Characteristics of Neuromuscular Ultrasound in Patients with Amyotrophic Lateral Sclerosis

Authors: Jieying Wu, Hongsong Song, Mukadas Arkin, Shuo Zhang, Xiao Huang, Dongsheng Fan, Yingsheng Xu

Abstract

Introduction

Neuromuscular ultrasound has been increasingly used in the detection and diagnosis of amyotrophic lateral sclerosis (ALS), commonly characterized by peripheral nerve atrophy, degeneration, and muscle fasciculations. The aim of this study was to assess the ultrasound characteristics of ALS patients.

Methods

A total of 67 consecutive patients with sporadic ALS and 19 with ALS mimics (63.16% peripheral neuropathy) were recruited. Ultrasound and electrophysiological examinations were performed; the peripheral nerve cross-sectional area (CSA) and fasciculation grades were compared between the groups, and correlations between ultrasound and electrophysiological data in ALS patients were determined.

Results

ALS patients had smaller proximal median and ulnar nerve CSAs than those of ALS mimics, who exhibited asymmetric changes. Fasciculation differences in the trapezius, triceps brachii, extensor digitorum communis, thenar, and first dorsal interosseous muscles were observed. In ALS patients, the CSA and fasciculation relative scores were correlated with electrophysiological indicators.

Conclusion

Ultrasound is a valuable tool for monitoring peripheral nerve CSA and muscle fasciculations, both of which correlate with electrophysiological indices, in ALS patients.

Introduction

Amyotrophic lateral sclerosis (ALS) is a sporadic, progressive, and degenerative disorder that characteristically affects both upper motor neurons (UMNs) and lower motor neurons (LMNs) [1, 2]. Classically, electrophysiological examinations are the primary tools for monitoring and diagnosing ALS; however, these methods are inherently invasive and require repeated assessments to confirm the diagnosis.

In the last 2 decades, high-frequency diagnostic ultrasound of peripheral nerves and muscles, known as neuromuscular ultrasound, has emerged as an alternative tool for evaluating individuals with neuromuscular conditions [3–6]. In nerve ultrasound studies, motor nerve atrophy has been observed in ALS patients [7, 8]. On the other hand, muscle ultrasound has revealed varying degrees of muscle atrophy, fasciculations, and alterations in muscle texture [9–11], with ALS patients exhibiting more frequent and intense fasciculations than non-ALS patients. To date, few studies have evaluated the correlation between electrophysiological data and ultrasound findings in individuals with ALS, and even fewer studies have emphasized the interrelationship between these diagnostic tools. Therefore, the aim of this study was to evaluate the application of neuromuscular ultrasound in ALS patients and to correlate the results with electrophysiological characteristics.

Methods

Participants

Patients diagnosed with ALS (n = 67) were recruited from the ALS clinic and wards of Peking University Third Hospital from November 2019 to March 2023. Nineteen patients who were diagnosed with other diseases that mimic ALS (e.g., brachial plexopathy, chronic inflammatory demyelinating polyneuropathy, frozen shoulder syndrome, and ALS overlap syndrome) were also recruited [12]. All ALS patients met the diagnostic criteria for ALS (“probable,” “laboratory-supported probable,” or “definite” ALS according to the Revised El Escorial Criteria) [13]. Baseline clinical information, including sex, age of onset, site of onset, disease duration, and relevant clinical scores for UMNs and LMNs, was collected. Participants were excluded if they met any of the following (a) inability to cooperate with the ultrasound examination; (b) presence of comorbidities such as diabetes or hypothyroidism; (c) preexisting symptoms of muscle fasciculations; or (d) history of peripheral nerve trauma, surgery or anxiety. Informed consent was obtained from each patient prior to the study. The study was approved by the Research Ethics Committee of Peking University Third Hospital (M2022513).

Neuromuscular Ultrasound

All 86 participants (67 with ALS and 19 with ALS mimics) underwent nerve ultrasound, and 69 of these participants (59 with ALS and 10 with ALS mimics) also underwent muscle ultrasound (Fig. 1). All ultrasounds were performed by the same physician (J.W.). A GE Vivid Q ultrasound system (GE Healthcare, Chicago, IL, USA) with a 12 MHz linear array transducer was used for the study. The participants lay in the supine position with the ultrasonographer facing them, and all imaging was performed bilaterally. Other than depth, the initial settings were kept constant during all the examinations.

Nerve Ultrasound

The transducer was adjusted to be perpendicular to the nerve to obtain the minimal cross-sectional image. The cross-sectional area (CSA, mm^2^) at a predetermined site of each nerve was measured by tracing just inside the hyperechoic rim of the nerve. Data were collected bilaterally from the median and ulnar nerves at 10 points on each nerve, moving from the distal to the proximal end (Fig. 2) [14]. The cervical fifth to seventh roots (excluding the eighth root due to its depth) were measured at their exits from the intervertebral foramina, and the upper, middle, and lower trunks of the brachial plexus were measured around Erb’s point between the anterior and middle scalene muscles. To account for differences in the proportion of motor fibers at different segments of the peripheral nerve, the ratio of distal/proximal CSAs at noncompressed sites of the peripheral nerves was also calculated [15].

Muscle Ultrasound

The participants were instructed to keep their muscles relaxed during the ultrasound. For each patient, ultrasound was performed on the following tongue, masseter, trapezius, and sternocleidomastoid in the bulbar region; deltoid, biceps brachii, triceps brachii, flexor carpi radialis (FCR), extensor digitorum communis (EDC), thenar (TH) muscles, and first dorsal interosseous (FDI) muscles in the cervical region; rectus abdominis (RA) and paravertebral muscles at the T10 level (in the prone position) in the thoracic region; and rectus femoris (RF), vastus lateralis, biceps femoris, tibialis anterior, and gastrocnemius muscles in the lumbosacral region (Fig. 3). Each muscle was visualized transversely at a depth and width of 4 cm under B mode. The skin temperature of the limbs and trunk was maintained above 32°C. Each muscle was observed on ultrasound for 60 s to assess the presence of fasciculations [16], defined as visible, localized muscle twitching with a highly irregular pattern during complete relaxation. Fasciculations were graded on a five-point grade no fasciculations observed; grade one site of fasciculations with ≤5 occurrences in 60 s; grade one site with >5 and ≤15 occurrences in 60 s or two sites with ≤10 occurrences; grade one site with >15 and ≤25 occurrences in 60 s or two to three sites with >10 and ≤20 occurrences; grade more than two sites with >20 occurrences in 60 s. This grading system was developed by modifying existing assessment criteria from previous fasciculation studies [10, 17]. Additionally, a relative fasciculation score was calculated for each region or site as relative fasciculation score = (total fasciculation grades of all monitored muscles in the region or site)/(number of muscles × 4). The score ranges from 0 to 1, with higher scores indicating a greater degree of fasciculation.

Neurophysiological Studies

Patients underwent routine nerve conduction studies using a keypoint four-channel electromyography evoked potentiometer (Medtronic, USA). All tests were performed by the same operator (S. Zhang). The negative amplitudes of the compound muscle action potentials (CMAPs) for motor nerve conduction were recorded in the median, ulnar, peroneal, and tibial nerves of ALS patients (median and ulnar nerves in the wrist, peroneal, and tibial nerves in the ankle). Additionally, the distal motor latency and F-wave frequency for the ulnar and median nerves were measured. These nerves were scored based on the decrease in the amplitude of each nerve CMAP [18]: 0 (CMAP ≥ [X-2 s]), 1 (50% [X-2 s] < CMAP < [X-2 s]), 2 (30% [X-2 s] < CMAP <50% [X-2 s]), and 3 (CMAP ≤30% [X-2 s]). The values of X and S are derived from Peking University Third Hospital, as listed in online supplementary Data S1 (for all online suppl. material, see https://doi.org/10.1159/000546425). The decreases in the amplitudes of the CMAPs of the 8 nerves were summed (0–24) to reflect the overall LMN axonal damage in each ALS patient. Since some patients had incomplete nerve conduction study data, a relative score (relative score = absolute score/[3 × number of nerves with motor conduction data], 0–1) was used as a proxy for overall LMN axonal damage [18]. The neurophysiological index (NI) for the ulnar and median nerves was calculated using the De Carvalho formula as NI = (CMAP amplitude/distal motor latency) × F-wave frequency [19]. A decrease in both CMAP and NI values indicated a greater degree of axonal damage or LMN involvement.

Statistical Analysis

The data were assessed for normality using the Shapiro-Wilk test. For normally distributed data, Student’s t test was used, whereas the Mann-Whitney U test was used for nonparametric data. The χ^2^ test was used for categorical variables. Associations between parameters were analyzed using Pearson’s correlation coefficient or Spearman’s rank correlation coefficient. Normally distributed data are expressed as the mean ± standard error of the mean, and nonnormally distributed data are expressed as the median (interquartile range). A p < 0.05 was considered statistically significant.

Results

Clinical Characteristics

No significant differences in demographic or clinical characteristics, including upper motor neuron scores (UMNS), lower motor neuron scores (LMNS), or electrophysiological data (Table 1), were detected between the two groups.

Nerve Ultrasound

CSA Differences in Peripheral Nerves between ALS and ALS Mimics

The CSAs of the bilateral peripheral nerves in each group were compared (Table 2). In the ALS group, asymmetry was observed in the median and ulnar nerves of the forearm (i.e., the distal upper limb) (M3: p = 0.035; M5: p = 0.035; U2: p = 0.036), whereas the CSAs of the nerves in the upper arm (i.e., the proximal upper limb) were generally symmetrical. Conversely, in the ALS mimics group, asymmetry was mainly observed in the nerves of the upper arm (U9: p = 0.022; U10: p = 0.033; U: p = 0.026).

A comparison of the CSA values between the groups revealed significant differences, primarily in the proximal peripheral nerves of both upper limbs (mainly at sites 8–10) and at the cervical nerve roots and trunks. Additionally, the distal-to-proximal ratio of the left ulnar nerve also differed between the groups, as detailed in online supplementary Data S2.

Correlation Analysis between CSA and Electrophysiological Data

Because of the differences in proximal peripheral nerves between the two disease groups, CSA data for the proximal peripheral nerves were included in the correlation analysis using Spearman’s correlation and controlling for disease duration, site of onset, LMNS, and BMI. Significant correlations were found between various sites of the right peripheral nerves and electrophysiological indicators in the ALS group (Fig. 4), whereas no significant correlations were found in the ALS mimics group.

Association of the Site of Onset with Upper Limb CSA

To avoid the impact of different onset regions on upper limb peripheral nerve ultrasound data, we analyzed data from participants with upper limb onset (41 ALS cases and 13 ALS mimic cases; online suppl. Data S3). There were no significant differences in the demographic data between the two groups. However, electrophysiological data revealed a significant difference in the right NI between the groups. In the paired analysis of bilateral peripheral nerves within each group, no significant differences were found in the CSA of bilateral nerves in the ALS group, whereas the ALS mimics group still exhibited asymmetrical changes in the ulnar nerve. Intergroup comparisons revealed that the bilateral proximal nerves remained significantly different.

Muscle Ultrasound

Differences in Muscle Fasciculation between ALS and ALS Mimics

A paired analysis of the grade of fasciculations in the bilateral muscles (excluding the tongue) within each group revealed asymmetry only in the RA in the ALS group. In the intergroup comparison, differences in the grade of fasciculations were found in the following bilateral trapezius, bilateral triceps brachii, bilateral EDC, bilateral TH muscles, and left FDI muscle. These muscles had higher rates of positive fasciculations in the ALS group (L. trapezius 80%, R. trapezius 86%; L. triceps brachii 92%, R. triceps brachii 87%; L. EDC 87%, R. EDC 89%; L. TH 89%, R. TH 71%; L. FDI 73%; online suppl. Data S4). Additionally, fasciculation relative scores in the bulbar and cervical regions showed intergroup differences, indicating that the degree of muscle fasciculation in the ALS group was greater than that in the ALS mimics group (online suppl. Data S5).

Muscle Fasciculations in ALS Patients

Given the specificity of fasciculations in ALS, the correlation of fasciculation severity across different sites in ALS patients was assessed. The results revealed a positive correlation between the relative fasciculation scores at various sites (Fig. 5). In the correlation analysis between upper limb muscle ultrasound data from ALS patients and the results of nerve ultrasound and electrophysiological assessments, the findings revealed a negative correlation between the NI of the right median nerve and the fasciculation relative score of the right upper limb (Table 3). However, no significant correlation was found between the overall fasciculation score and the King’s College Staging System (KCSS) stage (Pearson correlation analysis, r = 0.124, p = 0.349).

From an anatomical perspective, the correlation between fasciculations in specific muscles and the CSAs of the nerves that innervate them was assessed. The muscles examined included the deltoid (innervated by C5 and C6), biceps brachii (innervated by C5 and C6), triceps brachii (innervated by C6, C7, and C8), FCR and TH muscles (both innervated by the median nerve), and FDI muscles (innervated by the ulnar nerve). The results indicated that the proximal CSA of the left median nerve was positively correlated with the fasciculation grades of the FCR and TH muscles (FCR with M7 and M8; TH with M7 to M9; online suppl. Data S6).

Additionally, the fasciculation grades of antagonistic muscles within the ALS group were compared. Differences were observed between the bilateral sternocleidomastoid and trapezius muscles, with higher fasciculation grades in the trapezius. Similar differences were found between the right EDC and FCR, with the EDC showing a higher fasciculation grade (online suppl. Data S7).

Discussion

Neuromuscular ultrasound has been used in diagnosing neuromuscular diseases, offering new insights into the characteristics of ALS. This study revealed that the proximal CSAs of the upper limb peripheral nerves are smaller in ALS patients than in ALS mimics. Research has also demonstrated that ALS patients exhibit atrophy of the upper limb peripheral nerves, particularly in the proximal segments [7]. These findings suggest that proximal changes in the peripheral nerves are more pronounced in individuals with ALS. In peripheral nerves, motor fibers, as the primary myelinated nerves, are more predominant in the proximal segments, whereas sensory fibers, especially those that are unmyelinated or lightly myelinated, are more common in the distal segments. Thus, the more pronounced changes observed in the proximal peripheral nerves of ALS patients are likely due to the greater proportion of motor fibers [22]. Additionally, these results support the occurrence of motor neuron anterograde degeneration (or “Wallerian degeneration”) following damage to anterior horn cells, resulting in proximal peripheral nerve damage. Given the selective involvement of motor fibers in ALS, neuromuscular ultrasound is less likely to detect significant changes in the distal peripheral nerves. Moreover, the proximal peripheral nerves in ALS patients are correlated with neurophysiological data. Smaller proximal CSAs are associated with lower CMAPs and NIs, predominantly in the right median nerve. These observations are consistent with previous studies indicating that the median nerve is more affected than the ulnar nerve is [7], possibly due to a splitting mechanism [23]. These findings underscore the need for further clinical practice and research focusing on proximal peripheral nerves.

The superior noninvasiveness and flexibility for free detection of muscle ultrasound allow for a more comprehensive and detailed assessment of fasciculations. For example, muscle ultrasound can detect subtle fasciculations that electrophysiology examination might miss [11]. In this study, we identified significant differences in fasciculations between the ALS and ALS mimics groups, particularly in the bilateral trapezius, triceps brachii, EDC, and TH muscles and left FDI muscle. These muscles exhibited a higher rate of fasciculations in ALS patients than in ALS mimics, suggesting that evaluating these specific muscles could help distinguish between the two diseases. Notably, the muscles innervated by the thoracic and lumbosacral regions did not differ between the groups. Although the reason for this difference is unclear, it suggests that muscle ultrasound should focus on muscles innervated by the bulbar and cervical regions. Fasciculations are a common sign of ALS, and patients often complain of muscular twitching. The correlation of fasciculation relative scores across various sites was analyzed and revealed significant positive correlations at all sites. This finding suggests a lack of site selectivity for fasciculations in ALS patients, implying that in a time-constrained evaluation, assessing fasciculations in just one region may be sufficient. However, the heatmap (online suppl. Data S4) shows varying degrees of fasciculation among different muscles, indicating differential involvement, which is consistent with clinical observations of selective muscle involvement (e.g., the splitting phenomenon) [24]. However, paired bilateral muscles exhibit consistent levels of fasciculations, which is supported by previous studies on the consistency of homologous muscle involvement [25]. Correlation analysis with electrophysiology indices revealed a negative correlation between the fasciculation relative score of the right upper limb and the RM-NI, suggesting that more severe damage to the right median nerve is associated with more pronounced muscle fasciculations. This finding, along with the correlation between nerve ultrasound and electrophysiology indices in the right median nerve, indicates that the right limb, as the dominant limb, may be more susceptible to ALS pathogenesis [26]. Additionally, these results suggest that RM-NI is an electrophysiological indicator worthy of further exploration.

In this study, the correlation between nerve and muscle ultrasound findings in ALS patients was explored. The results revealed a positive correlation between the CSA of the proximal left median nerve and the grade of fasciculations in the FCR and TH muscles, indicating that as the CSA decreases, the intensity of fasciculation also decreases. Interestingly, no similar correlation was observed in the right median nerve. On the one hand, since the overall fasciculation relative score does not correlate linearly with KCSS staging, we suggest that the intensity of muscle fasciculations increases and peaks in early stages and then decreases as neurodegeneration worsens, exhibiting an inverted U-shaped relationship with a possible floor effect in later stages. Muscle fasciculations are widely recognized to be influenced by both UMNs and LMNs [27–29], and their combined effect is far from linear. On the other hand, we suggest that in ALS patients, the CSAs of peripheral nerves decrease unidirectionally due to degenerative changes. Therefore, given that the ultrasound findings of nerves and muscles do not always align over time, some sites may show nonsignificant correlations.

Our preliminary observations demonstrate that neuromuscular ultrasound serves as a complementary diagnostic tool. Furthermore, subsequent clinical investigations should refine the four-region fasciculation monitoring protocol and establish systematic correlations with electrophysiological parameters; this multimodal approach could increase the sensitivity of current diagnostic criteria.

The limitations of this study include the small sample size, but given that ALS is a rare disease, these findings can still provide clinically valuable results. Additionally, the lack of follow-up in this study limits the understanding of the dynamic changes in muscle fasciculations over time. We hope to address these limitations with a larger sample size and a longitudinal cohort. Furthermore, the study did not include a normal control group, making changes in peripheral nerves in ALS patients unclear. However, proximal peripheral nerves in ALS mimics exhibited asymmetric changes. The ALS mimics in this study were primarily patients with peripheral neuropathy, potentially due to asymmetric myelin damage and regeneration or axonal damage caused by primary or secondary factors [30–32]. Nerve ultrasound was performed only on the peripheral nerves of the upper limbs. A more comprehensive assessment should include the lower limbs. However, the lower limb peripheral nerves were not examined owing to their deeper location and the difficulty in clearly defining their boundaries. Therefore, peripheral nerve MRI may be a better assessment option [31]. Additionally, we did not explore the correlation between fasciculation potentials in needle EMG and fasciculation grades because our preliminary analysis identified multiple confounding factors (such as disease duration and the relative contributions of UMN and LMN involvement) that could affect the interpretability of the results. On the basis of these findings, we will optimize our study design to longitudinally assess the relationship between electrophysiological and ultrasonographic fasciculation characteristics in future follow-up studies. Finally, evaluating muscles affected by ALS could benefit from further analyses, such as muscle texture analysis and the application of machine learning techniques.

Conclusion

In summary, our data indicate that ALS and ALS mimics exhibit distinct peripheral nerve and muscle characteristics on ultrasound. The proximal CSAs of upper limb peripheral nerves and the intensity of fasciculation in muscles innervated by the bulbar and cervical regions are key differentiators. In ALS, both the proximal CSA of the right median nerve and the fasciculation relative score of the right upper limb are correlated with the RM-NI, which is consistent with the clinical characteristics of ALS. The results of this study suggest that assessing the proximal CSAs of upper limb nerves and fasciculations in the bulbar and cervical muscles can be effective in evaluating and differentiating ALS patients.

Acknowledgments

We thank all the participants involved in this study. The authors also extend their gratitude to Drs. Mingsheng Liu and Jingwen Niu from Peking Union Medical College Hospital for their technical support.

Statement of Ethics

The study was approved by the Research Ethics Committee of Peking University Third Hospital (M2022513). Written informed consent was obtained from all participants.

Conflict of Interest Statement

The authors report that they have no conflicts of interest.

Funding Sources

This work was supported by the National Natural Science Foundation of China (81873784, 82071426 and 81641079), the Clinical Cohort Construction Program of Peking University Third Hospital (BYSYDL2019002), and the Clinical Key Project of Peking University Third Hospital (BYSYZD2021004).

Author Contributions

Y.X. and D.F. designed and directed the study. J.W., H.S., and Y.X. performed the experiments and wrote the manuscript. J.W., H.S., M.A., S.Z., and X.H. collected the data. J.W. and Y.X. performed the data analysis. H.S. and S.Z. provided technical assistance. All the authors have read and approved the final version of the manuscript.