Studies of Nerve Conduction
The main laboratory technique for the study of peripheral nerve function involves the transcutaneous stimulation of motor or sensory nerves and recording of the elicited action potentials in the muscle (CMAP) and the sensory nerve action potential (SNAP). The results of these motor and sensory nerve conduction studies, expressed as amplitudes, conduction velocities, and distal latencies, yield certain quantitative information and additional qualitative observations regarding the waveform and dispersion of electrical neural and muscular impulses.
Hodes and coworkers, in 1948, were the first to describe nerve conduction studies in patients, and the techniques used currently are not much changed. An accessible nerve is stimulated through the skin by surface electrodes, using a stimulus that is large enough to recruit (cause a discharge in) all the available nerve fibers. The resulting action potential is recorded by electrodes on the skin (1) over the muscle distally in the case of motor fibers stimulated in a mixed or motor nerve (CMAP), (2) over the nerve more distally, using antidromic techniques for sensory nerve conduction studies (this has technical advantages over orthodromic techniques), and (3) over the nerve more proximally for mixed (sensory and motor) nerve conduction studies (Fig. 45-3). These techniques are the ones used most often in clinical work. An alternative but much more demanding technique uses "near-nerve" needle electrodes to record action potentials as they course through the nerve. The main characteristics of the conventional nerve conduction studies are described below.
The median nerve is stimulated percutaneously (1) at the wrist and (2) in the antecubital fossa with the resultant compound muscle action potential recorded as the potential difference between a surface electrode over the thenar eminence (arrow) and a reference electrode (Ref.) more distally. Sweep 1′ on the display depicts the stimulus artifact followed by the compound muscle action potential. The distal latency, A′, is the time from the stimulus artifact to the take-off phase of the compound muscle action potential and corresponds to conduction over distance A. The same is true for sweep 2′, where stimulation is at 2 and the time from the artifact to the response is A′ + B′. The maximum motor conduction velocity over segment B is calculated by dividing distance B by the time B′.
Distal (Terminal) Latencies, Conduction Times, and Conduction Velocities
The conduction times from the most distal stimulating electrode to the recording site over a muscle, in milliseconds, as determined by the latency from the stimulus artifact to the onset and to the peak of the CMAP, are termed the distal (or terminal) and peak motor latencies, respectively (see Fig. 45-3). The former is the one used more often as a reflection of conduction time in routine work. A stimulus may then be applied to the nerve at a second site more proximally (or if recording electrodes can be placed more proximally in the case of sensory fibers), and a conduction time can be measured over a longer segment of nerve. When the distance (in millimeters) between the two sites of stimulation is divided by the difference in conduction times (in milliseconds), one obtains a conduction velocity (in meters per second), which describes the maximal velocity of propagation of the action potentials in the largest-diameter and fastest-conducting nerve fibers. These velocities in normal subjects vary from a minimum of 40 or 45 m/s to a maximum of 65 to 75 m/s, depending upon which nerve is studied (e.g., slower in the legs than in the arms; Table 45-1). Values are lower in infants, reaching the adult range by the age of 2 to 4 years, and declining again slightly with advancing age. Conduction velocity also is diminished with exposure to cold, a potentially important factor if these recordings are taken when the patient's skin is cool; consequently, measurement of skin temperature is routinely done prior to performing the nerve conduction tests.
Table 45-1 Normal Values for Representative Nerve Conduction Values at Various Sites of Stimulation (Mean Values ± 2 SD for Adults 16 to 65 Years of Age) ||Download (.pdf)
Table 45-1 Normal Values for Representative Nerve Conduction Values at Various Sites of Stimulation (Mean Values ± 2 SD for Adults 16 to 65 Years of Age)
Motor Nerve Conduction Studies
DISTAL STIMULATION SITE
OTHER STIMULATION SITES
ONSET LATENCY (ms)
F-WAVE LATENCY (ms)
Sensory Nerve Conduction Studiesb
DISTANCE STIMULATION SITES
ONSET LATENCY (ms)
PEAK LATENCY (ms)
Normal values have been established for distal latencies from the usual sites of stimulation on various mixed nerves to the appropriate muscles. Stimulating the median nerve at the wrist, for example (see electrode 1 and segment A in Fig. 45-3), has a latency for motor conduction through the carpal tunnel to the median-innervated thenar muscles of less than approximately 4.5 ms in healthy adults. Similar normal values have been compiled for orthodromic and antidromic sensory conduction velocities and for distal latencies in all the main peripheral nerves (see Table 45-1).
Disease processes that preferentially injure the fastest-conducting, large-diameter fibers in peripheral nerves reduce the maximal conduction velocity because the remaining thinner fibers conduct more slowly. In most neuropathies, all of the axons are affected either by a fairly uniform "dying-back" phenomenon or by wallerian degeneration as described in Chap. 46, and nerve conduction velocities are then less informative. This is true, for example, in typical alcoholic-nutritional, carcinomatous, uremic, diabetic, and other metabolic neuropathies, in which conduction velocities range from the low-normal range to mildly slowed. In these "axonal neuropathies," the motor and sensory nerve amplitudes are diminished.
By contrast, demyelinating neuropathies of the acute (Guillain-Barré) and chronic types, such as chronic inflammatory, metachromatic leukodystrophy, and the common type of Charcot-Marie-Tooth disease, show marked slowing of conduction and, in the case of the acquired demyelinating diseases, there is also dispersion of the motor action potential and a highly characteristic conduction block (see later).
Amplitude of the Compound Muscle Action Potential
In addition to the study of distal latency and conduction velocity, the amplitude of the evoked muscle action potential (CMAP) yields valuable information about peripheral nerve function. These amplitudes are a semiquantitative measure of the number of nerve fibers that respond to a maximal stimulus (and of the innervated volume of muscle). Demyelinative lesions affecting the large, fast-conducting fibers are detected by the finding of differential slowing among various caliber fibers that causes a dispersal of the CMAP response. Reduction in motor and sensory amplitudes is a more specific and sensitive indicator of axonal loss than is slowing of conduction velocity or prolongation of distal latencies. Conversely, prolonged distal latencies and slowed motor conduction velocities, as well as conduction blocks and dispersed responses (described below), are the hallmarks of demyelinative lesions. Table 45-1 shows the range of normal amplitudes for the CMAPs that are elicited by stimulation of the main motor nerves.
It is usually possible to obtain a reliable motor conduction study as long as some functioning nerve fibers remain intact. The conduction velocities then reflect the status of the surviving axons, and the velocity may be normal or nearly so despite widespread axonal degeneration. This is most apparent following incomplete transection of a nerve; the maximal motor conduction velocity may be normal in the few remaining fibers, although the muscle involved is almost paralyzed and the compound muscle potential recorded from it is very low.
Sensory Nerve Action Potentials
When motor fibers in a mixed nerve are stimulated, an amplified CMAP of many hundreds of microvolts can easily be recorded from electrodes on the skin over the muscle. However, when one attempts to measure sensory potentials, where activity must be recorded from nerve fibers themselves, the "amplification" provided by many motor units is not available and electronic amplification is required. Sensory potentials are sometimes very small or absent even when powerful computer-averaging techniques are used, and sensory conduction measurements may then be difficult to determine. Table 45-1 gives the range of normal values for sensory nerve action potential amplitudes and velocities.
By stimulating a motor nerve at multiple sites along its course, it is possible to demonstrate segments in which conduction is partially "blocked" or is differentially slowed. From such data one infers the presence of a multifocal demyelinative process in motor nerves. This contrasts with the findings in certain of the inherited and metabolic demyelinating neuropathies, in which all parts of the nerve fiber are altered to more or less the same degree, i.e., there is uniform slowing and reduction in amplitude and no conduction block.
As a technical matter, conduction block is demonstrated by a reduction in the amplitude of the CMAP elicited from the proximal site along the motor nerve, compared to stimulation at a distal site. Generally, a 40 percent reduction in amplitude over a short distance of nerve, or 50 percent over a longer distance, qualifies as a block, one possible exception being along the tibial nerve, in which there is some degree of physiologic dispersion (it is also difficult to stimulate all the motor nerve fibers of this nerve in obese patients); therefore a slight drop in amplitude over the length of the nerve is normally expected. It is important to be sure that any reduction in amplitude along the course of the nerve is not solely a result of dispersion of the waveform. The presence of a conduction block can also be inferred from the finding of poor recruitment of muscle action potentials and the concurrent absence of active denervation (see further on). The finding of conduction block is a main feature of a number of acquired immune demyelinating neuropathies including Guillain-Barré syndrome, chronic inflammatory demyelinating neuropathy, and multifocal conduction block associated with the GM1 antibody, which are discussed in Chap. 46.
Focal conduction block may be caused simply by nerve compression at certain common sites (fibular head, across the elbow, flexor retinaculum at the wrist, etc.) rather than to an intrinsic disease of the peripheral nerves. Focal compression of nerve, as occurs in these entrapment syndromes, produces localized slowing or blocks in conduction, perhaps because of segmental demyelination, only at the site of compression. The demonstration of such localized changes of conduction affords ready confirmation of nerve entrapment; for example, if the distal latency of the median nerve (see A, Fig. 45-3) exceeds 4.5 ms while that of the ulnar nerve remains normal, compression of the median nerve in the carpal tunnel is likely. Similar focal slowing or partial block of conduction may be recorded from the ulnar nerve at the elbow and from the peroneal nerve at the fibular head.
Electrodiagnostic Studies of Nerve Roots and Spinal Segments (Late Responses, Blink Responses, Evoked Responses)
Information about the conduction of impulses through the proximal segments of a nerve is provided by the study of the H reflex and the F wave. In 1918, Hoffmann, after whom the H reflex was named, showed that submaximal stimulation of mixed motor–sensory nerves, insufficient to produce a direct motor response, nonetheless induces a muscle contraction (H wave) after a latency that is far longer than that of the direct motor response. This reflex is based on the activation of afferent fibers from muscle spindles (the same axons that conduct the afferent volley of the tendon reflex), and the long delay reflects the cumulative time required for the impulses to reach the spinal cord via the sensory fibers, synapse with anterior horn cells, and to be transmitted along motor fibers to the muscle (see Fig. 3-1). Thus the H reflex is the electrical representation of the tendon reflex circuit and is a useful measure because the impulse traverses both the posterior and anterior spinal roots. The H reflex is particularly helpful in the diagnosis of S1 radiculopathy and of polyradiculopathies. Its status generally parallels that of the clinically elicited Achilles reflex. However, it is difficult to obtain an H reflex from nerves other than the tibial. Stimuli of increasing frequency but low intensity cause a progressive depression and finally obliteration of H waves. The latter phenomenon has been used to study spasticity, rigidity, and cerebellar ataxia, in which there are differences in the frequency-depression curves of H waves. In parallel with the Achilles tendon reflex, the H-reflex is transiently obliterated in spinal shock (see Chap. 44).
The F response, so named because it was initially elicited in the feet, was first described by Magladery and McDougal in 1950. It is evoked by a supramaximal stimulus of a mixed motor–sensory nerve. After a latency that is longer than for the direct motor response (latencies of 28 to 32 ms in arms, 40 to 59 ms in legs), a second small muscle action potential is recorded. This F wave is the result of the impulses that travel antidromically in motor fibers to the anterior horn cells, a small number of which are activated and produce an orthodromic response that is recorded in a distal muscle. The F response is a more reliable test than the H wave of proximal motor nerve and root conduction in that the F wave traverses only the ventral root and can be elicited from a number of muscles. The combination of a normal F response and an absent H reflex is found in diseases of sensory nerves and roots. Both of these "late responses" find their main use as corroborative tests that are interpreted in the context of the entire nerve conduction examination (see Wilbourn). As with the H reflex, the F wave may be absent in the state of spinal shock (see Chap. 44). Table 45-1 gives the normal F wave response latencies.
This special nerve conduction test is not in frequent clinical use but it serves a purpose in the diagnosis of certain demyelinating neuropathies and in any process that affects the trigeminal or facial nerve. The supraorbital (or infraorbital) nerve is stimulated transcutaneously and the reflex closure of both orbicularis oculi muscles is recorded with surface electrodes. Two CMAP bursts are observed: the first (R1) appears ipsilaterally 10 ms after the stimulus and the second (R2), ipsilaterally at 30 ms and contralaterally up to 5 ms later. The amplitudes of the responses vary considerably and are not in themselves clinically important. The first response is not visible as a muscular contraction but may serve some preparatory function by shortening the blink reflex delay. R1 is mediated by an oligosynaptic pontine circuit consisting of one to three neurons located in the vicinity of the main sensory nucleus; R2 uses a broader and less-well-defined reflex pathway in the pons. It has been established that R1 and R2 are generated by the same facial motor neurons.
The elicitation of blink reflexes establishes the integrity of the afferent trigeminal nerve, the efferent facial nerve, and the interneurons in the pons (R1) and caudal medulla (related to the bilateral R2 response). The test may also be helpful in identifying a demyelinating neuropathy when the facial and oropharyngeal muscles are affected and those of the limbs are relatively spared, thereby leaving conventional nerve studies normal. In such cases, the blink responses are delayed ipsilaterally and contralaterally as a result of conduction block in the proximal facial nerve. Direct facial nerve stimulation often fails to demonstrate this block because only the distal segment of the nerve is amenable to study. Although the blink responses are rarely necessary for diagnosis, most patients with hereditary neuropathy have blink response abnormalities. In Bell's palsy there is a delay or absence of R1 and R2 responses only on the affected side. Large acoustic neuromas (vestibular schwannomas) may interfere with the afferent trigeminal portion of the pathway and give rise to abnormal responses on the affected side. Diseases of the brainstem have yielded inconsistent responses. It is noteworthy that the test is normal in patients with trigeminal neuralgia.
Segmental Motor, Cranial, and Somatosensory Evoked Potentials (See also Chap. 2)
These techniques find use in diseases that affect the spinal roots and in studying central pathways.
By applying a magnetic stimulus, which induces an electrical impulse, or by a directly delivered electrical stimulus over the lower cervical or lumbar spine, it is possible to activate the motor (anterior) roots and to measure the time required to elicit a muscle contraction (see review by Cros and Chiappa). These root stimulation tests can be quite uncomfortable for the patient as a result of the contraction of muscles surrounding the stimulation site.
Transcranial magnetic stimulation of the cerebral cortex permits measurement of the latency of muscle contraction after excitation of motor neurons in the cortex. Thus, the integrity of the entire corticospinal system, from the cortical motor neurons through spinal tracts, anterior horn cells, and the peripheral motor nerve can be determined. By combining this technique with the above-described root stimulation, it becomes possible to measure central and peripheral motor conduction times. These forms of motor testing have their main use in the study of multiple sclerosis, amyotrophic lateral sclerosis (ALS), and related disorders.
As described in Chap. 2, by applying repetitive electrical stimuli to a peripheral nerve, the sensory evoked responses can be recorded from sites along the nerve and plexus as well as in central pathways (the thalamus and somatosensory cortex). These somatosensory evoked potential tests find their main use in the diagnosis of multiple sclerosis and in disorders of the sensory nerve roots as discussed in Chaps. 2 and 36. Discussion of magnetic stimulation, collision techniques, and quantitative EMG, among other topics, can be found in several monographs, such as the ones by Kimura, by Aminoff, and by Brown and Bolton.
Repetitive Motor Nerve Stimulation (See also Chap. 49)
This test of the neuromuscular junction is based on Jolly's observation in 1895 that in myasthenia gravis the strength of muscular contractions progressively declines in response to a train of stimuli. By adjusting the amplitude of a stimulus over a nerve to the supramaximal range, a maximal CMAP may be obtained for each stimulus. With repeated stimuli, each response will have the same waveform and amplitude until normal fatigue supervenes. In a healthy individual, a muscular response follows each stimulus with rates of stimulation up to 25 per second for periods of 60 s or more before a decrement of the CMAP appears. In certain disorders, notably myasthenia gravis, a train of 4 to 10 stimuli at rates of 2 to 5 per second (optimally 2 to 3 per second), the amplitude of the motor potentials decreases and then, after four or five further stimuli, may increase slightly (Fig. 45-4A). A progressive reduction in amplitude is most likely to be found in proximal muscles, but these are not easily stimulated for which reason the locations most commonly used for clinical testing are the accessory nerve in the posterior triangle of the neck (trapezius), the ulnar nerve (hypothenar muscle), the median nerve at the wrist (thenar muscle), and the facial nerve (orbicularis oculi muscle). A decrement of 10 percent or more denotes a failure of a proportion of the neuromuscular junctions.
Compound action potentials evoked in hypothenar muscles by electrical stimulation of the ulnar nerve at the wrist. A. Patient with myasthenia gravis—typical pattern of decrement in first four responses followed by slight increment. At this rate of stimulation (4 per second), the decrement in response does not continue to zero. B. Patient with Lambert-Eaton syndrome and oat cell carcinoma—typical marked increase toward normal amplitude with rapid repetitive stimulation (20 per second). Horizontal calibration: 250 ms.
The sensitivity of the procedure is improved by first exercising the tested muscle for 30 to 60 s; a form of posttetanic potentiation. (The full procedure consists of testing the muscle with a train of stimuli before and immediately after exercise [or maximal voluntary contraction] and at 30-s intervals for several minutes. The posttetanic potentiation partially compensates for the depletion of ACh during slow rates of stimulation; this is followed by a decrease in the excitability of the neuromuscular junction during the approximately 2 to 4 min after exercise.) The induced failure of neuromuscular transmission in myasthenia is similar to the one produced by curare and other nondepolarizing neuromuscular blocking agents, and both cases can be partially corrected with anticholinesterase drugs such as neostigmine and edrophonium. Similar but lesser decremental responses may occur in poliomyelitis, ALS, and certain other diseases of the motor unit or motor nerve, particularly those resulting in the growth of reinnervating nerve twigs.
The Lambert-Eaton myasthenic syndrome, often associated with oat cell carcinoma of the lung, as discussed in Chap. 49, is characterized by a presynaptic blockage of acetylcholine release and produces the opposite defect of neuromuscular transmission to the one recorded in myasthenia gravis. During tetanic stimulation (20- to 50-per-second repetitive stimulation of nerve), the muscle action potentials, which are small or practically absent with the first stimulus, increase in voltage with each successive response until a more nearly normal amplitude is attained (see Fig. 45-4B). Exercising the muscle for 10 s before stimulation will cause a similar posttetanic facilitation in patients with the Lambert-Eaton syndrome (200-fold increases are not uncommon). A less important decremental response to slow stimulation may occur, but it is difficult to discern because of the greatly diminished amplitude of the initial responses. Neostigmine has little effect on this phenomenon, but it is reversed by guanidine and 3,4-diaminopyridine, which stimulate the presynaptic release of ACh. The effects of botulinum toxin and of aminoglycoside antibiotics are similar, i.e., being active at the presynaptic membrane, they produce an incremental response at high rates of stimulation.
The single-fiber EMG, discussed in a later section, is an even more sensitive method of detecting failure of the neuromuscular junction.
Needle Examination of Muscle (Electromyography)
This technique requires the use of monopolar or concentric bipolar needle electrodes, which are inserted into the body of the muscle to record the electrical activity generated by contraction. With concentric electrodes, the tip of the wire that runs in the hollow of the needle is in proximity to many muscle fibers belonging to several different overlapping motor units; this is the active recording electrode. The shaft of the needle, in contact over most of its length with intercellular fluid and many other muscle fibers, serves as the reference electrode. Monopolar electrodes use the uninsulated needle tip as the active electrode, while the reference electrode may be another monopolar needle electrode placed elsewhere in subcutaneous tissue or a surface electrode on the skin overlying the muscle. Patients almost invariably find this portion of the test uncomfortable and should be prepared by a description of the procedure. Rapid and brief needle insertion by the skilled examiner makes the test more tolerable.
As the electrical impulse travels along the surface of the muscle toward the recording electrode, a positive potential is recorded on the oscilloscope, i.e., the recorded signal is deflected downward by convention (at A in Fig. 45-5). When the depolarized zone moves under the recording electrode, it becomes relatively negative and the beam is deflected upward (at B). As the depolarized zone continues to move along the sarcolemma, away from the recording electrode, the current begins to flow outward through the membrane toward the distant depolarized region, and the recording electrode becomes relatively positive again (at C). There is then a return to the resting isopotential position. The net result is a triphasic action potential, as in Fig. 45-5. This configuration is typical of the firing of a single fiber.
The shaded area represents the zone of the action potential, which is negative to all other points on the fiber surface. It is shown at three points in its course (from left to right) along the fiber. At each point, the correspondingly lettered portion of the triphasic muscle action potential displayed on the display screen reflects the potential difference between the active (vertical arrow) and reference (Ref.) electrodes. Polarity in this and subsequent figures is negative upward as depicted. The time calibration is on the screen.
The electrical activity of various muscles is recorded both at rest and during active contraction by the patient. As indicated earlier, muscle fibers do not normally discharge until activated together in motor unit activity. This involves the almost simultaneous contraction of all the muscle fibers innervated by a single anterior horn cell. Although the typical configuration of a motor unit potential (MUP) is triphasic, up to 10 percent of normal MUPs consist of four or more phases (polyphasic potentials); however, an excess of polyphasic potentials beyond this is pathologic.
Normal muscle in the resting state should be electrically silent; the small tension spoken of as muscle tone has no EMG equivalent. There are, however, two closely related types of normal spontaneous activities and another that is induced by the insertion of the needle itself. One is a low-amplitude, 10- to 20-μV monophasic (negative) potential of very brief (0.5 to 1 ms) duration. These represent single or synchronized miniature end plate potentials (MEPPs) because of the small number of ACh quanta that are being released all the time. They are normally sparse but are most evident when the recording needle electrode is placed near a motor endplate ("endplate noise"). Fortuitous placement of the needle electrode very close to or in contact with the endplate gives rise to a second type of normal spontaneous activity. That is characterized by irregularly discharging high-frequency (50- to 100-Hz) biphasic spike discharges, 100 to 300 μV in amplitude (i.e., large enough to cause an isolated muscle action potential). These potentials have been termed endplate spikes and represent discharges of single muscle fibers excited by spontaneous activity in nerve terminals. They must be distinguished from fibrillation potentials (see later). Finally, insertion of the needle electrode into the muscle injures and mechanically stimulates a number of fibers, causing a burst of potentials of short duration (300 ms). This is referred to as normal insertional activity, but the extent of this activity is greatly raised in certain pathologic states as noted below.
When a muscle is voluntarily contracted, the action potentials of motor units begin to appear. One can observe a pattern of force build up by watching the progressive recruitment of MUPs; the initial ones, representing smaller motor units, firing at rates of 5 to 10 per second. With increased force of contraction, there is recruitment of larger, previously inactive motor units as well as an increased rate of firing (40 to 50 per second; Fig. 45-6A). Because individual MUPs can no longer be distinguished during maximal voluntary contraction, this activity is referred to as a complete interference pattern (Fig. 45-6A, right). This is seen not only as a summated signal pattern but is also heard as a mixed high-frequency clicking when the electrical activity is made audible. As muscles relax, an increasing number of units drop out. If a muscle is weakened by denervation or if electrical conduction is blocked, there obviously will be fewer MUPs, but the firing rate is still rapid (reduced recruitment; see Fig. 45-6B). In contrast, with poor voluntary effort and with upper motor neuron lesions, the MUPs fire in decreased numbers, at slower rates, and often in an irregular pattern (termed poor activation).
Patterns of motor unit recruitment. A. Normal. With each increment of voluntary effort, more and larger units are brought into play until, with full effort at the extreme right, a complete "interference pattern" is seen in which single units are no longer recognizable. B. After denervation, only a single motor unit is recorded despite maximal effort. It is seen to fire repetitively. C. With myopathic diseases, a normal number of units are recruited on minimal effort, though the amplitude of the pattern is reduced. Calibrations: 50 ms (horizontal) and 1 mV in A and B; 200 μV in C (vertical).
In the usual EMG examination, a plan for the study is made based on detailed knowledge of muscular innervation and focusing on the regions affected by weakness. In some patients, as in those with motor neuron diseases or polymyositis, a wider sampling of muscles is required to detect changes in asymptomatic regions.
The Abnormal Electromyogram
Clinically important deviations from the normal EMG include (1) increased or decreased activity upon insertion of the needle; (2) the occurrence of abnormal "spontaneous" activity during the relaxed state (fibrillation potentials, positive sharp waves, fasciculation potentials, cramp potentials, myotonic discharges, myokymic potentials); (3) abnormalities in the amplitude, duration, and shape of single MUPs; (4) a decrease in the number of MUPs and changes in their firing pattern; (5) variation in amplitude and number of phases of MUPs during voluntary contraction; and (6) the demonstration of special phenomena, such as electrical silence during shortening of the muscle (physiologic contracture and states of continuous muscle fiber activity).
At the moment the needle is inserted into muscle, there is a brief burst of action potentials that ceases once the needle is stable, provided that it is not in a position to irritate a nerve terminal. Increased insertional activity is seen in most instances of denervation as well as in many forms of primary muscle disease and in disorders that dispose to muscle cramps. In cases of advanced denervation or myopathy, in which muscle fibers have been largely replaced by connective tissue and fat, insertional activity may be decreased and there is a palpable increase in the mechanical resistance to the insertion of the needle.
Abnormal "Spontaneous" Activity
With the muscle at rest, spontaneous activity of single muscle fibers and of motor units, known respectively as fibrillation potentials and fasciculation potentials, is abnormal. The two phenomena are often confused. Fibrillation is the spontaneous contraction of a single muscle fiber. It occurs when the muscle fiber has lost its nerve supply and is ordinarily not visible through the skin (but may be visible in the tongue). Fasciculation represents the spontaneous firing of an entire motor unit, causing contraction of a group of muscle fibers, and may be visible through the skin. The irregular firing of a number of motor units, seen as a rippling of the skin, is called myokymia.
Destruction of a motor neuron or interruption of its axon causes the distal part of the axon to degenerate, a process that takes several days or more. The muscle fibers formerly innervated by the branches of the dead axon—that is, the motor unit—are thereby disconnected from the nervous system. By mechanisms that are still obscure, the chemosensitive region of the sarcolemma at the motor endplate "spreads" after denervation to involve the entire surface of the muscle fiber. Then, 10 to 25 days after death of the axon, the denervated fibers develop spontaneous activity; each fiber contracts at its own rate and without relation to the activity of neighboring fibers. This spontaneous activity is associated with a random conglomeration of brief di- or triphasic fibrillation potentials (Fig. 45-7A) having a duration of 1 to 5 ms and rarely exceeding 300 μV in amplitude. When brief spontaneous fibrillation potentials of this sort are observed firing regularly at two or three different locations (outside the endplate zone) of a resting muscle, one may conclude that the fibers are denervated. Usually, fibrillation potentials discharge at an almost regular rate. In some early lesions (less than 6 to 8 weeks), irregularly firing fibrillation potentials may be observed.
A. Fibrillations and positive sharp waves. This spontaneous activity was recorded from a totally denervated muscle—no motor unit potentials were produced by attempts at voluntary contraction. The fibrillations (above arrow) are 1 to 2 ms in duration, 100 to 300 mV in amplitude, and largely negative (upward) in polarity following an initial positive deflection. A typical positive sharp wave is seen above the star. B. Fasciculation. This spontaneous motor unit potential was recorded from a patient with amyotrophic lateral sclerosis. It has a serrated configuration and it fired once every second or two. Calibrations: 5 ms (horizontal) and 200 mV in A; 1 μV in B (vertical).
Diseases such as poliomyelitis, which damage spinal motor neurons, or injuries of the peripheral nerves or anterior spinal roots, frequently produce only partial denervation of the involved muscles. In such muscles, one electrode placement may record fibrillation potentials at rest from denervated fibers and normal potentials during voluntary contraction from nearby healthy fibers. Fibrillation potentials continue until the muscle fiber is reinnervated by progressive proximal-distal regeneration of the interrupted nerve fiber or by the outgrowth of new axons from nearby healthy nerve fibers (collateral sprouting), or until the atrophied fibers degenerate and are replaced by connective tissue, a process that may take many years. In addition, fibrillation potentials may take the form of positive sharp waves, i.e., spontaneous, initially positive diphasic potentials of longer duration and slightly greater amplitude than the spikes of fibrillation potentials (see Fig. 45-7A).
Fibrillation potentials, while characteristic of neurogenic denervation, are not altogether specific; for example, they are seen in muscle diseases such as polymyositis and inclusion body myopathy which presumably damage the muscle fiber and make its membrane electrically unstable.
As stated earlier, a fasciculation is the spontaneous or involuntary contraction of a motor unit or part of a motor unit. Such contractions may cause a visible dimpling or twitching under the skin, although ordinarily they are of insufficient force to move a joint. Large distal fasciculations can briefly displace a finger or toe. They occur irregularly and infrequently, and prolonged inspection of the skin overlying a muscle may be necessary to detect them. The accompanying electrical form of an individual fasciculation potential is relatively constant. Typically, a fasciculation potential will have 3 to 5 phases (i.e., they are "polyphasic" as described later, in contrast to normal biphasic muscle activity), a duration of 5 to 15 ms (longer than normal but somewhat less in the facial muscles), and an amplitude of several millivolts (see Fig. 45-7B). Fasciculation potentials are evidence of motor nerve fiber irritability. Thus, the combination of fibrillations and fasciculations indicates active denervation combined with more chronic reinnervation of muscle.
The precise source of fasciculation is still contested. Forster and colleagues challenged the original belief that the discharge originated in anterior horn cells by demonstrating that fasciculations persisted after nerve block in ALS and ceased only with the appearance of fibrillation potentials, signifying wallerian (axonal) degeneration. These observations favored a distal site of generation. Other physiologic and pharmacologic evidence pointed to the first segment of the motor axon, or to the distal axon, or even to the motor point (the site of insertion of the nerve into muscle), involving elements of the postsynaptic muscle membrane (particularly in the case of benign fasciculations) as the source of the spontaneous electrical activity. It seems that several regions of the axon are capable of spontaneous impulse generation, depending on the underlying disease. Most of the diseases that produce fasciculations involve the anterior horn cell or the motor root, but more distal sites in the motor axon are spontaneously active in cases of nerve compression and polyneuropathy.
Occasional fasciculation potentials, particularly in the calves, hands, and periocular or paranasal muscles, occur in many normal persons. They can be almost constant for days or weeks on end, or even for years in some individuals, without weakness or wasting; therefore they need not be taken as evidence of disease ("benign fasciculations"). Certain quantitative features of fasciculations, such as brief duration and a consistent pattern and location of firing, favor benign over pathologic discharges. Shivering induced by low temperature and twitchings associated with low serum calcium levels are other forms of fasciculatory activity.
Fasciculation potentials occur with great frequency in chronic, slowly advancing, destructive diseases of the anterior horn cells, such as ALS and progressive spinal muscular atrophy. In these diseases, both voluntary MUPs and fasciculation potentials may be of long duration (more than 15 ms) and of increased amplitude, indicating chronic denervation and reinnervation. They are seen often in the early stages of poliomyelitis but only occasionally in the chronic phase of the disease, perhaps because the affected cells die rapidly. When anterior horn cells degenerate once again in older individuals who had had poliomyelitis (postpolio syndrome), fasciculations may return. Occasionally, they are seen in one muscle as a result of a compressive anterior root lesion, such as those caused by a protruded intervertebral disc. Large numbers of axons may be affected in this case, with the result that the fasciculations (or even cramps) may be even more prominent than with disease of anterior horn cells but they are restricted to the territory of innervation of the root or nerve. Fasciculation potentials in lesser numbers are also observed with chronic nerve entrapments, e.g., ulnar neuropathy at the elbow and other peripheral nerve lesions and some polyneuropathies. In all these cases, the damaged neuron or its axon seems to leave intact axons in a state of hyperirritability. The blocking of axon conduction by local anesthesia does not abolish fasciculations, but curare-like drugs do so.
Other Types of Spontaneous and Elicited Electrical Activity
(See Chaps. 48 and 50) These various phenomena, including some discussed above, can be classified according to their generating source. The muscle fibers themselves are the source of fibrillations, positive sharp waves, and complex repetitive discharges (CRDs). The motor axons produce fasciculation potentials, myokymic discharges, neuromyotonia, and cramp syndromes; and the central nervous system is the source of complex ensembles of continuous motor activity such as occur in the stiff man syndrome.
The common phenomenon of complex repetitive discharges, referred to in the past as bizarre high-frequency discharges, consists of repetitive spontaneous potentials created by numerous single muscle fibers that fire in near synchrony; there is often an erratic configuration and abrupt starting and stopping of the discharges. They are seen in some myopathies, in hypothyroidism, and in certain denervating disorders, and are a mark of chronicity (lesions more than 6 months old). High-frequency coupling of action potentials into doublets, triplets, or higher multiples of single units, indicating instability in repolarization of the nerve fiber to a muscle, occurs in tetany and in the early stages of myokymia.
Myokymia is a persistent quivering and rippling of muscles at rest (colloquially called "live flesh"). The EMG picture is distinctive. The spontaneously firing MUPs are called myokymic potentials, or discharges and consist of groups of repetitive discharging units, each firing at its own rate, quasirhythmically, usually several times per second, followed by a briefer period of silence. The small motor unit discharges may occur singly or as doublets, triplets, or multiplets.
The site of generation of this activity has been contested, possibly because it may arise from several sites along the motor nerve. The discharge corresponds to an alteration in the calcium concentration in the microenvironment of the axon. Spontaneous discharges arising in large myelinated fibers have been implicated in the genesis of myokymia; indeed, demyelinating polyneuropathies are among the conditions that give rise to this phenomenon. Myokymia is also caused by peripheral nerve hyperexcitability because of both K channel mutations and antibodies against the channels. This activity may be blocked by lidocaine infusion around the peripheral nerve and may be diminished by carbamazepine or phenytoin. Central forms of myokymia also occur, as in multiple sclerosis; the mechanism is similar to the peripheral form, namely irritation or demyelination of the motor nerve, but in its fascicular (central) course.
Focal and segmental myokymias differ in small ways from the generalized form of myokymia with regard to the timing and duration of the discharges. The focal types refer mainly to facial myokymia, seen most often in multiple sclerosis, Guillain-Barré syndrome, large cerebellopontine angle tumors, or compression of the facial nerve by a small aberrant blood vessel, but it may follow any peripheral nerve injury and regeneration. The EMG patterns are complex, either high-frequency (30- to 100-Hz) recurrent bursts or brief lower-frequency bursts. Segmental myokymia is a common occurrence in demyelination and in radiation injuries of the brachial plexus. The EMG bursts tend to be longer and less frequent than in generalized myokymia, and the interburst frequency is highly variable. The origin of these discharges is probably in the distal peripheral nerve, where activity of afferent fibers, possibly via ephaptic transmission, irregularly excites distal motor terminals. Segmental myokymia refers to similar activity in the distribution of one or more adjacent motor roots again, usually related in some way to demyelination. This activity persists during sleep and general anesthesia.
The phenomenon of generalized myotonia, or neuromyotonia denotes a failure of voluntary relaxation of muscle because of sustained firing of the muscle membrane (see Chaps. 48 and 50), is characterized by high-frequency repetitive discharges generally having a positive sharp waveform. These myotonic discharges wax and wane in amplitude and frequency, producing a "dive-bomber" sound on the audio monitor. The discharges can be elicited mechanically by percussion of the muscle or movement of the needle electrode and are also seen following voluntary contraction or electrical stimulation of the muscle via its motor nerve. The MUPs may appear normal during voluntary contraction, but they are not followed by the silence that normally occurs on relaxation; instead, there is a "prolonged afterdischarge" consisting of long trains of fibrillation-like potentials that may take as long as several minutes to subside (Fig. 45-8A). These EMG findings can be seen with any myotonic disorder. If the muscle is activated repeatedly at short intervals, the late discharge becomes briefer and briefer and eventually disappears (see Fig. 45-8B), as the patient becomes able to relax the exercised muscle ("warmup" effect).
A. Myotonia congenita (Thomsen disease). The five lines are a continuous record of activity in the biceps brachii following a tap on the tendon. The initial response is within normal limits, but it is followed by a prolonged burst of rapid activity, gradually subsiding over a period of many seconds or minutes. B. Same electrode placement as in A. Response to the fifth of a series of tendon taps. "Warmup" has occurred, and the characteristic prolonged myotonic activity is no longer evident. (See Chap. 48 for a description of the disease.)
In paradoxical myotonia the myotonia worsens after each of a succession of voluntary contractions. This is the converse of what happens in myotonia congenita (Thomsen disease). As shown by single-fiber EMG studies, myotonia is generated by single muscle fibers and the mechanism of the membrane instability, at least in some forms, seems to involve changes in the chloride conductance. These disorders are discussed in subsequent chapters.
The cramp-like contracture of McArdle disease and phosphofructokinase deficiency is associated with electrical silence of contracting muscle. This feature is an important part of the definition of true physiologic muscle contracture (as distinguished from chronic shortening of a muscle and its tendon which, strictly speaking, is a pseudocontracture).
In the syndrome of continuous muscle fiber activity or Isaacs syndrome (see Chap. 50), which is a generalized form of myokymia, the EMG discloses high-frequency (up to 300-Hz) repetitive discharges of varying waveforms. In the stiff man syndrome, painful muscle spasms and stiffness are generated by a spinal mechanism; the EMG potentials resemble normal motor units but are abnormal by virtue of continuous firing at rest (see Chap. 50).
Abnormalities in Amplitude, Duration, and Shape of Motor Unit Potentials
Figures 45-9 (schematically) and 45-10 depict the ways in which disease processes affect the motor unit and the appearance of the MUP in the EMG.
The colored muscle fibers are functional members of one motor unit, whose axon enters from the upper left and branches terminally to innervate the appropriate muscle fibers. The action potential produced by each motor unit is seen in the upper right: its duration is measured between the two vertical lines. The normal-appearing but uncolored fibers belong to other motor units. A. Hypothetical situation, with five muscle fibers in the active unit. B. In this myopathic unit, only two fibers remain active; the other three (shrunken) were affected by one of the primary muscle diseases. C. Four fibers that originally belonged to other motor units and had been denervated are now reinnervated by terminal sprouting from an undamaged axon. Both the motor unit and its action potential are now larger than normal. Note that only under these abnormal circumstances do fibers in the same unit lie next to one another.
Single voluntary motor unit potentials. A. Normal. B. Prolonged polyphasic potential seen with reinnervation. C. "Giant unit"—normally shaped but of much greater amplitude than normal. D. Brief, low-amplitude "myopathic" units. Calibrations: 5 ms (horizontal) and 1 mV in A and B; 5 mV in C; 100 mV in D (vertical).
Motor Unit Potentials in Denervation
Early in the course of denervation, many motor units with functional connections to the spinal cord are unaffected, and although the number of MUPs appearing during contraction is reduced, the configurations of the remaining ones are quite normal. In time, the remaining MUPs often increase in size and in electrical amplitude, perhaps two to three times normal, and become longer in duration and sometimes polyphasic (more than four phases).
Such large and sometimes giant polyphasic potentials (see Figs. 45-9C and 45-10B) arise from motor units containing more than the usual number of muscle fibers that are spread out over a greatly enlarged territory within the muscle. Presumably, new nerve twigs have sprouted from nodal points and terminals of undamaged axons and have reinnervated previously denervated muscle fibers, thus adding them to their own motor units. Soon after reinnervation, the MUPs generated will be low in amplitude, extremely prolonged, and polyphasic, findings that constitute a transitional configuration of early reinnervation. These amplitudes disappear as the motor unit is reestablished. Increased amplitude is usually associated with very chronic, proximal axon loss, for example, with remote poliomyelitis and chronic radiculopathy. These MUPs are to be differentiated from (1) polyphasic potentials of normal duration, which, as has been mentioned, make up as much as 10 percent of the total number of MUPs in normal muscle, and (2) polyphasic MUPs of short duration and low amplitude, which are characteristic of most myopathies and of myasthenia gravis and other disorders of neuromuscular transmission.
The Motor Unit Potential in Myopathy
As Fig. 45-9B shows, diseases such as polymyositis, the muscular dystrophies, and other myopathies that randomly destroy muscle fibers or render them nonfunctional, and obviously reduce the population of muscle fibers per motor unit. Therefore, when such a unit is activated, its potential is of lower voltage and shorter duration than normal, and it may also appear polyphasic as the compound MUP becomes fragmented into its constituent single-fiber potentials. Slowing of the propagated muscle fiber action potential in affected muscle fibers also contributes to the changes in the "myopathic" MUP. When most of the muscle fibers are affected, the MUPs are very small and of short duration and are recruited out of proportion to the tension generated, so-called early recruitment. Both types of alterations produce a characteristic high-pitched crackling sound from the audio monitor that has been likened to rain falling on a tin roof. They occur in all forms of chronic myopathies. Identical MUP changes are seen occasionally with other processes that cause disintegration of the motor unit, for example, early Guillain-Barré syndrome (because of conduction block along some of the terminal nerve fibers), and rarely with disorders of neuromuscular transmission (myasthenia gravis, other myasthenic syndromes), but they are most characteristic of primary muscle disease.
As mentioned earlier, fibrillation potentials, while typical of denervation, are sometimes seen in the myositides, and the rapidly progressive muscular dystrophies, perhaps because the muscle membrane has been made unstable; in the past the finding was attributed to isolation of a segment of the fiber from its nerve supply. In myasthenia gravis, where transmission of impulse fails at the neuromuscular junction, a single MUP may vary in amplitude during sustained weak contraction and some fibers cease to function. Electromyographic recordings of single muscle fibers belonging to the same motor unit disclose varying interpotential intervals on successive discharges; this phenomenon is called "jitter" and increases to the point of actual block, with deficits in neuromuscular transmission within a motor unit (see below and Chap. 50).
Abnormalities of the Interference Pattern
Diseases that reduce the population of functional motor neurons or axons within the peripheral nerve decrease the number of motor units that can be recruited in the affected muscles. The decreased number of motor units available for activation produces a low-amplitude interference pattern with only a few remaining units firing at a moderate to rapid rate. A severe reduction in the interference pattern may result in the recruitment of only a single unit (see Fig. 45-6B). Structural damage to nerve, as well as demyelinating block, can produce this pattern of reduced recruitment; indeed, a reduced recruitment pattern coupled with the absence of denervation indicates a conduction block.
If muscle power is reduced in diseases such as polymyositis or muscular dystrophy, in which individual muscle fibers are affected, there may be little or no reduction in the number of motor units available for recruitment until the process is far advanced and entire MUPs have been lost as a result of random loss of all their constituent muscle fibers. Nonetheless, each motor unit will consist of fewer muscle fibers than normal, so more motor units must be activated to reach a certain degree of force. A modest effort can thus produce a full interference pattern despite marked weakness (increased recruitment). Because fewer muscle fibers are firing, the amplitude of the pattern will be reduced from normal. This type of full, highly complex interference pattern of less-than-usual amplitude in the face of dramatic weakness is the hallmark of myopathy (see Fig. 45-6C).
Motor Unit Quantification
This experimental technique, developed by McComas and colleagues estimates the size of motor units and is thus exquisitely sensitive to changes of denervation and reinnervation. It is carried out by applying a weak stimulus to a motor nerve or motor point and increasing it gradually as the evoked muscle response is recorded. Each quantal increase in the compound-evoked response is presumed to be caused by the addition of a single motor unit. In reinnervated muscles, the additional units are reduced in number and are abnormally large. The technique is used mainly for the investigation of motor neuron disorders. When a normal number and configuration of motor units is found, it has been helpful in distinguishing benign fasciculations from those of serious diseases.
This is a special technique for the recording of single-muscle-fiber action potentials that has found utility in measuring muscle fiber density and in detecting so-called jitter in disease of the neuromuscular function, particularly myasthenia gravis. Fiber density is an index of the number and distribution of muscle fibers within a motor unit. Jitter is the variability of the interpotential interval of successive discharges of two single muscle fibers belonging to the same motor unit. This phenomenon is largely a result of the very slight variability of delay at the branch points in the distal axon and by synaptic delay at the neuromuscular junction, especially in myasthenia gravis where it has found its main clinical use. Fiber density and jitter may, however, also be increased in neuropathic disorders that cause denervation with reinnervation. Both are usually normal or only slightly increased in myopathic disorders.
Testing for jitter is carried out by having the patient voluntarily contract a muscle to the slightest degree possible so as to activate only one motor unit (requiring a great deal of cooperation by the patient) or by stimulating an intramuscular nerve twig (requiring great patience on the part of the examiner). The EMG needle is advanced until two muscle fibers from the same motor unit are recorded. If the oscilloscope sweep is triggered by the firing of the first fiber, a fluctuating latency of the second fiber potential can be seen on the screen as a movement (jitter) of the second peak. The degree of jitter can be quantitated by measuring the interval between the activation of the two muscle fibers (the result of slightly differing lengths of the terminal axons) from which a mean interpeak interval is determined. Approximately 20 fiber pairs are sampled, and an average of the mean consecutive intervals can be derived. In a muscle such as the extensor digitorum communis, the average variation should be no more than 34 ms. The acceptable average is lower for large proximal muscles. Also, in disease of the neuromuscular junction, one muscle fiber in a pair may fail to fire intermittently as a result of a blocking of conduction. Further details of this technique and its clinical applications are discussed by Stålberg and Trontelj.
Imaging of Muscle and Nerve
Imaging techniques—CT, MRI, and ultrasonography—enable one to measure muscle volume and to recognize qualitative changes in muscle structure (see review of DeVisser and Reimers). Such methods are finding some clinical and research use in the diagnosis of disorders of muscle and in gauging the effects of treatment. CT scans of dystrophic muscle show foci of decreased attenuation, representing masses of fat cells. The fatty masses spread gradually from multiple foci and eventually replace muscle fibers. The original shape of the muscle is retained; indeed, an enlarged weak muscle containing mostly fat confirms the clinical impression of pseudohypertrophy. In denervative atrophy, the muscles are obviously small and contain multiple punctate areas of decreased attenuation, which represent interstitial fat. Eventually, large portions of chronically denervated muscle may be replaced by fat. Blood, blood products, and calcium deposits are expressed by increased attenuation in CT. This may be helpful in the diagnosis of muscle trauma, myositis ossificans, and dermato- and polymyositis.
Fat and bone marrow have a high signal intensity in MR images, whereas fascia, ligaments, and cortical bone lack signal intensity. In T1-weighted images, normal muscle has a low signal and dystrophic muscle, a slightly increased signal; in T2-weighted images, dystrophic muscle has a slightly enhanced signal. Given its sensitivity to these dystrophic changes in muscle, MRI is particularly effective in determining the distribution of muscle involvement in a dystrophy. Transverse MR sections, for example, can help distinguish the topographic patterns of such disorders as the proximally predominant Becker dystrophy and the distally predominant Miyoshi myopathy, or subtle subtypes of Emery-Dreifuss dystrophy (Mercuri et al). Spectroscopic MRI in metabolically determined myopathies has the capacity to quantitate levels of selected biochemical constituents of muscle, including intracellular pH and levels of metabolic intermediates such as phosphocreatine. This technique is particularly effective in demonstrating subnormal generation of intracellular acidosis after a limb is exercised in disorders of glycogenolysis and of glycolysis. Some individuals with mitochondrial disease of muscle will demonstrate rapid depletion of energy supplies and profound delays in recovery that can be quantified and used as an end point for treatments.
New magnetic resonance techniques are also being developed that allow the imaging of nerves. This may be an aid in assessing traumatic nerve injury and in demonstrating neuromas, hypertrophy or atrophy of a nerve trunk, as well as neural tumors.