Imaging Outcome Measures in Neuromuscular Diseases

Imaging technologies, with uses ranging from diagnostics to disease progression monitoring to assessment of treatment effects, play a key role in the evaluation of neuromuscular diseases. For neurologists, knowledge of neuromuscular imaging outcome measures is imperative for accurate patient assessment, effective integration into care strategies, and understanding of clinical trial results. 

This article focuses on the ability of MRI to assess 4 metrics: replacement of muscle with fat, muscle size, muscle “edema,” and nerve changes. Their relative importance depends on the clinical context. For example, replacement of muscle with fat is a key measure in facioscapulohumeral dystrophy (FSHD), muscle size in motor neuron disease, muscle edema in inflammatory myositis, and nerve imaging in acquired neuropathies.

The aim of this review is to equip neurologists with a practical understanding of neuromuscular MRI—including optimal sequence selection, anatomic coverage, and imaging intervals—to increase the integration of imaging into clinical practice and improve understanding of neuromuscular disease research using imaging outcome measures.

Replacement of Muscle with Fat 

The replacement of muscle tissue with fat, sometimes referred to as fat infiltration, is a hallmark of chronic muscle degeneration, whether attributable to denervation or a myopathic process. The pattern of muscle involvement on MRI scans can play a crucial role in diagnosis of inherited myopathies, as distinct patterns of muscle replacement with fat have been identified. For example, in collagen VI (Ullrich and Bethlem) myopathies, a characteristic “concentric fat replacement with central sparing” pattern is observed.¹ Whereas standard T1-weighted images allow this qualitative assessment of muscle replacement with fat, MRI can precisely quantify fat percentage by exploiting the different hydrogen resonant frequencies in fat and water molecules, as is used in the Dixon method. These techniques offer robust and reproducible measurements of muscle replacement with fat and have been shown to be a responsive measure of disease progression in a wide range of neuromuscular disorders.

In Duchenne muscular dystrophy (DMD), replacement of muscle with fat follows a characteristic pattern, beginning in proximal lower limb muscles and progressing distally.^2,3 MRI studies of lower limb muscles have highlighted correlations between fat fraction and motor function scores, demonstrating validity as an outcome measure.⁴

In FSHD, changes in fat fraction measurements have proven to be responsive outcome measures in clinical trials, such as those investigating ACE-083 and losmapimod.^5,6 However, heterogeneity in muscle involvement presents a challenge for standardizing imaging biomarkers and selecting uniform outcome measures which can complicate longitudinal monitoring in FSHD. Scapular stabilizers, hamstrings, and tibialis anterior commonly exhibit early changes,^7,8 and a whole-body protocol is used to capture all relevant muscles with intermediate fat replacement (10%–50%), which are the most vulnerable to disease progression.^9,10

The pattern of fat replacement in neuropathies presents distally in a reticular “muscle islands” pattern and differs from the more homogeneous patterns seen in muscular dystrophies.¹¹ In Charcot-Marie-Tooth disease, early fat infiltration is prominent in the calf and foot muscles and correlates with functional impairment. Longitudinally, calf muscle fat fraction is a more responsive biomarker than any clinical measure. This has been demonstrated for all common CMT subtypes (eg, CMT1A, CMT1B, CMTX1, CMT2A).^12–16a

Fat fraction measurements provide a direct measure of disease severity and progression across a wide range of neuromuscular disorders. The optimal coverage and analysis method depends on the disease distribution. Muscle fat fraction is used as an outcome measure in a number of clinical trials across a wide range of neuromuscular disorders.

Muscle Size

Quantifying muscle size is a fundamental component of imaging outcome measures and can be undertaken on any sequence with good tissue distinction. Muscle atrophy often reflects disease progression and correlates with functional impairment, making it an important marker for clinical and research purposes.¹⁶ Axial T1-weighted MRI is effective for capturing muscle morphology, whereas 3D imaging, including Dixon imaging, can provide precise volumetric measurements that surpass traditional clinical assessments.¹⁷

In DMD, MRI scans reveal both atrophy and compensatory hypertrophy, with muscle size changes not consistently reflecting muscle strength.¹⁸ In contrast, in motor neuron disease, muscle volume quantification by MRI is more sensitive than the Amyotrophic Lateral Sclerosis Functional Rating Scale score, which makes it a more objective biomarker.¹⁹ This is particularly evident in tracking atrophy in weak muscles which show progressive volume loss. A recent trial demonstrated the potential of longitudinal MRI to detect muscle cross-sectional area reductions as early as at 1 month, highlighting its role in tracking progression and assessing therapeutic efficacy.²⁰

Challenges in standardizing protocols and managing variability persist, but advances such as automated segmentation and artificial intelligence (AI)–based analysis could enhance the accuracy and efficiency of muscle size quantification.²¹

Muscle Edema

Muscle edema is a broad term used to describe changes in muscle water distribution and may be seen in a wide variety of processes, including active inflammation (ie, myositis) and acute denervation, but is also observed in muscular dystrophies (eg, DMD, sarcoglycanopathies, dysferlinopathies).^21a This dynamic feature of various neuromuscular diseases often precedes chronic changes, such as muscle replacement with fat. Muscle edema serves as an early marker of disease activity and a tool for monitoring therapeutic responses. T2-weighted MRI sequences with fat suppression, such as short tau inversion recovery (STIR), are effective for visualizing changes in muscle water distribution, with emerging T2 quantification techniques providing precise and longitudinal assessments.²²

In inflammatory myopathies, such as dermatomyositis, inclusion body myositis, immune-mediated necrotizing myopathy, and anti-synthetase syndrome, edema reflects active inflammation. Dermatomyositis typically shows symmetric pelvic and shoulder girdle involvement with a reticular honeycomb pattern, whereas inclusion body myositis displays asymmetric distal quadriceps and finger flexor edema, which are detectable before clinical onset. High STIR signal in the muscle fasciae may indicate anti-synthetase syndrome.²²

Edema is also observed in DMD and serves as a biomarker of DMD disease activity and progression. Studies have demonstrated that muscle edema can predict disease progression²³ and is responsive to corticosteroid treatment, highlighting its potential as a marker for therapeutic efficacy.²⁴ In limb-girdle muscular dystrophies (LGMDs), such as LGMD-R2 (ie, dysferlinopathy) and LGMD-R12 (ie, anoctamin-5 deficiency), STIR changes are more localized compared with inflammatory myopathies.^1,22,25 In addition, transient edema can occur in FSHD and juvenile dermatomyositis during exacerbations or after exercise.^7,26 Although trends in edema correlate with functional outcomes, its utility as a trial endpoint is limited by confounding variables, reducing its reliability for assessing therapeutic effects.

Nerve Changes

Peripheral nerve imaging requires extremely high resolution to visualize the small structures within peripheral nerves and detect subtle morphologic changes, making it a challenging yet crucial area of study. Advances in magnetic resonance neurography techniques have positioned nerve imaging as an invaluable complement to muscle imaging in peripheral nerve disorders, particularly in focal nerve lesions, chronic inflammatory demyelinating polyneuropathy, and hereditary neuropathies.²⁷

High-resolution magnetic resonance neurography, including diffusion tensor imaging (DTI), enables detailed visualization of the brachial and lumbosacral plexuses, as well as peripheral nerves, revealing structural abnormalities and pathologic changes. T2-weighted sequences with fat suppression are commonly used to detect nerve edema or inflammation, whereas DTI provides quantitative insights into nerve microstructure. Emerging techniques (eg, magnetization transfer ratio imaging and myelin water fraction analysis) are being investigated for their ability to quantify myelin integrity in vivo to differentiate demyelination versus axonal degeneration.²⁷

In Charcot-Marie-Tooth disease, nerve imaging consistently shows nerve hypertrophy and hyperintensity, along with decreased magnetization transfer ratio and altered diffusion metrics on DTI. These findings correlate with disease and neuropathy severity as measured by nerve conduction studies, offering potential biomarkers for monitoring disease progression^{27,28, 28a} However, to date, there are no longitudinal studies to determine the responsiveness of these measures. In chronic inflammatory demyelinating polyneuropathy, nerve and plexus thickening and edema are indicative of active inflammation or demyelination, which can guide therapeutic decisions.^29-31

Practical Considerations

The integration of muscle MRI into clinical practice requires careful correlation with clinical measures, including the presentation, examination findings, nerve conduction studies, EMG, creatine kinase levels, relevant serologic tests (eg, myositis-specific antibodies), and the results of genetic tests. Clinical context is essential for the interpretation of imaging findings, ensuring that MRI data complement rather than replace traditional diagnostic tools. Neurologists should also recognize that muscle MRI can reveal additional clinically relevant findings, such as fascia or skin abnormalities in inflammatory myopathies or incidental nerve changes that may refine diagnostic impressions. The Table includes an overview of key imaging findings for various neuromuscular diseases.

As with any imaging modality, the key questions include “What sequences should be performed?” and “What coverage is required?” For clinical purposes, the standard sequences recommended are axial T1-weighted sequences and fat-suppressed T2-weighted sequences, such as STIR. T1-weighted sequences characterize chronic muscle changes (eg, fat replacement and muscle atrophy), whereas the STIR sequence is sensitive to muscle water changes, which reflects disease activity. 

Although MRI is sensitive, it may lack specificity. The tendency to label any STIR hyperintensity in muscle as myositis should be avoided. STIR hyperintensity also may be seen in muscular dystrophies and with acute denervation, so integration with all clinical information is crucial. Similarly, fat infiltration and muscle atrophy can be seen on MRI in any chronic neuromuscular disease, but the pattern of involvement within and between muscles may give clues as to the underlying etiology. Quantitative sequences, such as Dixon, and the use of gadolinium contrast are generally not necessary for routine clinical scans, but may have applications in specific clinical scenarios.

The optimal anatomic coverage of imaging depends on the distribution of the disease. For instance, in inflammatory myopathies, pelvic and thigh muscle coverage is prioritized. In muscular dystrophies, the pattern of thigh and calf muscle involvement is most commonly reported in the literature and therefore provides the most useful diagnostic information. However, in myopathies where the predominant weakness is outside the lower limbs, additional anatomic coverage can be helpful, such as the periscapular muscles in FSHD. Whole-body muscle MRI is less widely used in standard practice in the United Kingdom due to limited added value, longer imaging times, and tolerance concerns, particularly in those with severe weakness or respiratory compromise.

Longitudinal imaging can provide valuable insights into disease progression and treatment response in both clinical practice and clinical trials. The timing and frequency of repeat scans should be tailored to the clinical context, balancing the need for detailed tracking against concerns about cost and accessibility. Challenges such as motion artefacts and intercenter technical variability must be addressed to ensure reliable longitudinal data, and neurologists should weigh the practical limitations of MRI against other modalities. Although MRI is particularly valuable for assessing deeper muscles and nerves, ultrasound is an alternative imaging modality which provides high-resolution assessment of superficial structures, making it useful for evaluating peripheral nerves and detecting changes in more acessible muscle groups. Ideally, MRI outcome measures should use quantitative sequences such as fat quantification using Dixon or similar sequences, and T2 mapping to measure disease activity. These methods are particularly valuable in clinical trials because they are highly sensitive to pathologic changes, enabling early detection of disease progression and treatment effects. However, using these methods in clinical practice has historically been limited by the requirement for extensive postacquisiton analysis, including time-consuming manual muscle segmentation. Emerging AI-driven segmentation tools hold promise in reducing the time required for postacquisiton analysis by enabling automated fat quantification and large-scale data analysis, while also reducing observer bias and enhancing reproducibility.

Muscle MRI results have been used as secondary outcome measures in all phases of clinical trials and as primary outcome measures in phase 2 clinical trials. Further evidence is needed for MRI results to be fully accepted as surrogate outcome measures. MRI outcome measures should be optimized by using a trial design that maximizes the size of the expected change through knowledge of the natural history of the disease on imaging as well as measurement reliability through careful site setup, quality control, and unbiased analysis. If chosen and executed well, imaging biomarkers may offer the elusive combination of both validity and responsiveness. 

Alternative imaging modalities, such as ultrasound and positron emission tomography, offer complementary advantages. Ultrasound is portable, accessible, and cost-effective, with applications in muscle architecture visualization and functional assessment. However, its operator dependency limits reproducibility compared with MRI. Other imaging modalities are emerging as potential biomarkers for neuromuscular diseases (eg, ultrasound shear wave elastography, functional ultrasound, bioelectrical impedance), but they require further validation for widespread use. The Figure includes examples of multimodal MRI scans.

Figure. Examples of multimodal MRI scans (T1-weighted, T2-weighted, short tau inversion recovery [STIR], and fat fraction [FF] maps) of muscle replacement with fat (T1-weighted thigh and FF maps), muscle size (Dixon acquisition thigh), and muscle edema (T2-weighted STIR calf and quantitative T2 mapping calf). Examples of healthy (blue top row) compared with diseased (red bottom rows) muscle at baseline (red row, top) and interval progression (red row, bottom).
Abbreviations: CSA, cross-sectional area; MG, medial head of gastrocnemius.

Conclusion

Muscle MRI has transformed the evaluation of neuromuscular diseases, providing detailed insights into muscle and nerve pathology that complement data from other diagnostic tools. Challenges including cost, accessibility, and the need for standardization remain, but advances in quantitative imaging and AI-driven analysis promise to enhance reliability and reproducibility. By measuring muscle size, fat infiltration, edema, and nerve abnormalities, MRI enables precise monitoring of disease progression and therapeutic effects. As technology continues to evolve, muscle MRI is poised to play an increasingly central role in advancing patient care and neuromuscular research.