CMAP variability as a function of electrode site and size

Revisiting the compound muscle action potential (CMAP)

The compound muscle action potential (CMAP) is among the first recorded waveforms in clinical neurography and one of the most common in clinical use. It is derived from the summated muscle fiber action potentials recorded from a surface electrode overlying the studied muscle following stimulation of the relevant motor nerve fibres innervating the muscle. Surface recorded motor unit potentials (SMUPs) are the fundamental units comprising the CMAP. Because it is considered a basic, if not banal signal, what it represents is often underappreciated. In this review we discuss current concepts in the anatomy and physiology of the CMAP. These have evolved with advances in instrumentation and digitization of signals, affecting its quantitation and measurement. It is important to understand the basic technical and biological factors influencing the CMAP. If these influences are not recognized, then a suboptimal recording may result. The object is to obtain a high quality CMAP recording that is reproducible, whether the study is done for clinical or research purposes. The initial sections cover the relevant CMAP anatomy and physiology, followed by how these principles are applied to CMAP changes in neuromuscular disorders. The concluding section is a brief overview of CMAP research where advances in recording systems and computer-based analysis programs have opened new research applications. One such example is motor unit number estimation (MUNE) that is now being used as a surrogate marker in monitoring chronic neurogenic processes such as motor neuron diseases.

Comparing Motor-Evoked Potential Characteristics of NEedle versus suRFACE Recording Electrodes during Spinal Cord Monitoring-The NERFACE Study Part I

Muscle-recorded transcranial electrical stimulation motor-evoked potentials (mTc-MEPs) are used to assess the spinal cord integrity. They are commonly recorded with subcutaneous needle or surface electrodes, but the different characteristics of mTc-MEP signals recorded with the two types of electrodes have not been formally compared yet. In this study, mTc-MEPs were simultaneously recorded from the tibialis anterior (TA) muscles using surface and subcutaneous needle electrodes in 242 consecutive patients. Elicitability, motor thresholds, amplitude, area under the curve (AUC), signal-to-noise ratio (SNR), and the variability between mTc-MEP amplitudes were compared. Whereas amplitude and AUC were significantly higher in subcutaneous needle recordings (p < 0.01), motor thresholds and elicitability were similar for surface and subcutaneous needle recordings. Moreover, the SNRs were >2 in more than 99.5% of the surface and subcutaneous needle recordings, and the variability between consecutive amplitudes was not significantly different between the two recording electrode types (p = 0.34). Surface electrodes appear to be a good alternative to needle electrodes for spinal cord monitoring. They are non-invasive, can record signals at similar threshold intensities, have adequately high SNRs, and record signals with equivalent variability. Whether surface electrodes are non-inferior to subcutaneous needle electrodes in detecting motor warnings is investigated in part II of the NERFACE study.

Comparison of nerve conduction parameters in canine studies using recording needle and surface electrodes

Surface electrodes have been used in electromyography and nerve conduction studies in human and veterinary medicine, but comparisons have not been made between surface and needle electrode recordings in dogs. Our aim in this method comparison study was to determine whether surface electrodes captured larger compound motor action potentials (CMAP) than needle electrodes. Tibial nerve CMAP from 25 dogs with normal limb function was acquired using both surface and needle recording electrodes; the stimulus was elicited with monopolar concentric needles. Paired Wilcoxon signed rank test (if data was not normally distributed) or a paired two tailed t-test was used if data were normally distributed; significance was set at P<0.05. Mean CMAP amplitude (P=0.009), area (P=0.045) and latency (P=0.02) recorded with needle electrodes were larger compared with surface recording. CMAP duration was not significantly longer when recorded with surface electrodes (P=0.898). Needle electrode recordings are suitable for canine studies, although surface electrodes could also be considered. Low CMAP amplitudes recorded with surface electrodes should be verified with needle electrodes.

Intramuscular Injection of Bone Marrow Stem Cells in Amyotrophic Lateral Sclerosis Patients: A Randomized Clinical Trial

Background

Preclinical studies suggest that stem cells may be a valuable therapeutic tool in amyotrophic lateral sclerosis (ALS). As it has been demonstrated that there are molecular changes at the end-plate during the early stages of motorneuron degeneration in animal models, we hypothesize that the local effect of this stem cell delivery method could slow the progressive loss of motor units (MUs) in ALS patients.

Methods

We designed a Phase I/II clinical trial to study the safety of intramuscularly implanting autologous bone marrow mononuclear cells (BMMCs), including stem cells, in ALS patients and their possible effects on the MU of the tibialis anterior (TA) muscle. Twenty-two patients participated in a randomized, double-blind, placebo-controlled trial that consisted of a baseline visit followed by one intramuscular injection of BMNCs, follow-up visits at 30, 90, 180, and 360 days, and an additional year of clinical follow-up. In each patient, one TA muscle was injected with a single dose of BMMCs while the contralateral muscle was given a placebo; the sides were selected randomly. All visits included a complete EMG study of both TA muscles.

Results

Our results show that (1) the intramuscular injection of BMMCs is a safe procedure; (2) ALS patients show heterogeneities in the degree of TA injury; (3) a comparison of placebo-injected muscles with BMMC-injected muscles showed significant differences in only one parameter, the D50 index used to quantify the Compound Muscle Action Potential (CMAP) scan curve. This parameter was higher in the BMMC-injected TA muscle at both 90 days (placebo side: 29.55 ± 2.89, n = 20; experimental side: 39.25 ± 3.21, n = 20; p < 0.01) and 180 days (placebo side: 29.35 ± 3.29, n = 17; experimental side: 41.24 ± 3.34, n = 17; p < 0.01).

Conclusion

This procedure had no effect on the TA muscle MU properties, with the exception of the D50 index. Finding differences in just this index supports the fact that it may be much more sensitive than other electrophysiological parameters when studying treatment effects. Given the low number of patients and their heterogeneity, these results justify exploring the efficacy of this procedure in further patients and other muscles, through Phase II trials.

Clinical Trial Registration

www.clinicaltrials.gov (identifier NCT02286011); EudraCT number 2011-004801-25.

Assessing Rat Forelimb and Hindlimb Motor Unit Connectivity as Objective and Robust Biomarkers of Spinal Motor Neuron Function

Sensitive and objective biomarkers of neuronal injury, degeneration, and regeneration can help facilitate translation of experimental findings into clinical testing. Whereas measures of upper motor neuron connectivity have been readily established, functional assessments of lower motor neuron (LMN) innervation of forelimb muscles are lacking. Compound muscle action potential (CMAP) and motor unit (MU) number estimation (MUNE) are well-established methods that allow longitudinal MU integrity monitoring in patients. In analogy we refined CMAP and MUNE methods for assessing spinal MU input in the rat forelimb and hindlimb. Repeated CMAP and MUNE recordings are robust (coefficients of variability: 4.5-11.3%), and MUNE measurements from forelimb wrist flexor muscles (415 ± 8 [SEM]) align with back-traced anatomical LMN counts (336 ± 16 [SEM]). For disease validation, cross-sectional blinded electrophysiological and muscle contractility measurements were obtained in a cohort of G93A SOD1 mutant overexpressing rats and compared with controls. Longitudinal assessment of mutant animals demonstrated progressive motor unit decline in the hindlimb to a greater extent than the forelimb. Hindlimb CMAP and MUNE demonstrated strong correlations with plantarflexion muscle contractility. Cross-species assessment of upper/fore- limb and lower/hind- limb motor units using objective electrophysiological CMAP and MUNE values as biomarkers will guide and improve bi-directional translation.

Global motor unit number index sum score for assessing the loss of lower motor neurons in amyotrophic lateral sclerosis

INTRODUCTION

We propose a motor unit number index (MUNIX) global sum score in amyotrophic lateral sclerosis (ALS) to estimate the loss of functional motor units.

METHODS

MUNIX was assessed for 18 ALS patients and 17 healthy controls in 7 muscles: the abductor pollicis brevis (APB), abductor digiti minimi (ADM), tibialis anterior (TA), deltoid, trapezius, submental complex, and orbicularis oris.

RESULTS

MUNIX was significantly lower in ALS patients than in healthy controls for the APB, ADM, TA, and trapezius muscles. The MUNIX sum score of 4 muscles (ADM + APB + trapezius + TA) was lower in ALS patients (P = 0.01) and was correlated with clinical scores.

DISCUSSION

The global MUNIX sum score proposed in this study estimates the loss of lower motor neurons in several body regions, including the trapezius, and is correlated with clinical impairment in ALS patients. Muscle Nerve 56: 202-206, 2017.

Responsiveness to botulinum toxin type A in muscles of complex regional pain patients with tonic dystonia

Tonic dystonia of the limbs in complex regional pain syndrome (CRPS) is associated with considerable disability. Treatment options are scarce. Botulinum toxin (BoNT) is sometimes used, but the effect is often said to be disappointing. However, this notion stems from case reports and clinicians' opinions but has never been formally studied. We therefore investigated responsiveness to BoNT in CRPS patients with tonic dystonia. We injected the extensor digitorum brevis (EDB) muscle with BoNT-A in 17 patients with CRPS and tonic dystonia to compare the response between affected and unaffected legs. We also investigated the right legs of 17 healthy controls. Responsiveness was defined as a decrease of the amplitude of the compound muscle action potential (CMAP) of >20% from baseline 2 weeks after BoNT-A injection. We controlled for a temperature effect on BoNT efficacy by measuring skin temperature hourly directly above the EDB muscle in the first 2 weeks. CMAP amplitude decreased >20% after injection on the affected side in 16 of 17 CRPS patients, similar to the response in unaffected legs (12/13) or legs of controls (17/17). The degree of CMAP reduction was significantly smaller in patients than in controls (56.0 ± 22.3 vs. 70.6 ± 14.6%; p = 0.031). This may be due to a lower physical activity level and a greater difficulty to localize the EDB muscle properly in affected legs. The decrease in CMAP amplitude was not related to skin temperature. Contrary to the prevailing opinion, BoNT-A has a normal, although perhaps slightly lower efficacy in CRPS patients with dystonia.

Pitfalls in using electrophysiological studies to diagnose neuromuscular disorders

Electrodiagnostic testing is used widely for the full characterization of neuromuscular disorders and for providing unique information on the processes underlying the pathology of peripheral nerves and muscles. However, such testing should be considered as an extension of anamnesis and physical examination, not as pathognomonic of a specific disease entity. There are many pitfalls that could lead to erroneous interpretation of electrophysiological study results when the studies are not performed properly or if they are performed in the presence of anatomical aberrations. The diagnostic reliability of electrodiagnostic studies can be improved and the associated pitfalls overcome if the physician is familiar with all of those possible pitfalls. In this article we discuss the most common and important pitfalls associated with electrodiagnostic medicine.

Motor unit number index (MUNIX) versus motor unit number estimation (MUNE): a direct comparison in a longitudinal study of ALS patients

OBJECTIVE

To evaluate how the motor unit number index (MUNIX) is related to high-density motor unit number estimation (HD-MUNE) in healthy controls and patients with amyotrophic lateral sclerosis (ALS).

METHODS

Both MUNIX and HD-MUNE were performed on the thenar muscles in 18 ALS patients and 24 healthy controls. Patients were measured at baseline, within 2 weeks, and after 4 and 8 months. Clinical evaluation included Medical Research Council (MRC) scale and the ALS functional rating scale (ALSFRS).

RESULTS

There was a significant positive correlation between MUNE and MUNIX values in ALS patients (r=0.49 at baseline; r=0.56 at 4 months; r=0.56 at 8 months, all p<0.05), but not in healthy controls. After 8 months, both MUNE and MUNIX values of the ALS patients decreased significantly more compared to MRC scale, ALS functional rating scale (ALSFRS) and compound muscle action potential (CMAP) (p<0.05). There was no significant difference in relative decline of MUNIX and HD-MUNE values.

CONCLUSIONS

In ALS patients, MUNIX and HD-MUNE are significantly correlated. MUNIX has an almost equivalent potential in detecting motor neuron loss compared to HD-MUNE.

SIGNIFICANCE

MUNIX could serve as a reliable and sensitive marker for monitoring disease progression in ALS.

Influence of motor unit properties on the size of the simulated evoked surface EMG potential

The purpose of the study was to quantify the influence of selected motor unit properties on the simulated amplitude and area of evoked muscle potentials detected at the skin surface. The study was restricted to a motor unit population simulating a hand muscle whose potentials were recorded on the skin over the muscle. Peak-to-peak amplitude and area of the evoked potential were calculated from the summed motor unit potentials and compared across conditions that simulated variation in different motor unit properties. The simulations involved varying the number of activated motor units, muscle fiber conduction velocities, axonal conduction velocities, neuronal activation times, the shape of the intracellular action potential, and recording configurations commonly used over hand muscles. The results obtained for the default condition simulated in this study indicated that ~7% of the motor unit potentials were responsible for 50% of the size of the evoked potential. Variation in the amplitude and area of the evoked muscle potential was directly related to the number of active motor units only when the stimulus activated motor units randomly, and not when activation was based on a parameter such as motor unit size. Independent adjustments in motor unit properties had variable effects on the size of the evoked muscle potential, including when the stimulus activated only a subpopulation of motor units. These results provide reference information that can be used to assist in the interpretation of experimentally observed changes in the size of evoked muscle potentials.

‘Neurographic’ palmaris brevis sign in type II° ulnar neuropathy at wrist

OBJECTIVE

To investigate the source of an unusual and previously unreported volume conducted potential on motor nerve conduction studies. In a case of subacute ulnar neuropathy at wrist (UNW) selectively involving the deep motor branch, we recorded from the hypothenar eminence a large positive wave (2.5 ms-2 mV) preceding the negative takeoff of the delayed distal ulnar motor response.

METHODS

We performed multiple channels motor and sensory ulnar nerve (UN) conduction studies; these included selective electrical stimulation and anaesthetic block of UN branches and also selective recording of motor responses by single fibre needles; data were confirmed by an intraoperative neurophysiological study and correlated with MRI and surgical findings.

RESULTS

Detailed neurophysiological investigation demonstrated the generation of this waveform from the palmaris brevis (PB) muscle. MRI and surgical exploration documented a hypertrophy of this muscle.

CONCLUSIONS

In type II degrees UNW, depolarization of a spared palmaris brevis muscle may be recorded as a positive wave preceding the delayed abductor digiti minimi motor response.

SIGNIFICANCE

We underline the peculiar localizing value of this volume conducted 'meaningful artefact' in that particular setting. It actually represented an early neurographic analogue of what is known as the clinical 'Palmaris Brevis Sign' in long standing type II degrees UNW.

Repetitive nerve stimulation: effects of recording site and the nature of 'pseudofacilitation'

OBJECTIVES

To describe changes in the waveform of the compound muscle action potential (CMAP) during repetitive nerve stimulation for various recording sites.

METHODS

Responses to trains of 10 stimuli given at 0.1, 1, 3, 5, 10 and 30 Hz to the ulnar nerve were recorded simultaneously from 8 hand sites in 15 healthy subjects. Percentile changes of amplitude, duration and area of both negative and positive phases were analyzed.

RESULTS

Duration consistently decreased during the trains. At 30 Hz, the mean amplitude of the negative phase increased on 5 sites but decreased on 3. Area consistently decreased, but least for hypothenar sites. Repeated stimulation causes an alteration in the waveform of the CMAP that consists of 4 elements: (1) shorter duration; (2) changed amplitude of the negative phase (up or down); (3) merging of bifid peaks; (4) changes were more pronounced for positive than negative phases.

CONCLUSIONS

As the term 'pseudofacilitation' implies an increase in amplitude, it is often not appropriate. Increased muscle fiber conduction velocity can explain most of the waveform alterations. Movement and shortening of muscles may play additional roles. Consequences for diagnostic yield await a comparison with disease groups.

'Rules of conduct': some practical guidelines for testing motor-nerve conduction

The testing of nerve conduction using electromyography (EMG) is a frequently used diagnostic method for the identification of various neuropathies. The present article illustrates a variety of conditions on the basis of clinical data, and suggests how one can obtain the best results by observing a few simple rules.

Effect of electrode dimensions on motor unit potentials

Different effects of longitudinal and transversal electrode dimensions on nerve or muscle single fibre action potentials detected monopolarly, were reported in the literature. The results were contradictory. We studied motor unit potentials (MUPs) detected at a large distance (typical of surface recording) on the basis of a mathematical model without source simplification. The MUPs were calculated as a single convolution of the first temporal derivative of a realistic intracellular action potential and MU impulse response. The spatial averaging of the MUPs by rectangular plate electrodes was performed through analytical integration of the MU impulse response over the electrode area. The effects of longitudinal dimension of the electrode were stronger than those of a transversal one. The effects were distance dependent. The longitudinal dimension of the electrode influenced the main phases (that reflected the excitation origin and propagation) more than the terminal phases (that reflected the excitation extinction at the muscle fibers' ends). This was due to differences in the character of the potential fields (quadrupole or dipole) during generation of individual MUP phases. It was shown that the relative weight of the individual MUP phases could be stressed or suppressed by a proper choice of electrode dimensions, position and orientation.

CMAP amplitude cartography of muscles innervated by the median, ulnar, peroneal, and tibial nerves

The spatial and temporal distribution of compound muscle action potential (CMAP) amplitudes was mapped using 1 x 1-cm grids over thenar, hypothenar, dorsal foot, and foot sole muscles (seven maps each). The high-amplitude zone (HAZ, area where amplitudes were over 80% of the maximum amplitude) denoted susceptibility to changes in recording site. Thenar maps had one peak (spatially and temporally) with a HAZ of 3.5 +/- 2.3 cm2. Hypothenar maps had two peaks (spatially and temporally) with a HAZ of 7.7 +/- 3.6 cm2. Dorsal foot maps had one temporal peak, which could be split up spatially; the HAZ was smallest, at 1.7 +/- 1.7 cm2. Foot sole muscles had one peak (spatially and temporally), with the largest HAZ at 18.4 +/- 6.1 cm2. Wave-form differences were ascribed to differences in muscle anatomy, architecture, and variability. These explain differences in amplitude reproducibility between nerves and the differing effect that increasing electrode size has on reproducibility.

Influence of electrode site and size on variability of magnetic evoked potentials

Successive magnetic evoked potentials (MEPs) concern varying motor neurons. We investigated whether this MEP-specific source of variability depends on electrode site and size. Amplitude variability (standard deviation) was largest over the center of the hypothenar muscles. Latencies were longer at distal and proximal sites than at the center site. Large electrodes (10 cm2) did not decrease this source of amplitude variability compared with EEG electrodes, in contrast to other sources of variability.

Acquired slow-channel syndrome: a form of myasthenia gravis with prolonged open time of the acetylcholine receptor channel

A 32-year-old female presented with a 2-year history of fluctuating generalized weakness including extraocular, bulbar, and limb muscles, suggesting myasthenia gravis, but with poor response to pyridostigmine and unusual electromyographic findings. After rest, power increased on repeated maximal contractions, followed by progressive weakness. There were decremental responses at low-frequency stimulation, but incremental responses at high frequencies, and single stimuli evoked repetitive compound muscle action potentials. Plasmapheresis was ineffective. In a conventional assay, antibodies against acetylcholine receptors (AChRs) were borderline. However, in an assay using cells expressing mainly adult-type human AChRs, the patient's serum was positive. Thymectomy revealed a hyperplastic thymus. An intercostal muscle specimen revealed small miniature end-plate potentials, 0.22+/-0.02 mV instead of 0.56+/-0.05 mV in controls. The number of 125I-alpha-bungarotoxin binding sites was normal. The decay time constant of end-plate potentials was increased from 5.3+/-0.6 msec in controls to 23+/-3.6 msec in the patient. Ultrastructurally, there was no destruction of the end plate. Transfer of the patient's plasma to mice in vivo produced similar physiological changes in their diaphragms. We conclude that the patient has an immune-mediated disorder, in which an antibody specific to the adult form of the AChRs alters the channel properties, reducing total current and slowing the closure. We propose the name "acquired slow-channel syndrome" for this variant of myasthenia gravis.

[Peripheral neuropathy investigation in the early stage of bone marrow transplantation]

Forty-three patients with several hematological diseases in the early stage of bone marrow transplantation were prospectively studied. All patients underwent a complete neurological examination, vibratory sense perception test and nerve conduction study both before and after the bone marrow transplantation. The following nerves were studied: median, peroneal, posterior tibial (motor), median and sural (sensitive). Most patients were in busulphan, cyclophosphamide, cyclosporine A, methotrexate and corticosteroids. Although the results showed statistical differences in the nerve conduction study, all of these differences were not sufficient to induce or diagnose peripheral nerve injury. Therefore, it was concluded that drugs used in the bone marrow transplantation do not induce peripheral neuropathy.

Reproducibility of normal facial motor nerve conduction studies and their relevance in the electrophysiological assessment of peripheral facial paralysis

To determine the intra-examiner intertrial reproducibility of normal facial motor nerve conduction studies (FNCS) and their relevance in electrophysiological assessments of peripheral facial paralysis, 52 patients with acute unilateral Bell's palsy were examined on two separate occasions 1 months apart. Three electroneurographic methods were assessed. On the unaffected side of the face, FNCS are reliable when performed by a single examiner over time. Nevertheless, compound muscle action potential (CMAP) baseline-to-peak and peak-to-peak amplitude showed a rather high intertrial variability. Reproducibility of the assessed surface electrode recording procedures was similar. Regarding the affected side, in patients with mild axonotmesis of the facial nerve variations of electroneurographic parameters 1 months apart fell within the range of normal intertrial variability. In patients with severe or moderate axonotmesis, the distal latency and the M wave amplitude variations showed significant intertrial variations. Reproducibility of FNCS appears to be similar to that found in limb motor nerves. Normal variability curtails the sensitivity of FNCS in detecting mild facial nerve axonotmesis, although this technique remains useful in severe cases.

Anatomical and electrophysiological determinants of the human thenar compound muscle action potential

Clinical interpretation of the compound muscle action potential (CMAP) requires a precise understanding of its underlying mechanisms. We recorded normal thenar CMAP5 and motor unit action potentials using different electrode configurations and different thumb positions. Computer simulations show that the CMAP has four parts: rising edge, negative phase, positive phase, and tail which correspond to four distinct stages of electrical activity in the muscle: initiation at the end-plate, propagation, termination at the muscle/tendon junctions, and slow repolarization. The shapes of volume-conducted signals recorded beyond the muscle are also explained by these four stages. Changes in CMAP shape associated with thumb abduction are due to changes in termination times resulting from changes in muscle-fiber lengths. These findings demonstrate that the negative and positive phases of the CMAP are due to different mechanisms, and that anatomical factors, particularly muscle-fiber lengths, play an important role in determining CMAP shape.

Repetitive CMAPs: mechanisms of neural and synaptic genesis

Repetitive compound muscle action potentials (R-CMAPs) occur when a single nerve shock excites muscle fibers repeatedly. "Double discharges" are due to intramuscular nerve reexcitation. "Synaptic" R-CMAPs, due to excess acetylcholine in the neuromuscular synapse, can occur in congenital myasthenia, the slow-channel syndrome, and acetylcholinesterase inhibition. Secondary nerve excitation can reexcite muscle fibers. Synaptic R-CMAPs in a patient consisted of two discharges. The second diminished during repetitive stimulation and began 3.5-4.0 ms after the first, which is slightly longer than the synapse-muscle refractory period. Neural R-CMAPs, due to ectopic nerve activity, occur in neuromyotonia (NMT). R-CMAPs in a patient consisted of about 20 discharges at 200-300 Hz. Studies in healthy subjects showed that such trains represent added single CMAPs. Impulse frequency in the patient lied close to the threshold of refractoriness. Refractoriness of the synapse-muscle cell assembly determines the characteristics of R-CMAPs regardless of the primary cause.

Large electrodes improve nerve conduction repeatability in controls as well as in patients with diabetic neuropathy

Recording site is an important cause of variability of compound muscle action potential (CMAP) and conduction parameters, which can be reduced by using large electrodes. Repeatability of CMAP and conduction parameters of conventional and large electrodes was compared in 16 controls and 17 diabetic neuropathic patients, using defined recording sites linked to anatomical landmarks. Right-sided median, peroneal, and tibial nerves were investigated twice with a 1-2 week interval by the same examiner. Compared to previous studies, conventional electrodes on strictly defined recording sites resulted in better repeatability: intraindividual coefficients of variation (CV) varied between 4% and 14.4% for all parameters. CV of conduction parameters, not published previously, was smaller than CV of CMAP parameters. The use of large electrodes improved repeatability further: large electrodes resulted in substantially smaller CV for duration, amplitude, area, and changes of amplitude and area over a length of nerve, which were reduced by 10%, 31%, 29%, 27%, and 16%, respectively. Patients had higher CV than controls; large electrodes reduced patient CV more than control CV, resulting in less contrast between groups. Strictly defined recording sites and large electrodes improve repeatability of motor conduction studies to relevant degrees: all CMAP and conduction parameters are suitable for longitudinal studies of neuropathic patients.