Supraspinatus Tendon Retraction and Histologic Myofiber Size are Associated with Supraspinatus Muscle Atrophy on MRI

Abstract

Background:

Atrophy of the rotator cuff is a negative prognostic indicator after rotator cuff repair. While full-thickness rotator cuff tears accompanied by tendon retraction are commonly associated with decreased muscle cross-sectional area (CSA) on MRI, it is unclear whether this is accompanied by histological atrophy of rotator cuff myofibers.

Hypothesis/Purpose:

To evaluate the impact of supraspinatus tendon retraction and myofiber size on supraspinatus atrophy on MRI.

Study Design:

Cross-sectional study

Methods:

Supraspinatus muscle biopsies were obtained from consecutive patients undergoing arthroscopic shoulder surgery. Rotator cuff tears were classified according to their size. Preoperative MRI was used to measure tendon retraction and CSA of the supraspinatus in the Y-shaped view. Occupation ratio of the supraspinatus was calculated by dividing the supraspinatus CSA by the CSA of the supraspinatus fossa. Muscle biopsies were examined using LAMININ to quantify myofiber CSA. The association between supraspinatus tear size and measures of histologic and MRI muscle atrophy were compared using standard statistical tests. A multivariable logistic regression analysis was used to identify independent predictors of muscle atrophy on MRI.

Results:

Thirty-eight patients were included (8 no tear, 14 partial-thickness, and 16 full-thickness). Increasing tear size was associated with a greater distance of tendon retraction (p<0.001), smaller mean histologic myofiber size (p=0.004), lower mean supraspinatus CSA on MRI (p<0.001), lower occupation ratio (0.73 vs. 0.66 vs. 0.53 vs. 0.38 in controls, partial, small-to-medium full-thickness, and large to massive full-thickness, respectively, p<0.001). On Pearson correlation analysis, tendon retraction demonstrated strong correlation with occupation ratio (−0.727, p<0.001), and weak correlation to myofiber size (−0.438, p=0.006), while occupation ratio showed moderate correlation with myofiber size (0.563, p<0.001). Multivariable linear regression analysis demonstrated increasing tendon retraction (p<0.001), age (p=0.034) and smaller histologic myofiber CSA (p=0.047) to be independently associated with greater supraspinatus atrophy on MRI.

Conclusion:

Supraspinatus muscle atrophy appreciated on MRI is independently associated with patient age, tendon retraction, and atrophy of the supraspinatus myofibers at the histological level.

Introduction:

Atrophy of the rotator cuff is associated with higher risk of tendon repair failure and poor functional outcomes following rotator cuff repair.^6,9,11,15,18 Therefore, assessing for rotator cuff atrophy is an important component of evaluating patients with rotator cuff tears. Multiple different methodologies have been proposed to quantify the severity of rotator cuff atrophy, which typically utilize the scapular Y-view (most lateral slice where the scapular spine contacts the body) on sagittal-oblique reconstruction of a magnetic resonance imaging (MRI) study.^3,18–20,22 However, there is concern that two-dimensional assessment of the supraspinatus on cross-sectional imaging may not be an accurate representation of the true atrophy of the muscle fibers of the rotator cuff.^4,10,21

Specifically, rotator cuff tendon retraction has been shown to falsely overestimate muscle atrophy appreciated on MRI,^4,21 which can have clinically significant implications for determining the repairability and functional prognosis after rotator cuff repair. While MRI is a clinically useful non-invasive method of assessing muscle atrophy, muscle biopsy has been performed to directly measure myofiber cross-sectional area (CSA) histologically.^5,13,16 Prior histopathological studies on human rotator cuff muscle have demonstrated full thickness supraspinatus tears are associated with corresponding decreased myofiber size compared to partial-thickness tears.^5,13,16

Despite the growing knowledge on the risk factors and pathogenesis of muscle atrophy following rotator cuff tear, there is limited evidence investigating whether the atrophy appreciated on MRI is associated with decreased histologic myofiber size of the muscle, which may bear importance in clinical outcomes. Additionally, since tendon retraction can result in decreased apparent size of muscle on MRI, it is important to understand the interplay between myofiber size, tendon retraction, and CSA assessed on MRI. While full-thickness rotator cuff tears accompanied by tendon retraction are commonly associated with decreased muscle CSA on MRI, it is unclear whether this is accompanied by histological atrophy of rotator cuff myofibers. Further understanding this can have significant implications for determining repairability of tears and the prognosis of rotator cuff repair. Therefore, the purpose of our study is to evaluate the impact of supraspinatus tendon retraction and myofiber size on supraspinatus occupation ratio on MRI. We hypothesize that both tendon retraction and histologic myofiber size will be significantly associated with supraspinatus occupation ratio.

Methods:

This was a prospective case-control study at a single academic institution between September 2020 and May 2021. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)²² statement was used in reporting this study. The study gained institutional review board approval. Patients indicated for arthroscopic shoulder surgery were consented, enrolled and surgically managed by the senior author. Patients undergoing rotator cuff repairs comprised the treatment group whereas patients undergoing surgery for a pathology other than rotator cuff tears comprised the control group (Table 1). Inclusion criteria included adult patients >18 years, preoperative magnetic resonance imaging (MRI), and had supraspinatus predominant pathology in patients indicated for rotator cuff repair. Supraspinatus tendon tear predominance was defined if either a) supraspinatus was the only tendon torn, or b) supraspinatus had the largest tear in the setting of a multi-tendon tear. Demographics including age, BMI, gender, ethnicity were collected. Patient-reported outcomes including visual analog score for pain (VAS), American Shoulder and Elbow Surgeons Survey (ASES), and Single Assessment Numeric Evaluation (SANE) were obtained preoperatively. The primary outcome of the current study is to determine variables (age, tendon retraction, myofiber size) associated with occupation ratio of the supraspinatus measured on MRI.

Table 1 –

Surgical indications and arthroscopic procedure performed in control group

Procedure	Number of patients
Labral repair	3
Biceps tenodesis	2
Lysis of adhesions/debridement	1
Distal clavicle excision	1
Subacromial decompression	1

Fifty-two total patients were enrolled in the current study. Of the fifty-two patients, there were nine poor quality biopsy samples were excluded, three patients without access to preoperative MRI images, and two infraspinatus predominant rotator cuff tears were excluded. Therefore, thirty-eight patients were included (8 no tear, 14 partial-thickness, and 16 full-thickness) in our final cohort (Figure 1).

Figure 1 –

Study cohort selection with inclusion and exclusion criteria.

Magnetic Resonance Imaging

Preoperative MRI was used to measure tendon retraction, cross-sectional area of the supraspinatus in the Y-shaped view, presence of the tangent sign, and Goutallier stage. MRI shoulder protocol includes imaging obtained on a 1.5T or 3T machine, with patient supine, arm externally rotated. The sagittal oblique axis is defined as perpendicular to the plane of the scapula, and the sagittal T1 FSE sequence is obtained with the following parameters: 160mm field of view, repetition time 767ms, echo time 11ms, 0.5mm x 0.5mm x 4mm resolution, 4mm slice thickness and 1mm slice gap, 320x192 matrix. The Y-shaped view was defined as the most lateral slice on the sagittal oblique reconstructions where the scapular spine contacts the body.^3,18–20,22 Occupation ratio of the supraspinatus in the supraspinatus fossa allowed for a standardized measurement of muscle atrophy, and a value less than 0.5 defined atrophy.^2,18,19 Briefly, the occupation ratio was measured by dividing the CSA of the supraspinatus on the scapular Y view by the CSA of the supraspinatus fossa.¹⁹ Figure 2 and Figure depict representative occupation ratio measurements in subjects with a partial-thickness supraspinatus tear and massive full-thickness tear, respectively.

Figure 2 –

Sagittal oblique magnetic resonance imaging (MRI) view demonstrating the occupation ratio (OR) measurements at the scapular-Y position for a patient with a partial-thickness supraspinatus tear. The supraspinatus muscle belly is outlined and cross-sectional area represented as 932 mm², while the supraspinatus fossa is outlined and measures 1031 mm². Calculated OR = 932 mm²/1031 mm² = 0.90, representing good muscle bulk.

Rotator Cuff Tear Severity

Rotator cuff tears (RCTs) were classified according to the partial or full-thickness involvement of the tendon. Size of full thickness tears were further classified intraoperatively using the DeOrio and Cofield classification system which groups RCTs on the basis of greatest dimension as either small (<1 cm), medium (1–3 cm), large (3–5 cm), or massive (>5 cm).¹ In this manner, patients were categorized on severity of rotator cuff tendon tears into four groups: 1) control (no tear) patients, 2) partial-thickness tears, 3) small to medium full-thickness tears, and 4) large to massive full-thickness tears.

Muscle Biopsy and Processing

All patients enrolled underwent en bloc supraspinatus muscle biopsy according to the same methodology published previously.¹⁶ In order to standardize the location of arthroscopic muscle biopsy across patients, the biopsies were collected after rotator cuff repair. The lateral border of the scapular spine was arthroscopically skeletonized with bipolar radiofrequency ablation to differentiate supraspinatus (anterior to scapular spine) from infraspinatus (posterior to scapular spine), and care was taken to ensure that muscle was clearly visualized medial to the musculotendinous junction. Using 2.4-mm arthroscopic biter, a muscle biopsy was obtained just medial to the musculotendinous junction and stored in 30% sucrose solution at 4°C. Muscle biopsies were then embedded in Tissue-Tek^© and flash-frozen in isopentane supercooled by liquid nitrogen. Muscle biopsies were then cryosectioned and fixed in 4% paraformaldehyde before immunofluorescent staining.

Myofiber size: LAMININ Staining and Analysis

Muscle biopsies were examined using LAMININ to quantify myofiber CSA. LAMININ is an extracellular protein found in the basement membrane of myofibers. As such, LAMININ can be used to outline muscle fibers and can be used to determine myofiber size. Detailed methodology on our LAMININ staining process can be found in a previously published study.¹⁶ Once stained, muscle fiber CSAs were quantified using the BioVoxxel toolbox plugin for ImageJ.¹⁷ LAMININ immunofluorescent staining was analyzed and depicted as heat maps, allowing for quantification of myofiber size. Biopsies which did not stain appropriately with LAMININ or lacked clear muscle boundaries were excluded due to their inability to be processed within ImageJ. Of the nine biopsies that did not stain appropriately, there were 2 controls (no tears), 3 partial-thickness tears, 3 small to medium full-thickness tears, and 2 large to massive full-thickness tears. Therefore, the higher severity of tears did not appear to predispose those muscles to inadequate staining.

Statistical Analysis

All statistical analyses were performed using SPSS 28 (IBM SPSS Statistics, Version 28.0. Armonk, NY: IBM Corp). Primary outcome variables including occupation ratio, myofiber size, and tendon retraction were assessed for normality using the Shapiro-Wilk test. Occupation ratio (p=0.287) and myofiber size (p=0.113) demonstrated no significant deviation from normal distribution, while tendon retraction (p<0.001) demonstrated nonparametric distribution. The demographics, patient-reported outcomes, and muscle atrophy measurements of the four study groups were compared using ANOVA for normally distributed continuous variables, Kruskal-Wallis for nonparametric continuous variables, and Chi-Square test for categorical variables. Post-hoc analysis for the continuous variables included Tukey honest significant difference test. Pearson correlation analysis was performed to evaluate the correlation between tendon retraction, myofiber size, and occupation ratio. Correlation strength was determined, where <0.3 is no to very weak, 0.3 to 0.5 is weak, 0.5 to 0.7 is moderate, and >0.7 is strong.¹⁴ Additionally, multivariable logistic regression was performed to assess which variables (age, myofiber size, tendon retraction) are independently associated with occupation ratio. Post-hoc power analysis was performed and demonstrated a power of 1.0, calculated using 38 samples, an alpha of 0.05, and a total effect size of 1.87 computed from R² = 0.652 from our linear regression model. Cross-sectional area of supraspinatus and occupation ratio measurements on MRI were performed by two authors, and the inter-rater reliability demonstrated an intraclass correlation of 0.982 (95% CI 0.966 to 0.991, p<0.001, demonstrating excellent agreement. P value <0.05 was considered statistically significant.

Results:

The only significant difference in patient demographics between the study groups was that groups with increasing severity of rotator cuff tear had a higher mean age (p=0.004). All other demographics were found to be similar between the groups. (Table 2)

Table 2 –

Patient demographics

	Control	Partial Thickness Tear	Small to Medium Full-Thickness Tear	Large to Massive Full-Thickness Tear	p values
Number of patients	8	14	8	8
Age	44.50 ± 17.05	55.29 ± 12.71	66.38 ± 9.55	63.75 ± 6.09	0.004*
BMI	28.43 ± 5.27	29.11 ± 4.85	29.07 ± 7.59	32.24 ± 4.64	0.513
Male Gender	6 (75.0%)	10 (71.4%)	4 (50.0%)	6 (75.0%)	0.652
White Ethnicity	6 (75.0%)	5 (35.7%)	4 (50.0%)	5 (62.5%)	0.315
VAS Pain	5.50 ± 2.07	6.86 ± 2.41	4.29 ± 0.95	5.43 ± 2.57	0.094
ASES	45.71 ± 16.87	39.57 ± 11.90	53.43 ± 9.95	47.43 ± 15.34	0.172
SANE	55.00 ± 20.49	54.93 ± 15.72	45.00 ± 14.83	41.43 ± 15.74	0.260

^*p<0.05

Tear Severity and Supraspinatus Atrophy

Increasing tear severity from no tear to large-to-massive full-thickness tear was associated with lower mean histologic myofiber CSA (p=0.004) (Figure 4), lower mean supraspinatus CSA on MRI (p<0.001), lower occupation ratio (0.73 vs. 0.66 vs. 0.53 vs. 0.38 in controls, partial, small-to-medium full-thickness, and large-to-massive full-thickness, respectively, p<0.001), and higher rate of tangent sign present on MRI (0% vs. 0% vs. 25% vs. 75% in controls, partial, small-to-medium full-thickness, and large-to-massive full-thickness, respectively p<0.001). There was also an increasing proportion of patients with supraspinatus atrophy (defined by occupation ratio >0.5) as severity of rotator cuff tendon tear increased (0% vs. 7.1% vs. 37.5% vs. 87.5% in controls, partial, small-to-medium full-thickness, and large-to-massive full-thickness, respectively p<0.001). (Table 3)

Figure 4 –

Fiber size decreased with increasing tear severity from no-tear to large-to-massive full thickness tears. Top row demonstrates muscle biopsy specimens stained for laminin. Scale bars represent 200 µm. Bottom row demonstrates myofibers color-coded by cross-sectional area from sections stained for laminin using the BioVoxxel toolbox plug-in for ImageJ.

Table 3 –

Supraspinatus muscle atrophy on histologic and magnetic resonance imaging analysis

	Control	Partial Thickness Tear	Small to Medium Full-Thickness Tear	Large to Massive Full-Thickness Tear	P values
Number of patients	8	14	8	8
Retraction (cm)	0	0	1.55 ± 0.78	3.70 ± 1.52	<0.001**ϕ
Histologic
Mean Supraspinatus Myofiber Size (μm)	2,249.0 +/− 713.8a	2,257.1 +/− 876.2b	1,485.0 +/− 523.8	1,218.3 +/− 441a,b	0.004 A=0.029 B=0.011
Magnetic Resonance Imaging (MRI)
Mean CSA of Supraspinatus (mm2)	801.25 ± 218.88a,b	748.11 ± 211.75c	507.81 ± 246.02a	377.38 ±162.12b,c	<0.001** A=0.043* B=0.002* C=0.002*
Mean Supraspinatus Fossa CSA (mm2)	1094.19 ± 245.40	1,120.46 ± 227.65	935.50 ± 232.88	975.69 ± 172.39	0.214
Mean Occupation Ratio	0.7313 ± 0.1192a,b	0.6625 ± 0.1205c	0.5266 ± 0.1477a	0.3798 ± 0.1138b,c	<0.001** A=0.012* B<0.001** C<0.001**
Occupation Ratio < 0.5	0 (0%)	1 (7.1%)	3 (37.5%)	7 (87.5%)	<0.001**
Tangent Sign Present	0 (0%)	0 (0%)	2 (25.0%)	6 (75.0%)	<0.001**
Goutallier Stage	0	0	0.5	2.0	0.002*ϕ

A,B,C represent p values for post-hoc Tukey honest significant difference test

CSA Cross-sectional area

^*p<0.05

^**p<0.001

^ϕKruskal-Wallis test performed for nonparametric data

Goutallier stage reported as median value.

Association Between Myofiber Size, Tendon Retraction, and Muscle Atrophy on MRI

On Pearson correlation analysis, tendon retraction demonstrated a strong negative correlation with occupation ratio (−0.725, p<0.001) and a moderate negative correlation to myofiber CSA (−0.529, p<0.001), while occupation ratio showed a moderate positive correlation with myofiber CSA (0.593, p<0.001) (Table 4).

Table 4 –

Pearson correlation between tendon retraction, myofiber size, and occupation ratio

	Tendon retraction (cm)	Myofiber size (μm)	Occupation Ratio	Age
Tendon retraction (cm)	--	−0.529** (Moderate) (p=0.006)	−0.725** (Strong) (p<0.001)	0.337 (Weak) (p=0.039)
Myofiber size (μm)	−0.529** (Moderate) (p=0.006)	--	0.593** (Moderate) (p<0.001)	−0.437* (Weak) (p=0.006)
Occupation Ratio	−0.725** (Strong) (p<0.001)	0.593** (Moderate) (p<0.001)	--	−0.512** (Moderate) (p<0.001)
Age	0.337 (Weak) (p=0.039)	−0.437* (Weak) (p=0.006)	−0.512** (Moderate) (p<0.001)	--

^*p<0.05

^**p<0.001

Multivariable linear regression analysis was used to assess whether tendon retraction and histologic myofiber size are independently associated with supraspinatus occupation ratio on MRI. The results demonstrated that tendon retraction (B coefficient −5.711, 95% confidence interval [CI] −8.187 to −3.235; p<0.001), histologic myofiber CSA (B coefficient 0.005, 95% CI 0.000 to 0.010; p=0.047), and age (B coefficient −0.305, 95% CI 0.138 to −0.244, p=0.034) are independently correlated with the occupation ratio (Table 5). Therefore, for each additional centimeter of tendon retraction there is an associated decrease in occupation ratio by 5.7% (i.e. occupation ratio is predicted to decrease from 65% to 48% on the scapular Y view for 3 cm of tendon retraction). Tendon retraction had the strongest correlation to occupation ratio with a standardized Beta coefficient of −0.540 (moderate strength correlation). These findings suggest that while patient age and myofiber size contribute to occupation ratio, tendon retraction has the strongest correlation with muscle atrophy appreciated on MRI when using the occupation ratio.

Table 5 –

Multivariable linear regression to identify the association between myofiber size and tendon retraction on supraspinatus atrophy (occupation ratio) on MRI

	B coefficient (95% CI)	Standardized Beta Coefficient	p value
Myofiber size (μm)	0.005 (0.000 to 0.010)	0.240	0.047*
Tendon retraction (cm)	−5.711 (−8.187 to −3.235)	−0.540	<0.001**
Age	−0.305 (0.138 to −0.244)	−0.244	0.034*

Occupation ratio in this model is multiplied by 100 to improve interpretability of this regression analysis

Variance inflation factor (VIF) of myofiber size = 1.3, VIF of tendon retraction = 1.3, and VIF of age = 1.2, which demonstrates no issues with multicollinearity.

^*p<0.05

^**p<0.001

Discussion:

Rotator cuff atrophy is a critical feature in the natural history of rotator cuff tears and is associated with poor functional outcomes after rotator cuff repair.^6,9,11,15,18 As such, the importance of quantifying atrophy of the rotator cuff has been an area of considerable research,^{3,8,18–20,22} although there is limited prior data investigating whether quantitative atrophy appreciated on MRI is secondary to a true decrease in myofiber size. A previous study confirmed that tendon status (full-thickness vs partial or no tear) was correlated with lower mean myofiber CSA.¹⁶ However, the impact of tendon retraction and myofiber size on supraspinatus atrophy observed on MRI remains largely unknown. This study demonstrates that there is increasing atrophy of the supraspinatus muscle, both histologically and on MRI, with increasing size of full-thickness supraspinatus tendon tears. Additionally, our results show that myofiber size, tendon retraction, and patient age are independently associated with supraspinatus occupation ratio on MRI, and therefore both contribute to the diminished CSA observed clinically on MRI.

Our findings are supported by previous studies suggesting that histologic myofiber atrophy accompanies rotator cuff tears. Lundgreen et al¹³ evaluated supraspinatus muscle biopsies from 24 patients undergoing arthroscopic repair of partial- or full-thickness supraspinatus tendon tears, and found that patients with full-thickness tears had more substantial atrophy of myofibers, as well as reduced density of satellite cells and fewer proliferating cells, suggesting lower regenerative potential. A prior study¹⁶ evaluated 39 rotator cuff biopsies in patients undergoing arthroscopic shoulder surgery showed that full-thickness rotator cuff tears had a smaller mean myofiber CSA compared to partial-thickness and no tear control patients. This prior literature substantiates our current findings that increasing severity of supraspinatus tendon tear is associated with decreased myofiber CSA.

Despite the evidence that histologic atrophy of the myofibers occurs in rotator cuff tears, there has been limited previous literature correlating these findings with quantitative atrophy of the muscle appreciated on MRI. Since MRI is the current gold-standard, non-invasive tool to evaluate rotator cuff pathology, understanding the correlation between MRI findings and histologic atrophy is important to validate current MRI measurements. Additionally, understanding the correlation will help determine the relative impact of tendon retraction on perceived muscle atrophy on MRI and further elucidate whether atrophy of the rotator cuff is reversible following successful rotator cuff repair.

One variable that has consistently been shown to impact muscle CSA on MRI is the distance of rotator cuff tendon retraction.^4,21 Fukuta et al.⁴ evaluated 76 shoulders in patients undergoing arthroscopic surgery for rotator cuff pathology and found that supraspinatus CSA and occupation ratio decreased in conjunction with increasing size of tendon tear. Additionally, they found that tendon retraction is negatively correlated with the supraspinatus CSA observed. Further supporting the theory that tendon retraction influences the observed muscle CSA, Jo and Shin¹² performed an MRI-based study investigating rotator cuff atrophy preoperatively and 3 days postoperatively, which demonstrated that the CSA of the supraspinatus increased 21.6%. They suggested that when the retracted tendon is reduced to its footprint on the greater tuberosity, the muscle of the supraspinatus is shifted laterally and may therefore increase the observed CSA on the Y-shaped view on MRI. Therefore, comparing the occupation ratio or CSA of the supraspinatus on preoperative MRI to post-operative MRI, one must account for the tendon retraction present in the preoperative study.

Our results help clarify the relative correlations and contributions of tendon retraction and myofiber size on supraspinatus occupation ratio on MRI. The present study demonstrated a strong correlation between tendon retraction and occupation ratio (0.725, p<0.001), and a moderate correlation between myofiber size and occupation ratio (0.593, p<0.001), suggesting that tendon retraction likely contributes more to decreased occupation ratio on MRI than myofiber size. Additionally, when including both tendon retraction and myofiber size in a multivariable linear regression model, both variables remain significant predictors of occupation ratio and therefore likely independently contribute to atrophy appreciated on MRI. Using our results from the regression analysis, we can predict the expected impact of tendon retraction on occupation ratio. For each additional centimeter of tendon retraction there is an associated decrease in occupation ratio by 5.7% (i.e. occupation ratio is predicted to decrease from 65% to 48% on the scapular Y view for 3 cm of tendon retraction), while each increasing year of age is associated with a 0.3% decrease in occupation ratio. Using our model to determine an adjusted occupation ratio for the supraspinatus accounting for patient age and tendon retraction can help further classify the degree of true muscle atrophy using MRI. These findings can help further elucidate the repairability of rotator cuff tears and aid our interpretation of the reversibility of muscle atrophy on MRI.

Limitations

Although our study compares quantitative supraspinatus atrophy on MRI to histologic atrophy of the myofibers, our study is not without limitations. Due to the natural spatial variation of muscle quality within an individual muscle belly,⁷ it is possible the biopsy samples were not fully representative of the muscle being sampled. Additionally, using the scapular Y view of the MRI to evaluate supraspinatus CSA and occupation ratio has its own unique limitations with regards to inconsistencies in scanning axis of the reconstructed image.¹⁰ Furthermore, there is an inability to perform volumetric analysis of muscle atrophy on standard shoulder MRI due to the limited field of view that does not include the entire muscle belly of the supraspinatus. Despite this, the methodology used in our study including occupation ratio, supraspinatus CSA, and tangent sign are consistent with the current standard for evaluating atrophy in the clinical setting. Therefore, our findings are clinically relevant and provide data to support the relative contributions of tendon retraction, myofiber size, and age to the calculated occupation ratio of the supraspinatus on MRI.

Conclusions

Our study supports the belief that supraspinatus atrophy appreciated on MRI is independently associated with both atrophy of the supraspinatus myofibers at the histological level, as well as tendon retraction. These findings highlight the need for future research to distinguish the relative impact of tendon retraction and true atrophy of the muscle fibers on clinical outcomes and reversibility of atrophy following rotator cuff repair.

Figure 3 –

Sagittal oblique magnetic resonance imaging (MRI) view demonstrating the occupation ratio (OR) measurements at the scapular-Y position for a patient with a massive full-thickness supraspinatus tear. The supraspinatus muscle belly is outlined and cross-sectional area represented as 271 mm², while the supraspinatus fossa is outlined and measures 1066 mm². Calculated OR = 271 mm²/1066 mm² = 0.25, representing an atrophied muscle.

Clinical Relevance:

Assessing supraspinatus muscle atrophy on scapular Y view has limitations and differentiating between tendon retraction and true myofiber atrophy may have substantial implications for determining chronicity and repairability of tears.

What is known about the subject:

Decreased myofiber size and reduced cross-sectional area on MRI have been shown to occur in full-thickness rotator cuff tears, although no quantitative assessment of muscle atrophy on MRI has been correlated to myofiber size. Additionally, when assessing for muscle atrophy of the supraspinatus on MRI, it is known that a single cross-sectional slice on the scapular Y view may overestimate the degree of muscle atrophy due to muscle-tendon retraction in full-thickness tears.

What this study adds to existing knowledge:

This is the first study to show correlation between quantitative muscle CSA on MRI and histologic myofiber size. This study also adds the first evidence that both muscle-tendon retraction and decreased myofiber size both independently contribute to muscle cross-sectional area observed on MRI, and that tendon retraction is more strongly correlated with atrophy appreciated on MRI. Additionally, we have quantified the impact of tendon retraction on the occupation ratio, which can be used to normalize calculations for muscle atrophy while adjusting for the impact of tendon retraction. This can help surgeons determine whether the supraspinatus truly is atrophied, or if the decreased CSA is secondary to retraction.