QUANTITATIVE ULTRASOUND OF DENERVATED HAND MUSCLES NEIL G. SIMON, MBBS, FRACP,1,2,3 JEFFREY W. RALPH, MD,1 CATHERINE LOMEN-HOERTH, MD, PhD,1 ANN N. PONCELET, MD,1 STEVE VUCIC, PhD, FRACP,2,4 MATTHEW C. KIERNAN, DSc, FRACP,5 and MICHEL KLIOT, MD6 1 Department of Neurology, University of California, San Francisco, San Francisco, California, USA 2 Neuroscience Research Australia, PO Box 1165, Randwick NSW 2031, Australia 3 Prince of Wales Clinical School, University of New South Wales, Australia 4 Westmead Clinical School, C24 Westmead Hospital, The University of Sydney, NSW 2006, Australia 5 Brain and Mind Research Institute, The University of Sydney, Mallett St, Camperdown, Australia 6 Department of Neurological Surgery, Northwestern Feinberg School of Medicine, Chicago, Illinois, USA Accepted 10 November 2014 ABSTRACT: Introduction: Presentations to the neuromuscular clinic commonly involve hand muscle denervation, but few studies have evaluated hand muscle ultrasound. Methods: Ultrasound studies of abductor pollicis brevis, first dorsal interosseous, and abductor digit minimi were prospectively performed in a cohort of 34 patients (77 muscles) with electromyography (EMG)-confirmed denervation, compared with 58 healthy control subjects. Results: In control subjects, muscle thickness was highly reproducible [intraclass correlation coefficient (ICC) 5 0.88–0.98], and echogenicity was moderately reproducible (ICC 5 0.542–0.686). Age, gender, and body mass index influenced muscle thickness and echogenicity. Ultrasound changes in denervated muscles correlated with the severity of EMG abnormalities. A z-score cutoff of 0 identified denervated muscles with a sensitivity of 100% and 89% for echogenicity and muscle thickness, respectively. Conclusions: Hand muscle ultrasound provides a noninvasive method to quantify muscle denervation and may be useful as a screening tool before EMG studies. Muscle Nerve 52: 221–230, 2015
Diagnostic ultrasound has made substantial inroads into neuromuscular and electrodiagnostic clinics. Presently, ultrasound has developed utility in the diagnosis of compressive neuropathy1 and is now considered by many clinicians to be an adjunct to clinical and neurophysiological diagnosis.2,3 Diagnostic muscle ultrasound detects changes of muscle denervation,4–6 and these changes have been used to quantify progression in diseases associated with progressive muscle denervation, such as amyotrophic lateral sclerosis (ALS).7–9 However, studies of quantitative muscle ultrasound are generally limited to large proximal muscles, such as biceps brachii, and to date there have been no comprehensive studies that have explored quantitative ultrasound techniques in hand muscles. Abbreviations: ADM, abductor digiti minimi; ALS, amyotrophic lateral sclerosis; APB, abductor pollicis brevis; BMI, body mass index; CSA, cross-sectional area; EMG, electromyography; FDI, first dorsal interosseous; ICC, intraclass correlation coefficient; MRC, medical research council; MUNE, motor unit number estimation; MUP, motor unit potential; PSW, positive sharp wave; ROI, region of interest; SD, standard deviation; SEM, standard error of the mean. Key words: amyotrophic lateral sclerosis; denervation; muscle; nerve injury; ultrasound Correspondence to: N. G. Simon; e-mail: [email protected] C 2014 Wiley Periodicals, Inc. V
Published online 11 November 2014 in Wiley Online Library (wileyonlinelibrary. com). DOI 10.1002/mus.24519
Hand Muscle Ultrasound
Evaluating hand muscles with quantitative ultrasound may be of particular relevance, as these muscles are involved frequently in patients who present to neuromuscular or electrodiagnostic clinics with conditions such as mononeuropathy 10 and motor neuron diseases.11,12 In addition, many existing techniques used to quantitate upper limb nerve injury, such as nerve conduction studies and motor unit number estimation (MUNE), rely on measurements obtained from the small muscles of the hand, typically abductor pollicis brevis (APB), first dorsal interosseous (FDI), and abductor digiti minimi (ADM).13–15 This study was undertaken to evaluate the role of quantitative ultrasound in the assessment of neuromuscular disorders involving denervation of hand muscles. Specific aims of the study were to evaluate the test–retest reliability of quantitative measurements and to determine demographic and morphometric factors that contribute to variation in ultrasound measurements in a healthy population. Furthermore, this study aimed to delineate the ultrasound changes observed in denervated hand muscles to establish the role of ultrasound in discriminating between normal and abnormal hand muscles. METHODS
We prospectively studied the ultrasound characteristics of the APB, FDI, and ADM in healthy control subjects and patients with muscle denervation. Quantitative ultrasound measurements were compared with EMG and clinical parameters in patients with muscle denervation. All subjects provided written informed consent. The study was approved by the University of California, San Francisco, Committee on Human Research. A healthy control cohort was studied consisting of 58 subjects recruited from relatives and caregivers of patients attending the University of California, San Francisco, Electrodiagnostic and Neuromuscular Clinics and members of the general community. Control subjects were excluded if they had evidence on clinical examination or past history of peripheral polyneuropathy, median or
Subjects.
MUSCLE & NERVE
August 2015
221
ulnar neuropathy, brachial plexopathy, cervical radiculopathy, diabetes mellitus, or previous significant hand trauma. A cohort of 34 patients with hand muscle denervation confirmed on EMG studies were included for analysis. A total of 49 limbs had 1 or more muscles with evidence of denervation on EMG (56% of muscles demonstrated active denervation with fibrillation potentials and positive sharp waves, and 44% of muscles demonstrated chronic neurogenic motor unit potential changes without fibrillation potentials). Disease processes were ALS in 10 patients, carpal tunnel syndrome in 7 patients, ulnar neuropathy in 12 patients, brachial plexopathy involving the lower trunk in 5 patients, and peripheral polyneuropathy in 1 patient. Patients were excluded if there was a history of significant hand trauma or arthritis. Clinical Measurements. A complete neurological examination was performed in all patients with suspected nerve injury. Muscle strength of finger and thumb abduction was recorded using the Medical Research Council (MRC) grading system.16 Height, weight, and handedness were recorded for all subjects. Body mass index (BMI) was calculated. Ultrasound Examinations. A Mindray M7 ultrasonic imaging system was used with a linear array transducer (L14-6s, Mindray, Shenzen, China; nominal frequency range, 6–14 MHZ), with peak frequency set at 12 MHZ. For muscle imaging, uniform device settings were used for each subject and stored as a preset (acoustic power 98%, line density set at medium, dynamic range set at 65, persistence set at 1, iClear set at 4). The time-gain compensation settings were held fixed in the midline, and gain was set at 94% for all images. All ultrasound studies were performed by the same investigator (N.G.S.) who was blinded to electrodiagnostic studies and clinical details. The process of acquiring the ultrasound images necessarily meant that this investigator was not blinded to gross muscle atrophy evident on inspection. Images were acquired with the transducer applied lightly to the skin to avoid muscle compression. Three hand muscles were imaged in both hands of each control subject (Fig. 1): APB, FDI, and ADM. In the muscle denervation cohort APB and FDI (not ADM) imaging data were analyzed, as EMG data were available for these muscles. Images were acquired in the cross-sectional and longitudinal planes. For cross-sectional image acquisition, the ultrasound transducer was oriented perpendicular to the underlying bone (first, second, and fifth metacarpals, respectively) by adjusting it to yield the brightest and most narrow bone interface. For longitudinal image acquisition, the 222
Hand Muscle Ultrasound
angle of the transducer was optimized using the first metacarpal for APB and the hyperechoic fascia underlying FDI and ADM. Images for analysis were acquired with an image depth of 1.8 cm and a single focal point fixed 1cm deep within the image. In subjects with larger muscles, the image depth was increased to enable accurate measurement of the muscle, but the image acquired with the original depth settings was used for quantitative grayscale analysis. The APB muscle was imaged with the forearm supinated and the thumb held in the neutral position. A mark was placed at the midpoint of the line connecting the volar aspect of the first metacarpophalyngeal joint and the volar prominence of the scaphoid bone. Cross-sectional images were acquired perpendicular to this line with the mark as the midpoint of the image. Longitudinal images were acquired with the transducer held along the line and midpoint corresponding to the location of the cross-sectional image acquisition. The FDI muscle was imaged with the forearm in the neutral position, the fingers extended, and the fingers and thumb adducted. A line was made perpendicular to the axis of the second metacarpal at the level of the first metacarpophalyngeal joint. Cross-sectional images were acquired with the transducer held on this line and perpendicular to the axis of the second metacarpal. Longitudinal images were acquired with the transducer held parallel to the second metacarpal with the midpoint over the marked line. The ADM muscle was imaged with the forearm pronated and the fingers extended and adducted. A mark was made at the midpoint of the line connecting the fifth metacarpophalyngeal joint and pisiform bone over the bulk of the muscle. Crosssectional images were acquired perpendicular to this line with the mark as the midpoint of the image. Longitudinal images were acquired with the transducer held along the line and midpoint corresponding to the location of the cross-sectional image acquisition. Muscle thickness on both cross-sectional and longitudinal images was calculated using the “distance” measurement function of the ultrasound device. The cross-sectional area (CSA) of the FDI and ADM muscles was measured on the crosssectional images using the “trace area” measurement function of the ultrasound device. CSA of APB was not measured as the outer margins of the muscle were not always distinct. Muscle grayscale analysis was performed using ImageJ software (National Institutes of Health, Bethesda, Maryland). The images were converted to 8-bit, attributing every pixel a gray scale value ranging from 0 (black) to 255 (white). A freehand MUSCLE & NERVE
August 2015
FIGURE 1. Representative ultrasound images of normal APB (axial image, A, longitudinal image, B), FDI (axial image, C; longitudinal image, D), and ADM (axial image, E; longitudinal image, F). The fascial plane delineating APB from opponens pollicis (OP) is shown (white arrows). Corresponding metacarpal (MC) bones are marked.
region of interest (ROI) was selected by tracing just within the hyperechoic superficial fascia of the muscle, and mean grayscale values were measured. Test–retest Measurements. A subgroup of 32 healthy control subjects underwent 3 separate examinations as part of a test–retest reliability study. Two ultrasound examinations were separated by approximately 20 min in the first testing session, and a third examination was performed on a separate day. Quantitative ultrasound measurements from 3 sessions were analyzed for the following parameters: CSA of FDI and ADM, thickness and echogenicity of the axial image of each muscle, and Hand Muscle Ultrasound
thickness and echogenicity of the longitudinal image of each muscle. Electrodiagnostic Studies. Patients with suspected muscle denervation underwent a standardized electrodiagnostic examination consisting of median and ulnar motor and sensory nerve conduction studies, including F-waves. Electrodiagnostic studies were performed by 2 investigators (J.W.R. and A.N.P.), and values were compared with the University of California San Francisco electrodiagnostic laboratory normal values, with unique normal values categorized by age (5 to 40; 41 to 60; 61 to 70; 71 to 80). Standardized needle EMG studies were performed, which included APB, FDI, biceps, triceps, and deltoid using a conventional approach.15 MUSCLE & NERVE
August 2015
223
Additional muscle sampling was determined by the clinical presentation. EMG studies were graded for each muscle. Fibrillation potentials and positive sharp waves (PSWs) were scored from 0 to 4 depending on their density and distribution within the muscle.17 Motor unit morphology was graded qualitatively according to the extent of neurogenic motor unit rearrangement, based on quantitative analysis of the duration of individual motor unit potentials (MUPs) and determined by the reporting neurophysiologist. In the absence of an established score, a novel EMG grading scale was devised for the study. Duration was considered normal if 10% of MUPs demonstrated duration >14 ms (0 points).18,19 Abnormal MUPs were scored as mild prolongation affecting
Diagnostic ultrasound has made substantial inroads into neuromuscular and electrodiagnostic clinics. Presently, ultrasound has developed utility in the diagnosis of compressive neuropathy1 and is now considered by many clinicians to be an adjunct to clinical and neurophysiological diagnosis.2,3 Diagnostic muscle ultrasound detects changes of muscle denervation,4–6 and these changes have been used to quantify progression in diseases associated with progressive muscle denervation, such as amyotrophic lateral sclerosis (ALS).7–9 However, studies of quantitative muscle ultrasound are generally limited to large proximal muscles, such as biceps brachii, and to date there have been no comprehensive studies that have explored quantitative ultrasound techniques in hand muscles. Abbreviations: ADM, abductor digiti minimi; ALS, amyotrophic lateral sclerosis; APB, abductor pollicis brevis; BMI, body mass index; CSA, cross-sectional area; EMG, electromyography; FDI, first dorsal interosseous; ICC, intraclass correlation coefficient; MRC, medical research council; MUNE, motor unit number estimation; MUP, motor unit potential; PSW, positive sharp wave; ROI, region of interest; SD, standard deviation; SEM, standard error of the mean. Key words: amyotrophic lateral sclerosis; denervation; muscle; nerve injury; ultrasound Correspondence to: N. G. Simon; e-mail: [email protected] C 2014 Wiley Periodicals, Inc. V
Published online 11 November 2014 in Wiley Online Library (wileyonlinelibrary. com). DOI 10.1002/mus.24519
Hand Muscle Ultrasound
Evaluating hand muscles with quantitative ultrasound may be of particular relevance, as these muscles are involved frequently in patients who present to neuromuscular or electrodiagnostic clinics with conditions such as mononeuropathy 10 and motor neuron diseases.11,12 In addition, many existing techniques used to quantitate upper limb nerve injury, such as nerve conduction studies and motor unit number estimation (MUNE), rely on measurements obtained from the small muscles of the hand, typically abductor pollicis brevis (APB), first dorsal interosseous (FDI), and abductor digiti minimi (ADM).13–15 This study was undertaken to evaluate the role of quantitative ultrasound in the assessment of neuromuscular disorders involving denervation of hand muscles. Specific aims of the study were to evaluate the test–retest reliability of quantitative measurements and to determine demographic and morphometric factors that contribute to variation in ultrasound measurements in a healthy population. Furthermore, this study aimed to delineate the ultrasound changes observed in denervated hand muscles to establish the role of ultrasound in discriminating between normal and abnormal hand muscles. METHODS
We prospectively studied the ultrasound characteristics of the APB, FDI, and ADM in healthy control subjects and patients with muscle denervation. Quantitative ultrasound measurements were compared with EMG and clinical parameters in patients with muscle denervation. All subjects provided written informed consent. The study was approved by the University of California, San Francisco, Committee on Human Research. A healthy control cohort was studied consisting of 58 subjects recruited from relatives and caregivers of patients attending the University of California, San Francisco, Electrodiagnostic and Neuromuscular Clinics and members of the general community. Control subjects were excluded if they had evidence on clinical examination or past history of peripheral polyneuropathy, median or
Subjects.
MUSCLE & NERVE
August 2015
221
ulnar neuropathy, brachial plexopathy, cervical radiculopathy, diabetes mellitus, or previous significant hand trauma. A cohort of 34 patients with hand muscle denervation confirmed on EMG studies were included for analysis. A total of 49 limbs had 1 or more muscles with evidence of denervation on EMG (56% of muscles demonstrated active denervation with fibrillation potentials and positive sharp waves, and 44% of muscles demonstrated chronic neurogenic motor unit potential changes without fibrillation potentials). Disease processes were ALS in 10 patients, carpal tunnel syndrome in 7 patients, ulnar neuropathy in 12 patients, brachial plexopathy involving the lower trunk in 5 patients, and peripheral polyneuropathy in 1 patient. Patients were excluded if there was a history of significant hand trauma or arthritis. Clinical Measurements. A complete neurological examination was performed in all patients with suspected nerve injury. Muscle strength of finger and thumb abduction was recorded using the Medical Research Council (MRC) grading system.16 Height, weight, and handedness were recorded for all subjects. Body mass index (BMI) was calculated. Ultrasound Examinations. A Mindray M7 ultrasonic imaging system was used with a linear array transducer (L14-6s, Mindray, Shenzen, China; nominal frequency range, 6–14 MHZ), with peak frequency set at 12 MHZ. For muscle imaging, uniform device settings were used for each subject and stored as a preset (acoustic power 98%, line density set at medium, dynamic range set at 65, persistence set at 1, iClear set at 4). The time-gain compensation settings were held fixed in the midline, and gain was set at 94% for all images. All ultrasound studies were performed by the same investigator (N.G.S.) who was blinded to electrodiagnostic studies and clinical details. The process of acquiring the ultrasound images necessarily meant that this investigator was not blinded to gross muscle atrophy evident on inspection. Images were acquired with the transducer applied lightly to the skin to avoid muscle compression. Three hand muscles were imaged in both hands of each control subject (Fig. 1): APB, FDI, and ADM. In the muscle denervation cohort APB and FDI (not ADM) imaging data were analyzed, as EMG data were available for these muscles. Images were acquired in the cross-sectional and longitudinal planes. For cross-sectional image acquisition, the ultrasound transducer was oriented perpendicular to the underlying bone (first, second, and fifth metacarpals, respectively) by adjusting it to yield the brightest and most narrow bone interface. For longitudinal image acquisition, the 222
Hand Muscle Ultrasound
angle of the transducer was optimized using the first metacarpal for APB and the hyperechoic fascia underlying FDI and ADM. Images for analysis were acquired with an image depth of 1.8 cm and a single focal point fixed 1cm deep within the image. In subjects with larger muscles, the image depth was increased to enable accurate measurement of the muscle, but the image acquired with the original depth settings was used for quantitative grayscale analysis. The APB muscle was imaged with the forearm supinated and the thumb held in the neutral position. A mark was placed at the midpoint of the line connecting the volar aspect of the first metacarpophalyngeal joint and the volar prominence of the scaphoid bone. Cross-sectional images were acquired perpendicular to this line with the mark as the midpoint of the image. Longitudinal images were acquired with the transducer held along the line and midpoint corresponding to the location of the cross-sectional image acquisition. The FDI muscle was imaged with the forearm in the neutral position, the fingers extended, and the fingers and thumb adducted. A line was made perpendicular to the axis of the second metacarpal at the level of the first metacarpophalyngeal joint. Cross-sectional images were acquired with the transducer held on this line and perpendicular to the axis of the second metacarpal. Longitudinal images were acquired with the transducer held parallel to the second metacarpal with the midpoint over the marked line. The ADM muscle was imaged with the forearm pronated and the fingers extended and adducted. A mark was made at the midpoint of the line connecting the fifth metacarpophalyngeal joint and pisiform bone over the bulk of the muscle. Crosssectional images were acquired perpendicular to this line with the mark as the midpoint of the image. Longitudinal images were acquired with the transducer held along the line and midpoint corresponding to the location of the cross-sectional image acquisition. Muscle thickness on both cross-sectional and longitudinal images was calculated using the “distance” measurement function of the ultrasound device. The cross-sectional area (CSA) of the FDI and ADM muscles was measured on the crosssectional images using the “trace area” measurement function of the ultrasound device. CSA of APB was not measured as the outer margins of the muscle were not always distinct. Muscle grayscale analysis was performed using ImageJ software (National Institutes of Health, Bethesda, Maryland). The images were converted to 8-bit, attributing every pixel a gray scale value ranging from 0 (black) to 255 (white). A freehand MUSCLE & NERVE
August 2015
FIGURE 1. Representative ultrasound images of normal APB (axial image, A, longitudinal image, B), FDI (axial image, C; longitudinal image, D), and ADM (axial image, E; longitudinal image, F). The fascial plane delineating APB from opponens pollicis (OP) is shown (white arrows). Corresponding metacarpal (MC) bones are marked.
region of interest (ROI) was selected by tracing just within the hyperechoic superficial fascia of the muscle, and mean grayscale values were measured. Test–retest Measurements. A subgroup of 32 healthy control subjects underwent 3 separate examinations as part of a test–retest reliability study. Two ultrasound examinations were separated by approximately 20 min in the first testing session, and a third examination was performed on a separate day. Quantitative ultrasound measurements from 3 sessions were analyzed for the following parameters: CSA of FDI and ADM, thickness and echogenicity of the axial image of each muscle, and Hand Muscle Ultrasound
thickness and echogenicity of the longitudinal image of each muscle. Electrodiagnostic Studies. Patients with suspected muscle denervation underwent a standardized electrodiagnostic examination consisting of median and ulnar motor and sensory nerve conduction studies, including F-waves. Electrodiagnostic studies were performed by 2 investigators (J.W.R. and A.N.P.), and values were compared with the University of California San Francisco electrodiagnostic laboratory normal values, with unique normal values categorized by age (5 to 40; 41 to 60; 61 to 70; 71 to 80). Standardized needle EMG studies were performed, which included APB, FDI, biceps, triceps, and deltoid using a conventional approach.15 MUSCLE & NERVE
August 2015
223
Additional muscle sampling was determined by the clinical presentation. EMG studies were graded for each muscle. Fibrillation potentials and positive sharp waves (PSWs) were scored from 0 to 4 depending on their density and distribution within the muscle.17 Motor unit morphology was graded qualitatively according to the extent of neurogenic motor unit rearrangement, based on quantitative analysis of the duration of individual motor unit potentials (MUPs) and determined by the reporting neurophysiologist. In the absence of an established score, a novel EMG grading scale was devised for the study. Duration was considered normal if 10% of MUPs demonstrated duration >14 ms (0 points).18,19 Abnormal MUPs were scored as mild prolongation affecting
