Introduction

Suprascapular neuropathy is a relatively uncommon condition which manifests clinically as shoulder pain and disability including weakness of forward flexion, abduction, and external rotation [2, 40, 52]. Dysfunction of the suprascapular nerve can occur as the result of a compression or traction mechanism. The aetiology of this condition is variable and includes massive rotator cuff tears [32], rotator cuff tears associated with muscle atrophy and/or fatty infiltration [54], labral tears with associated para-labral cysts [13, 35, 53], mass lesions impinging the suprascapular and/or spinoglenoid notches [19], and repetitive overhead activities [25]. Overall, suprascapular neuropathy is reported to be the cause of 1–2% of all shoulder pain, with a prevalence of up to 33% in overhead athletes [7, 60].

While activity modification, non-steroidal anti-inflammatory drugs, and physical therapy account for the major non-operative treatment modalities [33], surgical options aim to correct the cause of compression. These strategies range from aspiration and decompression of ganglion cysts, release of transverse scapular ligament at the suprascapular notch and/or spinoglenoid ligament at the spinoglenoid notch, repairing labral tears, and repairing rotator cuff tears. While historically many of these procedures had been performed via open procedures, numerous techniques have been described to approach the decompression of the suprascapular nerve arthroscopically [7].

Shoulder arthroscopy as a minimally invasive procedure has generally become increasingly popular as it lends many benefits over open surgery, including decreased risk of infection, less violation of soft tissue, improved post-operative pain, and shorter hospital stays [4]. In its application for suprascapular neuropathy, individual reports have reported positive outcomes [23, 37]. However, there has been no systematic review that has summarized the existing clinical literature on this topic. Therefore, the purpose of this study was to systematically assess the arthroscopic management of suprascapular neuropathy, including the aetiology, surgical decision-making, clinical outcomes, and complications associated with the procedure. It was hypothesized that shoulder arthroscopy provides patients with suprascapular neuropathy significant improvement in pain and function of the shoulder with relatively minimal complications.

Materials and methods

Search strategy

The methodology used in the present study is similar to that used in other systematic reviews conducted at the authors’ institution [22]. The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) statement was used for the reporting of study selection [34]. The online databases PubMed, EMBASE, and Ovid (Medline) were used to search for the literature addressing the use of shoulder arthroscopy for suprascapular neuropathy from database inception until 17 February 2017. The search terms “shoulder”, “arthroscop*”, and “suprascapular” were used (Appendix Table 4). No attempt was made to evaluate the grey literature since this evidence is not subject to the same transparency requirements and scrutiny as studies in peer-reviewed journals.

Study screening

Two reviewers independently screened the titles, abstracts, and full-text articles. Any disagreements were discussed between reviewers and a senior author (ORA) to determine study inclusion when necessary. The references of the eligible, included studies were then screened to include any additional articles that may not have been captured by the initial search strategy. The search strategy is outlined in Fig. 1.

Fig. 1
figure 1

PRISMA flow diagram demonstrating the systematic review of the literature for arthroscopic management of suprascapular neuropathy

Full size image

Assessment of study eligibility

The research questions and inclusion and exclusion criteria were determined a priori. The inclusion criteria included English-language studies, human studies, and studies investigating arthroscopic management of suprascapular neuropathy. Studies of all levels of evidence and all study designs that reported any outcomes including pain, range of motion, complications, and outcome scores were included. The exclusion criteria included cadaveric studies, conference papers, book chapters, review articles, and technical reports with no outcome data.

Data abstraction

Two reviewers collected data in duplicate and recorded them in a Microsoft Excel 2007 spreadsheet (version 2007, Microsoft Inc, Redmond, Washington, USA). Data regarding study year of publication, author, study design, sample size, per cent female patients, age, follow-up, aetiology, history, physical examination, diagnostic investigations, conservative management, surgical decision-making, arthroscopic management, outcomes, and complications were abstracted from all included studies. Two reviewers independently evaluated the abstracts and assigned a level of evidence (levels I–IV) to each included study using the AAOS classification scheme [57].

Quality assessment

The two reviewers independently assessed the level of evidence (level I–level IV) of the included studies using the American Academy of Orthopedic Surgeons (AAOS) classification system [57]. The MINORS (Methodological Index for Nonrandomized Studies) checklist was used as well in order to assess the methodological quality of the included studies [49]. The MINORS checklist grades comparative studies out of a maximum score of 24 and non-comparative studies out of 16. The senior author was consulted for any discrepancy among the reviewers. All eligible studies were included in this review regardless of level of evidence and study quality.

Assessment of agreement

Inter-reviewer agreement for the title, abstract, and full-text articles was calculated using a kappa (κ) statistic. An intraclass correlation coefficient (ICC) was calculated for the quality assessment using the MINORS criteria. Agreement was categorized a priori as follows: κ/ICC of 0.61 or greater was considered substantial agreement; κ/ICC of 0.21–0.60, moderate agreement; and κ/ICC of 0.20 or less, slight agreement [26].

Statistical analysis

The primary outcome assessed was pain (assessed via VAS and NRS scores), particularly the change in pain from the pre-operative level to the final follow-up post-operatively. Secondary outcomes including paraesthesia, nerve, and muscle function were recorded as well. Due to the limited reporting, these data were not combined in a meta-analysis and are summarized descriptively. Descriptive statistics calculated from the data included means, proportions, standard deviations, and ranges. All statistics were calculated using Minitab® statistical software (version 17, Minitab Inc., State College, Pennsylvania, USA).

Results

Search strategy

The initial search of three databases resulted in 4759 total studies. A total of 1896 studies were immediately removed as duplicates resulting in 2863 studies. A systematic screening approach removed articles failing to meet inclusion criteria and resulted in 40 available full-text articles for review (Fig. 1). There was almost perfect agreement among reviewers at the title (κ = 0.863; 95% CI 0.815–0.911), abstract (κ = 0.857; 95% CI 0.724–0.990), and full-text (κ = 1.00; 95% CI 1.00–1.00) screening stages.

Study characteristics

In total, 259 patients (261 shoulders) with a mean age of 47.0 years (range 16–88) underwent arthroscopic decompression for suprascapular neuropathy. Of the included patients, 37.8% were female and they were followed up for a mean of 23.7 months (range 3–92) (Table 1). None of the included studies performed an a priori power analysis for sample size calculation.

Table 1 Characteristics of included studies and patients
Full size table

Study quality

Of the 40 included studies, 17 case reports and 20 case series were identified, which were level IV evidence, 2 retrospective comparative studies were identified, which were level III evidence, and 1 prospective comparative study was identified, which was level II evidence. For non-comparative studies, the median MINORS score was 9 out of 16 (range 7–12), whereas the median MINORS score for comparative studies was 14 out of 24 (range 12–15). Overall, 100% of studies had a clearly stated aim, 92.5% had appropriate endpoints, 90% had an appropriate follow-up period, and 95% had loss of follow-up less than 5%. However, only 7.5% of studies had prospective collection of data and no studies had unbiased assessment of study endpoints. The overall inter-rater agreement for the MINORS score was high, with an ICC of 0.86 (95% CI 0.78–0.94).

Aetiology and surgical management decision-making

Suprascapular nerve (SN) compression was due to various aetiologies including a cyst at the spinoglenoid notch (110 shoulders, 42.1%), the transverse scapular ligament at the suprascapular notch (54 shoulders, 20.7%), a massive rotator cuff tear (50 shoulders, 19.2%), the spinoglenoid ligament at the spinoglenoid notch (16 shoulders, 6.1%), a cyst at the suprascapular notch (10 shoulders, 3.8%), a bony, stenotic suprascapular foramen (3 shoulders, 1.1%), branches of the suprascapular artery (1 shoulder, 0.4%), an intraosseous glenoid cyst (1 shoulder, 0.4%), and unreported aetiologies (46 shoulders, 17.6%). In 32 (12.3%) shoulders, the aetiology was a combination of the aforementioned pathologies. With respect to surgical decision-making, in 18 studies (45%), the decision to pursue arthroscopic management was made after failure of non-operative management strategies, in 19 studies (47.5%) the decision was based on clinical findings and investigations only, and in 3 studies (7.5%) arthroscopic management was pursued after failed ultrasound-guided aspiration of ganglion cysts (Tables 2 and 3). Diagnostic imaging modalities are summarized in Table 2, and non-operative management modalities and arthroscopic techniques are summarized in Table 3.

Table 2 Aetiology, clinical presentation, and investigations for suprascapular neuropathy
Full size table
Table 3 Management of suprascapular neuropathy and outcomes following arthroscopic surgery
Full size table

Outcomes

Of the 259 patients from the 17 case reports, 20 case series, and 3 comparative studies, 248 patients (95.8%) reported complete resolution or improvement in pre-operative symptoms. Specifically, of the 186 patients for whom pain outcomes were reported, 82 patients experienced complete resolution of pain, 94 patients experienced improvement in pain symptoms, and 10 patients experienced no pain relief. The visual analogue scale (VAS) was used to measure pain in 63 patients, and the results demonstrated a reduction in the VAS score from 7.3 pre-operatively to 1.6 post-operatively. Of the 84 patients for whom strength outcomes were reported, 28 patients regained complete shoulder strength, 54 patients achieved improvement in shoulder strength, and 2 patients experienced no improvement in pre-operative weakness, chiefly with regard to external rotation strength. Of the 32 patients for whom muscle atrophy outcomes were reported, 17 patients obtained complete resolution of muscle atrophy, while 15 patients experienced improvement in muscle atrophy, including the supraspinatus and infraspinatus muscles. Notably, assessment of muscle atrophy was based on gross inspection in 7 studies and MRI findings in 1 study. Of the 32 patients for whom range of motion outcomes were reported, 18 patients achieved full shoulder range of motion, while 14 patients obtained improvement in shoulder range of motion. Specifically, the methods regarding range of motion assessment were not reported in the included studies. Of the 51 patients for whom post-operative EMG results were reported, 25 patients’ EMGs demonstrated resolution of suprascapular neuropathy, while 26 patients’ EMGs demonstrated improvement in suprascapular neuropathy. Post-operative EMGs were completed at a mean time of 6.9 months (range 3–48) after the surgical procedure. Of the 74 patients for whom post-operative MRI results were reported, 70 patients’ MRIs demonstrated complete resolution of cysts, while 4 patients’ MRIs demonstrated persistent cysts. Of the 2 patients with post-operative ultrasound imaging, both patients demonstrated complete resolution of cysts on ultrasound. Of the 76 patients for whom return to activity outcomes were reported, 37 patients returned to daily activities and work by an average time of 12.3 months (range 3–22) post-operatively, while 39 patients returned to sports by an average time of 9.4 months (range 3–28) post-operatively. Moreover, of the 93 patients for whom the Constant score was reported, the mean Constant score improved from 61.4 pre-operatively to 91.4 post-operatively. Of the 28 patients for whom a subjective shoulder value (SSV) was reported, the mean SSV improved from 37.7 pre-operatively to 67.7 post-operatively. Of the 24 patients for whom an American Shoulder and Elbow Surgeons (ASES) score was reported, the mean ASES score improved from 36.5 pre-operatively to 70.6 post-operatively. Of the 22 patients for whom a simple shoulder test (SST) was reported, the mean SST improved from 5.5 pre-operatively to 11.3 post-operatively. Of the 57 patients for whom a University of California Los Angeles Shoulder (UCLA) score was reported, the mean UCLA score improved from 11.0 pre-operatively to 28.5 post-operatively. Of the 4 patients for whom a quick disabilities of the arm, shoulder and hand (qDASH) score was reported, the mean qDASH score improved from 22.7 pre-operatively to 1.1 post-operatively. Of the 9 patients for whom a Short Form Health Survey (SF-36) score was reported, the mean SF-36 score was 122.9 post-operatively. Finally, of the 28 patients for whom a Rowe score was reported, the mean Rowe score improved from 52.8 pre-operatively to 94.7 post-operatively. Collectively, these results were reported at a mean final follow-up period of 23.7 months (range 2–92) (Table 3).

Complications

Of all 259 patients in this review, one patient (0.4%) had a superficial soft tissue infection, which was successfully treated with oral antibiotics. Another patient (0.4%, 1/259) had post-operative pain along the direction of the long head of the biceps due to a partial rupture, which resolved after performing a subpectoral biceps tenodesis. One patient (0.4%, 1/259) experienced loss of sensation around the skin incision, which resolved at the final follow-up visit. Additionally, one patient (0.4%, 1/259) developed adhesive capsulitis of the shoulder post-operatively. Finally, of the 50 patients who underwent rotator cuff repair, there were 6 (12.0%, 6/50) rotator cuff tears that failed to heal, and 5 (10.0%, 5/50) of these patients were dissatisfied with their functional outcome (Table 3).

Discussion

The most important finding of the present study was that 96% of the included patients reported significant improvement in or complete resolution of their pre-operative symptoms. Specifically, there was significant improvement in pain (measures using NRS and VAS scores), shoulder strength, and subjective scoring scales (including the Constant, SSV, ASES, SST, and UCLA shoulder scales). Moreover, there was a low overall complication rate associated with the arthroscopic procedures. With regard to surgical decision-making, in approximately half of the included studies, the decision to pursue arthroscopic management was made after failure of non-operative management strategies, whereas the decision was based on clinical findings and investigations only in the other half of studies. The included studies were primarily of lower levels of evidence with only 3 comparative studies identified in the literature.

There are several sources of debate across the literature investigating arthroscopic surgery for the management of suprascapular neuropathy, including surgical timing. Specifically, the present systematic review included multiple studies reporting that shortened time from symptom onset to arthroscopic surgery correlated with improved outcomes. Specifically, Bilsel et al. [6] and Pillai et al. [39] attribute the delay to arthroscopic management resulting in greater muscle weakness pre-operatively, which resulted in poorer VAS scores and less improvements in ER strength. Shah et al. reported that patients receiving surgery within a year of symptom onset were significantly more likely to experience good and excellent outcomes, which allows surgeons to more accurately predict prognosis in patients planning on undergoing arthroscopic management. On the other hand, the results reported by Kim et al. [23] disagreed with the aforementioned studies, as they found no correlation between the duration of pre-operative symptoms and post-operative resolution of muscle atrophy. These findings warrant future research into the effect of surgical timing on patient outcomes.

Another important concept relates to outcomes following labrum repair or debridement performed with or without cyst decompression. Specifically, a comparative study by Kim et al. [23] demonstrated no difference in outcomes in patients who received a SLAP repair versus those who received a SLAP repair in combination with cyst decompression. They attributed these results to a one-way valve mechanism created by repetitive use or trauma, causing the labrum to avulse off the glenoid rim, allowing joint fluid to drain through the lesion, and forming a cyst [31]. As such, they described that arthroscopic cyst resection cannot ensure elimination or non-recurrence of the cyst and that the technical challenges of arthroscopic cyst resection may preclude one from routinely performing this procedure. However, a comparative study by Piatt et al. [38] demonstrated that patients receiving a labral repair in combination with cyst decompression achieved better outcomes than those who received isolated labral treatment. It was theorized by Hashiguchi et al. [17] that patients without a formal cyst decompression might experience poor outcomes due to a delay in achieving cyst resolution, and therefore, persistent compression by the cyst in the post-operative period, contributing to pain and additional symptoms.

Another method of suprascapular nerve decompression, first described by Costouros et al. [12] in 2007, involves isolated rotator cuff repair. The pathology they described included suprascapular nerve traction and compression around the base of the spine of the scapula as a result of the inferior and medial retraction of the supraspinatus and infraspinatus tendons in rotator cuff tears. After performing a rotator cuff repair, the suprascapular nerve moved laterally and away from the scapular spine, relieving any tension on the nerve. Additionally, in 2016, Savoie et al. [45] compared patients receiving an isolated rotator cuff repair to those receiving an additional suprascapular nerve decompression procedure, and they found superior outcomes in the latter group. In an extension to the description by Costouros et al. [12], they explained that over time, as scar tissue develops at the site of the rotator cuff tear, the laterally retracted nerve is held out of its anatomical location in part by local adhesions that form. Therefore, surgical release of these adhesions likely allows the nerve to achieve an even more accurate anatomical position, accounting for the improved recovery within this group.

The primary limitations in this review included the lack of high-quality studies investigating arthroscopic management for suprascapular neuropathy. The included studies were mostly case reports or case series, with limited sample sizes and retrospective collection of data. The paucity of comparative studies precludes definitive conclusions on the efficacy of arthroscopic management of suprascapular neuropathy in comparison to other procedures. Further, the included non-randomized and non-comparative studies may be prone to selection bias in the patients that are recruited to be participants in the study. Moreover, the reporting of certain data within the studies was not always consistent, including information regarding clinical presentation, conservative management regimens, arthroscopic technique, and clinical outcome scores. Finally, the overwhelming positive results found in this review to be associated with arthroscopic management of suprascapular neuropathy may be due to expertise bias and consequently have limited generalizability to a broader surgeon audience. Specifically, since arthroscopic decompression of the suprascapular nerve is a technically challenging procedure with multiple potential risks, including neurovascular injury, the excellent results achieved within the included studies may be influenced by expert surgeons who perform the operation and publish on the subject more frequently.

The clinical relevance of this study is that patients with suprascapular neuropathy who are managed with arthroscopic surgery may expect significant improvements in pain and functional outcomes. Additionally, the likelihood of experiencing a complication as a result of surgery is low.

Conclusions

While most studies evaluating arthroscopic management of suprascapular neuropathy are uncontrolled studies with lower levels of evidence, results indicate that such management provides patients with significant improvements in pain, strength, and subjective function of the shoulder, and has a low incidence of complications.