Multiparametric quantitative magnetic resonance imaging of skeletal muscle in CKD

Abstract

Skeletal muscle dysfunction causes functional decline and disability in patients with chronic kidney disease (CKD). Identification of muscle pathology before significant loss of physical function would be a major advance. Multiparametric, quantitative magnetic resonance imaging (qMRI) of seven leg muscle groups (3 thigh and 4 calf) was conducted in patients with CKD stages 4-5 (n = 6), end-stage kidney disease (ESKD, n = 3), and healthy controls (n = 10) using a 3 Tesla MRI scanner. Measurements included T1 relaxation time in the rotating frame (T1ρ) and transverse relaxation time (T2) mapping, Dixon imaging of intramuscular fat content, diffusion tensor imaging (DTI) for muscle structure, and ¹H-MR spectroscopy for intra- and extra-myocellular lipid (IMCL and EMCL, respectively) and physiologically relevant muscle metabolites. T1ρ and T2 times were prolonged and fat fraction (FF) was higher in patients with CKD compared with controls (differences of 4.99 ms (95% CI 1.71-8.27), 6.72 ms (95% CI 3.78-9.66), and 6.67% (95% CI 0.65-12.68), respectively). T1ρ and FF were similarly elevated across muscle groups, whereas T2 differences may have been greater in calf muscles. T1ρ and T2 were lower in patients with ESKD compared with CKD and similar to controls, consistent with prior histologic assessment of muscle fibrosis. No significant differences by CKD status were observed for DTI parameters. Compared with controls, IMCL was higher in patients with CKD, and trimethylamine and creatine concentrations were lower. In sum, multiparametric qMRI of skeletal muscle in patients with CKD noninvasively identified differences in metrics associated with fibrosis, fat infiltration, and metabolic dysregulation.NEW & NOTEWORTHY In this study, we demonstrate that multiparametric, quantitative magnetic resonance imaging (qMRI) can quantify multiple distinct anatomic and pathologic features of skeletal muscle pathology in patients with CKD before significant functional decline. qMRI metrics of fibrosis and fat infiltration were elevated, and muscle metabolite concentrations were reduced, in patients with CKD compared with controls. This noninvasive approach offers a valuable alternative to traditional muscle biopsies for evaluating muscle health in patients with CKD.

Keywords: MRI; chronic kidney disease; sarcopenia; skeletal muscle.

Figure 1.. Flow diagram of study participation.

Figure 2.. Representative quantitative magnetic resonance imaging maps.

T1-weighted images and T1ρ, fat fraction, and T2 maps of thigh muscles for control, CKD, and ESKD participants. Abbreviations: FF, fat fraction.

Figure 3.. Quantitative magnetic resonance imaging metrics by person and muscle group.

Values for each muscle group for an individual participant are plotted relative to that individual’s values for the vastus lateralis muscle. For each metric, the range of Spearman correlation coefficients across muscle groups is displayed in the upper left corner. n=19 for all metrics except T1ρ and T2 (n=18). Abbreviations: VM, vastus medialis; BFL, biceps femoris longus; TA, tibialis anterior; S, soleus; GM, gastrocnemius medialis; GL, gastrocnemius lateralis; ADC, apparent diffusion coefficient.

Figure 4.. Quantitative magnetic resonance imaging metrics by muscle group and CKD status.

Data are presented as mean ± SE. Mixed effects models including random intercepts were fit to account for clustering within an individual. Pairwise comparisons of CKD and control participants within muscle groups were performed using tests of contrast. n=19 for all metrics except T1ρ and T2 (n=18). Abbreviations: VL, vastus lateralis; VM, vastus medialis; BFL, biceps femoris longus; TA, tibialis anterior; S, soleus; GM, gastrocnemius medialis; GL, gastrocnemius lateralis. *p<0.05, **p<0.01, ***p<0.001.

Figure 5.. Correlations of quantitative magnetic resonance imaging metrics.

Scatterplots denoting values for selected metrics within each muscle group. Spearman correlation coefficients and associated p-values are presented separately for CKD, ESKD, and control participants. n=19 for all metrics except T1ρ and T2 (n=18).

Figure 6.. Skeletal muscle lipid content and metabolite concentrations measured by ¹H-magnetic resonance spectroscopy.

Mixed effects models including random intercepts were fit to account for clustering within an individual. Pairwise comparisons of CKD and control participants within muscle groups were performed using tests of contrast. Concentrations are presented referenced to the water peak. n=9 control, n=6 CKD. Abbreviations: IMCL, intramyocellular lipid content; EMCL, extramyocellular lipid content; TMA, trimethylamine. *p<0.05, **p<0.01, ***p<0.001.