Abstract
Background:
A recent report suggests that semi-quantitative two-dimensional Goutallier grade (2D-GG) correlates poorly with quantitative three-dimensional Dixon fat fraction (3D-Dixon-FF) on magnetic resonance imaging (MRI).
Purpose:
To determine whether the finding of poor correlation of supraspinatus 3D-Dixon-FF with 2D-GG is reproducible, and to determine the strength of the correlation of 3D-Dixon-FF with quantitative 2D Dixon fat fraction (2D-Dixon-FF).
Material and Methods:
Ten adults aged ≥60 years were recruited prospectively received shoulder MRI. 2D-Dixon-FF and 3D-Dixon-FF were measured on 6-point Dixon fat fraction maps. 2D-GG was rated on T1-weighted images.
Results:
The mean age of participants was 70.7 ± 3.7 years. The mean 3D-Dixon-FF was 8.3% ± 5.7%; the mean 2D-Dixon-FF was 7.3% ± 4.4%; and the mean 2D-GG was 0.9 ± 0.7. There was a strong correlation for 3D-Dixon-FF with 2D-Dixon-FF (rho = 0.90; P < 0.001) and with 2D-GG (rho = 0.73; P = 0.017). Excellent inter-observer reliability was found for Dixon fat fraction (intraclass correlation coefficient = 0.946), which was good for Goutallier grade (weighted kappa = 0.634).
Conclusion:
Supraspinatus 3D-Dixon-FF had a strong correlation with 2D-Dixon-FF and 2D-GG in our study population.
Keywords: Intramuscular fatty infiltration, magnetic resonance imaging, quantitative, rotator cuff, shoulder
Introduction
Rotator cuff (RC) tear is a common disorder in clinical patients, with most involving the supraspinatus tendon (1-3). Generally, orthopedic surgeons consider several factors when determining if RC repair surgery is indicated, and intramuscular fatty infiltration (FI) is a key metric for clinical decision making (4-6).
Magnetic resonance imaging (MRI) is a leading modality to diagnosis RC tear and to quantify FI (7,8). Imaging evaluation of RC muscles facilitates clinical decision making in the setting of symptomatic RC tear, since the degree of preoperative FI is associated with postoperative shoulder function, healing, and re-tear rates after RC repair surgery (6-9).
Estimation of RC FI by the semi-quantitative Goutallier classification system has been the dominant technique of image analysis for >20 years in routine clinical practice (6,9-11). Orthopedic surgeons typically consider Goutallier grade ≥3 (≥50% FI) to be a relative contraindication to RC repair surgery (8,11). Quantitative MR chemical shift imaging (CSI) techniques, such as multiecho Dixon MRI or iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL), have been gaining acceptance as a more reliable alternative to Goutallier grade for the estimation of RC FI (7,12-14).
Recently, investigators have reported that 2D single slice semi-quantitative Goutallier grade on MRI poorly correlates with quantitative volumetric three-dimensional (3D) FI content in the supraspinatus muscle (15). The aims of the present study were to determine if a poor correlation of semi-quantitative 2D Goutallier grade with quantitative volumetric 3D supraspinatus FI (3D fat fraction) could be reproduced, and to determine the strength of the correlation of quantitative 2D Dixon fat fraction (2D fat fraction) with quantitative volumetric 3D supraspinatus FI. We hypothesized that 2D Goutallier grade and 2D fat fraction would have moderate or better correlation with 3D fat fraction.
Material and Methods
Study population
The present study was approved by the authors’ institutional review board and complied with the Health Insurance Portability and Accountability Act. Each study participant provided written informed consent. This cross-sectional study included a convenience sample of 10 community-dwelling older adults (age range = 65–77 years), who received prospective evaluation from September 2020 to November 2020 at our institution’s MRI research center following recruitment by self-referral in response to local advertisement. The study sample of 10 participants was related to the level of funding support provided by the authors’ institution for this pilot study. All participants received shoulder MRI with the same imaging protocol and answered self-reported questionnaires on the same day as the MRI examination. The exclusion criteria included the following: contraindications to MRI; ipsilateral upper extremity paralysis; active worker’s compensation claim; fracture of the ipsilateral clavicle, scapula, or humerus 12 months before enrollment; a known history of malignancy at the ipsilateral shoulder; or a history of RC repair surgery or joint replacement at the ipsilateral shoulder.
MRI
Shoulder MRI was performed at 3.0 T (Magnetom Prismafit; Siemens Healthcare, Erlangen, Germany) with a four-channel flexible coil. The imaging protocol included a 3D oblique sagittal 6-point Dixon volumetric sequence; a 2D oblique sagittal 6-point Dixon fat fraction map, and 2D oblique sagittal T1-weighted (T1W), oblique sagittal short tau inversion recovery (STIR)-weighted, oblique coronal STIR-weighted, and axial STIR-weighted sequences (7).
MRI analysis
A musculoskeletal radiologist and a medical student/research assistant each independently quantified the 2D fat fraction of the supraspinatus muscle on de-identified DICOM images for each participant by manual segmentation with MIPAV software (version 10, National Institutes of Health, Bethesda, MD, USA) for the single oblique sagittal Dixon fat fraction map MR image, corresponding to the Y-shaped view MR image where the lateral-most aspect of the scapular spine contacts the scapular body (7,16). In addition, each rater independently quantified the 2D fat fraction by manual segmentation on the next nine contiguous oblique sagittal Dixon fat fraction map slices medial to the Y-shaped view MR image, corresponding to a volume of supraspinatus muscle spanning 31.5 mm mediolaterally (Figs. 1 and 2). The mean 2D fat fraction between raters was used for analysis. A second musculoskeletal radiologist and a musculoskeletal fellow each independently determined a 2D Goutallier grade for the supraspinatus muscle on de-identified DICOM images for each participant on the single oblique sagittal T1W MR image corresponding to the oblique sagittal Dixon fat fraction map Y-shaped view image using MIPAV software. The 5-point Goutallier classification system included: grade 0 = no fat; grade 1 = streaks of fat; grade 2 = muscle > fat; grade 3 = muscle = fat; and grade 4 = muscle < fat (10,11,17). The mean 2D Goutallier grades between raters was used for analysis. A musculoskeletal radiologist in a separate analysis independently reviewed each participant’s shoulder MRI to assign one of three classifications for the supraspinatus tendon: full-thickness tear, partial-thickness tear, or no tear (intact or tendinopathy).
Fig. 1.
Representative supraspinatus muscle images from a 72-year-old female study participant showing technique for 2D 6-point Dixon fat fraction and Goutallier grade analyses. (a) Oblique sagittal 6-point Dixon fat fraction map corresponding to the Y-shaped view MR image. A gold ROI is placed at the inner border of the supraspinatus muscle to determine quantitative 2D fat fraction. (b) Oblique sagittal T1-weighted Y-shaped view MR image, rated for semi-quantitative 2D Goutallier grade. 2D, two-dimensional; MR, magnetic resonance imaging; ROI, region of interest.
Fig. 2.
Representative supraspinatus muscle images from the same study participant as in Fig. 1 show the technique for 6-point Dixon fat fraction 3D volume analysis. (a–d) Example of oblique sagittal 6-point Dixon fat fraction map images medial to the Y-shaped view MR image showing placement of gold ROIs to determine mean 3D fat fraction. 3D, three-dimensional; MR, magnetic resonance; ROI, region of interest.
Clinical evaluation
Each study participant completed medical questionnaires on the same day as shoulder MRI to facilitate a description of the pilot study population, which included the Charlson Co-morbidity Index, American Shoulder and Elbow Surgeon (ASES; 0 = worst, 100 = best), and Disability of the Arm, Shoulder and Hand (DASH; 0 = best, 100 = worst) surveys (7,18). Height and weight were recorded to determine body mass index (BMI).
Statistical analysis
Statistical analysis was performed with SAS version 9.4 statistical software (Cary, NC, USA). Descriptive statistics were performed to determine the characteristics of the study population. Spearman rank order correlation (rho) with 95% confidence intervals was used to determine the correlation of mean 3D fat fraction with 2D measures of fat fraction and Goutallier grade. The interpretation of rho correlation was as follows: 0.00–0.35 = weak; 0.36–0.67 = moderate; and 0.68–1.00 = strong (19). Inter-observer reliability for Dixon fat fraction and Goutallier grade was determined by the intraclass correlation coefficient (ICC) and weighted kappa statistic, respectively. Inter-observer reliability was interpreted as follows: 0.00–0.39 = poor; 0.40–0.59 = fair; 0.60–0.74 = good; and 0.75–1.00 = excellent (20). P < 0.05 was considered statistically significant.
Results
The characteristics of the study population are shown in Table 1. Study participants were slightly obese on average with a relatively low overall co-morbidity, with the mean Charlson Co-morbidity Index mostly a function of a mean age of about 70 years. The ASES and DASH surveys showed that study participants on average had a self-perception of functional limitation at the shoulder.
Table 1.
Characteristics of the study population (n=10).
| Age (years) | 70.7 ±3.7 |
| Male (%) | 60 |
| BMI (kg/m2) | 30.1 ±6.7 |
| Charlson Co-Morbidity Index | 3.0±0.7 |
| ASES score (100 = best, 0 = worst) | 78.8 ±11.9 |
| DASH score (0 = best, 100 = worst) | 22.3 ±15.8 |
| Supraspinatus tendon status (%) | |
| No tear (%) | 50 |
| Partial-thickness tear (%) | 40 |
| Full-thickness tear (%) | 10 |
| Supraspinatus muscle 2D Y-shaped view 6-point Dixon fat fraction (%) | 7.3±4.4 |
| Supraspinatus muscle 2D Y-shaped view Goutallier grade (0 = best, 4 = worst) | 0.9±0.7 |
| Supraspinatus muscle 3D 6-point Dixon fat fraction (%) | 8.3±5.7 |
Values are given as mean±SD unless otherwise indicated.
2D, two-dimensional; 3D, three-dimensional; ASES, American Shoulder and Elbow Surgeon; BMI, body mass index; DASH, Disabilities of the Arm, Shoulder and Hand.
Most study participants lacked a supraspinatus tendon full-thickness tear, with no medial retraction of the supraspinatus muscle belly. The single participant with a full-thickness tendon tear had medial retraction of only 0.9 cm, consistent with a small tear. The average Goutallier grade was <1. The mean difference between quantitative mean 2D fat fraction and quantitative mean 3D fat fraction was 1.0%, with a maximum difference up to 5.5%.
Supraspinatus 3D fat fraction showed statistically significant strong correlation with both 2D fat fraction and 2D Goutallier grade, although correlation was superior for 2D fat fraction relative to 2D Goutallier grade (Table 2). Inter-observer reliability for 6-point Dixon fat fraction was excellent (ICC = 0.946). Good inter-observer reliability was present for Goutallier grade (weighted kappa = 0.634).
Table 2.
Correlation of supraspinatus quantitative mean 3D 6-point Dixon fat fraction with quantitative mean 2D 6-point Dixon fat fraction and semi-quantitative mean Goutallier grade.
| Quantitative 3D fat fraction |
|
|---|---|
| Quantitative 2D Dixon fat fraction | rho=0.90 |
| P < 0.001 | |
| 95% CI = 0.63–0.98 | |
| Semi-quantitative 2D Goutallier grade | rho=0.73 |
| P=0.017 | |
| 95% CI = 0.18–0.93 |
2D, two-dimensional; 3D, three-dimensional; CI, confidence interval.
Discussion
The findings of the present study suggest that the estimation of 3D supraspinatus intramuscular fatty infiltration by a single slice 2D MR image may have merit for continuation in clinical practice. Our study did not reproduce the findings of a recent study that concluded that 2D estimation of supraspinatus intramuscular fat on MRI poorly correlates with 3D intramuscular fat content in the supraspinatus muscle. The mean quantitative 2D Dixon fat fraction showed a strong correlation with mean volumetric 3D 6-point Dixon fat fraction for the supraspinatus muscle in our study population, with only a 1% difference observed between mean 2D and 3D supraspinatus fat fractions. The mean semi-quantitative 2D Goutallier grade also showed a strong correlation with mean volumetric 3D Dixon fat fraction for the supraspinatus muscle, although with inferior correlation relative to mean quantitative 2D Dixon fat fraction.
Our study showed good inter-observer reliability for Goutallier grade. The major drawback of the semi-quantitative Goutallier grade is reproducibility, which ranges from poor to good inter-observer reliability (4,7,11,12). Alternatively, quantitative MR CSI techniques offer excellent reproducibility, and our study is in line with prior studies showing that Dixon fat fraction has superior reliability relative to Goutallier grade (7,12,14,21).
Recently, the rationale to estimate RC volumetric 3D FI with single-slice 2D MR images has come into question (15). Vidt et al. studied the relationship between semi-quantitative single-slice 2D Goutallier grade at the Y-shaped view on T1W MRI and quantitative volumetric 3D fat fraction by 3-point Dixon MRI produced from contiguous axial images for the supraspinatus muscle (15). These investigators found no significant relationship between 2D Goutallier grade and 3D Dixon fat fraction (R2=0.32; P=0.116) for the supraspinatus muscle, and concluded that single-slice 2D supraspinatus FI estimates poorly correlate with volumetric 3D supraspinatus FI. However, our pilot study was unable to reproduce these findings and instead suggests a strong correlation between 2D and 3D measures of supraspinatus FI.
The strengths of our study included testing supraspinatus 2D 6-point Dixon fat fraction, as well as 2D Goutallier grade, for correlation with supraspinatus volumetric 3D 6-point Dixon fat fraction. This is in contrast to Vidt et al. who tested only for correlation between 2D Goutallier grade and volumetric 3D 3-point Dixon fat fraction (15). Our use of the 6-point Dixon sequence was also an advantage. Protocols with 6-point echo MRI CSI techniques provide accurate estimation of FI by reducing T1 effects and T2* decay through a multipeak fat spectral model (14). Grimm et al. reported that methods using three or more echoes should be used to quantify FI, since 2-point echo MRI CSI methods incompletely separate fat from water and possibly overestimate the amount of intramuscular fat in skeletal muscle (22). Of the study participants in the present study, 90% had no medial retraction of the supraspinatus muscle belly and were an ideal population to evaluate 2D FI at the Y-shaped view, since they lacked a supraspinatus full-thickness tendon tear. Only one study participant had a small full-thickness tear, measuring <1 cm of medial retraction, consistent with minimal expected impact on the estimation of supraspinatus FI at the Y-shaped view, and thus with all study participants receiving single image 2D Goutallier grade and 2D Dixon fat fraction assessment at the same approximate location within the supraspinatus muscle (23,24). This observation is one plausible factor contributing to our results differing from Vidt et al. who suggested that single image 2D Goutallier grade poorly correlates to 3D Dixon fat fraction in the supraspinatus muscle. One potential limitation in the methodology for Vidt et al. affecting the strength of correlation was that all single-image 2D Goutallier grade assessments occurred in a static location on MRI: the Y-shaped view (15). Since multiple participants in their study had supraspinatus full-thickness tendon tears without reports of tendon tear size or how medial retraction of the supraspinatus muscle related to medium, large, or massive full-thickness tendon tears was controlled for, the possibility is raised that the 2D Goutallier grade assessments may not have occurred in the same approximate intramuscular location for each participant, thus introducing potential unintended error when determining the relationship between 2D and 3D supraspinatus FI measurements.
Lastly, we tested for correlation between mean supraspinatus 2D FI and 3D FI in the oblique sagittal plane. Fuchs et al. suggested that testing for the correlation between RC FI obtained in different imaging planes may potentially reduce the strength of the correlation and favored use of the oblique sagittal plane when evaluating RC FI on MRI (11). This is a second plausible factor as to why Vidt et al. concluded that 2D Goutallier grade poorly correlates with 3D Dixon fat fraction, since the former was acquired in the oblique sagittal plane and the latter in the axial plane (15).
The present study has some limitations. These include the small sample size, although our pilot study design was able to identify statistically significant correlations to address the hypothesis. The study population is a convenience sample prospectively recruited by local advertisement, and therefore may be limited by selection bias. The study’s purpose did not include comparing between groups or testing for correlation between supraspinatus FI and clinical outcomes or study participant characteristics. Future larger studies are warranted to compare the differences between groups and to determine the relationships with study participant characteristics, utilizing 3D as well as 2D FI measures. Our study may not be generalizable to populations with a Goutallier grade ≥3, since the supraspinatus FI of study participants did not include this range. Future studies are necessary to test the reliability for 2D and 3D MRI CSI fat fraction in supraspinatus muscles with a Goutallier grade ≥3. However, patients with a Goutallier grade <3 are the ideal population for study, since this group is the most relevant in the context of clinical decision making for RC repair surgery (8,11,15). Although 2D estimation of supraspinatus FI on the Y-shaped view MR image showed a strong correlation with mean volumetric 3D supraspinatus FI in our study population, lending support to the theory that 2D images offer a feasible and time-efficient method for evaluation of FI at the RC in clinical practice, future research is warranted to establish which 2D oblique sagittal MR image(s) have the best utility to estimate FI in populations of patients with medially retracted full-thickness RC tear, since our study participants lacked significant medial retraction from full-thickness tear.
In conclusion, 2D single-image estimation strongly correlated with 3D supraspinatus intramuscular fatty infiltration in our pilot study population. This finding suggests that the 2D single-image estimation of FI may still have potential merit for use in clinical practice for shoulder MRI, but future larger research studies with more varied study populations are needed to verify this observation. Future research is also warranted to identify which oblique sagittal MR image(s) would be most beneficial for the 2D estimation of volumetric 3D FI in patients with significant medial retraction from medium, large, and massive full-thickness tears.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the University of Maryland Baltimore, Institute for Clinical & Translational Research and the National Center for Advancing Translational Sciences, National Institute on Aging, University of Maryland Claude D. Pepper Older Americans Independence Center (grant nos. 1UL1TR003098, NIA 3P30AG028747).
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References
- 1.Yamamoto A, Takagishi K, Osawa T, et al. Prevalence and risk factors of a rotator cuff tear in the general population. J Shoulder Elbow Surg 2010;19:116–120. [DOI] [PubMed] [Google Scholar]
- 2.Melis B, Nemoz C, Walch G. Muscle fatty infiltration in rotator cuff tears: descriptive analysis of 1688 cases. Orthop Traumatol Surg Res 2009;95:319–324. [DOI] [PubMed] [Google Scholar]
- 3.McElvany MD, McGoldrick E, Gee AO, et al. Rotator cuff repair: published evidence on factors associated with repair integrity and clinical outcome. Am J Sports Med 2015;43:491–500. [DOI] [PubMed] [Google Scholar]
- 4.Davis DL, Gilotra MN, Calderon R, et al. Reliability of supraspinatus intramuscular fatty infiltration estimates on T1-weighted MRI in potential candidates for rotator cuff repair surgery: full-thickness tear versus high-grade partial-thickness tear. Skeletal Radiol 2021;50:2233–2243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kweon C, Gagnier JJ, Robbins CB, et al. Surgical versus non-surgical management of rotator cuff tears: predictors of treatment allocation. Am J Sports Med 2015;43:2368–2372. [DOI] [PubMed] [Google Scholar]
- 6.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. [DOI] [PubMed] [Google Scholar]
- 7.Davis DL, Zhuo J, Almardawi R, et al. Association of patient self-reported shoulder scores to quantitative and semi-quantitative MRI measures of rotator cuff intramuscular fatty infiltration: a pilot study. AJR Am J Roentgenol 2019;213:1307–1304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ashry R, Schweitzer ME, Cunningham P, et al. Muscle atrophy as a consequence of rotator cuff tears: should we compare the muscles of the rotator cuff with those of the deltoid? Skeletal Radiol 2007;36:841–845. [DOI] [PubMed] [Google Scholar]
- 9.Deniz G, Kose O, Tugay A, et al. Fatty degeneration and atrophy of the rotator cuff muscles after arthroscopic repair: does it improve, halt or deteriorate? Arch Orthop Trauma Surg 2014;134:985–990. [DOI] [PubMed] [Google Scholar]
- 10.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. [PubMed] [Google Scholar]
- 11.Fuchs B, Weishaupt D, Zanetti M, et al. 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. [DOI] [PubMed] [Google Scholar]
- 12.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. [DOI] [PubMed] [Google Scholar]
- 13.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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.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. Euro Radiol 2016;26:3719–3727. [DOI] [PubMed] [Google Scholar]
- 15.Vidt ME, Santago AC 2nd, Tuohy CJ, et al. Assessments of fatty infiltration and muscle atrophy from a single magnetic resonance image slice are not predictive of 3-dimensional measurements. Arthroscopy 2016;32:128–139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Davis DL, Almardawi R, Henn RF, et al. Correlation of quantitative versus semiquantitative measures of supraspinatus intramuscular fatty infiltration to shoulder range of motion and strength: a pilot study. Curr Probl Diagn Radiol 2021;50:629–636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Davis DL, Gilotra MN, Hovis JP, et al. Association of rotator cuff tear patterns and intramuscular fatty infiltration on magnetic resonance imaging. J Clin Imaging Sci 2019;9:38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.McClure PW, Michener L. Measures of adult shoulder function. Arthritis Rheum 2003;49:S50–S58. [Google Scholar]
- 19.Vidt ME, Santago AC 2nd, Hegedus EJ, et al. Can self-report instruments of shoulder function capture functional differences in older adults with and without a rotator cuff tear? J Electromyogr Kinesiol 2016;29:90–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cicchetti DV. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychol Assess 1994;6:284–290. [Google Scholar]
- 21.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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Grimm A, Meyer H, Nickel MD, et al. Evaluation of 2-point, 3-point, and 6-point Dixon magnetic resonance imaging with flexible echo timing for muscle fat quantification. Eur J Radiol 2018;103:57–64. [DOI] [PubMed] [Google Scholar]
- 23.Lee YB, Yang CJ, Li CZ, et al. Can a single sagittal magnetic resonance imaging slice represent whole fatty infiltration in chronic rotator cuff tears at the supraspinatus? Clin Orthop Surg 2018;10:55–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Yoo HJ, Choi JY, Hong SH, et al. Quantifying rotator cuff atrophy and fatty degeneration at the supraspinatus origin in the scapular fossa. Knee Surg Sports Traumotol 2015;23:399–407. [DOI] [PubMed] [Google Scholar]