Abstract
Background:
Modern rotator cuff repair techniques demonstrate favorable early and midterm outcomes, but long-term results have yet to be reported.
Purpose:
To determine 10-year outcomes and survivorship after arthroscopic double-row transosseous-equivalent (TOE) rotator cuff repair.
Study Design:
Case series; Level of evidence 4.
Methods:
The primary TOE rotator cuff repair procedure was performed using either a knotted suture bridge or knotless tape bridge technique on a series of patients with 1 to 3 tendon full-thickness rotator cuff tears involving the supraspinatus. Only patients who were 10 years postsurgery were included. Patient-reported outcomes were collected pre- and postoperatively, including American Shoulder and Elbow Surgeons (ASES), 12-Item Short Form Health Survey (SF-12), Single Assessment Numeric Evaluation (SANE), shortened version of the Disabilities of the Arm, Shoulder and Hand (QuickDASH), and satisfaction. Kaplan-Meier survivorship analysis was performed. Failure was defined as progression to revision surgery.
Results:
A total of 91 shoulders (56 men, 31 women) were included between October 2005 and December 2009. Mean follow-up was 11.5 years (range, 10.0-14.1 years). Of 91 shoulders, 5 (5.5%) failed and required revision surgery. Patient-reported outcomes for patients who survived were known for 80% (69/86). Outcomes scores at final follow-up were as follows: ASES, 93.1 ± 10.8; SANE, 87.5 ± 14.2; QuickDASH, 11.1 ± 13.5; and SF-12 physical component summary (PCS), 49.2 ± 10.1. There were statistically significant declines in ASES, SANE, and SF-12 PCS from the 5-year to 10-year follow-up, but none of these changes met the minimally clinically important difference threshold. Median satisfaction at final follow-up was 10 (range, 3-10). From this cohort, Kaplan-Meier survivorship demonstrated a 94.4% survival rate at a minimum of 10 years.
Conclusion:
Arthroscopic TOE rotator cuff repair demonstrates high patient satisfaction and low revision rates at a mean follow-up of 11.5 years. This information may be directly utilized in surgical decision making and preoperative patient counseling regarding the longevity of modern double-row rotator cuff repair.
Arthroscopic rotator cuff repair (ARCR) has dramatically evolved with improved surgical techniques and implant technology from open techniques to arthroscopic double-row (DR) linked constructs. Traditional DR rotator cuff repair involves knotted suture anchors placed on the medial and lateral aspects of the rotator cuff footprint. The additional fixation with the lateral row has biomechanically been shown to increase repair strength, and postoperative magnetic resonance imaging (MRI) studies have demonstrated improved healing compared with single-row repairs. 6
With advances in implant technology and techniques, DR rotator cuff repairs through a transosseous-equivalent (TOE) technique have gained traction in the field of shoulder arthroscopy. This technique is a self-reinforcing construct that allows for axial compression along the entire rotator cuff footprint through the sutures from the medial row across the bursal surface of the tendon footprint to the lateral row.11,17 This decreases the forces seen at the medial row, allows for a more forgiving construct, and provides improved healing capacity with higher vascular in-flow to the tendon.2,10,12-14 This increases healing and decreases the risk for retear.1,17 Despite promising biomechanical studies and excellent early to midterm outcomes with this technique,5,7,10-14,21 long-term results have not yet been reported for this relatively new technology.
The purpose of this study was to report a minimum of 10-year outcomes of primary arthroscopic DR rotator cuff repair utilizing the TOE technique. It was hypothesized that there would be low rates of revision and high patient satisfaction at long-term follow-up in those treated with DR ARCR utilizing the TOE technique.
Methods
Study Cohort
In this institutional review board–approved (Vail Health IRB 2019-54; Vail, CO), retrospective study of prospectively collected data of a single surgeon (P.J.M.), patients who met the following criteria were included: (1) full-thickness tear of the supraspinatus tendon with or without anterior and/or posterior tear propagation into the subscapularis or infraspinatus tendon; (2) operative management using a DR-TOE ARCR with a knotted suture bridge or a knotless tape bridge construct; (3) age of at least 18 years at the time of surgery; and (4) a minimum of 10 years’ follow-up after surgery. Risks, benefits, and treatment alternatives were discussed with all patients before they elected to undergo surgery. Physical therapy was recommended for all patients with chronic tears. Early surgery was recommended for those patients who had full-thickness, acute traumatic tears. Exclusion criteria were (1) previous rotator cuff repair; (2) partial-thickness tears; (3) patch augmentation of the rotator cuff; (4) isolated tears of the teres minor, infraspinatus, or subscapularis; (5) fractures about the shoulder; (6) bony defects of the glenohumeral joint; (7) existence of an os acromiale; (8) rheumatoid arthritis; (9) open distal clavicle excision; and (10) concomitant reconstructive procedures such as shoulder stabilizations or coracoclavicular ligament reconstructions; or (11) patient refusal to participate or death before initiation of the current study. Patients who received open distal clavicle excisions were excluded to minimize confounding variables that could have affected the results. Furthermore, it was required at the time of surgery that the remaining rotator cuff tendon be at least 10 mm in length from lateral to medial to allow footprint coverage and that the remaining rotator cuff tendons were able to be mobilized adequately to cover the footprint. Interval slides were not used.
Surgical Technique
All surgeries were performed by a single shoulder surgeon (P.J.M.) using a DR-TOE rotator cuff repair. In the earlier period of this study, a knotted suture bridge construct was utilized, which was then replaced by a knotless tape bridge construct as biomechanical evidence emerged supporting knotless tape bridge repairs.11,20,21 Before inclusion of the first patient in this study, the senior author (P.J.M.) performed over 500 ARCRs, thus minimizing variability in technical skills.
After the induction of general anesthesia and placement of an interscalene nerve block, patients were placed in the beach-chair position. The procedure started with a diagnostic glenohumeral arthroscopy via a standard posterior viewing portal. Most patients subsequently underwent glenohumeral debridement to remove any frayed or torn labral tissue and to treat synovitis with a shaver and radiofrequency ablation device. All patients underwent subacromial decompression. Acromioplasty was performed in all cases utilizing a modified cutting block technique to achieve a type 1 acromion. In patients with a type 1 acromion, a simple undersurface decortication was performed to provide bleeding and theoretic biologic benefits for tendon healing. Biceps tenotomy was performed if there were degenerative lesions of the long head of the biceps tendon (eg, fraying, synovitis), unstable biceps anchor lesions (eg, SLAP type 2 and higher), subscapularis tears, or pulley lesions. After the rotator cuff was repaired, a subpectoral tenodesis of the long head of the biceps using an interference screw was performed according to the technique described by Pogorzelski et al. 18
After the treatment of concomitant pathologies, attention was then drawn to the ARCR. Soft tissues were removed from the greater tuberosity and punctate bleeding was created using an arthroscopic shaver. Care was taken to preserve the underlying cortex to achieve optimal fixation of the suture anchors.
For the knotted suture bridge repair, the medial-row anchors (4.5 mm Corkscrew; Arthrex) were inserted adjacent to the lateral articular margin. A curved suture shuttling device (SutureLasso, 90° straight; Arthrex) was then used to pass the 2 limbs of the No. 2 suture (FiberWire; Arthrex) of the first medial-row anchor through 2 adjacent points of the tendon lateral to the musculotendinous junction. Care was taken to place the sutures in a manner such that an anatomic tendon footprint restoration was achieved based on tear pattern. The medial row was then fixed using a sliding Weston knot and further secured with 4 alternating half-hitches. This was then repeated for the second medial-row anchor. Next, 1 suture strand of each individual medial-row anchor was then grasped, tensioned, and inserted into the eyelet of the first lateral-row anchor (4.75 mm biocomposite SwiveLock or 4.5 mm Pushlock; Arthrex). The first lateral-row anchor was then inserted onto the anterolateral rim of the greater tuberosity. The same technique was performed for the second lateral-row anchor, which was placed onto the posterolateral rim of the greater tuberosity, creating a DR-TOE suture bridge repair.
For the knotless tape bridge repair, a prefabricated kit (SpeedBridge Kit using 4.75 mm biocomposite SwiveLock anchors; Arthrex) comprising 2 anchors for the medial row loaded with No. 2 tape suture and 2 sutureless lateral-row anchors (Arthrex) were used. The repair was performed in a similar technique to that used in the suture bridge repair with the difference being that the fixation of the medial row did not use knot-tying, suture tape was used, the 2 limbs of tape were passed in 1 single pass through the tendon, and laterally threaded anchors were used. The 2 suture tapes of each medial-row anchor were shuttled through the tendon together, and 1 tape strand of each individual medial-row anchor was then inserted into the eyelet of the lateral-row anchors. The lateral-row anchors were then inserted at the anterolateral and posterolateral rims of the greater tuberosity, creating a DR-TOE tape bridge construct.
Postoperative Rehabilitation
Immediately postoperatively, patients used a sling with an abduction pillow and were encouraged to start movement of the fingers, wrist, and elbow. Early passive range of motion was started on postoperative day 1 and was continued for 6 weeks. Between 4 and 6 weeks postoperatively, patients began active-assisted range of motion for an additional 2 to 3 weeks. Active range of motion began around 4 to 6 weeks postoperatively once all active-assisted shoulder movements could be performed without pain. 6 Strengthening of the rotator cuff muscles was started between postoperative weeks 8 to 10. Resisted elbow flexion was avoided for 6 weeks in patients who underwent additional biceps tenodesis. Full activities were typically permitted at 14 to 16 weeks. 22
Outcome Scores
Preoperatively, as well as at 2 years, 5 years, and minimum 10 years follow-up, patient-reported outcomes were collected, which included the American Shoulder and Elbow Surgeons (ASES) score, Single Assessment Numeric Evaluation (SANE) score, shortened version of the Disabilities of the Arm, Shoulder and Hand (QuickDASH) score, the physical component summary (PCS) of the 12-Item Short Form Health Survey (SF-12) questionnaire, and patient satisfaction (1-10; 10 = highly satisfied). Of note, before 2010, the SANE and the QuickDASH scores were not routinely collected preoperatively. Complications and further surgical interventions were assessed. Failure was defined as revision rotator cuff repair or conversion to reverse total shoulder arthroplasty.
At the minimum follow-ups of 2, 5, and 10 years postoperatively, questionnaires with the aforementioned scores were sent to the patients electronically. If patients did not return their questionnaires, they were contacted via telephone or email and were asked to complete the shoulder survey. No questions regarding the patient-reported outcomes were asked via telephone, to avoid response bias. The outcomes of some patients in the current study have been reported in previously published 2-year and 5-year outcome studies.9,18,21
Statistical Analysis
Statistical analysis was performed with SPSS Version 11.0 (IBM). A priori power analysis was not performed because of a fixed sample size; however, assuming the study design, minimum sample sizes per comparison of 57, 2-tailed testing, and an alpha level of .05, an effect size of d = 0.38 was detectable with 80% statistical power. Furthermore, assuming an SD of 10 points, 16 this effect size corresponds to a between–time point mean difference of 3.8 points on the ASES scale. Thus, based on a previously reported minimally clinically important difference of 11.1 points on the ASES scale, 3 nonsignificant comparisons in this study can reasonably rule out magnitude changes that are clinically relevant.
Categorical data are presented as number and percentage and continuous data as mean ± SD. For normally distributed variables, an independent or paired t test was used for univariate analysis. For nonparametric data, a Mann-Whitney or Kruskal-Wallis test was performed. The Fisher exact test was used for discrete covariates. Survivorship analysis was performed using Kaplan-Meier survival curves for progression to revision rotator cuff repair or shoulder arthroplasty on the index shoulder as an endpoint. The level of significance was set at P < .05.
Results
Between October 2005 and December 2009, the senior surgeon (P.J.M.) performed 175 DR-TOE rotator cuff repair procedures. Of those who underwent surgical repair, 121 procedures were eligible for a 10-year follow-up after application of the exclusion criteria. At the time of the final follow-up, 30 patients refused study participation or were deceased. The final study population of 87 patients (91 primary repairs; 4 bilateral) included 56 men and 31 women with a mean age of 61.5 years (range, 38-80 years) at the time of the index surgery (Figure 1). Patient characteristics, including demographics, injury, and surgical information, are listed in Table 1. Of the 91 surgeries included, all shoulders underwent preoperative MRI. Due to computer archiving, preoperative MRI to study Goutallier muscle fatty infiltration was only available at the time of the study for 41 shoulders (45%). Of these shoulders, 23 were grade 0, 15 were grade 1, 1 was grade 2, and 3 were grade 3. No patient-reported outcomes were associated with preoperative Goutallier grade (P > .05).
Flow diagram of the study cohort. DCE, distal clavicle excision; DR-TOE, double-row transosseous-equivalent; RC, rotator cuff; SSC, subscapularis tendon; TM, teres minor tendon.
Patient Characteristics: Demographics, Injury, and Surgical Information a
Data are reported as n/N (%) or mean ± SD. Acute tears were defined as repair <90 days after injury; chronic tears were defined as repair >90 days after injury or insidious onset without specific injury. ISP, infraspinatus tendon; M, male; SSC, subscapsularis tendon; SSP, supraspinatus tendon.
Data points missing for 2 patients.
Failures and Complications
Patients who subsequently underwent revision rotator cuff repair after primary surgery were defined as having failed results. Five patients (5.5%) required revision surgery; 4 were revised under 5 years and 1 was revised 12 years after primary repair. Of the 5 patients, 3 were revised by the primary surgeon. One patient suffered repair failure secondary to infection at 38 days and underwent irrigation and debridement with revision cuff repair. Two patients suffered retears at 49 days and 4.1 years and were revised with margin convergence and double-row rotator cuff repair, respectively. All went on to achieve satisfactory outcomes after revision surgery. Postrevision outcomes were unknown for 2 patients, as revision surgery was performed elsewhere.
Three patients underwent subsequent shoulder surgery to treat additional conditions. Two patients underwent capsular release and lysis of adhesions after developing stiffness. One patient underwent 2 consecutive debridement and decompression surgeries for persistent pain and subacromial impingement.
Clinical Outcomes
Minimum 10-year patient-reported outcomes were obtained for 69 of 86 patients (80%) at a mean follow-up of 11.5 ± 1.1 years (range, 10.0-14.1 years). The remaining 17 patients were lost to follow-up despite many efforts to contact them via multiple phone calls and emails. Additionally, previously obtained subjective outcome data were available for 66 of 86 patients (77%) and 59 of 86 patients (69%) at a minimum of 2 years and 5 years after surgery, respectively. The mean follow-up was 2.8 ± 0.8 years at a minimum 2-year follow-up and 6.5 ± 1.5 years at a minimum 5-year follow-up.
The mean 10-year postoperative ASES score was 93.1 ± 10.8 and mean SF-12 PCS score was 49.2 ± 10.1, both of which improved significantly from preoperative baseline (Table 2) (P < .001). Ten-year postoperative ASES and SF-12 PCS scores significantly decreased compared with 5-year scores of 95.2 ± 8.2 (P = .017) and 52.1 ± 8.5 (P = .029), respectively. At long-term follow-up, patients reported a high median satisfaction of 10 out of 10 (range, 3-10). Although the mean values were identical, there was a statistically significant improvement from the median 2-year patient satisfaction of 10 out of 10 (range, 1-10). The progression of shoulder-specific patient-reported functional outcomes in the study cohort are illustrated in Figure 2. The progression of general health patient-reported functional outcomes in the study cohort are illustrated in Figure 3.
Patient-Reported Outcomes after DR-TOE Rotator Cuff Repair a
Preoperative SANE, QuickDASH, and satisfaction were not collected preoperatively, thus P values were unavailable. SANE and QuickDASH scores were not routinely collected before 2010. ASES, American Shoulder and Elbow Surgeons; DR-TOE, double-row transosseous-equivalent; QuickDASH, shortened version of the Disabilities of the Arm, Shoulder and Hand score; SANE, Single Assessment Numeric Evaluation; SF-12 PCS, 12-Item Short Form Health Survey physical component summary.
Indicates significant (p < .05) difference between preop PRO versus the 10-year scores.
Indicates significant (p < .05) difference between the PRO at 5 years versus 10-year scores.
Shoulder-specific clinical outcomes from preoperatively to 2, 5, and ≥10 years postoperatively. ASES, American Shoulder and Elbow Surgeons; PRO, patient-reported outcome; QuickDASH, shortened version of the Disabilities of the Arm, Shoulder and Hand score; RCR, rotator cuff repair; SANE, Single Assessment Numeric Evaluation.
General health functional outcomes from preoperatively to 2, 5, and ≥10 years postoperatively. SF-12 PCS, 12-Item Short Form Health Survey physical component summary.
Survivorship
Postoperative Kaplan-Meier survivorship showed 94.4% survival at a minimum of 10 years (Figure 4). One patient underwent revision rotator cuff repair >10 years postoperatively (11.7 years) and thus was not represented as a failure on the curve.
Kaplan-Meier survivorship curve showing 94.4% survival at a minimum of 10 years postoperatively after surgical treatment using a DR-TOE RCR technique for full-thickness supraspinatus tears. DR-TOE, double-row transosseous-equivalent; RCR, rotator cuff repair.
Discussion
This study demonstrates excellent long-term outcomes and low revision rates after TOE rotator cuff repairs in patients with full-thickness rotator cuff tears. Kaplan-Meier analysis estimated a survivorship probability—defined by no revision rotator cuff repair nor conversion to reverse total should arthroplasty—of 94.4% at 10 years after surgery. This value is much lower than in previously documented literature on rotator cuff repair.8,9,15,16 These results are quite promising, although there were no long-term structural data. It is possible that there were patients in our cohort with minimally symptomatic or asymptomatic retears that did not require further surgery; however, the long-term nature of the study makes these types of failure less significant. This information is valuable when counseling patients about the risk for future surgery, which is a patient-centric endpoint.
Although there are a number of studies evaluating rotator cuff repair outcomes at a 5-year minimum follow-up, very few report on long-term results.4,5 In contradistinction to the excellent outcomes reported in this present study, Millett et al 10 demonstrated a 17% revision rate at 10 years after open rotator cuff repair. Modern arthroscopic techniques have appeared to decrease the revision rates. For example, Plachel et al 16 demonstrated a revision rate of 11% (6/56) at a minimum 10-year follow-up using a transosseous anchorless technique. In their study, mean ASES score (92 ± 10), as well as other outcome scores not used in the current study, showed high values at a mean 15-year follow-up (range, 12-18 years). The ASES value reported in that study is similar to that reported in the current study (93.1 ± 10.8) as were patient demographics of age, sex, and tear type. Plachel et al 16 also demonstrated high patient satisfaction at final follow-up. The current study’s results suggest that modern DR-TOE techniques using anchors may have lower failure rates (5.5% vs 11%) at long-term minimum 10-year follow-up, while having similar improvements in patient-reported outcomes and satisfaction.
In another more recent study by Plachel et al 15 where modern techniques were used, 16 single-row vs 11 double-row rotator cuff repairs were compared at 11- to 14-year follow-up. Five patients underwent revision, making the overall revision rate 8.1%. MRI was obtained at the 2-year follow-up and revealed full-thickness retears in 6 of 22 patients (27%) who did not undergo revision, while 9 of 20 patients (45%) had full retears at 12 years postoperatively. When compared with the current study, Plachel et al’s 15 group has fewer patients (11 vs 69) but a similar follow-up time and technique. Their revision rate was also higher (9.1% vs 5.5%) than in the current study, but the small sample size also makes this comparison difficult. However, Plachel et al did report long-term outcomes over 10 years as well as on advanced imaging, which is useful for surgeon and patient knowledge.
The senior author has previously published results with minimum 2-year and 5-year outcomes.9,17 In similar patient cohorts, Millett et al 9 demonstrated a 5.1% revision rate at a mean 2.9 years, and Pogorzelski et al 17 documented a 7.8% revision rate at a mean follow-up of 6.6 years. The current study demonstrates that the results are durable with minimal changes from the 5-year to 10-year time point. However, while there were statistically significant declines in ASES and SF-12 scores between the 5- and 10-year follow-up, there was no decrease in the SANE, QuickDASH, or satisfaction scores. It should also be noted that the changes in ASES or SF-12 scores between 5 and 10 years did not reach the minimal clinically important difference. 3 Thus, although a statistically significant decrease was seen in ASES and SF-12 scores between the 5-year and 10-year follow-up, the clinical implications appear minimal, indicating that TOE rotator cuff repair is likely durable at long-term follow-up. Further, patient satisfaction, SANE, and QuickDASH scores did not decrease between the 5-year and 10-year follow-up.
This study has several advantages and limitations. This is one of the first studies to publish long-term results with modern DR-TOE techniques for rotator cuff repair that utilized stringent criteria for patient selection allowing comparisons to be meaningful. However, structural follow-up was not available at final follow-up. Thus, while patients had low revision rates, a portion of the population may have had asymptomatic rotator cuff retears or incomplete healing. However, if a retear remains asymptomatic after 10 years, we believe that this should likely be considered a successful outcome. Therefore, this information may be used in counseling patients about the need for future surgery but does not reflect structural healing rates. In addition, all surgeries were performed by a single surgeon, and therefore, it is unknown whether the results seen can be extrapolated or generalized to other centers due to both the technical aspects of how the surgery was performed and the patient population that was treated. Over 50% of the patient population had isolated supraspinatus tears, which may have biased the results toward more favorable outcomes. In addition, due to the long follow-up period, some patient-specific variables such as patient comorbidities were not readily available, which could have provided further information on risk factors for failure. Overall, there were very few smokers and in general the population was active and healthy. We were able to obtain preoperative imaging for fatty infiltration classification for 45% of our study population, but a higher percentage of preoperative imaging may have allowed for better understanding of the study population. Although all were DR-TOE rotator cuff repairs, some included in the study were knotless tape bridge constructs while others were knotted suture bridge constructs. Because of the patient numbers, a comparison between the 2 groups was not feasible, but a preliminary analysis demonstrated no significant difference between techniques at 10 years. 19 Further, the knotted constructs were used earlier in the study before knotless technology existed. The authors do not believe bias related to this is significant, as the senior author was skilled in ARCR for many years before patients were included in the current study. Last, this is a retrospective review of prospectively collected data. Since many patients were either deceased, refused to participate, or were otherwise lost to follow-up before study initiation, there is some uncertainty as to whether the reported outcomes are generalizable to all patients who were 10 years out from a primary rotator cuff repair.
In conclusion, arthroscopic TOE rotator cuff repair demonstrates high patient satisfaction and low revision rates at a minimum of 10 years’ follow-up. This information may be directly utilized in surgical decision making and preoperative patient counseling regarding longevity of modern double-row rotator cuff repair.
Footnotes
Submitted June 26, 2020; accepted February 7, 2021.
One or more of the authors has declared the following potential conflict of interest or source of funding: This research was supported by the Steadman Philippon Research Institute, which is a 501(c)(3) nonprofit institution supported financially by private donations and corporate support. Steadman Philippon Research Institute (SPRI) exercises special care to identify any financial interests or relationships related to research conducted here. During the past calendar year, SPRI has received grant funding or in-kind donations from Arthrex, US Department of Justice, DJO, Major League Baseball, Ossur, Siemens, Smith & Nephew, and XTRE. J.W.A. has received a grant from DJO and hospitality payments from Smith & Nephew. P.-C.N. has been supported by AGA, via Arthrex, for 1 year. P.J.M. has received royalties and consulting fees from Arthrex; research support from Ossur, Siemens, and Smith & Nephew; royalties from Medbridge and Springer Publishing; and hospitality payments from Merz Pharmaceuticals GMBH, Arthrosurface Incorporated, Gemini Mountain Medical LLC, Stryker Corporation, and Sanofi-Aventis; and holds stock in VuMedi. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
