In the evaluation of muscle performance and motor control, we are concerned with the integrity of both central and peripheral mechanisms. The assessment of electrophysiological properties of nerve and muscle provides essential information to understand and assist in the diagnosis of neuromuscular disease or trauma, identify the location of a lesion within the PNS, and establish reasonable prognosis or rate of healing or decay. In order to appreciate deficiencies in peripheral neuromuscular function, clinicians must have a basic understanding of the anatomy and physiology of the nerve and muscle cells. Requirements for peripheral motor response include the successful ability to create an action potential at the lower motor neuron, transmit that signal down the motor neuron and across the neuromuscular junction and result in subsequent muscle contraction. There is a potential for failure in each of these “phases” of neuromuscular transmission due to pathology. Such disorders can result in weakness or lack of motor coordination in movement, resulting in disruption to feedback and motor control mechanisms. These conditions may be related to disorders or disease processes affecting the peripheral nerve (sensory and motor), muscle, or the neuromuscular junction.98,99
Electrodiagnosis (EDX) of muscle and nerve is a specialization applied to evaluate the scope of a neuromuscular disorder through assessment of muscle and nerve activity using EMG and NCS. EMG assesses muscle function at rest and during activity, and NCS determines the speed and strength with which a peripheral motor or sensory nerve conducts an impulse. Together, data from EMG and NCS tests assist with establishing goals and expected outcomes for patients with musculoskeletal and neuromuscular disorders. Practitioners providing EDX should have thorough knowledge of anatomy and physiology of muscle and nerve, the biophysical understanding of the instruments used to collect signals, and an understanding of the pathophysiology of nerve and muscle disorders. EDX is only part of a complete patient examination, which will include a thorough understanding of the patient’s history and clinical findings. For example, the therapist would also examine muscle strength, pain, reflexes, fatigue, sensory function, and the presence of atrophy, as well as functional abilities and special tests related to the suspected condition. The findings from the clinical examination will suggest which muscles and/or nerves will be tested with EMG and NCS. EDX findings are not diagnostic in isolation and must be considered in relation to other clinical findings and findings from other physical therapy, medical, and physiological tests and measurements.99
Neuromuscular Transmission, Injury, and Repair
Nerve cells (neurons) communicate with one another and to other organs, such as muscle and sensory organs, in order to transmit signals that allow motor and sensory function. Neurons are small in diameter, usually much smaller than a hair, and can be as long as the distance from your back to the tip of your toe! A motor unit is composed of one anterior horn cell, one axon, its neuromuscular junction, and all the muscle fibers innervated by that axon (Fig. 5.1). The cell body or nucleus of the neuron is the location of many protein particles and organelles that provide critical neuron function to keep the cell alive and allow repair. Without communication with the cell body, a peripheral nerve will die. This process is called Wallerian degeneration. Dendrites are proximal terminals that receive input from other cells. The axons project away from the cell bodies and carry electrical signals toward targeted organs such as muscles.
Nerves are comprised of groups of neurons that can provide sensory, motor, or autonomic signals to and from other nerves and transmitting organs. Peripheral nerves (Fig 5.2) are typically surrounded by myelin, which is made of water, fat, and protein and provides electrical insulation to the neurons; this allows for neurons to conduct impulses at increased speed. The main cell of myelin in the peripheral nervous system is called the Schwann cell, and spaces between the cells are called the nodes of Ranvier. Nerves are like coated cables that are comprised of bundles of neurons. Each bundle is called a fascicle, and the nerve trunk is a collection of fascicles. Connective tissue structures assist in the organization and protection of neurons and are known as the endoneurium, perineurium, and epineurium. The endoneurium surrounds individual axons and provides a conduit for nerves to grow and travel. Perineurium envelopes the fascicles and epineurium surrounds the entire nerve trunk.100
The Neuromuscular Junction
The neuromuscular junction (NMJ) (Fig. 5.3) is the location of electrical transmission from the nerve to the muscle cell and is comprised of three distinct parts. The synaptic terminal is the end bulb of the lower motor neuron. Within the synaptic terminal are large numbers of synaptic vesicles that contain acetylcholine (ACH), which is a substance necessary for electrical neurotransmission. The motor end plate is the part of the muscle cell in closest proximity to the synaptic terminal, where there is a high concentration of post-synaptic terminals. These terminals are arranged in folds of muscle membrane and increase the area to which ACH can bind. The space between the synaptic terminal and motor end plate is known as the synaptic cleft.100
The neuromuscular junction and muscle cell.
Striated skeletal muscle cell is comprised of bundles of elongated myofibrils (actin and myosin) surrounded by a cellular membrane known as the sarcolemma. The sarcolemma is enveloped in capillaries, t-tubules, nuclei, and sarcoplasmic reticulum, which collectively comprise the muscle fiber. Muscle fibers are covered in a connective tissue matrix known as endomysium and are bundled similarly to peripheral nerve in a cable-like fashion of fascicles. The connective tissue surrounding individual fascicles is known as perimysium. The epimysium surrounds the total bundle of many fascicles. These connective tissue structures continue beyond the contractile structures of the muscle cell as the tendons, which then attach onto the bone100 (see Fig 5.3).
When descending/ascending electrical input reaches an axon, there is a disruption in the resting membrane potential, which is a balance of sodium and potassium concentrations within and outside the cell. Resting membrane potential is variable, from –90 to –70 mV. When descending electrical input reaches the lower motor neuron, voltage-gated sodium channels open and there is an exchange where sodium can rush into the cell in an attempt to balance the concentrations of these ions. Smaller positively charged potassium ions are subsequently forced out the cell, and the resting membrane polarity becomes more positive. If the electrical potential reaches a threshold for depolarization (approximately 20 mV), an action potential is created and an impulse travels down the axon. The speed and success of the action potential conduction is dependent on the thickness of the axon and the presence of myelin. Action potentials travel down the myelinated axons through saltatory conduction, in which the action potential jumps from one node of Ranvier to the next, where there are high concentrations of sodium/potassium channels.100 When myelin or axon thickness is disrupted in certain pathologies, the ability of the axon to transmit at normal speeds and strengths can be affected and can lead to abnormal motor or sensory function. When nerve injury affects myelin, it is considered demyelination, and when there is Wallerian degeneration of the axon, it is considered axonal loss. Collectively, injury or disease to the nerve itself is called neuropathy and can occur in individual nerves, isolated areas, or more widespread throughout the peripheral nerves.99
When the nerve action potential reaches the terminal axon, calcium channels open and acetylcholine (ACH) is released into the NMJ. ACH binds to ACH receptors on the motor end plate, and an action potential is generated; this causes the release of calcium ions and finally triggers contraction of the muscle cell. Certain disease processes affecting the release of calcium at the presynaptic terminal or binding of ACH at the post-synaptic terminal can diminish neuromuscular transmission and subsequently impair muscle function. A condition affecting neuromuscular transmission would be considered a neuromuscular junction disorder.100
During muscle contraction, proteins and structural cells slide over one another to create a shortening of the muscle. Calcium released from the “sarcoplasmic reticulum” at the motor end plate opens binding sites on actin filaments, which bind to myosin. Myosin heads undergo a pivoting action to shorten the cell and then detaches when adenosine triphosphate (ATP) binds to the myosin.100 The muscle cell is secured to the extracellular matrix through a series of protein molecules, including dystrophin, which allows for mechanical stability and provides a reasonable anchor for contraction. In certain disease or inflammatory processes, such as the muscular dystrophies or myositis, the muscle cell loses structural integrity and leads to poor muscle performance. This destructive muscle condition leads to variability of muscle fiber size and infiltration of fibrous connective tissue. This is considered a myopathy.
Nerve injury is generally characterized by demyelination or axon loss and can be partial or complete. Demyelination can result from inflammation, immune disorders, genetic predisposition, or exposure to toxic chemicals and/or mechanical forces such as compression or shearing. Disturbance of the myelin can reduce myelin thickness or invagination that leads to a widening of the nodes of Ranvier with sodium channel redistribution. This can occur focally, segmentally, or uniformly along a nerve. When this occurs, saltatory conduction is altered and results in slowing, altered shapes of nerve potentials and/or blocking of nerve transmission altogether (conduction block).99
When an axon loses continuity with its cell body, Wallerian degeneration occurs. Also known as axon loss, this process begins immediately with one or two nodes of Ranvier dying back and with progressive degeneration distally along the neuron. This degenerative process typically requires 5 to 7 days to complete in motor neurons and 7 to 10 days in sensory neurons. As long as the cell body has continuity with its terminal nerve branch, the branch will remain alive. For example, if an upper motor neuron is damaged due to stroke, the lower motor neuron (whose cell body lies within the anterior horn of the spinal cord) remains intact, as does the remainder of the lower motor neuron. Even though the patient cannot fire the nerve because of the central damage in the brain, the lower motor neuron remains alive and can carry information normally. Similarly, imagine a patient with cervical radiculopathy who has weakness and sensory changes in a specific nerve root distribution. There may be weakness and Wallerian degeneration involving the lower motor neuron but not in the sensory neurons because the primary sensory neuron cell bodies reside in the dorsal root ganglia, which are distal to the typical site of nerve root compression in cervical radiculopathy. Injury to nerve proximal to the dorsal root ganglia is often called a preganglionic lesion. This is a very important concept in understanding EDX, evaluating data, and formulating a reasonable diagnostic impression.99
In an attempt to organize nerve structure and progression of injury, an early publication by Seddon101 proposed classifications of nerve injury, which was later expanded by Sunderland102 to highlight involvement of the connective tissue support system of the axon. Seddon’s classification of nerve injury by neuropraxia, axonotmesis, and neurotmesis and Sunderland’s first- to fifth-degree injury classification have become common methods of describing nerve damage (Fig 5.4). It is important to recognize that neuropraxia and first-degree injury involve demyelination and more specifically conduction block. Axonotmesis/neurotmesis and second- to fifth-degree injury involve axon loss with gradual loss of supportive connective tissue elements in the nerve. Ability of a nerve to recover is largely dependent on the preservation of a regenerating neuron’s connective tissue support structure, specifically preservation of the endoneurium. In axonotmesis and second-degree injuries, the endoneurium remains intact and serves as a conduit for the regenerating axon, which improves the likelihood of nerve recovery. In contrast, with neurotmesis or third- to fifth-degree injuries, the endoneurium, perineurium, or epineurium is lost, and no conduit exists for neuron regeneration.
Seddon and Sunderland classification of nerve injury.
Nerve healing occurs through either remyelination or reinnervation by way of axonal sprouting or axonal regeneration (Fig 5.5). In remyelination and with optimal conditions of nutrition, removal or control of disease, and/or cessation of compression, Schwann cells produce new myelin, and action potential speeds and strengths will improve and can even return to normal. Reinnervation due to collateral axon sprouting requires the presence of uninjured axons within the injured nerve bundle. Similar to surviving branches on a tree that has been pruned, uninjured neurons will sprout new branches to reinnervate the target organs of the injured neurons. Earliest collateral sprouts require 6 to 8 weeks to be documented. It is generally agreed that a nerve will require about 20% to 25% of uninjured axons to achieve recovery without residual functional weakness. This occurs from collateral axon sprouting, as each terminal neuron branch can grow up to five collateral sprouts. This means that a peripheral nerve can lose quite a large percentage of axons and still recover well. Reinnervation owing to axonal regeneration is a slower process, and under optimal conditions, the injured nerve bulb can regenerate along an intact endoneurium at a rate of about 1 mm per day, or 1 inch (25.4 mm) per month. Age, other disease processes, and distance to the target organ significantly influence a neuron’s successful regeneration. EDX can assist in identifying nerve injury through localization, determining the extent of injury and involvement of myelin versus axon loss, and assessing reinnervation by way of collateral axon sprouting versus axon regeneration.99
Remyelination and reinnervation.
Concepts of Electrodiagnosis
Both EMG and NCS recordings are made using biomedical equipment that can record and store evoked potential data elicited from the stimulation of nerves and electrical activity from within the muscle (Fig 5.6A). Electrodes are placed on, in, or near muscle or nerve to record bioelectric charges. Recordings are transported to an amplifier for signal processing. Following amplification, the equipment filters ambient noise that would otherwise distort the intended bioelectric charges, and the digitized analog signal is displayed on an oscilloscope for analysis. Audio amplification with speakers also permits acoustic monitoring of these nerve and muscle potentials.
EMG machine, EMG, and NCS setup. (A) EMG machine. (Courtesy of Cadwell Laboratories, Kennewick, WA.) (B) Sensory NCS of the superficial radial nerve. (C) Needle EMG of the anterior tibialis.
A stimulator is used in NCS and is electronically synchronized with the oscilloscope to fire the nerve. Intensity of the stimulator is dependent upon the voltage, amperage, and pulse duration of the stimulus.
There are three types of electrodes used for recordings. A ground electrode serves as a zero voltage reference and reduces extraneous electrical noise and interference. The recording electrode (pickup electrode) records the intended potential and extraneous signals, and the reference electrode records extraneous signals and different aspects of the intended potential.
In NCS, small surface electrodes are usually used to record the evoked potential from the test muscle or directly from the nerve under study. NCS recordings can also be made using intramuscular or subcutaneous needles. The recording electrode is placed over the belly of the test muscle and a reference electrode is taped over the tendon of the muscle (see Fig 5.6B).
For EMG, the most common types of needle electrodes are bipolar and monopolar. A bipolar electrode is a hypodermic needle, through which a single wire made of platinum or silver is threaded. The cannula shaft and wire are insulated from each other, and only their tips are exposed. The wire and the needle cannula act as recording and reference electrodes, and the difference in potential between them is recorded in volts. A monopolar needle electrode is composed of a single fine needle, insulated except at the tip. A second surface electrode placed on the skin near the site of insertion serves as the reference electrode. These electrodes are generally less painful than bipolar electrodes because they are smaller in diameter and do not have a cutting edge (see Fig. 5.6C).
Bioelectric potentials travel through body fluids in all directions, not just in the direction of the recording electrode. Fibrous tissue, fat, and blood vessels act as insulators in this process. Therefore, the actual pattern of the flow of electrical activity is not predictable. The signals that do reach the electrode are transmitted to an amplifier. The activity produced by all the individual fibers of nerve or muscle at any one time is summated, reaching the electrode almost simultaneously. Electrodes only record potentials they pick up, without differentiating their origin. Therefore, if two motor units contract at the same time, from the same or adjacent muscles, the activity from fibers of both units will be summated and recorded as one large potential.
The size and shape of the evoked potentials can be affected by several variables. The proximity of the electrodes to the nerve and muscle that are firing will affect the amplitude and duration of the recorded potential. Targets that are farther away will contribute less to the recorded potential. The number and size of the muscle fibers or neurons will influence the evoked potential size. Finally, the distance between the fibers will affect the output, because if the fibers are very spread out, less of their total activity is likely to reach the electrodes.
In addition to these variables, many excess signals, or artifacts, can be recorded and processed simultaneously with the EMG signal. An artifact is any unwanted electrical activity that arises outside of the tissues being examined. These artifacts can be of sufficient voltage to distort the output signal markedly, such as those coming from other electrical equipment or fluorescent lights. Electromyographers will usually observe the output signal on an oscilloscope or computer screen to monitor artifacts, then employ troubleshooting techniques to minimize unwanted noise.
EMG is the recording of the electrical activity of muscle while at rest and during motor unit activation. Data are visualized on an oscilloscope with audio feedback that allows the examiner to distinguish between normal and abnormal activity.
Initially, the patient is asked to relax the muscle to be examined during insertion of the needle electrode. Insertion into a contracting muscle is uncomfortable but tolerable. At this time, the electromyographer will observe a spontaneous burst of potentials, called insertional activity, which is possibly caused by the needle breaking through muscle fiber membranes. This normally lasts less than 300 milliseconds (msec).103 Insertional activity can be described as normal, reduced, absent or increased (Fig 5.7). Increased insertional activity is representative of muscle membrane instability and is frequently seen in nerve injury or muscle degeneration. Reduced insertional activities generally represent chronic neuropathy or myopathy in which there has been significant atrophy of muscle and/or infiltration of fat and connective tissue within the muscle. Decreased insertional activity implies non-viable muscle tissue, and the examiner may feel increased resistance when advancing the needle. The needle electrode will be moved to different areas and depths of each muscle. This is necessary because of the small area from which a needle electrode will pick up electrical activity and because the effects of pathology may vary within a single muscle. Up to 25 different points within a single muscle may be examined by moving and redirecting the needle electrode.
Insertional activity. (A) Normal insertional activity. (B) Increased insertional activity.
Following cessation of insertional activity, a normal relaxed muscle will exhibit electrical silence, which is the absence of electrical potentials. Observation of silence in the relaxed state is an important part of the EMG examination. Potentials arising spontaneously during this period can be seen in normal muscle if the needle electrode is close to a neuromuscular junction. Miniature end plate potentials are thought to represent small releases of ACH at a NMJ that are not sufficient to propagate a muscle fiber action potential, and end plate spikes/end plate potentials represent sufficient release of ACH at the NMJ to cause depolarization of individual muscle fibers. End plate potentials are irregular in their firing rates and are often uncomfortable to the patient when encountered.99 They are not indicative of pathology and are seen frequently in normal muscle. When encountered, it is important to confirm what they are and then move the needle away from the painful area (Fig. 5.8A,B).
Resting activity. (A) Normal resting activity. (B) End plate potentials. (C) End plate spikes. (D) Fibrillation potentials. (E) Positive sharp waves.
There are several characteristic potentials observed during the relaxed state that are considered abnormal. Fibrillation potentials are believed to arise from spontaneous depolarization of a single muscle fiber. They are not visible through the skin. Fibrillation potentials are biphasic spikes and are indicative of muscle membrane instability. They are seen in neuropathic disorders, such as peripheral nerve lesions, anterior horn cell disease, radiculopathies, and polyneuropathies with axonal degeneration (see Fig 5.8C). They are also found in muscle degenerative conditions such as muscular dystrophy, dermatomyositis, polymyositis, and less frequently in myasthenia gravis. Their sound is a high-pitched click, which has been likened to the sound of rain falling on a roof or rhythmic wrinkling tissue paper. Positive sharp waves have been observed in denervated muscle at rest, often accompanied by fibrillation potentials, and they, too, are reported in primary muscle disease, especially muscular dystrophy and polymyositis. The waves are typically biphasic, with a sharp initial positive deflection (below baseline) followed by a slow negative phase. The negative phase is of much lower amplitude than the positive phase and of much longer duration, sometimes up to 100 msec. The peak-to-peak amplitude may be variable, with voltages from 50 microvolts (uV) to 2 millivolts (mV). The sound has been described as a dull thud (see Fig. 5.8D). While morphology (shape) is different in positive sharp waves and fibrillations, it is thought that they represent the same phenomena of spontaneous discharge of a muscle fiber but look differently due to the orientation of the needle electrode with respect to the discharging muscle fiber.99 In fact, it is common to see a fibrillation or positive sharp wave morph from one to the other during the examination. Fibrillations and positive sharp waves can be observed in EMG within a week, particularly in proximal muscles, but require up to 3 weeks following axon loss injury to become more widespread.
Complex repetitive discharges may be seen with lesions of the anterior horn cell and peripheral nerves and with myopathies. The discharge is characterized by an extended train of potentials with the same or nearly the same waveform. The feature that distinguishes these discharges from other spontaneous potentials is their regular and repetitive waveform. The frequency usually ranges from 5 to 100 impulses per second. Myotonic discharges are repetitive potentials that increase and decrease in amplitude in a waxing and waning fashion. They are found in myotonic disorders such as myotonic dystrophy, as well as other myopathies. The sound is highly characteristic and sounds like a dive-bomber. High-frequency discharges are probably triggered by movement of the needle electrode within unstable muscle fibers or by volitional activity.99
Fasciculations are spontaneous motor unit action potentials (MUAPs) seen with irritation or degeneration of the anterior horn cell, chronic peripheral nerve lesions, nerve root compression, and muscle spasms or cramps. They are believed to represent the involuntary asynchronous firing of motor units. Their sound has been described as a low-pitched thump. Fasciculations are often visible through the skin and can be seen as a small twitch. They are not by themselves a definitive abnormal finding, as they are often seen in normal individuals, particularly in muscles of the calves, eyes, hands, and feet.104 Myokymia also represents spontaneous bursts of motor unit firing that is more uniform and can fire fast at rates from 2 Hz to 60 Hz.99 Like fasciculations, myokymic potentials are not by themselves indicative of pathology and may be seen in normal or fatigued muscle.
Motor Unit Morphology and Recruitment
After observing the muscle at rest, the patient is asked to contract the muscle minimally. When a muscle is contracted to produce force, a single axon conducts an impulse to all its muscle fibers, causing them to depolarize at relatively the same time. This depolarization produces electrical activity that is manifested as a MUAP and can be recorded and displayed graphically. A MUAP is actually the summation of electrical potentials from all the fibers of that unit close enough to the electrodes to be recorded. The amplitude (voltage) is affected by the number of fibers involved or by the motor unit territory. The duration and shape are functions of the distance of the fibers from the recording electrodes, the more distant fibers contributing to terminal phases of the potential. Because of these variables, each motor unit will have a distinctive shape and vary in amplitude and number of phases, where a phase represents a section of a potential crossing above or below the baseline. MUAPs are examined with respect to morphology (amplitude, duration, shape) and frequency of firing. These parameters are the essential characteristics that distinguish normal from abnormal potentials. Initially, a type I MUAP is recruited at approximately 5 Hz. As resistance is increased, that first MUAP increases frequency until it reaches approximately 10 Hz, when another type I MUAP is recruited. The pattern of increasing frequency and addition of more MUAPs continues until type II MUAPs are recruited. Typically, only one to three MUAPs can visibly be assessed qualitatively at any given time for morphology and frequency. Gradually increasing the force of contraction will allow the electromyographer to observe the pattern of recruitment in the muscle but not qualitatively assess morphology. With greater effort, the increased numbers of potentials are summated and can no longer be recognized, but an overall interference pattern can be estimated. The characteristics of the MUAP will change when there is damage to either the nerve or muscle. EMG equipment has improved over the years and most models have features to provide quantitative MUAP analysis that can improve the objective assessment of MUAP characteristics. In normal muscle, the peak-to-peak amplitude of typical MUAP may range from 150 microvolts uV to 5 millivolts mV. The duration of the potential is a measure of time from onset to cessation of the electrical potential, typically from 2 to 14 msec105 (Fig 5.9A). The typical shape of a MUAP is biphasic or triphasic and generally not displaying greater than four phases.
Motor unit morphology and recruitment. (A) Normal motor unit. (B) Polyphasic motor unit. (C) Larger than normal amplitude motor unit with neuropathic recruitment. (D) Short duration, low amplitude motor units with myopathic recruitment.
Polyphasic potentials, which are MUAPs having five or more phases, are generally considered abnormal when seen in large numbers (see Fig 5.9B). It is normal to observe small numbers of polyphasic potentials; however, when polyphasic potentials represent approximately 20% or more of a muscle’s output, it may be an abnormal finding. However, caution is warranted in identifying pathology based solely on the presence of polyphasic MUAPs, as some authors have reported their appearance in greater than 30% of normal subjects.99 Polyphasic potentials may also be seen during degeneration and after regeneration of a peripheral nerve. Polyphasic potentials with longer than normal durations are considered a sign of motor neuropathy and may be a result of the asynchronous firing of muscle fibers within a MUAP that is undergoing reinnervation owing to collateral sprouting. This phenomenon is probably due to the difference in the length of the terminal branches and maturity of myelin in the sprouting axons extending to each muscle fiber. As some muscle fibers become reinnervated by collateral sprouts, they will generate action potentials along with the other muscle fibers within that motor unit. The result is a normal amplitude MUAP with larger number of phases and longer duration than a normal motor unit.
Following axonal degeneration, collateral sprouts mature. Successfully reinnervated muscle fibers typically demonstrate larger than normal amplitude MUAP, also known as a Giant MUAP (see Fig 5.9C). Generally, MUAPs exceeding 5 mV are considered larger than normal. These potentials may be seen in post-polio syndrome and other neuropathic conditions such as chronic radiculopathy or focal nerve entrapments like carpal tunnel syndrome.106-108
In a complete axon loss nerve injury, reinnervation can occur due to axonal regeneration. Recall that the presence of an endoneurial tube is critical for successful axonal regeneration. As the first regenerating axons reach their intended muscle fibers, the potential will look very similar to a fibrillation potential, and as more terminal branches reach muscle fibers, the MUAP will be small in amplitude and highly polyphasic. Early MUAPs representing axonal regeneration have been termed nascent motor units. As these MUAPs mature, the number of phases will decrease and the morphology and duration will become more normal.
In primary muscle disease (myopathies), polyphasic potentials are seen but are generally of smaller amplitude than normal motor units and are typically of shorter duration. These multiphasic changes occur because of a decrease in the number of active muscle fibers within the individual motor units due to pathology. Although the entire unit will fire during voluntary contraction, fewer fibers are available in each unit to contribute to the total voltage and the duration of the potential. The result is a MUAP of lower than normal amplitude and potentially shorter duration (see Fig. 5.9D).
Firing frequencies of any given MUAP are generally less than 12 Hz to 15 Hz if only one to three MUAPs are isolated on an oscilloscope. In a condition of motor axon loss where there are less available MUAPs for recruitment to produce force, firing frequencies may be seen in excess of 15 Hz and can be accompanied by a less than full interference pattern. The MUAP morphological changes consisting of long duration polyphasia or larger than normal amplitudes with overall reduced and fast firing frequencies is consistent with neuropathic recruitment (see Fig 5.9C). Conversely and as discussed in primary muscle generation, motor unit morphology can show MUAPs with shorter duration and low amplitude polyphasia. The total availability of motor units remains normal, so overall interference patterns will be normal; however, muscle degeneration will result in less viable muscle tissue producing resistance to demanded force during the EMG. The resulting interference pattern will occur sooner than anticipated and is considered early recruitment. The combination of MUAP morphology changes consisting of low amplitude, short-duration polyphasia with early recruitment is consistent with myopathic recruitment (see Fig 5.9D).
Nerve conduction studies (NCS) involve direct stimulation to initiate an impulse in motor or sensory nerves. The conduction time is measured by recording the evoked potential either from the muscle innervated by the motor nerve or from the sensory nerve itself. NCS can be tested on any peripheral nerve that is superficial enough to be stimulated through the skin at two different points. The most commonly tested motor nerves are the ulnar, median, fibular (peroneal), tibial, radial, femoral, and sciatic nerves. Commonly tested sensory nerves include the median, ulnar, radial, sural, and superficial fibular nerves, but this is not at all a comprehensive list of available nerves to examine. Complete guidelines for performing NCS tests are available in comprehensive references.99,103,105,109,110
Motor Nerve Conduction Studies
Because a peripheral nerve trunk houses both sensory and motor fibers, recording potentials directly from a peripheral nerve makes monitoring of purely sensory or motor nerves impossible. Therefore, to isolate the potentials conducted by motor axons of a mixed nerve, the evoked potential is recorded from a distal muscle innervated by the nerve under study. Although the stimulation of the nerve will evoke sensory and motor impulses, only the motor fibers contribute to the contraction of the muscle. For example, to test the ulnar nerve, the test muscle is typically the abductor digiti minimi. Other examples are the following: for the median nerve, the abductor pollicis brevis; for the fibular nerve, the extensor digitorum brevis; and for the tibial nerve, the abductor hallucis. Of course, motor evoked potentials can be recorded from any muscle if its target nerve can be stimulated and a recording electrode can be properly placed on or within the muscle.
For the purposes of illustration, the test procedure for the motor NCV of the median nerve will be described (Fig. 5.10). The technique is basically the same for all nerves, except for the sites of stimulation and placement of the electrodes. For this example, the recording electrode is taped over the abductor pollicis brevis. The stimulating electrode is placed over the median nerve at multiple sites, including the palm, wrist, elbow, and axilla, and a recording is made at each site. At the moment the stimulus is produced, the stimulus artifact is seen at the left of the oscilloscope screen. A trigger mechanism controls this and it will, therefore, always appear in the same spot on the screen, facilitating consistent measurements. This spike is purely mechanical and does not represent any muscle activity. The stimulus intensity starts out low and is slowly increased until the evoked potential is clearly observed. When the stimulating electrode is properly placed over the nerve, all muscles innervated distal to that point will contract and the patient will see and feel the hand jump. The intensity is then increased until the evoked response no longer increases in size. At that time, the intensity is increased further to be sure that the stimulus is supramaximal. Because the intensity must be sufficient to reach the threshold of all motor fibers in the nerve, a supramaximal stimulus is required. It is also essential that the stimulator be properly placed over the nerve trunk so that the stimulus reaches all the motor axons.
Sites of stimulation for median motor nerve conduction.
As in the EMG signal, the potentials seen on the screen represent the electrical activity detected by the recording electrode. The signal will represent the difference in electrical potential between the recording and reference electrodes. When the supramaximal stimulus is applied to the median nerve at the wrist, all the axons in the nerve will depolarize and begin conducting an impulse, transmitting the signal across the motor end plate, initiating depolarization of the muscle fibers. During these events, the recording electrodes do not record a difference in potential because no activity is taking place beneath the electrodes. When the muscle fibers begin to depolarize, the electrical potentials are transmitted to the electrodes, and a deflection is seen on the oscilloscope. This is the evoked potential, which is called the M wave. The M wave is also referred to as the motor action potential (MAP) or compound motor action potential (CMAP). The CMAP represents the summated activity of all motor units in the muscle that responded to stimulation of the nerve trunk. The amplitude of this potential is, therefore, a function of the total voltage produced by the contracting motor units. The initial deflection of the CMAP is the negative portion of the wave, above the baseline. Conduction parameters of interest include latency, nerve conduction velocity, amplitude, and morphology.99,103
Calculation of Motor Nerve Conduction Velocity
The point at which the CMAP leaves the baseline indicates the time elapsed from the initial propagation of the nerve impulse to the depolarization of the muscle fibers beneath the electrodes. This is called the response latency. The latency is measured in milliseconds from the stimulus artifact to the onset of the CMAP. This time alone is not a valid measurement of nerve conduction because it incorporates time related to other events besides pure nerve conduction—namely, transmission across the NMJ and generation of the muscle action potential. Therefore, these extraneous factors must be eliminated from the calculation of the motor NCV, so that the measurement reflects only the speed of conduction within the nerve trunk.
To account for these distal variables, the nerve is stimulated at a second, more proximal point. This will produce a response similar to that seen with distal stimulation. The stimulus artifact will appear in the same spot on the screen, but the CMAP will originate in a different place because the time for the impulses to reach the muscle would, obviously, be longer. Subtraction of the distal latency from the proximal latency will determine the conduction time for the nerve trunk segment between the two points of stimulation. Nerve conduction velocity (NCV) is determined by dividing the distance between the two points of stimulation (measured along the surface) by the difference between the two latencies (velocity = distance/time).
NCV = Conduction distance/(Proximal latency – Distal latency).
NCV is always expressed in meters per second (m/s), although distance is usually measured in centimeters and latencies in milliseconds. These units must be converted during calculation. To compute the motor NCV, the latencies are determined for each stimulation site along the nerve by measuring the time from the stimulus artifact to the initial CMAP deflection. The segment conduction time is calculated by taking the difference between latencies of adjacent stimulation sites. Conduction distance is then determined by measuring the length of the nerve between the two points of stimulation. For example in Figure 5.11:
Median motor nerve conduction.
Interpretation of the motor NCV is made in relation to normal values, which are usually expressed as mean values, standard deviations, and ranges. Many investigators in different laboratories have determined normal values.110 Even so, average values seem to be fairly consistent. The motor NCV for the UE has a fairly wide range, with values reported from 50 to 70 m/s. The average normal value is about 60 m/s. For the LE, the average value is about 50 m/s. Distal latencies and average normal amplitudes of CMAPs are also available in such tables, but these must be viewed with caution, because technique, electrode setup, instrumentation, and patient size can affect these values. Age and temperature can also influence NCS measurements, decreasing after age 35 and with lower temperature.98 The reader is referred to more comprehensive discussions for complete details about techniques for studying various nerves and for tables of normal values.105,109,110
It is important to note that the value calculated as the conduction velocity is actually a reflection of the speed of the fastest axons in the nerve. Although all axons are stimulated at the same point in time, and supposedly fire at the same time, their conduction rates vary with their size. Not all motor units will contract at the same time; some receive their nerve impulse later than others. Therefore, the initial CMAP deflection represents the contraction of the motor unit, or units, with the fastest conduction velocity. The curved shape of the CMAP is reflective of the progressively slower axons reaching their motor units at a later time. Figure 5.12A demonstrates slower than normal wrist to palm conduction velocity and slower wrist to abductor pollicis latency in a patient with carpal tunnel syndrome.
Abnormal motor nerve conduction studies. (A) Mild median slowing focally across the wrist. (B) Temporal dispersion of tibial motor nerve. (C) Partial motor conduction block of ulnar nerve at elbow. (D) Severe median slowing across the wrist with partial axon loss.
The CMAP can also provide useful information about the integrity of the nerve or muscle. The shape and configuration of the CMAP should be examined along the course of the nerve tested and changes duly noted. Temporal dispersion is a phenomenon representing asynchronous recording of muscle fibers contributing to the CMAP due to varying speeds of individual motor neurons. While increasing the length of a segment will accentuate some disparity among neuron speed, demyelinating conditions often cause significant temporal dispersion in CMAPs.99,103 Figure 5.12B demonstrates temporal dispersion in the tibial motor CMAP of a patient with diabetes-related peripheral polyneuropathy. Notice that the distal latency and leg velocity measures fall just outside of normal limits, but the CMAP morphology is longer in duration with smaller amplitude when comparing the knee to ankle CMAPs.
In instances of partial neuropraxia or conduction block, distal CMAP amplitude is normal, but stimulation proximal to the lesion is comparatively reduced without significant temporal dispersion. Figure 5.12C shows slowing with partial conduction block of just over 30% occurring at or just distal to the medial epicondyle in the ulnar motor NCS of a patient with cubital tunnel syndrome. To identify a conduction block, stimulation must occur distally and proximally to the nerve lesion. However, there are scenarios where this is not possible. For example, in assessing a brachial plexus disorder, difficulty may be experienced stimulating the axillary nerve distal to the clavicle. In contrast, stimulating the radial nerve is generally easily accessible both distally and proximally. This is an important concept, particularly when recognizing the improved prognostic implications of a neuropathy demonstrating conduction block as opposed to axon loss.
In motor axon loss, the CMAP distal to the lesion is expected to reflect lower than normal CMAP amplitude. Figure 5.12D depicts the motor NCS in a patient with severe carpal tunnel syndrome. In addition to slowed latency from wrist to the abductor pollicis brevis and slowed NCV from wrist to palm, the amplitude of the CMAP is well below the lower limit of normal. These parameters reflect the summated voltage over time produced by all the contracting motor units within the test muscle. Therefore, as this muscle is partially denervated, fewer motor units are contracting after nerve stimulation. This will cause the CMAP amplitude to decrease. Temporal dispersion may accentuate, depending on the conduction velocity of the intact units. Similar changes may also be evident in myopathic conditions, in which all motor units are intact but fewer fibers of variable size are available in each motor unit.
The shape of the CMAP can also be variable. Deviation from a smooth curve need not be abnormal, and it is often useful to compare the proximal and distal CMAPs with each other as well as with the contralateral side if indicated. They should be similar. In abnormal conditions, changes in shape may be the result of a significant slowing of conduction in some axons, repetitive firing, or asynchronous firing of axons after a single stimulus. Anatomic variations in innervation of muscle can also influence distal to proximal CMAP morphology, so it is important that EDX providers have a thorough understanding of neuromuscular anatomy and physiology.
Sensory Nerve Conduction Studies
Sensory neurons demonstrate the same physiological properties as motor neurons, and NCV can be measured in a similar way. However, some differences in technique are necessary to differentiate between sensory and motor axons. Although sensory fibers can be tested using orthodromic conduction (physiological direction) or antidromic conduction (opposite to normal conduction), antidromic measurements appear to be more common. For the same reason that motor axons are examined by recording over muscle, sensory axons are either stimulated or recorded from digital sensory nerves. This minimizes the activity of the motor axons from the recorded potentials.
The stimulating electrode used for sensory NCS tests is typically provided by ring, surface, or needle electrodes placed around the base of the digit innervated by the nerve, directly over or subcutaneously near the anatomic location of the nerve. Again, a comprehensive understanding of surface anatomy is paramount. Sensory potentials for the median and ulnar nerves can be recorded antidromically by stimulating at the wrist, elbow, and upper arm. Typically the sensory study of these nerves is limited to stimulation at the wrist. Other sensory nerves can be studied in the UEs and include the superficial radial, medial antebrachial cutaneous, lateral antebrachial cutaneous, and the dorsal cutaneous branch of the ulnar nerve. In the LEs, the sensory nerves most commonly studied are the sural nerve and the superficial fibular (peroneal) nerves. Other nerves that have been studied include the lateral femoral cutaneous, saphenous, and deep fibular (peroneal). Sensory evoked potentials are also called sensory nerve action potentials (SNAPs). Like motor nerve conduction, the sensory conduction parameters of interest include latency, amplitude, NCV, and morphology. Normal sensory NCV ranges between 40 and 75 m/s. Amplitude, measured with surface electrodes, are variable and can range from 2 uV to 120 uV, and duration should be short, generally less than 2 msec. Sensory evoked potentials are usually sharp, not rounded like the CMAP. Sensory NCVs are slightly faster than motor NCVs because of the uniformly larger diameter of sensory nerves contributing to the SNAP.99,103 Figure 5.13A depicts normal antidromic sensory NCS of a median nerve to digit III. In this figure, latency is measured to the negative peak of the potential and SNAP amplitude from baseline to negative peak. Figure 5.13B depicts abnormal sensory NCS from the median nerve of a patient with mild carpal tunnel syndrome.
Sensory nerve conduction study. (A) Normal median sensory nerve conduction study. (B) Mild focal slowing of median sensory nerve at wrist.
The H reflex is a useful diagnostic measure for radiculopathy and peripheral neuropathy. Its most common application is in testing the integrity of the sensory and motor monosynaptic pathways of S1 nerve roots via the tibial nerve and to a lesser extent at C6–C7 and L3–4 via the median or femoral nerves.111 In traditional testing, a submaximal stimulus is applied to the tibial nerve at the popliteal fossa, and a motor response is recorded from the medial portion of the soleus muscle. The action potentials travel along the IA afferent neurons toward the spinal cord, synapsing onto interneurons at the level of the spinal cord, then alpha motor neurons within the anterior horn. The consequent activation of the motor neuron leads to an impulse traveling peripherally to the soleus muscle, resulting in a muscle contraction. Because the stimulus causes impulses to travel both distally and proximally within a mixed motor and sensory neuron, the latency of this response comprises a measure of the integrity of both sensory and motor fibers (Fig. 5.14A).
Late responses. (A) Reflex of tibial nerve. (B) Waves of ulnar nerve.
An average tibial H reflex latency is around 30 msec but is affected by limb length, age, and temperature.112 A slowed latency with otherwise normal distal NCS parameters is indicative of abnormal proximal function, often from a herniated disc or other nerve root impingement. Because of more proximal involvement, the peripheral motor and sensory NCS would not be affected. This latency may also identify nerve root compression before obvious EMG changes occur.
F waves are a form of NCS test that allows for study of proximal nerve segments that would otherwise be inaccessible to routine nerve conduction studies. F wave abnormalities can be an indicator of peripheral nerve pathology or demyelination. The F wave ratio compares the conduction in the proximal half of the total pathway with the distal and may be used to determine the site of conduction slowing—for example, to distinguish a root lesion from a distal generalized neuropathy.113 The F wave is elicited by the supramaximal stimulus of a peripheral nerve at a distal site, leading to propagation of impulses in both directions. While the orthodromic impulse travels to the distal muscle, the antidromic response travels to the anterior horn cell, depolarizing the axon hillock, leading to depolarization of dendrites, which in turn depolarizes the axon hillock once again, generating an orthodromic volley back to the muscle. No synapse is involved, so the F wave is not considered a reflex, but rather a measure of motor neuron conduction (see Fig 5.14B).
The F wave has some merit to assist other nerve conduction and EMG measures and is most helpful in the diagnosis of conditions where the most proximal portion of the axon is involved, such as Guillain-Barré syndrome, thoracic outlet syndrome, brachial plexus injuries, and radiculopathies with more than one nerve root involved.105 The latency of the F wave is generally about 30 seconds in the upper limb and less than 60 seconds in the lower limb and is influenced by age and limb length. Only a small percent of motor neurons actually participate in the F response.99 Because it is an inconsistent response, it must be calculated on the basis of at least 10 successive trials.103
Repetitive Nerve Stimulation
Repetitive nerve stimulation (RNS) is a technique used to evaluate for suspected NMJ disorders. RNS requires technical proficiency, attention to electrode placement, immobilization of the limb, and temperature control.114-117 A baseline CMAP is recorded from a target muscle (preferably one that is clinically weak). Then a protocol of repetitive supramaximal stimulations is delivered in a train of 5 to 10 CMAPs at a low rate of stimulus frequency, around 2 Hz to 5 Hz. Comparisons of amplitude are made typically between the first and fourth or first and fifth CMAPs. If there is an amplitude decrement greater than 10%, it is considered abnormal. Following the first train of stimuli, a protocol series of exercise and periods of rest are performed between trains of stimuli. If there is decrement at low rates of stimulation, either fast RNS (20–50 Hz) or brief isometric contraction (to mimic tetany) is performed followed by another train of stimuli. In NMJ disorders, the response to slow and fast RNS can help determine pre- versus post-synaptic NMJ disorders (Fig. 5.15)
Repetitive nerve stimulation of ulnar nerve.
Disorders of Peripheral Nerve
Electrophysiological findings usually correlate with clinical signs in patients with neuropathic or myopathic involvement. As discussed, lesions of peripheral nerve fall into two categories, demyelination and axonal loss. Lesions involving peripheral nerve can occur focally, diffusely, segmentally or along a specific nerve root distribution. They can also involve motor nerves sparing sensory nerves or sensory nerve sparing motor nerves. A skilled electrodiagnostician should have the capabilities to perform and interpret findings from the EMG and NCS in order to concisely convey an impression of the neurophysiological state of a patient’s condition.
In focal neuropathy, NCS tests can detect evidence of degeneration and slowing of fibers across the site of compression but may be normal above and below that site. For example, patients with carpal tunnel syndrome may have abnormal motor and sensory NCS findings across the wrist with normal findings in the median nerve proximal to the carpal tunnel and in other nerves of the same limb. EMG abnormalities of increased insertion, fibrillations, and positive sharp waves may be noted in the thenar muscles but not in proximal median innervated muscles or other muscles of the same limb. Recruitment in the thenar muscles may show reduced interference pattern with fast-firing, long-duration polyphasic and/or larger than normal amplitude MUAPs.
Radiculopathy may involve sensory and motor abnormalities on clinical examination, but generally, the motor and sensory NCS is fairly normal unless motor axon loss is severe, at which point low amplitude CMAPs may be evident. SNAP amplitudes are generally preserved, despite the patient complaining of altered sensation, as most nerve root compression due to radiculopathy occurs proximal to the dorsal root ganglia and the cell body is in continuity with the distal axon. This is considered a preganglionic lesion and is an important distinguishing EDX characteristic in radiculopathy. H reflexes or F waves could be absent or delayed in the nerves supplied by the involvement spinal root level. EMG may show evidence of neuropathy with insertion, rest, and volitional testing in the distribution of the radiculopathy with sparing of muscles supplied by other nerve root levels.
Plexopathy refers to neuropathy involving components of the brachial, cervical, lumbar, or lumbosacral plexuses. It can be related to a number of different causes, including trauma, immune-mediated inflammation, birth-related injury, tumors, surgical complications, and unknown etiology (idiopathic). Plexopathy can occur in isolated trunks, divisions, or cords with or without extension of the injury distal into the terminal nerve branches. Isolated lesions involving a single trunk, division, or cord is rare. Slowed NCV across the plexus can be recognized. SNAP amplitudes with EMG changes in the distribution of the plexus components involved are the most sensitive indicators of sensory and motor axon loss in plexopathy.118,119 Sparing of EMG changes in the paraspinal muscles helps rule out nerve root pathology.
Polyneuropathy affects multiple nerves and typically results in sensory changes, distal weakness, and hyporeflexia. It can be related to medical conditions, such as diabetes, alcoholism, or renal disease, as well as secondary complications related to cancer and its treatments.99,103 These conditions manifest as axonal damage and/or demyelination and sensory and/or motor nerve involvement. Polyneuropathy can occur diffusely, segmentally, or peripherally. While EMG/NCS findings are not specific to a single diagnosis, the pattern of involvement can assist with narrowing the cause of the polyneuropathy.120
Motor neuron disorders typically involve degeneration of the anterior horn cell, such as in poliomyelitis, or diseases that involve both UMNs and LMNs, such as ALS. Abnormal resting potentials are classically seen with these disorders, as well as reduced recruitment, allowing single motor unit potentials to be visible even with an interference pattern. CMAPs may be reduced in amplitude. Motor NCV can be slightly slowed, depending on the distribution of degeneration and propensity for larger, faster fibers to undergo loss earlier in the disease. EMG will show neuropathic patterns with normal sensory NCS findings.
In myopathy, such as the muscular dystrophies and inflammatory myopathies (e.g., myositis, polymyositis, dermatomyositis), the motor unit remains intact, but degeneration of muscle fibers is evident. Motor latencies and NCVs should be normal, although the CMAP amplitudes can be reduced in more weak and involved muscles. Sensory NCS will also be normal. In early stages, EMG will show prolonged insertion activity, fibrillations, and positive sharp waves at rest and often complex repetitive discharges. Recruitment will show full interference patterns with minimal recruitment (early recruitment) with motor unit morphology demonstrating short-duration, low-amplitude polyphasic potentials with voluntary activity. In advanced stages, insertional activity becomes reduced with increased resistance to needle advancement with little electrical activity due to fibrosis of muscle tissue.
Neuromuscular Junction Disorders
In NMJ disorders, such as botulism toxicity, myasthenia gravis (MG), and Lambert-Eaton myasthenic syndrome (LEMS), there is a disturbance of neuromuscular transmission occurring at the NMJ at either the pre- or post-synaptic terminals. EDX abnormalities are most commonly seen using repetitive nerve stimulation and with observed CMAP changes to brief exercise. In post-synaptic NMJ disorders such as MG, baseline CMAPs are generally normal with no change following brief contraction. In slow RNS of 2 Hz to 5 Hz, a decrement may be seen.116 With fast RNS of 20 Hz to 30 Hz, decrements persist or do not change. In LEMS, baseline CMAPs are generally low with significant improvement following brief exercise. Slow RNS shows decrement, but fast RNS demonstrates significant increase in CMAP amplitude.115 Sensory NCS and EMG is generally normal in NMJ disorders, although there may be some MUAP amplitude variability. In LEMS, one could expect reduced recruitment.114
Assessing Severity and Estimating Prognosis in Neuropathy
Seddon and Sunderland offer a reasonable concept of nerve injury classification and lay the foundation for the progression of nerve injury from demyelination to axon loss.101,102 Table 5.13 illustrates a proposed alternative classification of injury-predicted recovery.99,121,122 It identifies severity of nerve injury from very mild to complete. Characteristics of the EDX in both the EMG and NCS can assist in determining if an injury involves predominantly myelin, axons, or both. Typically, more mild cases of neuropathy involve myelin and are characterized by slowing of motor/sensory latencies and NCV with possible conduction block. Nerve injury is considered more severe as axonal loss occurs. Axonal loss is recognized in the NCS as decrements in SNAP and CMAP amplitudes, and in the EMG with abnormal resting potentials, reduced recruitment, and motor unit morphological changes representing evidence of collateral axonal sprouting. Preservation of motor axons is essential for muscle function and is therefore a critical characteristic of interest. When there is significant loss of motor axons, the timing required for healing can be protracted with the possibility of residual deficit that can be influenced by age, other illnesses, and distance from the injury to the target muscles.
Table 5.13Alternative Classification of Nerve Injury Progression and Prognosis ||Download (.pdf) Table 5.13 Alternative Classification of Nerve Injury Progression and Prognosis
|Severity ||Type of Injury ||EDX Findings ||Recovery ||Time to Recover ||Prognosis |
|Very Mild || |
Mild slow NCV/latencies
Preserved SNAPs/CMAPs Normal EMG
|Remyelination ||2–12 weeks ||Excellent |
|Mild || |
Advanced slow NCV/Latencies Partial CB, preserved
SNAP/CMAP Normal EMG
|Remyelination ||2–12 weeks ||Excellent |
|Moderate ||Demyelination/Mild Axon Loss || |
Reduced SNAP, Preserved CMAPs
CB, mild EMG changes
|2–6 months ||Excellent |
|Severe || |
Absent SNAP, Low CMAP
Advanced EMG changes
|Collateral Sprouts ||2–6 months ||Good |
|Profound ||Severe Axon Loss || |
CMAP 80%–100% reduced
Advanced EMG changes
|up to 18 months ||Guarded/Fair |
|Complete ||Severed Nerve || |
Advanced EMG changes
|Surgery required ||Protracted ||Guarded at Best |