Assessment of Fat Content in Supraspinatus Muscle with Proton MR ...

Radiology

Musculoskeletal Imaging Christian W. A. Pfirrmann, MD Marius R. Schmid, MD Marco Zanetti, MD Bernhard Jost, MD Christian Gerber, MD Juerg Hodler, MD

Index terms: Fat Magnetic resonance, (MR) spectroscopy, 414.12145 Muscles, atrophy, 41.4813, 41.564 Muscles, MR spectroscopy, 41.12145 Shoulder, injuries, 414.4813 Shoulder, MR, 414.4813 Published online 10.1148/radiol.2323030442 Radiology 2004; 232:709 –715 Abbreviation: CI ⫽ confidence interval 1

From the Departments of Radiology (C.W.A.P., M.R.S., M.Z., J.H.) and Orthopedic Surgery (B.J., C.G.), University Hospital, Balgrist, Forchstrasse 340, CH8008 Zurich, Switzerland. From the 2002 RSNA scientific assembly. Received March 22, 2003; revision requested June 13; final revision received January 11, 2004; accepted February 4. Supported in part by the VISION foundation. Address correspondence to C.W.A.P. (e-mail: christian @pfirrmann.ch). Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, C.W.A.P.; study concepts, C.W.A.P., C.G.; study design, C.W.A.P., J.H., M.Z.; literature research, C.W.A.P., M.R.S.; clinical studies, M.R.S., B.J.; data acquisition, M.R.S., B.J.; data analysis/interpretation, C.W.A.P., M.R.S., M.Z., C.G.; statistical analysis, C.W.A.P., J.H.; manuscript preparation, C.W.A.P., B.J.; manuscript definition of intellectual content, C.W.A.P., M.R.S., C.G.; manuscript editing, J.H., C.W.A.P.; manuscript revision/ review, J.H., M.Z.; manuscript final version approval, J.H., M.Z., C.G. ©

RSNA, 2004

Assessment of Fat Content in Supraspinatus Muscle with Proton MR Spectroscopy in Asymptomatic Volunteers and Patients with Supraspinatus Tendon Lesions1 PURPOSE: To evaluate proton magnetic resonance (MR) spectroscopy in the assessment of lipid content of the supraspinatus muscle in asymptomatic volunteers and patients with supraspinatus tendon lesions. MATERIALS AND METHODS: Single-voxel proton MR spectroscopy was used to assess lipid content of the supraspinatus muscle in asymptomatic volunteers (n ⫽ 30) and patients with partial-thickness supraspinatus tendon tears (n ⫽ 30), newly diagnosed full-thickness supraspinatus tendon tears (n ⫽ 30), and chronic full-thickness supraspinatus tendon tears (n ⫽ 30). The apparent lipid content of the supraspinatus muscle measured with proton MR spectroscopy was related to its appearance on sagittal-oblique T1-weighted spin-echo MR images (grades 0 – 4). One-way analysis of variance was performed to test for significant differences, and the Tukey honestly significant difference procedure was performed for post hoc comparisons. RESULTS: Mean apparent lipid content was 13.7% (95% confidence interval [CI]: 11.5%, 15.8%) for asymptomatic volunteers, 29.5% (95% CI: 25.1%, 34.0%) for patients with partial-thickness tears, 48.6% (95% CI: 41.3%, 55.9%) for patients with full-thickness tears, and 66.1% (95% CI: 57.7%, 74.5%) for patients with chronic tears. Values were significantly different (analysis of variance, P ⬍ .001; P ⬍ .001–.002 for all post hoc pairwise comparisons). Mean apparent lipid content for the supraspinatus muscle was as follows: grade 0, 19.6% (95% CI: 16.7%, 22.6%); grade 1, 36.8% (95% CI: 33.2%, 40.4%); grade 2, 53.6% (95% CI: 43.1%, 64.2%); grade 3, 67.5% (95% CI: 52.6%, 82.3%); and grade 4, 79.2% (95% CI: 73.2%, 85.3%). With analysis of variance (P ⬍ .001), all post hoc pairwise comparisons were significant (P ⱕ .001) except between grades 2 and 3 (P ⫽ .112) and between grades 3 and 4 (P ⫽ .261). In 14 (25%) subjects who had grade 0 appearance on T1-weighted images, lipid content values were greater than the upper range of values in the volunteers. CONCLUSION: Proton MR spectroscopy is suitable in the assessment of apparent lipid content of rotator cuff muscles. ©

RSNA, 2004

Fatty atrophy of the rotator cuff muscle occurs as early as 6 weeks after the initial rotator cuff tear (1). The degree of fatty atrophy is an important predictor of the success of surgical reconstruction (2–5). Replacement of more than half of the muscle cross section with fat has been considered to be a relative contraindication for surgical reconstruction (6). It seems preferable to treat large tears with surgery before irreversible muscular damage occurs (1). Therefore, preoperative assessment of fatty atrophy of the rotator cuff muscles with imaging examinations is important. 709

Radiology Figure 1. Illustration of voxel positioning. A 10 ⫻ 10 ⫻ 10-mm voxel was positioned in the center of the supraspinatus muscle at the site of the largest diameter. Position of voxel is demonstrated on (a) a sagittal-oblique T1-weighted spin-echo MR image (600/12), (b) a coronal-oblique intermediate-weighted fast spin-echo MR image with fat saturation (3300/14), and (c) a transverse T1-weighted spin-echo MR image (600/12).

Computed tomography (CT) and magnetic resonance (MR) imaging have both been used to assess the degree of fatty atrophy of the rotator cuff muscles (1,7– 9). CT correlates poorly with MR imaging for quantification of fatty atrophy of the rotator cuff muscles (9). Both methods allow a semiquantitative assessment of fatty atrophy only. Proton MR spectroscopy can be used to quantify the lipid content of muscle tissue noninvasively (10 –14). The accuracy of proton MR spectroscopy in the measurement of the intramuscular fat content appears to be comparable with that of biochemical measurements (12). The purpose of our study was to evaluate proton MR spectroscopy in the assessment of supraspinatus muscle fat content in asymptomatic volunteers and patients with supraspinatus tendon lesions.

MATERIALS AND METHODS Volunteers and Patients Thirty asymptomatic volunteers (21 men with a mean age of 27.0 years and an age range of 21–39 years and nine women with a mean age of 30.8 years and an age range of 21– 45 years) were included in the study if they (a) had no shoulder pain, (b) had no history of shoulder trauma, (c) did not participate in overhead throwing sports, nor were they high-performance athletes, (d) had never seen a doctor for shoulder problems, and (e) never had to stop work because of shoulder problems. Thirty consecutive clinical patients with a partial-thickness tear of the supraspinatus tendon (15 men with a mean 710

䡠

Radiology

䡠

September 2004

age of 49.9 years and an age range of 33– 82 years and 15 women with a mean age of 50.3 years and an age range of 29 – 63 years) and 30 consecutive clinical patients with a full-thickness supraspinatus tendon tear (17 men with a mean age of 61.8 years and an age range of 40 –79 years and 13 women with a mean age of 68.8 years and an age range of 50 – 80 years) were also included in the study. Both groups consisted of patients with a partial- or full-thickness tear of the supraspinatus tendon that had been recently diagnosed with MR arthrography. Thirty consecutive study patients with a chronic full-thickness supraspinatus tear (18 men with a mean age of 68.1 years and an age range of 53– 83 years and 12 women with a mean age of 72.5 years and an age range of 58 – 87 years) were included, as well. All patients with a chronic supraspinatus tendon tear were evaluated at least 2 years after a fullthickness rotator cuff tear was initially diagnosed with MR imaging. All clinical patients were referred from one outpatient shoulder clinic of a university hospital for assessment of the integrity of the rotator cuff with MR arthrography. The patients with chronic tears were referred for MR imaging for a different clinical research project. The study protocol was approved by the institutional review board. Informed consent was obtained from all patients.

MR Imaging and Proton Spectroscopy MR imaging and proton spectroscopy were performed with a 1.5-T imager (Symphony; Siemens Medical Solutions,

Erlangen, Germany). A shoulder array coil was used for MR imaging and proton MR spectroscopy. All asymptomatic volunteers and study patients with chronic supraspinatus tears underwent routine MR imaging before they underwent spectroscopy of the supraspinatus muscle to allow us to assess the integrity of the supraspinatus tendon. In the coronal-oblique plane, T2-weighted and intermediate-weighted fast spin-echo MR images were obtained with fat saturation 3300/ 95, 14 (repetition time msec/echo time msec); a section thickness of 4 mm; a field of view of 160 ⫻ 100 mm; and a matrix of 256 ⫻ 512. In the transverse plane, T2-weighted and intermediateweighted fast spin-echo MR images were obtained with fat saturation (3300/95, 14; 4-mm section thickness; 160 ⫻ 160-mm field of view; 256 ⫻ 512 matrix). In the sagittal-oblique plane, T1-weighted spinecho MR images were obtained (600/12, 4-mm section thickness, 160 ⫻ 100-mm field of view, 512 ⫻ 320 matrix). All clinical patients (30 with newly diagnosed partial-thickness tears and 30 with newly diagnosed full-thickness tears) underwent MR arthrography of the shoulder after injection of 12 mL of gadoteridol (ProHance; Bracco Diagnostics, Princeton, NJ) solution with a concentration of 2 mmol/L. MR arthrography was used in the diagnosis of partial-thickness tears because it performed significantly better than standard MR imaging in the diagnosis the rotator cuff (15). MR arthrography is the standard imaging modality used in the evaluation of the rotator cuff at our institution. Informed patient consent was obtained before MR arthrograPfirrmann et al

TABLE 1 Intrasubject Reproducibility of MR Spectroscopy Measurements

Radiology

Mean Values of Measurement Parameter

First Measurement

Second Measurement

Mean Difference

Mean Difference (%)

Apparent lipid content* Peak ratio† Area ratio†

12.0 0.1 0.1

12.0 0.1 0.1

1.1 (1.1) 0.01 (0.01) 0.02 (0.02)

9.2 (7.2) 9.7 (8.7) 10.6 (8.6)

Image Analysis and Analysis of Spectroscopic Data

Note.—Data in parentheses are standard deviations. * Data are percentages of the apparent lipid content. † Data are spectroscopy signal ratios.

TABLE 2 Comparison of the Integrity of the Supraspinatus Tendon with the Apparent Lipid Integrity of the Supraspinatus Tendon Apparent lipid content* Healthy Partial-thickness tear Full-thickness tear Chronic tear Peak ratio† Healthy Partial-thickness tear Full-thickness tear Chronic tear Area ratio† Healthy Partial-thickness tear Full-thickness tear Chronic tear

No. of Patients

Mean

95% CI

30 30 30 30

13.7 (5.8) 29.5 (11.9) 48.6 (19.6) 66.1 (22.5)

11.5, 15.8 25.1, 34.0 41.3, 55.9 57.7, 74.5

30 30 30 30

0.1 (0.0) 0.3 (0.2) 1.5 (3.3) 4.2 (6.0)

0.09, 0.12 0.2, 0.4 0.3, 2.8 2.0, 6.5

30 30 30 30

0.2 (0.1) 0.5 (0.3) 1.9 (3.7) 5.1 (7.1)

0.1, 0.2 0.4, 0.6 0.5, 3.3 2.4, 7.7

Note.—Data in parentheses are standard deviations. CI ⫽ confidence interval. * Data are percentages of the apparent lipid content, unless otherwise indicated. † Data are spectroscopy signal ratios, unless otherwise indicated.

phy. For the MR arthrography protocol, the coronal-oblique T2-weighted and intermediate-weighted fast spin-echo images with fat saturation and the sagittal-oblique T1weighted spin-echo images were identical to those obtained with the standard MR protocol described previously. In addition, coronal-oblique T1-weighted spinecho images with fat saturation (777/20, 3-mm section thickness, 160 ⫻ 100-mm field of view, 512 ⫻ 320 matrix) and transverse T1-weighted spin-echo images (600/12, 3-mm section thickness, 160 ⫻ 160-mm field of view, 512 ⫻ 512 matrix) were obtained. A full-thickness tendon tear was diagnosed if the intraarticular contrast material communicated through a transmural defect with the subdeltoid bursa. A partial-thickness tendon tear was diagnosed in patients with an articular- or bursal-sided partial defect of the tendon without communication of contrast material from the articular space to the bursal space (15). Single-voxel spin-echo proton MR specVolume 232

䡠

Number 3

ibility was assessed by investigating 10 consecutive volunteers as follows: After the first measurement was obtained, the volunteer was removed from the imager and asked to sit upright; the volunteer was then repositioned in the imager to perform a second spectroscopic examination.

troscopy without water saturation of the supraspinatus muscle was performed (1500/135, 128 signals averaged). Imaging time was 3 minutes 12 seconds. The same dedicated receive-only shoulder coil that was used for routine MR imaging of the shoulder was used for proton MR spectroscopy. A voxel of 10 ⫻ 10 ⫻ 10 mm was positioned in the center of the supraspinatus muscle at the site of the largest diameter (Fig 1). A coronaloblique MR image and a sagittal-oblique MR image from the preceding routine MR examination of the shoulder were used for appropriate positioning of the voxel. The automated local three-dimensional shimming procedure and proton MR spectroscopic sequence were launched with a single command. Proton MR spectroscopy could be performed in all volunteers and patients with spectroscopic data suitable for quantitative evaluation. The imaging time, including the automated shimming procedure, was 4 minutes 20 seconds. Intrasubject reproduc-

After Fourier transformation of the spectroscopic data, postprocessing included Gauss filtering, a zero-order phase correction, and a baseline correction. For curve fitting, the proprietary software of the spectroscopic package of the MR imager was used. A Gaussian curve-fitting algorithm was applied. The peak height and area of the water and lipid peaks were calculated. The lipid-water ratio was calculated for the peak height (peak ratio) and area (area ratio) (11,16). The fat content is given as relative signal (Sfat, area of the fat peak) in percent of the total signal (Sfat, area of fat and Swater, area of the water peak) in the following equation: Sfat/(Sfat ⫹ Swater) ⫻ 100% (17,18). Ratios rather than direct spectroscopic signal measurements were used because the absolute amplitude of a signal peak varies from patient to patient and depends on coil and shoulder geometry. Fatty atrophy of the supraspinatus muscle was graded on the T1-weighted sagittal-oblique MR images by two musculoskeletal radiologists (M.R.S., M.Z.) in consensus without knowledge of spectroscopic and clinical data. Their experience in MR imaging of the musculoskeletal system was 6 and 12 years, respectively, and both had routinely used the following grading system during clinical work: grade 0 ⫽ no intramuscular fat; grade 1 ⫽ some fatty streaks present; grade 2 ⫽ fat is evident, but there was less fat than muscle tissue; grade 3 ⫽ the amount of fat was equal to the amount of muscle tissue; and grade 4 ⫽ there was more fat than muscle tissue (1,9). One of the authors (C.W.A.P.) measured the coronal extent of the full-thickness tears of the supraspinatus tendon on the coronal-oblique T1weighted spin-echo MR images obtained with fat saturation.

Statistical Analysis Demographic data were analyzed for significant age differences between the four groups by using a one-way analysis of variance and between the sexes within

Proton MR Spectroscopy of the Supraspinatus Muscle

䡠

711

Radiology

groups by using an unpaired t test. A oneway analysis of variance was performed to test the groups for significant differences. The Tukey honestly significant difference procedure for multiple comparisons was used for post hoc comparisons. The apparent lipid content was analyzed with multiple regression, and patient category and age were used as regressors. For all statistical calculations, a dedicated software package (SPSS, version 10.0.7; SPSS, Chicago, Ill) was used. P values less than .05 were considered to indicate a statistically significant difference.

RESULTS Demographic Data No significant differences in age were found between the sexes within the four groups (P values were between .491 and .845). The mean age of patients in the four groups was significantly different, as calculated with an analysis of variance (P ⬍ .001). The post hoc comparison was significant (P ⬍ .001) for all pairs, except between both groups with full-thickness tears (P ⫽ .094). The apparent lipid content was analyzed with multiple regression analysis by using patient group and age as regressors. The regression was a good fit (R2 ⫽ 0.627), and the overall relationship was significant (F ⫽ 98.47, P ⬍ .001). With other variables held constant, apparent lipid content was positively related to patient group and age, increasing by 11.59% for every higher patient category and 0.44% for every extra year of age. The corrected effect of patient group was somewhat greater (t ⫽ 4.68, P ⬍ .001) than the corrected effect of patient age (t ⫽ 2.86, P ⫽ .005).

MR Imaging No supraspinatus tendon abnormalities were seen in the asymptomatic volunteers. Partial-thickness tears were all limited to the supraspinatus tendon. The average coronal extent of the full-thickness tears of the supraspinatus tendon in the group of clinical patients was 3.2 cm (range, 0.5– 6.0 cm). The average coronal extent of the full-thickness chronic tear of the supraspinatus tendon was 3.9 cm (range, 0.5– 6.0 cm). Almost all asymptomatic volunteers (n ⫽ 29) had grade 0 supraspinatus muscles; however, one volunteer had a grade 1 supraspinatus muscle. The supraspinatus muscles in clinical patients with partial tears were all graded as either 0 (n ⫽ 23) or 1 (n ⫽ 7). Clinical patients with full-thickness tears had a wide range of fatty atrophy of the su712

䡠

Radiology

䡠

September 2004

Figure 2. Box plots show comparison of the integrity of the supraspinatus tendon with the apparent lipid content of the supraspinatus muscle. Values of the four groups were significantly different for apparent lipid content (analysis of variance, P ⬍ .001; P ⬍ .001–.002 for all post hoc pairwise comparisons among means calculated with the Tukey honestly significant difference test). The line within the box represents the median value. Boxes represent 25th to 75th percentiles. Lines outside boxes represent 10th and 90th percentiles.

praspinatus muscle (grade 1, n ⫽ 17; grade 2, n ⫽ 6; grade 3, n ⫽ 1; grade 4, n ⫽ 6). The study group with chronic tears had the highest number of patients with very advanced fatty atrophy (grade 0, n ⫽ 3; grade 1, n ⫽ 1; grade 2, n ⫽ 6; grade 3, n ⫽ 6; grade 4, n ⫽ 14).

Applicability and Reproducibility The results of the reproducibility assessment are presented in Table 1. The mean intrasubject difference in the 10 volunteers for assessment of the apparent lipid content was 9.2%.

Comparison of the Results of Proton MR Spectroscopy with Different Degrees of Supraspinatus Tendon Lesions The results of the comparison between the different degrees of supraspinatus tendon lesions are displayed in Table 2 and Figure 2. The mean apparent lipid content was 13.7% (95% CI: 11.5%, 15.8%) for asymptomatic volunteers, 29.5% (95% CI: 25.1%, 34.0%) for patients with partial-thickness tears, 48.6% (95% CI: 41.3%, 55.9%) for patients with full-thickness tears, and 66.1% (95% CI: 57.7%, 74.5%) for patients with chronic tears. The values of the four groups were significantly different for the apparent

Figure 3. Box plots show comparison of the appearance of supraspinatus muscle tissue on T1-weighted MR images with the apparent lipid content of the supraspinatus muscle. Analysis of variance for the values of the five groups indicated a significant difference for apparent lipid content (P ⬍ .001). All post hoc pairwise comparisons among mean values with the Tukey honestly significant difference test were significant (P ⱕ .001), except between grades 2 and 3 (P ⫽ .112) and grades 3 and 4 (P ⫽ .261). The line within the box represents the median value. Boxes represent 25th to 75th percentiles. Lines outside boxes represent 10th and 90th percentiles. The x-axis represents grading of the appearance of the supraspinatus muscle tissue on T1-weighted MR images.

lipid content, which was calculated with analysis of variance (P ⬍ .001, P ⬍ .001– .002 for all post hoc pairwise comparisons among means by using the Tukey honestly significant difference test).

Comparison with Appearance of the Supraspinatus Muscle Tissue on T1-weighted MR Images The correlation between the grading of the fat content of the supraspinatus muscle on T1-weighted sagittal-oblique images and the spectroscopic data is displayed in Table 3 and Figure 3. Fifty-five of 90 patients demonstrated a normal appearance of the supraspinatus muscle (Fig 4). The supraspinatus muscle was classified as grade 1 in 26 patients, grade 2 in 12 (Fig 5), grade 3 in seven, and grade 4 in 20. The mean apparent lipid content of the supraspinatus muscle was as follows: grade 0, 19.6% (95% CI: 16.7%, 22.6%); grade 1, 36.8% (95% CI: 33.2%, 40.4%); grade 2, 53.6% (95% CI: 43.1%, 64.2%); grade 3, 67.5% (95% CI: 52.6%, 82.3%); and grade 4, 79.2% (95% CI: 73.2%, 85.3%). The analysis of variance was performed with the values of the five groups, and a significant differPfirrmann et al

Radiology Figure 4. Left image shows proton MR spectroscopy spectrum (1500/135) of the supraspinatus muscle in an asymptomatic volunteer. Note water peak (W) and lipid peak (L). The small peak between water peak and lipid peek represents the creatine peak. Right image shows corresponding sagittal-oblique T1-weighted spin-echo MR image (600/12). Supraspinatus muscle (arrowheads) demonstrates healthy muscle tissue without fatty atrophy.

Figure 5. Left image shows proton MR spectroscopy spectrum (1500/135) of the supraspinatus muscle of a patient with a fullthickness supraspinatus tendon tear. Note water peak (W) and lipid peak (L). Right image shows corresponding sagittal-oblique T1weighted spin-echo MR image (600/12). Supraspinatus muscle (arrowheads) demonstrates grade 2 fatty atrophy. Fat is evident, but there is less fat than muscle tissue.

TABLE 3 Comparison of the Appearance of the Supraspinatus Muscle Tissue on T1-weighted MR Images with the Apparent Lipid Content of the Supraspinatus Muscle Appearance of the Supraspinatus Muscle Tissue Apparent lipid content* Healthy Grade 1 Grade 2 Grade 3 Grade 4 Peak ratio† Healthy Grade 1 Grade 2 Grade 3 Grade 4 Area ratio† Healthy Grade 1 Grade 2 Grade 3 Grade 4

No. of Patients

Mean

95% CI

55 26 12 7 20

19.6 (11.1) 36.8 (9.0) 53.6 (16.6) 67.5 (16.0) 79.2 (12.9)

16.7, 22.6 33.2, 40.4 43.1, 64.2 52.6, 82.3 73.2, 85.3

55 26 12 7 20

0.2 (0.2) 0.4 (0.2) 1.2 (1.2) 2.1 (1.4) 6.8 (7.2)

0.1, 0.3 0.3, 0.5 0.4, 1.9 0.7, 3.4 3.4, 10.1

55 26 12 7 20

0.3 (0.2) 0.6 (0.2) 1.5 (1.1) 3.0 (2.2) 7.9 (8.4)

0.2, 0.3 0.5, 0.7 0.8, 2.1 0.9, 5.0 4.0, 11.8

Note.—Data in parentheses are standard deviations. For all three measurements (apparent lipid content, peak ratio, area ratio), differences among groups were statistically significant (P ⬍ .001). * Data are percentages of the apparent lipid content, unless otherwise indicated. † Data are spectroscopy signal ratios, unless otherwise indicated.

ence was found for apparent lipid content (P ⬍ .001). All post hoc pairwise comparisons among means with the Tukey honestly significant difference test were significant (P ⱕ .001), except between grades 2 and 3 (P ⫽ .112) and grades 3 and 4 (P ⫽ .261). In 14 (25%) of the 55 subjects with tears that were classified as grade 0 on T1-weighted images, the apparent lipid Volume 232

䡠

Number 3

content values were higher than values in the upper range of the normal volunteer group (Fig 6).

DISCUSSION Proton MR spectroscopy has been used to assess the lipid content in different tissues of the body and in different metabolic disorders, such as congenital lipo-

dystrophy (12) and acquired generalized lipoatrophy (19). Proton MR spectroscopy is used to analyze the lipid metabolism of muscle tissue noninvasively (14,20,21). Two pools of lipids are present in skeletal muscle; these are the intramyocellular lipids and the extramyocellular lipids. These two pools are kinetically different. The intramyocellular lipids are thought to be in a dynamic equilibrium and are used for rapid metabolism. The extramyocellular lipids turn over very slowly and serve as a long-term fat depot. The extramyocellular lipids are increased by fatty atrophy of the muscle. The anisotropic structure of the extramyocellular lipids results in a bulk magnetic susceptibility induced shift of its resonances (10,22). This shift is greatest when the muscle fibers are aligned with the constant magnetic induction field axis of the imager. This results in a maximal 0.2ppm separation between the extramyocellular lipids and the intramyocellular lipids; however, there is usually severe overlap of the intramyocellular lipids and the extramyocellular lipids, thus hampering accurate measurement (10,12). The intramyocellular and extramyocellular lipid peaks were therefore not analyzed separately in our study. The intrasubject variability for the assessment of the fat content of the supraspinatus muscle was approximately 10%. This variability is comparable with reported values of other studies of single voxel spectroscopy in muscle tissue, which vary between 6% and 20% (10,23). Slight variations of voxel positioning might account for this intrasubject vari-


䡠

713

Radiology Figure 6. (a) Proton MR spectroscopy spectrum (1500/135) of the supraspinatus muscle of a patient with partial-thickness supraspinatus tendon tear and a corresponding sagittal-oblique T1-weighted spin-echo MR image (600/12 msec). The sagittal-oblique T1-weighted MR image corresponds to the location where the proton MR spectroscopy spectrum was obtained; although no fat is visible within the supraspinatus muscle (arrowheads), MR spectroscopy reveals a markedly elevated lipid peak (L). Proton MR spectroscopy is able to depict fatty muscle changes before they become visible on T1-weighted images. W ⫽ water peak. (b) Corresponding coronal-oblique T1-weighted spin-echo MR image with fat saturation (777/20) demonstrates an articular-sided partial-thickness tear (arrowheads) of the supraspinatus tendon.

ability. Regional differences in the content of intramyocellular lipids and extramyocellular lipids have also been reported for the soleus, tibialis posterior, and tibialis anterior muscles in humans by using in vivo hydrogen 1 spectroscopic imaging (23). No significant correlation of the muscular fat with the thickness of the subcutaneous fat layer was found (24). In addition, the body mass index does not appear to affect muscular fat content, except in very obese subjects (24). Results of phantom studies have shown a high accuracy of MR spectroscopy in the assessment of the lipid content (25). The assessment of fatty atrophy of the muscle of the rotator cuff is particularly important for the prognosis of surgical treatment. The repair of massive tears of the rotator cuff does not result in substantial reversal of muscular atrophy and fatty degeneration (26). The clinical outcome of a rotator cuff repair is significantly correlated with the fatty muscle degeneration of the rotator cuff muscle (4,5). Repair within 6 months of the first symptoms improves functional outcome and is the result of less involution of muscle and tendon tissue (27). The fatty changes in the supraspinatus muscles associated with a rotator cuff tear occur mainly around the tendon fibers and the vessels and are associated with the degree of retraction of the tendon fibers (28). An important limitation of the present study is the fact that the asymptomatic subjects were considerably younger than the subjects with supraspinatus lesions. 714

䡠

Radiology

䡠

September 2004

Notably, the overall physical condition of the subjects was not controlled or known. These factors may affect the findings of this study. However, the results of a study in which supraspinatus muscle quality was assessed in asymptomatic volunteers demonstrated that age has relatively little impact on muscle volume (8). No statistically significant influence on fatty atrophy was found for sex and age of volunteers aged 21–70 years (8). The period of time after a tear was diagnosed seems to be a much stronger influence on the muscle quality of the supraspinatus muscle. In the two groups with full-thickness tears, patients with chronic tears have a significantly higher apparent lipid content, whereas the mean patient age is not significantly different. The second study limitation is the reference standard. It is ethically problematic to obtain muscle biopsy samples in volunteers and patients for histologic or chemical analysis. Surgical correlation or arthroscopic evaluation does not provide information about muscle quality. MR imaging has good sensitivity in the assessment of full-thickness tears of the rotator cuff (29,30); however, MR imaging is not as adept in the assessment of partial-thickness tears (31–33). MR arthrography, which has been used to depict partial-thickness tears, is much more accurate in this regard (15). Agreement between MR arthrography and surgery for supraspinatus tendon lesions (both partial- and full-thickness tears) was 94% with our MR arthrography protocol (34).

Results of proton MR spectroscopy depend on reproducible voxel positioning. It is known that fatty changes after a tendon tear are not uniformly distributed within the muscle tissue. Histologically, fatty changes occur mainly around the tendon fibers and vessels (28). Although correct voxel position has been defined in our article as the center of the supraspinatus muscle at the site of the largest diameter, on the coronal-oblique and sagittal-oblique MR images, slight changes in voxel position will affect spectroscopic results. We conclude that routine assessment of the apparent lipid content of rotator cuff muscles is feasible with proton MR spectroscopy and provides the possibility to quantify fatty atrophy of the supraspinatus muscle. The apparent lipid content of the supraspinatus muscle depends on the presence of rotator cuff abnormalities. Long-standing tears of the rotator cuff result in greater degrees of fatty muscle changes. Proton MR spectroscopy is able to depict fatty muscle changes before they become visible on T1-weighted MR images. References 1. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fatty muscle degeneration in cuff ruptures: pre- and postoperative evaluation by CT scan. Clin Orthop 1994; 304:78 – 83. 2. Thomazeau H, Boukobza E, Morcet N, Chaperon J, Langlais F. Prediction of rotator cuff repair results by magnetic resonance imaging. Clin Orthop 1997; 344: 275–283. 3. Nove-Josserand L, Levigne C, Noel E, Pfirrmann et al

Radiology

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Walch G. Isolated lesions of the subscapularis muscle: apropos of 21 cases. Rev Chir Orthop Reparatrice Appar Mot 1994; 80:595– 601. [French] Jost B, Pfirrmann CW, Gerber C. Clinical outcome after structural failure of rotator cuff repairs. J Bone Joint Surg Am 2000; 82:304 –314. Schaefer O, Winterer J, Lohrmann C, Laubenberger J, Reichelt A, Langer M. Magnetic resonance imaging for supraspinatus muscle atrophy after cuff repair. Clin Orthop 2002; 403:93–99. Nove´ -Josserand L, Gerber C, Walch G. Lesions of the anterio-superior rotator cuff. In: Warner JJP, Iannotti JP, Gerber C, eds. Complex and revision problems in shoulder surgery. Philadelphia, Pa: LippincottRaven, 1997; 165–176. Thomazeau H, Rolland Y, Lucas C, Duval JM, Langlais F. Atrophy of the supraspinatus belly: assessment by MRI in 55 patients with rotator cuff pathology. Acta Orthop Scand 1996; 67:264 –268. Zanetti M, Gerber C, Hodler J. Quantitative assessment of the muscles of the rotator cuff with magnetic resonance imaging. Invest Radiol 1998; 33:163–170. Fuchs B, Weishaupt D, Zanetti M, Hodler J, Gerber C. Fatty degeneration of the muscles of the rotator cuff: assessment by computed tomography versus magnetic resonance imaging. J Shoulder Elbow Surg 1999; 8:599 – 605. Boesch C, Slotboom J, Hoppeler H, Kreis R. In vivo determination of intra-myocellular lipids in human muscle by means of localized 1H-MR-spectroscopy. Magn Reson Med 1997; 37:484 – 493. Schick F, Eismann B, Jung WI, Bongers H, Bunse M, Lutz O. Comparison of localized proton NMR signals of skeletal muscle and fat tissue in vivo: two lipid compartments in muscle tissue. Magn Reson Med 1993; 29:158 –167. Szczepaniak LS, Babcock EE, Schick F, et al. Measurement of intracellular triglyceride stores by H spectroscopy: validation in vivo. Am J Physiol 1999; 276:E977– E989. Kreis R, Boesch C. Spatially localized, one- and two-dimensional NMR spectroscopy and in vivo application to human

Volume 232

䡠

Number 3

14. 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

muscle. J Magn Reson B 1996; 113:103– 118. Schick F, Machann J, Brechtel K, et al. MRI of muscular fat. Magn Reson Med 2002; 47:720 –727. Palmer WE, Brown JH, Rosenthal DI. Rotator cuff: evaluation with fat-suppressed MR arthrography. Radiology 1993; 188: 683– 687. Schick F, Seitz D, Machann J, Lutz O, Claussen CD. Magnetic resonance bone densitometry: comparison of different methods based on susceptibility. Invest Radiol 1995; 30:254 –265. Kugel H, Jung C, Schulte O, Heindel W. Age- and sex-specific differences in the 1H-spectrum of vertebral bone marrow. J Magn Reson Imaging 2001; 13:263–268. Schellinger D, Lin CS, Fertikh D, et al. Normal lumbar vertebrae: anatomic, age, and sex variance in subjects at proton MR spectroscopy—initial experience. Radiology 2000; 215:910 –916. Brechtel K, Jacob S, Machann J, et al. Acquired generalized lipoatrophy (AGL): highly selective MR lipid imaging and localized (1)H-MRS. J Magn Reson Imaging 2000; 12:306 –310. Brechtel K, Dahl DB, Machann J, et al. Fast elevation of the intramyocellular lipid content in the presence of circulating free fatty acids and hyperinsulinemia: a dynamic 1H-MRS study. Magn Reson Med 2001; 45:179 –183. Bachmann OP, Dahl DB, Brechtel K, et al. Effects of intravenous and dietary lipid challenge on intramyocellular lipid content and the relation with insulin sensitivity in humans. Diabetes 2001; 50:2579 – 2584. Kreis R, Boesch C. Liquid-crystal-like structures of human muscle demonstrated by in vivo observation of direct dipolar coupling in localized proton magnetic resonance spectroscopy. J Magn Reson B 1994; 104:189 –192. Hwang JH, Pan JW, Heydari S, Hetherington HP, Stein DT. Regional differences in intramyocellular lipids in humans observed by in vivo 1H-MR spectroscopic imaging. J Appl Physiol 2001; 90:1267– 1274. Steidle G, Machann J, Claussen CD,

25.

26. 27.

28.

29.

30.

31.

32.

33.

34.

Schick F. Separation of intra- and extramyocellular lipid signals in proton MR spectra by determination of their magnetic field distribution. J Magn Reson 2002; 154:228 –235. Kostler H, Kenn W, Hummer C, Bohm D, Hahn D. 2D-SPLASH spectroscopy to determine the fat/water ratio in the muscle of the rotator cuff. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2002; 174:991–995. [German] Gerber C, Fuchs B, Hodler J. The results of repair of massive tears of the rotator cuff. J Bone Joint Surg Am 2000; 82:505–515. Warner JJ, Higgins L, Parsons IM 4th, Dowdy P. Diagnosis and treatment of anterosuperior rotator cuff tears. J Shoulder Elbow Surg 2001; 10:37– 46. Nakagaki K, Ozaki J, Tomita Y, Tamai S. Fatty degeneration in the supraspinatus muscle after rotator cuff tear. J Shoulder Elbow Surg 1996; 5:194 –200. Quinn SF, Sheley RC, Demlow TA, Szumowski J. Rotator cuff tendon tears: evaluation with fat-suppressed MR imaging with arthroscopic correlation in 100 patients. Radiology 1995; 195:497–500. Iannotti JP, Zlatkin MB, Esterhai JL, Kressel HY, Dalinka MK, Spindler KP. Magnetic resonance imaging of the shoulder: sensitivity, specificity, and predictive value. J Bone Joint Surg Am 1991; 73:17– 29. Balich SM, Sheley RC, Brown TR, Sauser DD, Quinn SF. MR imaging of the rotator cuff tendon: interobserver agreement and analysis of interpretive errors. Radiology 1997; 204:191–194. Robertson PL, Schweitzer ME, Mitchell DG, et al. Rotator cuff disorders: interobserver and intraobserver variation in diagnosis with MR imaging. Radiology 1995; 194:831– 835. Wnorowski DC, Levinsohn EM, Chamberlain BC, McAndrew DL. Magnetic resonance imaging assessment of the rotator cuff: is it really accurate? Arthroscopy 1997; 13:710 –719. Zanetti M, Jost B, Lustenberger A, Hodler J. Clinical impact of MR arthrography of the shoulder. Acta Radiol 1999; 40:296 – 302.


䡠

715