Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis

Clinical Neurophysiology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph

Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis A. Grimm a,⇑, T. Prell b, B.F. Décard a, U. Schumacher c, O.W. Witte b, H. Axer b,1, J. Grosskreutz b,1 a

Department of Neurology, University Hospital Basel, Basel, Switzerland Hans Berger Department of Neurology, Jena University Hospital, Jena, Germany c Center for Clinical Studies (CCS), Jena University Hospital, Germany b

a r t i c l e

i n f o

Article history: Accepted 27 June 2014 Available online xxxx Keywords: Myosonography Amyotrophic lateral sclerosis Electromyography Neuromuscular Ultrasound

h i g h l i g h t s  Muscle ultrasonography (MUS) is highly sensitive in detecting fasciculations—even in the muscles of

full strength—and fibrillations, especially in the bulbar muscles of patients with amyotrophic lateral sclerosis (ALS).  MUS shows significantly increased echo intensity and fasciculations in clinically and electrophysiologically affected and unaffected muscles in ALS patients.  When combined with electromyography (EMG), MUS can provide additional information about specific muscles and increase the diagnostic yield.

a b s t r a c t Objective: We aimed to determine the utility of muscle ultrasonography (MUS) in addition to electromyography (EMG) in the diagnosis of amyotrophic lateral sclerosis (ALS). Methods: In all, 60 patients with ALS and 20 with other neuromuscular disorders underwent MUS and EMG. In addition, 30 healthy controls underwent only MUS. Occurrence of fasciculations and fibrillations was evaluated. Ultrasonic echogenicity was graded semiquantitatively. Results: The incidence of fasciculations was significantly higher in patients undergoing MUS than in those undergoing EMG (p < 0.05), even in muscles of full strength (p < 0.001). However, EMG was more sensitive in detecting fibrillations (p < 0.05). MUS had an overall higher sensitivity in detecting spontaneous activity in the tongue (p < 0.05). Patients with ALS showed significantly increased muscle echo intensity (EI) compared to patients who were initially suspected as having ALS and normal controls (p < 0.05), irrespective of the clinical or electrophysiological status. Conclusion: Our results showed that the sensitivity and specificity of MUS in diagnosing ALS was almost equivalent to those of EMG, using the Awaji criteria. Combination of MUS and EMG enhances the diagnostic accuracy compared to EMG alone (p < 0.05). Significance: The combination of EMG and MUS can be used to evaluate the lower motor neuron affection by reducing the use of the often painful and uncomfortable EMG examinations but without decreasing the diagnostic sensitivity and specificity. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Abbreviations: ALS, amyotrophic lateral sclerosis; ALSFRS-R, ALS functional rating scale in its revised form; APB, abductor pollicis brevis muscle; BB, biceps brachii muscle; EI, echo intensity; EMG, electromyography; ER, extensor muscles of the forearm; MND, motor neuron disease; MMN, multifocal motor neuropathy; MUNE, motor unit number estimation; PSW, positive sharp wave; RA, rectus abdominis muscle; RF, rectus femoris muscle; SMA, spinal muscle atrophy; TA, tibialis anterior muscle. ⇑ Corresponding author. Address: Department of Neurology, Basel University Hospital, University of Basel, Petersgraben 4, CH-4031 Basel, Switzerland. Tel.: +41 61 5565130. E-mail address: [email protected] (A. Grimm). 1 Contributed equally to the work. http://dx.doi.org/10.1016/j.clinph.2014.06.052 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

2

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx

1. Introduction

2. Materials and methods

Diagnosis of amyotrophic lateral sclerosis (ALS) is based on the anamnesis of progressive muscular weakness, clinical examination, and exclusion of other conditions. Various clinical onsets and courses often delay the diagnosis of ALS (Kraemer et al., 2010). However, early diagnosis is important for the initiation of treatment and enrollment into clinical trials. Electromyography (EMG) is essential to identify the loss of lower motor neurons (LMN), typically seen as fibrillations and positive sharp waves (PSWs), i.e., pathological spontaneous activity. Another typical phenomenon in ALS is the presence of fasciculations, which are not classified as pathological spontaneous activity when present in healthy muscles (Brooks, 1994). However, with the further revision of the revised El-Escorial criteria (Brooks et al., 2000) in 2006 to the Awaji criteria, detection of fasciculations has become an important step in the diagnosis of ALS (Hardiman et al., 2011). Fasciculations are similar to PSWs and fibrillations if they occur in muscles showing neurogenic changes (De Carvalho et al., 2008). This modification increases the sensitivity of diagnosis without increasing the false-positive rate (de Carvalho and Swash, 2009; Boekestein et al., 2010). A meta-analysis showed that the use of the Awaji criteria increased the number of patients diagnosed with probable or definitive ALS by up to 23%, without a decrease in specificity, compared with that diagnosed using the revised ElEscorial criteria (Costa et al., 2012) that emphasizes on the importance of refined electrodiagnostic studies. However, in daily clinical routine, EMG is invasive, painful, and time consuming (Mills, 2011). In recent years, muscle ultrasonography (MUS) has become a promising tool for diagnosing neuromuscular disorders (Pillen et al., 2008; Boon et al., 2012; Mayans et al., 2012; Grimm et al., 2013). In this study, we employed both MUS and EMG to diagnose ALS and other neuromuscular disorders (ALS-mimicking disorders). Healthy controls underwent only MUS. The primary end points were detection of spontaneous activity on MUS in one hand and evaluation of muscle structure on the other. The aim was to determine whether MUS was more sensitive than EMG for the detection of fasciculations, as demonstrated previously (Walker et al., 1990; Reimers et al., 1996a; Scheel and Reimers, 2004; Misawa et al., 2011), and whether fibrillations could also be detected using MUS, as suggested by some studies (Dengler, 2009; Pillen et al., 2009a; Pillen and van Alfen, 2011; van Alfen et al., 2011), which would enhance the role of MUS in the diagnosis of neuromuscular disorders and facilitate the diagnosis of ALS. The study also intended to evaluate whether MUS would determine fasciculations as the leading symptom of ALS in clinically and/or electrophysiologically unaffected muscles. To determine the degradation of muscle structure, we focused on echo intensity (EI) as a diagnostic parameter because several studies have shown that affected muscles show structural changes in different aspects of myosonography compared to healthy muscles and clinically unaffected muscles having normal strength (Pillen et al., 2007, 2008; Walker et al., 2004; Arts et al., 2008, 2010, 2011a, 2012). Muscle atrophy, fatty infiltration, and intramuscular fibrosis can be detected using ultrasonography (Pillen et al., 2009b). The differences between patients with ALS, patients with other neuromuscular disorders, and healthy controls were analyzed. The second end point of the study was to determine whether the sensitivity and specificity of the combination of EMG and MUS (included in a reasonable clinical program) was equivalent to that of EMG alone.

2.1. Subjects In all, 80 patients with suspected ALS who were aged 18– 90 years and who visited our hospital or the ALS special unit between October 2011 and March 2013 were included in this observational study. Patients underwent clinical examination, EMG, and MUS. The study also included 30 healthy age-matched controls who underwent the same MUS protocol as the patients. The study was registered with the German Clinical Trials Register (DRKS-ID: DRKS00004322) and was approved by the local ethics committee (No. 3519-07/12). Informed consent was obtained from all patients and controls before their inclusion in the study. 2.2. Ultrasonography MUS was performed using real-time linear array scanner (Siemens Acuson, Erlangen, Germany) with a 9–13-MHz probe. Initial settings (such as contrast) were kept constant for all examinations, except depth, which was changed for each examination to observe the complete muscle, e.g., rectus femoris (RF). Different muscles of the upper (cervical) and lower limbs (lumbar), muscles of the trunk (thoracic), and bulbar muscles were examined bilaterally (biceps brachii muscle [BB], extensor [ER] muscles of the forearm, rectus femoris [RF] muscle, tibialis anterior muscle [TA], rectus abdominis muscles [RA], and the tongue). Four anatomical regions were defined (bulbar = tongue, cervical = BB and ER, thoracic = RA, and lumbar = RF and TA). Patients were examined in the supine position, with arms and legs extended and muscles relaxed as recommended (Reimers and Kellner, 1996b; Scholten et al., 2003; Arts et al., 2010). Muscles were scanned in the axial and longitudinal planes, and each muscle was measured at the standardized anatomical points; in detail: BB and RF at the midline between the origin and insertion, ER at the first third of the distance between the elbow and processus styloideus radii, TA at the first third of the distance between the knee and malleolus lateralis, and RA periumbilical at 2 cm laterally to the midline. The tongue was examined from the submandibular direction, and patients were asked to stop swallowing and breathing for 10 s.

Fig. 1. Ultrasonic cross sections through the left tibialis anterior muscle (TA) showing a fasciculation using M-mode.

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx

One major parameter was the ultrasonic detection of spontaneous activity in the muscle. Each muscle was scanned over a 10-s period at 3 different sites, without moving the probe as described before (Reimers and Kellner, 1996b; Scheel and Reimers, 2004). Fasciculations were visible as short twitches of muscle fibers. They were registered only if they appeared at least twice. Fasciculations can be easily distinguished from other movements such as arterial pulsations (unifocal, rhythmic) or voluntary movements, which involve the entire muscle (Arts et al., 2011a). Fasciculations were stored as videos by using M-mode (Fig. 1). Fibrillations are short, fast vibrations of single muscle fibers, with a worm-like movement of the larger parts of the muscle appearing frequently (Pillen et al., 2009a). They were classified as pathological spontaneous activity if they appeared during the whole analyzing period (3 times 10 s) even in muscles showing normal EI. In our initial experiments, when EMG was performed in parallel with MUS, such ultrasonic worm-like muscle movements were classified as fibrillations. However, MUS and EMG were not simultaneously performed in this study for practical reasons. Misinterpretation of pseudotremor as a fibrillation was avoided by relaxing patients and by avoiding the measurement of nearby bones, as recommended in the literature (Pillen et al., 2009a). The second parameter was ultrasonic echogenicity of the muscle. Compression of the tissue and oblique scanning were avoided, which may have increased the echotexture of the muscle (Reimers and Kellner, 1996b). Modifications in transducer frequency and signal gain were strictly prevented. EI was graded according to that mentioned by Heckmatt et al. (1982). This score differentiates ultrasonic echogenicity semiquantitatively into 4 grades (Table 1, Fig. 2). Increased EI mainly indicates fatty degeneration, edema, fibrosis, and/or inflammation (Pillen et al., 2008). However, it is unclear whether a higher grade of echotexture along with reduced or lost bone signal is correlated with the severity of pathological muscle impairment in ALS (Arts et al., 2011b). We did not measure muscle atrophy parameters such as diameter, area, and volume of the muscles as well as homogeneity because we intended to find a fast and easily applicable procedure. These parameters may be more important when focusing on aspects of the disease course, which was not the aim of this study. Analysis of MUS data was done online as well as offline. The examiner was blinded for the supposed diagnosis. Ultrasonic examination of each patient required approximately 25 min. A second examiner evaluated all the measurements offline for the second time. Both the examiners were blinded to the electrophysiological measurements of the patients. 2.3. Electromyography In all, 70 patients (52 with ALS and 18 with other neuromuscular disorders) underwent standardized needle EMG using a standard electroneurophysiological device (Synergy 15.0, VIASYS Healthcare UK Ltd.). Ten patients did not undergo EMG because of anticoagulation, lack of cooperation, or being unable to tolerate the pain of examination. The TA, RF, abductor pollicis brevis muscle (APB), BB, tongue, and paravertebral muscles of the thoracic spine on different sides were subjected to standardized investigation with EMG. We defined the 4 anatomical regions according to the

3

El-Escorial criteria as follows: bulbar = tongue, cervical = BB and APB, thoracic = paravertebral muscles, and lumbar = RF and TA muscles. EMG was performed by an experienced neurologist/ neurophysiologist. Spontaneous activity such as PSWs or fibrillations and fasciculations were documented. Spontaneous activity was examined for 1 min at 3 different locations of each muscle because the probability of detecting fasciculation potentials increases with the examination period (Mills, 2011). At least 20 MUAPs were analyzed quantitatively in each muscle with respect to mean amplitude, mean duration, mean area, and number of turns and phases, as described in literature (Buchthal and Pinelli, 1952; Stålberg et al., 1996). In addition, recruitment and interference patterns were analyzed. Other sophisticated EMG techniques such as motor unit number estimation (MUNE), single fiber analysis, or motor unit stability analysis described by de Carvalho and Swash (2013) were not performed. 2.4. Interpretation and comparison of both the techniques According to the Awaji criteria, fibrillations and PSWs were always classified as pathological spontaneous activity; fasciculations were classified as pathological spontaneous activity only if the muscle also showed chronic neurogenic changes. By using MUS, chronic denervation was determined as increased echogenicity grade 2 according to the Heckmatt score (Heckmatt et al., 1982). By using EMG, chronic denervation was determined as mentioned above. According to these criteria, we defined fasciculations as a sign of LMN affection in MUS as well as in EMG (De Carvalho et al., 2008). 2.5. Statistical analysis Data were analyzed using SPSS Statistics for Windows (Version 19.0 Armonk, NY). Mann–Whitney U test was used to calculate statistical differences with respect to age between men and women, between ALS and ALS-mimicking disorders, between patients with ALS and controls, and between disease durations in both the patient groups. Differences in the detection of fasciculations and fibrillations between MUS and EMG were calculated using the McNemar test for all comparable muscles. Diagnostic differences in EI and MUAP analysis between ALS and ALS-mimicking disorders and between patients with ALS and controls were calculated using ANOVA. Statistical differences between EMG, MUS and the combination of the 2 methods with respect to the Awaji criteria were calculated using the McNemar test. For all the tests, the statistical significance was set at p < 0.05. Pearson’s correlation was used to calculate the correlation between muscle echogenicity, disease severity, as assessed by the ALS-Functional Rating Scale in its revised version (ALSFRS-R) and disease duration. Intraclass correlation coefficients (ICC) were calculated to evaluate interrater and intrarater reliability for the detection of fibrillations and fasciculations with MUS as well as for the EI. 3. Results 3.1. Subjects

Table 1 Heckmatt Score: visual grading scale to classify muscle echo intensity in ultrasound (Heckmatt et al., 1982). Grading

Characteristics

Grade Grade Grade Grade

Normal echo intensity with starry-night aspect Increased echo intensity with distinct bone signal Increased echo intensity with reduced bone signal Increased echo intensity and loss of bone signal

1 2 3 4

In all, 80 patients were included in the study. Of these, 60 patients (38 men and 22 women; age range, 27–90 years; mean age, 63.2 ± 11.2 years; mean ALSFRS-R, 35.6 ± 8.7; mean disease duration, 27.0 ± 12.5 months) were clinically and electrophysiologically diagnosed as having ALS according to the Awaji criteria. The remaining 20 patients (12 men and 8 women; age range, 18– 84 years; mean age, 62.6 ± 11.8 years; mean disease duration,

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

4

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx

Fig. 2. Ultrasonic cross sections through the left tibialis anterior muscle (TA) showing different echogenicity: (A) normal (starry night) (healthy control), (B) increased echo intensity with normal bone echo (patient with polyneuropathy), (C) increased echo intensity with reduced bone signal (ALSFRS-R 46), (D) increased echo intensity and loss of bone signal (ALSFRS-R 10).

27.2 ± 10.8 months) showing clinical symptoms similar to those of ALS (ALS-mimicking disorders) were diagnosed as having other neuromuscular disorders (3 patients with axonal polyneuropathy, 2 patients with generalized and 1 patient with bulbar myasthenia gravis, 3 patients with cervical myelopathy, 1 patient with cramps, 2 patients with inclusion body myositis, 2 patients with Kennedy syndrome, 2 patients with spinal muscular atrophy [SMA], 2 patients with polymyositis, 1 patient with plexus neuritis, and 1 patient with multifocal motor neuropathy [MMN]) according to their clinical course, response to therapy, and additional diagnostic tests such as blood samples, genetic testing, or muscle/nerve biopsies. The 30 healthy controls (15 men and 15 women; age range, 18– 82 years; mean age, 59.8 ± 15.2 years) underwent the same MUS protocol as the patients. There were no significant differences between men and women with respect to age (Mann–Whitney U test, p = 0.19) and between patients with ALS and ALS-mimicking disorders (p = 0.92) and controls (p = 0.53). Moreover, the mean disease duration was not significantly different between patients with ALS and those with ALS-mimicking disorders (p = 0.98). The clinical characteristics along with their mean and standard deviation (SD) are shown in Table 2. 3.2. Spontaneous activity in MUS and EMG The interrater ICC for offline MUS measurements of muscles was 0.982 for fasciculations and 0.690 for fibrillations. The intrarater ICC was 0.996 for fasciculations and 0.833 for fibrillations.

3.2.1. Fasciculations Both EMG and MUS detected fasciculations in most patients. MUS detected fasciculations in 292 out of 391 (74.7%) examined muscles of the limbs and the trunk and EMG detected fasciculations in 182 out of 335 (54.3%) muscles. Comparison of corresponding muscles (BB, RF, and TA) in patients with ALS showed that MUS detected fasciculations in 137 out of 156 (87.9%) examined muscles while EMG detected fasciculations in 108 (69.2%) muscles, with an overall significant difference between both the methods (McNemar test p < 0.05). These differences were significant for TA and RF (McNemar test, p < 0.01 and p = 0.022) and not significant for BB (p = 0.057). MUS detected fasciculations in the RA of 33 (56.9%) patients, while EMG detected fasciculations in the paravertebral muscles of 15 (28.8%) patients (results are shown in Table 3). In patients with ALS, 37 out of 233 examined muscles (BB, ER, RF, and TA) showed a normal strength (15.9%), defined as Medical Research Council grade force 5/5. The tongue and RA were not analyzed because muscle strength was not tested in these muscles. MUS detected fasciculations in 30 of these 37 (81.1%) notatrophic muscles showing normal strength while 12 of these muscles with fasciculations showed a normal EI. In all, 32 of these muscles were analyzed using both the techniques. EMG detected fasciculations in 16 of the 32 muscles, and MUS detected fasciculations in 27 of the 32 muscles, with MUS showing high significance in detecting fasciculations (50% vs. 84.4%; McNemar test, p < 0.001; Table 3).

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

5

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx Table 2 Study population. n

Mean age (y)

Mean disease duration (m)

Mean ALSFRS-R

Male: female ratio

ALS (all)

60

15

Spinal onset

45

ALS mimicking disorders

20

35.6 ± 8.7 (range 13–48) (median 36.0) 33.2 ± 10.0 (range 13–46) (median 34.0) 36.5 ± 8.2 (range 14–48) (median 39.0) –

12:8

Controls

30

27.0 ± 12.5 (range 3–73) (median 18.0) 29.9 ± 11.9 (range 4–60) (median 20.0) 25.9 ± 13.1 (range 3–73) (median 18.0) 27.2 ± 10.8 (range 6–54) (median 24.0) –

38:22

Bulbar onset

63.2 ± 11.2 (range 27–90) (median 66.0) 62.7 ± 11.9 (range 39–90) (median 70.0) 63.4 ± 11.1 (range 27–81) (median 67.0) 62.6 ± 11.8 (range 18–84) (median 66.5) 59.8 ± 15.2 (range 18–82) (median 63.0)

–

15:15

5:10

33:12

± standard deviation. ALSFRS-R = ALS functional rating scale in its revised form, y = years, m = month.

Table 3 Fasciculations in MUS vs. EMG (52 patients) in ALS.

BB RF TA RA vs. paravert ER vs. APB Tongue 32 Muscles with MRC-grade 5/5

MUS positive

EMG positive

p-Value (McNemar)

45 47 45 33 43 37 27

37 38 33 15 35 27 16

p = 0.057 p = 0.022 p < 0.01 n.d. n.d. p = 0.041 p < 0.001

MUS = muscle ultrasound; EMG = electromyography; BB = biceps brachii; RF = rectus femoris; TA = tibialis anterior; RA = rectus abdominis; ER = extensors of the forearm; APB = abductor pollicis brevis; n.d. = not done. McNemar test was set significant with p < 0.05. Significant results are shown in bold print. Not comparable muscles of the same anatomical regions in MUS and EMG are shown in cursive.

3.2.2. Fibrillations Both the techniques detected fibrillations in patients with ALS and those with ALS-mimicking disorders, with equal results in the majority of patients and controls. MUS detected fibrillations in 120 out of 391 (30.6%) examined muscles while EMG detected fibrillations or PSWs in 152 out of 335 (45.4%) examined muscles. In BB, RF, and TA of patients with ALS, fibrillations were detected in 99 out of 156 muscles by using EMG and in 72 muscles by using MUS (63.5% vs. 46.2%; McNemar test, p < 0.05). EMG detected more fibrillations than MUS in all the muscles, with a significant difference in TA (McNemar test, p < 0.01) and by trend in RF and BB (McNemar test p = 0.169 and p = 0.302). In the tongue, fibrillations and fasciculations were generally regarded as pathological spontaneous activity. MUS detected spontaneous activity in 37 (63.8%) patients while EMG detected spontaneous activity in 27 (51.8%) patients, with significant difference being observed with MUS (McNemar test, p = 0.041).

women (p > 0.05). EI of the tongue was not evaluated. The median EI of each muscle in each group is shown in Table 4. EMG detected chronic neurogenic changes in 161 out of 205 (78.4%) examined muscles (most prominently in BB) in patients with ALS. ANOVA showed no significant differences between patients with ALS and those with ALS-mimicking disorders of neurogenic origin with respect to MUAP of TA (1.9 ± 1.0 vs. 1.3 ± 0.7 mV, p = 0.059), RF (2.1 ± 1.2 vs. 1.5 ± 0.9 mV, p = 0.125), BB (1.6 ± 1.1 vs. 1.1 ± 0.9 mV, p = 0.172), and APB (2.1 ± 1.1 vs. 1.6 ± 0.8 mV, p = 0.11). MUS detected increased EI in 25 out of 32 (78.1%) muscles that were clinically unaffected and showed no signs of chronic or acute denervation in EMG (McNemar test, p = 0.143). A weak negative correlation was observed between echogenicity and ALSFRS-R in the TA ( 0.367, p = 0.009), RF ( 0.301, p = 0.030), and RA ( 0.406, p = 0.003). No correlation was observed between echogenicity and disease duration.

3.3. EI and chronic neurogenic changes in EMG 3.4. Impact on classification by using a combination of MUS and EMG Ultrasonic echogenicity was graded semiquantitatively by using the Heckmatt score (Table 1). Fig. 2 shows some examples of different grades of echo texture in the TA muscle. Evaluation of EI demonstrated an interrater ICC of 0.915 and an intrarater ICC of 0.972. Increased EI was observed in 247 out of 295 (83.7%) analyzed muscles (most prominently in TA and RF) in patients with ALS. EI was significantly higher in patients with ALS and ALS-mimicking disorders than in healthy controls for all the investigated muscles (ANOVA, p < 0.001). EI of TA (p = 0.012) and RA (p = 0.025) was significantly higher in patients with ALS than in patients with ALS-mimicking disorders. In contrast, EIs of ER (p = 0.163), BB (p = 0.060), and RF (p = 0.063) showed no significant differences. EI was not significantly different between men and

Sensitivity and specificity of diagnosing ALS according to the Awaji criteria by using EMG, MUS, and the combination of both were calculated. Fibrillations and PSWs (detected by EMG) and fasciculations in chronic neurogenically affected muscles were defined as the pathological involvement of the LMN. EMG showed that 45 out of 51 patients had signs of LMN in 2 or more anatomical regions, suggesting that they could be diagnosed as having probable or definitive ALS; 6 patients with ALS did not fulfill these criteria (sensitivity, 88.2%). Two patients with ALSmimicking disorders fulfilled the Awaji criteria for the signs of LMN in EMG; however, none of these patients showed the signs of UMN involvement.

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

6

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx

Table 4 Echo intensity (EI) according to semiquantitative grading (Heckmatt et al., 1982). Group ALS ALS-mimicking disorders Controls p-Value ALS vs. ALS-mimicking disorders p-Value ALS vs. control

Median Number Median Number Median Number

TA

RF

BB

ER

RA

2.62 ± 0.93 58 2.00 ± 0.92 20 1.23 ± 0.45 30 p = 0.012 p < 0.001

2.38 ± 0.78 60 2.00 ± 0.79 20 1.27 ± 0.43 30 p = 0.063 p < 0.001

2.24 ± 0.73 59 1.85 ± 0.93 20 1.03 ± 0.18 30 p = 0.060 p < 0.001

2.02 ± 0.69 58 1.75 ± 0.85 20 1.00 ± 0 30 p = 0.163 p < 0.001

1.98 ± 0.76 58 1.50 ± 0.78 18 1.03 ± 0.18 30 p = 0.025 p < 0.001

80 Patients (60 ALS, 20 ALS-mimics) and 30 healthy control persons got muscle ultrasound (MUS) examination of the TA, RF, BB, ER and RA. Median EI (± standard deviation) of different muscles differs in each group with highest median EI in ALS patient in each muscle followed by ALS-mimics and healthy control persons. Significance level of ANOVA was p < 0.05. Significant differences are shown in bold print.

MUS showed that 2 patients with ALS had no spontaneous activity, leading to a false-negative ALS diagnosis; moreover, 4 patients were classified as having possible ALS. The remaining 54 patients with ALS were classified as having probable or definitive ALS (sensitivity, 90.0%). In contrast, 4 patients with ALS-mimicking disorders showed a false-positive ALS diagnosis with respect to LMN affection, without any signs of UMN involvement. In the control group, 4 individuals showed fasciculations in 2 anatomical regions (cervical and lumbar); however, none of them showed fasciculations in the bulbar and thoracic regions and none had fibrillations. Based on the hypothesis that MUS might be more sensitive in detecting acute and/or chronic neurogenic affection in the tongue and RA (compared to paravertebral muscles) and that EMG might be more accurate in detecting acute and/or chronic neurogenic affection, especially fibrillations and PSWs, in the RF and TA, a diagnostic procedure involving a combination of the 2 examination methods, with a reduction in the number of examined muscles (tongue, RA, and ER with MUS; BB, RF, and TA with EMG) was evaluated. The sensitivity and specificity of ALS diagnosis were calculated exclusively by using these measurements. The combination of MUS and EMG diagnosed probable or definitive ALS in 49 out of 51 patients (sensitivity, 96.0%). Four patients detected as having possible ALS with EMG were diagnosed as having probable ALS by the combination of MUS and EMG with a significant difference (McNemar test, p = 0.039). Only 3 out of 20 patients with ALS-mimicking disorders showed the signs of LMN in more than 2 anatomical regions but did not show the signs of UMN. 4. Discussion and conclusion EMG is the technical gold standard for assessing LMN function in ALS (de Carvalho et al., 2008). However, ultrasonography is easily applicable and noninvasive and therefore may serve as an additional diagnostic instrument that can evaluate large muscle regions in a short time. To the best of our knowledge, this is the first study that compared the sensitivity and specificity of EMG with MUS for detecting the signs of acute (spontaneous activity) and chronic denervation in patients with ALS and ALS-mimicking disorders and healthy controls according to the Awaji criteria.

as ultrasonic correlate of structural muscle changes, fasciculation may be a reliable sign of LMN affection similar to EMG. Fasciculations were observed in 16 out of 32 muscles of full strength by using EMG and in 27 out of 32 muscles by using MUS (50% vs. 84.4%). Thus, in our study population, MUS was more sensitive than EMG in detecting fasciculations in clinically unaffected muscles, which may be crucial in diagnosing ALS. De Carvalho and Swash (2013) showed that detection of fasciculations is an important finding in the early stages of disease in muscles of full strength. Thus, the advantage of MUS might be based on its higher sensitivity to detect fasciculations than that of EMG. To our knowledge, the disadvantage of MUS is its inability to differentiate between complex, instable fasciculations and stable fasciculations, which may be an additional criterion in the diagnosis of motor neuron disease and which can be only done using EMG thus far (De Carvalho and Swash, 2013). 4.2. Fibrillations Recently, high-resolution ultrasonography was used to visualize fibrillations (Pillen et al., 2009a; van Alfen et al., 2011). Fibrillations were detected in bulbar, cervical, thoracic, and lumbar muscles of patients with ALS. However, EMG was more sensitive than MUS in detecting fibrillations. Detection of fibrillations by using MUS may be rather limited because fibrillations of single muscle fibers represent only small localized movements in the muscle (in contrast to contractions of whole motor units in case of fasciculations). In previous studies, detection of fibrillations by using MUS has been proven to be less sensitive than that with EMG (approximately 33% vs. 63%), depending on the analyzed muscle (Pillen et al., 2009a; van Alfen et al., 2011). Nevertheless, with further technical advances in ultrasonography devices, the advantage of EMG over MUS with respect to resolution may further decrease (Pillen et al., 2009a). MUS might be superior to EMG for evaluating the spontaneous activity in the tongue because it can scan larger muscle regions than EMG and because use of EMG might be hampered by missing muscle relaxation due to pain sensations. As a limitation, MUS and EMG were not performed simultaneously in this study because results of each method had to be blinded to those of the other method to compare the results. 4.3. EI and chronic neurogenic changes in EMG

4.1. Fasciculations This study showed that MUS was more sensitive in detecting muscle fasciculations than EMG, which was also reproducible with good intra- and interrater ICC. This result is in line with those of previous studies (Reimers et al., 1996a; Misawa et al., 2011). However, fasciculations were also detected in healthy persons, especially in the TA. Nevertheless, in combination with increased EI

Ultrasonic EI was increased in patients with ALS compared to healthy controls and patients with ALS-mimicking disorders. However, it must be considered that muscle EI increases with age (Pillen et al., 2009b). Thus, the differences between patient groups and the healthy control group could have been influenced by slightly increased mean age in patients with ALS and ALS-mimicking disorders.

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx

In contrast to spontaneous activity, ultrasonic EI of the muscle is correlated to the structural muscle changes, for example, due to chronic denervation. It has been well-established that chronic denervated muscles show structural changes in the different aspects of MUS compared to those in healthy controls (Walker et al., 2004; Pillen et al., 2007, 2008; Arts et al., 2008, 2010, 2011a,b, 2012; Grimm et al., 2013). Muscle atrophy, fatty infiltration, and intramuscular fibrosis are clearly detectable. However, the extent of muscular changes in ALS may depend on the stage of the disease and disease severity. A previous study found a correlation between ultrasonic EI and survival in patients with ALS (Arts et al., 2011b). In our study population, EI was correlated with reduced ALSFRS-R. However, a clear differentiation between myogenic and neurogenic changes in the muscles might be difficult by using ultrasonography alone. Myopathic muscles often show a homogeneously increased EI pattern with increased vascularization (Reimers, 2005). In this study, we did not focus on such subtle differences because muscle changes in ALS were clearly neurogenic in origin and the number of patients with myopathic changes in the control group was too small to draw reliable conclusions. Comparison of MUAP amplitude showed no significant differences between patients with ALS and those with ALS-mimicking disorders without the myopathies. In most patients, the signs of chronic denervation in EMG corresponded well with ultrasonic changes in EI. However, ultrasonography showed slight and heterogeneous increased echogenicity in 20% of the muscles that were clinically unaffected and that did not show typical pathological findings on EMG. It can be hypothesized that such patchy changes may be an early sign of motor neuron degeneration. As a caveat, sophisticated EMG [such as single fiber analysis (Cui et al., 2004), MUNE (Bromberg et al., 1993), or motor unit stability studies (Payan, 1978; De Carvalho and Swash, 2013)], which are considered to be useful in evaluating chronic neurogenic muscle damage, were not performed in this study. The overall role of increased EI in diagnosing ALS in the early stages is not yet clear and must be evaluated in further studies. 4.4. Value of the combination of MUS and EMG The major practical conclusion of the study is the suggestion to include ultrasonic parameters into the clinical evaluation of LMN affection according to the Awaji criteria, especially for the bulbar and thoracic regions. The combined examination of 3 muscles (the BB, RF, and TA) by using EMG and 3 muscles (the tongue, APB, und RA) by using MUS increased the diagnosis rate of probable or definitive ALS in our study population compared to examination with EMG only, without a reduction of specificity. Although the direct consequence on the treatment and enrollment of patients into clinical studies would be low, the combination of both the methods could reduce the number of painful and uncomfortable EMG procedures. The Awaji criteria have been criticized because they are based on a consensus meeting and are not supported by evidence (Benatar and Tandan, 2011). However, a meta-analysis by Costa et al. (2012) showed that the Awaji criteria can evaluate MND with higher clinical impact than that evaluated with the revised El-Escorial criteria, without decreasing specificity. Especially, the increased number of patients who could be enrolled in clinical trials is the most impressive and relevant. 4.5. Study limitations The study has several limitations. MUS and EMG were not performed in the same muscles of the upper extremities and thoracic spine (MUS in ER vs. EMG in APB; MUS in RA vs. EMG in

7

paravertebral), and a direct comparison of the distal arm and trunk muscles could not be done. The examination protocol was chosen to achieve convenience in studying different muscles and differed between the methods. The number of patients with ALS-mimicking disorders was relatively small; however, these patients had diseases that had relevant differential diagnosis to ALS, e.g., cervical myelopathy, Kennedy disease, and MMN. Disease duration in some patients was relatively long (mean duration, 27 months) and could have influenced the echogenic changes in the study population as well as the number of patients fulfilling criteria for probable or definitive ALS (90% using MUS) as Misawa et al. (2011) reported lower rates in their study (79% using MUS). In addition, it can be criticized that MUS detected signs of LMN affection in 4 patients with other neuromuscular disorders, thus fulfilling the criteria for probable or definitive ALS diagnosis with respect to LMN pathology. All these patients had other forms of MND (2 with Kennedy disease and 2 with SMA). In these patients, the role of electrophysiology is beyond controversy because experienced neurophysiologist can differentiate between these disorders but ultrasonography probably cannot. For diagnosing ALS, the results of study examinations have to be evaluated with care; therefore, additional information from MUS may be a confirmative. 5. Conclusions Despite its limitations, this observational study showed a promising potential of MUS for diagnosing ALS in addition to other standard neurophysiological examinations. However, further studies are needed. In a succeeding study, identical muscles of all anatomical regions should be evaluated using ultrasonography and distinguished electrophysiological examinations. Moreover, the study population should include patients in the early stages of the disease, with regular follow-up examinations. Therefore, it would be reasonable to address whether additional use of MUS facilitates the early diagnosis of ALS and whether MUS monitors the progression of the disease. Acknowledgments This project is supported by the German Bundesministerium für Bildung und Forschung (BMBF) grant SOPHIA to J.G. under the aegis of the EU Joint Programme – Neurodegenerative Disease Research (JPND – www.jpnd.eu), and a BMBF grant to J.G. within the framework of the ERANET E-Rare program (PYRAMID). The authors thank Nicholas Sanderson and Nasim Kroegel for her help with English language editing. Conflict of interest statement: The authors report no conflicts of interests. References Arts IMP, van Rooij FG, Overeem S, Pillen S, Janssen HMHA, Schelhaas HJ, et al. Quantitative muscle ultrasonography in amyotrophic lateral sclerosis. Ultrasound Med Biol 2008;34:354–61. Arts IMP, Pillen S, Schelhaas HJ, Overeem S, Zwarts MJ. Normal values for quantitative muscle ultrasonography in adults. Muscle Nerve 2010;41:32–41. Arts IM, Overeem S, Pillen S, Schelhaas HJ, Zwarts MJ. Muscle ultrasonography to predict survival in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 2011a;82:552–4. Arts IMP, Overeem S, Pillen S, Schelhaas HJ, Zwarts MJ. Muscle changes in amyotrophic lateral sclerosis: a longitudinal ultrasonography study. Clin Neurophysiol 2011b;122:623–8. Arts IMP, Overeem S, Pillen S, Kleine BU, Boekestein WA, Zwarts MJ, et al. Muscle ultrasonography: a diagnostic tool for amyotrophic lateral sclerosis. Clin Neurophysiol 2012;123:1662–7.

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052

8

A. Grimm et al. / Clinical Neurophysiology xxx (2014) xxx–xxx

Benatar M, Tandan R. The Awaji criteria for the diagnosis of amyotrophic lateral sclerosis: have we put the cart before the horse? Muscle Nerve 2011;43: 461–3. Boekestein WA, Kleine BU, Hageman G, Schelhaas HJ, Zwarts MJ. Sensitivity and specificity of the ‘Awaji’ electrodiagnostic criteria for amyotrophic lateral sclerosis: retrospective comparison of the Awaji and revised El Escorial criteria for ALS. Amyotroph Lateral Scler 2010;11:497–501. Boon AJ, Smith J, Harper CM. Ultrasound applications in electrodiagnosis. PM R 2012;4:37–49. Bromberg MB, Forshew DA, Nau KL, Bromberg J, Simmons Z, Fries TJ. Motor unit number estimation, isometric strength, and electromyographic measurements in amyotrophic lateral sclerosis. Muscle Nerve 1993;16:1213–9. Brooks BR. El Escorial World Federation of Neurology criteria for the diagnosis of amyotrophic lateral sclerosis. Subcommittee on motor neuron diseases/ amyotrophic lateral sclerosis of the world federation of neurology research group on neuromuscular diseases and the El Escorial ‘‘clinical limits of amyotrophic lateral sclerosis’’ workshop contributors. J Neurol Sci 1994;124(Suppl.):96–107. Brooks BR, Miller RG, Swash M, Munsat TL. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler 2000;1:293–9. Buchthal F, Pinelli P. Analysis of muscle action potentials as a diagnostic aid in neuro-muscular disorders. Acta Med Scand Suppl 1952;266:315–27. de Carvalho M, Dengler R, Eisen A, England JD, Kaji R, Kimura J, et al. Electrodiagnostic criteria for diagnosis of ALS. Clin Neurophysiol 2008;119:497–503. de Carvalho MD, Swash M. Awaji diagnostic algorithm increases sensitivity of El Escorial criteria for ALS diagnosis. Amyotroph Lateral Scler 2009;10:53–7. de Carvalho M, Swash M. Fasciculation potentials and earliest changes in motor unit physiology in ALS. J Neurol Neurosurg Psychiatry 2013;84:963–8. Costa J, Swash M, Carvalho M. Awaji criteria for the diagnosis of amyotrophic lateral sclerosis. Arch Neurol 2012;69:1410–6. Cui LY, Liu MS, Tang XF. Single fiber electromyography in 78 patients with amyotrophic lateral sclerosis. Chin Med J (Engl) 2004;117:1830–3. Dengler R. Dynamic muscle ultrasound conquers another domain of needle EMG, the detection of fibrillations. Clin Neurophysiol 2009;120:843–4. Grimm A, Teschner U, Porzelius C, Ludewig K, Zielske J, Witte OW, et al. Muscle ultrasound for early assessment of critical illness neuromyopathy in severe sepsis. Crit Care 2013;17:R227. Hardiman O, Van den Berg LH, Kiernan MC. Clinical diagnosis and management of amyotrophic lateral sclerosis. Nat Rev Neurol 2011;7:639–49. Heckmatt JZ, Leeman S, Dubowitz V. Ultrasound imaging in the diagnosis of muscle disease. J Pediatr 1982;101:656–60.

Kraemer M, Buerger M, Berlit P. Diagnostic problems and delay of diagnosis in amyotrophic lateral sclerosis. Clin Neurol Neurosurg 2010;112:103–5. Mayans D, Cartwright MS, Walker FO. Neuromuscular ultrasonography: quantifying muscle and nerve measurements. Phys Med Rehabil Clin N Am 2012;23:133–48. Mills KR. Detecting fasciculations in amyotrophic lateral sclerosis: duration of observation required. J Neurol Neurosurg Psychiatry 2011;82:549–51. Misawa S, Noto Y, Shibuya K, Isose S, Sekiguchi Y, Nasu S, et al. Ultrasonographic detection of fasciculations markedly increases diagnostic sensitivity of ALS. Neurology 2011;77:1532–7. Payan J. The blanket principle: a technical note. Muscle Nerve 1978;1:423–6. Pillen S, Verrips A, van Alfen N, Arts IMP, Sie LTL, Zwarts MJ. Quantitative skeletal muscle ultrasound: diagnostic value in childhood neuromuscular disease. Neuromuscul Disord 2007;17:509–16. Pillen S, Arts IMP, Zwarts MJ. Muscle ultrasound in neuromuscular disorders. Muscle Nerve 2008;37:679–93. Pillen S, Nienhuis M, van Dijk JP, Arts IMP, van Alfen N, Zwarts MJ. Muscles alive: ultrasound detects fibrillations. Clin Neurophysiol 2009a;120:932–6. Pillen S, Tak RO, Zwarts MJ, Lammens MMY, Verrijp KN, Arts IMP, et al. Skeletal muscle ultrasound: correlation between fibrous tissue and echo intensity. Ultrasound Med Biol 2009b;35:443–6. Pillen S, van Alfen N. Skeletal muscle ultrasound. Neurol Res 2011;33:1016–24. Reimers CD, Ziemann U, Scheel A, Rieckmann P, Kunkel M, Kurth C. Fasciculations: clinical, electromyographic, and ultrasonographic assessment. J Neurol 1996a;243:579–84. Reimers CD, Kellner H. Muscle ultrasound: muscle imaging in health and disease. New York: Springer; 1996b. Reimers CD . In: Kaps M, von Reutern G, et al., editors. Ultraschall in der neurologie, Thieme, vol. 2. Auflage; 2005. p. 248–50. Scheel AK, Reimers CD. Detection of fasciculations and other types of muscular hyperkinesias with ultrasound. Ultraschall Med 2004;25:337–41. Scholten RR, Pillen S, Verrips A, Zwarts MJ. Quantitative ultrasonography of skeletal muscles in children: normal values. Muscle Nerve 2003;27:693–8. Stålberg E, Nandedkar SD, Sanders DB, Falck B. Quantitative motor unit potential analysis. J Clin Neurophysiol 1996;13:401–22. van Alfen N, Nienhuis M, Zwarts MJ, Pillen S. Detection of fibrillations using muscle ultrasound: diagnostic accuracy and identification of pitfalls. Muscle Nerve 2011;43:178–82. Walker FO, Donofrio PD, Harpold GJ, Ferrell WG. Sonographic imaging of muscle contraction and fasciculations: a correlation with electromyography. Muscle Nerve 1990;13:33–9. Walker FO, Cartwright MS, Wiesler ER, Caress J. Ultrasound of nerve and muscle. Clin Neurophysiol 2004;115:495–507.

Please cite this article in press as: Grimm A et al. Muscle ultrasonography as an additional diagnostic tool for the diagnosis of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.06.052