Fig 1.
Representative T2 maps, 1H MRS voxel placement, and 1H MR spectra.
A) T2 maps were created from the 40, 60, 80, and 100ms TE spin echo images of the lower leg and thigh. Utilizing the popliteus and biceps femoris short head as standardized anatomical landmarks to guide lower leg and thigh slice selection, the TA, PER, TP, SOL, MG, VL, BFLH, and GRA muscles were traced on three contiguous slices to determine mean T2 for each muscle. B) T1-weighted images demonstrate the location of voxel placement for MR spectroscopy used to quantify FF. Voxels were sized as large as possible while staying completely within the SOL and VL muscles, avoiding the fascia. The voxel position from prior years was referenced to position voxels at an anatomically similar location at follow-up visits. C-F) T1-weighted images of the calf and thigh with corresponding SOL and VL spectra and FFs from a representative participant in each of four functional groups (based on the 10m walk/run test). Participants have increasing levels of SOL and VL fat infiltration (appreciated visually on the T1-weighted images and quantified using spectroscopy) with decreasing ambulatory ability. 10m walk/run times were 3.56s, 7.16s, and 11.32s for the example participants in panels C, D, and E, respectively. (TA = tibialis anterior, PER = peroneus longus and brevis, TP = tibialis posterior SOL = soleus, MG = medial gastrocnemius, VL = vastus lateralis, BFLH = biceps femoris long head, GRA = gracilis, FF = fat fraction).
Table 1.
Cohort characteristics and demographics.
Fig 2.
Scatterplots showing the correlations between VL MRI T2 and FF and functional endpoints.
The qMR biomarkers most strongly correlated to function were VL MRI T2 and VL FF. A) VL MRI T2 versus 10m walk/run, B) VL FF versus 10m walk/run, C) VL MRI T2 versus 6MWD, D) VL FF vs 6MWD. Participants who lost the ability to perform the timed tests were given the maximum time of 45 seconds (open circles) for the first visit in which they were unable to complete the test. Participants who lost the ability to perform the 6 MWT have distances of 0 m (open circles). As indicated in Table 2, all correlations were significant with the 95% bootstrap confidence intervals excluding zero. (VL = vastus lateralis, 6MWD = six minute walk distance, 6MWT = six minute walk test).
Table 2.
Correlation between MRI T2, muscle FF, and functional endpoints.
Fig 3.
FF and quantitative and qualitative measures of walking/running ability.
A and B) FF was assessed based on 10m walk/run performance. Data points from participants who could complete the test were divided into bins of <5 sec (n = 152), 5–8 sec (n = 184), and >8 sec (n = 127). The LOA group represents data points from subjects that could no longer complete the 10m walk/run at the time of the visit (n = 35). The amount of SOL and VL fatty infiltration was significantly different between functional performance groups. C) Groups were created based on the qualitative rating of participants’ 10m walk/run test at each time point. Mean VL FF was significantly different between each qualitative grade group. * = significantly different; bootstrap confidence intervals excluded zero. (SOL = soleus, VL = vastus lateralis).
Fig 4.
Loss of function in participants with different VL FFs.
Data points were binned by VL FF and further divided by the number of functional skills the participants could complete at the time of the visit. Black = ambulatory, able to climb 4 stairs, able to move supine to stand. Dark gray = ambulatory, able to climb 4 stairs, unable to rise from supine. Light gray = ambulatory, unable to perform stairs, unable to rise from supine. White = nonambulatory. As FF increased, there was a progressive decrease in the percent of participants who could perform the functional tasks. At VL FFs less than 0.19, 99% of participants could perform all functional tests. Apart from one outlier, participants began losing ambulation at VL FFs ≥ 0.40. (VL = vastus lateralis, STS = supine to stand).
Fig 5.
qMR biomarkers of muscle pathology at the first visit following loss of ambulation.
A) MRI T2 and B) MRS FF are plotted for each muscle at the first time point at which participants were nonambulatory (n = 35). Whiskers represent the 5th-95th percentiles with outliers represented by the filled dots. MRI T2 was highest in the BFLH and VL and lowest in the GRA, TA, and TP. The MRI T2 and FF of the VL were higher than that of the SOL as expected based on the proximal to distal pattern of involvement in DMD.