34.3 points in the RCR cohort; p = 0.027), suggesting that combined rotator cuff and SLAP injuries may be more symptomatic than isolated rotator cuff tears. It is important to note that the ASES scoring sheet includes a pain rating scale (ranging from 0 to 10 for severity) that constitutes 50% of the total ASES score. Furthermore, the patients in the RCR + SLAP cohort had higher postoperative mean Constant and normalized Constant scores than did the patients in the RCR cohort (mean, 91.0 compared with 85.0 points [p = 0.002] and 101.0 compared with 95.8 points [p = 0.006], respectively), although clinically this difference is probably not relevant.
All patients in both cohorts had significant improvement in their range of motion after the surgery, and we could not identify a clinically relevant or statistically significant difference in the range of motion between the two cohorts. Some surgeons are reluctant to perform concomitant rotator cuff and SLAP lesion repairs for fear of causing postoperative shoulder stiffness. Although our study design cannot show specifically the effect of an early intensive range-of-motion program since we employed the same program for both groups, the comparable results obtained in the two cohorts may have resulted from our rehabilitation routine. There is concern that intensive early-range-of-motion protocols may result in re-tears after rotator cuff and SLAP lesion repairs. Our study did not address the issue of whether the SLAP lesions and/or rotator cuff tears healed, since postoperative imaging was not performed. Although good outcomes were achieved, it is well documented that that is possible even if the rotator cuff repair does not heal20,21.
To our knowledge, the present study is unique in that it is the first comparison of a consecutive series of patients who had combined repairs of a SLAP lesion and a rotator cuff tear with a cohort of patients who had only arthroscopic rotator cuff repair during the same time period. Previous studies of repairs of type-II SLAP lesions have excluded patients with rotator cuff tears18,22-25. Voos et al., however, reported on the outcome of combined arthroscopic rotator cuff and labral repairs as part of a case series in which fourteen patients had a combined SLAP lesion and rotator cuff repair and sixteen patients had a combined Bankart and rotator cuff repair; the average postoperative ASES score in that study was 94.3 points after a mean duration of follow-up of 2.7 years6.
More recently, a randomized controlled study was performed to compare repair of a concomitant rotator cuff tear and type-II SLAP lesion with a repair of a rotator cuff tear combined with a tenotomy of the long head of the biceps1. The authors observed significantly better results in terms of function, active forward flexion, and satisfaction in the latter cohort. The patients in that study, however, were restricted from overhead stretching for six weeks postoperatively to avoid damaging the repair. It is possible that this may have contributed to stiffness in the cohort treated with the combined SLAP lesion and rotator cuff repair. A decreased range of motion was not observed in our study, perhaps as a result of the initiation of passive range-of-motion and shoulder isometric exercises at two weeks postoperatively.
Some of the weaknesses of this study are a consequence of its retrospective nature. The long-term follow-up of seven patients was completed by means of a telephone interview. Because we did not analyze early postoperative data, we cannot comment on whether there were differences between the two cohorts in the early rehabilitative phase within six months after the surgery. Furthermore, there was no control group for the rehabilitation protocol.
Another potential weakness is the fact that all operations were performed by one surgeon, thus arguably decreasing the generalizability of the study's results. The fact that there was no postoperative imaging prevents an analysis of whether the SLAP lesions and/or the rotator cuff tears healed. Furthermore, no power analysis was performed for the secondary outcome factors, and there may have been differences between the two cohorts that we did not detect because of the small number of patients studied.
The SLAP lesion may occur as an isolated injury but frequently it is observed in association with other pathological conditions, such as a partial or full-thickness rotator cuff tear. The results of this study suggest that a concomitant repair of a rotator cuff tear and a SLAP lesion can achieve satisfactory outcomes in middle-aged patients. It may be that these encouraging results are related to the fact that patients with combined lesions have higher premorbid levels of function, have smaller rotator cuff tears and less retraction, and have more acute lesions that are thus more amenable to surgical repair. On the basis of this study, it appears that combined SLAP lesion and rotator cuff repair provides results comparable with those of rotator cuff repair alone. Additional studies are required to determine whether the rehabilitation methods used to obtain these results do not compromise the surgical repair.