- Email: [email protected]
Does direct measurement of forearm mixed nerve conduction velocity reflect actual nerve conduction velocity through the carpal tunnel?
Clinical Neurophysiology 113 (2002) 1236–1240 www.elsevier.com/locate/clinph
Does direct measurement of forearm mixed nerve conduction velocity reflect actual nerve conduction velocity through the carpal tunnel? Ming-Hong Chang a,b,c,*, Shiew-Jue Wei a,b,c, Hui-Ling Chiang a,b,c, Hui-Mei Wang a,b,c, Peiyuan F. Hsieh a,b,c, San-Yong Huang a,b,c a
Section of Neurology, Taichung Veterans General Hospital, No 160 Chung-Kang Road, Section 3, Taichung ,40705, Taiwan b Department of Neurology, National Yang-Ming University, Taipei, Taiwan c Department of Neurology, Chung Shan Medical University, Taichung, Taiwan Accepted 21 May 2002
Abstract Objectives: The purpose of this study was to determine whether forearm (wrist–elbow) mixed nerve conduction velocity (W–Emix) represents the actual nerve conduction velocity (CV) of nerve fibers passing through the carpal tunnel. Background: W–Emix is presumed to reflect the actual forearm CV through the carpal tunnel. However, it has been argued that W–Emix chiefly originates from the nerve fibers passing outside the carpal tunnel. Therefore, the direct measurement of W–Emix cannot be used to assess retrograde axonal atrophy in carpal tunnel syndrome (CTS). Subjects and methods: Thirty patients with clinical signs and symptoms of CTS were recruited and the diagnosis was confirmed with standard electrodiagnosis. Fifty age-matched volunteers served as control. Recording electrodes were placed over the elbow and index finger for mixed nerve and sensory nerve conduction studies, respectively. Stimulation was applied at the palm and wrist for the measurement of mixed nerve wrist–palm CV (W–Pmix), wrist–elbow CV (W–Emix), and elbow–palm CV (E–Pmix). Stimulation was applied at the elbow, wrist, and palm for the measurement of wrist–elbow sensory CV (W–Esen), wrist–palm CV (W–Psen), and elbow–palm CV (E–Psen). Comparisons were made between W–Pmix and W–Psen, W–Emix and W–Esen, and E–Pmix and E–Psen. Results: Correlations between W–Emix and W–Esen, E–Pmix and E–Psen, and W–Pmix and W–Psen were good in the control. In the patient group, there was a strong positive correlation between W–Pmix and W–Psen, and between E–Pmix and E–Psen. However, W–Esen correlated weakly with W–Emix, suggesting that W–Emix chiefly represents the CV of fibers passing outside the carpal tunnel. Therefore, the direct measurement of W–Emix cannot be used to assess retrograde axonal atrophy. Furthermore, the reduction in W–Psen was more marked than the reduction in W–Esen, implying that a conduction block at the wrist is the least likely cause of proximal slowing in CTS. Conclusions: W–Emix does not reflect the actual CV of the nerve fibers passing through the carpal tunnel. In addition, retrograde axonal atrophy appears to be the primary cause of decreased forearm CV in CTS. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Carpal tunnel syndrome; Electrodiagnosis; Conduction block; Retrograde axonal atrophy
1. Introduction Standard electrodiagnostic tests of patients with carpal tunnel syndrome (CTS) typically show decreased conduction velocity (CV) in the median wrist–palm segment (Buchthal and Rosenfalck, 1971; Kimura, 1978, 1979; Felsenthal, 1979; Mills, 1985; Jackson and Clifford, 1989; Stevens, 1997). Occasionally, some patients also have a slowing of forearm median nerve CV (Thomas et al., 1967; Melvin et al., 1973; Buchthal et al., 1974; Stoehr et al., 1978; Pease et al., 1990; Uchida and Sugioka, 1993). * Corresponding author. Tel.: 1886-4-2374-1302; fax: 1886-4-23584403. E-mail address: [email protected] (M.-H. Chang).
This slowing is due to one of two possible mechanisms. The first is a conduction block of the fastest myelinating fibers in the carpal tunnel. In this situation, CV, as measured by electrodiagnostic methods, reflects the CV of only the slower axons (Thomas and Fullerton, 1963; Melvin et al., 1973; Stevens, 1997; Wilson, 1998). The other proposed mechanism, observed in animal models, is retrograde axonal atrophy, which may result in an increase in electrical resistance and a decrease in CV (Anderson et al., 1970; Dyck et al., 1981; Devor and Govrin-Lippmarn, 1986). A measurement of the forearm (wrist–elbow) mixed nerve CV (W–Emix) could point to the actual mechanism (Pease et al., 1990; Uchida and Sugioka, 1993; Wilson, 1998). However, because W–Emix does not come about as a result of the
1388-2457/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88- 2457(02)0015 7-8
CLINPH 2001746
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
action potentials of propagation through the carpal tunnel, it would not be affected by a conduction block which is distal to the forearm. Thus, a decrease in mixed nerve CV in the forearm segment would suggest retrograde axonal atrophy (Hansson, 1994). Some authors argue that W–Emix does not reflect the actual CV of mixed nerve fibers passing through the carpal tunnel, though most but not all of the fibers of the median nerve located in the forearm could be compressed in the carpal tunnel (Hansson, 1994). If there is retrograde axonal atrophy, wrist compression will exert a large influence on the nerve fibers in the carpal tunnel and the palmar cutaneous branch, allowing the other unimpaired branches of the median nerve to dominate in the measurement of W– Emix (Hansson, 1994). Therefore, W–Emix may not reflect the actual CV of impaired fibers through the carpal tunnel. The purpose of this study was to determine if the direct measurement of forearm mixed nerve CV reflects actual nerve CV through the carpal tunnel. In addition, we attempted to elucidate the primary cause of decreased forearm CV in CTS. 2. Patients and methods 2.1. Patients CTS was diagnosed clinically based on the presence of the following primary symptoms in a median nerve distribution: numbness, paresthesia, nocturnal awakening, weakness/clumsiness, and pain. The diagnosis was often supported by a positive Tinel’s sign, or by thenar muscle atrophy, but these signs were not required. Subjects with a history or physical examination suggestive of a neuromuscular disorder other than CTS were excluded. All patients with clinically diagnosed CTS demonstrated median neuropathy at the wrist, which was confirmed by the presence of one or more of the following standard electrophysiological criteria: (1) prolonged motor distal latency to APB (abnormal $ 4.5 ms, stimulation over the wrist, 8 cm proximal to the active electrode); (2) prolonged antidromic sensory onset latency to the second digit (abnormal $ 3.0 ms; stimulation over the wrist, 14 cm proximal to the active electrode); and (3) prolonged orthodromic mixed nerve or sensory nerve onset latency in the median palm–wrist segment at a distance of 8 cm (abnormal $ 2.0 ms). Thirty consecutive patients with a clinical diagnosis of CTS and a proven median neuropathy at the wrist, were enrolled in the study. All patients had mild symptoms of CTS. Subjects were excluded if they had muscle atrophy, weakness, or persistent sensory deficits, because sensory and mixed nerve responses are difficult to elicit in advanced CTS.
1237
with stimulation was carried out for all studies. Recording electrodes were two 1 cm diameter stainless steel disk electrodes for mixed nerve studies, and saline-soaked Velcro ring electrodes for antidromic sensory studies. The surface temperature of the hand was measured and maintained at or above 328C. 2.3. Sensory nerve conduction studies Median sensory nerve action potentials (SNAPs) were obtained by antidromically stimulating the mid-palm, 60 mm proximal to the index finger, the wrist, 140 mm proximal to the index finger, and the elbow. The recording electrodes were placed over the metacarpal joint (active) and distal interphalangeal joint (reference). Supramaximal responses were obtained and latencies to the onset, as well as amplitude from baseline to negative peak were measured. The distance from the stimulating cathode to the active electrode over the metacarpal joint was measured. All sensory responses were averaged 5 times to obtain clear onset latencies. Based on the latency differences of the median nerve at the wrist, palm, and elbow stimulation, the median sensory wrist–palm CV (W–Psen), wrist– elbow CV (W–Esen), and elbow–palm CV (E–Psen), were calculated. 2.4. Mixed nerve conduction studies Mixed nerve conduction studies were performed by stimulating the mid-palm and wrist, followed by a recording at the elbow. Supramaximal responses were obtained, and latencies to the onset, as well as amplitude from baseline to negative peak, were measured. All mixed nerve responses were averaged 5 times to obtain clear onset latencies. The distance from the stimulating cathode to the active recording electrode on the elbow was measured. Based on the latency differences at the mid-palm and wrist stimulation, wrist–palm mixed nerve CV (W–Pmix), wrist–elbow mixed nerve CV (W–Emix), and elbow–palm mixed nerve CV (E– Pmix), were calculated. 2.5. Statistical analysis Linear regression was used to determine the correlations between the mixed nerve CV and sensory nerve CV of each segment. A paired t test or Wilcoxon signed rank test was used to compare segment CVs of patients and the control, which were determined by sensory or mixed nerve stimulation. A P value less than 0.05 was considered significant. 3. Results
2.2. Electrophysiological protocol
3.1. Normal controls
The studies were performed using a Nicolet Viking IV or Dantec Key Point 4 electromyograph. Surface recording
Fifty volunteers consented to receiving the electrophysiological protocol. There were good correlations between W–
1238
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
3.2. Patients Thirty CTS patients, with a mean age of 48.3 years, consented to participate in this study. The electrophysiological data are summarized in Table 1. W–Pmix and E–Pmix showed good correlations with W–Psen (R ¼ 0:813, R2 ¼ 0:661, P , 0:0001) and E–Psen (R ¼ 0:909, R2 ¼ 0:825, P , 0:0001), respectively. However, there was no correlation between W–Emix and W–Esen (R ¼ 0:122, R2 ¼ 0:015, P ¼ 0:522). Moreover, a relatively lesser reduction in W–Emix, and a relatively greater decrease in W–Esen, was found. This suggested that W– Emix mainly represented the CV of nerve fibers passing outside the carpal tunnel, and possibly the CV of undamaged nerve fibers. Therefore, W–Emix should not replace W–Esen for the assessment of retrograde axonal atrophy. The summarized results are shown in Figs. 1–3, and in Table 1. Furthermore, the decrease in W–Esen (the difference between the W–Esen of patients and the mean value of the control W–Esen: 3.85 ^ 3.4) was less marked than the decrease in W–Psen (the difference between the W–Psen of patients and the mean value of the control W–Psen: 24.2 ^ 7.2), suggesting that the reduction in wrist–palm CV does not parallel the reduction in wrist–elbow CV. If a conduction block with only the slower axons passing through the carpal tunnel is the primary cause of proximal slowing, the reduction in wrist–palm CV should parallel the reduction in wrist–elbow CV. However, the reverse was true in our study. Fig. 1. Linear regression of wrist–palm conduction velocity (W–P) by mixed nerve stimulation (W–Pmix) versus sensory nerve stimulation (W– Psen) in patients and controls.
Pmix and W–Psen (R ¼ 0:721, R2 ¼ 0:520, P , 0:0001), between W–Emix and W–Esen (R ¼ 0:618, R2 ¼ 0:382, P , 0:0001), and between E–Pmix and E–Psen (R ¼ 0:745, R2 ¼ 0:556, P , 0:0001). The cumulative results are shown in Figs. 1–3 and in Table 1.
4. Discussion Stoehr et al. first introduced the direct measurement of forearm mixed nerve CV to represent the actual CV of damaged nerve fibers in CTS (Stoehr et al., 1978). Since then, numerous other studies have used this method for the assessment of median forearm conduction slowing in CTS (Pease et al., 1990; Uchida and Sugioka, 1993; Wilson, 1998). Some authors support the hypothesis that retrograde
Table 1 Summarized electrophysiological data a Controls
Wrist–elbow conduction velocity (W–E) Wrist–palm conduction velocity (W–P) Elbow–palm conduction velocity (E–P) a
Patients
Mixed nerve
Sensory nerve
Mixed nerve
Sensory nerve
63.51 ^ 3.46
61.98 ^ 3.5
63.04 ^ 3.41
57.92 ^ 3.34*
54.58 ^ 5.41
60.7 ^ 5.64
31.11 ^ 6.44**
33.49 ^ 7.51**
60.59 ^ 3.57
61.54 ^ 3.54
47.97 ^ 5.02**
47.56 ^ 5.4**
*P , 0:01; **P , 0:0001.
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
Fig. 2. Linear regression of the wrist–elbow conduction velocity (W–E) by mixed nerve stimulation (W–Emix) versus sensory nerve stimulation (W– Esen) in patients and controls.
axonal atrophy causes the slowing of CV in the forearm segment in CTS (Pease et al., 1990; Uchida and Sugioka, 1993; Chang et al., 2000); however, others suggest that retrograde axonal atrophy dose not play a role in reduced CV (Stevens, 1997; Wilson, 1998). Possible explanations for this controversy are the use of inappropriate methods, or incorrect interpretations. There are major differences in the results, which are secondary to different study populations. In severe CTS patients, wrist compression may have a markedly large effect on the nerve fibers in the carpal tunnel, and allow the unimpaired branches passing outside the carpal tunnel to dominate in the measurement of W–Emix. Therefore, W– Emix might only reflect the CV of the nerve fibers outside the carpal tunnel. If a group with severe axonal degeneration and lower CMAP amplitude were to be selected, it would appear that mixed nerve CV and motor or sensory forearm CV do not equally decrease with a relative sparing of the mixed nerve CV, when compared with control. This finding would suggest that conduction block, rather than retrograde axonal atrophy, is the major cause of proximal slowing. If a mildly afflicted group were to be tested, the opposite results
1239
Fig. 3. Linear regression of the elbow–palm conduction velocity (E–P) by mixed nerve stimulation (E–Pmix) versus sensory nerve stimulation (E– Psen) in patients and controls.
would be obtained, with a parallel reduction in W–Esen and W–Emix. Therefore, Hansson argued that the direct measurement of forearm mixed nerve CV does not reflect the actual CV of damaged nerve fibers through the carpal tunnel, but rather the CV of nerve fibers in the forearm that do not pass through the carpal tunnel. He recommended that the motor and sensory CVs, but not the mixed nerve CV, in the forearm be used to estimate possible retrograde impairment in CTS (Hansson, 1994). In his study, he did not measure the wrist–palm CV or elbow–palm CV for mixed and sensory nerves. In the present study, the correlations between mixed and sensory CVs were good in each segment, among the normal control. However, in the CTS patients, W– Emix was markedly faster than W–Esen, suggesting that in a diseased state, W–Emix is dominated by the undamaged nerve fibers and W–Esen reflects the actual CV of the nerve fibers through the carpal tunnel. Furthermore, E– Pmix correlated well with E–Psen, implying E–Pmix would be preferred over W–Emix for the evaluation of retrograde axonal atrophy.
1240
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
If a conduction block of the large myelinating fibers at the wrist, with only slower conduction axons persisting through the carpal tunnel, is the primary cause of decreased forearm CV, the reduction in wrist–palm CV should parallel the reduction in wrist–elbow CV. If retrograde axonal atrophy is the primary cause, the reduction in wrist–elbow CV would not correlate with the reduction in wrist–palm CV, depending on the severity of the retrograde axonal atrophy. Based on our results, the reduction in wrist–palm CV was more marked than the reduction in wrist–elbow CV, suggesting that retrograde axonal atrophy is the more likely cause of the decrease in forearm CV. This finding was consistent with animal studies (Dyck et al., 1981; Devor and Govrin-Lippmarn, 1986). Retrograde axonal atrophy increases the electrical resistance to longitudinal current flow, resulting in a slowing of the proximal CV. We conclude that a direct measurement of forearm mixed nerve CV, would only reflect the CV of the nerves fibers passing outside the carpal tunnel, and that for an assessment of retrograde axonal atrophy, mixed or sensory conducting techniques through the wrist are more appropriate. Furthermore, retrograde axonal atrophy is the most likely explanation for the forearm conduction slowing in CTS, based on our findings. Acknowledgements This study was supported by grants NSC 89-2314-B075A-029, NSC90-2314-B-075A-005, NSC91-2314-B075A-004, TCVGH-903402C and TCVGH-903405N and TCVGH-913404C, awarded to Dr Ming-Hong Chang. References Anderson MJ, Fullerton PM, Gilliatt RW, Hern JE. Changes in the forearm associated with median nerve compression at the wrist in the guinea pig. J Neurol Neurosurg Psychiatry 1970;33:70–79. Buchthal F, Rosenfalck A. Sensory conduction from digit to palm and from palm to wrist in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1971;34:243–252.
Chang MH, Chiang HT, Ger LP, Yang DA, Lo YK. The cause of slowed forearm median conduction velocity in carpal tunnel syndrome. Clin Neurophysiol 2000;111:1039–1044. Buchthal F, Rosenfalck A, Trojaborg W. Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 1974;37:340–360. Dyck PJ, Lais AC, Karnes JL, Sparks M, Hunder H, Windebank AJ. Permanent axotomy, a model of axonal atrophy and secondary segmental demyelination and remyelination. Ann Neurol 1981;9:575–583. Devor M, Govrin-Lippmarn R. Retrograde slowing of conduction in sensory axons central to a sciatic nerve neuroma. Exp Neurol 1986;92:522–532. Felsenthal G, Spindler H. Palmar conduction time of median and ulnar nerves of normal subjects and patients with carpal tunnel syndrome. Am J Phys Med 1979;25:131–138. Hansson S. Does forearm mixed nerve conduction velocity reflect retrograde changes in carpal tunnel syndrome? Muscle Nerve 1994;17:725– 729. Jackson DA, Clifford JC. Electrodiagnosis of mild carpal tunnel syndrome. Arch Phys Med Rehabil 1989;70:199–204. Kimura J. A method for determining median nerve conduction velocity across the carpal tunnel. J Neurol Sci 1978;38:1–10. Kimura J. The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 1979;102:619–639. Melvin JA, Schuchmann JA, Lanese RR. Diagnostic specificity of motor and sensory nerve conduction variables in carpal tunnel syndrome. Arch Phys Med Rehabil 1973;54:69–74. Mills KR. Orthodromic sensory action potential from palmar stimulation in the diagnosis of carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1985;48:250–255. Pease WS, Lee HH, Johnson EW. Forearm median nerve conduction velocity in carpal tunnel syndrome. Electromyogr Clin Neurophysiol 1990;30:299–302. Stoehr M, Petruch F, Scheglmann K, Schilling K. Retrograde changes of nerve fibers with the carpal tunnel syndrome. J Neurol 1978;218:287– 297. Stevens JC. The electrodiagnosis of carpal tunnel syndrome. Muscle Nerve 1997;20:1447–1486. Thomas PK, Fullerton PM. Nerve fiber size in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1963;26:520–527. Thomas JE, Lambert EH, Cseuz KA. Electrodiagnostic aspects of the carpal tunnel syndrome. Arch Neurol 1967;16:635–641. Uchida Y, Sugioka Y. Electrodiagnosis of retrograde changes in carpal tunnel syndrome. Electromyogr Clin Neurophysiol 1993;33:55–58. Wilson JR. Median mixed nerve conduction studies in the forearm: evidence against retrograde demyelination in carpal tunnel syndrome. J Clin Neurophysiol 1998;15:541–546.
Does direct measurement of forearm mixed nerve conduction velocity reflect actual nerve conduction velocity through the carpal tunnel? Ming-Hong Chang a,b,c,*, Shiew-Jue Wei a,b,c, Hui-Ling Chiang a,b,c, Hui-Mei Wang a,b,c, Peiyuan F. Hsieh a,b,c, San-Yong Huang a,b,c a
Section of Neurology, Taichung Veterans General Hospital, No 160 Chung-Kang Road, Section 3, Taichung ,40705, Taiwan b Department of Neurology, National Yang-Ming University, Taipei, Taiwan c Department of Neurology, Chung Shan Medical University, Taichung, Taiwan Accepted 21 May 2002
Abstract Objectives: The purpose of this study was to determine whether forearm (wrist–elbow) mixed nerve conduction velocity (W–Emix) represents the actual nerve conduction velocity (CV) of nerve fibers passing through the carpal tunnel. Background: W–Emix is presumed to reflect the actual forearm CV through the carpal tunnel. However, it has been argued that W–Emix chiefly originates from the nerve fibers passing outside the carpal tunnel. Therefore, the direct measurement of W–Emix cannot be used to assess retrograde axonal atrophy in carpal tunnel syndrome (CTS). Subjects and methods: Thirty patients with clinical signs and symptoms of CTS were recruited and the diagnosis was confirmed with standard electrodiagnosis. Fifty age-matched volunteers served as control. Recording electrodes were placed over the elbow and index finger for mixed nerve and sensory nerve conduction studies, respectively. Stimulation was applied at the palm and wrist for the measurement of mixed nerve wrist–palm CV (W–Pmix), wrist–elbow CV (W–Emix), and elbow–palm CV (E–Pmix). Stimulation was applied at the elbow, wrist, and palm for the measurement of wrist–elbow sensory CV (W–Esen), wrist–palm CV (W–Psen), and elbow–palm CV (E–Psen). Comparisons were made between W–Pmix and W–Psen, W–Emix and W–Esen, and E–Pmix and E–Psen. Results: Correlations between W–Emix and W–Esen, E–Pmix and E–Psen, and W–Pmix and W–Psen were good in the control. In the patient group, there was a strong positive correlation between W–Pmix and W–Psen, and between E–Pmix and E–Psen. However, W–Esen correlated weakly with W–Emix, suggesting that W–Emix chiefly represents the CV of fibers passing outside the carpal tunnel. Therefore, the direct measurement of W–Emix cannot be used to assess retrograde axonal atrophy. Furthermore, the reduction in W–Psen was more marked than the reduction in W–Esen, implying that a conduction block at the wrist is the least likely cause of proximal slowing in CTS. Conclusions: W–Emix does not reflect the actual CV of the nerve fibers passing through the carpal tunnel. In addition, retrograde axonal atrophy appears to be the primary cause of decreased forearm CV in CTS. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Carpal tunnel syndrome; Electrodiagnosis; Conduction block; Retrograde axonal atrophy
1. Introduction Standard electrodiagnostic tests of patients with carpal tunnel syndrome (CTS) typically show decreased conduction velocity (CV) in the median wrist–palm segment (Buchthal and Rosenfalck, 1971; Kimura, 1978, 1979; Felsenthal, 1979; Mills, 1985; Jackson and Clifford, 1989; Stevens, 1997). Occasionally, some patients also have a slowing of forearm median nerve CV (Thomas et al., 1967; Melvin et al., 1973; Buchthal et al., 1974; Stoehr et al., 1978; Pease et al., 1990; Uchida and Sugioka, 1993). * Corresponding author. Tel.: 1886-4-2374-1302; fax: 1886-4-23584403. E-mail address: [email protected] (M.-H. Chang).
This slowing is due to one of two possible mechanisms. The first is a conduction block of the fastest myelinating fibers in the carpal tunnel. In this situation, CV, as measured by electrodiagnostic methods, reflects the CV of only the slower axons (Thomas and Fullerton, 1963; Melvin et al., 1973; Stevens, 1997; Wilson, 1998). The other proposed mechanism, observed in animal models, is retrograde axonal atrophy, which may result in an increase in electrical resistance and a decrease in CV (Anderson et al., 1970; Dyck et al., 1981; Devor and Govrin-Lippmarn, 1986). A measurement of the forearm (wrist–elbow) mixed nerve CV (W–Emix) could point to the actual mechanism (Pease et al., 1990; Uchida and Sugioka, 1993; Wilson, 1998). However, because W–Emix does not come about as a result of the
1388-2457/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88- 2457(02)0015 7-8
CLINPH 2001746
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
action potentials of propagation through the carpal tunnel, it would not be affected by a conduction block which is distal to the forearm. Thus, a decrease in mixed nerve CV in the forearm segment would suggest retrograde axonal atrophy (Hansson, 1994). Some authors argue that W–Emix does not reflect the actual CV of mixed nerve fibers passing through the carpal tunnel, though most but not all of the fibers of the median nerve located in the forearm could be compressed in the carpal tunnel (Hansson, 1994). If there is retrograde axonal atrophy, wrist compression will exert a large influence on the nerve fibers in the carpal tunnel and the palmar cutaneous branch, allowing the other unimpaired branches of the median nerve to dominate in the measurement of W– Emix (Hansson, 1994). Therefore, W–Emix may not reflect the actual CV of impaired fibers through the carpal tunnel. The purpose of this study was to determine if the direct measurement of forearm mixed nerve CV reflects actual nerve CV through the carpal tunnel. In addition, we attempted to elucidate the primary cause of decreased forearm CV in CTS. 2. Patients and methods 2.1. Patients CTS was diagnosed clinically based on the presence of the following primary symptoms in a median nerve distribution: numbness, paresthesia, nocturnal awakening, weakness/clumsiness, and pain. The diagnosis was often supported by a positive Tinel’s sign, or by thenar muscle atrophy, but these signs were not required. Subjects with a history or physical examination suggestive of a neuromuscular disorder other than CTS were excluded. All patients with clinically diagnosed CTS demonstrated median neuropathy at the wrist, which was confirmed by the presence of one or more of the following standard electrophysiological criteria: (1) prolonged motor distal latency to APB (abnormal $ 4.5 ms, stimulation over the wrist, 8 cm proximal to the active electrode); (2) prolonged antidromic sensory onset latency to the second digit (abnormal $ 3.0 ms; stimulation over the wrist, 14 cm proximal to the active electrode); and (3) prolonged orthodromic mixed nerve or sensory nerve onset latency in the median palm–wrist segment at a distance of 8 cm (abnormal $ 2.0 ms). Thirty consecutive patients with a clinical diagnosis of CTS and a proven median neuropathy at the wrist, were enrolled in the study. All patients had mild symptoms of CTS. Subjects were excluded if they had muscle atrophy, weakness, or persistent sensory deficits, because sensory and mixed nerve responses are difficult to elicit in advanced CTS.
1237
with stimulation was carried out for all studies. Recording electrodes were two 1 cm diameter stainless steel disk electrodes for mixed nerve studies, and saline-soaked Velcro ring electrodes for antidromic sensory studies. The surface temperature of the hand was measured and maintained at or above 328C. 2.3. Sensory nerve conduction studies Median sensory nerve action potentials (SNAPs) were obtained by antidromically stimulating the mid-palm, 60 mm proximal to the index finger, the wrist, 140 mm proximal to the index finger, and the elbow. The recording electrodes were placed over the metacarpal joint (active) and distal interphalangeal joint (reference). Supramaximal responses were obtained and latencies to the onset, as well as amplitude from baseline to negative peak were measured. The distance from the stimulating cathode to the active electrode over the metacarpal joint was measured. All sensory responses were averaged 5 times to obtain clear onset latencies. Based on the latency differences of the median nerve at the wrist, palm, and elbow stimulation, the median sensory wrist–palm CV (W–Psen), wrist– elbow CV (W–Esen), and elbow–palm CV (E–Psen), were calculated. 2.4. Mixed nerve conduction studies Mixed nerve conduction studies were performed by stimulating the mid-palm and wrist, followed by a recording at the elbow. Supramaximal responses were obtained, and latencies to the onset, as well as amplitude from baseline to negative peak, were measured. All mixed nerve responses were averaged 5 times to obtain clear onset latencies. The distance from the stimulating cathode to the active recording electrode on the elbow was measured. Based on the latency differences at the mid-palm and wrist stimulation, wrist–palm mixed nerve CV (W–Pmix), wrist–elbow mixed nerve CV (W–Emix), and elbow–palm mixed nerve CV (E– Pmix), were calculated. 2.5. Statistical analysis Linear regression was used to determine the correlations between the mixed nerve CV and sensory nerve CV of each segment. A paired t test or Wilcoxon signed rank test was used to compare segment CVs of patients and the control, which were determined by sensory or mixed nerve stimulation. A P value less than 0.05 was considered significant. 3. Results
2.2. Electrophysiological protocol
3.1. Normal controls
The studies were performed using a Nicolet Viking IV or Dantec Key Point 4 electromyograph. Surface recording
Fifty volunteers consented to receiving the electrophysiological protocol. There were good correlations between W–
1238
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
3.2. Patients Thirty CTS patients, with a mean age of 48.3 years, consented to participate in this study. The electrophysiological data are summarized in Table 1. W–Pmix and E–Pmix showed good correlations with W–Psen (R ¼ 0:813, R2 ¼ 0:661, P , 0:0001) and E–Psen (R ¼ 0:909, R2 ¼ 0:825, P , 0:0001), respectively. However, there was no correlation between W–Emix and W–Esen (R ¼ 0:122, R2 ¼ 0:015, P ¼ 0:522). Moreover, a relatively lesser reduction in W–Emix, and a relatively greater decrease in W–Esen, was found. This suggested that W– Emix mainly represented the CV of nerve fibers passing outside the carpal tunnel, and possibly the CV of undamaged nerve fibers. Therefore, W–Emix should not replace W–Esen for the assessment of retrograde axonal atrophy. The summarized results are shown in Figs. 1–3, and in Table 1. Furthermore, the decrease in W–Esen (the difference between the W–Esen of patients and the mean value of the control W–Esen: 3.85 ^ 3.4) was less marked than the decrease in W–Psen (the difference between the W–Psen of patients and the mean value of the control W–Psen: 24.2 ^ 7.2), suggesting that the reduction in wrist–palm CV does not parallel the reduction in wrist–elbow CV. If a conduction block with only the slower axons passing through the carpal tunnel is the primary cause of proximal slowing, the reduction in wrist–palm CV should parallel the reduction in wrist–elbow CV. However, the reverse was true in our study. Fig. 1. Linear regression of wrist–palm conduction velocity (W–P) by mixed nerve stimulation (W–Pmix) versus sensory nerve stimulation (W– Psen) in patients and controls.
Pmix and W–Psen (R ¼ 0:721, R2 ¼ 0:520, P , 0:0001), between W–Emix and W–Esen (R ¼ 0:618, R2 ¼ 0:382, P , 0:0001), and between E–Pmix and E–Psen (R ¼ 0:745, R2 ¼ 0:556, P , 0:0001). The cumulative results are shown in Figs. 1–3 and in Table 1.
4. Discussion Stoehr et al. first introduced the direct measurement of forearm mixed nerve CV to represent the actual CV of damaged nerve fibers in CTS (Stoehr et al., 1978). Since then, numerous other studies have used this method for the assessment of median forearm conduction slowing in CTS (Pease et al., 1990; Uchida and Sugioka, 1993; Wilson, 1998). Some authors support the hypothesis that retrograde
Table 1 Summarized electrophysiological data a Controls
Wrist–elbow conduction velocity (W–E) Wrist–palm conduction velocity (W–P) Elbow–palm conduction velocity (E–P) a
Patients
Mixed nerve
Sensory nerve
Mixed nerve
Sensory nerve
63.51 ^ 3.46
61.98 ^ 3.5
63.04 ^ 3.41
57.92 ^ 3.34*
54.58 ^ 5.41
60.7 ^ 5.64
31.11 ^ 6.44**
33.49 ^ 7.51**
60.59 ^ 3.57
61.54 ^ 3.54
47.97 ^ 5.02**
47.56 ^ 5.4**
*P , 0:01; **P , 0:0001.
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
Fig. 2. Linear regression of the wrist–elbow conduction velocity (W–E) by mixed nerve stimulation (W–Emix) versus sensory nerve stimulation (W– Esen) in patients and controls.
axonal atrophy causes the slowing of CV in the forearm segment in CTS (Pease et al., 1990; Uchida and Sugioka, 1993; Chang et al., 2000); however, others suggest that retrograde axonal atrophy dose not play a role in reduced CV (Stevens, 1997; Wilson, 1998). Possible explanations for this controversy are the use of inappropriate methods, or incorrect interpretations. There are major differences in the results, which are secondary to different study populations. In severe CTS patients, wrist compression may have a markedly large effect on the nerve fibers in the carpal tunnel, and allow the unimpaired branches passing outside the carpal tunnel to dominate in the measurement of W–Emix. Therefore, W– Emix might only reflect the CV of the nerve fibers outside the carpal tunnel. If a group with severe axonal degeneration and lower CMAP amplitude were to be selected, it would appear that mixed nerve CV and motor or sensory forearm CV do not equally decrease with a relative sparing of the mixed nerve CV, when compared with control. This finding would suggest that conduction block, rather than retrograde axonal atrophy, is the major cause of proximal slowing. If a mildly afflicted group were to be tested, the opposite results
1239
Fig. 3. Linear regression of the elbow–palm conduction velocity (E–P) by mixed nerve stimulation (E–Pmix) versus sensory nerve stimulation (E– Psen) in patients and controls.
would be obtained, with a parallel reduction in W–Esen and W–Emix. Therefore, Hansson argued that the direct measurement of forearm mixed nerve CV does not reflect the actual CV of damaged nerve fibers through the carpal tunnel, but rather the CV of nerve fibers in the forearm that do not pass through the carpal tunnel. He recommended that the motor and sensory CVs, but not the mixed nerve CV, in the forearm be used to estimate possible retrograde impairment in CTS (Hansson, 1994). In his study, he did not measure the wrist–palm CV or elbow–palm CV for mixed and sensory nerves. In the present study, the correlations between mixed and sensory CVs were good in each segment, among the normal control. However, in the CTS patients, W– Emix was markedly faster than W–Esen, suggesting that in a diseased state, W–Emix is dominated by the undamaged nerve fibers and W–Esen reflects the actual CV of the nerve fibers through the carpal tunnel. Furthermore, E– Pmix correlated well with E–Psen, implying E–Pmix would be preferred over W–Emix for the evaluation of retrograde axonal atrophy.
1240
M.-H. Chang et al. / Clinical Neurophysiology 113 (2002) 1236–1240
If a conduction block of the large myelinating fibers at the wrist, with only slower conduction axons persisting through the carpal tunnel, is the primary cause of decreased forearm CV, the reduction in wrist–palm CV should parallel the reduction in wrist–elbow CV. If retrograde axonal atrophy is the primary cause, the reduction in wrist–elbow CV would not correlate with the reduction in wrist–palm CV, depending on the severity of the retrograde axonal atrophy. Based on our results, the reduction in wrist–palm CV was more marked than the reduction in wrist–elbow CV, suggesting that retrograde axonal atrophy is the more likely cause of the decrease in forearm CV. This finding was consistent with animal studies (Dyck et al., 1981; Devor and Govrin-Lippmarn, 1986). Retrograde axonal atrophy increases the electrical resistance to longitudinal current flow, resulting in a slowing of the proximal CV. We conclude that a direct measurement of forearm mixed nerve CV, would only reflect the CV of the nerves fibers passing outside the carpal tunnel, and that for an assessment of retrograde axonal atrophy, mixed or sensory conducting techniques through the wrist are more appropriate. Furthermore, retrograde axonal atrophy is the most likely explanation for the forearm conduction slowing in CTS, based on our findings. Acknowledgements This study was supported by grants NSC 89-2314-B075A-029, NSC90-2314-B-075A-005, NSC91-2314-B075A-004, TCVGH-903402C and TCVGH-903405N and TCVGH-913404C, awarded to Dr Ming-Hong Chang. References Anderson MJ, Fullerton PM, Gilliatt RW, Hern JE. Changes in the forearm associated with median nerve compression at the wrist in the guinea pig. J Neurol Neurosurg Psychiatry 1970;33:70–79. Buchthal F, Rosenfalck A. Sensory conduction from digit to palm and from palm to wrist in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1971;34:243–252.
Chang MH, Chiang HT, Ger LP, Yang DA, Lo YK. The cause of slowed forearm median conduction velocity in carpal tunnel syndrome. Clin Neurophysiol 2000;111:1039–1044. Buchthal F, Rosenfalck A, Trojaborg W. Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 1974;37:340–360. Dyck PJ, Lais AC, Karnes JL, Sparks M, Hunder H, Windebank AJ. Permanent axotomy, a model of axonal atrophy and secondary segmental demyelination and remyelination. Ann Neurol 1981;9:575–583. Devor M, Govrin-Lippmarn R. Retrograde slowing of conduction in sensory axons central to a sciatic nerve neuroma. Exp Neurol 1986;92:522–532. Felsenthal G, Spindler H. Palmar conduction time of median and ulnar nerves of normal subjects and patients with carpal tunnel syndrome. Am J Phys Med 1979;25:131–138. Hansson S. Does forearm mixed nerve conduction velocity reflect retrograde changes in carpal tunnel syndrome? Muscle Nerve 1994;17:725– 729. Jackson DA, Clifford JC. Electrodiagnosis of mild carpal tunnel syndrome. Arch Phys Med Rehabil 1989;70:199–204. Kimura J. A method for determining median nerve conduction velocity across the carpal tunnel. J Neurol Sci 1978;38:1–10. Kimura J. The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 1979;102:619–639. Melvin JA, Schuchmann JA, Lanese RR. Diagnostic specificity of motor and sensory nerve conduction variables in carpal tunnel syndrome. Arch Phys Med Rehabil 1973;54:69–74. Mills KR. Orthodromic sensory action potential from palmar stimulation in the diagnosis of carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1985;48:250–255. Pease WS, Lee HH, Johnson EW. Forearm median nerve conduction velocity in carpal tunnel syndrome. Electromyogr Clin Neurophysiol 1990;30:299–302. Stoehr M, Petruch F, Scheglmann K, Schilling K. Retrograde changes of nerve fibers with the carpal tunnel syndrome. J Neurol 1978;218:287– 297. Stevens JC. The electrodiagnosis of carpal tunnel syndrome. Muscle Nerve 1997;20:1447–1486. Thomas PK, Fullerton PM. Nerve fiber size in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 1963;26:520–527. Thomas JE, Lambert EH, Cseuz KA. Electrodiagnostic aspects of the carpal tunnel syndrome. Arch Neurol 1967;16:635–641. Uchida Y, Sugioka Y. Electrodiagnosis of retrograde changes in carpal tunnel syndrome. Electromyogr Clin Neurophysiol 1993;33:55–58. Wilson JR. Median mixed nerve conduction studies in the forearm: evidence against retrograde demyelination in carpal tunnel syndrome. J Clin Neurophysiol 1998;15:541–546.