Neuromuscular imaging in inherited muscle diseases

Ultrasound (US) is a well-established and validated diagnostic imaging method in the evaluation of patients with suspected muscle disorders [10]. It is a relatively cheap and easily applicable method, allowing the visualisation of striated muscle with a high temporal resolution (>0.1 mm). The major advantage of US is the lack of any radiation exposure, which makes it the perfect imaging method for the evaluation of children. It even allows dynamic imaging of contracting muscles and can visualise pathological muscle activity such as fasciculation [11‐13].

A major drawback of US is that its application is limited to superficial muscle groups. Sound wave reflection and absorption lead to difficulties in displaying the deeper structures. This effect becomes even more pronounced when multiple muscle groups overlap. Another disadvantage is the relatively low inter-observer agreement and intra-observer agreement (depending on the level of experience), which makes a strictly standardised examination (e.g. according to certain anatomical landmarks) crucial [2, 14]. However, ultrasound is a reliable method concerning the measurement of muscle thickness and muscle echo intensity. In addition, by measuring muscle echo intensity it is possible to further characterise age-related or pathological changes of the striated muscles with special regard to dystrophic changes in terms of fatty degeneration and replacement of muscle by connective tissue (Fig. 1). In order to quantify the degree of fat deposition there are several ratings scales available (e.g. the Heckmatt score) as well as computed-assisted quantification methods of muscle echo intensity [15‐19].

Ultrasound applications are widely and routinely used in neuromuscular disorders in terms of assessment of changes in muscle morphology (atrophy, hypertrophy, changes in muscle architecture). In particular it is a useful screening tool during the initial diagnostic phase, especially in children. Depending on the disease entity, the sensitivity of detecting dystrophic changes ranges from 25% in non-dystrophic myopathies up to 100% in dystrophic myopathies (Duchenne muscular dystrophy) [10, 19, 20]. The detection of pathological changes can be helpful in guiding muscle biopsy, and the description of the muscle involvement pattern might help in the differential diagnosis [10, 21].

Computed tomography (CT) has been widely used in the past in order to evaluate the presence and extent of change in the striated muscles in patients with hereditary neuromuscular disorders [2, 22‐24]. CT is a fast imaging method that is easy to apply and allows a good and standardised assessment of the aspect and shape of the muscles as well as dystrophic changes (in particular fatty degeneration). CT is also relatively operator-independent and allows the evaluation of deeper muscle groups, and newer CT methods using multi-detector rows provide improved imaging possibilities in terms of spatial resolution and multi-planar reconstructions. However, CT has substantial drawbacks leading to an almost complete replacement of this imaging technique by US and MRI. One of the most relevant disadvantages of CT compared with US and MRI is the relatively high radiation dose, which makes the application obsolete, especially in children. Because of the high radiation dose, whole-body applications in order to describe the pattern of muscle involvement are not desired. Another drawback of CT is the limited soft tissue contrast, which substantially impairs the sensitivity in the detection of inflammatory changes (e.g. oedema) that can precede muscle dystrophy.

MRI is increasingly being used in the evaluation of patients with suspected or proven inherited or metabolic neuromuscular disorders. MRI provides a high soft tissue contrast allowing excellent assessment of striated muscles concerning shape, volume (hypotrophy, hypertrophy) and tissue architecture [1, 2]. Because of the lack of ionising radiation, MRI has become a valuable imaging method in children, although sometimes sedation might be necessary. Basically, MRI is performed as a multi-sequence imaging protocol including T1-weighted (T1W) and T2-weighted (T2W) (turbo) spin echo as well as fat-suppressed (short tau inversion recovery or spectral fat suppression techniques) T2-weighted sequences (T2WFS). The image acquisition is performed in the axial plane with a slice thickness of 5-7 mm. If necessary, additional images in other anatomical planes (coronal, sagittal) can be easily acquired.

Dystrophic changes such as fatty degeneration can be easily and sensitively detected using the T1W and T2W sequences. In addition, inflammatory changes such as muscle oedema can be depicted on the T2WFS sequences (Fig. 2). It has been conclusively shown that MRI has a higher sensitivity in the detection of dystrophic changes compared with CT [22, 23]. MRI can be performed and rated in a standardised manner suggesting a good inter-rater agreement and intra-rater (during follow-up) agreement. The degree of muscular dystrophy in inherited muscle diseases is rated according to rating scales [25‐27]. Most of the established rating scales are based on the amount of fatty degeneration ranging from normal appearance to complete fatty degeneration (Table 1). The evaluation of the muscle MRI using standardised rating scales allows a fast and reproducible assessment of the degree of involvement of each muscle. Initially, muscle MRI protocols were developed to evaluate certain anatomical areas such as the lower extremities and pelvis. More recent whole-body imaging protocols have been established allowing the evaluation of almost all relevant striated muscle groups. Pattern recognition of muscle involvement is sometimes helpful in narrowing the differential diagnosis and leading to the most probable diagnosis before muscle biopsy. Because of the lack of any radiation exposure, muscle MRI has become an important diagnostic technique for the evaluation of children [3]. In addition, whole-body MR imaging protocols allow evaluation of tissues and organs beyond the striated muscle such as the parenchymatous organs in the abdomen, the oesophagus and heart, all of which can be affected in patients with inherited neuromuscular diseases [28, 29] (Fig. 3).

Using T2WFS sequences, MRI can sensitively detect discrete toxic, metabolic and inflammatory changes that are reflected by muscle oedema, which normally precedes dystrophic changes in terms of fatty degeneration. Therefore, MRI is able to detect ongoing disease activity in the muscle tissue and can be helpful in guiding muscle biopsy (Fig. 2).

The diagnostic value of contrast-enhanced (CE) MR imaging protocols has not been investigated so far. Late contrast enhancement may be a valuable option for the assessment of connective tissue that replaces normal muscle tissue in degenerative myopathies. However, this assessment is associated with a substantial increase in examination time. Therefore, CE MR imaging should not been performed on a regular basis. Recently performed animal studies with more specific contrast agents entering affected muscle tissue have shown that CE MRI might be helpful in distinguishing normal non-affected muscles from damaged muscles and further describe the extension and mechanism of muscular damage in dystrophic disorders [30, 31]. However, more prospective studies are necessary before the clinical application of such imaging protocols in humans during the diagnostic workup.

Quantitative MRI methods such as T2 relaxation time measurements, muscle fat quantification using the 3-point Dixon technique, magnetic resonance spectroscopy and perfusion imaging might be helpful to further analyse the degree of pathological changes in the striated muscles [32‐34]. In particular blood flow measurements in the striated muscle based on either dynamic contrast-enhanced T1-weighted MRI, arterial spin labelling or blood oxygen-dependent MRI can be helpful to distinguish between inflammatory (increase of microvascular perfusion) and degenerative/dystrophic changes [4, 35]. However, there have been only a limited number of studies with low numbers of included subjects available up to now. Therefore, these techniques remain experimental for this purpose and should not be considered a standard imaging tool in the clinical routine setting.

In western and central Europe, LGMD2I, LGMD2A and LGMD2B are probably the most common LGMD forms. A muscle biopsy including protein expression analysis is usually necessary to distinguish among LGMD subtypes [36, 37]. LGMDs also present with different patterns of muscle involvement on imaging that might help in the genetic diagnosis. An illustration of different muscle imaging findings in LGMD patients is provided in Fig. 4.

MFMs are histopathologically characterised by aberrant desmin aggregation and ultrastructurally by myofibrillar degeneration, which led to the introduction of the term `myofibrillar myopathy’ (MFM). Mutations in the human desmin gene (DES) were first shown to be associated with MFM; another form is associated with mutations in the gene encoding αB-crystallin (CRYAB). More recently, mutations in the human Z-disc proteins myotilin (MYOT), ZASP (LDB3) and filamin C (FLNC) have also been shown to cause MFM [51]. Practically, a definitive determination of the MFM subtype can only be established by direct gene sequencing. However, muscle imaging in combination with clinical information is very helpful for separation of distinct MFM subtypes and in scheduling of genetic analysis.

Recently, we performed muscle imaging in 46 MFM patients (19 desminopathy, 12 myotilinopathy, 11 filaminopathy, 1 αB-crystallinopathy and 3 ZASPopathy patients) and observed two major characteristic patterns of muscle involvement [26]. In desminopathy patients, at the thigh the semitendinosus was the most affected muscle, being more affected than the biceps femoris and semimembranosus muscles. Furthermore, the sartorius and gracilis muscles were often involved earlier and more severely than other thigh muscles. At the lower legs, peroneal muscles were more involved than the tibialis anterior muscle or at least equally as involved. Muscle imaging findings in a patient with a R120G αBC mutation were very similar to those in desminopathy patients. The most frequently involved muscles were the gluteus maximus, semitendinosus, sartorius, gracilis and the peroneal muscles.

Highly contrarily, all other MFM patients (caused by mutations of the Z-disc protein-encoding genes myotilin, filamin c and ZASP) showed an opposite posterior thigh affliction with more involvement of the biceps femoris, semimembranosus and the adductor magnus than the semitendinosus muscle. In filaminopathy, the gracilis and sartorius muscles were often relatively equally spared, while in myotilinopathy the sartorius muscle was slightly more involved than the gracilis muscle. In the lower legs, the soleus and the medial gastrocnemius were the most frequently affected muscles, and in the anterior compartment the tibialis anterior muscle was the most frequently affected. Differences among these three MFM subforms were much more subtle, but detectable on statistical analysis. This revealed highly sensitive and specific criteria (Fig. 5) that may be very useful for the detection of individual MFM forms.

Diagnosis of common muscular dystrophies such as myotonic dystrophy type 1, facio-scapulo-humeral muscular dystrophy and oculopharyngeal muscular dystrophy is usually based on classical clinical findings. Therefore, in these muscular dystrophies the role of muscle imaging has yet to be defined. Recent studies performed on these myopathies suggested that some characteristic findings are present, even if there is less diagnostic value compared with CM, LGMD and MFM (Fig. 6).

CMs show distinctive morphological abnormalities such as rods, an abnormal number of central nuclei or (central or multiple) cores. Often these findings guide the physician to the correct genetic diagnosis. However, some pathological overlap between genetically distinct disorders has been described, making a genetic diagnosis sometimes difficult. Recently, different muscle imaging patterns in CM have been identified, which can be of help in the differential diagnosis.

Muscle imaging in patients with central core disease (CCD) caused by mutations in the ryanodine receptor 1 gene (RYR1) shows a characteristic pattern with main involvement of the gluteus maximus in the pelvis. At the thigh level, the most severe changes are observed in the medial compartment (adductor magnus) and in the anterior compartment muscles (vastus lateralis, vastus intermedius). Sparing of the adductor longus, gracilis and biceps femoris muscles is common. In the distal lower leg muscles, the soleus and the lateral head of the gastrocnemius muscles are most severely involved [61, 62]. This pattern seems to be highly characteristic, as muscular involvement in other congenital myopathies caused by mutations in the SEPN1, ACTA1, NEB or collagen VI encoding (COL6A1, COL6A2 and COL6A3) genes is different [1]. In addition to RYR1, patients with Ullrich CMD or Bethlem myopathy related to collagen VI encoding genes also present with dominant affliction of the anterior thigh compartment. These patients show an early and selective involvement of the centre of the rectus femoris and vastus lateralis muscles. In addition, in the lower legs there is often a rim of degenerative changes between the soleus and the gastrocnemius muscle, a pattern not observed in other CMs [63, 64].

Contrary to RYR1- and collagen 6A-related disorders, which have marked affliction of the anterior thigh, in most other CM forms the posterior rather than anterior thigh compartment muscles are affected. SEPN1 patients, often presenting as a rigid spine (multi-minicore) myopathy, typically show a selective affliction of the sartorius and posterior compartment muscles and a relative sparing of the quadriceps and the gracilis muscle [25]. In patients with dynamin 2-related autosomal dominant centronuclear myopathy (DNM2-CNM), muscle imaging shows predominantly distal lower leg muscle affliction (medial head of the gastrocnemius, soleus), milder involvement of the posterior thigh compartment (mostly the biceps femoris and semimembranosus muscles) and the gluteus minimus muscle. The sartorius and gracilis are often spared [65]. Congenital (nemaline) myopathies related to ACTA1 gene mutations show predominantly distal anterior lower leg and mild diffuse thigh compartment involvement. On the other hand, nebulin gene-associated nemaline myopathies often spare the thigh muscles and show selective distal involvement of the soleus and tibial anterior muscles [66]. A useful flowchart for the differential diagnosis of CM using muscle imaging is provided in Fig. 7.

Compared to the increasing data regarding the diagnostic value of neuromuscular MRI in patients presenting with dystrophic myopathies or non-dystrophic congenital myopathies, data concerning MRI findings in muscle channelopathies or metabolic myopathies are rather limited. A recently published study focussed on patients with myotonia congenital type Becker, a non-dystrophic generalised myotonia caused by mutations in the muscle chloride gene. Although all patients presented with a severe disabling myotonia, no fatty degeneration or muscle oedema could be detected by using a whole-body high-field MRI protocol [67]. These findings suggest that conventional MRI techniques are less sensitive in the detection of changes in channelopathies. Recent experimental studies using ²³Na-MRI instead of or in combination with ¹H-MRI have shed some light on the muscle cell function and disease process in Na-channelopathies by demonstrating Na-accumulation during episodes of weakness [33, 68].