Median Neuropathy at the Wrist

Median-Versus-Ulnar Palm-to-Wrist Mixed Nerve Studies

This technique takes advantage of measuring the mixed nerve potential. Mixed nerve potentials consist of both motor and sensory fibers. The sensory fibers in the mixed nerve potential carry both cutaneous sensory fibers, which are measured in routine sensory studies, as well as muscle sensory fibers, which are not measured in routine sensory studies. This is important because the muscle sensory fibers include the Ia afferents from muscle spindles, which are the largest and fastest-conducting fibers and hence have the greatest quantity of myelin sheath. These fibers are very susceptible to demyelination, the primary pathology in CTS. The mixed nerve study also takes advantage of conducting over a very short distance of 8 cm. Because such a short distance is used, most of the conduction time is computed over the area of pathology. Only a short length of normal nerve is included that potentially could dilute any slowing present across the carpal tunnel.

The technique is performed by stimulating the median nerve in the palm, recording the median nerve at the wrist, and comparing it with the ulnar nerve stimulated in the palm and recorded over the ulnar nerve at the wrist ( Fig. 20.8 ). Each nerve is stimulated supramaximally in the palm at a distance of 8 cm from its respective recording electrodes. The median nerve is stimulated in the palm on a line connecting the median nerve in the middle of the wrist to the web space between the index and middle fingers. The ulnar nerve is stimulated in the palm on a line connecting the ulnar nerve at the medial wrist (lateral to the flexor carpi ulnaris tendon) to the web space between the ring and little fingers. Supramaximal responses are obtained for each nerve, and the difference between the onset or peak latencies is calculated.

Median-Versus-Ulnar Digit 4 Sensory Latencies

The technique of comparing median-versus-ulnar digit 4 sensory latencies takes advantage of the fact that, in most individuals, the sensory innervation to the fourth digit (ring finger) is split, with the lateral half innervated by the median nerve and the medial half innervated by the ulnar nerve ( Fig. 20.9 ). Thus, if identical distances are used, the latencies stimulating each nerve can be directly compared. The antidromic technique is performed by stimulating the median and ulnar nerves at the wrist, one at a time, with recording ring electrodes placed over digit 4 (G1 over the metacarpophalangeal joint and G2 over the distal interphalangeal joint). Identical distances must be used for both (range 11–13 cm). Supramaximal responses are obtained, and the difference between the median and ulnar onset or peak latencies is recorded. The study also can be done orthodromically, stimulating with the ring electrodes over digit 4 as just described and recording the median and ulnar nerves at the wrist at identical distances. We do not recommend the latter method because, with orthodromic stimulation at digit 4, co-stimulation of the median and ulnar nerves cannot be avoided, and spread of the potential from the adjacent nerve may contaminate the recorded SNAP at the wrist.

Median Second Lumbrical-Versus-Ulnar Interossei Distal Motor Latencies

The technique of comparing the second lumbrical (2L)-versus-interosseous distal motor latencies takes advantage of two facts: (1) motor fibers are easy to record and more resistant to compression than sensory fibers, and (2) the median 2L muscle lies just above the ulnar INT. In some cases of generalized polyneuropathy with superimposed CTS, the SNAPs and mixed nerve potentials may be absent. In severe cases, the routine median CMAP recording the APB may also be absent, whereas the motor fibers to the second lumbrical and ulnar INT are still recordable.

CMAPs from both the median-innervated 2L and the ulnar-innervated INT can easily be recorded by placing an active electrode (G1) slightly lateral and distal to the midpoint of the third metacarpal, with the reference electrode over the proximal interphalangeal joint of the second digit, and stimulating the median and ulnar nerves at the wrist, respectively ( Fig. 20.10 ). The motor point to the 2L is identified when the active recording electrode has been placed such that stimulation of the median nerve at the wrist elicits a waveform with the fastest rise time and an initial negative deflection. Because the 2L cannot be seen or palpated, moving the active electrode slightly may be necessary to ensure the electrode is optimally placed. In some individuals, if the sensitivity is increased, a small mixed nerve potential will be seen slightly before the onset of the 2L CMAP. This is a normal finding, especially in younger patients. If this small mixed nerve potential is present, the latency should be measured from the onset of the 2L CMAP, not from the onset of the mixed nerve potential. The ulnar nerve is then stimulated supramaximally at the wrist, at the same distance, leaving the recording electrodes in place . A CMAP from the underlying ulnar INT muscles will be easily elicited. The ulnar CMAP is generally larger than the median CMAP. Identical distances (range 8–10 cm) must be used to compare the difference between the distal latencies.

The normal values for the three median-versus-ulnar comparison studies are given in Table 20.3 . In our laboratory, the palmar mixed nerve peak latency difference is the most sensitive study, followed closely by the digit 4 sensory and 2L-INT motor studies. However, there is a very high degree of correlation among the results of the three studies. In one comparison study, two of the three studies yielded abnormal results in 97% of all patients with mild CTS. In a patient in whom only one of the median-versus-ulnar comparison studies is abnormal, one should be hesitant to make a definite electrodiagnosis of CTS (see Chapter 9 ).

Wrist-to-Palm Versus Palm-to-Digit Sensory Conduction Velocity (Segmental Sensory Conduction Studies Across the Wrist)

Another extremely sensitive internal comparison study includes the wrist-to-palm versus palm-to-digit sensory conduction velocity (segmental sensory conduction studies across the wrist). This test is more technically challenging than the others listed previously but is extremely sensitive in detecting CTS. This technique compares the sensory conduction velocity along the median nerve at two segments of identical distance: the wrist-to-palm segment and the palm-to-digit segment. Digit 3 is the preferred finger to record from due to its longer length. The recording electrodes (G1, G2) are placed at the proximal and the distal interphalangeal joints, respectively. Placing G1 at the proximal interphalangeal instead of the metacarpal-phalangeal joint allows more distance between it and the stimulator (and less stimulus artifact). The median nerve is then stimulated at the wrist at a fixed distance to G1. The median nerve is next stimulated at the palm, with the recording ring electrodes left in place, at half the wrist-to-digit distance ( Fig. 20.11 ). If the EMG machine is set up to automatically compute the wrist-to-palm and palm-to digit conduction velocities, any distance can be used. Otherwise, making the palm-to-digit distance half that of the wrist-to-digit distance (e.g., 7 cm and 14 cm) greatly simplifies the mathematical equation. In this situation, the wrist-to-palm conduction velocity is then computed by multiplying the palm-to-digit conduction velocity by the wrist-to-digit conduction velocity, and then dividing it by the quantity of the palm-to-digit conduction velocity times two, minus the wrist-to-digit conduction velocity ( Fig. 20.12 ). In normal subjects, the wrist-to-palm segment (i.e., the segment across the carpal tunnel) is equal to or faster than the distal segment (palm-to-digit) because proximal nerve normally conducts faster than distal segments, secondary to larger nerve diameter and warmer temperatures. In CTS, there is a reversal of this normal pattern; the proximal segment (wrist-to-palm) conducts more slowly than the distal palm-to-digit segment. In general, any slowing of more than 10 m/s is considered abnormal.

Inching Across the Wrist and Palmar Stimulation

Another technique useful in demonstrating CTS, first described by Kimura and later by others, involves segmental stimulation (“inching”) of the median nerve across the carpal tunnel ( Fig. 20.13 ). One looks for an abrupt change in latency or increase in amplitude above normal control values, recording either a median CMAP at the APB or a median digital SNAP at the index or middle finger.

Kimura’s method begins at 4 cm proximal to the distal wrist crease and continues to 6 cm distal to the wrist crease, with segmental stimulation at 1-cm increments. For each 1-cm increment, latency usually increases 0.2 to 0.3 ms. Any abrupt change in latency greater than this is highly suggestive of focal demyelination. Although the inching technique has the advantage of showing the exact site of the lesion, its effectiveness often is limited by difficulty stimulating the nerve at the sites just distal to the wrist crease. The technique is particularly difficult to perform recording the median CMAP because stimulation of motor fibers at 1-cm increments following the course of the recurrent thenar branch of the median nerve can be quite difficult. Furthermore, stimulation in the palm often requires rotation of the anode to prevent excessive stimulus artifact ( Fig. 20.14 ).

Rather than measuring a change in latency, comparing the CMAP or SNAP amplitudes stimulating at the wrist and palm can be technically easier and can yield additional information about the underlying pathophysiology ( Fig. 20.15 ). Wrist and palmar stimulation can be performed for either median motor or sensory studies. Only single palm and wrist stimulations are required, whereas inching requires stimulation at multiple 1-cm increments. Several technical factors must be taken into account. First, as noted earlier for motor studies, the anatomy of the recurrent thenar motor branch is such that for stimulating the motor branch in the palm, the stimulator often must be placed beyond the thenar eminence with the anode rotated distally to prevent excessive stimulus artifact ( Fig. 20.14 ). Second, the examiner must be aware of normal values when comparing amplitudes proximal and distal to the carpal tunnel. There is always some drop in amplitude proximally compared with distally due to greater temporal dispersion and phase cancellation with proximal stimulation. The effects of normal temporal dispersion and phase cancellation are always greater for sensory fibers than for motor fibers. In normal median nerves, the ratio of the distal to proximal CMAP amplitude does not exceed 1.2, whereas the ratio of distal to proximal SNAP amplitude does not exceed 1.6. Larger ratios suggest some element of conduction block ( Fig. 20.16 ). This assumption presumes that both stimulations are supramaximal, that there is no co-stimulation of adjacent nerves, and that the baseline is not obscured by shock artifact or noise that precludes an accurate amplitude measurement.

In CTS, if wrist stimulation yields a low CMAP or SNAP amplitude, there are two possible explanations: (1) there is conduction block secondary to demyelination across the carpal tunnel with the underlying axon intact, or (2) there has been secondary axonal loss ( Fig. 20.17 ). Comparing the amplitudes obtained with wrist and palmar stimulation can easily sort out these two possibilities. Take the following example:

In both cases, when the median nerve is stimulated at the wrist, the recorded CMAP is low (normal value >4.0 mV). When the palm is stimulated in case A, however, the CMAP amplitude increases by an additional 200%; the distal-to-proximal amplitude ratio is 3.0, signifying conduction block. In contrast, there is no change in amplitude in case B, signifying that the low amplitude is secondary to axonal loss.

Median-Versus-Radial Digit 1 Sensory Latencies

Comparison of the median-versus-radial digit 1 sensory latencies takes advantage of the fact that, in most individuals, digit 1 (the thumb) is innervated by both the median and radial nerves ( Fig. 20.18 ). The basic concept is the same as in the median-versus-ulnar digit 4 sensory study: the median and radial nerves are stimulated at the wrist, using identical distances, with recording ring electrodes over digit 1 (G1 over the metacarpophalangeal joint and G2 over the interphalangeal joint). The radial nerve is stimulated at the wrist along the lateral border of the radial bone. Using the same distance, the median nerve is stimulated at the wrist in the usual location. Supramaximal responses are obtained at each stimulation site, and the onset or peak latencies are compared. Although this technique is popular in some laboratories, stimulating the nerves at identical distances may be difficult because the median nerve travels to the thumb at an angle, which can hinder measurement of its true distance. Any difference between the median and radial latencies greater than or equal to 0.5 ms is considered abnormal.

Median-Versus-Ulnar Minimum F-Wave Latencies

This technique compares the minimum F-wave latency stimulating the median and ulnar nerves at the wrist, recording the APB and abductor digiti minimi muscles, respectively. In normal individuals, the minimum F-wave latency from the median nerve is approximately 1–2 ms shorter than the minimum F-wave latency from the ulnar nerve. A reversal of this pattern is considered abnormal ( Fig. 20.19 ). This test is nonspecific, however, because the F wave measures conduction along the entire length of nerve, from the recording electrode to the spinal cord. Although this study can confirm a problem with the median nerve, it cannot localize the lesion to the wrist. It is generally used only as confirmatory evidence for a diagnosis of CTS, in conjunction with abnormalities noted using more sensitive techniques.