IN THE ELECTRODIAGNOSTIC evaluation of nerve lesions,

Transcription

1 190 Median Pseudoneurapraxia at the Wrist: Reassessment of Palmar Stimulation of the Recurrent Median Nerve Tracy A. Park, MD, John A. Welshofer, MD, William W. Dzwierzynski, MD, Scott J. Erickson, MD, David R. Del Toro, MD ABSTRACT. Park TA, Welshofer JA, Dzwierzynski WW, Erickson SJ, Del Toro DR. Median pseudoneurapraxia at the wrist: reassessment of palmar stimulation of the recurrent median nerve. Arch Phys Med Rehabil 2001;82: Objective: To determine the occurrence and effect of incidental deep ulnar nerve (DUN) costimulation during palmar stimulation of the recurrent median nerve (RMN). Design: Observational. Setting: Electromyography laboratory. Participants: Seventeen asymptomatic adult volunteers (34 hands) and 4 fresh cadaver hands. Main Outcome Measures: Median nerve stimulation at the wrist and careful incremental surface and subcutaneous (needle) palmar stimulation were performed while recording thenar and first dorsal interosseous manus compound muscle action potentials. Thenar palm-to-wrist amplitude difference (P W ) and palm-to-wrist amplitude ratio (P/W) were compared with published values. Correlation of DUN costimulation with falsely elevated P W and P/W values was assessed. Multiplanar magnetic resonance imaging (MRI) and subsequent dissection of 4 fresh cadaver hands was done to measure the distance between the RMN and DUN at the palmar stimulation site. Results: Two groups emerged: Group I (63%), with DUN stimulation, and II (37%), with no DUN stimulation. When DUN costimulation occurred (Group I), there was a statistically significant increase inp W (p.002 percutaneous, p.02 subcutaneous) and P/W (p.004 percutaneous, p.007 subcutaneous) when compared with previously published data. Combining all trials on all hands, 53% and 25% had P W values and P/W values that exceeded previously published normative data, respectively. The mean distance between the DUN and RMN at the palmar stimulation site was determined by dissection (1.2cm) and MRI (1.5cm). Conclusions: Close proximity of the DUN to the RMN causes frequent and often unavoidable DUN activation during palmar stimulation of the RMN. This inadvertent stimulation may cause a false diagnosis of median neurapraxia at the wrist. Key Words: Carpal tunnel syndrome; Electrodiagnosis; Median nerve; Rehabilitation by the American Congress of Rehabilitation Medi- From the Departments of Physical Medicine and Rehabilitation (Park, Welshofer, Del Toro), Plastic and Reconstructive Surgery (Dzwierzynski), and Radiology (Erickson), Medical College of Wisconsin, Milwaukee, Wisconsin. Welshofer is currently affiliated with Miller Orthopaedic Clinic, Charlotte, NC. Accepted in revised form May 2, Presented in part at the American Association of Electrodiagnostic Medicine s 41st Annual Meeting, Oct 1, 1994, San Francisco, CA. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Tracy A. Park, MD, Medical Rehabilitation Associates, 5000 W Chambers St, Milwaukee, WI /01/ $35.00/0 doi: /apmr cine and the American Academy of Physical Medicine and Rehabilitation IN THE ELECTRODIAGNOSTIC evaluation of nerve lesions, neurapraxia may carry a more favorable prognosis for neurologic improvement than axonotmesis. 1 Patients with neurapraxia present with nonconducting nerve fibers that are still in continuity and that may recover conduction if the source of the conduction block is removed. 2 In the electrodiagnostic evaluation of suspected carpal tunnel syndrome (CTS), palmar stimulation of the recurrent median nerve (RMN) has been proposed as a technique for detecting and quantifying neurapraxia of the median nerve in the carpal tunnel. 3-7 If the amplitude of the compound muscle action potential (CMAP) recorded over the thenar eminence is larger with palmar stimulation of the RMN than it is with wrist stimulation of the median nerve, then median neurapraxia may be present in the carpal tunnel. Thus, palmar stimulation has been recommended as part of the routine evaluation of CTS because it may provide both diagnostic and prognostic information. 6,7 However, review of the palm s anatomy suggests another reason for an increased thenar CMAP amplitude with palmar stimulation. The deep ulnar nerve (DUN) lies in close proximity to the RMN, particularly at the site of palmar stimulation. Frequent communications between these 2 nerves (known as Riche-Cannieu anastomoses), 8-18 as well as commonly overlapping innervation patterns, 11,15,17-19 have been extensively documented. If palmar stimulation inadvertently depolarizes the DUN or any communicating or anomalous branches, the resulting thenar CMAP may be larger than a CMAP produced by median nerve stimulation at the wrist. This falsely increased thenar CMAP amplitude could lead to the misdiagnosis of a neurapraxic lesion of the median nerve in the carpal tunnel. Although incidental stimulation of the DUN has been mentioned, 5,6,8,14 no study has examined this aspect of palmar stimulation of the RMN. The primary goals of the present study were to investigate the possibility of incidental DUN excitation by palmar stimulation intended for the RMN and to identify any association between DUN activation and an increased thenar CMAP amplitude after palmar stimulation. These goals were accomplished using 2-channel recording over the thenar eminence and the first dorsal interosseus manus (FDIM) while stimulating the RMN in the palm, and comparing these recordings with findings obtained by stimulating the median nerve and the ulnar nerve at the wrist. An additional aim was to investigate the anatomic relationship between the RMN and DUN in the palm using cadaver hands, which were studied with dissection and magnetic resonance imaging (MRI) scans to verify the close proximity of the RMN to the DUN at the site of palmar stimulation. METHODS Seventeen healthy volunteers (8 men, 9 women; age range, 24 48yr) with no history of injury to the median nerve or

2 PALMAR STIMULATION, Park 191 Fig 1. Placement of percutaneous stimulating electrodes in the palm. The cathode is distal over the site of RMN stimulation, and the anode is place proximally over the median nerve. The recording electrode montage shown is for thenar CMAP recording, with E 1 over the thenar eminence and E 2 over the interphalangeal joint of the thumb. symptoms suggestive of CTS agreed to participate in the study. Physical examination showed no neurologic abnormalities or signs of CTS. Both hands were examined in each volunteer, for a total of 34 hands tested. Hand skin temperature was kept constant throughout the nerve conduction studies for each subject. We performed median and ulnar nerve transcarpal studies on each hand to rule out the possibility of subclinical median neuropathy at the wrist. A standardized technique with stimulation of the median nerve and ulnar nerve in the palm and recording at the wrist was used with a transcarpal distance of 10cm. 20 No hand had electrophysiologic evidence of median nerve slowing across the wrist (defined in our laboratory as a median nerve transcarpal latency at least 0.5ms greater than the ulnar nerve transcarpal latency). All motor conduction studies were performed using simultaneous 2-channel recording a with 9-mm disc electrodes b over the thenar eminence and FDIM. The thenar E 1 site was determined by a technique described elsewhere, 21 with E 2 over the Fig 2. Cathode placement of the 37-mm monopolar needle for subcutaneous palmar stimulation. Anode placement is medially over the midpoint of the fifth metacarpal (not shown). lateral aspect of the interphalangeal joint of the thumb. The FDIM E 1 site was over the palpated belly of the FDIM, with E 2 over the medial aspect of the interphalangeal joint of the thumb. The ground electrode was placed on the dorsum of the hand. We performed median and ulnar nerve stimulation at the wrist by means of percutaneous stimulation with a bipolar stimulator; CMAPs were recorded at 20% supramaximal stimulation intensity. The median nerve was stimulated at the wrist 8cm proximal to the thenar E 1 site 21 and at the antecubital Table 1: CMAP Amplitude Data for All Trials (n 34) Stimulation Site E 1 Site Mean SD (mv) Range (mv) Median nerve at wrist Thenar Ulnar nerve at wrist Thenar Median nerve at wrist FDIM Ulnar nerve at wrist FDIM Palm Surface Thenar Needle Thenar Surface FDIM Needle FDIM

3 192 PALMAR STIMULATION, Park Table 2: Palm-to-Wrist Thenar CMAP Amplitude Change and Palm-to-Wrist Thenar CMAP Amplitude Ratio For All Trials Stimulation No. P W p* P/W p Surface (5.3) (1.8).007 Needle (5.3) (1.7).005 Surface and needle (5.3) (1.8).0001 NOTE. P W : palm-to-wrist thenar CMAP amplitude difference in mv, as mean 2SDs. P/W: palm-to-wrist thenar CMAP ratio, as mean 2SDs. Boldface: Statistically significant p values. * Comparison to published P W mean of.56mv 7 (Student s t test). Comparison to published P/W mean of (Student s t test). crease. The ulnar nerve was stimulated at the wrist 8cm proximal to the midpoint of the hypothenar eminence 21 and in the ulnar groove. Palmar stimulation was performed both subcutaneously (with a Teflon -coated 37-mm monopolar needle electrode b ) and percutaneously (figs 1, 2). For both techniques, the cathode location (previously described by Johnson 3 ) was the point at which the tip of the ring finger flexed at the metacarpophalangeal and proximal interphalangeal joints touched the lifeline of the palm. 7,22 For surface stimulation, we placed the anode proximally over the median nerve in the carpal tunnel, thus minimizing spread of stimulus to the ulnar nerve or to the more distal RMN. We used the technique originally devised by Johnson 3 for subcutaneous needle stimulation. In this technique, only the tip of the stimulating needle (cathode) is placed superficially into the skin, and a 9-mm disc electrode (anode) is placed medially over the midpoint of the fifth metacarpal. For each trial, we first determined RMN threshold stimulus intensity, then performed stimulation while increasing the current in 0.5- to 4.0-mA increments until at least a 20% supramaximal thenar CMAP was achieved. A palm-to-wrist median thenar CMAP amplitude difference (P W ), defined as the difference between the thenar CMAP amplitude resulting from palmar stimulation of the RMN and that resulting from wrist stimulation of the median nerve, was calculated for each trial. As described by Lesser et al, 6 palm-to-wrist thenar CMAP amplitude ratio (P/W), defined as the ratio of the thenar CMAP amplitude resulting from palmar stimulation of the RMN to that resulting from wrist stimulation of the median nerve, was calculated for each trial. For all studies, we used the Viking II bipolar stimulator with a stimulus duration of 100 s. We recorded motor conduction studies with a bandpass of 2Hz to 10kHz, a sweep speed of 5ms/div, and a sensitivity of 5mV/div. Transcarpal studies were recorded with a bandpass of 20Hz to 2kHz, a sweep speed of 2ms/div, and a sensitivity of 20 to 50 V/div. Thenar and FDIM CMAP amplitudes were measured baseline to peak. Incidental DUN activation during attempted RMN stimulation at the palm was considered to have occurred if the FDIM CMAP amplitude exceeded 120% of that recorded with median nerve stimulation at the wrist. The presence of a Martin-Gruber anastomosis in the forearm was determined by the criteria of Crutchfield and Gutmann. 23 Four fresh cadaver hands underwent multiplanar MRI in axial, sagittal, coronal, and 3-dimensional views. Using axial cuts, an experienced neuroradiologist (SJE) identified each nerve and followed its course. The shortest distance between the RMN and DUN at a point corresponding to the site of palmar stimulation was calculated for each hand by applying the Pythagorean theorem to consecutive scans with a 2-mm separation between axial images. The 4 cadaver hands were then dissected by an experienced hand surgeon (WWD) who was blinded to the MRI results. The RMN and DUN were identified, and the shortest distance between these 2 nerves at a point corresponding to the palmar stimulation site was then measured. A comparison of the internerve distances for the cadaver hands was made between those obtained by MRI and anatomic dissection. For data analysis, we used the Student s t test, with p less than.05 considered statistically significant. Linear regression analysis was used to compare the thenar and FDIM CMAP amplitudes with respect to the incrementally increasing stimulus current. RESULTS The values for the CMAPs recorded over the thenar eminence and the FDIM from all stimulation sites are shown for all subjects in table 1. A Martin-Gruber anastomosis was documented in 24% (8/34) of the limbs studied. We found a statistically significant increase in meanp W for surface (1.3mV) and for total surface and needle (1.2mV) palmar stimulation when we compared our present data with previous published normative data (table 2). 7 However, no significant difference existed for needle palmar stimulation (1.1mV). The upper limit of normal (mean 2 standard Table 3: Group I and Group II Palm-to-Wrist Thenar CMAP Amplitude Change and Palm-to-Wrist Thenar CMAP Amplitude Ratio Group Stimulation No. P W p* P/W p Group I Surface (5.9) (1.8).004 Needle (6.7) (1.8).007 Total 43/ Group II Surface (1.4) (1.2) 0.4 Needle (2.3) (1.2) 0.3 Total 25/ NOTE. Group I: trials in which the DUN was stimulated by palmar stimulation; Group II: trials in which the DUN was not stimulated by palmar stimulation. P W : palm-to-wrist thenar CMAP amplitude difference in mv, as mean 2SDs. P/W: palm-to-wrist thenar CMAP ratio, as mean 2SDs. Boldface: Statistically significant p values. * Comparison to published P W mean of.56mv 7 (Student s t test). Comparison to published P/W mean of (Student s t test).

4 PALMAR STIMULATION, Park 193 Table 4: Group I-A and Group I-B Palm-to-Wrist Thenar CMAP Amplitude Change Group Stimulation No. % Mean r P W p Group I-A Surface (7.0).004 Needle (7.0).003 Total 25/ Group I-B Surface (3.9) 0.2 Needle (4.0) 0.6 Total 18/ NOTE. Group I-A: DUN was stimulated by palmar stimulation with contribution to thenar CMAP amplitude. Group I-B: DUN was stimulated by palmar stimulation with no contribution to thenar CMAP amplitude. Mean r : mean correlation coefficient as determined by linear regression analysis. P W : wrist-to-palm thenar CMAP amplitude difference (mean 2SD). p: comparison to published mean of.56mv 7 (Student s t test). Boldface: Statistically significant p values. deviations [SDs]) for P W was 5.3mV whether needle or surface stimulation was performed. We found a statistically significant increase in mean P/W for surface (1.2), needle (1.1), and total (1.2) palmar stimulation when we compared our present data with previous published normative data (table 2). 6 The upper limit of normal (mean 2SD) for P/W was 1.8 for surface stimulation, 1.7 for needle stimulation, and 1.8 for all hands. Based on the criteria for DUN stimulation defined in Methods, our data showed 2 distinct groups: Group I, 43 of 68 trials (63%), with electrophysiologic evidence of DUN stimulation, and Group II, 25 of 68 trials (37%), with no electrophysiologic evidence of DUN stimulation (table 3). For all Group I trials (surface, needle, total), mean P W and mean P/W were significantly higher than previously reported data. 6,7 In hands with no DUN stimulation (Group II), we found no significant difference in P/W for surface, needle, and total palmar stimulation from the mean reported by Lesser et al 6 of1.0.p W for Group II was significantly greater for surface and total (surface and needle) palmar stimulation, but not for needle stimulation, when compared with the mean reported by Pease et al 7 of.56mv. We performed regression analysis on Group I data to determine the correlation between increasing thenar CMAP amplitudes and increases in FDIM CMAP amplitudes resulting from incremental palmar stimulation (table 4). This calculation resulted in a strongly bimodal distribution, with 2 clear subgroups designated Group I-A (37%), with correlation coefficients greater than 0.5, and Group I-B (26%), with correlation coefficients less than 0.5. In other words, in trials with correlation coefficients greater than 0.5, the increase in FDIM CMAP amplitude was clearly associated with a concomitant increase in thenar CMAP amplitude. An example of Group I-A data for 1 hand is provided in figure 3, in which the increasing FDIM CMAP amplitude is clearly accompanied by an increase in the thenar CMAP amplitude. Figure 4 presents the data from 1 hand in Group I-B, in which the increasing FDIM CMAP amplitude was not accompanied by an increase in the thenar CMAP amplitude. For comparison, figure 5 shows data from 1 hand in Group II, in which there was no DUN stimulation, and therefore no increasing FDIM CMAP amplitude. The mean correlation coefficient was 0.8 for Group I-A and 0.1 for Group I-B. The Group I-A values for P W were significantly higher than previously reported values, 3,4,7 whereas the Group I-B values were not. Using published criteria for median neurapraxia at the wrist (P W 0.5mV), 3,4 we placed 53% (18/34) of surfacestimulated hands and the same percentage of needle-stimulated hands into the category of median nerve conduction block across the wrist (table 5). Interestingly, 21% (7/34) of all surface trials and 21% (7/34) of all needle trials met this criteria for median neurapraxia at the wrist, even though (1) no DUN activation existed (Group II), or (2) the DUN activation was not associated with an increasing thenar CMAP amplitude (Group I-B). Fig 3. Example of incremental surface palmar stimulation in which activation of the DUN contributed to the thenar CMAP amplitude, with a correlation coefficient of 0.9 (Group I-A). The thenar CMAP amplitude with median nerve stimulation of the wrist was 8.8mV. Note that as the FDIM CMAP amplitude increased sharply, a simultaneous sharp increase occurred in the thenar CMAP amplitude.

5 194 PALMAR STIMULATION, Park Fig 4. Example of incremental needle palmar stimulation in which activation of the DUN did not contribute to the thenar CMAP amplitude, with a correlation coefficient of 0.0 (Group I-B). The thenar CMAP amplitude with median nerve stimulation at the wrist was 7.5mV. Note that there was no increase in the thenar CMAP amplitude despite a 12-mV increase in the FDIM CMAP amplitude. Using the other published criteria for median neurapraxia at the wrist (P/W 1.2), 6 26% (9/34) of all surface trials and 24% (8/34) of all needle trials fell into the category of median nerve conduction block at the wrist. In the 4 fresh cadaver hands, the mean distance between the RMN and DUN at a point corresponding to the site of palmar stimulation was relatively small: cm by MRI (fig 6) and cm by anatomic dissection (fig 7). No statistically significant difference existed between the 2 techniques (p.08). DISCUSSION Few studies have addressed palmar stimulation of the RMN, and none has incorporated recording over FDIM to detect incidental DUN stimulation. Kimura 5 presented a segmental technique for studying the median nerve across the carpal tunnel and into the palm using surface stimulation. Although intrasubject data were not analyzed, Kimura reported an insignificant difference of 1.0mV between the mean thenar CMAP amplitude with median nerve stimulation at the wrist and the mean thenar CMAP amplitude with RMN stimulation in the palm. This value was similar to the statistically significant differences of 1.3mV for surface stimulation and 1.1mV for needle stimulation in the present study. Pease et al, 7 using needle stimulation of the RMN in the palms of 23 hands, found a mean intrasubject palm-to-wrist CMAP amplitude change (P W ) of.56.74mv. Although the upper limit of normal was not addressed, the mean 2SDs would be 2.04mV. Our study found the upper limit of normal for surface and needle stimulation was much higher, being 5.3mV for both. Comparing our present data to the reported mean of.56mv (see table 2), we found a statistically significant difference for P W for surface stimulation (p.03) and combined surface and needle stimulation trials (p.007), but not for needle stimulation alone (p 0.2). Lesser et al 6 evaluated for neurapraxia using the palm-towrist thenar CMAP amplitude ratio (P/W). He reported in control hands a mean P/W of with an upper limit of normal (mean 2SDs) of 1.2. Comparing our data with this reported mean, we found (see table 2) significantly higher P/W Fig 5. Example of incremental needle palmar stimulation in which no activation of the DUN occurred (Group II). The thenar CMAP amplitude with median nerve stimulation at the wrist was 9.4mV.

6 PALMAR STIMULATION, Park 195 Table 5: Number of Hands With Palm-to-Wrist Thenar CMAP Amplitude Values That Exceed Previously Published Values Stimulation Group Number of Hands With P W 0.5mV 3,4 Number of Hands With P/W Surface (n 34) Group I-A 11 8 Group I-B 51 Group II 2 0 Total 18/34 (53%) 9/34 (26%) Needle (n 34) Group I-A 11 6 Group I-B 3 1 Group II 4 1 Total 18/34 (53%) 8/34 (24%) NOTE. P W : palm-to-wrist thenar CMAP amplitude change in mv. P/W: palm-to-wrist thenar CMAP ratio. Group I-A: trials where DUN stimulation correlated with increasing thenar CMAP. Group I-B: trials where DUN stimulation did not correlate with increasing thenar CMAP. Group II: trials in which the DUN was not activated by palmar stimulation. values for surface (p.007), needle (p.005), and combined surface and needle stimulation (p.0001). The reported upper limit of normal P/W 6 was exceeded in 26% (9/34) of hands for surface stimulation and in 24% (8/34) of hands for needle stimulation (see table 5) in our study. Johnson 3,4 recommended an upper limit of normal for P W of 0.5mV, but did not detail how this value was determined. In our study (see table 5), 53% (18/34) of hand with surface stimulation and 53% (18/34) of hands with needle stimulation exceeded Johnson s recommended upper limit of normal for P W. The present findings call into question the use of palmar stimulation of the RMN for the assessment of median nerve entrapment at the wrist. In part, these findings may be secondary to incidental stimulation of the DUN in the midpalm. Our dissection and MRI analysis of fresh cadaver hands showed that the DUN and RMN were consistently less than 2.0cm apart at the site of palmar stimulation (see figs 6, 7). The purpose of the present study was to investigate the occurrence and effect of incidental DUN activation during palmar stimulation of the RMN. To eliminate the possibility of inadvertent overstimulation at the RMN site in each palm, we gradually increased the stimulus intensity in 0.5- to 4.0-mA increments. This method enabled us to determine precisely 20% supramaximal stimulus intensity for the RMN. A new aspect of the present study, not undertaken in any previous study of palmar stimulation of the RMN, was the addition of FDIM CMAP recording to detect and quantify incidental DUN activation with palmar stimulation intended for Fig 6. Axial MRI image of the palm showing the RMN (dark arrow) and DUN (light arrow) as it traverses the palm, passing at its closest detectable point to the RMN origin. The distance between the 2 nerves at the site of electric stimulation was measured between the DUN in this image and the RMN origin seen on a slightly more proximal axial image. Fig 7. Dissected cadaver hand showing the median nerve (MN) traversing the carpal tunnel beneath the resected edge of the flexor retinaculum (FR) before giving off the RMN. Also seen is the DUN as it traverses the palm. In this particular hand, the shortest distance between the RMN and the DUN is 1.1cm.

7 196 PALMAR STIMULATION, Park activation associated with increasing thenar CMAP amplitude. No other obvious change in the thenar CMAP waveform has occurred, and one cannot determine if this increase in thenar CMAP amplitude is secondary to DUN coactivation or if it simply represents a thenar CMAP that is caused by increasing submaximal stimulation of the RMN. Figure 9 shows an example of DUN activation that has no effect on the thenar CMAP amplitude. Again, there is no appreciable difference in the thenar CMAP waveform characteristics with DUN stimulation and despite a very large FDIM CMAP (18mV) the volume-conducted response at the thenar recording site is negligible. This pattern was seen in 13 of the 43 trials in which the DUN was stimulated. What figure 9 clearly shows is that simple volume conduction from DUN-innervated muscles does not fully explain the increase in thenar CMAP amplitude with palmar stimulation compared with wrist stimulation. This finding may reflect the complex and overlapping innervation patterns seen in normal hands. In the classic innervation of the intrinsic muscles of the lateral hand, as shown in basic anatomy and basic electrophysiology textbooks, 24,25 the muscles primarily responsible for the thenar CMAP the abductor Fig 8. Waveforms from a hand in which DUN activation occurred and was associated with an increasing thenar CMAP amplitude. Surface palmar stimulation was used with 2-channel recording. the RMN. In hands in which DUN stimulation occurred, this technique also allowed direct comparison of the incremental increases in the FDIM CMAP amplitude with those of the thenar CMAP amplitude so that linear regression analysis of these 2 CMAPs could be performed (see figs 3, 4). We found incidental stimulation of the DUN in 63% (43/68) of hands on which palmar stimulation of the RMN was performed. This DUN stimulation was not clinically apparent during electrophysiologic testing. Although the occurrence of a slight contraction of the adductor pollicis has been reported with palmar stimulation of the RMN, 5,6,8 we found this phenomenon to be an unreliable marker of DUN stimulation. The contraction of the thenar eminence often masks the cocontraction of the adductor pollicis, making this indicator of DUN activation difficult or impossible to discern. In addition, well-established overlapping and anomalous innervation patterns of intrinsic hand muscles 9,11,13,15,19 may mean this indicator is misleading. The flexor pollicis brevis may be ulnar innervated in 32% of hands or dually innervated by the ulnar and median nerve in 15.5% of hands. 19 Ulnar stimulation may cause flexor pollicis brevis contraction in a significant number of hands that may either hide or mimic contraction of the adductor pollicis. Additionally, DUN stimulation could not be detected with certainty by observing the thenar CMAP waveform. This point is illustrated by figure 8, which shows an example of DUN Fig 9. Waveforms from a hand in which DUN activation occurred but was not associated with an increasing thenar CMAP amplitude. Surface palmar stimulation was used with 2-channel recording.

8 PALMAR STIMULATION, Park 197 pollicis brevis, opponens pollicis, and superficial head of the flexor pollicis brevis are innervated by the median nerve through the RMN. However, more detailed anatomic and electrophysiologic investigations have found a 33% to 83% frequency of anomalous innervation patterns in the intrinsic muscles of the lateral hand. 9,11,13,15,19 Although palmar stimulation may result in activation of the DUN, it is not clear that this incidental stimulation is actually responsible for the increased thenar P W. CONCLUSION It is apparent from the wide variety of innervation patterns of the thenar muscles that palmar stimulation intended for the RMN could also excite branches from the DUN destined for the thenar muscles or communicating nerve fibers between the DUN and RMN, with or without actually stimulating the DUN itself. Because of the unpredictable course and destination of the motor nerve fibers subjacent to the site of palmar stimulation of the RMN, this technique yields information that may be more misleading than helpful, and it should not be considered a reliable indicator of neurapraxia of the median nerve at the wrist. References 1. Curtis R, Eversmann W. Internal neurolysis as an adjunct to the treatment of the carpal tunnel syndrome. J Bone Joint Surg Am 1973;55: Gilliatt RW. Acute compression block. In: Sumner AJ, editor. The physiology of peripheral nerve disease. Philadelphia: WB Saunders; p Johnson EW. Carpal tunnel syndrome. In: Johnson EW, editor. Practical electromyography. 2nd ed. Baltimore: Williams & Wilkins; p Johnson EW. Electrodiagnosis in prognosis [course handout]. In: Proceedings of the American Association of Electrodiagnostic Medicine Annual Meeting; 1993 Oct 7-10; New Orleans (LA). 5. Kimura J. The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 1979;102: Lesser EA, Venkatesh S, Preston DC, Logigian EL. Stimulation distal to the lesion in patients with carpal tunnel syndrome. Muscle Nerve 1995;18: Pease WS, Cunningham ML, Walsh WE, Johnson EW. Determining neurapraxia in carpal tunnel syndrome. Am J Phys Med Rehabil 1988;67: Brown WF, Bolton CF. Clinical electromyography. 2nd ed. Boston: Butterworth-Heinemann; p Dumitru D, Walsh NE, Weber CF. Electrophysiologic study of the Riche-Cannieu anomaly. Electromyogr Clin Neurophysiol 1988; 28: Gutmann L. AAEM Minimonograph #2: important anomalous innervations of the extremities. Muscle Nerve 1993;16: Harness D, Sekeles E. The double anastomotic innervation of thenar muscles. J Anat 1971;109: Kaplan EB, Spinner M. Normal and anomalous innervation patterns in the upper extremity. In: Omer GE, Spinner M, editors. Management of peripheral nerve problems. Philadelphia: WB Saunders; 1980, p Kimura I, Ayyar DR. The hand neural communication between the ulnar and median nerves: electrophysiologic detection. Electromyogr Clin Neurophysiol 1984;24: Kimura J. Electrodiagnosis in diseases of nerve and muscle: principles and practice. 2nd ed. Philadelphia: FA Davis; p 144, Mannerfelt L. Studies on the hand in ulnar nerve paralysis. A clinical experimental investigation in normal and anomalous innervation. Acta Orthop Scand Suppl 1966;87: Oh SJ. Clinical electromyography: nerve conduction studies. 2nd ed. Baltimore: Williams & Wilkins; p Seddon H. Surgical disorders of the peripheral nerves. 2nd ed. Edinburgh: Churchill Livingstone; p Sunderland S. Nerves and nerve injuries. 2nd ed. Edinburgh: Churchill Livingstone; p Rowntree T. Anomalous innervation of the hand muscles. J Bone Joint Surg Br 1949;31: Mills KR. Orthodromic sensory action potentials from palmar stimulation in the diagnosis of carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1985;48: DeLisa JA, Mackenzie K, Baran EM. Manual of nerve conduction velocity and somatosensory evoked potentials. 2nd ed. New York: Raven Pr; p 47-50, Johnson RK, Shrewsbury MM. Anatomical course of the thenar branch of the median nerve usually a separate tunnel through the carpal ligament. J Bone Joint Surg Am 1970;52: Crutchfield CA, Gutmann L. Hereditary aspects of median-ulnar nerve communications. J Neurol Neurosurg Psychiatry 1980;43: Delagi EF, Perotto A. Anatomic guide for the electromyographer. Springfield (IL): Thomas; p Sethi RK, Thompson LL. The electromyographer s handbook. 2nd ed. Boston: Little, Brown; p Suppliers a. Nicolet Biomedical, Inc, 5225 Verona Rd, Bldg 2, Madison, WI b. Oxford Instruments Medical Inc, 12 Skyline Dr, Hawthorne, NY