The median nerve is one of the main terminal nerves of the brachial plexus, formed by contributions from the lateral and medial cords (Figure C14-1). The lateral cord component, comprised of C6–C7 fibers, provides sensory fibers to the thumb and thenar eminence (C6), index finger (C6–C7), and middle finger (C7) and motor fibers to the proximal median innervated forearm muscles. The medial cord component, comprised of C8–T1 fibers, provides sensory fibers to the lateral half of the ring finger (C8) and motor fibers to the hand and distal median innervated forearm muscles.

Figure C14-1 Median nerve course and branches in the forearm and hand. 1 = the palmar cutaneous branch, and 2 = the terminal median sensory nerve.

(From Haymaker W, Woodhall B. Peripheral nerve injuries. Philadelphia, PA: WB Saunders, 1953, with permission.)

The median nerve descends with no branches in the arm. In the antecubital fossa, it passes between the two heads of the pronator teres and sends muscular branches to the pronator teres, flexor carpi radialis, flexor digitorum sublimis, and palmaris longus muscles. Soon after and while in the proximal forearm, the median nerve gives off the anterior interosseous nerve which is a pure motor nerve that innervates the flexor pollicis longus, medial head of the flexor digitorum profundus and the pronator quadratus muscles.

Right before entering the wrist, the median nerve gives off its first cutaneous branch, the palmar cutaneous branch, which runs subcutaneously (does not pass through the carpal tunnel) and innervates a small patch of skin over the base of the thumb and the thenar eminence (see Figure C14-1). Then, the main trunk of the median nerve, along with nine finger flexor tendons, enters the wrist through the carpal tunnel. The carpal bones form the floor and sides of the tunnel while the carpal transverse ligament, which is attached to the scaphoid, trapezoid, and hamate bones, forms its roof (Figure C14-2). The carpal tunnel cross-section is variable but is approximately 2.0 to 2.5 cm at its narrowest point in most individuals.

Figure C14-2 A cross-section of the wrist revealing the carpal tunnel and its contents.

(From Hollinshead WH. Anatomy for surgeons: the back and limbs, 3rd ed., vol 3. Philadelphia, PA: Harper and Row, 1982, with permission.)

Right after exiting the tunnel, the median nerve branches into motor and sensory branches. The motor branch innervates the first and second lumbricals and gives off the recurrent motor branch, which innervates the thenar muscles (abductor pollicis brevis, opponens pollicis, and half of the flexor pollicis brevis). The sensory branch divides into terminal digital sensory branches to innervate three and one-half fingers (thumb, index, middle finger, and lateral half of the ring finger) with the corresponding palm.

Clinical Features

Carpal tunnel syndrome (CTS) is the most common entrapment neuropathy. It is slightly more common in women and usually involves the dominant hand first. It is most prevalent after 50 years of age, but it may occur in younger patients, especially in association with pregnancy and certain occupations or medical conditions. Most cases of CTS are idiopathic, but many are associated with disorders that decrease the carpal tunnel space or increase the susceptibility of the nerve to pressure. Among the medical conditions with a high risk for CTS are pregnancy, diabetes mellitus, hypothyroidism, acromegaly, rheumatoid arthritis, sarcoidosis, and amyloidosis. Some patients have congenitally small carpal tunnels, while others have anomalous muscles, wrist fractures (Colles or carpal bone), or space occupying lesions (ganglia, lipoma, schwannoma). Occupational CTS, which has reached a near-epidemic level in the industrial world, is seen in patients whose jobs involve repetitive movements of the wrists and fingers. Although most cases of CTS are subacute or chronic in nature, it occasionally may be acute, such as after crush injury of the hand, fracture (Colles or carpal bone), or acute tenosynovitis.

The most common symptoms of CTS are episodic numbness and pain in the affected hand, mostly at night. A characteristic of CTS is frequent awakening at night because of hand paresthesias, hence the name, nocturnal acroparesthesia. Symptoms usually are relieved by shaking the affected hand. In addition, these symptoms are often exacerbated by certain activities, such as driving, holding a book, or knitting. There is wrist and hand pain, which may radiate proximally to the forearm and, less commonly, to the arm or shoulder. Weakness of the hand and loss of dexterity are common in more advanced cases.

Phalen sign (reproduction of paresthesias in a median nerve distribution after passive flexion of the hand at the wrist) is extremely sensitive, present in 80–90% of patients with CTS with rare false positives. Tinel sign (paresthesias in a median nerve distribution after percussion of the median nerve at the wrist) is less common sign, present in about 50% of patients and may be false positive. On examination, there is often relative hypesthesia throughout the median nerve distribution, particularly in the fingertips and excluding the skin over the thenar eminence. Sometimes, the sensory loss is more selective to one or two fingers. Fasciculations or myokymia of the thenar muscles is not uncommon. Atrophy of the thenar muscles with weakness of thumb abduction may be evident in advanced cases. Less common associated conditions include vasomotor skin changes and Raynaud phenomenon.

The differential diagnoses of CTS include:

1. Cervical radiculopathy, especially a C6 or C7 radiculopathy, which commonly results in numbness of the thumb, index finger, or middle finger. Sensory symptoms or signs above the wrist, unilateral radicular pain exacerbated by neck movements, segmental weakness in the arm and forearm, or myotatic reflex asymmetry favors a radiculopathy.

2. Neurogenic thoracic outlet syndrome, which frequently is associated with thenar atrophy. However, the pain and sensory manifestations in neurogenic thoracic outlet syndrome are in a C8/T1 distribution (ring and little fingers and medial aspect of the forearm).

3. Peripheral polyneuropathy, which may be associated with hand numbness. However, there are often sensory manifestations and motor weakness in the legs. Also, there usually is hyporeflexia or areflexia, especially at the ankles.

4. High median mononeuropathy, including the pronator syndrome, and compression at the ligament of Struthers in the distal arm. Both are rare syndromes and usually are associated with weakness of the long finger flexors.

5. Cervical myelopathy. In this condition, the numbness is not restricted to the median nerve, and there are frequently other pyramidal manifestations.

The treatment for CTS includes correcting the offending occupational factor or medical illness, wrist splinting at night, and the use of oral nonsteroidal anti-inflammatory agents or corticosteroids. Corticosteroid injection into the carpal tunnel area also is helpful to alleviate sensory symptoms and pain in patients with mild to moderate compression. Surgical decompression is indicated in patients with:

• Significant sensory loss.

• Significant thenar weakness and/or evidence of active denervation (fibrillation potentials).

• Acute median compression.

• Failure of conservative treatment.

Electrodiagnosis

Carpal tunnel syndrome (CTS) is the most common reason for referral to the EMG laboratory. Aims of the EDX studies are to confirm the diagnosis by assessing the status of the median sensory and motor fibers across the carpal tunnel, and to exclude other possible causes of the symptomatology, such as a cervical radiculopathy.

The main underlying pathophysiology in CTS early in the course is primarily paranodal demyelination. Hence, the electrophysiologic hallmark of CTS is focal slowing of conduction at the wrist, resulting in prolongation of the latencies of both motor and sensory fibers. In severe and advanced CTS, axonal loss dominates the picture.

Nerve Conduction Studies: Routine Studies

Historically, slowing of the median motor distal latency was the first described abnormality in CTS. Later, slowing of median sensory distal latencies was confirmed. These techniques, which include orthodromic or antidromic sensory conduction studies to the digits (particularly the index and middle fingers) and motor studies to the abductor pollicis brevis, are easy to perform and are reproducible. Delayed sensory distal latencies and/or delayed motor latencies usually confirm the clinical diagnosis of CTS in one half to two thirds of patients, with a high degree of sensitivity and specificity. It is important, however, to study neighboring nerves, such as the ulnar nerve, to establish that the abnormalities are restricted to the median nerve.

The routine median sensory and motor NCSs are the most widely used techniques in the diagnosis of CTS. The median sensory studies are usually more likely to reveal abnormalities before the motor studies. The distal latencies and amplitudes of the median sensory nerve action potentials (SNAP) are often lower than those of the compound muscle action potential (CMAP). Occasionally, the SNAPs are unelicitable whenever there is axonal loss (or occasionally significant conduction block caused by segmental demyelination). In some patients with CTS, the median motor proximal conduction velocities in the forearm may be mildly slowed without implying a proximal median neuropathy. In these cases, the median CMAP is usually low in amplitude, and the proximal slowing is best explained by a conduction block or axon loss of the fastest median motor fibers at the wrist.

Nerve Conduction Studies: Comparison Studies

It is now evident that the median motor and median sensory distal latencies are not sensitive enough in the diagnosis of CTS. Relying on these measurements only will fail to detect a significant number (up to one-third) of patients with mild CTS, particularly those with symptoms precipitated by certain hand activities (e.g., drilling, typing, etc.). In addition, as the syndrome has become well known to the medical community and to the general public, it has become common practice for EMG laboratories to test patients with very early symptoms of CTS. This has resulted in the design of several NCSs with higher sensitivity and specificity than the routine median sensory and motor studies (Table C14-1). These techniques rely on one or both of the following approaches:

• Measuring the conduction time of a short segment of the median nerve across the carpal tunnel. The rationale is that the slow-conducting segment of the median nerve in CTS usually is very short. If this short segment is included in a longer nerve segment, such as the wrist-to-index segment, a mild abnormality may become “diluted” by the normal conduction in the rest of the nerve, and results in an overall conduction time (i.e., latency) that may remain within normal limits in mild and early cases of CTS.

• Comparing the median distal latency to the distal latency of the ulnar nerve or radial nerve obtained from the same hand. With these internal comparison studies, median nerve conduction values are expressed in relative rather than absolute terms. Also, patients serve as their own controls because variables affecting nerve conduction studies, such as temperature and distance, are held constant, allowing direct comparison of distal latencies.

Table C14-1 Nerve Conduction Studies in the Diagnosis of Carpal Tunnel Syndrome

Table C14-2 lists and Figure C14-3 shows the most common internal comparison NCSs used in the diagnosis of CTS. Most of these procedures yield abnormal findings in symptomatic patients. However, the absence of a gold standard for the diagnosis of CTS precludes the determination of sensitivity, specificity, or predictive value for any of these tests. Applying these sensitive techniques in combination increases the diagnostic yield of EDX testing in the diagnosis of CTS to approximately 95%.

Table C14-2 Internal Comparison Nerve Conduction Studies in the Evaluation of Carpal Tunnel Syndrome (CTS)

Figure C14-3 Comparison studies in the diagnosis of carpal tunnel syndrome. S, stimulation site; R1 and R2, are active and reference recording sites, respectively.

(From Katirji B, Kaminski HJ, Preston DC et al., eds. Neuromuscular disorders in clinical practice. Boston, MA: Butterworth-Heinemann, 2002.)

Median-Ulnar Palmar Mixed Latency Difference “Palmar Study”

Trans-palmar mixed nerve conduction studies involve the elicitation of focal slowing of the median nerve between the palm and the wrist. Although abnormal absolute values were first considered to be a satisfactory indication in the diagnosis of CTS, comparison of median to ulnar latency with palmar stimulation has proved to be more sensitive and specific. The median nerve is stimulated in the midpalm between the second and third metacarpals, and the ulnar nerve is stimulated between the fourth and fifth metacarpals. Recording occurs at the wrist over the median and ulnar nerves 8 cm proximal to the midpalm cathode (Figure C14-3A). Extreme care must be given to measurements of nerve segments and latency analyses to prevent false-negative and false-positive results. Initial reports suggested that a median-ulnar palmar difference of greater than or equal to 0.2 ms is diagnostic of CTS; however, recent studies suggest that a difference of greater than or equal to 0.4 ms is needed for confirmation to prevent false-positive results (Figure C14-4A, A¹, and A²). A median-ulnar difference of 0.2 to 0.3 ms is considered borderline. It is estimated that palmar studies are abnormal in about 80% of symptomatic hands with CTS. In all published studies of CTS, palmar mixed nerve studies were far superior to the routine median sensory distal latency between the wrist and digit (index or middle finger).

Figure C14-4 Comparison studies in the diagnosis of carpal tunnel syndrome. The numbers shown are in ms and represent peak latencies. Palmar mixed study: (A) normal, (A¹) mildly abnormal and (A²) markedly abnormal. Digit 4 (median-ulnar) sensory study: (B) normal, (B¹) mildly abnormal and (B²) markedly abnormal. Second lumbrical/interosseous (median-ulnar) study: (C) normal, (C¹) mildly abnormal and (C²) markedly abnormal. Digit 1 sensory (median-radial) study: (D) normal, (D¹) mildly abnormal and (D²) markedly abnormal.

Median-Ulnar Sensory Latency Difference Between the Wrist and the Ring Finger

In this study, the median and ulnar sensory distal latencies recording the ring finger are compared. When the technique is performed antidromically at a 11 to 14 cm distance (Figure C14-3B), the difference in peak latencies of greater than or equal to 0.4 ms is abnormal (Figure C14-4, B, B¹, and B²). This test is abnormal 80 to 90% of patients with CTS. Its only disadvantage is that the median or ulnar SNAPs may be low in amplitude and difficult to evoke, due to the variable sensory innervation of the ring finger. When done orthodromically, the response in patients with CTS has a double hump, the first peak reflecting the volume-conducted ulnar fibers, and the second peak reflecting the slowed median fibers. It is preferable, in this situation, to also record over the ulnar nerve at the wrist to confirm that the first peak represents the ulnar fibers.

Median-Ulnar Motor Latency Difference Recording the Second Lumbrical/Interossei

This motor study compares the distal motor latency of the median nerve, recording the second lumbrical muscle, to the ulnar motor latency, recording the second intersossei. The recording surface electrode is placed just lateral to the midpoint of a line over the third metacarpal bone that connects the base of the middle finger to the middle of the distal wrist crease. The reference electrode is placed over the second proximal interphalangeal joint (Figure C14-3C). The lumbrical and interosseous CMAPs are recorded when the median and ulnar nerves are stimulated at the wrist, respectively. If a standard and equal distance of 8–10 cm is used for both nerves, a median-ulnar distal latency difference of greater than or equal to 0.6 ms is consistent with CTS (Figure C14-4C, C¹ and C²). This technique has several advantages: (1) the motor responses are generally more easily recorded than sensory responses; (2) this study is able to localize the lesion to the wrist in over 90% of cases of severe CTS resulting in absent routine median CMAPs and SNAPs; and (3) this study can still be easily performed in patients with CTS and advanced polyneuropathy associated with absent sensory responses in the hands.

Median-Radial Sensory Latency Difference Between the Wrist and the Thumb

This technique compares the distal sensory latency of the median nerve to the latency of the radial nerve while the thumb is recorded (Figure C14-3D). This can be performed antidromically or orthodromically, but can be difficult to perform because of movement or stimulus artifacts. At an 8–10 cm distance, a difference of greater than or equal to 0.4 ms of peak latencies is considered abnormal (Figure C14-4D, D¹, and D²). This study is abnormal in about 80% of hands with CTS.

Other tests have been advocated but that have not proved to be superior nor specific for CTS and have no localizing value. This includes assessment for fibrillation potentials, myokymia, or chronic neurogenic changes of the thenar muscles, and delay of median minimal F wave latencies when compared to the ulnar minimal F wave latencies.

Segmental Nerve Conduction Studies: “Inching” Studies

These studies, described by Kimura, consist of serial stimulations in 1 cm increments of the median nerve from the mid-palm to the distal forearm, recording antidromically from the index or middle finger (Figure C14-5). There usually is a latency change of 0.16 to 0.21 ms/cm between stimulation sites. In patients with CTS, there is an abrupt latency increase of greater than 0.4 to 0.5 ms across one or two adjoining segments (Figure C14-6). This most often occurs 2 to 4 cm distal to the distal wrist crease, the latter corresponding to the origin of the transverse carpal ligament. Although it is time consuming and subject to error (measurement error and volume conduction), the sensory study precisely localizes the lesion in more than 80% of symptomatic hands. Similar incremental study of the median nerve across the carpal tunnel, recording the abductor pollicis brevis, is also possible. However, unlike the sensory fibers, the median motor fibers are difficult to activate sequentially in 1 cm intervals because of the recurrent course of the motor branch to the thenar muscles; hence, the response is frequently contaminated by stimulus artifact because of the proximity of the recording electrode to the stimulating cathode.

Figure C14-5 The inching technique of Kimura. Twelve sites of stimulation in 1 cm increments along the length of the median nerve. The 0 level at the distal crease of the wrist corresponds to the origin of the transverse carpal ligament.

(From Kimura J. The carpal tunnel syndrome. Localization of conduction abnormalities within the distal segment of the median nerve. Brain 1979;102:619–635, with permission.)

Figure C14-6 The inching technique in a patient with carpal tunnel syndrome (CTS) (top is asymptomatic hand and bottom is symptomatic hand). The left side of the figure shows the results of 12 antidromically recorded sensory nerve action potentials (SNAPs), as shown in Figure C14-5. The right side of the figure graphs the successive time difference between traces.

(From Kimura J. The carpal tunnel syndrome. Localization of conduction abnormalities within the distal segment of the median nerve. Brain 1979;102:619–635, with permission.)

Needle EMG

Needle EMG examination in patients with CTS has two objectives:

1. To evaluate the thenar muscles for evidence of axonal loss or active (ongoing) denervation. Increased spontaneous activity and fibrillation potentials is consistent with ongoing axonal loss, and is long considered a strong indication for surgical intervention. Large MUAPs and impaired recruitment suggest chronic loss of axons. Myokymic discharges are seen sometimes in the thenar muscles of patients with CTS, particularly in those with chronic axonal loss.

2. To exclude other causes of hand numbness or pain, particularly cervical radiculopathy. It is estimated that one-quarter of patients with EDX evidence of CTS harbor a clinically significant cervical radiculopathy. Thus, a root search, particularly of muscles innervated by the C6 and C7 roots, should be sampled in all patients with CTS. Hence, needle EMG in patients with CTS should include, in addition to the abductor pollicis brevis, at least the pronator teres, biceps, and triceps muscles.

Special Situations

Severe Carpal Tunnel Syndrome

In severe CTS, which is more common in elderly patients, absent median sensory NCS, recording all digits, and median motor NCS, recording APB, is not uncommon. This renders EDX localization of the median nerve lesion to the wrist not possible, despite classic manifestations. Traditionally, these lesions were poorly localized, by needle EMG only, at or above the wrist. In these situations, the sensory and mixed internal comparison studies are equally absent. However, the second lumbrical-interosseous motor comparison study confirms the lesion at the wrist in more than 90% of cases by revealing that the median motor response recording second lumbrical is still evokable often with marked slowing of median distal latency (Figure C14-7). The relative preservation of the motor fibers to the lumbrical muscles as compared to the thenar muscles is best explained by the fascicular distribution of the median nerve fibers within the carpal tunnel: Fibers to the lumbrical muscles, which are more centrally located, tend to be relatively spared from axonal loss late in the course of the disease while the motor fibers to the thenar muscles, as well the sensory fibers, located in the periphery of the nerve, are destroyed earlier.

Figure C14-7 The second lumbrical-interossei study in a patient with severe CTS resulting in absent routine median SNAPs and median CMAP recording APB. Note that the median CMAP recording second lumbrical is recorded with very long latency (and relatively low amplitude) when compared to the ulnar.

Carpal Tunnel Syndrome With Peripheral Polyneuropathy

Carpal tunnel syndrome coexists not uncommonly in patients with underlying peripheral polyneuropathy, such as with diabetes. In mild axonal polyneuropathies, where the SNAPs in the hands are still recorded though low in amplitudes and delayed in latencies, sensory, motor, or mixed nerve internal comparison studies are useful in showing the preferential slowing of the median fibers at the wrist due to entrapment at the carpal tunnel. In instances of severe axonal polyneuropathy, where the SNAPs are often absent in the hands, the second lumbrical-interosseous motor comparison study is the most accurate by demonstrating preferential focal slowing of median motor fibers across the carpal tunnel. In all axonal polyneuropathy cases, selective large fiber axon loss may cause slowing; hence, borderline or minimal latency differences on internal comparison studies should be cautiously interpreted.

Carpal Tunnel Syndrome With Martin-Gruber Anastomosis

Carpal tunnel syndrome occasionally occurs in a patient with Martin-Gruber anastomosis, an anomalous connection between the median and the ulnar nerves in the forearm (see Chapter 2). When the anomalous fibers innervate the thenar muscles (usually adductor pollicis and deep head of flexor pollicis brevis), stimulation of the median nerve at the elbow activates the median nerve and the crossing ulnar fibers resulting in a large CMAP, with an initial positivity caused by volume conduction of action potential from ulnar thenar muscles to the median thenar muscles. This positive dip is not present at the wrist. Also, the median nerve conduction velocity in the forearm is spuriously fast in the presence of a CTS, since the CMAP onset represents different population of fibers at the wrist compared to the elbow (Figure C14-8). An accurate conduction velocity may be obtained by using specialized collision studies that abolish action potentials of the crossed fibers.