The impact of supraspinatus tear on subscapularis muscle atrophy and fatty infiltration

Patients

A total of 417 patients underwent arthroscopic RC repair from January 2020 to March 2022. Patients who received posterosuperior RC repair with no subscapularis tear on arthroscopy were included (n=189). The exclusion criteria were as follows: no preoperative MRI (n=0), large (greater than 3 cm, n=15) or massive (greater than 5 cm, n=2) supraspinatus tear [14], history of previous shoulder surgery (n=11), history of prior shoulder fracture (n=4), and inflammatory arthritis (n=4). Ultimately, 153 patients met the criteria and were analyzed in the study (Fig. 2). Patient demographics, including age, sex, dominant arm, and occupation as a heavy laborer, were reviewed using electronic charts.

Fig. 2.

Flow diagram of patient inclusion.

Radiologic Evaluations

Preoperative T2-weighted spin-echo MRI (3.0-T scanner, Gyroscan Intera Achieva; Philips Medical Systems) and PACS software (GE Centricity) were used for MRI evaluations. Oblique coronal and sagittal views were used to determine the type of supraspinatus tear (partial- or full-thickness tear). The oblique coronal and sagittal views were utilized for measuring the mediolateral and anteroposterior tear size of the supraspinatus/infraspinatus. In addition, supraspinatus/infraspinatus medial retraction was measured on the coronal oblique view based on a modification of Patte’s classification [15]. This assessment divides the distance from the greater tuberosity to the upper margin of the glenoid into 6 zones, resulting in a relative scoring scale from 0 to 6. For instance, a score of 6 indicates retraction up to the glenoid margin, while a score of 3 represents retraction over the upper tip of the humeral head, halfway between the greater tuberosity and the glenoid (Fig. 3) [15]. These scores were treated as continuous variables in our analysis.

Fig. 3.

Modified Patte’s classification for rotator cuff retraction. (A) The distance from the greater tuberosity to the upper margin of the glenoid is divided into six segments, with scores ranging from 0 to 6. This modification offers a more detailed assessment of supraspinatus tendon retraction relative to the original Patte’s classification, independent of patient size. (B) A retraction of 1/6 of the supraspinatus tendon just medial to the greater tuberosity footprint (asterisk) is assigned a score of 1. (C) A retraction of half the supraspinatus tendon, reaching the upper tip of the humeral head (asterisk), is assigned a score of 3.

The Y-view on sagittal oblique imaging was employed to assess the MAFI of RC muscles using the Goutallier classification [16]. Given the insufficient detail in previous studies regarding the evaluation of subscapularis Goutallier grade, this study provides a defined grading system based on Yoon et al. [17]: grade 1, a fatty streak observed between the subscapularis tendons; grade 2, less fat than muscle in the upper half of the muscle, presenting a cog-shaped appearance in the upper subscapularis muscle; grade 3, more fat than muscle in the upper half of the muscle, with normal or fatty streaks in the lower half; and grade 4, more fat than muscle throughout the upper half (Fig. 4). Goutallier grade was treated as a continuous variable during multivariable analysis. Additionally, the supraspinatus occupation ratio was calculated on the Y-view [18].

Fig. 4.

Goutallier classification grading. The corresponding areas for each grade are indicated by arrows: (A) Grade 1, a fatty streak observed between the subscapularis tendons. (B) Grade 2, less fat than muscle in the upper half of the muscle, presenting a cog-shaped appearance in the upper subscapularis muscle. (C) Grade 3, more fat than muscle in the upper half of the muscle, with normal or fatty streaks in the lower half. (D) Grade 4, more fat than muscle throughout the muscle.

A postoperative 6-month MRI was performed to evaluate the integrity of the repaired supraspinatus tendon using the Sugaya classification, and types 4 and 5 were considered retears [19]. Two junior shoulder surgeons (SCK and JHJ), blinded to the study populations, conducted all radiologic evaluations. Intra-observer and inter-observer reliability values were calculated. Mean values for continuous variables were used, and agreement for ordinal variables was achieved when disagreement arose between the surgeons.

Arthroscopic Findings and Surgery

Pathology of intra-articular subscapularis tendon insertion was evaluated using a probe. Posterosuperior RC tear was evaluated from the subacromial space and repaired by double-row suture bridge technique for full-thickness tears or single-row retensioning technique for partial-thickness tears [20]. No patient underwent subscapularis repair. Long head of the biceps tendon (LHBT) lesions, such as partial tears or subluxation, were recorded. Additionally, hyperemic synovitis observed during arthroscopy was documented. Hyperemic synovitis was characterized by the presence of elongated villi and reddish-colored blood vessels in the synovial membrane [21].

Clinical Evaluations

Clinical evaluations were performed at every routine outpatient visit (initial and final [minimum 1-year]). Active range of motion (ROM) including forward flexion, external rotation at the side, and internal rotation (IR) behind the back were evaluated. IR was graded based on the patient’s ability to reach the vertebral spinous process with the tip of the thumb as follows: T1–12, scored 1–12; L1–5, scored 13–17; and buttock, scored 18. Trained shoulder fellows who were blinded to the study measured the ROM.

Functional scores of pain visual analog scale (VAS), functional VAS, American Shoulder and Elbow Surgeons (ASES) score, and Constant score were evaluated. Scaption, external rotation, IR, and belly press strength were measured using a handheld myometer (FGJN-20; Nidec-Shimpo Co.) [4]. Shoulder scaption is lifting the arm in the scapular plane. Scaption strength was measured by applying upward force with the arm at 90° while seated. External and internal rotator strength were measured with the upper arm positioned at the side of the torso. Belly press strength was measured with the elbow positioned in front of the trunk and the palm pushed toward the abdomen [4]. All shoulder strength measurements were performed by pushing the gauge of a handheld myometer as far as possible for 5 seconds, and the peak value was recorded. We calculated the strength ratio as the ratio of power values of the operated and non-operated sides. All functional scores and strength measurements were performed by a shoulder specialized athletic trainer (SML) blinded to the study populations. Among the patients involved, 111 of 153 (72.5%) were followed up for over 1 year with clinical outcome evaluations, and 132 of 153 (86.3%) underwent postoperative 6-month MRI.

Rehabilitation

An abduction brace was applied for 4 weeks. Passive ROM exercises were started in the second postoperative week. Active ROM exercises were permitted when 80% ROM was achieved passively compared to the contralateral side. Strengthening exercises using an elastic band were initiated 3 months after surgery. Sports activity and heavy labor work were restricted until 6 months postoperatively. For patients with osteoporosis or large to massive tears, the duration of immobilization was extended by an additional 2 weeks, which consequently delayed rehabilitation.

Statistical Analysis

Statistical analyses were conducted using R version 4.0.3. The level of significance was set at P<0.05 with a two-tailed test. Continuous variables are presented as mean±standard deviation (range) or median (interquartile range [IQR]). Categorical variables are presented as number (%), unless otherwise indicated. In this study, subscapularis Goutallier grade 1 or 2 was defined as progressed subscapularis MAFI. Univariable logistic regression analysis was conducted to analyze the factors associated with progressed subscapularis MAFI. Student t-test or Mann-Whitney U-test was used for analyzing continuous variables, and the chi-square test or Fisher's exact test was employed for categorical variable analysis. Multivariable logistic regression analysis, using the Akaike information criterion to select the best model, was employed to identify predisposing factors for progression of subscapularis MAFI. In addition, preoperative and final clinical outcomes were compared according to the presence or absence of progressed subscapularis MAFI using the Mann-Whitney U-test. Cohen’s kappa statistics and interclass correlation coefficient were used for radiologic measurement reliability, rated as excellent, fair to good, and poor at kappa >0.75, 0.4 to 0.75, and <0.4, respectively. The intra- and inter-observer reliability values of radiologic measurement were excellent (Supplementary Table 1).

Subscapularis MAFI without Tendon Tear

During surgical planning, MAFI on Y-view provides valuable information for RC repair [11]. For posterosuperior RC repair, evaluating supraspinatus and infraspinatus MAFI can help predict the reparability of the tendon. Kim et al. [26] reported that infraspinatus Goutallier grades higher than 2 showed significant association with reparability. Also, supraspinatus Goutallier grade has been reported as a significant predictor for retear in revision RC repair [27].

Subscapularis MAFI also gives information for subscapularis tear [28]. Kilic et al. [9] reported that higher upper subscapularis Goutallier grade and coracohumeral distance narrowing are predictive of subscapularis tears in patients undergoing arthroscopic RC repair. They reported that subscapularis Goutallier grades 1 and 2 showed 77.1% and 98.7% of subscapularis tear, respectively. Seppel et al. [8] reported a new quantitative measurement method for subscapularis atrophy and compared it between normal patients and those with isolated subscapularis tears. They calculated the cross-sectional area ratio of the upper and entire subscapularis muscle from Y-view images and found that this ratio was significantly lower in the subscapularis tear group.

However, during practice, we have realized that not all subscapularis MAFIs indicate subscapularis tears. There have been cases where a subscapularis tear was not detected intraoperatively, despite progressed MAFI on preoperative MRI. In our study, this condition was common in patients with posterosuperior RC tears, occurring in 73.8% of cases, which had not been well reported previously. Additionally, we performed an analysis of the predisposing factors for this condition.

Age and Female Sex

RC tear is a primary degenerative disease of aging, so it might be a natural finding for age to be related to subscapularis MAFI [29]. Previously, age has been reported to be related to supraspinatus fatty infiltration [30]. Gueniche and Bierry reported [12] that subscapularis MAFI was observed with increasing age in 210 patients with an intact RC. They reported that 60% of patients aged 30–40 years had no subscapularis muscle fatty atrophy, while only 3% of patients older than 70 years had no such atrophy. Additionally, patients older than 70 years had a median Goutallier grade of 3 for subscapularis fatty atrophy, even in the absence of an RC tear. Also, Riem et al. [31] analyzed RC muscle volume using three-dimensional volume reconstructions from more than 200 CT and MRI scans and found that age had a significant impact on subscapularis muscle volume decrease.

Barry et al. [30] reported that female sex is associated with supraspinatus MAFI. This phenomenon can be explained by the differences in sarcopenia between genders [32]. Della Peruta et al. [32] reported that women are more susceptible to disuse muscle atrophy than men during aging. Various factors could contribute to this difference, and among them, a decrease in insulin-like growth factor-1 in women specifically contributes to the difference in sarcopenia between genders [32]. In an animal study, male and female mice aged 3, 9, 18, and 24 months were compared for RC contractile tension and muscle fatty infiltration [33]. Among various conditions, only the 24-month-old female mice showed significantly decreased contractile tension and increased fatty infiltration of the RC, which corresponds with our research [33].

Synovitis

Another contributing factor for subscapularis MAFI was identified as synovitis. Limited research has explored the relationship between synovitis and RC MAFI. Abrams et al. [34] observed a high level of synovitis in patients with full-thickness RC tears. They also noted elevated matrix metalloproteinase-3 in patients with synovial inflammation. Synovitis accompanying RC tears seems to induce chronic shoulder pain and may contribute to shoulder disuse. Shoulder disuse following an RC tear has been identified as a contributing factor to RC muscle atrophy and degeneration [35].

Posterosuperior RC Tear and Subscapularis MAFI

Some studies have reported mutual influence between posterosuperior RC and the subscapularis. Bergin et al. [13] analyzed the relation between subscapularis tear and supraspinatus MAFI and suggested that the latter was of diagnostic help for the former. They claimed that supraspinatus tear results in loss of the depressor function of the humeral head, which induces abnormal mechanical impingement of the subscapularis tendon insertion to the subcoracoid arch. Also, Cheung et al. [36] noted that torn supraspinatus tendons in elderly patients lead to fatty infiltration of neighboring RC muscles.

Mehta et al.’s prospective study [37] involved 354 patients with asymptomatic RC tears for which they underwent annual ultrasound examinations. The incidence of new subscapularis muscle tears was higher in those with posterosuperior RC tears compared to those without. This suggests that posterosuperior RC tears can affect subscapularis tear initiation, which aligns with our findings. The supraspinatus and subscapularis muscles are separated by the rotator interval but are interconnected via the comma tissue, which is pivotal for anterosuperior RC stability [38]. The subscapularis tendon performs different actions depending on its insertion. Biomechanical studies showed that the upper subscapularis contributes to shoulder elevation, while the lower subscapularis aids in shoulder adduction [25,39]. This demonstrates a biomechanical relationship between supraspinatus and subscapularis tendons [39].

Surgical Outcomes

In our study, preoperative subscapularis MAFI did not affect postoperative repaired tendon integrity. Supraspinatus retears at the 6-month follow-up MRI were similar between patients with and without subscapularis MAFI. This suggests that subscapularis MAFI does not compromise the structural outcome of RC repairs. Also, the final clinical outcome was not affected by subscapularis MAFI. Patients with progressed subscapularis MAFI had significantly lower preoperative ASES and Constant scores, as well as reduced scaption, IR strength, and belly press strength ratios compared to those without progressed MAFI. However, there were no significant differences in final ROM, functional scores, or shoulder strength, including subscapularis-specific tests, at the last follow-up.

Although the repair of a partial subscapularis tear remains in question, it is important to consider potential subscapularis pathology during surgical planning. This study is the first to evaluate the natural phenomenon of subscapularis MAFI in patients with posterosuperior RC tears without a subscapularis tendon tear. Surgeons should be aware of these natural changes, particularly in older patients, female patients, and those with combined posterosuperior RC tears. Furthermore, synovitis may accompany subscapularis MAFI, although preoperative evaluation of intra-articular synovitis can be challenging.

Limitations

This study has several limitations. First, the study has inherent limitations due to its retrospective design and single-surgeon case series. Additionally, the study participants were included solely based on specific surgical indications, which may have led to selection bias. Although partial supraspinatus tears can be managed conservatively, this study only includes patients who underwent surgical procedures, contributing to selection bias. Furthermore, the study focused on patients with smaller than medium-sized isolated supraspinatus tears, which may limit the generalizability of the results. For example, this study suggests that progressive subscapularis MAFI does not affect the postoperative integrity of the repaired tendon, while different effects might be observed in cases involving large supraspinatus and subscapularis atrophy. According to the cable theory, significant subscapularis atrophy can influence the outcomes of supraspinatus repair. Therefore, the specific inclusion criteria and conditions of this study are not generally applicable to all cases RC tear. Second, radiologic measurement errors are unavoidable. Assessing subscapularis MAFI using the Goutallier classification is inherently limited due to the absence of a bony boundary around the upper portion of the subscapularis muscle on the Y-view.

Previously, Yoon et al. [17] reported fair to good reliability in the subscapularis Goutallier classification measurements by raters and suggested a modified method for subscapularis MAFI. However, we established the measurement criteria for subscapularis Goutallier grade and demonstrated excellent reliability both within and between observers. Third, lack of detection does not guarantee absence of a subscapularis tear. Previous reports indicate that a concealed intratendinous lesion can be identified from the subacromial space by releasing the transverse humeral ligament and removing the LHBT. However, this procedure was not carried out for the patients in this study because it would damage the subscapularis bursal side insertion and connection with the supraspinatus tendon [40]. Fourth, the long-term clinical and structural outcomes of patients with subscapularis MAFI have not been studied. Clinical outcomes were evaluated for at least 1 year in 72.5% of patients, and structural outcomes were assessed using MRI at 6 months after surgery. The understanding of the natural course of this condition, particularly whether it serves as a precursor to subscapularis tear, could be clarified through an analysis of the long-term clinical and structural outcomes. Fourth, clinical relevance related to daily life or return to sports activities was not identified.