Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Oct 1.
Published in final edited form as: Neuromuscul Disord. 2016 Jul 28;26(10):650–658. doi: 10.1016/j.nmd.2016.07.013

Quantifying disease activity in fatty-infiltrated skeletal muscle by IDEAL-CPMG in Duchenne muscular dystrophy

Ami Mankodi 1, Courtney A Bishop 2, Sungyoung Auh 1, Rexford D Newbould 2, Kenneth H Fischbeck 1, Robert Janiczek 3
PMCID: PMC5062949  NIHMSID: NIHMS814544  PMID: 27593185

Abstract

Purpose

To explore the use of iterative decomposition of water and fat with echo asymmetry and least-squares estimation Carr-Purcell-Meiboom-Gill (IDEAL-CPMG) to simultaneously measure skeletal muscle apparent fat fraction (AFF) and water T2 (T2,w) in patients with Duchenne muscular dystrophy (DMD).

Materials and Methods

In twenty healthy volunteer (HV) boys and thirteen subjects with DMD, thigh muscle AFF was measured by Dixon and IDEAL-CPMG, with the IDEAL-CPMG also providing T2,w as a measure of muscle inflammatory activity. A subset of subjects with DMD was followed up during a 48-week clinical study. The study was in compliance with the Patient Privacy Act and approved by the Institutional Review Board.

Results

AFF in the thigh muscles of subjects with DMD was significantly increased compared to HV boys (p < 0.001). Dixon and IDEAL-CPMG AFF strongly correlated (r = 0.92) and were in good agreement. Muscle T2,w measured by IDEAL-CPMG was independent of changes in AFF. Muscle T2,w was higher in the biceps femoris and vastus lateralis muscles of subjects with DMD (p < 0.05). There was a strong correlation (p < 0.004) between AFF in all thigh muscles and six-minute walk distance (6MWD) in subjects with DMD.

Conclusions

IDEAL-CPMG allowed independent and simultaneous quantification of skeletal muscle fatty degeneration and disease activity in DMD. IDEAL-CPMG AFF and T2,w may be useful as biomarkers in clinical trials of DMD as the technique disentangles two competing biological processes.

Keywords: Duchenne muscular dystrophy, magnetic resonance imaging, IDEAL-CPMG, skeletal muscle, apparent fat fraction, fatty degeneration, T2 relaxation times, water T2

1. Introduction

Promising new treatments for Duchenne muscular dystrophy (DMD) are now being studied in clinical trials [1]. There is an urgent need for biomarkers to assess disease activity and treatment response in DMD. Since the disease pathology and treatment response may vary locally within a muscle, biomarkers that reflect the spatial distribution of the disease process within and across muscles are desired. Magnetic resonance imaging (MRI) enables non-invasive, repeatable, and objective assessment of individual muscles. In a preliminary study, muscle fat fractions quantified by Dixon correlated better with functional scores than measures of muscle strength [2]. Skeletal muscle T2 relaxation times were described as a biomarker of disease progression [3, 4] and treatment response in patients [5] and animal models of DMD [6]. However, increases in lipids and extracellular water contribute to lengthening of the apparent skeletal muscle T2, therefore the effects of inflammatory activity cannot be distinguished from fatty infiltration. Absence of muscle fat in mouse models makes apparent T2 and water T2 measurements almost equivalent [7]. In contrast increases in apparent T2 in patient muscle are proportional to increase in muscle fat fraction, and provide little information on inflammatory activity [7–9]. Simultaneous measurements of T2-weighted fat and T2-weighted water images independently within a single acquisition makes iterative decomposition of water and fat with echo asymmetry and least-squares estimation Carr-Purcell-Meiboom-Gill (IDEAL-CPMG) an attractive MRI technique for measuring inflammatory disease activity in the presence of fatty degeneration [7, 10]. The present study explored the use of IDEAL-CPMG in assessing disease activity in subjects with DMD participating in a clinical trial.

2. Materials and Methods

2.1 Participants and study design

The DMD Imaging Study was a 48-week prospective imaging study conducted between January 2012 and October 2013 as an optional sub-study of a multicenter phase 2, double blind, placebo-controlled trial of oligonucleotide therapy in ambulatory boys with DMD resulting from a mutation potentially corrected by skipping exon 51 (DMD114876; NCT01462292). The DMD imaging study was offered to all subjects eligible for the DMD114876 study before or during their first screening visit. Appendix E1 (Online) contains details of subject eligibility, recruitment, and inclusion and exclusion criteria. Subjects in the larger study received either drisapersen (3 or 6 mg/kg/week) or placebo for 24 weeks followed by a 24-week post-treatment phase. The subjects traveled to our Clinical Center during the screening phase, and at 12, 24, and 48 weeks. Subjects not eligible for randomization in the clinical trial had a one-time evaluation during the screening phase. Pilot data was also obtained from age-range matched healthy volunteer (HV) boys who were recruited from our Clinical Research Volunteer Program registry. All subjects were asked to avoid any excessive physical activity beyond their normal levels for a week prior to each study visit.

2.2 Standard protocol approvals, registrations, and patient consents

The study was registered on clinicaltrials.gov (NCT01451281) and was in compliance with the Patient Privacy Act and approved by the Institutional Review Board. Informed written assent and consent were obtained from the subject and parent or guardian before participation in the study.

2.3 MRI acquisition

MR images included a T1-weighted scan, an IDEAL-CPMG scan, and a DIXON scan acquired on a 3T Verio scanner (Siemens, Erlangen, Germany) at the Clinical Center. A Siemens 8 channel Body Matrix Coil Array was used for the thighs. All scans were performed on the same scanner. A parent or staff member was present in the scanner suite to alleviate subject anxiety. MR operators were trained to ensure consistency in the image acquisition across study visits.

Imaging was performed at mid-thigh, defined on the right femur as the midpoint between the inferior edge of the right medial femoral condyle and the inferior edge of the right femoral head. T1-weighted turbo spin echo images were acquired consisting of 21 5 mm slices with 5 mm gap and 1.5 mm in-plane resolution. The IDEAL-CPMG [10] acquisition measured 3 gradient echoes around each of N=12 spin echo times (10ms, 20ms, …, 120ms). The gradient echoes sampled three fat-water shifts at (−5π/6, π/2, 11π/6) [11] around each spin-echo time. Sixteen 5 mm slices with 5 mm gap were imaged in a 40×20 cm FOV with matrix size of 160×80, resulting in 2.5×2.5×5 mm voxel size with a TR of 5 s, resulting in a scan time of 6 m: 45 s. The Dixon acquisition consisted of 2 gradient echoes at 2.45 and 3.675 ms (nominally in-phase and out-of-phase at 3T) with a B0 field map in a 24 cm 3D slab with 48 5 mm sections in a 36×25 cm FOV with 256×256 matrix size, resulting in 1.4×1.4×5 mm voxels. The Dixon scan used a TR of 5.64 ms, flip angle 9°, and 2 averages in each of two abutting 3D slabs, resulting in 1 m: 40 s scan time.

2.4 MRI analysis

The T1-weighted scans were used as an anatomical reference image for the co-registration of all other within-visit scans for a given subject and for definition of the six thigh muscle regions-of-interest (ROIs): bicep femoris, rectus femoris, semitendinosus, vastus intermedius, vastus lateralis, and vastus medialis in the central five imaging slices of the right leg covering a distance of 50 mm. Definition of these muscle groups was performed manually using the Analyze 11.0 software package (AnalyzeDirect, Overland Park, KS). The co-registration process was a rigid-body (six degrees of freedom) transformation using the FSL tool FLIRT [12]. For two scans (one baseline and one 24-week scan), the T1-weighted image was incomplete/ unusable, so the in-phase Dixon scan was instead used as the anatomical reference image.

Dixon images were reconstructed using Siemens (Erlangen, Germany) product software producing a single water image (sw) and fat image (sf) generated from two gradient echo images and a field map [13]. The apparent fat fraction (AFF) measured using Dixon was then calculated as AFFDixon = 100 * |sf|/(|sf| + |sw|).

IDEAL-CPMG was reconstructed using in-house MATLAB code to generate a n = 1 … N series of T2-weighted water images, ρw[n], and T2-weighted fat images, ρf[n]. The water proton density (ρ0,w), fat proton density (ρ0,f), T2 of water (T2,w), T2 of fat (T2,f), were then fit according to ρw[n] = ρ0,w exp(−10ms · n/T2,w) + Nw and ρf[n] = ρ0,f exp(−10ms · n/T2,f) + Nf where Nw and Nf account for the non-zero floor due to sequence imperfections (e.g. B1-inhomogeneity, stimulated echoes) and the noise floor of magnitude data [14]. The T2-corrected AFF measured using IDEAL-CPMG was then calculated as AFFIDEAL−CPMG = 100 * |ρ0,f|/(|ρ0,f| + |ρ0,w|). Full details of the reconstruction and fitting can be found elsewhere [10].

For both the Dixon and the IDEAL-CPMG sequences, the reconstructed in-phase image was used for co-registration to the corresponding T1-weighted scan, with the computed co-registration parameters subsequently used to transform all of the remaining maps for that given imaging sequence (i.e., the T2 maps (T2,w, T2,f) and AFF maps for IDEAL-CPMG, and the AFF maps for Dixon). For each of the processed maps the mean of the signal in each of the muscle ROIs was generated using in-house MATLAB functions. All investigators were blinded to subject grouping and treatment during analysis.

2.5 Statistical analysis

A two-sample t-test, nonparametric (Mann-Whitney) test, and Fisher’s z transformation of sample correlation coefficients were used in cross-sectional comparisons. The level of significance was set at p < 0.05 without corrections for multiple comparisons. The degree of agreement was measured by the Bland-Altman method [15]. Pearson’s correlation was used for continuous variables. Spearman’s rank correlation was used for ordinal variables. Data analyses were done using Prism 6.0 (Graphpad Software, La Jolla, CA, USA).

3. Results

3.1 Subject Characteristics

Baseline MRI data were obtained from 13 boys with DMD and 20 HV boys. Both groups were similar in age (DMD: 6 –14 years; mean ± SD: 8.8 ± 2.4 years; and HV: 5–14 years; mean ± SD: 10 ± 2.1 years; p = 0.13) and age-adjusted height (p = 0.06) and weight (p = 0.17). Nine subjects with DMD, who were randomized to drisapersen 3 mg/kg/week (n=6) or placebo (n=3), returned for follow up visits at 12, 24, and 48 weeks during the DMD114876 trial. Three subjects did not have follow up visits because they were not randomized in the clinical trial, and one subject withdrew from the study after the baseline visit due to family/personal reasons.

3.2 Thigh muscle apparent fat fraction (AFF) and water T2 (T2,w) in age-range matched healthy volunteers and subjects with DMD

Figure 1 shows a T1-weighted image and IDEAL-CPMG images including T2-corrected AFF map and T2,w map from representative slices in three subjects with DMD. A variable degree of fatty replacement is seen in all subjects, but edema is less prominent and seen in isolated muscle groups in a subset of subjects (white arrow) on the T2,w map.

Figure 1.

Figure 1

Open in a new tab

Representative T1-weighted and IDEAL-CPMG images of the thigh muscles in three subjects with DMD. A T1-weighted image (A), T2-corrected fat fraction map (B), and water-T2 map (C) are shown representing subject anatomy, changes in muscle apparent fat fraction (AFFIDEAL-CPMG) and muscle water T2 (T2,w IDEAL-CPMG) respectively in the thigh muscles of subjects with DMD. Different severity of fatty degeneration is present in the thigh muscles of each subject, whereas inflammatory activity is sparse and seen in only few muscles (arrow).

The AFF in six thigh muscles measured by IDEAL-CPMG and Dixon was about 2 to 4 fold higher in subjects with DMD than the age-range matched HV boys (Fig. 2A–B, Table 1). A linear regression model using factor status, HV or DMD, and age as a covariate showed a significantly higher AFF in subjects with DMD compared to HV boys at all ages (p ≤ 0.001). The range of Dixon AFF in the thigh muscles was 4%–17% in HV boys and 8%–75% in subjects with DMD and that of IDEAL-CPMG AFF was 3%–20% in HV boys and 7%–84% in subjects with DMD (Table 1). There was a strong correlation between IDEAL-CPMG and Dixon for AFF measurements in the same muscle (n=180 muscles, Spearman r = 0.92; p < 0.0001). The mean difference, IDEAL-CPMG minus Dixon, was 0.95 (95% CI 0.29, 1.63). Figures 2C and 2D show agreement between the two methods. Bland and Altman plot shows that 95% of the differences in the AFF were within the limits of agreement (−7.97, 9.88) (Fig. 2D). We found significant correlations among all muscles of subjects with DMD for IDEAL-CPMG and Dixon AFF (Table 2).

Figure 2.

Figure 2

Open in a new tab

MRI quantification of thigh muscle apparent fat fraction (AFF) and water T2 (T2,w) in healthy volunteers and subjects with DMD. (A–B) Comparison of AFF measured by Dixon (A) and IDEAL-CPMG (B) in the biceps femoris (BF), rectus femoris (RF), semitendinosus (ST), vastus intermedius (VI), vastus lateralis (VL), and vastus medialis (VM) muscles of HV boys (blue circles) and subjects with DMD (red squares). Mean AFF measured by both methods was significantly higher in all the muscles in subjects with DMD relative to HV (p < 0.001). (C–D) Line of identity (C) and Bland-Altman plot (D) demonstrate excellent agreement between the two methods. Difference = AFFIDEAL-CPMG − AFFDixon. Dotted lines in D represent mean difference and limits of agreement (mean ± 1.96 S.D.) (E) Mean T2,w values were significantly higher in the BF (p = 0.02) and VL (p = 0.03) of subjects with DMD (red squares) relative to HV boys (blue circles). (F) Graph shows that T2,w measurements were independent of the variation in AFF in 240 muscles examined in HV boys (blue circles) and subjects with DMD (red squares). Error bars represented as mean ± SEM.

Table 1.

MRI measures of muscle apparent fat fraction (AFF) in the thigh muscles of age range-matched healthy volunteers (HV) and subjects with Duchenne muscular dystrophy (DMD)

Muscles IDEAL-CPMG
Median AFF (%) (range)		 Dixon
Median AFF (%) (range)
HV
(n = 14)		 DMD
(n = 10)		 P value
Mann-
Whitney		 HV
(n = 16)		 DMD
(n = 8)		 P value
Mann-
Whitney
Biceps femoris 7.4
(4.2–17.5)		 23.5
(16.4–80.8)		 <0.0001 10.5
(5.1–15.8)		 23.0
(17.3–74.7)		 <0.0001
Rectus femoris 6.5
(4.1–20.3)		 28.2
(11.8–84.0)		 <0.0001 6.6
(3.6–11.7)		 20.2
(13.4–71.7)		 <0.0001
Semitendinosus 8.3
(4.1–19.6)		 19.1
(11.8–49.9)		 <0.001 8.1
(5.1–13.3)		 15
(13.6–40.2)		 <0.0001
Vastus intermedius 3.6
(3–10.1)		 15.1
(7.3–79.4)		 <0.001 7
(4.8–10.7)		 15.5
(8.5–74)		 <0.0001
Vastus lateralis 5.2
(3.9–17.5)		 18.3
(8.3–83.9)		 <0.001 11
(5.7–17.4)		 18.8
(11.4–75.4)		 <0.001
Vastus medialis 5
(3.8–19.7)		 17.9
(8.0–80.5)		 <0.001 6.7
(4.8–11.4)		 14.1
(9.9–72.3)		 <0.0001
Open in a new tab

Table 2.

Spearman correlations between muscle AFFIDEAL-CPMG (a) and AFFDixon (b) in six thigh muscles of subjects with Duchenne muscular dystrophy (DMD)

a.
Muscles BF RF ST VI VL VM
BF 1 0.73* 0.74* 1**** 0.94*** 0.91***
RF 1 0.65 0.73* 0.81** 0.81**
ST 1 0.74* 0.83** 0.88**
VI 1 0.94*** 0.91***
VL 1 0.93***
VM 1
b.
Muscles BF RF ST VI VL VM
BF 1 0.90* 0.69 0.90** 0.90** 0.90**
RF 1 0.40 0.76* 0.71 0.71
ST 1 0.79* 0.74 0.74
VI 1 0.95** 0.95**
VL 1 1***
VM 1
Open in a new tab

BF = biceps femoris, RF = rectus femoris, ST = semitendinosus, VI = vastus intermedius, VL = vastus lateralis, VM = vastus medialis

(*p <0.05; **p <0.01; ***p <0.001; ****p <0.0001)

The IDEAL-CPMG T2,w of the biceps femoris (mean ± S.E.M 31.4 ± 0.4 v. 29.9 ± 0.5; p = 0.02) and vastus lateralis (mean ± S.E.M 32.3 ± 0.4 v. 31.3 ± 0.7; p = 0.03) muscles was significantly higher in subjects with DMD than HV boys (Fig. 2E, Table 3). The T2,w of other muscles was not significantly different between the two groups. There was no correlation between IDEAL-CPMG T2,w and AFF in the same muscle (n = 240 muscles, Spearman r = 0.04; p = 0.48) (Fig. 2F).

Table 3.

MRI measures of water-T2 (T2,w) in the thigh muscles of age range-matched healthy volunteers (HV) and subjects with Duchenne muscular dystrophy (DMD)

Muscles IDEAL-CPMG
Median muscle T2,w (ms) (range)
HV
(n = 14)		 DMD
(n = 10)		 P value
Mann-
Whitney
Biceps femoris 29.6 (27.6–33.6) 31.1 (29.7–33.6) 0.02*
Rectus femoris 38.8 (28.2–63.1) 34.5 (29.4–44.2) 0.6
Semitendinosus 28.3 (26.7–32.6) 29.6 (28.1–34.3) 0.2
Vastus intermedius 31.1 (28.9–40.6) 31.7 (30.0–33.5) 0.1
Vastus lateralis 30.5 (28.7–39.2) 32.3 (30.4–34.2) 0.03*
Vastus medialis 34.8 (30.3–39.5) 32.8 (30.1–38.1) 0.1
Open in a new tab
*

p < 0.05

3.3 Age correlations with thigh muscle IDEAL-CPMG AFF and T2,w in subjects with DMD

The AFF of the biceps femoris and vastus medialis, vastus intermedius and vastus lateralis muscles increased with age in subjects with DMD (Pearson r = 0.65–0.69; p < 0.05) (Fig. 3A). There was no significant correlation between age and AFF in the rectus femoris and semitendinosus muscles (Pearson r = 0.53–0.54; p = 0.1). The T2,w decreased with age in the rectus femoris muscle (Pearson r = −0.67, p < 0.05), whereas there was no significant correlation in the other muscles (Pearson r = −0.56–0.04; p = 0.09 – 0.4) of subjects with DMD (Fig. 3B). No correlations were found between age and the muscle AFF or T2,w in HV boys.

Figure 3.

Figure 3

Open in a new tab

Age correlations with thigh muscle IDEAL-CPMG apparent fat fraction (AFF) and water T2 (T2,w) in subjects with DMD. (A) There is a significant increase in AFF of the biceps femoris, vastus intermedius, vastus lateralis and vastus medialis muscles with age. (B) Conversely, there is a significant decrease in T2,w of the rectus femoris muscle with age. Pearson correlation coefficient (r) is shown. Dotted lines represent mean values of the age-range matched HV boys.

MRI scans were evaluated in 4 subjects with DMD randomized to drisapersen at 3 mg/kg/week (age range: 7–10 years) and 2 subjects (age: 6 years) randomized to placebo at baseline, 24 weeks, and 48 weeks during the DMD114876 trial. Images in the remaining 3 subjects were not suitable for quantitative longitudinal analysis due to motion artifacts or fat/water swaps across ROIs. The AFF of the thigh muscles ranged from 45% to 76% in the 10 year old subject and ≤ 25% in the younger subjects. The six year-old subjects (n=2) in the placebo group had ≤ 2% increase in mean muscle AFF, whereas the 7–10 year old subjects (n=4) assigned to treatment group had ≥ 6% increase in mean muscle AFF of the thigh muscles during the 48-week follow up. The T2,w range (29 – 44 ms) in the thigh muscles was not different between the subjects in the placebo and treatment groups. There was no apparent effect of the treatment on AFF or T2,w in this cohort.

3.4 Correlations between six-minute walk distance and thigh muscle AFFIDEAL-CPMG in subjects with DMD

All subjects showed a change in the six-minute walk distance (6MWD) over the 48-week period (Fig. 4A). The six and seven year-old subjects had an increase in the 6MWD, whereas DMD subjects age >7 years had a decline in the 6MWD, as shown previously [16], There was an exception in that the seven year-old subject had a lower 6MWD relative to the eight year-old subject throughout the study period. Notably, the seven year-old subject had a higher AFF in all six muscles compared to the eight-year old subject. We observed that changes in AFF of the thigh muscles over time had a trend opposite to that of the 6MWD in younger vs. older subjects with DMD (Fig. 4B; vastus lateralis muscle shown). We examined correlations between all available paired measurements of AFF in individual thigh muscles and 6MWD in subjects with DMD (n = 22 for each muscle; each pair was treated independently). Strong negative correlations were found between the 6MWD and AFF in all thigh muscles of subjects with DMD examined in this study (Spearman r = −0.75– −0.57; p ≤ 0.004) (Fig. 4C; vastus lateralis muscle shown). Subjects with ≤ 20% AFF in the vastus lateralis muscle walked > 400 m and conversely, those with > 20% AFF in the vastus lateralis muscle walked < 400 m in 6 minutes.

Figure 4.

Figure 4

Open in a new tab

Change in the six-minute walk distance (6MWD) and thigh muscle IDEAL-CPMG apparent fat fraction (AFF) and water T2 (T2,w) in subjects with DMD from baseline to 48 weeks (including 24-week treatment phase plus 24-week follow up) in the clinical trial. (A–B) Change in the 6MWD and the vastus lateralis muscle AFF from baseline to 48 weeks. The 6MWD increased in the six (light or dark blue) and seven (green) year-old subjects and decreased in the eight (brown), nine (purple) and 10 (red) year-old subjects (A). Dotted lines represent predictor cutoffs of decline in ambulation (age > 7 years and 6MWD < 350 m) [16]. The vastus lateralis muscle AFF increased in all subjects (B). Dotted line represents the mean AFF of the age-range matched HV boys. Time scale is shown from individual subject’s age at baseline. (C) Correlation between paired measurements of the 6MWD and the vastus lateralis muscle AFF. Spearman r is shown. Dotted horizontal line represents mean AFF of the age-range matched HV boys (7.5%) and dotted vertical line represents predictor cutoff of decline in ambulation (6MWD <350 m). (D–E) Change in muscle AFF and T2,w of the biceps femoris (BF), rectus femoris (RF), semitendinosus (ST), vastus intermedius (VI), vastus lateralis (VL) and vastus medialis (VM) muscles of subjects with increased 6MWD (D) and decreased 6MWD (E). Each bar represents an individual subject with color schema as in A and B.

The progression of muscle AFF (as in [17]) and muscle water T2 was different in individual thigh muscles followed over the 48-week period (Fig. 4D–E). In agreement with a recent publication [18], we found that the IDEAL-CPMG sequence detected modest increases in the thigh muscles AFF in the six and seven year-old subjects whose 6MWD improved over year (Fig. 4D). The greatest increase (>10%) in the muscle AFF was found in the biceps femoris and vastus lateralis muscles of the eight and nine year-old subjects whose 6MWD declined by >30 m (Fig. 4E). Overall, there was a greater increase in the muscle AFF in the older subjects whose 6MWD decreased compared to the younger subjects whose 6MWD increased during the 48-week follow up period. Conversely, the changes in muscle water T2 were more remarkable in the younger subjects with increased 6MWD compared to the older subjects with decreased 6MWD over the year (Fig. 4D–E).

4. Discussion

Quantitative MRI correlates of fatty degeneration and muscle inflammatory activity are potentially useful for monitoring disease progression and treatment response during a clinical trial. The MRI parameter of muscle inflammatory activity corresponds to water mobility within muscle tissue (T2,w) and may reflect loss of muscle fiber integrity in the dystrophic lesions and in the areas of inflammation and necrosis in skeletal muscle. Changes in T2,w are dynamic and sensitive to the underlying disease activity. Fatty replacement of muscle tissue occurs during later stages in the disease process. Dixon separates muscle fat and water based on phase difference (chemical shift) and provides measurement of muscle fat without any confounding effect of muscle edema. However, quantification of muscle inflammatory activity in fat-infiltrated dystrophic muscles has been challenging because increases in muscle water T2 are masked by the presence of muscle fat [7, 9]. We present IDEAL-CPMG as an MRI biomarker for simultaneous quantification of skeletal muscle T2,w and AFF in DMD. Our results show that IDEAL-CPMG reliably and efficiently separates fat and water components of T2 in the thigh muscles, which are affected early and more severely than other muscles in subjects with DMD. There was no relationship between T2,w and AFF (r = 0.04), indicating that there is little fat contamination within the IDEAL-CPMG water signal and that the level of inflammatory activity does not change with increases in AFF in the thigh muscles of subjects with DMD. This agrees with previous studies in DMD using spectroscopy in the lower leg muscles [3] and the forearm muscles [19] but contrasts to previous studies in hypokalemic periodic paralysis [20] where a weak correlation (r = 0.54) between T2,w and AFF in the lower leg muscles was observed.

MR spectroscopy (MRS) is the standard for noninvasive measurement of muscle fat. However, poor spatial resolution and coverage as well as challenges in the spectroscopy volume registrations consistently between imaging visits limit its use in multicenter clinical studies. IDEAL-CPMG and Dixon provide better spatial resolution and coverage, enabling assessment of multiple muscles simultaneously [10, 21–23]. This is pertinent in DMD because the disease varies locally within a muscle and between muscles in patients. Variations of Dixon have been used in AFF quantification in DMD [2, 22, 24, 25]. We observed a strong correlation (r = 0.92) between IDEAL-CPMG and Dixon across a broad range (3%–84%) of muscle fat values. The Dixon sequence is T1-weighted, with the level of fat overestimated from the shorter T1 of fat versus water. The IDEAL-CPMG sequence did not suffer from T1 saturation, but a magnetization transfer effect from the radiofrequency pulses would reduce only the water signal. Thus, both methods suffer from reduction of the water signal, but for different reasons. Further, the IDEAL-CPMG decomposition used a 7-peak fat multispectral model [10], whereas the Dixon formulation assumes a single fat peak. The implicit assumption of a single T2* value in the Dixon sequence may explain the increased difference between IDEAL-CPMG and Dixon at high fat fractions [26].

The AFF and T2,w changes had a different muscle pattern and evolution with age in subjects with DMD. Fatty degeneration was prominent and dystrophic inflammation or edema was relatively sparse in the thigh muscles of the DMD cohort with 11 of 13 subjects ≥ 7 years of age. The biceps femoris and vastus lateralis muscles were most responsive to disease progression consistent with previous observations [3,18]. The change in the T2,w of the thigh muscles was relatively less in the > 7 year-old subjects compared to younger ones. These findings support previous observations that MRS T2,w may be a sensitive marker of response to prednisone before skeletal muscles are loaded with fatty degeneration [5] and that MRS T2,w tends to normalize with disease progression in DMD [3, 19]. The T2,w was reliably measured in the hamstrings and vastus muscles (SE ±0.5ms to ± 0.8ms). However, B1 inhomogeneity due to the relatively high contribution from stimulated echoes within the rectus femoris muscle precluded accurate T2,w assessment (SE ± 2.4 ms). This could be addressed by using extended phase graph fitting as proposed by Lebel and Wilman [27, 28] instead of the monoexponential least squares fitting employed in this study. Other methodological limitations include a restricted spatial resolution demanded by time constraints in a pediatric study, and a simplistic fitting model for T2 decay in magnitude images [29]. Further studies are needed to more fully qualify these biomarkers, including determining the inter- and intra- rater and site variability.

We did not see a therapeutic response to drisapersen on MRI measures of fatty infiltration and edema in this exploratory imaging substudy, probably because 1) few subjects participated, 2) subjects receiving placebo and active agent had different age distributions due to randomization and 3) no subjects received the higher dose (6 mg/kg/week) found to be efficacious elsewhere [30], again due to randomization.

Our findings support the conclusions of earlier studies [18, 31] that MRI measures are sensitive to disease progression that cannot be identified by functional testing such as 6MWD. We found that subjects with DMD who walked about 350 m at baseline had a broad range of AFF in each muscle. The 6MWD and other functional assessments are influenced by patient motivation and cooperation, which may be challenging in a pediatric population. Furthermore, multiple muscle groups differently affected by the disease are recruited during the activity. In contrast, MRI biomarkers provide objective assessments of individual muscles to monitor disease progression and treatment response during clinical trial. Skeletal muscle AFF at baseline in subjects with DMD may serve as an objective indicator of disease severity and help to select or stratify subjects appropriately for clinical trials.

In summary, IDEAL-CPMG provides independent and simultaneous quantification of disease activity and fatty degeneration in multiple skeletal muscles of subjects with Duchenne muscular dystrophy. The T2,w changes are sparse in the thigh muscles of this age group in DMD, agreeing with MRS [3, 19] and in contrast with apparent muscle T2 relaxation times [3, 4]. IDEAL-CPMG and Dixon agree well on AFF measurements. IDEAL-CPMG AFF and T2,w may be useful as an imaging biomarker in future trials.

Supplementary Material

1

Highlights.

  • IDEAL-CPMG enables quantification of disease activity in fatty-infiltrated muscles

  • IDEAL-CPMG muscle AFF and T2,w are useful biomarkers in clinical trials of DMD

  • IDEAL-CPMG AFF at baseline may help stratify subjects with DMD in clinical trials

  • IDEAL-CPMG T2,w is an objective biomarker of disease activity in subjects with DMD.

Acknowledgments

The study was supported by NINDS intramural research funds. The cost of image analysis was covered in part by GlaxoSmithKline through a contract with Imanova Limited. Drs. Bishop and Newbould are employees of Imanova Limited. Dr. Fischbeck has served as an unpaid member of the Scientific Advisory Board for Prosensa (now BioMarin Nederland). Dr. Janiczek is an employee of GlaxoSmithKline.

The authors thank the study participants and their families, Mr. Donovan Stock, Ms. Elizabeth Hartnett, and Ms. Alice Schindler for help with coordinating study visits and scheduling MRI studies, Mr. Robert Evers and Dr. Larry Yao (Radiology, Clinical Center) for help with installing MRI sequence protocols on the scanner, and the Principal Investigators of the referring sites participating in the DMD114876 trial (Drs. Barry Russman, Portland, OR; Benjamin Renfroe, Gulf Breeze, FL; Brenda Wong, Cincinnati, OH; Douglas Sproule, New York, NY; Edward Smith, Durham, NC; and Kathryn Wagner, Baltimore, MD) for sharing study participants. We also thank GlaxoSmithKline and Prosensa/BioMarin, study sponsor of the DMD114876 trial, for sharing the 6MWD data.

Abbreviations

IDEAL-CPMG

Iterative decomposition of water and fat with echo asymmetry and least-squares estimation Carr-Purcell-Meiboom-Gill

AFF

Apparent fat fraction

T2,w

water T2

DMD

Duchenne muscular dystrophy

MRI

Magnetic resonance imaging

MRS

Magnetic resonance spectroscopy

ROI

Region of interest

6MWD

Six-minute walk distance

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Disclosures: Drs. Mankodi and Auh report no disclosures.

Authors’ contributions

Design and conceptualization of the study (AM, KF, RJ), data analysis and interpretation (AM, CB, RN, RJ), statistical analysis (AM, SA), drafting the manuscript and revising the manuscript for intellectual content (AM, RJ, CB, RN, KF). AM had access to all the data and takes responsibility for the data, accuracy of the data analysis, and the conduct of the research.

Contributor Information

Courtney A. Bishop, Email: Courtney.Bishop@imanova.co.uk.

Sungyoung Auh, Email: Sungyoung.Auh@nih.gov.

Rexford D. Newbould, Email: Rexford.Newbould@imanova.co.uk.

Kenneth H. Fischbeck, Email: Fischbeck@ninds.nih.gov.

Robert Janiczek, Email: Rob.X.Janiczek@gsk.com.

REFERENCES

  • 1.Seto JT, Bengtsson NE, Chamberlain JS. Therapy of Genetic Disorders-Novel Therapies for Duchenne Muscular Dystrophy. Curr Pediatr Rep. 2014;2:102–112. doi: 10.1007/s40124-014-0044-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wren TA, Bluml S, Tseng-Ong L, Gilsanz V. Three-point technique of fat quantification of muscle tissue as a marker of disease progression in Duchenne muscular dystrophy: preliminary study. Am J Roentgenol. 2008;190:W8–W12. doi: 10.2214/AJR.07.2732. [DOI] [PubMed] [Google Scholar]
  • 3.Forbes SC, Willcocks RJ, Triplett WT, et al. Magnetic resonance imaging and spectroscopy assessment of lower extremity skeletal muscles in boys with Duchenne muscular dystrophy: a multicenter cross sectional study. PLoS One. 2014;9:e106435. doi: 10.1371/journal.pone.0106435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Willcocks RJ, Arpan IA, Forbes SC, et al. Longitudinal measurements of MRI-T2 in boys with Duchenne muscular dystrophy: effects of age and disease progression. Neuromuscul Disord. 2014;24:393–401. doi: 10.1016/j.nmd.2013.12.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Arpan I, Willcocks RJ, Forbes SC, et al. Examination of effects of corticosteroids on skeletal muscles of boys with DMD using MRI and MRS. Neurology. 2014;83:974–980. doi: 10.1212/WNL.0000000000000775. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Walter G, Cordier L, Bloy D, Sweeney HL. Noninvasive monitoring of gene correction in dystrophic muscle. Magn Reson Med. 2005;54:1369–1376. doi: 10.1002/mrm.20721. [DOI] [PubMed] [Google Scholar]
  • 7.Carlier PG. Global T2 versus water T2 in NMR imaging of fatty infiltrated muscles: different methodology, different information and different implications. Neuromuscul Disord. 2014;24:390–392. doi: 10.1016/j.nmd.2014.02.009. [DOI] [PubMed] [Google Scholar]
  • 8.Azzabou N, Loureiro de Sousa P, Caldas E, Carlier PG. Validation of a generic approach to muscle water T2 determination at 3T in fat-infiltrated skeletal muscle. J Magn Reson Imaging. 2015;41:645–653. doi: 10.1002/jmri.24613. [DOI] [PubMed] [Google Scholar]
  • 9.Hollingsworth KG. Quantitative MRI in muscular dystrophy: An indispensable trial endpoint? Neurology. 2014;83:956–957. doi: 10.1212/WNL.0000000000000785. [DOI] [PubMed] [Google Scholar]
  • 10.Janiczek RL, Gambarota G, Sinclair CD, et al. Simultaneous T(2) and lipid quantitation using IDEAL-CPMG. Magn Reson Med. 2011;66:1293–1302. doi: 10.1002/mrm.22916. [DOI] [PubMed] [Google Scholar]
  • 11.Pineda AR, Reeder SB, Wen Z, Pelc NJ. Cramer-Rao bounds for three-point decomposition of water and fat. Magn Reson Med. 2005;54:625–635. doi: 10.1002/mrm.20623. [DOI] [PubMed] [Google Scholar]
  • 12.Jenkinson M, Smith S. A global optimisation method for robust affine registration of brain images. Med Image Anal. 2001;5:143–156. doi: 10.1016/s1361-8415(01)00036-6. [DOI] [PubMed] [Google Scholar]
  • 13.Jellus V. Phase correction method. US20100201364. US Patent. 2010 Aug 12;
  • 14.Milford D, Rosbach N, Bendszus M, Heiland S. Mono-exponential fitting in T2-relaxometry: relevance of offset and first echo. PLoS One. 2015;10:e0145255. doi: 10.1371/journal.pone.0145255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310. [PubMed] [Google Scholar]
  • 16.McDonald CM, Henricson EK, Abresch RT, et al. The 6-minute walk test and other endpoints in Duchenne muscular dystrophy: longitudinal natural history observations over 48 weeks from a multicenter study. Muscle Nerve. 2013;48:343–356. doi: 10.1002/mus.23902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Hollingsworth KG, Garrood P, Eagle M, Bushby K, Straub V. Magnetic resonance imaging in duchenne muscular dystrophy: Longitudinal assessment of natural history over 18 months. Muscle Nerve. 2013;48:586–588. doi: 10.1002/mus.23879. [DOI] [PubMed] [Google Scholar]
  • 18.Willcocks RJ, Rooney WD, Triplett WT, et al. Multicenter prospective longitudinal study of magnetic resonance biomarkers in a large duchenne muscular dystrophy cohort. Ann Neurol. 2016;79:535–547. doi: 10.1002/ana.24599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Wary C, Azzabou N, Giraudeau C, et al. Quantitative NMRI and NMRS identify augmented disease progression after loss of ambulation in forearms of boys with Duchenne muscular dystrophy. NMR Biomed. 2015;28:1150–1162. doi: 10.1002/nbm.3352. [DOI] [PubMed] [Google Scholar]
  • 20.Sinclair CD, Morrow JM, Janiczek RJ, et al. Reproducibility and sensitivity of muscle-water T2 determined independendtly of fat fraction with IDEAL-CPMG. Proc Intl Soc Mag Reson Med. 2015;23 0748. [Google Scholar]
  • 21.Kukuk GM, Hittatiya K, Sprinkart AM, et al. Comparison between modified Dixon MRI techniques, MR spectroscopic relaxometry, and different histologic quantification methods in the assessment of hepatic steatosis. Eur Radiol. 2015;25:2869–2879. doi: 10.1007/s00330-015-3703-6. [DOI] [PubMed] [Google Scholar]
  • 22.Triplett WT, Baligand C, Forbes SC, et al. Chemical shift-based MRI to measure fat fractions in dystrophic skeletal muscle. Magn Reson Med. 2014;72:8–19. doi: 10.1002/mrm.24917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bernard CP, Liney GP, Manton DJ, Turnbull LW, Langton CM. Comparison of fat quantification methods: a phantom study at 3.0T. J Magn Reson Imaging. 2008;27:192–197. doi: 10.1002/jmri.21201. [DOI] [PubMed] [Google Scholar]
  • 24.Fischmann A, Hafner P, Gloor M, et al. Quantitative MRI and loss of free ambulation in Duchenne muscular dystrophy. J Neurol. 2013;260:969–974. doi: 10.1007/s00415-012-6733-x. [DOI] [PubMed] [Google Scholar]
  • 25.Wokke BH, Bos C, Reijnierse M, et al. Comparison of dixon and T1-weighted MR methods to assess the degree of fat infiltration in duchenne muscular dystrophy patients. J Magn Reson Imaging. 2013;38:619–624. doi: 10.1002/jmri.23998. [DOI] [PubMed] [Google Scholar]
  • 26.Chebrolu VV, Hines CDG, Yu H, et al. Independent estimation of T*2 for water and fat for improved accuracy of fat quantification. Magn Reson Med. 2010;63:849–857. doi: 10.1002/mrm.22300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Lebel RM, Wilman AH. Transverse relaxometry with stimulated echo compensation. Magn Reson Med. 2010;64:1005–1014. doi: 10.1002/mrm.22487. [DOI] [PubMed] [Google Scholar]
  • 28.Marty B, Baudin PY, Reyngoudt H, et al. Simultaneous muscle water T2 and fat fraction mapping using transverse relaxometry with stimulated echo compensation. NMR Biomed. 2016;29:431–443. doi: 10.1002/nbm.3459. [DOI] [PubMed] [Google Scholar]
  • 29.Yin X, Shah S, Katsaggelos AK, Larson AC. Improved R2* measurement accuracy with absolute SNR truncation and optimal coil combination. NMR Biomed. 2010;23:1127–1136. doi: 10.1002/nbm.1539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Voit T, Topaloglu H, Straub V, et al. Safety and efficacy of drisapersen for the treatment of Duchenne muscular dystrophy (DEMAND II): an exploratory, randomised, placebo-controlled phase 2 study. Lancet Neurol. 2014;13:987–996. doi: 10.1016/S1474-4422(14)70195-4. [DOI] [PubMed] [Google Scholar]
  • 31.Willis TA, Hollingsworth KG, Coombs A, et al. Quantitative muscle MRI as an assessment tool for monitoring disease progression in LGMD2I: a multicenter longitudinal study. PLoS One. 2013;8:e70993. doi: 10.1371/journal.pone.0070993. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS814544-supplement-1.docx (18.8KB, docx)

RESOURCES