INTRODUCTION
Shoulder stiffness is commonly encountered in patients with rotator cuff tears [1-3]. In the surgical management of rotator cuff tears with shoulder stiffness, a two-stage treatment has been recommended, where stiffness is resolved conservatively before rotator cuff repair [4]. However, recent studies indicate that single-stage surgery can achieve the same outcomes in stiff and non-stiff shoulders [1,3,5-11]. In most previous studies, arthroscopic global capsular release combined with manipulation under anesthesia (MUA) was routinely performed to improve range of motion (ROM) after surgery [10,12,13]. However, we question whether the arthroscopic global capsular release, a standard surgical procedure for a primary frozen shoulder, is necessary to resolve shoulder stiffness associated with rotator cuff tears [11]. Oh et al. [7] found that many patients with rotator cuff tears showed localized stiffness. Ueda et al. [14] reported that severe and global loss of passive motion was rarely present in shoulders with full-thickness rotator cuff tears. Based on these findings, we propose that arthroscopic global capsular release is not always necessary to resolve shoulder stiffness associated with rotator cuff tear. In many cases, MUA aimed at improving elevation through the release of the anteroinferior capsule [7] may be sufficient. Although complications such as fractures and brachial plexus injuries have been reported with MUA, the benefits of MUA in avoiding unnecessary surgical intervention in many cases are considered to outweigh the associated risks.
Only a few studies [3,8] to date have reported the results of arthroscopic rotator cuff repair (ARCR) with MUA alone in shoulders with preoperative stiffness. In these studies, patients with preoperative stiffness who underwent ARCR with MUA alone achieved comparable clinical outcomes to those without stiffness. Whether preoperative stiffness is associated with better rotator cuff healing after ARCR remains controversial. Several recent studies have indicated that stiff shoulders are more likely to heal after ARCR than non-stiff shoulders. McGrath et al. [15] reported that patients with preoperative stiffness had no retears (0%) at 2 years after ARCR, while 34 retears (20%) were documented among patients in the non-stiffness group. Kim and Jung [1] and Jeong et al. [2] reported similarly favorable results among shoulders with preoperative stiffness.
The present study aimed to examine whether patients with shoulder stiffness who underwent ARCR combined with MUA but without arthroscopic capsular release would achieve comparable clinical results to those without shoulder stiffness. We also aimed to compare retear rates after ARCR between patients with and without shoulder stiffness. We hypothesized that clinical outcomes and retear rates would not differ significantly between the two groups.
METHODS
This study was approved by the Institutional Review Board of the Tokushima Red Cross Hospital (No. Tokubyoso-509). As the present study was categorized as a retrospective study, the Ethical Review Board Committee waived the patients’ informed consent requirement.
Patients
Four hundred eighty-five consecutive patients who underwent ARCR between January 2012 and May 2022 were enrolled. The inclusion criteria were as follows: (1) full-thickness posterosuperior rotator cuff tear confirmed by arthroscopy; (2) magnetic resonance imaging (MRI) examination performed 6 months after surgery; and (3) data from a minimum of 2 years follow-up, including clinical examinations at 3, 6, 12, and 24 months postoperatively. The exclusion criteria were osteoarthritis, instability, a history of previous shoulder surgery, partial-thickness tear, isolated subscapularis tendon tear, partial repair, less than 2 years of follow-up, and a lack of postoperative MRI scans or clinical examination during the follow-up period. Patients with subscapularis tears combined with posterosuperior full-thickness tears were included. Patients who required arthroscopic capsular release after MUA due to persistent motion restrictions were excluded from the study. The following preoperative patient characteristics were evaluated: age, sex, body mass index, duration from symptom onset to surgery, history of trauma (defined as a clear episode such as a fall or direct blow), presence of diabetes, smoking status, and follow-up period. Intraoperative characteristics included mediolateral and anteroposterior tear size measured with an arthroscopic ruler, tear classification according to DeOrio and Cofield [16], management of the long head of the biceps tendon (tenotomy or tenodesis), and repair technique (single row, double row, or suture bridge). The stiffness group included the patients who underwent MUA before ARCR, regardless of their preoperative ROM. All other patients were classified into the non-stiffness group.
Surgical Technique
Study participants were placed in a beach chair position under general anesthesia with an interscalene block. ARCR was performed by two senior surgeons (KF and NS) using the same technique. Before surgery, the passive ROM was assessed in each patient. MUA was performed before arthroscopy if the examiner experienced stiff resistance during forward flexion or abduction at any angle in the following standard sequence: application of flexion, abduction, external rotation, internal rotation, external and internal rotation at 90° of abduction, and cross-body adduction. The examiner observed a sudden marked improvement in most patients after experiencing crepitus during gentle forward flexion or abduction.
During glenohumeral arthroscopy, we observed that the anteroinferior part of the capsule was divided in patients as a result of MUA in almost all cases. Tenotomy or tenodesis was performed in cases with a biceps long head lesion. Thereafter, the scope was moved to the subacromial space, and subacromial bursectomy was routinely performed. After tendon edge debridement, tear size in the mediolateral and anteroposterior directions was measured with an arthroscopic ruler. Most rotator cuffs were repaired using a double-row technique until 2014, after which the suture-bridge technique became the primary method of repair. However, if the tension on the tendon remained high after sequential release [17], a single-row repair was performed by fixing the tendon at the medial aspect of the footprint.
Postoperative Protocol
All patients were prescribed a standard rehabilitation program after ARCR; each wore an abduction brace for 6 weeks, with passive motion exercise introduced at 1 week by a physical therapist. Thereafter, active motion was initiated at 6 weeks, and rotator cuff strengthening exercises began 12 weeks after surgery. Full return to sports or heavy labor was allowed after 6 months.
Clinical Evaluation
Clinical outcome measures included passive ROM preoperatively and at 3, 6, 12, and 24 months post-surgery, along with Japanese Orthopedic Association (JOA) and University of California, Los Angeles (UCLA), scores assessed preoperatively and at 6, 12, and 24 months postoperatively. These clinical outcomes were evaluated in the outpatient clinic by the senior author (YT), who was not the operating surgeon but who was aware of the patients’ clinical information. The shoulder ROMs of flexion, abduction, and external rotation were recorded in degrees, while internal rotation was graded based on the part of the spine the thumb could reach. For the statistical analysis of internal rotation, we converted the values into contiguously numbered groups (T1–T12 as 1–12, L1–L5 as 13–17, the sacrum as 18, and the buttocks as 19), as described by Oh et al. [7]. Postoperative rotator cuff integrity was evaluated using MRI 6 months after surgery. According to the Sugaya classification [18], tendon types 4 and 5 are considered torn. Preoperative and postoperative MRI evaluations were performed by an orthopedic surgeon (YK) who was blinded to the patient information.
Statistical Analyses
Preoperative and intraoperative factors were compared between the stiffness and non-stiffness groups. Student t-test or the Mann-Whitney U-test was used for continuous variables after checking their normality using the Shapiro-Wilk test. The chi-square test or Fisher’s exact test was used for categorical variables. The ROM and clinical scores between the two groups at each follow-up period were compared using Student t-test or the Mann-Whitney U-test. The retear rates of the two groups were compared using Fisher’s exact test. The retear rate was first compared overall, then further analyzed by dividing the cases into medium-sized tears and large and massive tears. One-way repeated-measures analysis of variance was performed to compare the serial changes in ROM and clinical scores from baseline to the final follow-up period using the Bonferroni post-hoc test. The significance level was set at P<0.05. Statistical analyses were performed using the Mac statistical analysis program, version 3.0 (ESUMI Co., Ltd.).
RESULTS
Of the 485 selected patients, 163 were excluded based on the exclusion criteria. The remaining 322 patients (322 shoulders) met our inclusion criteria and were divided into the stiffness group (n = 88) and the non-stiffness group (n = 234) (Fig. 1). The demographic data are shown in Table 1. The prevalence of both trauma history and diabetes was significantly higher in the stiffness group than in the non-stiffness group (P=0.008 and P=0.015, respectively). The tear size in the anteroposterior direction was also significantly larger in the stiffness group (P=0.002).
Preoperatively, forward flexion, abduction, and external and internal rotation in the stiffness group were significantly restricted compared to those in the non-stiffness group (P<0.001) (Fig. 2). Both groups showed significant improvements in ROM, except for external rotation in the non-stiffness group after ARCR. However, the ROM-improvement rate from baseline to the final follow-up was significantly higher in the stiffness group than in the non-stiffness group (flexion and abduction: P<0.001; external rotation: P=0.015; internal rotation: P=0.007) (Table 2). Although external rotation in the stiffness group remained lower than that in the non-stiffness group (P=0.005) at the final follow-up, flexion, abduction, and internal rotation were not significantly different between the two groups (Fig. 2). The preoperative JOA and UCLA scores of the stiffness group were significantly lower than those of the non-stiffness group (P<0.001) (Fig. 3). Both groups showed significant improvements in scores 6 months after surgery, and no significant differences were found in the scores 24 months after surgery.
Retears were found in nine of the 88 shoulders in the stiffness group (10.2%) and 37 of the 234 shoulders in the non-stiffness group (15.8%) (Table 3). There was no significant difference between the two groups (P=0.202). Among large and massive tears, no significant difference was observed between the stiffness (23.5%) and non-stiffness (28.4%) groups (P=0.604). However, among medium-sized tears, the retear rate was significantly lower in the stiffness group (1.9%) than in the non-stiffness group (10.8%) (P=0.042). No patients experienced complications related to MUA, including instability, fractures, or brachial plexus injuries.
DISCUSSION
This study demonstrated that preoperative shoulder stiffness treated with MUA alone significantly improved the ROM and clinical outcome scores in ARCR. At the final follow-up, the clinical outcome scores and ROMs, except for external rotation, were not significantly different between the stiffness and non-stiffness groups. Among the medium-sized tears, the retear rate was significantly lower in the stiffness group than in the non-stiffness group.
This study aimed to clarify the effectiveness of MUA without arthroscopic capsular release in patients with stiff shoulders and rotator cuff tears. We did not routinely perform arthroscopic capsular release for stiff shoulders associated with rotator cuff tears because prior studies suggest that their pathophysiology differs from that of primary frozen shoulders, including capsular fibrosis, inflammation, and chondrogenesis [19]. Conversely, the stiffness of rotator cuff tears may be primarily caused by several factors, such as adhesions within the extra-articular glenohumeral interface [1,4] or secondary muscular and capsular contractures caused by disuse due to pain and weakness. Therefore, the type and severity of ROM deficits in stiff shoulders with rotator cuff tears differ from those in frozen shoulders. Iwamoto et al. [20] reported that two-thirds of shoulders showed ROM deficits in only one or two directions in 138 shoulders with rotator cuff tears. Ueda et al. [14] reported that severe and global loss of passive motion was not found in shoulders with full-thickness rotator cuff tears in 379 patients with stiff shoulders. These studies suggested that global capsular release is not always necessary in patients with stiff shoulders and rotator cuff tears.
Although Cho and Rhee [8] reported that the anterosuperior capsule could rupture with crepitus during MUA, our experience aligns with that of Oh et al. [7], who found that MUA typically releases the inferior capsule but rarely the anterior capsule. Their findings are consistent with those of this study. In our experience, the differences in flexion and abduction between the stiffness and non-stiffness groups were insignificant at the final follow-up, while external rotation remained limited in the stiffness group. These results suggest that global capsular release is not always necessary for shoulder stiffness associated with rotator cuff tears. If necessary, only anterior capsular release would be enough for most stiff shoulders [21].
Both groups showed significant improvements in clinical scores and ROM, except for external rotation in the non-stiffness group. The degree of improvement was greater in the stiffness group. According to Xu et al.’s classification of patient satisfaction based on the change in the UCLA score after ARCR [22], the mean change in the UCLA score at 24 months in the stiffness group (17.9 points) indicated excellent satisfaction, while that in the non-stiffness group (13.4 points) corresponded to a level between very good and good. These findings suggest that patients with preoperative stiffness may achieve greater postoperative satisfaction following ARCR.
Whether shoulder stiffness is positively associated with better tendon healing after ARCR remains controversial. Several studies [7,23] reported that postoperative rotator cuff integrity showed no significant difference between the stiffness and non-stiffness groups after 1 year. In contrast, McGrath et al. [15] reported a retear rate of 0% in the stiffness group, which was significantly lower than that of 25% in the non-stiffness group. Similarly, Kim et al. [1] and Jeong et al. [2] reported significantly lower retear rates in the stiffness group (2.6% and 5.3%, respectively) compared to the non-stiffness group (14.7% and 12.3%, respectively). The present study found no significant difference in rates of large and massive tears between the stiffness and non-stiffness groups (23.5% vs. 28.4%, respectively). However, the retear rate was significantly lower in the stiffness group (1.9%) than in the non-stiffness group (10.8%) when considering medium tears.
Previous studies have proposed two explanations for why stiffness is associated with a lower retear rate. One suggestion is that shoulder immobility in the early postoperative period provides mechanical protection for the repaired tendon [2,24,25]. The other explanation attributes it to a strong biological healing response associated with stiffness [15,25-27]. Ko et al. [28] reported that patients with shoulder stiffness showed increased interleukin-1β expression and greater myofibroblast recruitment in the subacromial bursa compared to those without stiffness. It is conceivable that these factors related to stiffness would be canceled out by the other influential factors, such as repair tension [16] or fatty degeneration [29], in large and massive tears.
This study had some limitations. First, stiffness was defined by the need for MUA rather than preoperative ROM, limiting comparability with previous studies and introducing potential selection bias. However, no consensus currently exists on a standardized ROM threshold [30,31]. Second, we could not compare MUA alone with combined arthroscopic capsular release due to the small number of relevant cases available. Third, we did not compare the outcomes of patients with preoperative stiffness between those who underwent MUA and those who did not, so natural recovery cannot be excluded. Fourth, rotator cuff repair was performed using three techniques, although their distribution did not differ significantly between the two groups. Finally, retears were assessed by MRI at 6 months, which would not capture those occurring later.
CONCLUSIONS
Patients with preoperative stiff shoulders who underwent ARCR combined with MUA alone showed comparable clinical outcome scores and ROMs to those without preoperative stiffness. Furthermore, preoperative stiffness was significantly associated with better tendon healing in medium-sized tears.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
