Nerve Conduction Studies

NCS are the first component of the electrodiagnostic examination. During NCS, a peripheral nerve is stimulated resulting in an electrical response generated directly by the nerve itself (as in a sensory response) or the muscle it innervates (as in a motor response). The duration and intensity of the stimulus are gradually increased until a maximal response is generated. These responses are recorded using surface electrodes placed over the skin and then analyzed. During each study, valuable information is produced regarding the number of functioning nerve fibers, the speed of conduction along those fibers, and their relative rates of conduction.

Three basic types of NCS are available: motor, sensory, and mixed (Fig. 13–1). Motor and sensory NCS are generally performed on every patient. Mixed NCS are typically used in the evaluation of specific disorders, such as carpal tunnel syndrome, and are of limited value in the evaluation of spine-related nerve pathology. NCS protocols vary depending on the diagnosis in question and can be tailored to help exclude other diagnoses in the differential. Most electrodiagnostic laboratories have a routine protocol, however, for a general study of the upper extremity (Table 13–1) and lower extremity (Table 13–2).

FIGURE 13–1 Three basic types of nerve conduction studies: motor, sensory, and mixed (S₁ and S₂ are stimulation sites, whereas R is the recording site; X overlies the shock artifact.)

(Modified from Isle M, Krauss G, Levin K, et al: Electromyography/Electroencephalography. Redford, WA, Spacelabs Medical, 1993, p 4.)

TABLE 13–1 Nerve Conduction Studies in the Upper Limb

Note: On each line, the nerve being studied is listed first, followed after the colon by the recording site and then, in parentheses, the root innervation (motor) or derivation (sensory). Underlined root provides major innervation.

TABLE 13–2 Nerve Conduction Studies in the Lower Limb

Note: On each line, the nerve being studied is listed first, followed after the colon by the recording site and then, in parentheses, the root innervation (motor) or derivation (sensory).

* M component of H response.

† Studies technically difficult to perform.

Motor Nerve Conduction Studies

For motor NCS, the recording electrode is placed over the muscle belly, and the reference electrode is affixed over the tendon. The nerve supplying that muscle is stimulated, and the resulting motor nerve response is a compound muscle action potential (CMAP), a biphasic waveform that represents summated muscle fiber action potentials (Fig. 13–2). In routine motor NCS, small muscles of the hand and feet serve as recording muscles, and the nerves supplying them are stimulated at two separate points along their course. For the upper extremity, the wrist (distal) and elbow (proximal) are used as stimulation sites, and for the lower extremity, the ankle (distal) and knee (proximal) are used as stimulation sites.

FIGURE 13–2 Various components of motor nerve conduction study. (The median nerve is being assessed.)

(Modified from Isley M, Krauss G, Levin K, et al: Electromyography/Electroencephalography. Redford, WA, Spacelabs Medical, 1993, p 40.)

Numerous parameters are assessed with each CMAP obtained, including amplitude, latency, and conduction velocity (Fig. 13–3). The CMAP amplitude represents the number of nerve fibers that responded to the stimulus and are capable of conducting impulses to the recorded muscle.^1,2 It is measured from baseline to negative peak (negative being up) and reported in millivolts. The latency is the time interval between the instant the nerve was stimulated and the onset of CMAPs; these are reported in milliseconds. The conduction velocity is the speed of transmission over the fastest conducting nerve fibers assessed and is reported in meters per second. Conduction velocities are calculated by dividing the distance traveled along a nerve segment (as determined by surface measurements) by the latency difference between the responses to proximal and distal stimulation. Normal conduction velocity in the upper limb is greater than 50 m/sec; in the lower limb, it is greater than 40 m/sec.

FIGURE 13–3 Compound muscle action potential (CMAP). Distal latency is measured from stimulus to onset of negative response. Amplitude is measured from baseline to negative peak.

Late Responses (H Responses and F Waves)

Two special studies, the H response and the F wave, are NCS used to measure the time in which nerve impulses travel proximally to the spinal cord along the peripheral nerve trunk and then back down the limb to the recorded muscle after distal stimulation of the nerve. Because the potentials seen with both of these techniques are much delayed after nerve stimulation compared with potentials seen with standard NCS, they are referred to as late responses.

The H response is the electrophysiologic correlate of the Achilles tendon reflex and is named after Hoffmann, who first described it in 1918. To obtain the H response, the tibial nerve is stimulated in the popliteal fossa using low voltage to activate sensory fibers (as opposed to motor fibers), which carry the nerve impulse proximally to the spinal cord (Fig. 13–4). The fibers synapse there with motor neuron cells to complete a monosynaptic reflex arc. The nerve impulse travels down the motor efferent nerve to the gastrocnemius where the recording electrode captures the response. Although the amplitude and the latency of the H response are analyzed, the amplitude is more reliable for diagnostic purposes in the authors’ laboratory.

FIGURE 13–4 Sensory nerve action potential (SNAP). Peak latency is measured to onset of negative phase. Amplitude is measured from baseline to negative peak.

The F wave was first described by Magladery and McDougall in 1950 and was named the F wave because it was first recorded from muscles in the foot. In contrast to H responses, F waves are not a component of a reflex arc because the nerve impulses recorded travel only along motor axons. F waves are produced when, after distal motor nerve stimulation, some of the impulses passing antidromically up the motor axons cause a few of the motor cell bodies in the anterior horns to backfire; the resulting nerve impulses travel back down the motor axons to produce submaximal muscle activations that are recorded several milliseconds after the initial CMAP as F waves. Several consecutive responses from the same muscle are elicited, and the shortest latency time usually is used for diagnosis. Also in contrast to H responses, F waves can be elicited with any of the standard motor NCS with consistency.

Needle Electrode Examination

NEE is the second and oldest component of the basic electrodiagnostic examination. During this procedure, a recording needle electrode is inserted into various muscles, and the electrical activity being generated in them is evaluated on a visual and audio display system via a differential amplifier. NEE records activity in muscle (1) at rest during needle insertion, (2) at rest without needle movement, and (3) during voluntary muscle activation.

Nerve Conduction Studies

Routine Studies

Axon loss occurs when the axon is disconnected from its cell body. The motor cell body (anterior horn cell) resides in the anterior zone of the spinal cord; the sensory cell body (DRG) resides outside the spinal cord, either within individual intervertebral foramina or within the spinal canal (intradural and intra-arachnoid) (Fig. 13–5). A disc protrusion causing severe compression of a motor and sensory nerve root within the spinal canal disconnects the anterior horn cell from its motor axon, but if the DRG is distal to the point of compression, the extra spinal sensory axons remain connected to their DRG and do not undergo degeneration (Fig. 13–6). In that setting, motor NCS show amplitude loss, but sensory NCS are normal despite marked clinical sensory impairment with few exceptions.

FIGURE 13–5 Standard lower limb H response. A, With minimal stimulus strength, only the H wave is elicited. B, As stimulation strength increases, the M wave appears and becomes progressively larger, while the H wave progressively loses amplitude.

FIGURE 13–6 A and B, Cross-sectional views of cervical region, showing relationship of dorsal root ganglia (asterisk) to surrounding structures (A) and usual site of disc herniation (arrow) (B). Preganglionic sensory root fibers usually are compromised (B).

One exception is seen with nerve root pathology that extends beyond the intraspinal canal. A mass lesion (e.g., meningioma) or infiltrative process (e.g., malignancy, inflammatory cause, or infection) that progresses distally along the nerve root to involve the DRG can result in decreased SNAP amplitudes. The other exception is when the DRG resides inside the intraspinal canal, proximal to the intervertebral foramina; this has been found to occur in the lumbosacral region. Based on cadaveric, radiographic, and magnetic resonance imaging (MRI) studies, 3% of L3 and L4 DRG are intraspinal, about 11% to 38% of L5 DRG are intraspinal, and 71% of S1 DRG are intraspinal.¹¹ As a result, root lesions in the lower spine, particularly lesions involving the L5 root, can affect the corresponding SNAP amplitude, which in the case of an L5 lesion is the superficial peroneal SNAP (see later). SNAP peak latency and nerve conduction velocity are never involved in radiculopathy, however.

The CMAP amplitude is the only portion of motor NCS that may be significantly affected in radiculopathy. Because it is a measure of the number of viable, conducting nerve fibers, the CMAP amplitude can be decreased with severe motor axon loss lesions. The ulnar CMAP amplitude would be reduced in a severe C8 radiculopathy. In chronic lesions, reinnervation changes such as collateral sprouting can contribute to the CMAP amplitude and may lead to normal or near-normal values over time.

In many cases, motor NCS remain relatively unaffected in radiculopathies for two reasons. First, most radiculopathies result in only partial nerve injuries. For the CMAP amplitude to be significantly reduced on motor NCS, about half of the motor axons within the peripheral nerve trunk need to be lost or injured. Second, the myotomes of the affected nerve root must be accessible to stimulation and recording. The ulnar-innervated hand muscles may be examined for a C8 radiculopathy, and the biceps and deltoids are available for assessing a C5 radiculopathy. Muscles innervated by C6 and C7 nerve roots cannot be reliably examined with routine motor NCS, however, owing to technical factors and overlap in innervation.

Needle Electrode Examination

Because NCS and the late responses generally are normal with isolated root lesions (except for the H response with S1 radiculopathies), NEE usually is the sole component of the electrodiagnostic examination that is beneficial in detecting a radiculopathy. The diagnosis depends on finding abnormalities on NEE in a root, or myotome (all the muscles that receive innervation from a single spinal cord segment or root). These abnormalities include insertional positive sharp waves, fibrillation potentials, a reduced or neurogenic recruitment of motor units, and changes in the motor unit morphology (e.g., increased duration, amplitude, and polyphasia).

The most widely used criterion for diagnosing radiculopathies by NEE is that abnormalities should be found in two, and preferably more, limb muscles innervated by the same root but different peripheral nerves. In addition, muscles in the limb not innervated by the damaged root, but by the roots contiguous to it, should appear normal. A patient with a C7 radiculopathy should have fibrillation potentials or other signs of denervation in the triceps (radial nerve) and pronator teres (medial nerve), but not the abductor digiti minimi or deltoid muscles. Needle electromyography (EMG) not only should be tailored to the clinical question and the patient’s symptoms, but it also should include a comprehensive survey of a sufficient number of muscles (proximal and distal when possible) to make a reliable diagnosis of a radiculopathy.

Numerous myotome charts derived from radiographic, cadaveric, and electrodiagnostic studies have been established to help guide the electrodiagnostic physician in choosing the best muscles to examine for each patient (Figs. 13-7 through 13-9). A radiculopathy screen in the authors’ laboratory consists of an examination of at least seven muscles, including the paraspinals to help with localization in the upper extremity (Table 13–3) and lower extremity (Table 13–4). The presence of fibrillation potentials in the paraspinals is typically indicative of an axon loss lesion localized to or near the intraspinal canal, excluding the possibility of a plexopathy or more distal lesion. Paraspinal fibrillation potentials are most valuable for the support of radiculopathy when they are present at only one or two contiguous segmental levels and absent at levels above, below, and contralaterally.

FIGURE 13–7 Traditional myotome chart.

(From Wilbourn AJ, Aminoff MF: Radiculopathies. In Brown WF, Bolton CF [eds]: Clinical Electromyography, 2nd ed. Boston, Butterworth-Heinemnann, 1993, p 192.)

FIGURE 13–8 Lower limb myotome chart. Needle electrode examination results grouped by surgically defined root level of involvement. Numbers in the left column represent patients. AD, abductor digiti quinti; AH, abductor hallucis; AL, adductor longus; BFLH, biceps femoris long head; BFSH, biceps femoris short head; EDB, extensor digitorum brevis; EHL, extensor hallucis longus; GM, gluteus maximus; GMED, gluteus medius; IL, iliacus; LG, lateral gastrocnemius; MG, medial gastrocnemius; PL, peroneus longus; PSP, paraspinal; PT, posterior tibialis/flexor digitorum longus; RF, rectus femoris; ST, semitendinosus; TA, tibialis anterior; TFL, tensor fascia lata; VL, vastus lateralis; VM, vastus medialis.

(From Tsao BE, Levin KH, Bodner RA: Comparison of surgical and electrodiagnostic findings in single root lumbosacral radiculopathies. Muscle Nerve 27:61, 2003.)

FIGURE 13–9 Upper limb myotome chart. Needle electrode examination results grouped by surgically defined root level of involvement. Blue circle, positive waves or fibrillation, with or without neurogenic recruitment and motor unit changes; red circle, neurogenic recruitment changes only; green circle, normal examination. ADM, abductor digiti minimi; ANC, anconeus; APB, abductor pollicis brevis; BIC, biceps; BRAC, brachioradialis; DEL, deltoid; EDC, extensor digitorum communis; EIP, extensor indicis proprius; FCR, flexor carpi radialis; FDI, first dorsal interosseus; FPL, flexor pollicis longus; INF, infraspinatus; SUP, supraspinatus; PSP, paraspinal; PT, pronator teres; TRIC, triceps.

(From Levin KH, Maggiano HJ, Wilbourn AJ: Cervical radiculopathies: Comparison of surgical and EMG localization of single-root lesions. Neurology 46:1023, 1996.)

TABLE 13–3 Screening Needle Electrode Examination for the Arm

TABLE 13–4 Screening Needle Electrode Examination for the Leg

Many limitations can reduce the value of the paraspinal examination. First, there is overlapping innervation of most paraspinals, which prevents accurate localization of fibrillation potentials to one specific segment or root. Second, even in proven radiculopathies, fibrillation potentials may be absent owing to reinnervation or sampling error. Third, paraspinal fibrillation potentials may be seen in diabetic patients, in patients with a prior history of spine surgery, or in some asymptomatic elderly patients. Finally, paraspinal denervation is not specific to radiculopathy and is seen in other disorders, including diseases of the muscle (e.g., inflammatory myopathy) and the anterior horn cell (e.g., amyotrophic lateral sclerosis). Nonetheless, NEE of the paraspinal muscles is an integral portion of the electrodiagnostic examination and should be routinely performed in all patients with suspected nerve root disease.

The timing of needle EMG is also crucial. Fibrillation potentials do not appear in a denervated muscle until 2 to 3 weeks after the onset of the initial injury and in some patients may require 4 to 6 weeks to develop.¹¹ Consequently, the findings on NEE performed earlier than 3 weeks after onset of a radiculopathy are likely to be false-negative or, at best, indeterminate, even if subsequently they would be positive for a root lesion. It is optimal to wait at least 3 weeks after the onset of symptoms before performing NEE. Guidelines that help the clinician decide the best timing of a study to obtain maximal information are provided in Table 13–5. These are based on the neurophysiologic concepts of axon loss as described in Table 13–6.

TABLE 13–5 Appropriate Timing of the Electrodiagnostic Examination

NCS, nerve conduction studies; NEE, needle electrode examination.

TABLE 13–6 Timing of Nerve Pathology: Neurophysiologic Concepts

NEE, needle electrode examination.

Determining Duration of Radiculopathy: Acute versus Chronic

Information regarding the duration of a radiculopathy is often derived by findings on NEE. Whenever evidence of an isolated compressive root disorder of recent onset is detected on the electrodiagnostic examination, the typical combination of findings is as follows: (1) motor NCS are normal (unless the degree of axon loss is severe); (2) sensory NCS are normal; (3) with S1 root involvement, the H response usually is abnormal; and (4) NEE discloses fibrillation potentials in several muscles that are innervated by the compromised root unaccompanied by changes in the size and configuration of the MUP.

In contrast, when chronic neurogenic MUP changes (polyphasic configuration with increased duration and amplitude) are the prominent finding on NEE with only a few fibrillation potentials, the lesion is likely to be chronic. When the chronic neurogenic MUP changes are limited to distal muscles within a myotome in the absence of fibrillation potentials, the radiculopathy is likely to be static and remote.

Finally, when fibrillation potentials and chronic neurogenic MUP changes are found in a myotome distribution, the diagnostic possibilities include a chronic, progressive radiculopathy or an acute root lesion superimposed on a remote lesion. The latter possibility is the more likely choice if fibrillation potentials are found in proximal muscles (e.g., the glutei and hamstrings with L5 or S1 root lesions), in addition to more distal muscles in the same myotome.^7–9

When the previous factors are considered, it is apparent that whenever the classic NEE presentation of a radiculopathy is encountered—fibrillation potentials in most or all of the muscles constituting the myotome—the root lesion in question usually is of more recent onset, and motor root axon loss has been substantial. Whenever other circumstances prevail, as is far more commonly the case, fibrillation potentials usually are found in only some, if any, of the muscles of the myotome. They are typically seen in the more distal muscles. Fibrillation potentials generally are important only if they are present; their absence in any specific muscle does not exclude the diagnosis.^8,9

Determining Severity of Radiculopathy

The severity of a nerve root lesion is based on motor NCS and NEE. The degree of reduced MUP recruitment seen on NEE correlates with the degree of muscle weakness and, in combination with the CMAP amplitude reduction (in muscles that can be assessed with NCS), the degree of axon loss. The amount of fibrillation potentials seen in a muscle is a subjective measure and does not correlate as well with the degree of axon loss.

Cervical Radiculopathy

Lesions of the cervical nerve roots account for 36% of all radiculopathies.⁹ In clinical and radiographic studies, the most common root affected is at the C7 level (70% of the time) followed by C6 (19% to 25%), C8 (4% to 10%), and C5 (2%).^9,16–18 The electrodiagnostic examination presentations with C5 radiculopathies are typically manifested as abnormalities in the spinati, deltoid, biceps, and brachioradialis muscles. NCS are typically unhelpful because proximal muscles are not assessed during routine studies, although the biceps and deltoid muscles are amenable to NCS and may show reduced CMAP amplitudes when axon loss is sufficiently severe.

C6 radiculopathies do not have a single, discrete appearance. Rather, they have two very different ones, which imitate those of C5 and C7 root lesions. Manifestations of C5 root lesions may also be seen with some C6 radiculopathies.¹⁹ C7 lesions are diagnosed by the presence of abnormalities in some muscles innervated by radial and median nerves: the triceps and anconeus (radial) and the pronator teres and flexor carpi radialis (median). As stated before, NEE abnormalities sometimes are seen in the same combination of upper limb muscles with C6 root lesions as well.

In contrast, C8 radiculopathies have a very characteristic electrodiagnostic presentation, manifesting as abnormalities in ulnar-innervated muscles, the extensor indicis proprius, and the flexor pollicis longus.¹⁹ Nonetheless, they can sometimes be confused with combined axon loss lesions of the posterior interosseous nerve and the ulnar nerve whenever the ipsilateral ulnar SNAP is of low amplitude or cannot be elicited (e.g., because of advanced age or a coexisting polyneuropathy). For uncertain reasons, the axon loss that occurs with many C8 radiculopathies is exceptionally severe, so much so that the CMAPs recorded from the ulnar nerve–innervated hand muscles, particularly the hypothenar, are low in amplitude. Some of these patients never regain normal hand strength.

Thoracic Radiculopathy

Radiculopathies in this region are difficult to assess by electrodiagnostic examination because there are relatively few muscles in each myotome, and only some of them can be sampled. With suspected thoracic radiculopathies, only the paraspinal and abdominal muscles are sampled routinely; the intercostal muscles are typically not studied for fear of entering the pleural space. Generally, if NEE abnormalities are seen, no attempt is made to identify a specific root lesion. Instead, the localization is limited to upper thoracic, mid-thoracic, or lower thoracic root involvement. Most patients found to have thoracic radiculopathies have diabetes mellitus, and the pathology is probably root infarction or ischemia rather than compression. In any case, these radiculopathies often produce very severe axon loss and frequently apparently involve two or more adjacent roots.^9,20,21 T1 radiculopathies are quite rare and typically produce changes only in the lateral thenar muscles.²²

Lumbosacral Radiculopathy

Nerve root lesions are most commonly seen in the lumbosacral spine: More than two thirds of all radiculopathies occur in this region.⁷ In contrast to lesions involving the cervical roots, it is difficult sometimes to localize lumbosacral radiculopathies accurately to a vertebral level with the electrodiagnostic examination. This difficulty is primarily due to anatomic reasons. Given their long intraspinal course, lumbosacral nerve roots may be injured anywhere along their tract from the T12-L1 vertebral level where they are formed, down through the canal into the cauda equina, and the site where they exit from their respective foramina. The L5 nerve root can be compressed by a central disc herniation at the L3-4 level, a posterolateral disc herniation at the L4-5 level, or foraminal stenosis at the L5-S1 level. Additionally, when nerves are affected at the level of the cauda equina where the fibers are compact, a single lesion in this location can result in injury to multiple roots bilaterally. It is important to perform comparison NEE of the contralateral limb when any abnormalities are seen to exclude the possibility of subclinical nerve root involvement.

L2, L3, and L4 radiculopathies are generally considered together because of the myotome overlap of the thigh muscles and the paucity of muscles that are innervated solely by one individual nerve root. Localization of an L2 root lesion is difficult because only the iliacus muscle may show abnormalities on NEE. Lesions at these levels typically produce denervation changes in the quadriceps, thigh adductors, and iliacus. With L4 lesions, abnormalities may also be seen in the tibialis anterior occasionally.

The most common lumbosacral radiculopathies involve the L5 and S1 roots. Lesions of these two roots are most amenable to recognition on electrodiagnostic examination. In addition, the L5 nerve root is the most common single radiculopathy seen.³ L5 radiculopathies produce abnormalities in the tibialis anterior, flexor digitorum longus, and posterior tibialis in greater than 75% of surgically proven cases.²³ In a more recent study, 100% of patients with L5 radiculopathies, which were also surgically proven, showed abnormalities in the peroneus longus and tensor fascia lata.²⁴ Changes may also be seen in the extensor digitorum brevis, gluteus medius, and semitendinosis.

An exception to the rule that SNAPs are not affected in radiculopathies has been found to occur with some L5 root lesions. As stated before, SNAPs are typically spared in radiculopathies because the lesion is situated proximal to sensory cell bodies (DRG), which lie in the intervertebral foramina outside of the intraspinal canal. At the level of the lumbosacral spine, the DRG is sometimes found proximal to the intervertebral foramina, however, within the intraspinal canal, leaving them vulnerable to injury from a herniated disc or other degenerative spine condition. Based on cadaveric, radiographic, and MRI studies, 3% of L3 and L4 DRG are intraspinal, 11% to 38% of L5 DRG are intraspinal, and up to 71% of S1 DRG are intraspinal.^24–26

The L5 nerve root in some cases may be affected distal to the DRG, resulting in an abnormal superficial peroneal SNAP. In one retrospective study, six patients with clinical and radiographic evidence of an L5 radiculopathy were found to have reduced amplitude of the ipsilateral superficial peroneal SNAP along with denervation changes in the L5 myotome.²⁷ This condition has not been found with S1 nerve root lesions, in which the sural SNAP remains normal despite the higher percentage of DRG located within the intraspinal canal.

S1 radiculopathies are the second most common root lesion encountered. Needle EMG may show abnormalities in the gastrocnemii, abductor hallucis, abductor digit quinti pedis, glutei, and biceps femoris short head. In addition, the H response is either absent or reduced in amplitude.

Cauda Equina Syndrome

Multiple lumbosacral radiculopathies are encountered with some frequency. Typically, the involvement is bilateral and often asymmetrical.^10,20 Most of these lesions are attributable to midline lumbar disc protrusions or lumbar canal stenosis. Characteristically, S1 and S2 roots, being the most medial of the roots supplying the lower limbs, are affected. In many patients, more extensive lumbosacral root involvement occurs; a common combination is bilateral S1 and S2 root compromise accompanied by unilateral or bilateral L5 root involvement.

The electrodiagnostic findings most commonly seen consist of a mixture of low-amplitude CMAPs and normal SNAPs on NCS, along with fibrillation potentials and MUP dropout on needle EMG (Fig. 13–10). On NEE, the abnormalities often are more severe in muscles located distal to the knees. With some substantial lesions of recent onset, they are just as prominent in the more proximal muscles. Whenever the disorder is subacute or chronic, fibrillation potentials usually are accompanied by chronic neurogenic MUP changes. Low lumbar or high sacral paraspinal fibrillation potentials often are found bilaterally with more acute lesions, but are undetectable with many chronic ones. Typically, the H responses cannot be elicited, and even the M components of the H responses, recorded from the gastrocnemius/soleus muscles, are quite low in amplitude.

FIGURE 13–10 Coronal view of inferior spinal cord, cauda equina, and surrounding structures. Dorsal root ganglia are located in intervertebral foramina so that all the sensory fibers composing the cauda equina are “preganglionic.“ Axon loss lesion of cauda equina generally has no effect on lower limb sensory nerve conduction studies, regardless of its severity.

Lumbar Canal Stenosis

Lumbar canal stenosis has no single characteristic electrodiagnostic presentation. Rather, the findings are extremely variable, depending on the degree of axon loss affecting the lumbosacral motor roots. At one extreme are patients who experience only intermittent, short-lived symptoms that often can be relieved completely by various maneuvers (e.g., sitting, flexing at the waist). In these patients, the electrodiagnostic examination often is completely normal. At the opposite end of the spectrum are patients who have substantial cauda equina lesions with severe, bilateral, fixed motor and sensory deficits. These lesions produce the electrodiagnostic presentation described previously.

Between these two extremes are numerous different electrodiagnostic patterns: (1) two or more radiculopathies, far more often bilateral than strictly unilateral; (2) a single radiculopathy, typically S1, that is sometimes detected in the less symptomatic or asymptomatic limb; (3) unilateral or bilateral absent H responses alone; (4) NEE changes restricted to just one or two limb muscles, most commonly those innervated by the S1 roots; or (5) fibrillation potentials limited to the paraspinal muscles.^9,19

Myelopathy

The effect that a focal myelopathy has on the electrodiagnostic examination depends principally on whether the anterior horn cells or their existing fibers are compromised at the level of the lesion. If only the descending corticospinal tracts of the spinal cord are affected (upper motor neuron lesion), the only abnormality found on the electrodiagnostic examination concerns the MUP firing pattern of muscles receiving innervation from spinal cord segments caudal to the lesion; in these muscles, the MUPs show no or incomplete activation (i.e., they fire in decreased numbers at a slow to moderate rate).

In contrast, if the anterior horn cells or the intramedullary fibers derived from them are involved, the electrodiagnostic findings are those of a focal intraspinal canal lesion that are characteristically bilateral, but often asymmetric. How prominent the electrodiagnostic changes are with such focal disorders depends mainly on where the lesions are located along the spinal cord. Lesions situated in C5-T1 segments and L4-S2 segments produce substantial abnormalities on motor NCS and NEE and generally are readily recognized as intraspinal canal lesions. All such disorders result in low-amplitude CMAPs or CMAPs that cannot be elicited and normal SNAPs on NCS, accompanied by fibrillation potentials, MUP dropout, and, depending on lesion duration, chronic neurogenic MUP changes on needle EMG. Conversely, lesions involving T2 through L3 segments result only in NEE changes (i.e., motor NCS using various limb muscles as recording sites are normal). Finally, lesions involving the upper cervical cord segments (C1-4) have essentially no electrodiagnostic manifestations because that region of the spinal cord cannot be assessed.²⁷

Postlaminectomy Electrodiagnostic Findings

Electrodiagnostic examinations are obtained frequently on patients who have undergone neck or back surgery. The specific diagnostic benefit derived from such assessments varies considerably, depending on the reason for referral and the time that has elapsed since operation. Overall, such postoperative studies are of limited value, however, unless they are obtained after very remote surgery to diagnose a recent-onset lesion. In the immediate postoperative period (first 10 to 14 days after surgery), the electrodiagnostic examination can reveal preexisting abnormalities because any NEE changes observed during that period, with the exception of a reduced MUP recruitment, are caused by a lesion that predated the operation.

During the early postoperative period (3 weeks to 3 to 4 months after surgery), the electrodiagnostic examination is of considerable benefit in assessing patients with postoperative weakness, principally because a normal CMAP amplitude recorded from a weak muscle (e.g., the tibialis anterior, resulting in footdrop) 7 or more days after onset of symptoms virtually excludes motor axon loss as the cause. The remaining possibilities include a proximal conduction block (neurapraxia), an upper motor neuron lesion, or hysteria or malingering. In the rare patient who develops nonorganic weakness postoperatively, the electrodiagnostic examination can prove that the symptoms are not the result of significant nerve fiber damage.

The electrodiagnostic examination usually cannot answer reliably the early postoperative question: “Was the root adequately decompressed?” Axon loss features of radiculopathy persist for weeks to months or indefinitely. Even an electrodiagnostic examination performed 2 to 3 months postoperatively is not likely to show significant improvement compared with a preoperative study. An exception is radiculopathy resulting from conduction block at the root level, which may resolve rapidly after the pressure is relieved. With an S1 radiculopathy, an H response that could not be elicited preoperatively may reappear in the early postoperative period. Similarly, on NEE, reduced MUP recruitment (and clinical weakness) could resolve rapidly in the affected muscles postoperatively.

An electrodiagnostic examination can be valuable in identifying root damage as the cause of new or worsening weakness in the postoperative period. The extent, amount, and distribution of fibrillation potentials provide information when compared with the preoperative study.

Cervical Root Avulsion

Root avulsions, which are usually restricted to the cervical region, differ from the typical single compressive radiculopathy principally in the degree of axon loss that results. Because the entire motor supply from one or both roots innervating the particular muscle has been disrupted, that muscle is severely or totally denervated. If it is used as a recorded muscle during motor NCS, the CMAP obtained is of very low amplitude, if it can be elicited. Similarly, during needle EMG of that muscle, fibrillation potentials are abundant, and MUPs are either absent or, if present, quite sparse and show reduced recruitment. Sensory NCS responses derived from the same roots are normal because the sensory roots are interrupted proximal to their DRG. Fibrillation potentials are not found in the appropriate paraspinal muscles in patients with cervical avulsion injuries, so their absence does not exclude this diagnosis.⁹