Factors associated with pain in nonsurgically treated rotator cuff tears -A study with magnetic resonance imaging

A rotator cuff tear is a common cause of pain and disability of the shoulder among adults. Some rotator cuff tears are symptomatic, whereas others are asymptomatic [1, 2]. Treatment strategy for rotator cuff tears remains unclear [3]. Because nonsurgical treatment often improves the pain and range of motion in patients with symptomatic rotator cuff tears [4], it is typically preferred as the first-line treatment [5, 6]. Conversely, surgical treatment is occasionally required when symptoms do not improve. Success rates of nonsurgical treatment have been reported to vary widely, ranging 33–82% [6‐10]. Itoi et al. reported that patients with complete rotator cuff tears, and who had been treated conservatively, exhibited unsatisfactory results with limited active abduction angle and muscle weakness on abduction on first examination [8]. Tanaka et al. reported that an intact intramuscular tendon of the supraspinatus (SSP), showing little or no atrophy of the SSP muscle, with negative impingement signs, and with preserved motion in external rotation are factors related to the successful outcome of conservative treatments [6]. Boorman et al. reported that the rotator cuff quality-of-life index was predictive of the outcome of non-operative treatment in patients with a chronic full-thickness rotator cuff tear [11]. Dunn et al. reported that patient expectations regarding the role of rehabilitation were the strongest predictor of surgery and other factors associated with surgery were higher activity level and not smoking [12]. However, a failure of nonsurgical treatment was defined if the patient elected to have surgery after nonsurgical treatment in most previous studies. There may be a potential selection bias because some patients do not want to undergo surgery whereas others want to undergo surgery and failure of nonsurgical treatment is determined relatively early in previous studies. It is unclear about the difference between cases that become asymptomatic with nonsurgical treatment and cases that nonsurgical treatment is continued as symptomatic. Although most rotator cuff tears occur in the SSP tendon, other tendons of the rotator cuff may also be torn. In addition, the degree of fatty infiltration of rotator cuff muscles varies. The effect of nonsurgical treatments may depend on the severity of the rotator cuff tears. Therefore, the purpose of this study was to identify the factors associated with persisted pain in nonsurgically treated rotator cuff tears using magnetic resonance imaging (MRI).

There are several reports of predictors of outcome of nonsurgical treatment for rotator cuff tears [6, 8, 11, 12]. However, nonsurgical treatment outcome was defined as a success if surgical treatment was not required in most previous studies. Strengths of our study include re-assessing via MRI patients who did not have surgery with more than a year of follow-up, and we compared patients that became asymptomatic by nonsurgical treatment with patients that pain persisted and nonsurgical treatment is continued.

There are a few reports comparing MRI findings in symptomatic and asymptomatic rotator cuff tears. Some reports showed that there was no difference [15, 16], others showed that there were significant associations between symptoms and tear size, positive tangent sign, and fatty degeneration [17]. It was often reported that symptoms of pain do not correlate with rotator cuff tear severity [18‐20]. However, there are currently no reports to examine SSC tendon tears in previous studies, apart from a study conducted by Harris et al. who reported that involvement of the SSP, ISP, and SSC tendons was associated with a decreased American Shoulder and Elbow Surgeons score comparing isolated SSP tears in patients with symptomatic and traumatic full-thickness rotator cuff tears who are enrolled in a structured physical therapy program [19]. In our study, we evaluated factors associated with pain in nonsurgically treated SSP tendon tears by MRI. Bursal fluid accumulation, SSP tendon retraction, SSC tendon tears, ISP tendon tears, and Goutallier classification in SSC, SSP, and ISP were included as evaluation factors in the multivariate logistic regression analysis. Consequentially, bursal fluid decreased when pain improved. This suggested the subcoracoid–subacromial–subdeltoid bursal fluid was associated with the presence of pain. However, bursal fluid could not be a predictive factor for persistent pain in the initial MRI. On the other hand, rotator cuff tears with the involvement of SSC tendon tears on initial MRI were considered as predictive factors for persistent pain.

The balance between the SSC muscle and the ISP and teres minor muscles is often referred to as the rotator cuff “transversal force couple.” It generates stability in the glenohumeral joint by an antagonistic balance between the forces exerted in an anterior and posterior part [21]. This anterior–posterior force balance between the SSC and the ISP, and teres minor causes compression of the humeral head into the glenoid fossa in patients with isolated SSP tendon tears [22‐24]. Even when there are irreparable rotator cuff tears, shoulder function can be preserved if there is a balance between important force couples [25, 26]. Therefore, damage of transversal force couples is suggested as the cause of the correlation between SSC tendon tears and persistent pain. On the other hand, ISP tendon tears were not a factor associated with persistent pain in this study. This may be due to compensation by teres minor muscles. It has been reported that progression of ISP muscle atrophy induces compensatory hypertrophy of the teres minor muscles [27, 28].

The SSC has other crucial functions in the shoulder joint. The SSC stabilizes the long head of the biceps tendon. It was reported that most SSC tendon tears occurred on the most superior part and the articular side of the attachment of the humerus [29, 30]. Ide et al. reported that the SSC insertion to the humerus consisted of a proximal tendinous part and a distal muscular part in anatomical study [31]. Arai et al. reported that anatomically, the most superior part of the SSC tendon was attached to the upper margin of the lesser tuberosity and extended as a thin tendinous slip to the fovea capitis of the humerus [32]. The trochlea-like structure was composed of the superior-most insertion, the tendinous slip, and the lateral portion of the cranial part of intramuscular tendons supporting the long head of the biceps tendon. SSC tendon tears, involving injuries of the trochlea-like structure, causes instability and pathology of the long head of the biceps tendon [33, 34]. SSC tendon tears may induce lesion of the long head of biceps tendon and cause pain.

For the reasons stated above, in this study, SSC tendon tears may be associated with persistent pain for nonsurgically treated rotator cuff tears. Arthroscopic repair with use of a suture anchor technique has been reported to be a safe and effective procedure for the treatment of combined rotator cuff tears involving the SSC tendon [35]. It has also been demonstrated that improvement in functional outcome after arthroscopic repair of an SSC tendon tear is maintained long-term [36]. Therefore, if no effect is seen from nonsurgical treatment for rotator cuff tears with the involvement of SSC tendons, surgical repair can be considered.

There were several limitations to this study. First, the number of cases included is small. Second, nonsurgical treatment was not standardized as the frequency of injection varied. Third, there is no scale for pain evaluation. Conditions requiring continued nonsurgical treatments such as steroids, hyaluronic acid injections, and/or anti-inflammatory analgesics were defined as persistent pain. Fourth, interpretations of MRI are not always accurate [37, 38]. However, Adam et al. state that preoperative shoulder MRI scans interpreted by orthopedic surgeons via a systematic approach resulted in improved accuracy in diagnosing SSC tendon tears [39]. Yoshimura et al. focused on the anatomical structure of SSC and evaluated MRI findings of the most proximal portion of the SSC tendon on axial and oblique sagittal T2-weighted images [40]. Based on their report, SSC tendon tears were evaluated on MRI in this study.

We evaluated the factors associated with pain in nonsurgically treated rotator cuff tears by MRI. Bursal fluid accumulation decreased when pain improved, but the bursal fluid is not a reliable predictive factor for persistent pain in initial MRI. On the other hand, SSC tendon tears on initial and repeat MRI were significantly associated with persistent pain. Therefore, the involvement of SSC tendon tears may serve as a predictive factor in persistent pain. SSC function may be important since SSC contributes to “transversal force couple” and stabilizes the long head of the biceps tendon. When patients with rotator cuff tears including SSC tendon tears are treated nonsurgically, attention to the possibility of persistent pain should be given.