The role of tendon repair versus physiotherapy in treating small-to-medium-sized full-thickness rotator cuff tears is still under debate. Over the last decade, considerable efforts have been made to define evidence-based guidelines for treatment selection for patients with tears mainly affecting the supraspinatus tendon1–8. Some randomized studies have reported comparable treatment effects from primary tendon repair and physiotherapy, and a primary nonoperative approach is now often recommended. An important inadequacy of the majority of existing studies, however, is their limited duration of follow-up of only 1 or 2 years3,4,6,7. Studies of the natural course of unrepaired tears have shown that these tears are at risk for anatomic and functional deterioration over time9–14, and earlier published results of the present study showed an increase in outcome differences in favor of tendon repair from the 2 to 10-year follow-ups2,5. Further evidence from longer-term comparisons is necessary to better understand the full impact of our treatment decisions.
In the present study, we aimed to investigate the outcome differences between tendon repair and physiotherapy in the yet-unexplored interval from 10 to 15 years. We hypothesized that the results from tendon repair would remain stable while those from unrepaired tears would continue to worsen, and that this would lead to an increase in between-group differences at 15 years.
Materials and Methods
In the current report, we present the 15-year results of a single-center, randomized, 2-arm, parallel-group, single-blinded superiority trial performed at a secondary care institution in Norway. The 1, 5, and 10-year results were previously published1,2,5. The study was approved by the Committee for Medical and Health Research Ethics in Norway and was registered at ClinicalTrials.gov (NCT00852657). Written informed consent was obtained from all participants after study information was given orally and in writing.
Patients and Procedures
Between September 2004 and October 2007, a total of 103 patients referred from primary care services for the treatment of a full-thickness rotator cuff tear of not more than 3 cm in diameter were included in the study. Inclusion and exclusion criteria are given in Table I, and the patient flow through the study is shown in Figure 1.
TABLE I -
Inclusion and Exclusion Criteria
| Inclusion Criteria |
| Pain, at rest or during exercise, laterally on the shoulder |
| A painful motion arc38
|
| A positive impingement sign39,40
|
| Passive shoulder motion of at least 140° for abduction and flexion |
| Demonstration of a full-thickness tear of the rotator cuff by both sonography and MRI, with a tear size not exceeding 3 cm on sonography |
| Muscle atrophy not exceeding Thomazeau stage 2 on MRI41
|
| Exclusion Criteria |
| Patient age of <18 yr |
| Tears involving >25% of the width of the subscapularis tendon |
| Presence of other local or systemic diseases affecting shoulder function |
| A history of surgical treatment of the involved shoulder |
| A medical contraindication for surgery or anesthesia |
| An inability to understand written and spoken Norwegian |
;)
Fig. 1: Trial flowchart: screening, randomization, and primary outcome population. Patients who crossed over to secondary surgery remained in the physiotherapy group for analysis in accordance with the intention-to-treat principle.
Outcome Measures
Measurements of shoulder function at the 15-year follow-up were performed by the same blinded assessor (T.H.) as at the previous follow-ups. The 15-year result for the Constant score was the outcome of primary interest15. Secondary outcome measures were the self-report section of the American Shoulder and Elbow Surgeons (ASES) score16, the Short Form-36 (SF-36) Health Survey17, and assessments of pain on a 10-cm visual analog scale (VAS), strength (with a handheld spring balance), pain-free mobility of the shoulder (with a goniometer), and patient satisfaction. To measure patient satisfaction, patients were asked to respond to the question, “How satisfied are you with the treatment result of your shoulder?” using a VAS ranging from 0 (very unsatisfied) to 10 (very satisfied). The clinical relevance of the results for the Constant and ASES score assessments was investigated by using the patient acceptable symptom state (PASS) and the percentage of maximal outcome improvement (%MOI) as cutoffs in a proportion analysis18–20. Sonographic follow-up included assessment of tear widening in unrepaired tears and of repair integrity in repaired tears, and was performed blinded to the patient’s clinical data and the results of earlier ultrasound examinations. A Sonoline Antares (Siemens Medical; VFX13-5 linear transducer, 8.5 to 11.5 MHz) and Logiq S8 scanner (General Electric; ML6-15D linear transducer, 4 to 15 MHz) were used by an experienced sonographer whose accuracy in diagnosing rotator cuff tears has been documented21,22. The diagnostic criteria for rotator cuff tears and full- and partial-thickness retears that were used in the current study were previously described23–26.
Randomization
The randomization list (block length of 20 and 1:1 ratio) was generated by an external statistician before the start of the study. The randomization sequence was concealed from the study’s collaborators until interventions were assigned, and from the outcome assessor for the entire duration of the study.
Surgical Procedures and Postoperative Management
Tendon repair was performed using a mini-open or open approach, with the patient in the beach-chair position. Following diagnostic arthroscopy, the tear was exposed through a deltoid-splitting approach. An acromioplasty was performed as described by Neer27. The rotator cuff was mobilized until coverage of the footprint was achieved. The footprint was prepared to bleeding bone, and a tendon-to-bone repair was performed with transosseous sutures using a Mason-Allen technique. Tenodesis of the long head of the biceps to the intertubercular sulcus was performed when the tendon was found to be degenerated or partially torn. A detailed description of the surgical method was previously reported1,2.
Postoperatively, the arm was immobilized in a sling for 5 to 6 weeks, and passive range-of-motion exercises were started immediately. Active range-of-motion exercises were initiated 6 weeks after surgery, and were supplemented by strengthening exercises after 12 weeks.
Physiotherapy
The physiotherapy group received outpatient treatment at our hospital from 4 physiotherapists experienced in shoulder rehabilitation. A previously reported rehabilitation program, which was individually adapted, was followed2,28. During the first 12 weeks, treatment sessions of 40 minutes were given twice weekly, with decreasing frequency during the following 6 to 12 weeks. No supplementary treatment measures, such as anti-inflammatory or analgesic medication, were given.
Secondary Surgery
Patients with insufficient progress after at least 15 physiotherapy sessions could ask for reexamination at any point during the study. In cases of persistent clinical findings, secondary surgical treatment was offered.
Statistical Analysis
The original sample-size calculation was previously presented and was performed on the basis of a t test of the primary outcome1. It showed that a group size of 45 patients was needed to detect a 12-point difference in the Constant score with a power of 80% and a 2-sided significance level of 0.05.
We present categorical baseline data as the number of patients and continuous data as the mean and standard deviation. To handle missing data, we analyzed our primary and secondary outcome data by using a linear mixed model for repeated measurements under a “missing at random” assumption. All available data were used in the analysis. Analyses were adjusted for baseline differences in the respective dependent variable and patient age and were performed according to the intention-to-treat principle. We estimated the linear mixed model by using linear maximum likelihood and included a random intercept, the baseline value of the dependent variable, and patient age as covariates, and observation time after the intervention and the type of intervention as well as their interaction term as factors. We used mean differences between groups at the 15-year follow-up from the linear mixed model to assess differences between interventions. We considered a 2-sided p value of <0.05 to be significant. The same statistical method was used for a supplementary as-treated analysis comparing the 15-year results for the Constant score between patients treated by primary surgery, physiotherapy only, or physiotherapy followed by secondary surgery.
The change in the between-group difference in the Constant score from the 2-year to the 15-year follow-up was assessed by a linear mixed-model analysis by estimation of the linear combination of change between follow-up times.
We explored the clinical importance of the results by a between-group comparison of the proportions of patients meeting or exceeding the PASS value and the %MOI predicting treatment satisfaction for the Constant and the ASES scores18–20.
A paired samples t test was used to compare the increase in tear size from baseline to follow-up in the physiotherapy-only group, and an independent samples t test was used to compare results between intact repairs and retears and to investigate differences in patient satisfaction between the tendon repair group and the physiotherapy group. Normalized Constant scores were also calculated and are given in Appendix 129.
Results
At 15 years, 83 (81%) of the original 103 patients were available for follow-up (43 randomized to primary tendon repair and 40 randomized to physiotherapy with optional secondary repair). Patients with missing values were, on average, 5 years older than those with a complete data set but did not differ with respect to any of the other data registered at baseline. Over the 15-year follow-up period, 15 (29%) of the patients had crossed over from physiotherapy to secondary surgical repair. Patient characteristics at the time of enrollment are given in Table II. Adverse events and the need for additional therapeutic measures during follow-up are presented in Table III.
TABLE II -
Patient Characteristics at the Time of Enrollment
|
Primary Tendon Repair (N = 52) |
Physiotherapy with Optional Secondary Repair (N = 51) |
| Age*(yr)
|
59 ± 7.5 |
61 ± 7.6 |
| Male sex† |
37 |
36 |
| Right side affected† |
31 |
29 |
| Tear on dominant side† |
33 |
31 |
| Shoulder-demanding activities† |
26 |
28 |
| Duration of symptoms*(mo)
|
12.3 ± 18.7 |
9.8 ± 9.8 |
| Tear size on ultrasound*(mm)
|
|
|
| Anterior-posterior direction |
15.6 ± 6.7 |
14.3 ± 6.3 |
| Medial-lateral direction |
14.9 ± 5.7 |
14.7 ± 6.9 |
| Type of injury† |
|
|
| Acute-on-chronic‡ |
30 |
29 |
| Chronic§ |
22 |
22 |
| Occupational situation† |
|
|
| On job |
23 |
24 |
| On sick leave |
15 |
8 |
| Retired |
11 |
17 |
| Receiving disability benefit |
3 |
2 |
| Earlier treatment† |
|
|
| Physiotherapy |
28 |
21 |
| Cortisone injections |
5 |
10 |
| Nonsteroidal anti-inflammatory drug |
7 |
9 |
| None |
12 |
11 |
| Muscle atrophy on MRI41† |
|
|
| Grade 0 |
26 |
23 |
| Grade 1 |
12 |
18 |
| Grade 2 |
13 |
10 |
| MRI not available |
1 |
0 |
| Localization of tear on ultrasound† |
|
|
| Supraspinatus only |
37 |
40 |
| Supraspinatus and infraspinatus |
14 |
10 |
| Supraspinatus and subscapularis |
1 |
1 |
| Current smoking status† |
|
|
| Nonsmoker |
37 |
44 |
| ≤10 cigarettes per day |
10 |
3 |
| >10 cigarettes per day |
5 |
4 |
*The values are given as the mean and the standard deviation.
†The values are given as the number of patients.
‡A tear in a shoulder with an acute onset of pain following minor trauma.
§A tear in a shoulder with a gradual onset of pain in the absence of trauma.
TABLE III -
Adverse Events and Need for Additional Therapeutic Measures During Follow-up
|
Primary Tendon Repair |
Physiotherapy with Optional Secondary Surgery |
| Site other than the index shoulder |
|
|
| Medical event |
Polymyalgia rheumatica (n = 1)*
Cerebral apoplexy (n = 1)†, (n = 1)§
Chronic lymphocytic leukemia (n = 1)‡
Lymphoma (n = 1)§
Myocardial infarction (n = 1)§
Implantation of pacemaker (n = 1)§
Atrial fibrillation (n = 1)§
Renal failure (n = 1)§ |
Polymyalgia rheumatica (n = 1)*
Herpes zoster (n = 1)*
Lymphoma (n = 1)*
Leukemia (n = 2)‡ |
| Surgical event |
Operation for abdominal aortic aneurysm (n = 1)†
Hepatic transplantation (n = 1)‡
Lumbar discectomy (n = 1)‡
Operation for prostate cancer (n = 1)§
Tendon repair in the contralateral shoulder (n = 3)‡
Hip replacement (n = 1)§
Knee replacement (n = 1)§ |
Acromioplasty in the contralateral shoulder (n = 1)‡
Tendon repair in the contralateral shoulder (n = 1)‡
Operation for spinal stenosis (n = 2)§
Operation with artificial heart valve (n = 2)§
Hip replacement (n = 1)§
Knee replacement (n = 2)§ |
| Musculoskeletal event |
Lateral humeral epicondylitis (n = 1)*
Low back pain (n = 1)†
Cervical radiculopathy (n = 1)*, (n = 1)†, (n = 1)§ |
Cervical radiculopathy (n = 1)* |
| Index shoulder |
|
|
| Need for additional therapeutic measures |
Physiotherapy (n = 1)*
Reoperation with acromioplasty and biceps tenotomy (n = 1)† |
Physiotherapy (n = 3)†, (n = 1)‡
Glenohumeral arthrosis, conservatively treated (n = 1)‡ |
| New shoulder trauma |
Fracture of the humerus, conservatively treated (n = 1)*
Contusion of the shoulder (n = 1)†, (n = 2)‡
Fracture of the humerus, conservatively treated (n = 1)‡
Traumatic retear of the rotator cuff, surgically treated (n = 2)§ |
Contusion of the shoulder (n = 2)*, (n = 1)†, (n = 2)‡, (n = 2)§
Shoulder dislocation (n = 1)§
Spontaneous tear of the long head of the biceps tendon (n = 1)§ |
*Time of occurrence before 2-year follow-up.
†Time of occurrence between 2 and 5-year follow-ups.
‡Time of occurrence between 5 and 10-year follow-ups.
§Time of occurrence between 10 and 15-year follow-ups.
Treatment Effects
Primary Outcome
At 15 years, the mean Constant score was 79.9 points in the primary tendon repair group and 68.5 points in the physiotherapy with optional secondary repair group. The between-group difference of 11.8 points (95% confidence interval [CI], 5.1 to 18.5 points) was significant (p = 0.001) (Table IV, Fig. 2-A). The difference between the treatment groups increased significantly from 2 to 15 years (by 8.8 points [95% CI, 2.0 to 15.7 points]; p = 0.01). The clinical relevance of the result was supported by a significantly higher percentage of patients in the tendon repair group who met or exceeded the reported values for the PASS (70% versus 43%; p = 0.02) and %MOI for treatment satisfaction (86% versus 53%; p = 0.002) (see Appendices 2a and 2b)18,20.
TABLE IV -
Summary of Primary and Secondary Outcomes by Treatment Group
|
Primary Tendon Repair (N = 52/51/51/51/51/48/43)* |
Physiotherapy with Optional Secondary Repair (N = 51/51/51/50/50/43/40)* |
Between-Group Difference |
| Mean (95% CI)† |
P Value‡ |
| Primary |
|
|
|
|
| Constant score (points)
|
|
|
|
|
| Baseline |
35.3 ± 13.2 |
38.4 ± 14.2 |
|
|
| 6 mo |
65.6 ± 16.3 |
63.9 ± 20.2 |
2.7 (−3.6 to 9.0) |
|
| 1 yr |
77.7 ± 13.4 |
70.3 ± 19.1 |
8.3 (2.0 to 14.5) |
|
| 2 yr |
79.3 ± 13.6 |
77.7 ± 14.9 |
3.0 (−3.3 to 9.3) |
|
| 5 yr |
79.8 ± 15.0 |
74.2 ± 20.3 |
6.8 (0.5 to 13.1) |
|
| 10 yr |
80.5 ± 9.8 |
71.8 ± 17.8 |
10.0 (3.5 to 16.5) |
|
| 15 yr |
79.9 ± 9.7 |
68.5 ± 22.2 |
11.8 (5.1 to 18.5) |
0.001 |
| Secondary |
|
|
|
|
| ASES score§(points)
|
|
|
|
|
| Baseline |
45.5 ± 14.5 |
48.2 ± 14.4 |
|
|
| 6 mo |
85.3 ± 13.7 |
75.4 ± 20.2 |
10.7 (4.2 to 17.2) |
|
| 1 yr |
93.6 ± 12.5 |
83.6 ± 18.3 |
10.8 (4.3 to 17.2) |
|
| 2 yr |
93.1 ± 13.9 |
88.0 ± 14.9 |
6.4 (0.1 to 12.9) |
|
| 5 yr |
92.8 ± 13.3 |
85.4 ± 21.0 |
8.4 (1.9 to 14.9) |
|
| 10 yr |
94.0 ± 9.5 |
80.0 ± 20.2 |
15.2 (8.4 to 22.0) |
|
| 15 yr |
92.0 ± 10.9 |
78.7 ± 27.8 |
13.9 (6.9 to 20.9) |
<0.001 |
| VAS pain (cm)
|
|
|
|
|
| Baseline |
5.6 ± 2.0 |
5.3 ± 1.9 |
|
|
| 6 mo |
1.1 ± 1.3 |
2.7 ± 2.2 |
1.6 (0.9 to 2.3) |
|
| 1 yr |
0.5 ± 1.2 |
1.6 ± 1.6 |
1.1 (0.4 to 1.8) |
|
| 2 yr |
0.7 ± 1.5 |
1.4 ± 1.4 |
0.8 (0.1 to 1.5) |
|
| 5 yr |
0.6 ± 1.4 |
1.6 ± 1.6 |
1.1 (0.4 to 1.8) |
|
| 10 yr |
0.6 ± 1.3 |
2.3 ± 2.4 |
1.8 (1.1 to 2.5) |
|
| 15 yr |
0.6 ± 1.1 |
2.4 ± 3.0 |
1.8 (1.1 to 2.6) |
<0.001 |
| Pain-free abduction (deg)
|
|
|
|
|
| Baseline |
73.7 ± 28.0 |
81.9 ± 29.8 |
|
|
| 6 mo |
135.4 ± 41.7 |
135.4 ± 47.9 |
2.4 (−12.1 to 16.9) |
|
| 1 yr |
158.4 ± 33.7 |
143.8 ± 43.9 |
17.1 (2.6 to 31.6) |
|
| 2 yr |
161.7 ± 30.8 |
163.6 ± 32.6 |
1.2 (−13.3 to 15.8) |
|
| 5 yr |
167.3 ± 30.6 |
155.1 ± 41.2 |
15.2 (0.7 to 29.8) |
|
| 10 yr |
169.1 ± 23.8 |
151.7 ± 40.9 |
19.9 (4.8 to 35.1) |
|
| 15 yr |
168.1 ± 30.4 |
152.4 ± 48.4 |
16.2 (0.6 to 31.8) |
0.04 |
| Pain-free flexion (deg)
|
|
|
|
|
| Baseline |
86.8 ± 41.3 |
88.6 ± 32.1 |
|
|
| 6 mo |
147.3 ± 34.5 |
146.6 ± 46.3 |
1.3 (−11.1 to 13.7) |
|
| 1 yr |
166.1 ± 27.5 |
155.6 ± 38.4 |
11.1 (−1.3 to 23.5) |
|
| 2 yr |
168.5 ± 26.1 |
170.5 ± 23.0 |
−0.6 (−13.1 to 11.9) |
|
| 5 yr |
170.6 ± 27.9 |
163.5 ± 35.4 |
8.3 (−4.2 to 20.8) |
|
| 10 yr |
175.8 ± 12.0 |
162.0 ± 35.5 |
15.1 (2.2 to 28.1) |
|
| 15 yr |
176.4 ± 16.2 |
153.8 ± 45.3 |
22.4 (9.0 to 35.9) |
0.001 |
| Strength (kg)
|
|
|
|
|
| Baseline |
7.5 ± 5.5 |
8.1 ± 5.8 |
|
|
| 6 mo |
8.0 ± 4.6 |
10.6 ± 5.4 |
−2.6 (−4.3 to −0.8) |
|
| 1 yr |
11.1 ± 4.0 |
11.9 ± 5.1 |
−0.7 (−2.5 to 1.0) |
|
| 2 yr |
11.9 ± 4.3 |
12.8 ± 5.3 |
−0.8 (−2.6 to 1.0) |
|
| 5 yr |
12.1 ± 4.7 |
11.4 ± 5.4 |
0.9 (−0.9 to 2.6) |
|
| 10 yr |
11.7 ± 4.5 |
10.2 ± 5.6 |
1.7 (0.1 to 3.5) |
|
| 15 yr |
10.2 ± 5.4 |
8.4 ± 5.8 |
1.8 (0.0 to 3.6) |
0.05 |
*The values are the raw measurement data, given as the mean and the standard deviation. N = number of patients at the 7 measurement points.
†The values were adjusted for baseline measurements of the variable and patient age; positive values indicate a better result for primary tendon repair.
‡A p value of <0.05 indicates a significant between-group difference at the 15-year follow-up.
§Self-report section of the ASES score.
Figs. 2-A through 2-F Graphs showing mean results at baseline and all follow-up times for the Constant score (Fig. 2-A); ASES score (Fig. 2-B); VAS score for pain (Fig. 2-C); active, pain-free abduction and flexion (Figs. 2-D and 2-E); and strength (Fig. 2-F) in the 2 study groups. The cumulative number of patients who had crossed over from physiotherapy to secondary surgical treatment was 3 patients at 6 months, 9 at 1 year, 12 at 2 and 5 years, 14 at 10 years, and 15 at 15 years. Analysis is by intention to treat, with the results from secondary surgery included in the physiotherapy group. Whiskers represent the 95% confidence interval. Higher values on the y axis represent better results, except for the VAS for pain, for which lower values represent a better result.
Secondary Outcomes
Significant differences in favor of primary tendon repair were found for the ASES score (difference, 13.9 points [95% CI, 6.9 to 20.9 points]; p < 0.001), the 10-cm VAS for pain (difference, 1.8 cm [95% CI, 1.1 to 2.6 cm]; p < 0.001), and active pain-free shoulder abduction (difference, 16.2° [95% CI, 0.6° to 31.8°]; p = 0.04) and flexion (difference, 22.4° [95% CI, 9.0° to 35.9°]; p = 0.001). For shoulder strength, the difference of 1.8 kg in favor of tendon repair was just below the level of significance (95% CI, 0.0 to 3.6 kg; p = 0.05) (Table IV, Figs. 2-B through 2-F). Patient satisfaction as measured on a 10-cm VAS was superior in the tendon repair group compared with the physiotherapy group (mean, 9.2 compared with 7.9 cm; difference, 1.3 cm [95% CI, 0.3 to 2.3 cm]; p = 0.02). Regarding the SF-36 score for quality of life, the between-group differences for the 8 component scales and the 2 summary scales for physical and mental health did not reach significance (see Appendix 3). The clinical relevance of the results for the ASES score was supported by the larger number of patients in the tendon repair group who met or exceeded the corresponding values for the PASS and %MOI for treatment satisfaction (see Appendices 4a and 4b).
Secondary Surgical Treatment
Fifteen (29%) of the 51 patients in the physiotherapy group did not achieve a satisfactory result and crossed over to secondary tendon repair (13 patients) or acromioplasty (2 patients): 9 during the first year, 3 during the second year, 2 between 5 and 10 years, and 1 between 10 and 15 years. Twelve patients with a secondary repair attended the 15-year follow-up, and they had a mean Constant score that did not differ from that of the primary tendon repair group but was 16.5 points superior to that of patients with physiotherapy only (95% CI, 9.4 to 23.6 points; p < 0.001) (Fig. 3).
Fig. 3: Results from a post-hoc, as-treated analysis for the Constant score at baseline and all follow-up times for the primary tendon repair group, the physiotherapy-only group, and the secondary surgery group. Mean scores are shown; n = number of patients in each group at the 7 measurement points.
Tear-Size Increase and Shoulder Function
Repeated ultrasound measurements of tear size were performed in 26 of 28 patients who were treated by physiotherapy only. The mean tear-size increase from baseline to 15 years was from 16.2 to 31.6 mm in the anterior-posterior direction (difference, 15.4 mm [95% CI, 10.2 to 20.6 mm]; p < 0.001) and from 16.0 to 26.9 mm in the medial-lateral direction (difference, 11.0 mm [95% CI, 6.7 to 15.2 mm]; p < 0.001) (Table V). This was associated with a drop in the Constant score in the physiotherapy-only group from 78.8 points at 2 years to 63.7 points at 15 years, which was 16.5 points below the result for primary tendon repair at the last follow-up (95% CI, 9.4 to 23.6 points; p < 0.001) (Fig. 3; see also Appendix 5).
TABLE V -
Results from Repeated Sonographic Tear-Size Measurements and Constant Scores in 26 Unrepaired Rotator Cuff Tears
*
|
Baseline |
5 Yr |
10 Yr |
15 Yr |
| Tear size in the anterior-posterior direction (mm)
|
16.2 ± 5.8 |
20.8 ± 10.2 |
28.0 ± 13.9 |
31.6 ± 14.0 |
| Tear size in the medial-lateral direction (mm)
|
16.0 ± 7.0 |
19.4 ± 8.9 |
23.1 ± 11.0 |
26.9 ± 11.1 |
| Constant score (points)
|
37.9 ± 13.0 |
73.2 ± 21.3 |
69.5 ± 17.1 |
62.5 ± 25.3 |
*The values are given as the mean and the standard deviation.
Among the 43 patients in the primary tendon repair group who attended the 15-year follow-up, 10 had been diagnosed with a recurrent tear (4 full-thickness and 6 partial-thickness) on magnetic resonance imaging (MRI) at 1 year after surgery, and an additional 6 were diagnosed with a later retear on ultrasound (2 after an adequate traumatic event). Only the 2 traumatic retears were treated by repeat surgery, but tendon healing was not achieved in either case. At 15 years, results for those with intact repairs were superior, with a mean Constant score of 82.2 points compared with 76.1 points for those with a retear (difference, 6.1 points [95% CI, 0.1 to 12.0 points]; p = 0.05).
Discussion
This study found significant and clinically relevant between-group differences in favor of primary tendon repair at 15-year follow-up. The findings are consistent with previous reports from the same study, which showed better outcomes for tendon repair compared with physiotherapy at 5 and 10-year follow-ups. The study implies that while both treatment options may yield comparable results in the short term, the differences between the groups increase over time in favor of tendon repair, mainly because of a gradual deterioration in physiotherapy outcomes.
Clinical decision-making regarding nonoperative or operative treatment for small-to-medium-sized rotator cuff tears is challenging. Better long-term results in the present study may support a primary operative approach. However, physiotherapy together with optional secondary repair also resulted in improvement compared with baseline, and most physiotherapy patients did not require surgery. Both treatment options should be discussed during counseling, but patients should be informed that while physiotherapy can be effective, long-term results from primary tendon repair are superior.
Our finding of a time-dependent decrease in the results from physiotherapy is supported by reports from studies of the natural course of unrepaired tears showing that they are at risk for anatomic and functional deterioration10,11,30–35. According to the literature, tear widening is likely to occur in approximately 50% of unrepaired rotator cuff tears and may progress over many years, highlighting the need for long-term monitoring36. In the present study, tears in the physiotherapy group that remained unrepaired for 15 years had progressively increased in size, and this was associated with a decrease in the Constant score. The observation that these patients had little interest in supplementary treatment measures at their last follow-up indicates that they may have developed effective coping strategies that enabled them to live well despite restrictions in shoulder function.
Over the 15-year follow-up period, 29% of the patients initially assigned to physiotherapy had crossed over to secondary surgical treatment but, in accordance with the intention-to-treat principle, they remained in the physiotherapy group for analysis. This led to a more favorable outcome in the physiotherapy group, since unsatisfactory results from physiotherapy were replaced by better results from secondary surgery, thereby reducing the difference between the groups. However, this reflects current clinical practice, where crossover is an option after an unsuccessful trial of physiotherapy. It is noteworthy that the crossover group’s results at 15 years for the first time equaled those of the primary surgery group, which is in contrast to the 5 and 10-year results. However, since only 12 crossover patients attended the final follow-up, this finding should be interpreted with caution.
Strengths and Limitations
The strengths of the study are its long observation time and high follow-up rate of 81% at 15 years. The study provides new insights into the previously unexplored period from 10 to 15 years after operative and nonoperative treatment of small-to-medium-sized rotator cuff tears.
The following limitations need to be considered. First, loss to follow-up is inevitable in long-term studies and may have affected our study outcomes. However, the study’s follow-up rate of 81% exceeds the reported cutoff of 80% for an acceptable follow-up in randomized trials37, and with 83 attendees at the most recent follow-up, the study retained sufficient statistical power for the main analyses.
Second, over 15 years, external factors other than the treatment selection may have affected the results. However, we regularly monitored the study patients and urged them to contact us immediately in the event of any shoulder-related incident. Additionally, we recorded all other medical and surgical events during routine follow-ups and found them evenly distributed between the 2 groups.
Third, the non-inclusion of 63% of the patients assessed for eligibility who did not match the study’s eligibility criteria or declined to participate may have weakened the study’s external validity.
Conclusions
This 15-year follow-up report, which compared the effects of tendon repair and physiotherapy for small-to-medium-sized rotator cuff tears, showed that initially small between-group differences increased over time in favor of tendon repair and reached statistical significance and clinical relevance. The results provide a better understanding of the long-term impact of our treatment decisions and will help to select the treatment that best meets patients’ present and future shoulder demands.
Appendix
Supporting material provided by the authors is posted with the online version of this article as a data supplement at jbjs.org (https://links.lww.com/JBJS/I155).
Data Sharing
A data-sharing statement is provided with the online version of the article (https://links.lww.com/JBJS/I156).
References
1. Moosmayer S, Lund G, Seljom U, Svege I, Hennig T, Tariq R, Smith HJ. Comparison between surgery and physiotherapy in the treatment of small and medium-sized tears of the rotator cuff: A randomised controlled study of 103 patients with one-year follow-up. J Bone Joint Surg Br. 2010 Jan;92(1):83-91.
2. Moosmayer S, Lund G, Seljom US, Haldorsen B, Svege IC, Hennig T, Pripp AH, Smith HJ. Tendon repair compared with physiotherapy in the treatment of rotator cuff tears: a randomized controlled study in 103 cases with a five-year follow-up. J Bone Joint Surg Am. 2014 Sep 17;96(18):1504-14.
3. Lambers Heerspink FO, van Raay JJ, Koorevaar RC, van Eerden PJ, Westerbeek RE, van ’t Riet E, van den Akker-Scheek I, Diercks RL. Comparing surgical repair with conservative treatment for degenerative rotator cuff tears: a randomized controlled trial. J Shoulder Elbow Surg. 2015 Aug;24(8):1274-81.
4. Kukkonen J, Joukainen A, Lehtinen J, Mattila KT, Tuominen EK, Kauko T, Äärimaa V. Treatment of nontraumatic rotator cuff tears: A randomized controlled trial with two years of clinical and imaging follow-up. J Bone Joint Surg Am. 2015 Nov 4;97(21):1729-37.
5. Moosmayer S, Lund G, Seljom US, Haldorsen B, Svege IC, Hennig T, Pripp AH, Smith HJ. At a 10-year follow-up, tendon repair is superior to physiotherapy in the treatment of small and medium-sized rotator cuff tears. J Bone Joint Surg Am. 2019 Jun 19;101(12):1050-60.
6. Ranebo MC, Björnsson Hallgren HC, Holmgren T, Adolfsson LE. Surgery and physiotherapy were both successful in the treatment of small, acute, traumatic rotator cuff tears: a prospective randomized trial. J Shoulder Elbow Surg. 2020 Mar;29(3):459-70.
7. Cederqvist S, Flinkkilä T, Sormaala M, Ylinen J, Kautiainen H, Irmola T, Lehtokangas H, Liukkonen J, Pamilo K, Ridanpää T, Sirniö K, Leppilahti J, Kiviranta I, Paloneva J. Non-surgical and surgical treatments for rotator cuff disease: a pragmatic randomised clinical trial with 2-year follow-up after initial rehabilitation. Ann Rheum Dis. 2021 Jun;80(6):796-802.
8. Kukkonen J, Ryösä A, Joukainen A, Lehtinen J, Kauko T, Mattila K, Äärimaa V. Operative versus conservative treatment of small, nontraumatic supraspinatus tears in patients older than 55 years: over 5-year follow-up of a randomized controlled trial. J Shoulder Elbow Surg. 2021 Nov;30(11):2455-64.
9. Maman E, Harris C, White L, Tomlinson G, Shashank M, Boynton E. Outcome of nonoperative treatment of symptomatic rotator cuff tears monitored by magnetic resonance imaging. J Bone Joint Surg Am. 2009 Aug;91(8):1898-906.
10. Safran O, Schroeder J, Bloom R, Weil Y, Milgrom C. Natural history of nonoperatively treated symptomatic rotator cuff tears in patients 60 years old or younger. Am J Sports Med. 2011 Apr;39(4):710-4.
11. Moosmayer S, Gärtner AV, Tariq R. The natural course of nonoperatively treated rotator cuff tears: an 8.8-year follow-up of tear anatomy and clinical outcome in 49 patients. J Shoulder Elbow Surg. 2017 Apr;26(4):627-34.
12. Kim YS, Kim SE, Bae SH, Lee HJ, Jee WH, Park CK. Tear progression of symptomatic full-thickness and partial-thickness rotator cuff tears as measured by repeated MRI. Knee Surg Sports Traumatol Arthrosc. 2017 Jul;25(7):2073-80.
13. Ranebo MC, Björnsson Hallgren HC, Norlin R, Adolfsson LE. Clinical and structural outcome 22 years after acromioplasty without tendon repair in patients with subacromial pain and cuff tears. J Shoulder Elbow Surg. 2017 Jul;26(7):1262-70.
14. Yamamoto N, Mineta M, Kawakami J, Sano H, Itoi E. Risk Factors for Tear Progression in Symptomatic Rotator Cuff Tears: A Prospective Study of 174 Shoulders. Am J Sports Med. 2017 Sep;45(11):2524-31.
15. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop Relat Res. 1987 Jan;(214):160-4.
16. Richards RR, An KN, Bigliani LU, Friedman RJ, Gartsman GM, Gristina AG, Iannotti JP, Mow VC, Sidles JA, Zuckerman JD. A standardized method for the assessment of shoulder function. J Shoulder Elbow Surg. 1994 Nov;3(6):347-52.
17. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992 Jun;30(6):473-83.
18. Kanto K, Lähdeoja T, Paavola M, Aronen P, Järvinen TLN, Jokihaara J, Ardern CL, Karjalainen TV, Taimela S. Minimal important difference and patient acceptable symptom state for pain, Constant-Murley score and Simple Shoulder Test in patients with subacromial pain syndrome. BMC Med Res Methodol. 2021 Mar 6;21(1):45.
19. Cvetanovich GL, Gowd AK, Liu JN, Nwachukwu BU, Cabarcas BC, Cole BJ, Forsythe B, Romeo AA, Verma NN. Establishing clinically significant outcome after arthroscopic rotator cuff repair. J Shoulder Elbow Surg. 2019 May;28(5):939-48.
20. Beck EC, Gowd AK, Liu JN, Waterman BR, Nicholson KF, Forsythe B, Yanke AB, Cole BJ, Verma NN. How Is Maximum Outcome Improvement Defined in Patients Undergoing Shoulder Arthroscopy for Rotator Cuff Repair? A 1-Year Follow-Up Study. Arthroscopy. 2020 Jul;36(7):1805-10.
21. Moosmayer S, Smith HJ. Diagnostic ultrasound of the shoulder—a method for experts only? Results from an orthopedic surgeon with relative inexpensive compared to operative findings. Acta Orthop. 2005 Aug;76(4):503-8.
22. Moosmayer S, Heir S, Smith HJ. Sonography of the rotator cuff in painful shoulders performed without knowledge of clinical information: results from 58 sonographic examinations with surgical correlation. J Clin Ultrasound. 2007 Jan;35(1):20-6.
23. Mack LA, Nyberg DA, Matsen FR 3rd, Kilcoyne RF, Harvey D. Sonography of the postoperative shoulder. AJR Am J Roentgenol. 1988 May;150(5):1089-93.
24. Middleton WD, Teefey SA, Yamaguchi K. Sonography of the Shoulder. Semin Musculoskelet Radiol. 1998;2(3):211-22.
25. Bouffard JA, Lee SM, Dhanju J. Ultrasonography of the shoulder. Semin Ultrasound CT MR. 2000 Jun;21(3):164-91.
26. Gilat R, Atoun E, Cohen O, Tsvieli O, Rath E, Lakstein D, Levy O. Recurrent rotator cuff tear: is ultrasound imaging reliable? J Shoulder Elbow Surg. 2018 Jul;27(7):1263-7.
27. Neer CS 2nd. Anterior acromioplasty for the chronic impingement syndrome in the shoulder: a preliminary report. J Bone Joint Surg Am. 1972 Jan;54(1):41-50.
28. Lund G, Seljom U, Moosmayer S, Haldorsen B, Svege IC. [Physiotherapy of the shoulder]. Oslo: Legeforlaget; 2013. Norwegian.
29. Katolik LI, Romeo AA, Cole BJ, Verma NN, Hayden JK, Bach BR. Normalization of the Constant score. J Shoulder Elbow Surg. 2005 May-Jun;14(3):279-85.
30. Yamaguchi K, Tetro AM, Blam O, Evanoff BA, Teefey SA, Middleton WD. Natural history of asymptomatic rotator cuff tears: a longitudinal analysis of asymptomatic tears detected sonographically. J Shoulder Elbow Surg. 2001 May-Jun;10(3):199-203.
31. Zingg PO, Jost B, Sukthankar A, Buhler M, Pfirrmann CW, Gerber C. Clinical and structural outcomes of nonoperative management of massive rotator cuff tears. J Bone Joint Surg Am. 2007 Sep;89(9):1928-34.
32. Mall NA, Kim HM, Keener JD, Steger-May K, Teefey SA, Middleton WD, Stobbs G, Yamaguchi K. Symptomatic progression of asymptomatic rotator cuff tears: a prospective study of clinical and sonographic variables. J Bone Joint Surg Am. 2010 Nov 17;92(16):2623-33.
33. Moosmayer S, Tariq R, Stiris M, Smith HJ. The natural history of asymptomatic rotator cuff tears: a three-year follow-up of fifty cases. J Bone Joint Surg Am. 2013 Jul 17;95(14):1249-55.
34. Keener JD, Galatz LM, Teefey SA, Middleton WD, Steger-May K, Stobbs-Cucchi G, Patton R, Yamaguchi K. A prospective evaluation of survivorship of asymptomatic degenerative rotator cuff tears. J Bone Joint Surg Am. 2015 Jan 21;97(2):89-98.
35. Hsu J, Keener JD. Natural History of Rotator Cuff Disease and Implications on Management. Oper Tech Orthop. 2015 Mar 1;25(1):2-9.
36. Kwong CA, Ono Y, Carroll MJ, Fruson LW, More KD, Thornton GM, Lo IKY. Full-Thickness Rotator Cuff Tears: What Is the Rate of Tear Progression? A Systematic Review. Arthroscopy. 2019 Jan;35(1):228-34.
37. Fewtrell MS, Kennedy K, Singhal A, Martin RM, Ness A, Hadders-Algra M, Koletzko B, Lucas A. How much loss to follow-up is acceptable in long-term randomised trials and prospective studies? Arch Dis Child. 2008 Jun;93(6):458-61.
38. Kessel L, Watson M. The painful arc syndrome. Clinical classification as a guide to management. J Bone Joint Surg Br. 1977 May;59(2):166-72.
39. Neer CS 2nd. Impingement lesions. Clin Orthop Relat Res. 1983 Mar;(173):70-7.
40. Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. Am J Sports Med. 1980 May-Jun;8(3):151-8.
41. Thomazeau H, Rolland Y, Lucas C, Duval JM, Langlais F. Atrophy of the supraspinatus belly. Assessment by MRI in 55 patients with rotator cuff pathology. Acta Orthop Scand. 1996 Jun;67(3):264-8.
