TYPE Original Research PUBLISHED 10 October 2022 DOI 10.3389/fneur.2022.980746 Shear wave elastography OPEN ACCESS combined with electromyography to assess the EDITED BY Wen-Shiang Chen, National Taiwan University, Taiwan REVIEWED BY Chueh-Hung Wu, effect of botulinum toxin on National Taiwan University, Taiwan Masachika Niimi, Nihon University, Japan spastic dystonia following *CORRESPONDENCE Lucio Marinelli stroke: A pilot study lucio.marinelli@unige.it SPECIALTY SECTION This article was submitted to William Campanella1,2 , Angelo Corazza3 , Luca Puce1 , Neurorehabilitation, Laura Privitera1 , Riccardo Pedrini1 , Laura Mori1,2 , a section of the journal Frontiers in Neurology Leonardo Boccuni4,5,6 , Giovanni Turtulici7 , Carlo Trompetto1,2 RECEIVED 28 June 2022 and Lucio Marinelli1,8* ACCEPTED 16 September 2022 1 PUBLISHED 10 October 2022 Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genova, Genova, Italy, 2 IRCCS Ospedale Policlinico San Martino, Department of CITATION Neuroscience, Division of Neurorehabilitation, Genova, Italy, 3 Unità di Radiologia Diagnostica ed Campanella W, Corazza A, Puce L, Interventistica Istituto Ortopedico Galeazzi di Milano, Milan, Italy, 4 Institut Guttmann, Institut Privitera L, Pedrini R, Mori L, Boccuni L, Universitari de Neurorehabilitació adscrit a la UAB, Barcelona, Spain, 5 Universitat Autònoma de Turtulici G, Trompetto C and Marinelli L Barcelona, Cerdanyola del Vallès, Spain, 6 Fundació Institut d’Investigació en Ciències de la Salut (2022) Shear wave elastography Germans Trias i Pujol, Barcelona, Spain, 7 S.C. Radiodiagnostica Ospedale Evangelico Internazionale combined with electromyography to di Genova, Genova, Italy, 8 IRCCS Ospedale Policlinico San Martino, Department of Neuroscience, assess the effect of botulinum toxin on Division of Clinical Neurophysiology, Genova, Italy spastic dystonia following stroke: A pilot study. Front. Neurol. 13:980746. doi: 10.3389/fneur.2022.980746 COPYRIGHT Background: Shear wave elastography (SWE) is a method for carrying out a © 2022 Campanella, Corazza, Puce, quantitative assessment of the mechanical properties of soft tissues in terms Privitera, Pedrini, Mori, Boccuni, of stiffness. In stroke survivors, the paretic muscles may develop hypertonia Turtulici, Trompetto and Marinelli. This is an open-access article distributed due to both neural-mediated mechanisms and structural alterations with under the terms of the Creative consequent muscular fibrous-fatty remodeling. Commons Attribution License (CC BY). The use, distribution or reproduction Methods: Fourteen adult patients with spastic dystonia following stroke in other forums is permitted, provided were recruited. Muscle hypertonia was assessed using the modified Ashworth the original author(s) and the copyright owner(s) are credited and that the scale (MAS). Muscle activation was measured by surface electromyography original publication in this journal is (sEMG) with the selected muscle in shortened (spastic dystonia) and stretched cited, in accordance with accepted (dynamic stretch reflex) positions. SWE was performed on a selected paretic academic practice. No use, distribution or reproduction is permitted which muscle and on the contralateral non-paretic one to calculate shear wave does not comply with these terms. velocities (SWV) along and across muscular fibers. The modified Heckmatt scale (MHS) pattern was also determined. All evaluations were performed shortly before BoNT-A injections (T0) and one month later (T1). Results: All SWV on paretic muscles were higher than contralateral non-paretic ones (p < 0.01). After BoNT-A injection, a significant reduction in MAS (p = 0.0018), spastic dystonia (p = 0.0043), and longitudinal SWE measurements, both in shortened (p = 0.001) and in stretched muscular conditions (p = 0.0029), was observed. No significant changes in SWV on non-paretic muscles were observed. Higher SWV resulted along the direction of muscular fibers vs. across them (p = 0.001). No changes resulted from the MHS evaluations after BoNT-A. There was a positive correlation between MHS scores and SWV values Frontiers in Neurology 01 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 while the muscle was in the shortened position, but not with spastic dystonia recorded by sEMG. Conclusions: This is the first study evaluating the effect of BoNT-A on muscle hypertonia following stroke, assessed by both SWE and sEMG. These findings support SWE as a useful method to disclose intrinsic muscular remodeling, independently of the effect of spastic dystonia, in particular, while muscles were assessed in a neutral position. SWE measurements of muscle stiffness cannot tell apart neural-mediated and intrinsic muscle hypertonia. Interestingly, when sEMG activity is very limited, as in spastic muscles kept in a shortened position, SWE can provide a measurement of stiffness due almost completely to intrinsic muscle changes. Alongside sEMG, SWE could aid clinicians in the assessment of responses to treatments. KEYWORDS shear wave, elastosonography, spasticity, muscle tone, muscle hypertonia, stiffness Introduction spasticity in clinical practice (7). However, the MAS is based on a subjective and non-quantitative assessment and this restricts Following a stroke, paretic limbs often develop muscle its application. Moreover, considering the validity, reliability, hypertonia within the so-called upper motor neuron syndrome. and sensitivity, MAS is not considered an ideal scale for the Spasticity is the most frequently investigated cause of increased assessment of muscle hypertonia (8). muscle tone, occurring in about 38% of post-stroke patients Conventional B-mode ultrasonography (US) is widely (1). Spasticity can be clinically appreciated as a velocity- utilized for musculoskeletal pathology as a first-line approach dependent increase in muscle tone during passive stretching, mainly because of real-time access and relatively low costs (9). due to increased excitability of the stretch reflex resulting Sonoelastography can assess the mechanical properties of soft from upper motor neuron lesions (2, 3). Spastic dystonia tissues by US imaging. In particular, a further development is less commonly evaluated, but recent studies suggest that of such technique, called “shear wave elastography” (SWE), it may occur more frequently than spasticity (4). Like in provides a quantitative evaluation of elastic properties by muscles with spasticity, those with spastic dystonia show a measuring the propagation velocity distribution of the stretch reflex during passive muscle lengthening, but, differently directional shear waves, produced by an ultrasound pulse (10). from spasticity, electromyographic activity reflecting muscle In the recent years, SWE has been introduced as a promising activation is present also at rest, with the muscle in a diagnostic tool in musculoskeletal pathology to assess muscle neutral or shortened position, despite relaxation attempts (2). and tendon elastic properties related to degeneration, injury, Spastic dystonia probably shares the same neural substrate as and healing (11, 12). Recently some authors demonstrated spasticity, but likely contributes more importantly to muscle good to excellent inter-observer reproducibility and intra- shortening and disability (5, 6). Surface electromyography observer repeatability in radiologists, validating the feasibility (sEMG) is considered the gold standard to detect electric muscle and reliability of SWE to assess mechanical properties of skeletal activation and is therefore ideal to measure spasticity and spastic muscle tissue in the upper and lower extremities (12–14). dystonia, which represent the main contributors to muscle In recent years, many studies (15–24) used SWE with tone increase in post-stroke patients. Importantly, reduced limb different approaches and protocols, to quantify muscular mobility affects muscle structure: sarcomere loss, an increase stiffness related to spasticity. of connective tissue, and muscle contracture produce intrinsic Shear wave velocity (SWV) increases with increasing tissue hypertonia that sums up with spasticity and spastic dystonia but stiffness and is therefore influenced either by muscle fibrosis cannot be directly assessed by sEMG. (intrinsic hypertonia) or muscle contraction (active or reflex). Clinicians should discriminate with greater precision the Compared to healthy muscles, paretic muscles therefore should increase in muscle tone after stroke, distinguishing the neuro- demonstrate increased stiffness at rest (mostly because of mediated reflex hypertonia (from now on called “reflex intrinsic hypertonia, particularly if the muscle is in a shortened hypertonia”) from the intrinsic one, to define the correct position) and even more while stretched or kept elongated treatment and assess the therapeutic impact. The modified (intrinsic hypertonia + reflex hypertonia) (6). An increase of Ashworth scale (MAS) is the most feasible measurement of SWV is therefore expected in resting paretic muscles compared Frontiers in Neurology 02 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 to the healthy ones as well as in stretched compared to shortened reduced distance from the ultrasound transducer (such as biceps conditions. Botulinum toxin should only reduce stiffness due brachii muscle) were preferred. to reflex hypertonia (recordable with sEMG), but not intrinsic All patients underwent sEMG, conventional B-mode US for hypertonia, as it is ineffective on fibro-adipose degeneration of the qualitative identification of muscle echogenicity applying the muscle. the modified Heckmatt Scale (MHS) and the quantitative In this context, SWE could gain an important role due to its evaluation of SWV, reflecting muscle stiffness, through SWE potential capability to quantitatively assess the contribution of ultrasound evaluation. The latter was also performed at the viscoelastic properties to soft tissue stiffness. The opportunity to contralateral homolog non-paretic muscle. The ultrasound match the sEMG evaluation of spastic dystonia before and after evaluations were performed by an experienced radiologist botulinum neurotoxin type A (BoNT-A) treatment, paralleled (Corazza). The examiners evaluated the patients independently with SWE evaluation, can provide unprecedented information and were blind to each other. about the contribution of reflex hypertonia to SWV. All MAS, sEMG, MHS, and SWV evaluations were carried Based on these assumptions, in this pilot study, the main out just before the injection of BoNT-A (T0) and then repeated objective was to assess the effect of botulinum toxin on post- 4 weeks (range: 22–32 days) later (T1) according to the same stroke patients with spastic dystonia using both sEMG and SWE protocols: this time interval was chosen to intercept the time to confirm that muscle hypertonia improvement can be reflected when the maximum effect of BoNT-A is expected after injection. by the reduction of both muscle electromyographic activity and SWV. Furthermore, we aimed to investigate the capability of SWE to estimate muscular changes in terms of elasticity in Electromyography paretic limbs, due to the contribution of reflex hypertonia to overall stiffness. Self-adhesive electrodes (Neuroline 700, Ambu A/S, Ballerup, Denmark) were placed 1 cm apart over the muscle belly for bipolar recording following SENIAM (Surface Electromyography for Non-Invasive Assessment of Muscles) Materials and methods guidelines (25). Signals were acquired by a BIOAMP LT unit Population (Vertigo, Genoa, Italy) and underwent a 20 to 500 Hz band-pass filter for offline processing using custom software running in From September 2018 to December 2019, we recruited 14 Labview (National Instrument, Austin TX, USA). Subjects were patients (5 females) with a median age of 61 years (range 42– lying supine on the examination table in a quiet room while 78). Inclusion criteria were as follows: (1) presence of spastic attempting to remain relaxed (the prone position has been dystonia (presence of EMG activity at rest that cannot be preferred in the case of triceps surae muscle). voluntarily silenced by the subject) (3) in at least one muscle Spastic dystonia was assessed by recording sEMG, while group (e.g., elbow flexors) of only one side of the body due the muscle remained in a neutral position, before performing to ischaemic or hemorrhagic stroke occurred at least 3 months passive limb movements. Each subject entered the analysis with before, (2) unaffected contralateral body side, (3) for patients the average rectified value (ARV) during 30 s of recording. A treated with botulinum toxin, at least 4 months must have dynamic stretch reflex was also assessed in the selected muscle passed since last injection, (4) age of 18 years or more, and (5) by recording sEMG during 10 passive stretches performed ability to understand and sign the informed consent. Exclusion at a reproducible speed, according to a previously validated criteria were concomitant parkinsonism and history of lesions methodology (26). The selected muscle was stretched with affecting the muscle to be evaluated. The institutional review linear (ramp and hold) passive movements following the pacing board at IRCCS Ospedale Policlinico San Martino (Genoa, Italy) provided by an emulated metronome. The beats per minute pace approved the present study and all patients provided written (BPM) was decided for each patient according to the muscle and informed consent before data collection and assessment. the amount of hypertonia and maintained the same at both time points. The ARV was calculated for each movement only during the dynamic phase of muscle stretching. Each subject entered the analysis with the mean ARV obtained from 10 consecutive Assessment stretches before and after BoNT-A injections. All enrolled patients underwent a neurological examination and assessment of muscle hypertonia applying the MAS. Among Sonoelastography those muscles where botulinum toxin injection was indicated, a muscle with MAS > 0 was selected for the consequent sEMG A LOGIQ E9 ultrasound system equipped with a 9L linear and SWE assessments. Muscles characterized by large bulk and array transducer with Shear Wave Elastography technology (GE Frontiers in Neurology 03 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 Healthcare, General Electric Company, Chicago, Illinois, US) injected according to the current guidelines based on muscle size was used to acquire B-mode images and SWE measurements. and amount of hypertonia. Generally, each muscle was injected Such technology uses a “comb-push” excitation source (multiple with 20 to 100 IU and each patient received no more than 400 pushing beams transmitted on the transducer simultaneously in IU. All BoNT-A injections were performed under US guidance. a comb-like pattern), time-interleaved shear wave tracking (data interpolation), directional filters, and a time-of-flight algorithm which is used to estimate the local SWV at every location in the Statistical analysis region of interest. Patients have been placed in the supine position; the Wilcoxon signed-rank test was used for all comparisons: paretic arm (P) of the subject was first held in a shortened MAS and MHS (T0-T1), electromyographic ARV for spastic (sh) position. Axial scans were obtained with the US probe dystonia (T0-T1), and SWE measures (T0-T1, P-NP, Ax-Lo, positioned on the skin surface on the short axis (Ax) of the sh-st). Spearman rank test was used to assess the correlation muscle being examined. Standard machine settings included a between the Heckmatt scale and SWV measurements at T0. maximum image depth of 10 cm, scanning frequency of 7 MHz, Continuous variables have been reported using mean ± SD, single image focus, tissue harmonic imaging, speckle reduction while non-continuous variables have been reported as median function turned off, and setting meters per second (m/s) as the (1st to 3rd quartiles). Statistical significance was set for p < unit of measurement of the SWV (15). Particular attention was 0.01. This is a pilot study: no preliminary data were available to paid to applying the least possible pressure on the underlying estimate the sample size. tissues, to avoid any additional, albeit minimal, compression on the muscle by increasing the risk of obtaining overestimated measurements. The quantitative values of the SWV for muscle Results stiffness estimation were measured using a rectangular region of interest (ROI) with variable dimensions (maximum dimensions: All 14 patients were evaluated on paretic and contralateral 2 x 2 cm), set to a depth between 1 and 5 cm from the skin surface healthy selected muscles for a total of 28 muscles (Table 1). and displayed on the screen as an encoded color map of the None reported discomfort or any side effect during or after SWVs detected. The objective of the procedure was to achieve sEMG and SWE assessments. One subject (#14) had a severe reproducible SWV measurements, avoiding the detection of reduction in the range of motion of the paretic limbs aponeurosis and/or thickest intramuscular septa within the ROI (MAS=4). Given the very limited joint excursion, in this and by searching, at the same time, to cover the wider area patient, dynamic stretch reflex recording and stretched SWV of muscle fibers (27). Further, analogous SWE measurements assessments were unfeasible. were repeated by rotating the probe on longitudinal planes of the muscle (Lo) (Figures 1, 2). This procedure was repeated on the same muscle in a stretched condition (st), while a second Modified Ashworth scale operator was keeping the arm of the patient passively extended. Subsequently, the protocol was performed on the contralateral MAS values at T0 were 2 (2–3) and at T1 significantly homolog non-paretic muscle (NP). For each condition, 10 decreased to 1.75 (1-2) (p = 0.018) (Figure 3). measurements were collected and the median value entered consequent analyses. Surface electromyography Modified Heckmatt scale Spastic dystonia ARV was 8.07 ± 8.24 µV at T0 and 2.65 ± 1.62 µV at T1; dynamic stretch ARV was 18.13 ± 22.61 µV at Meantime, since the SWE evaluation was provided based on T0 and 7.30 ± 3.70 µV at T1. Both spastic dystonia (p = 0.0043) a conventional B-Mode US image, an MHS pattern (28) of the and dynamic stretch (p = 0.0019) significantly decreased at T1 selected muscle was assessed by the same operator following the after BoNT-A injection (Figure 4). different degrees of muscle echogenicity. Shear wave elastography Botulinum toxin a injections A total of 224 SWV values resulted considering 14 patients In total, 100-unit vials of incobotulinumtoxinA (Merz evaluated in 16 conditions resulting from the combination of the Pharma, Frankfurt am Main, Germany) were diluted with 2 ml four binary variables: time point (T0 vs. T1), body side (P: paretic of preservative-free normal saline (50 U/ml). Muscles were limb vs. NP: non-paretic limb), transducer orientation (Ax: axial Frontiers in Neurology 04 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 1 Scheme illustrating SWE acquisition on the short axis of muscular fibers: the transducer is placed on the muscle belly performing consecutive axial scans to assess the maximum amount of muscle (left); in each scan, SW velocities are measured in the positioned rectangular ROI generating the relative color-coded elastogram superimposed to the B-mode image (right). The stiffness at any location within the ROI is sampled using the measurement tool (dotted circles). A final median value is then expressed. vs. Lo: longitudinal), and muscle condition (st: stretched vs. Further, SWV measurements obtained in stretched sh: shortened). conditions were consistently higher than measures assessed in Overall, SWV measurements on paretic muscles assessed shortened conditions (p = 0.001 for all comparisons) (Figure 6). with a longitudinal positioning of the probe showed statistically SWV measurements obtained with a longitudinal significant reduction at T1 vs. T0 both in shortened (p = positioning of the transducer were also higher than measures 0.001) and in stretched (p = 0.0029) conditions (Figures 5, 6). performed with the axial orientation of the probe (p = 0.001 for Conversely, SWV assessments with an axial orientation of the all comparisons) (Figures 6, 7). probe did not show a statistically significant reduction between the two time points in stretched (p = 0.66) and shortened (p = 0.03) conditions (Figure 6). Modified Heckmatt scale SWV measurements on non-paretic muscles did not significantly decrease between the two time points (Figure 6). During the execution of the conventional B-mode US scan, All SWV measurements on paretic muscles were higher the MHS was applied for the paretic side (Figure 8). MHS scores than contralateral non-paretic muscles apart from shortened [T0 and T1: 2 (2, 3)] did not change between T0 and T1. There longitudinal at T1 (p = 0.016 for shLoT1, p = 0.002 for shAxT0 was a positive correlation between MHS scores and SWV values and shAxT1, and p = 0.001 for the other five comparisons) at T0 and also at T1, more clear with the muscle in a shortened (Figure 6). position (T0: Axial Rho = 0.8, p = 0.004; Longitudinal Rho = Frontiers in Neurology 05 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 2 Scheme illustrating SWE acquisition on the long axis of muscular fibers: the transducer is placed on the muscle belly performing consecutive longitudinal scans to assess the maximum amount of muscle; in each scan, SW velocities are measured in the positioned rectangular ROI generating the relative color-coded elastogram superimposed to the B-mode image (below). The stiffness at any location within the ROI is sampled using the measurement tool (dotted circles). A final median value is then expressed. TABLE 1 Baseline patients’ characteristics and assessed muscle. Patient Gender Age (y) Affected side of the body Time since stroke (y) Target muscle 1 M 49 Hemorrhagic/left 5 Flexor Digitorum Sup 2 M 78 Ischemic/right 3 Brachialis 3 M 46 Ischemic/right 7 Soleus 4 F 48 Hemorrhagic/right 1 Gastrocnemii 5 F 57 Ischemic/right 6 Biceps Brachii 6 M 65 Hemorrhagic/right 3 Brachialis 7 M 62 Ischemic/left 5 Biceps Brachii 8 F 46 Ischemic/right 14 Brachialis 9 M 71 Ischemic/right 8 Gastrocnemii 10 F 42 Ischemic/right 3 Gastrocnemii 11 M 62 Ischemic/right 4 Biceps Brachii 12 M 62 Ischemic/left 4 Brachialis 13 F 60 Ischemic/left 2 Soleus 14 M 69 Ischemic/left 15 Biceps Brachii Frontiers in Neurology 06 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 3 Chart illustrating the significant difference in MAS score assessments of spastic muscles before (T0) and 4 weeks after (T1) BoNT-A injection. Correlation with stroke age The time elapsed from the stroke to the baseline evaluation (time after stroke, see Table 1) was positively correlated with MHS (Rho = 0.7, p = 0.007) but no correlation emerged with SWV, EMG, and MAS values. Discussion To our knowledge, this is the first attempt to quantitatively assess the structural and neural-mediated components in post- FIGURE 4 Chart illustrating the average rectified values (ARV) of stroke patients with spastic dystonia by sEMG and SWE electromyographic activity (µV) measured in muscles with combined assessment. Two main observations resulted from spastic dystonia while kept in a shortened position before (T0) this study. First, SWE, sEMG, and MAS were all able to and 4 weeks after (T1) BoNT-A injection. See text for statistical comparison. detect the effect of BoNT-A on paretic muscles in terms of a significant decrease in their values, as previously demonstrated (26, 27). Second, SWE can differentiate paretic from non- paretic muscles. Following BoNT-A injection, both SWV and EMG measurements decreased because both are affected by reflex hypertonia. Such results were expected and should be 0.8, p = 0.006 – T1: Axial Rho = 0.8, p = 0.004; Longitudinal related to the fact that SWE estimates muscular stiffness Rho = 0.8, p = 0.006) but to a lesser extent also with muscle by measuring its viscoelastic behavior; hence, it is unable stretched (T0: Axial Rho = 0.6, p = 0.04; Longitudinal Rho = to properly differentiate stiffness due to muscle fibrosis or 0.7, p = 0.01 – T1: Axial Rho = 0.6, p = 0.04; Longitudinal Rho muscle activity (two simultaneous phenomena which sum up = 0.8, p = 0.008). No correlation emerged between MHS and in patients with muscle hypertonia due to upper motor neuron spastic dystonia ARVs. syndrome) (29). Frontiers in Neurology 07 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 5 SWE reduction between T0 and T1. In the upper-right quadrant is illustrated an example of an elastogram (superimposed to the B-mode image) generated by an ROI placed in the biceps brachii muscular belly along the direction of the fibers at T0; in the lower right image, the same evaluation at T1 shows lower values (lower left corners of the images). Images on the upper- and lower-left quadrant are the respective B-mode conventional US image automatically paired on the screen by the SEL software. Regarding SWE, a significant decrease at T1 was more shear waves generated into such tissue propagate in different evident for measurements obtained with a longitudinal ways with respect to the orientation of muscular fibers. It may orientation of the probe with respect to the direction of be argued that shear waves which propagate on a transverse muscular fibers. Such findings reflect the physical properties axis with respect to the direction of the muscular fibers, of skeletal muscle and are comparable with results from encounter a greater number of tissue interfaces, resulting in published studies (12, 17, 23, 27, 30). Because of the well-known more heterogeneous measurements. Conversely, shear waves physical assumption that skeletal muscle is an anisotropic that propagate along the direction of the muscular fibers could (orthotropic), non-linearly viscoelastic, and deformable tissue, generate more homogeneous measurements (29). Further, Frontiers in Neurology 08 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 6 Boxplot reporting median and inter-quartile ranges of shear wave elastography velocity (SWV) values obtained during the 16 assessments. NP, non-paretic limb (white background); P, paretic limb (gray background); sh, shortened muscle (dotted outline); st, stretched muscle (solid outline); Ax, axial probe position (horizontal stripes); Lo, longitudinal probe position (oblique stripes); T0 (left box), T1 (right plot). For example, NPshAxT0 represents the median SWV values among the selected unaffected (non-paretic) muscles of the 14 patients kept in a shortened position with the ultrasound probe in the axial position at T0. Horizontal bars indicate significant comparisons (p < 0.01). orthotropic physical properties of skeletal muscle are also allowed the evaluation of a wide spectrum of muscle changes, responsible for the fact that shear waves travel faster along the considering that muscle changes, particularly the fibrous-fatty direction of the fibers than they do when perpendicular to them, substitution of muscle fibers, increase with time. A direct which is another point resulted in the present study and already correlation between stroke age and SWV values has been demonstrated by previous papers (12, 15, 31, 32). demonstrated in three studies (19, 22, 36). The lack of Statistically significant differences emerged between correlation in this study can be probably attributed to the small velocities detected on paretic muscles and respective sample size, considering also many other nuisance variables contralateral ones, in particular, SWV from paretic muscles that could have influenced muscle stiffness increase after the was consistently higher. This is also an expected finding stroke event, such as the severity of the paresis, the amount and since stiffness is increased in paretic muscles compared to the efficacy of rehabilitative treatments, how long the limb remained contralateral unaffected side because of higher fibrosis (33) immobilized, and so on. and the presence of EMG activity in a neutral or shortened Conversely, a good correlation emerged between time since position (spastic dystonia) and while kept stretched (static phase stroke and MHS values, confirming that the structural changes of spasticity or spastic dystonia occurring in flexor muscles). occurring to paretic muscles over time can be detectable with Furthermore, the greater reduction of SWV detected at T1 the conventional US. As expected, MHS scores did not change in muscles in the stretched condition in comparison to the between T0 and T1, confirming that BoNT-A does not affect shortened position was an expected finding due to the exclusive structural fibrotic changes of paretic muscle, at least after a single effect of BoNT-A on reflex hypertonia. Conversely, SWV in treatment. Most importantly, MHS values showed a positive non-paretic muscles did not consistently change between the correlation with SWV measurements while the muscle was in two time points T0-T1: this crucial finding indirectly reinforces the shortened position, but not with spastic dystonia ARV. data resulting from paretic muscles. In 2015, Kesikburun This finding supports SWE as a useful method to estimate found that SWV was increased in the gastrocnemius muscle on muscle fibrosis independently of the effect of spastic dystonia. the affected side compared to the unaffected side in subjects It is important to underline that SWE measurement of muscle with unilateral stroke (34). Some other authors (16, 21, 35) stiffness cannot tell apart neural-mediated and not neural- did not show a significant difference between the paretic and mediated muscle hypertonia. non-paretic limbs, but they showed a difference between limbs Therefore, as reported in previous studies (17, 24), and control groups supporting the fact that hypertonic muscles SWE is sensitive enough to detect stiffness reduction tend to have greater SWV than non-hypertonic ones. following BoNT-A treatment, but, differently from EMG Recruitment of patients with quite different time intervals assessment, cannot demonstrate that SWV reduction is between the stroke and the current assessment (1–15 years) due to decreased muscle activity. Conversely, sEMG is Frontiers in Neurology 09 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 7 Example of SWE difference between axial and longitudinal assessments. In the upper-right quadrant is illustrated an example of an elastogram (superimposed to the B-mode image) generated by an ROI placed in a non-paretic, shortened biceps brachii muscular belly across the direction of the fibers; in the lower right image, the elastogram generated along the direction of muscular fibers of the same muscle shows higher values expressed both qualitatively with different-colored map and quantitatively in m/s (lower left corners of the images). Images on the upper- and lower-left quadrant are the respective B-mode conventional US image automatically paired on the screen by the SEL software (axial US scan in the upper-left quadrant and longitudinal US scan in the lower-left quadrant). unable to provide information about muscle fibrosis, subjects and hypertonic muscles from patients with other but when sEMG activity is very limited, as in spastic conditions determining upper motor neuron lesions muscles kept in a shortened position, SWE can provide a (such as multiple sclerosis and spinal cord injuries). measurement of stiffness due almost completely to intrinsic SWE measurements were obtained only by a single muscle changes. US-experienced operator. SWE has the same issues as This study presents many limitations. Our preliminary ultrasonography, which is operator-dependent, and, to date, findings should be reproduced in a large number of it lacks a methodological standardization (12, 15, 27, 37). Frontiers in Neurology 10 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 FIGURE 8 Representative MHS grade II. Conventional US appearance of biceps brachii and brachialis muscles on non-paretic (NP) and paretic side (P) following stroke, as assessed with the axial US scans: on the left image normal US appearance of muscle with regular echogenicity of fibers and normal muscular fibers—intramuscular septa differentiation; the right image shows volume reduction of muscle and increase in background echogenicity with intramuscular septa more represented and still definable. Indeed, we tried to reduce measurement variability by reached the same level of muscle relaxation/contraction providing for each condition a final median SWE value during the two measures. Finally, the number of patients resulting from 10 SWE measurements over the entire enrolled in the study was relatively small. Therefore, muscular volume. Even if several studies report good future studies assessing different muscle groups in a reliability of SWE in the assessment of skeletal muscle larger population with post-stroke muscle hypertonia, stiffness (12–14), given the technical difficulties, SWE using multiple US scanners with standardized techniques, requires being properly performed, and less experienced are warranted. examiners could not achieve the same results. Further, In conclusion, our study confirms that SWE is capable all SWE measurements were conducted with a single US of detecting muscle stiffness related to spastic dystonia and equipment: this could have been a limitationparticularly the effect of botulinum toxin treatment. The longitudinal considering that several manufacturers had implemented probe position for SWE assessment on spastic muscles their algorithms and post-processing calculation to appears the most suitable to detect stiffness changes. Most assess mechanical properties quantifying the shear interestingly, we showed that muscle hyperechogenicity wave speed. measured with the MHS positively correlated with SWE in Electromyographic activity was measured at rest with a shortened muscle independently of the contribution of the examined jointly in a neutral position (spastic dystonia) neural-mediated components. and while being stretched (dynamic stretch reflex), while SWV was calculated with the muscle in a shortened (not necessarily neutral) and while kept stretched (not during the Data availability statement dynamic phase of the stretching movement). It is possible that the ARV values obtained with the limb in a neutral The raw data supporting the conclusions of this article will position could be slightly higher than those obtained with be made available by the authors, without undue reservation. the examined muscle in a shortened position; similarly, the ARV values recorded during the dynamic phase are likely higher than that appearing while the muscle Ethics statement is kept stretched. Furthermore, since EMG and SWE assessments have been performed in different sessions and The studies involving human participants were reviewed environmental conditions, we cannot be sure that patients and approved by Comitato Etico Regionale della Liguria. The Frontiers in Neurology 11 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 patients/participants provided their written informed consent to Conflict of interest participate in this study. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could Author contributions be construed as a potential conflict of interest. WC performed the electromyographic recording and drafted the manuscript. AC performed the ultrasonographic Publisher’s note acquisitions and drafted the manuscript. LPu analyzed electromyographic data. LPr and RP performed the All claims expressed in this article are solely those of the electromyographic recording and collected all data. LMo, authors and do not necessarily represent those of their affiliated LB, GT, and CT critically revised the work for important organizations, or those of the publisher, the editors and the intellectual content. LMa conceived and designed the study, reviewers. Any product that may be evaluated in this article, or performed statistical analyses, and drafted the manuscript. All claim that may be made by its manufacturer, is not guaranteed authors approved the final version. or endorsed by the publisher. References 1. Watkins CL, Leathley MJ, Gregson JM, Moore AP, Smith TL, 14. Šarabon N, Kozinc Ž, Podrekar N. Using shear-wave elastography in skeletal Sharma AK. Prevalence of spasticity post stroke. Clin Rehabil. (2002) muscle: a repeatability and reproducibility study on biceps femoris muscle. PLoS 16:515–22. doi: 10.1191/0269215502cr512oa ONE. (2019) 14:e0222008. doi: 10.1371/journal.pone.0222008 2. Trompetto C, Marinelli L, Mori L, Pelosin E, Currà A, Molfetta L, et al. 15. Eby S, Zhao H, Song P, Vareberg BJ, Kinnick R, Greenleaf JF, et al. Pathophysiology of spasticity: implications for neurorehabilitation. Biomed Res Int. Quantitative evaluation of passive muscle stiffness in chronic stroke. Am J Phys (2014) 2014:354906. doi: 10.1155/2014/354906 Med Rehabil. (2016) 95:899–910. doi: 10.1097/PHM.0000000000000516 3. Trompetto C, Marinelli L, Puce L, Mori L, Serrati C, Fattapposta F, 16. Gao J, He W, Du L-J, Chen J, Park D, Wells M, et al. Quantitative et al. “Spastic dystonia” or “Inability to voluntary silence EMG activity”? ultrasound imaging to assess the biceps brachii muscle in chronic post- Time for clarifying the nomenclature. Clin Neurophysiol. (2019) 130:1076– stroke spasticity: preliminary observation. Ultrasound Med Biol. (2018) 44:1931– 7. doi: 10.1016/j.clinph.2019.03.009 40. doi: 10.1016/j.ultrasmedbio.2017.12.012 4. Trompetto C, Currà A, Puce L, Mori L, Serrati C, Fattapposta F, et al. 17. Gao J, Rubin JM, Chen J, O’Dell M. Ultrasound elastography to assess Spastic dystonia in stroke subjects: prevalence and features of the neglected botulinum toxin A treatment for post-stroke spasticity: a feasibility study. phenomenon of the upper motor neuron syndrome. Clin Neurophysiol. (2019) Ultrasound Med Biol. (2019) 45:1094–102. doi: 10.1016/j.ultrasmedbio.2018.10.034 130:521–7. doi: 10.1016/j.clinph.2019.01.012 18. Jakubowski KL, Terman A, Santana RVC, Lee SSM. Passive material 5. Trompetto C, Currà A, Puce L, Mori L, Pallecchi I, Gazzola P, properties of stroke-impaired plantarflexor and dorsiflexor muscles. Clin Biomech et al. Ghost spasticity in multiple sclerosis. J Electromyogr Kinesiol. (2020) (Bristol, Avon). (2017) 49:48–55. doi: 10.1016/j.clinbiomech.2017.08.009 51:102408. doi: 10.1016/j.jelekin.2020.102408 19. Lee SSM, Jakubowski KL, Spear SC, Rymer WZ. Muscle material properties in passive and active stroke-impaired muscle. J Biomech. (2019) 83:197– 6. Puce L, Currà A, Marinelli L, Mori L, Capello E, Di Giovanni R, et al. 204. doi: 10.1016/j.jbiomech.2018.11.043 Spasticity, spastic dystonia, and static stretch reflex in hypertonic muscles of patients with multiple sclerosis. Clin Neurophysiol Pract. (2021) 6:194– 20. Le Sant G, Nordez A, Hug F, Andrade R, Lecharte T, McNair PJ, 202. doi: 10.1016/j.cnp.2021.05.002 Gross R. Effects of stroke injury on the shear modulus of the lower leg muscle during passive dorsiflexion. J Appl Physiol (1985). (2019) 126:11– 7. Meseguer-Henarejos A-B, Sánchez-Meca J, López-Pina J-A, Carles-Hernández 22. doi: 10.1152/japplphysiol.00968.2017 R. Inter- and intra-rater reliability of the modified Ashworth scale: a systematic review and meta-analysis. Eur J Phys Rehabil Med. (2018) 54:576– 21. Mathevon L, Michel F, Aubry S, Testa R, Lapole T, Arnaudeau LF, et al. Two- 90. doi: 10.23736/S1973-9087.17.04796-7 dimensional and shear wave elastography ultrasound: A reliable method to analyse spastic muscles? Muscle Nerve. (2018) 57:222–8. doi: 10.1002/mus.25716 8. Zúñiga LDO, López CAG, González ER. Ultrasound elastography in the assessment of the stiffness of spastic muscles: a systematic review. Ultrasound Med 22. Wu C-H, Ho Y-C, Hsiao M-Y, Chen W-S, Wang T-G. Evaluation of Biol. (2021) 47:1448–64. doi: 10.1016/j.ultrasmedbio.2021.01.031 post-stroke spastic muscle stiffness using shear wave ultrasound elastography. Ultrasound Med Biol. (2017) 43:1105–11. doi: 10.1016/j.ultrasmedbio.2016.12.008 9. Klauser AS, Miyamoto H, Bellmann-Weiler R, Feuchtner GM, Wick MC, Jaschke WR. Sonoelastography: musculoskeletal applications. Radiology. (2014) 23. Leng Y, Wang Z, Bian R, Lo WLA, Xie X, Wang R, et al. Alterations 272:622–33. doi: 10.1148/radiol.14121765 of elastic property of spastic muscle with its joint resistance evaluated from shear wave elastography and biomechanical model. Front Neurol. (2019) 10. Bamber J, Cosgrove D, Dietrich CF, Fromageau J, Bojunga J, Calliada F, 10:736. doi: 10.3389/fneur.2019.00736 et al. EFSUMB guidelines and recommendations on the clinical use of ultrasound elastography. Part 1: Basic principles and technology. Ultraschall Med. (2013) 24. Cao J, Xiao Y, Qiu W, Zhang Y, Dou Z, Ren J, et al. Reliability 34:169–84. doi: 10.1055/s-0033-1335205 and diagnostic accuracy of corrected slack angle derived from 2D-SWE in quantitating muscle spasticity of stroke patients. J NeuroEng Rehabil. (2022) 11. Ryu J, Jeong WK. Current status of musculoskeletal application of shear wave 19:15. doi: 10.1186/s12984-022-00995-8 elastography. Ultrasonography. (2017) 36:185–97. doi: 10.14366/usg.16053 25. Hermens HJ. European recommendations for surface ElectroMyoGraphy: 12. Creze M, Nordez A, Soubeyrand M, Rocher L, Maître X, Bellin M-F. results of the SENIAM project. In: BV (eds) Roessingh Research and Development. Shear wave sonoelastography of skeletal muscle: basic principles, biomechanical Enschede: Roessingh Research and Development (1999). 122 p. concepts, clinical applications, and future perspectives. Skeletal Radiol. (2018) 47:457–71. doi: 10.1007/s00256-017-2843-y 26. Marinelli L, Trompetto C, Mori L, Vigo G, Traverso E, Colombano F, et al. Manual linear movements to assess spasticity in a clinical 13. Phan A, Lee J, Gao J. Ultrasound shear wave elastography in assessment setting. PLoS ONE. (2013) 8:e53627. doi: 10.1371/journal.pone.00 of skeletal muscle stiffness in senior volunteers. Clin Imaging. (2019) 58:22– 53627 6. doi: 10.1016/j.clinimag.2019.06.006 Frontiers in Neurology 12 frontiersin.org Campanella et al. 10.3389/fneur.2022.980746 27. Lehoux M-C, Sobczak S, Cloutier F, Charest S, Bertrand-Grenier A. 32. Miyamoto N, Hirata K, Kanehisa H, Yoshitake Y. Validity of measurement of Shear wave elastography potential to characterize spastic muscles in stroke shear modulus by ultrasound shear wave elastography in human pennate muscle. survivors: Literature review. Clin Biomech (Bristol, Avon). (2020) 72:84– PLoS ONE. (2015) 10:e0124311. doi: 10.1371/journal.pone.0124311 93. doi: 10.1016/j.clinbiomech.2019.11.025 33. Liu J, Pan H, Bao Y, Zhao Y, Huang L, Zhan W. The value of real-time shear 28. Moreta MC, Fleet A, Reebye R, McKernan G, Berger M, Farag J, et al. wave elastography before and after rehabilitation of upper limb spasm in stroke Reliability and validity of the modified heckmatt scale in evaluating muscle patients. Biomed Res Int. (2020) 2020:1–7. doi: 10.1155/2020/6472456 changes with ultrasound in spasticity. Arch Rehabil Res Clin Transl. (2020) 34. Kesikburun S, Yaşar E, Adigüzel E, Güzelküçük Ü, Alaca R, Tan AK. 2:100071. doi: 10.1016/j.arrct.2020.100071 Assessment of spasticity with sonoelastography following stroke: a feasibility study. 29. Miller T, Ying M, Sau Lan Tsang C, Huang M, Pang MYC. Reliability and PM&R. (2015) 7:1254–60. doi: 10.1016/j.pmrj.2015.05.019 validity of ultrasound elastography for evaluating muscle stiffness in neurological 35. Rasool G, Wang AB, Rymer WZ, Lee SSM. Shear waves reveal viscoelastic populations: a systematic review and meta-analysis. Physical Therapy. (2021) changes in skeletal muscles after hemispheric stroke. IEEE Trans Neural Syst 101:pzaa188. doi: 10.1093/ptj/pzaa188 Rehabil Eng. (2018) 26:2006–14. doi: 10.1109/TNSRE.2018.2870155 30. Gennisson J-L, Deffieux T, Macé E, Montaldo G, Fink M, Tanter M. 36. Lee SSM, Spear S, Rymer WZ. Quantifying changes in Viscoelastic and anisotropic mechanical properties of in vivo muscle tissue material properties of stroke-impaired muscle. Clin Biomech. (2015) assessed by supersonic shear imaging. Ultrasound Med Biol. (2010) 36:789– 30:269–75. doi: 10.1016/j.clinbiomech.2015.01.004 801. doi: 10.1016/j.ultrasmedbio.2010.02.013 37. Shin HJ, Kim M-J, Kim HY, Roh YH, Lee M-J. Comparison of 31. Dorado Cortez C, Hermitte L, Ramain A, Mesmann C, Lefort T, Pialat JB. shear wave velocities on ultrasound elastography between different machines, Ultrasound shear wave velocity in skeletal muscle: a reproducibility study. Diagn transducers, and acquisition depths: a phantom study. Eur Radiol. (2016) 26:3361– Interv Imaging. (2016) 97:71–9. doi: 10.1016/j.diii.2015.05.010 7. doi: 10.1007/s00330-016-4212-y Frontiers in Neurology 13 frontiersin.org