Advertisement

Longitudinal MR Quantification of the Fat Fraction within the Supraspinatus and Infraspinatus Muscles in Patients with Shoulder Pain

Open AccessPublished:March 22, 2022DOI:https://doi.org/10.1016/j.acra.2022.02.011

      Rationale and Objectives

      Knowing the natural history of fatty degeneration of rotator cuff muscles is important for estimating the risk and rate of progression to cuff tear arthropathy (CTA). The purpose of this study was to investigate the changes in rotator cuff muscle fatty degeneration over time quantitatively in patients treated conservatively for shoulder pain.

      Materials and Methods

      Thirty patients with a baseline and follow-up shoulder MRI, including a 2-point Dixon sequence, which were performed at least 1 year apart, were included. We classified patients into 3 groups: “full-thickness tear” (n = 7), “partial-thickness tear” (n = 13), and “no-tear” (n = 10) groups. The fat fraction in the supra- and infraspinatus muscles, and the rate of change in the fat fraction (ΔFfr) were calculated using the formula “fat fraction of follow-up MRI/fat fraction of initial MRI.” We investigated the difference in ΔFfr among the 3 groups and the degree of progression to CTA.

      Results

      Statistically significant differences in ΔFfr within the supraspinatus and infraspinatus muscles were found among full-thickness, partial-thickness, and no-tear groups (2.54 vs 1.02 vs 0.75, p < 0.001 and 1.96 vs 1.07 vs 0.73, p = 0.021, respectively). Overall, 71.4% of the full-thickness tear group showed progression of CTA, and 28.6% of the full-thickness tear group needed reverse shoulder arthroplasty within an average follow-up period of 34 months.

      Conclusions

      MR quantification, together with the knowledge of change in fatty degeneration over time, may be useful for the management of patients with shoulder pain.

      Keywords

      Abbreviations:

      CTA (cuff tear arthropathy), ΔFfr (the rate of change in the fat fraction)

      INTRODUCTION

      Arthroscopic repair of the rotator cuff has now become common practice and thus preoperative evaluation of the rotator cuff muscles, including fatty degeneration, has become an important part of work-up for determining the therapeutic strategy and predicting prognosis in patients with shoulder injuries (
      • Williams Jr., GR
      • Rockwood Jr., CA
      • Bigliani LU
      • et al.
      Rotator cuff tears: why do we repair them?.
      ,
      • Goutallier D
      • Postel JM
      • Zilber S
      • et al.
      Shoulder surgery: from cuff repair to joint replacement. an update.
      ). The term “fatty degeneration” was originally described by Goutallier et al. 1994, as follows, “Fatty degeneration was characterized on CT scan by an infiltration of rotator cuff muscles by areas of fat, a fact that could be confirmed histologically” (
      • Goutallier D
      • Postel JM
      • Bernageau J
      • et al.
      Fatty muscle degeneration in cuff ruptures. pre- and postoperative evaluation by CT scan.
      ). Typically, evaluation of fatty degeneration in rotator cuff muscles has been performed qualitatively using the Goutallier or modified Goutallier classification on MRI at present. Recently, several methods of quantifying the fat amount in tissues using MRI have been reported; these methods enable more objective evaluation (
      • Gilbert F
      • Meffert RH
      • Schmalzl J
      • et al.
      Grade of retraction and tendon thickness correlates with MR-spectroscopically measured amount of fatty degeneration in full thickness supraspinatus tears.
      ,
      • Matsumura N
      • Oguro S
      • Okuda S
      • et al.
      Quantitative assessment of fatty infiltration and muscle volume of the rotator cuff muscles using 3-dimensional 2-point Dixon magnetic resonance imaging.
      ,
      • Davis DL
      • Kesler T
      • Gilotra MN
      • et al.
      Quantification of shoulder muscle intramuscular fatty infiltration on T1-weighted MRI: a viable alternative to the Goutallier classification system.
      ,
      • Iijima Y
      • Matsuki K
      • Hoshika S
      • et al.
      Relationship between postoperative retear and preoperative fatty degeneration in large and massive rotator cuff tears: quantitative analysis using T2 mapping.
      ,
      • Nardo L
      • Karampinos DC
      • Lansdown DA
      • et al.
      Quantitative assessment of fat infiltration in the rotator cuff muscles using water-fat MRI.
      ,
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Predicting retear after repair of full-thickness rotator cuff tear: two-point dixon mr imaging quantification of fatty muscle degeneration-initial experience with 1-year follow-up.
      ,
      • Horiuchi S
      • Nozaki T
      • Tasaki A
      • et al.
      Reliability of MR quantification of rotator cuff muscle fatty degeneration using a 2-point dixon technique in comparison with the goutallier classification: validation study by multiple readers.
      ,
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Quantification of fatty degeneration within the supraspinatus muscle by using a 2-point dixon method on 3-T MRI.
      ,
      • Agten CA
      • Rosskopf AB
      • Gerber C
      • et al.
      Quantification of early fatty infiltration of the rotator cuff muscles: comparison of multi-echo Dixon with single-voxel MR spectroscopy.
      ,
      • Lee S
      • Lucas RM
      • Lansdown DA
      • et al.
      Magnetic resonance rotator cuff fat fraction and its relationship with tendon tear severity and subject characteristics.
      ). Previous reports using the preoperative 2-point Dixon method have shown that patients who experience re-rupture after rotator cuff repair have a higher rate of fatty degeneration of the muscles, and also show progression of fatty degeneration after the re-rupture (
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Predicting retear after repair of full-thickness rotator cuff tear: two-point dixon mr imaging quantification of fatty muscle degeneration-initial experience with 1-year follow-up.
      ). Functional prognosis after rotator cuff repair for patients with severe fatty degeneration is poor, and these patients may eventually need reverse shoulder arthroplasty (
      • Eajazi A
      • Kussman S
      • LeBedis C
      • et al.
      Rotator cuff tear arthropathy: pathophysiology, imaging characteristics, and treatment options.
      ,
      • Rugg CM
      • Gallo RA
      • Craig EV
      • et al.
      The pathogenesis and management of cuff tear arthropathy.
      ). The natural history of fatty degeneration of rotator cuff muscles without surgery is not well understood yet and is important for estimating the risk and rate of progression to cuff tear arthropathy (CTA). It may also allow determination of the appropriate timing of surgical intervention, as well as detection of high-risk groups for which early treatment is required. However, while there are numerous reports about the outcome of the treatments of rotator cuff tears with surgical intervention, there have been only a handful of reports that have conducted qualitative evaluations of the natural history of rotator cuff muscle fatty degeneration (
      • Melis B
      • Wall B
      • Walch G.
      Natural history of infraspinatus fatty infiltration in rotator cuff tears.
      ,
      • Melis B
      • DeFranco MJ
      • Chuinard C
      • et al.
      Natural history of fatty infiltration and atrophy of the supraspinatus muscle in rotator cuff tears.
      ,
      • Hebert-Davies J
      • Teefey SA
      • Steger-May K
      • et al.
      Progression of fatty muscle degeneration in atraumatic rotator cuff tears.
      ). Furthermore, there have been no reports that have evaluated this quantitatively.
      Thus, the purpose of this study was to investigate the change in fatty degeneration of rotator cuff muscles over time in patients treated conservatively for shoulder pain, including cases with rotator cuff tears.

      MATERIALS AND METHODS

      Patients

      The study was approved by our institutional review board (#19-R166), and the requirement for informed consent was waived because of the retrospective nature of the study. 200 patients underwent follow-up shoulder MRIs with the 2-point Dixon sequence for the evaluation of fatty degeneration of rotator cuff muscles between January 2011 and September 2019 at our institution. 45 patients were treated conservatively, without surgery. Of these, 30 patients (mean age 64.9 ± 9.3 years; range 45-82 years; 14 males and 16 females) who had undergone follow-up MRI performed at least 1 year after the initial MRI were included in this study (Fig. 1).

      MRI

      All MRI examinations were performed with a 3.0-T unit (Magnetom Verio, version VB17; Siemens, Erlangen, Germany) using a 4-channel flex coil. Standard MR examinations included fast spin-echo proton density-weighted imaging (PDWI) in the oblique coronal and sagittal planes (matrix, 320 × 320; repetition time/echo time, 2500 ms/19 ms; field-of-view, 180 mm; number of acquisitions, one; section thickness, 3 mm); fat-saturated fast spin-echo PDWI in the oblique coronal and sagittal planes (matrix, 320 × 320; repetition time/echo time, 2500/37; field-of-view, 180 mm; number of acquisitions, one; section thickness, 3 mm), and a T2*-weighted sequence in the axial plane (matrix, 320 × 320; repetition time/echo time, 900 ms/13 ms; field-of-view, 180 mm; number of acquisitions, one; section thickness, 3 mm). We also performed a 3-dimensional 2-point Dixon volumetric interpolated breath-hold examination sequence in the oblique sagittal plane. The 2-point Dixon sequence used an acquisition matrix of 128 × 128; repetition time of 6.5 ms; echo time 1.225/2.45 ms; flip angle, 10°; field-of-view, 196 mm; number of acquisitions, one; section thickness, 2.5 mm; and acquisition time, 2 minutes 30 seconds. From the 2-point Dixon sequence, water-only, fat-only, in-phase (water and fat), and out-of-phase (water minus fat) images were produced (
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Predicting retear after repair of full-thickness rotator cuff tear: two-point dixon mr imaging quantification of fatty muscle degeneration-initial experience with 1-year follow-up.
      ,
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Quantification of fatty degeneration within the supraspinatus muscle by using a 2-point dixon method on 3-T MRI.
      ).

      MR Image Interpretation

      Evaluation of rotator cuff and fatty degeneration of the initial and follow-up MR images was performed independently by 2 board-certified radiologists (S.A. and S.H., with 7 and 8 years of experience, respectively), who were blinded to clinical information. One of the 2 readers (S.A.) read all studies twice with a 1-month interval in order to evaluate the intra-rater reliability. We categorized the type of tendon tears into 3 groups: no-tears, partial-thickness tears, and full-thickness tears, according to the Ellman classification for partial-thickness tears and the Cofield grading system for full-thickness tears (
      • Ellman H.
      Diagnosis and treatment of incomplete rotator cuff tears.
      ,
      • Cofield RH.
      Subscapular muscle transposition for repair of chronic rotator cuff tears.
      ). We also assessed the presence or absence of CTA using Hamada's classification (
      • Hamada K
      • Fukuda H
      • Mikasa M
      • et al.
      Roentgenographic findings in massive rotator cuff tears. a long-term observation.
      ,
      • Philips VK.
      Cuff tear arthropathy.
      ).

      Quantification of Fatty Degeneration

      The degree of fatty degeneration was quantified using 2-point Dixon MR images (
      • Dixon WT.
      Simple proton spectroscopic imaging.
      ) on a PACS system (Centricity, version 4.0, GE Healthcare, Chicago, IL, USA). A region-of-interest (ROI) was placed manually over the supraspinatus and infraspinatus muscles at the most lateral section of the scapula Y-view appearance in the oblique sagittal plane (Fig. 2). When muscular atrophy was noted, the peri-muscular fat tissue was excluded from the ROI, according to a prior study (
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Quantification of fatty degeneration within the supraspinatus muscle by using a 2-point dixon method on 3-T MRI.
      ). The signal intensities of fat images and in-phase images obtained using the 2-point Dixon sequence were denoted as S(Fat) and S(In). The amount of fat contained in each rotator cuff muscle was calculated with the following equation: fat fraction = S(Fat)/S(In). The rate of change in the fat fraction (ΔFfr) in the supraspinatus and infraspinatus muscles, respectively, was then calculated using the formula as “fat fraction of follow-up MRI/fat fraction of initial MRI.”
      Fig 2
      Fig. 2Seventy-seven-year-old female with a full-thickness tear of the superior rotator cuff. (a) MRI shows discontinuity of the superior rotator cuff and fluid retention in the subacromial bursa (arrow) on an oblique coronal fat-saturated proton density-weighted image at the initial examination. (b) Oblique sagittal proton density-weighted MR image shows the muscle region and segmented area (dotted lines) at the initial examination. (c. d) Measurements of signal intensity within the regions-of-interest over the supraspinatus and infraspinatus muscles were performed on oblique sagittal images on (c) an in-phase image and on (d) a fat image acquired with the 2-point Dixon sequence. Each signal intensity value was represented as S(In) and S(Fat), and we calculated the fat fraction in supraspinatus and infraspinatus muscles with the formula S(Fat)/S(In). Fat fraction was calculated as 0.087 in the supraspinatus muscle and as 0.170 in the infraspinatus muscle. (e) After 25 months of follow-up, MRI shows progression of the superior rotator cuff tear and fluid retention in the subacromial bursa (arrowheads) on an oblique coronal fat-saturated proton-density weighted image. (f) Oblique sagittal proton density-weighted MRI shows progression of atrophy and fatty degeneration of rotator cuff muscles, particularly in the supraspinatus muscle. (g, h) Fat fraction within the rotator cuff muscles increased to 0.31 in the supraspinatus muscle and 0.25 in the infraspinatus muscle. The ΔFfr was 0.31 / 0.087 = 3.56 in the supraspinatus and 0.25/0.17 = 1.47 in the infraspinatus muscles.

      Statistical Analysis

      The Kruskal-Wallis test was used for a comparison of the difference in the ΔFfr among the 3 groups. The Steel‒Dwass method was used for the post hoc analysis. The Mann-Whitney U test was used to compare the ΔFfr between the “full-thickness tear” group and the “no full-thickness tear” (i.e., partial-thickness tear group + no-tear group) group. We also evaluated the relationship between CTA and the ΔFfr during the course of this study using the Mann-Whitney U test. We also conducted a sub-analysis by comparing the difference in the ΔFfr between the mild partial-thickness tear (≤50%) and severe partial-thickness tear (>50%) groups using the Mann-Whitney U test. The level of significance for all calculations was defined at p < 0.05. Intra- and inter-class correlation coefficients (ICCs) were calculated to evaluate the reproducibility of the fat fraction values. All statistical analyses were performed by using R for Windows software, version 3.0.2 (R Development Core Team, Vienna, Austria).

      RESULTS

      The patients were classified into 3 groups according to the condition of the superior rotator cuff on MR images at the initial examination. Their characteristics are described in Table 1. The “no-tear” group at the initial examination included 2 patients with a frozen shoulder, 2 with calcific tendinitis, 5 with rotator cuff tendinopathy, and 1 with no damage. There was no significant difference in the mean age among the 3 groups (p = 0.75).
      Table 1Demographic Data and the ΔFfr In the Supraspinatus and Infraspinatus Muscles Among the Full-Thickness Tear, Partial-Thickness Tear, And No-Tear Groups
      CharacteristicsFull-Thickness Tear Group (n = 7)Partial-Thickness Tear Group (n = 13)No-Tear Group (n = 10)p-value
      Age (y), mean ± SD (range)66.9 ± 9.065.1 ± 7.563.3 ± 12.20.75
      (58-82)(53-75)(45-80)
      Sex (no. of patients)
       Male392
       Female448
      Follow-up MRI interval (m), mean ± SD (range)34.0 ± 22.230.8 ± 18.425.8 ± 12.7
      (15-74)(12-77)(12-48)
       More than 2 years (no. of patients)585
       More than 1 year and less than 2 years (no. of patients)255
      ΔFfr in the supraspinatus muscle, mean ± SD
      Total2.54 ± 1.301.02 ± 0.360.75 ± 0.15<0.001
      p < 0.05.
       Male1.49 ± 0.471.00 ± 0.390.85 ± 0.07
       Female3.33 ± 1.141.06 ± 0.340.72 ± 0.16
       Follow-up MRI interval was more than 2 years2.64 ± 2.221.29 ± 0.320.72 ± 0.09
       Follow-up MRI interval was more than 1 year and less than 2 years2.50 ± 1.130.85 ± 0.290.77 ± 0.20
      ΔFfr in the infraspinatus muscle, mean ± SD
      Total1.96 ± 1.711.07 ± 0.410.73 ± 0.190.021
      p < 0.05.
       Male0.97 ± 0.161.13 ± 0.110.74 ± 0.29
       Female2.71 ± 2.031.04 ± 0.520.72 ± 0.19
       Follow-up MRI interval was more than 2 years3.36 ± 3.361.13 ± 0.110.71 ± 0.25
       Follow-up MRI interval was more than 1 year and less than 2 years1.41 ± 0.451.04 ± 0.520.75 ± 0.15
      low asterisk p < 0.05.
      We also classified the patients into 2 groups: 7 patients in the “full-thickness tear” group and 23 patients in the “no full-thickness tear” group, where we combined the no-tear and partial-thickness tear groups (Table 2). There was also no significant difference in the mean age between the full-thickness tear group and no full-thickness tear group (p = 0.58).
      Table 2Demographic Data and the ΔFfr In the Supraspinatus and Infraspinatus Muscles Between the Full-Thickness Tear and No Full-Thickness Tear Groups
      CharacteristicsFull-Thickness Tear Group (n = 7)No Full-Thickness Tear Group (n = 23)p-value
      Age (y), mean ± SD (range)66.9 ± 9.0(58-82)64.3 ± 9.5(45-80)0.58
      Sex (no. of patients)
       Male311
       Female412
      Follow-up MRI interval (m), mean ± SD (range)34.0 ± 22.2(15-74)28.6 ± 16.0(12-77)
       More than 2 years (no. of patients)513
       More than 1 year and less than 2 years (no. of patients)210
      ΔFfr in the supraspinatus muscle, mean ± SD2.54 ± 1.300.90 ± 0.32<0.001
      p < 0.05.
      ΔFfr in the infraspinatus muscle, mean ± SD1.96 ± 1.700.92 ± 0.370.013
      p < 0.05.
      low asterisk p < 0.05.
      Of 13 patients in the partial-thickness tear group, 8 had articular-sided partial-thickness tears, 3 had bursal-sided partial-thickness tears, and 2 had interstitial partial-thickness tears. We classified these patients into 2 groups according to the depth of the tear: 8 patients were classified into the mild partial-thickness tear group with tear depth equal to or less than 50% (which was equal to Ellmann grade 1 or 2), and 5 patients were classified into the severe partial-thickness tear group with tear depth greater than 50% (which was equal to Ellmann grade 3) (Table 3). No significant difference was found in the mean age between the 2 groups (p = 0.61).
      Table 3Demographic Data and the ΔFfr In the Supraspinatus and Infraspinatus Muscles Between the Partial-Thickness Tear Group with Tear Depth Greater Than 50% And Partial-Thickness Tear Group with Tear Depth Equal to Or Less Than 50%
      CharacteristicsPartial-Thickness Tear (>50%) group (n = 5)Partial-Thickness Tear Group (≤50%) (n = 8)p-value
      Age (y), mean ± SD (range)64.0 ± 8.3(54-74)65.8 ± 7.4(53-75)0.61
      Sex (no. of patients)
       Male27
       Female31
      Follow-up MRI interval (m), mean ± SD (range)28.6 ± 28.0(12-77)32.4 ± 11.5(21-50)
       More than 2 years (no. of patients)26
       More than 1 year and less than 2 years (no. of patients)32
      ΔFfr in the supraspinatus muscle, mean ± SD1.29 ± 0.40.84 ± 0.180.09
      ΔFfr in the infraspinatus muscle, mean ± SD1.02 ± 0.31.11 ± 0.490.94
      *p < 0.05.
      Statistically significant differences in the ΔFfr in the supraspinatus and infraspinatus muscles were found among the full-thickness, partial-thickness, and no-tear groups (p < 0.001 and p < 0.01, respectively) (Fig. 3 and Table 1). The ΔFfr in the supraspinatus and infraspinatus muscles were also significantly higher in the full-thickness tear group than in the no full-thickness tear group (p < 0.001 and p = 0.013) (Fig. 4 and Table 2). Namely, the fatty degeneration progressed significantly in the full-thickness tear group. Regarding the depth of partial-thickness tears, the ΔFfr of the supraspinatus muscle tended to be slightly higher in the severe partial-thickness tear group than that in the mild partial-thickness tear group, although it was not statistically significant (Table 3).
      Fig 3
      Fig. 3The ΔFfr in the (a) supraspinatus and (b) infraspinatus muscles among full-thickness, partial-thickness, and no-tear groups. A box-plot graph shows a comparison of the ΔFfr in the supraspinatus and infraspinatus muscles among full-thickness, partial-thickness, and no-tear groups. There was a statistically significant difference among 3 groups (2.54 vs. 1.02 vs. 0.75, p < 0.001 and 1.96 vs 1.07 vs 0.73, p = 0.021, respectively).
      Fig 4
      Fig. 4The ΔFfr in the (a) supraspinatus and (b) infraspinatus muscles between the full-thickness and no full-thickness tear (i.e., partial-thickness tear group + no-tear group) groups. A box-plot graph shows a comparison of the ΔFfr in the supraspinatus and infraspinatus muscles between the full-thickness and no full-thickness tear groups. The ΔFfr in the supraspinatus and infraspinatus muscles was significantly higher in the full-thickness tear group than in the no full-thickness tear group with statistically significant difference (2.54 vs. 0.90, p < 0.001 and 1.96 vs 0.92, p = 0.013, respectively).
      In the full-thickness tear group, the ΔFfr tended to be higher in females than in males for both the supra- and infraspinatus muscles, although there was no statistically significant difference (3.33 vs 1.49 in the supraspinatus muscle, p = 0.114 and 2.71 vs 0.97 in the infraspinatus muscle, p = 0.057).
      Figure 5 shows the correlation between the follow-up MRI interval (x) and the ΔFfr (y). Only the full-thickness tear group showed a positive slope of a linear-fit trendline in both the supraspinatus and infraspinatus muscles (supraspinatus slope: 0.002724, infraspinatus slope: 0.001823) (Fig. 5a, d). On the other hand, the gradient of the linear approximation was close to zero in both the partial-thickness and no-tear groups (Fig. 5b, c, e, and f).
      Fig 5
      Fig. 5Relationship between the follow-up MRI interval (x) and fat fraction of full-thickness, partial-thickness, and no-tear groups in the supraspinatus and infraspinatus muscles (y). (a, d) Only the full-thickness tear group showed a positive slope for the linear-fit trendline between the follow-up interval and the ΔFfr in both the supraspinatus and the infraspinatus muscles (supraspinatus slope: 0.002724, infraspinatus slope: 0.001823). Five patients showed progression of CTA during the follow-up MRI intervals (dashed lines), and 2 of these patients underwent reverse shoulder arthroplasty afterwards (Bold dashed lines). (b,c,e,f) The gradient of the linear approximation was close to zero in both the partial-thickness and no-tear groups (Color version of figure is available online.)
      Of 30 patients, 5 patients showed progression of CTA during the follow-up MRI intervals (dashed lines in Fig. 5a and d). According to the Hamada classification, 2 patients had progressed from grade 1 to 2, 1 had progressed from grade 1 to 4, 1 had progressed from grade 2 to 4, and 1 had progressed from grade 3 to 5. In all of these cases, they had full-thickness tears at the initial MR examination, and 2 of these patients underwent reverse shoulder arthroplasty afterwards (Bold dashed lines in Fig. 5a, d and Figure 6). 71.4% of the full-thickness tear group showed progression of CTA, and 28.6% of the full-thickness tear group needed reverse shoulder arthroplasty with an average follow-up period of 34 months. The ΔFfr in the supraspinatus and infraspinatus muscles were significantly higher in patients with progression to CTA than in patients without progression (2.95 ± 1.31 vs 0.95 ± 0.36 in the supraspinatus muscle, p < 0.001 and 2.39 ± 1.89 vs 0.92 ± 0.36 in the infraspinatus muscle, p < 0.01). Among the full-thickness tear group, the ΔFfr tended to be higher in the patient group with progression to CTA than in the patient group without progression (2.95 ± 1.31 vs. 1.53 ± 0.66 in the supraspinatus muscle, p = 0.38 and 2.39 ± 1.89 vs. 0.90 ± 0.13 in the infraspinatus muscle, p = 0.09), although there were no statistically significant differences.
      Fig 6
      Fig. 6Progression to CTA in an 82-year-old female. (a, b) Oblique coronal and sagittal proton-density weighted images, respectively, shows full-thickness rotator cuff tear (arrow) and mild atrophy and fatty degeneration of the supraspinatus and infraspinatus muscles (dotted lines) at the initial examination. (c) Radiography shows sufficient joint space in the glenohumeral joint. (d) After 15 months of follow-up, chondral loss (arrows) and osteophyte formation (arrowhead) are shown in humeral head and glenoid fossa on oblique coronal proton density-weighted image. (e) Oblique sagittal proton density-weighted MR image shows significant atrophy and fatty degeneration of the supraspinatus and infraspinatus muscles (dotted lines). The ΔFfr was 0.57 / 0.13 = 4.2 in the supraspinatus and 0.54 / 0.094 = 5.7 in the infraspinatus muscle. (f) Radiography shows uplift and collapse of the humeral head (arrow), suggestive of progression to CTA. (g) This patient underwent a reverse shoulder arthroplasty.
      One of 10 patients without a rotator cuff tear had initially developed a full-thickness tear corresponding to a “small tear” in Cofield classification during the course of this study. One patient without a rotator cuff tear at the initial MR examination had progression of osteoarthritic change, suggestive of primary osteoarthritis of the shoulder.
      Regarding the reproducibility of the quantitative values of fatty degeneration in this study, the intra-observer correlation coefficient was 0.964 (95% confidence interval: 0.915, 0.985) and the inter-observer correlation coefficient was 0.928 (95% confidence interval: 0.798, 0.972). Both these values indicted excellent reproducibility.

      DISCUSSION

      This study quantitatively investigated the changes in fatty degeneration of the rotator cuff muscles over time in patients who received conservative treatment, rather than surgery, for shoulder pain. We found statistically significant differences in ΔFfr within the supraspinatus and infraspinatus muscles in individuals with full-thickness, partial-thickness, or no tears in the rotator cuff. Almost three-quarters of the full-thickness tear group showed progression of CTA within an average follow-up period of 34 months.
      Previous studies have reported a relationship between the progression of rotator cuff muscle fatty degeneration and the severity of rotator cuff tears (
      • Melis B
      • Wall B
      • Walch G.
      Natural history of infraspinatus fatty infiltration in rotator cuff tears.
      ,
      • Melis B
      • DeFranco MJ
      • Chuinard C
      • et al.
      Natural history of fatty infiltration and atrophy of the supraspinatus muscle in rotator cuff tears.
      ,
      • Hebert-Davies J
      • Teefey SA
      • Steger-May K
      • et al.
      Progression of fatty muscle degeneration in atraumatic rotator cuff tears.
      ). In our quantitative study, we found a significant increase in the fat fraction of the rotator cuff muscles over time in the full-thickness tear group as compared with the no full-thickness tear group. The correlation between the follow-up MRI interval and the ΔFfr also showed a positive slope for the linear fit trendline in the full-thickness tear group, whereas the slopes of the linear approximation in the partial-thickness and no-tear groups were very close to zero. In other words, a full-thickness tear of the rotator cuff can be considered as a risk factor for increased fatty degeneration of the rotator cuff muscle. While, the rotator cuff fat fraction changed little over time in the no-tear group and in the partial-thickness tear group. However, regarding the degree of partial-thickness tear, the rate of change of the supraspinatus muscle in fatty degeneration tended to be slightly higher in more severe partial-thickness tears, although there were only a few cases. It is, therefore, necessary to confirm these findings with a larger cohort, although these preliminary findings may help surgeons intervene earlier to prevent cuff repair failures.
      In this study, there was a trend for a higher rate of change in the fat fraction in females than in males in the full-thickness tear group. Previous reports have shown that sex is an important predictor of supraspinatus fatty infiltration (
      • Barry JJ
      • Lansdown DA
      • Cheung S
      • et al.
      The relationship between tear severity, fatty infiltration, and muscle atrophy in the supraspinatus.
      ) and that fatty degeneration may progress faster in female patients than in male patients (
      • Nozaki T
      • Tasaki A
      • Horiuchi S
      • et al.
      Quantification of fatty degeneration within the supraspinatus muscle by using a 2-point dixon method on 3-T MRI.
      ). The present results also indicate that sex is a factor that accelerates fat fraction.
      Several reports have qualitatively examined the natural history of rotator cuff muscle fatty infiltration (
      • Melis B
      • Wall B
      • Walch G.
      Natural history of infraspinatus fatty infiltration in rotator cuff tears.
      ,
      • Melis B
      • DeFranco MJ
      • Chuinard C
      • et al.
      Natural history of fatty infiltration and atrophy of the supraspinatus muscle in rotator cuff tears.
      ), with moderate fatty infiltration (Goutallier stage 2) occurring from the onset of non-traumatic shoulder pain at an average of 54.1 months in the supraspinatus muscle and 56.4 months in the infraspinatus muscle. These reports described a relationship between the number of torn tendons and the rate of progression of fatty infiltration, but the relationship between the depth of the rotator cuff tear and the rate of fatty infiltration has not yet been elucidated. In the present study, both supra- and infraspinatus muscles showed faster progression of fatty degeneration in the full-thickness tear group than previously reported, whereas the progression of fatty degeneration in the no full-thickness tear group was markedly slower than those described in the prior reports. On the other hand, there was no significant difference in the ΔFfr between groups with a follow-up MRI interval of less than or more than 2 years. It has been reported that fatty degeneration progresses within 1 year after surgery in cases of re-tear after rotator cuff repair (
      • Gerber C
      • Schneeberger AG
      • Hoppeler H
      • et al.
      Correlation of atrophy and fatty infiltration on strength and integrity of rotator cuff repairs: a study in thirteen patients.
      ,
      • Gladstone JN
      • Bishop JY
      • Lo IK
      • et al.
      Fatty infiltration and atrophy of the rotator cuff do not improve after rotator cuff repair and correlate with poor functional outcome.
      ), and it is likely that some patients with full-thickness rotator cuff tears develop severe fatty degeneration of the rotator cuff muscles in the early post-tear period. In our study, the fact that there was no significant difference in the ΔFfr even with an extended follow-up MRI interval may indicate that the rate of increase in fat fraction is not constant, but rather increases at a relatively rapid rate in the early period after rotator cuff tear and then slows down afterwards.
      We also showed quantitatively that fatty degeneration of the rotator cuff muscles progressed significantly over time in the group of patients with CTA. Reverse shoulder arthroplasty is the current surgical treatment of choice for advanced CTA, but its narrow indication and high complication rate, and highly invasive nature, are still problematic (
      • Eajazi A
      • Kussman S
      • LeBedis C
      • et al.
      Rotator cuff tear arthropathy: pathophysiology, imaging characteristics, and treatment options.
      ). Quantification of the fat fraction with an understanding of the natural course of fatty degeneration of rotator cuff muscles may simultaneously predict the progression to CTA, and lead to the selection of early treatment for rotator cuff tears in order to avoid the progression to irreversible arthropathy.
      Our study had several limitations. First, we included only a small number of patients. Case accumulation of these patients is difficult because many of them do not have long-term visits for follow-up when surgery is not an option. In particular, a larger sample size is needed to evaluate the full-thickness tear group that shows progression of fatty degeneration over time, and also the severe partial-thickness tear group. Second, because this was a retrospective study, the follow-up MRI interval was not uniform. The present study suggested that the ΔFfr may not be constant, but observation at many time points at shorter intervals is needed to investigate this issue more precisely.

      CONCLUSION

      We have shown changes in fatty degeneration within the rotator cuff muscles over time in patients with shoulder pain using quantitative MRI and found that fatty degeneration within the rotator cuff is significantly and quantitatively more progressive in the full-thickness tear group than in the no full-thickness tear group. We also provided quantitative support for previous findings that conservative observation in full-thickness tears of rotator cuff is likely to lead to progressive fatty degeneration of the rotator cuff, reducing the adaptability to primary rotator cuff repair and increasing the risk of CTA. Understanding the natural history of fatty degeneration of the rotator cuff muscles revealed in this study may be important for estimating changes in shoulder joint function after rotator cuff tear and deciding on the appropriate timing of therapeutic intervention.

      References

        • Williams Jr., GR
        • Rockwood Jr., CA
        • Bigliani LU
        • et al.
        Rotator cuff tears: why do we repair them?.
        J Bone Joint Surg Am. 2004; 86: 2764-2776
        • Goutallier D
        • Postel JM
        • Zilber S
        • et al.
        Shoulder surgery: from cuff repair to joint replacement. an update.
        Joint Bone Spine. 2003; 70: 422-432
        • Goutallier D
        • Postel JM
        • Bernageau J
        • et al.
        Fatty muscle degeneration in cuff ruptures. pre- and postoperative evaluation by CT scan.
        Clin Orthop Relat Res. 1994; 304: 78-83
        • Gilbert F
        • Meffert RH
        • Schmalzl J
        • et al.
        Grade of retraction and tendon thickness correlates with MR-spectroscopically measured amount of fatty degeneration in full thickness supraspinatus tears.
        BMC Musculoskelet Disord. 2018; 19: 197
        • Matsumura N
        • Oguro S
        • Okuda S
        • et al.
        Quantitative assessment of fatty infiltration and muscle volume of the rotator cuff muscles using 3-dimensional 2-point Dixon magnetic resonance imaging.
        J Shoulder Elbow Surg. 2017; 26: e309-e318
        • Davis DL
        • Kesler T
        • Gilotra MN
        • et al.
        Quantification of shoulder muscle intramuscular fatty infiltration on T1-weighted MRI: a viable alternative to the Goutallier classification system.
        Skeletal Radiol. 2019; 48: 535-541
        • Iijima Y
        • Matsuki K
        • Hoshika S
        • et al.
        Relationship between postoperative retear and preoperative fatty degeneration in large and massive rotator cuff tears: quantitative analysis using T2 mapping.
        J Shoulder Elbow Surg. 2019; 28: 1562-1567
        • Nardo L
        • Karampinos DC
        • Lansdown DA
        • et al.
        Quantitative assessment of fat infiltration in the rotator cuff muscles using water-fat MRI.
        J Magn Reson Imaging. 2014; 39: 1178-1185
        • Nozaki T
        • Tasaki A
        • Horiuchi S
        • et al.
        Predicting retear after repair of full-thickness rotator cuff tear: two-point dixon mr imaging quantification of fatty muscle degeneration-initial experience with 1-year follow-up.
        Radiology. 2016; 280: 500-509
        • Horiuchi S
        • Nozaki T
        • Tasaki A
        • et al.
        Reliability of MR quantification of rotator cuff muscle fatty degeneration using a 2-point dixon technique in comparison with the goutallier classification: validation study by multiple readers.
        Acad Radiol. 2017; 24: 1343-1351
        • Nozaki T
        • Tasaki A
        • Horiuchi S
        • et al.
        Quantification of fatty degeneration within the supraspinatus muscle by using a 2-point dixon method on 3-T MRI.
        AJR Am J Roentgenol. 2015; 205: 116-122
        • Agten CA
        • Rosskopf AB
        • Gerber C
        • et al.
        Quantification of early fatty infiltration of the rotator cuff muscles: comparison of multi-echo Dixon with single-voxel MR spectroscopy.
        Eur Radiol. 2016; 26: 3719-3727
        • Lee S
        • Lucas RM
        • Lansdown DA
        • et al.
        Magnetic resonance rotator cuff fat fraction and its relationship with tendon tear severity and subject characteristics.
        J Shoulder Elbow Surg. 2015; 24: 1442-1451
        • Eajazi A
        • Kussman S
        • LeBedis C
        • et al.
        Rotator cuff tear arthropathy: pathophysiology, imaging characteristics, and treatment options.
        AJR Am J Roentgenol. 2015; 205: W502-W511
        • Rugg CM
        • Gallo RA
        • Craig EV
        • et al.
        The pathogenesis and management of cuff tear arthropathy.
        J Shoulder Elbow Surg. 2018; 27: 2271-2283
        • Melis B
        • Wall B
        • Walch G.
        Natural history of infraspinatus fatty infiltration in rotator cuff tears.
        J Shoulder Elbow Surg. 2010; 19: 757-763
        • Melis B
        • DeFranco MJ
        • Chuinard C
        • et al.
        Natural history of fatty infiltration and atrophy of the supraspinatus muscle in rotator cuff tears.
        Clin Orthop Relat Res. 2010; 468: 1498-1505
        • Hebert-Davies J
        • Teefey SA
        • Steger-May K
        • et al.
        Progression of fatty muscle degeneration in atraumatic rotator cuff tears.
        J Bone Joint Surg Am. 2017; 99: 832-839
        • Ellman H.
        Diagnosis and treatment of incomplete rotator cuff tears.
        Clin Orthop Relat Res. 1990; : 64-74
        • Cofield RH.
        Subscapular muscle transposition for repair of chronic rotator cuff tears.
        Surg Gynecol Obstet. 1982; 154: 667-672
        • Hamada K
        • Fukuda H
        • Mikasa M
        • et al.
        Roentgenographic findings in massive rotator cuff tears. a long-term observation.
        Clin Orthop Relat Res. 1990; : 92-96
        • Philips VK.
        Cuff tear arthropathy.
        J Bone Joint Surg Am. 1984; 66: 807-808
        • Dixon WT.
        Simple proton spectroscopic imaging.
        Radiology. 1984; 153: 189-194
        • Barry JJ
        • Lansdown DA
        • Cheung S
        • et al.
        The relationship between tear severity, fatty infiltration, and muscle atrophy in the supraspinatus.
        J Shoulder Elbow Surg. 2013; 22: 18-25
        • Gerber C
        • Schneeberger AG
        • Hoppeler H
        • et al.
        Correlation of atrophy and fatty infiltration on strength and integrity of rotator cuff repairs: a study in thirteen patients.
        J Shoulder Elbow Surg. 2007; 16: 691-696
        • Gladstone JN
        • Bishop JY
        • Lo IK
        • et al.
        Fatty infiltration and atrophy of the rotator cuff do not improve after rotator cuff repair and correlate with poor functional outcome.
        Am J Sports Med. 2007; 35: 719-728