Pediatric Radiology

, Volume 43, Issue 10, pp 1327–1335

Analysis of fatty infiltration and inflammation of the pelvic and thigh muscles in boys with Duchenne muscular dystrophy (DMD): grading of disease involvement on MR imaging and correlation with clinical assessments

Authors

    • Department of RadiologyCincinnati Children’s Hospital Medical Center
  • Arnold C. Merrow
    • Department of RadiologyCincinnati Children’s Hospital Medical Center
  • Sahar Shiraj
    • Department of RadiologyCincinnati Children’s Hospital Medical Center
  • Brenda L. Wong
    • Department of NeurologyCincinnati Children’s Hospital Medical Center
  • Paul S. Horn
    • Department of NeurologyCincinnati Children’s Hospital Medical Center
    • Department of Mathematical SciencesUniversity of Cincinnati
  • Tal Laor
    • Department of RadiologyCincinnati Children’s Hospital Medical Center
Original Article

DOI: 10.1007/s00247-013-2696-z

Cite this article as:
Kim, H.K., Merrow, A.C., Shiraj, S. et al. Pediatr Radiol (2013) 43: 1327. doi:10.1007/s00247-013-2696-z

Abstract

Background

Prior reports focus primarily on muscle fatty infiltration in Duchenne muscular dystrophy (DMD). However, the significance of muscle edema is uncertain.

Objective

To evaluate the frequency and degree of muscle fat and edema, and correlate these with clinical function.

Materials and methods

Forty-two boys (ages 5–19 years) with DMD underwent pelvic MRI. Axial T1- and fat-suppressed T2-weighted images were evaluated to grade muscle fatty infiltration (0–4) and edema (0–3), respectively. Degree and frequency of disease involvement were compared to clinical evaluations.

Results

Gluteus maximus had the greatest mean fatty infiltration score, followed by adductor magnus and gluteus medius muscles, and had the most frequent and greatest degree of fatty infiltration. Gluteus maximus also had the greatest mean edema score, followed by vastus lateralis and gluteus medius muscles. These muscles had the most frequent edema, although the greatest degree of edema was seen in other muscles. There was correlation between cumulative scores of fatty infiltration and all clinical evaluations (P < 0.05).

Conclusion

In DMD, the muscles with the most frequent fatty infiltration had the greatest degree of fatty infiltration and correlated with patient function. However, the muscles with the most frequent edema were different from those with the greatest degree of edema. Thus, edema may not predict patient functional status.

Keywords

MRIMuscleDuchenne muscular dystrophyChildren

Introduction

Duchenne muscular dystrophy (DMD) is one of the most common fatal genetic diseases. It is an X-linked recessive muscular dystrophy in children affecting one in 4,700 boys [1]. DMD is characterized by progressive muscle weakness and subsequent early death from respiratory and/or cardiac failure during the 3rd decade. It results from mutation of the dystrophin gene, which leads to deficient and fragile muscle membranes. These membranes are vulnerable to injury from eccentric muscle contractions [2, 3], resulting in transient edema and subsequent irreversible fatty infiltration and fibrosis [3, 4]. Early treatment with steroids can slow the disease progression; however, the disease remains incurable [5].

Pelvic and lower extremity muscle weakness in early childhood is usually the first clinical manifestation of DMD. An elevation in serum creatine kinase (CK) accompanies the clinical weakness. Advanced genetic analyses have enabled early diagnosis of DMD, including in utero detection of the disease [6]. Prior MRI studies have shown a characteristic disease distribution in the pelvic and thigh muscles of patients with DMD [79]. These findings included fatty infiltration involving the gluteal and adductor muscles with sparing of the gracilis, sartorius, rectus femoris and semi-tendinosus muscles. Muscle edema most commonly involved the quadriceps and adductor muscles [79]. The pelvic and thigh muscles were also more severely involved than lower leg muscles. Boys with DMD were also found to have 27% less muscle volume than normal controls [10]. However, disease involvement in DMD is unevenly distributed in a given patient and has varying degrees of disease severity within each muscle [11, 12]. A few studies have examined the severity of fatty infiltration of the pelvic and thigh muscles in DMD and demonstrated a correlation of the degree of muscular fatty infiltration with muscle weakness [11, 12].

To our knowledge, this is a unique study that grades edema of the pelvic and thigh muscles in boys with DMD. Furthermore, it is uncertain whether edema of the muscle, which precedes and/or is associated with fatty infiltration, is correlated to muscle weakness. Therefore, we have evaluated the distribution, frequency and degree of disease involvement of the pelvic and thigh muscles in boys with DMD by grading fatty infiltration and edema in each muscle. The muscles with the greatest degree of fatty infiltration and edema were determined. The cumulative scores of fatty infiltration and edema of all pelvic and thigh muscles were compared to clinical evaluations of motor function.

Materials and methods

This study was approved by the Institutional Review Board and written consent was obtained from all participants and/or their parents or guardians. This study was compliant with the Health Insurance Portability and Accountability Act.

Boys with DMD in whom the diagnosis of DMD was confirmed by means of genetic analysis and/or muscle biopsy were recruited from July 2010 to September 2011. Forty-two boys ranging in age from 5 to 19 years (median: 8.96, mean: 9.73 ± 2.62) were included. This patient cohort group is different from that previously evaluated [11]. All 42 boys with DMD underwent clinical evaluation and MRI examinations. All clinical evaluations, including timed functional tests and measurements of serum creatine kinase (CK), were performed within 3 days of the MR imaging.

Clinical evaluations

All children were recruited by a pediatric neurologist with special expertise in neuromuscular diseases (one of the coauthors with pediatric neurology fellowship training and 14 years’ experience). Physical characteristics of patients including age, height (cm) and weight (kg) were measured and recorded. Body surface area (BSA) and body mass index (BMI) were calculated using the DuBois formula and standard formula respectively (\( \mathrm{BSA}\;\left[ {{{\mathrm{m}}^2}} \right] = 0.007184 \times \mathrm{height}\ {{\left[ {\mathrm{cm}} \right]}^{0.725 }} \times \mathrm{weight}\;{{\left[ {\mathrm{kg}} \right]}^{0.425 }} \); \( \mathrm{BMI} = {{{\mathrm{weight}\;\left[ {\mathrm{kg}} \right]}} \left/ {{{{{\left[ {\mathrm{height}\;\left( \mathrm{m} \right)} \right]}}^2}}} \right.} \)).

Clinical assessments were comprised of the clinical functional scale (CFS) as well as timed functional tests that included the timed Gower score (seconds), time to run 30 feet (9.144 m) (seconds), and time to climb up four steps (seconds). CFS was graded from 1 (normal) to 8 (severe dysfunction) using the motor functional ability scale by Swinyard et al. [13]. The timed Gower score was the time the patient needed to rise from a sitting position on the floor to a standing position. Timed Gower’s, 30-feet run, and climb up four steps tests were performed only in patients whose CFS was less than grade 3; ambulatory subjects, as the boys with significant muscle weakness (grade 3 or higher; non-ambulatory subjects) could not perform the muscle strength tests. Serum CK also was recorded for each patient.

MR imaging

Axial T1-weighted images without fat suppression and axial fluid-sensitive sequence (T2-W images with fat suppression) were obtained using a 3.0-T MR scanner (Achieva 3T X-series; Philips Healthcare, Best, The Netherlands) equipped with a 32-channel cardiac/torso coil. The position of the coil was such that the top part of the coil was approximately at the level of the iliac crest, covering the pelvic to mid-thigh muscles. No intravascular contrast medium or sedation was used.

MR sequence parameters were as follows: axial T1-W images without fat suppression (repetition time [ms]/echo time [ms], 320–460/10; matrix 320 × 249 to 460 × 357; echo train length, 8; slice thickness, 5 mm; and field of view, 32–46 cm), and axial fast spin-echo T2-W images with fat suppression (echo train length, 19; 3,000 or 4,045/85; matrix, 260 × 197 to 400 × 311; slice thickness, 5 mm; and field of view, 30–46 cm). To achieve maximal homogeneous fat suppression on T2-W images, spectral attenuated inversion recovery (SPAIR) fat suppression with pencil-beam volume shimming was applied. SPAIR fat suppression was chosen because this technique has very low sensitivity to radiofrequency field inhomogeneity and is also independent of B1 field [14]. We scanned an area spanning from the iliac crest superiorly to the mid-thigh inferiorly. We selected five to six images in approximately the same locations in each child. These locations included: (1) the level of sciatic notch; (2) the level of the hip joint where the femoral head appears spherical; (3) the level of the greater trochanter-ischial tuberosity; (4) the level of the most proximal femoral diaphysis, (5) 5 cm below #4 and 6) 10 cm below #4. The total acquisition time for T1- and T2-W images with fat-suppression sequences was approximately 10 min (4.5 min for T1-W images and 5 min for fat-suppressed T2-W).

MR image interpretation

MR images were interpreted by two radiologists (one of the coauthors with pediatric radiology fellowship training and 8 years’ experience, and the leading author with both pediatric and musculoskeletal radiology fellowship training and 9 years’ experience) by consensus. The radiologists were blinded to any clinical information at the time of image review. Eighteen muscles in the right side of the pelvic girdle and right thigh were evaluated. The sections of the slices that were chosen for interpretation were the ones that contained the largest area of visible muscle (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00247-013-2696-z/MediaObjects/247_2013_2696_Fig1_HTML.gif
Fig. 1

Images in an 11-year-old boy with Duchenne muscular dystrophy (DMD). Grading of fatty infiltration and edema was performed on T1-weighted images (left) and T2-weighted images with fat suppression (right), respectively. The grading of each muscle was done on the single slice, which contained the largest area of visible muscle. Grading of muscle fatty infiltration and edema is provided on the figure after each muscle abbreviation as follows: Section (a) IP iliopsoas; (b) GMa gluteus maximus, GMi gluteus minimus, GMe gluteus medius, OI obturator internus; (c) TF tensor fascia, Pe pectineus; (d) AB adductor brevis, AL adductor longus; (e) Sa sartorius, RF rectus femoris, VL vastus lateralis, VI vastus intermedius, VM vastus medialis, St semitendinosus, Sm semimembranosus, BF biceps femoris, AM adductor magnus, Gr gracilis

Grading of fatty infiltration

On axial T1-W, grading of fatty infiltration of the right pelvic and thigh muscles was performed using a scale originally described by Mercuri et al. [15] and modified by Kim et al. [11]. Grades on a scale from 0 to 4 are as follows: grade 0: homogeneous low T1 muscle signal with no fatty infiltration, grade 1: minimal scattered high T1 fat signal intensity, grade 2: mild fatty infiltration with areas of patchy fatty infiltration involving <30% of muscle, grade 3: moderate fatty infiltration with 30%–60% muscle involvement and preserved interface between the muscle and subcutaneous fat, and grade 4: severe fatty infiltration of >60% of the muscle with loss of interfaces between the muscle and subcutaneous fat.

Grading of edema

On axial T2-W with fat suppression, grading of muscle edema of the right pelvic and thigh muscles was performed using a scale originally described by Carlo et al. [16] and modified by Kim et al. [11]. Grades on a scale from 0 to 3 are as follows: grade 0, no high T2 fluid signal intensity in the muscle; grade 1, minimal interfascicular edema; grade 2, minimal inter- and intrafascicular edema, and grade 3, moderate inter- and intrafascicular edema.

The mean grading score values of fatty infiltration and edema were obtained for each of the 18 pelvic and thigh muscles in each child to determine the distribution of the involved muscles. The frequency of fatty infiltration and edema (involvement with grade ≥1) was documented for each muscle as well. The muscles with the greatest degree of fatty infiltration and edema were identified. The cumulative fat score (sum of the fatty infiltration grading scores) and cumulative edema score (sum of the edema grading scores) for all of the pelvic and thigh muscles were calculated and compared to clinical evaluations for each child.

Statistical analysis

The frequency of fatty infiltration and edema of each muscle were compared using Student’s t-test and Wilcoxon rank sum test. The Spearman correlation coefficients model was utilized to evaluate the correlation between clinical evaluations (age, height, weight, BSA, BMI, CFS, timed Gower score, time to run 30 feet, time to climb up four steps and serum CK) and cumulative scores of fat and edema of the pelvic and thigh muscles from MR images. Positive and negative correlations were evaluated and considered statistically significant if the P-value was less than 0.05. All statistical analyses were performed by using statistical software (SAS, version 9.1; SAS Institute, Cary, NC).

Results

Fatty infiltration

The gluteus maximus muscle had the greatest mean grading score, followed by the adductor magnus and gluteus medius muscles (Fig. 2). These three muscles were most frequently involved and showed at least minimal fatty infiltration (≥grade 1) in all 42 boys (100%) (Fig. 3). The greatest degree of fatty infiltration was generally seen in those muscles with the highest frequency of involvement: there was grade 4 fatty infiltration of the gluteus maximus muscle in 9 boys (21.4%), adductor magus muscle in 6 boys (14.3%), and gluteus medius muscle in 5 boys (11.9%) (Fig. 4). The gracilis muscle had the lowest mean grading score of fatty infiltration, followed by the iliacus and sartorius muscles (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00247-013-2696-z/MediaObjects/247_2013_2696_Fig2_HTML.gif
Fig. 2

Bar graph shows the mean grading score of fatty infiltration in 18 pelvic and thigh muscles. The gluteus maximus muscle has the greatest mean grading score, followed by the adductor magnus and gluteus medius muscles

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-013-2696-z/MediaObjects/247_2013_2696_Fig3_HTML.gif
Fig. 3

Bar graph showing the frequency of involvement of fatty infiltration and edema for each of the pelvic and thigh muscles in all patients. Overall, the frequency of fatty infiltration was significantly higher than the incidence of edema

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-013-2696-z/MediaObjects/247_2013_2696_Fig4_HTML.gif
Fig. 4

Bar graph showing the cumulative grading scores of fatty infiltration in the 18 pelvic and thigh muscles

Edema

The gluteus maximus muscle had the greatest mean grading score of edema, followed by the vastus lateralis and gluteus medius muscles (Fig. 5). The frequency of involvement with at least minimal edema (≥grade 1) was highest in those three muscles: edema was noted in the gluteus maximus muscle in 33 boys (78.6%), vastus lateralis muscle in 24 (57.1%) and the gluteus medius muscle in 22 (52.4%) (Fig. 3). However, the greatest degree of edema (grade 3) was noted in other muscles, namely the gluteus minimus muscle (n = 1), tensor fascia muscle (n = 1), adductor longus muscle (n = 1), vastus lateralis muscle (n = 1) and vastus intermedius muscle (n = 1) (Fig. 6). The adductor brevis muscle had the lowest mean grading score of edema, followed by the adductor longus muscle (Fig. 5). No reactive soft tissue edema was seen around the edematous muscles.
https://static-content.springer.com/image/art%3A10.1007%2Fs00247-013-2696-z/MediaObjects/247_2013_2696_Fig5_HTML.gif
Fig. 5

Bar graph showing the mean grading score of edema in 18 pelvic and thigh muscles. The gluteus maximus muscle has the greatest mean grading score, followed by the vastus lateralis and gluteus medius muscles

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-013-2696-z/MediaObjects/247_2013_2696_Fig6_HTML.gif
Fig. 6

Bar graph showing the cumulative grading scores of edema in the 18 pelvic and thigh muscles

In summary, the gluteus maximus muscle had the greatest mean score of fatty infiltration and edema; the gluteus medius had the third greatest mean score of fatty infiltration and edema. The other 16 muscles had varying amounts of fatty infiltration and edema that did not appear to correlate with each other (Figs. 2, 5). Overall, the frequency of involvement of fatty infiltration was significantly higher than the frequency of involvement of edema of the same muscles (P < 0.0001) (Fig. 3).

Clinical evaluations of all the children and normal range for each assessment are summarized in Table 1.
Table 1

Summary of clinical evaluations and correlation with cumulative scores of fatty infiltration and inflammation

Clinical evaluations

Age (year)

Clinical functional scale (1–7) (n = 42)

Full timed Gower’s score (s) (n = 30)

30-feet run time (s) (n = 35)

Climb up four steps (s) (n = 34)

Serum creatine kinase (n = 41)

Body weight (kg) (n = 42)

Height (cm) (n = 42)

Body surface area (n = 42)

Body mass index (n = 42)

Normal range

 

Grade 1

<2 s

Depends on age < 3 s

1 s

60–400 IU/L

    

Mean

9.73

Grade1: 19

4.00

4.32

2.63

11,649.65

33.39

126.01

 

21.16

Grade2 : 15

Standard deviation

±2.62

Grade 3: 1

3.98

1.56

1.74

8,299.31

12.77

13.61

 

6.44

Grade 4: 2

Minimum

4.99

Grade 5: 3

1.20

2.50

0.90

1,314.00

16.70

105.80

 

13.71

Grade 6: 2

Maximum

18.87

Grade 7: 0

21.7

9.22

8.70

49,120.00

80.60

163.00

 

54.70

Grade 8: 0

Median

8.96

 

2.50

3.80

2.10

10,740.00

29.00

123.15

 

19.70

Spearman correlation coefficients between clinical assessments and cumulative scores (P-values)

Cumulative scores of fatty infiltration

0.73* (P < 0.0001)

0.67* (P < 0.0001)

0.66* (P < 0.0001)

0.62* (P < 0.0001)

0.57* (P < 0.05)

−0.40* (P < 0.05)

0.71* (P < 0.0001)

0.66* (P < 0.0001)

0.73* (P < 0.0001)

0.31* (P < 0.05)

Cumulative scores of inflammation

0.23 (P = 0.2198)

0.17 (P = 0.2632)

0.27 (P = 0.1299)

0.14 (P = 0.1430)

0.15 (P = 0.3848)

−0.10 (P = 0.5130)

0.21 (P = 0.1608)

0.14 (P = 0.3639)

0.18 (P = 0.2255)

0.23 (P = 0.1239)

*Statistically significant

Statistically significant correlation was seen between the cumulative scores of fatty infiltration and all clinical evaluations; positive correlation with age, height, weight, BSA, BMI, CFS, timed Gower score, time to run 30 feet, and time to climb up four steps (P < 0.05) and negative correlation with serum CK (P < 0.05). No statistically significant correlation was seen between cumulative scores of edema and any of the clinical evaluations (Table 1).

Discussion

Two MRI patterns of abnormal signal are recognized as the main manifestations of skeletal muscle disease: fatty infiltration (resulting in increased T1 signal intensity within muscle) and edema (resulting in increased T2 signal intensity within muscle on water-sensitive images) [17]. With technical advancements, objective measurements of fatty infiltration and edema of the muscles have been made possible. T2 relaxation time mapping [11, 18], Dixon imaging [19] and MRI spectroscopy [20] have allowed measurement of very small amounts of fat and edema in the muscles that were not appreciable using conventional T1- or T2-W images [20].

In our study, the frequency of edema in children with DMD was significantly lower than the frequency of fatty infiltration. Furthermore, edema of the muscle did not correlate with clinical evaluations. Hence, it is unlikely that edema has a significant effect on muscle weakness. Fatty infiltration is well-recognized to be a marker of disease progression in neuromuscular disorders, and is an independent factor related to patient muscle weakness and morbidity [19]. However, fatty infiltration is irreversible and most therapeutic trials are aimed at ameliorating muscle edema, a potentially reversible stage of disease activity. Therefore, the significance of muscle edema should not be underestimated in DMD [5]. MRI is currently the only described noninvasive way to determine the presence of muscle edema.

Prior studies have demonstrated that neuromuscular disorders have a characteristic distribution of fatty infiltration and edema [79]. However, the degree of edema of each muscle has not been described. A prior study in children with DMD showed that the gluteal and adductor muscles were most commonly affected by fatty infiltration and that the quadriceps and adductor muscles most commonly demonstrated edema [79]. However, disease involvement in neuromuscular disorders is heterogeneous in a given patient and the degree of disease involvement differs between each muscle group and compartment. Even within the same muscle group, disease severity may vary between muscles [11] and may even differ in each muscle fascicle at the microscopic level.

In measuring disease severity in DMD, advanced MR techniques may be useful because they can measure fatty infiltration and edema objectively and assign a numeric value to the degree of each disease manifestation. However, to apply advanced MRI techniques practically for quantitative analysis of disease involvement in a clinical setting, it is important to determine the muscles that are most frequently and most severely involved. In our current study, the gluteus maximus muscle was the most frequently involved muscle and also the muscle that demonstrated the greatest degree of fatty infiltration, as was documented in a prior study [11]. Similarly, the adductor magnus and gluteus medius muscles had the second- and third-highest frequencies of involvement and greater degrees of fatty infiltration. We have found that, in DMD, the fatty infiltration follows a general pattern in which the most frequently involved muscle is also the most severely affected on MR images. However, edema does not follow this pattern. The gluteus maximus, vastus lateralis and gluteus medius muscles were most frequently involved with edema, but the greatest degree of edema was noted in other muscles (namely, the gluteus minimus, tensor fascia, adductor longus, vastus lateralis and vastus intermedius muscles). Therefore, to determine disease severity, the degree of fatty infiltration should be assessed in one of the three frequently affected muscles (the gluteus maximus, adductor magnus or gluteus medius muscles). However, the degree of edema should be assessed on a case-by-case basis according to the imaging findings in each child.

In our study, the gluteus maximus and gluteus medius muscles demonstrated both fatty infiltration and edema in the same order of frequency (the greatest and third-highest frequencies of fatty infiltration and edema, respectively). Such concurrence within the same muscles is not surprising since muscle edema leads to fatty infiltration as shown in prior histological studies [21, 22].

It could be argued that some of the high T2 signal on T2-W images with fat suppression may have resulted from muscle fat due to inhomogeneous fat suppression and lack of use of inversion recovery techniques. Inversion recovery is very sensitive for the evaluation of edema, but the signal intensity of normal muscle is usually higher on inversion recovery images and subtle edema may be more difficult to detect [23]. Therefore, to optimize homogeneous fat suppression in our study, we used the SPAIR technique as it has a very low sensitivity to radiofrequency field homogeneity [14].

The mean grading scores for fatty infiltration and edema varied in all of the muscles other than the gluteus maximus and medius. For example, the adductor magnus muscle had the second-highest mean grading score for fatty infiltration but the 12th-highest mean grading score for edema. The opposite was also seen: The semitendinosus muscle had a relatively lower grade of fatty infiltration than other muscles (the 13th-highest mean grading score for fatty infiltration) but a relatively higher grade of edema than other muscles (the 5th-highest mean grading score for edema). These results again support the conclusion that the high T2 signal intensity in the muscle is a result of edema rather than inhomogeneous fat suppression.

Our study has some limitations. First, we did not include sequences with intravenous contrast, which are the most sensitive sequences for the detection of skeletal muscle damage and edema as proven in animal studies but not in humans [24, 25]. Second, the sample size of our study (n = 42) is a small number cohort to definitely determine which muscle has the most severe edema, as this process is seen in a relatively small number of patients.

It is not surprising that there is different muscle group involvement between edema and fatty infiltration. It may be that the various muscle groups are imaged during different time points of disease activity. It is as yet unknown whether edema directly precedes fatty infiltration, and whether the muscle signal can normalize following edema. Furthermore, we did not stratify the boys with DMD according to their activity level, such as involvement in sports or level of ambulation, which could contribute to muscle edema. We also did not evaluate muscle volume; because muscle expansion accompanies the fatty changes in DMD, this would be an important variable to examine in future studies. Quantitative MR techniques, such as T2 relaxation mapping, can detect muscle inflammation before it manifests as signal abnormality on fluid-sensitive sequences; therefore, future studies using quantitative MR techniques to evaluate muscle edema might be useful for early detection of DMD while the disease is reversible. Finally, longitudinal studies that correlate disease progression with MRI changes will be necessary.

Conclusion

Fatty infiltration of the muscles in DMD follows a general pattern, with the most frequently involved muscles generally having the most severe degree of fatty infiltration. Edema of the muscles did not follow this pattern, however, as the most frequently involved muscles were different from the most severely involved muscles. Fatty infiltration was more frequent than edema of the muscles and correlates with muscle weakness in boys with DMD.

Acknowledgments

This research project was supported by RSNA seed grant (The Fujifilm Medical System RSNA Research Seed Grant 2010–2011) and SPR seed grant 2009–2010.

Conflicts of interest

None

Copyright information

© Springer-Verlag Berlin Heidelberg 2013