In patients with suspected carpal tunnel syndrome (CTS), the findings on nerve conduction studies (NCS) during electrodiagnostic (EDX) testing help to confirm the diagnosis and provide information regarding the pathophysiology and severity of nerve damage. Patients with mild CTS typically demonstrate features of focal demyelination, whereas those with more severe or longstanding nerve compression often exhibit features of axonal loss. Various EDX grading scales have been reported to grade the severity of CTS.1,2 Although these scales focus primarily on motor and sensory NCS abnormalities, some also incorporate needle electromyography (EMG) findings in thenar muscles. The presence of fibrillation potentials or neurogenic motor unit potentials (MUPs) provides evidence of loss of axons and reinnervation, and suggests a more severe stage of nerve injury. However, the utility of thenar muscle needle EMG in CTS and the degree to which it provides useful information beyond the NCS are controversial, and practices vary in the inclusion of needle EMG of thenar muscles.3,4
The aim of this study was to prospectively determine the frequency of needle EMG abnormalities in the thenar muscles in patients with less severe CTS, defined by an absolutely normal median motor compound muscle action potential (CMAP) amplitude. We aimed to compare patients using a simpler cutoff measure of whether the median motor distal latency (DL) was below or above the absolute laboratory reference value. We hypothesized that when the median motor NCSs are normal or only demonstrate a prolonged DL, the yield of finding needle EMG abnormalities in thenar muscles would be low.
Patients referred to the EMG laboratory in which the EDX study demonstrated findings of only CTS were prospectively studied. Patients with any other peripheral nervous system or neuromuscular disorder (e.g., other mononeuropathy, radiculopathy, polyneuropathy, myopathy, motor neuron disease, etc.) on the basis of abnormalities of any other NCS apart from the median nerves or needle EMG abnormalities in any nonthenar muscle were excluded.
To exclude patients with more severe CTS, only subjects with a median neuropathy at the wrist in which the median motor CMAP amplitude was >4.0 mV (which is the lower limit of normal for our laboratory) were included. Patients were stratified into two groups solely on the basis of the median motor DL, those with an absolutely normal median motor DL (<4.5 ms) (NL-DL) and those with a prolonged median motor DL (≥4.5 ms) (ABN-DL). Patients with low-median CMAP amplitudes (<4.0 mV) were expected to have more severe disease and likely needle EMG abnormalities, and were not evaluated in this study.
In patients included in the study, needle EMG of a thenar muscle (abductor pollicis brevis or opponens pollicis) was performed by one of the investigators, both of whom are experienced electromyographers. Standard needle EMG technique was used for all patients, including at least 2 passes through the muscle while assessing the findings at several sites with each pass. Needle EMG findings recorded included the presence and grade of fibrillation potentials, MUP recruitment, and long duration and/or polyphasic MUPs or MUPs with increased proportion of turns. Assessment of fibrillation potentials, recruitment, and MUP duration and phases was made semiquantitatively and abnormalities graded on a 0 (none) to 4 (severe) scale. Fibrillation potentials were graded as 0 if none were present, “±” if only one or two fibrillation potentials were present in a single location in the muscle, “1+” if a few fibrillation potentials were recorded in more than one area of muscle. Grades of “2+” to “4+” indicate a higher density of fibrillation potentials in most areas of the muscle. Motor unit potential duration was also graded according to a 0 (normal) to 4+ (severely long duration and high amplitude) scale based on the degree of abnormalities from the mean of the parameters. The percent of MUPs that were polyphasic (>4 phases) was recorded (0, 25%, 50%, 75%, or 100%). The number of MUPs with more than three turns was graded as 0 (<20%), 1+ (21%–30%), 2+ (31%–40%), 3+ (41%–50%), or 4+ (>50%). The study was approved by the IRB.
Statistical analysis comparing NCS parameters in the ABN-DL CTS patients with needle EMG abnormalities to those without needle EMG abnormalities was performed using the Mann–Whitney U test. The statistical significance level was set at P < 0.05.
Median neuropathy at the wrist was identified in 154 patients of EMGs performed by the authors during the study period; 50 with NL-DL, 52 with ABN-DL, and 52 with low CMAP amplitudes (and were not included in the study). Table 1 shows the demographics and EDX findings in patients with NL-DL and ABN-DL. In the NL-DL group, orthodromic (palmar) sensory studies were performed in 41/50 patients, whereas the other nine underwent antidromic sensory studies. In the ABN-DL group, antidromic sensory studies were performed in 49/52 patients, whereas the other three had orthodromic (palmar) sensory studies. Nine patients in the ABN-DL group had absent sensory responses. The mean median motor DLs were significantly longer (P = 0.0001) and motor amplitudes lower (P = 0.0124) in the ABN-DL patients compared with NL-DL patients.
Abnormalities on needle EMG of the abductor pollicis brevis or opponens pollicis were found in 23/102 (22.5%) of all patients. In those with NL-DL, 6/50 (12%) demonstrated minimal, equivocal abnormalities. One patient was found to have a single fibrillation potential in one area of the muscle, but no fibrillation potentials were found in any other patient. Five patients demonstrated slightly increased proportion of MUPs with turns, but no patients in this group demonstrated abnormal MUP recruitment, duration, or phases.
In patients with ABN-DL, 17/52 (32.6%) had abnormalities in the abductor pollicis brevis or opponens pollicis muscles. Eight patients (15.4%) had fibrillation potentials, which were graded as ± in five and 1+ in three. Thirteen patients (25.0%) demonstrated abnormalities of MUP, which included mildly long duration in five, mildly reduced recruitment in three, and 25% to 50% polyphasic MUP in 5.
There was no statistically significant difference in median sensory DLs or amplitudes, or median motor DLs in the ABN-DL patients with needle EMG abnormalities compared to those without needle EMG abnormalities. However, patients with needle EMG abnormalities were significantly older (mean 64.7 vs. 54.0 years, P = 0.0295) and had significantly lower CMAP amplitudes compared to those without needle EMG abnormalities. (6.7 vs. 8.7 mV, P = 0.0057).
Needle EMG of thenar muscles in CTS provides complementary information to NCS regarding localization to the distal median nerve and determining chronicity and pathophysiology of nerve injury. However, it remains questionable as to whether performing needle EMG in thenar muscles, which is more painful than in other muscles, for the assessment of median nerve function when CTS is already identified on NCSs provides additional value.4,5 Although the severity and treatment decisions in CTS are often based on the patient's clinical features, these may not correlate with the degree of demyelination or loss of axons within the nerve. In some cases, patients with subjectively severe symptoms will only demonstrate minimal features of demyelination, whereas other asymptomatic patients will demonstrate evidence of severe axonal loss. In patients with severe CTS and low-median CMAP amplitudes, the degree of axonal loss is reflected by the presence of fibrillation potentials in the thenar muscles and the chronicity and extent of reinnervation by the degree of reduction in recruitment and MUP enlargement. Identifying the degree of axonal loss may be valuable in determining treatment options and may help to predict the extent and rate of potential recovery after intervention.
With milder nerve compression in which only sensory NCS abnormalities or focal slowing in the distal median motor axons are present, abnormalities on needle EMG in thenar muscles would not be expected. However, a normal median CMAP amplitude may still be reduced for an individual patient. In addition, when the median nerve is compressed gradually over time with ongoing reinnervation, a loss of motor units may occur with preservation of the CMAP amplitude. Therefore, in those cases, needle EMG may still provide evidence of axonal loss that is not apparent on NCS.
Several studies assessing the electrophysiologic findings in CTS have attempted to correlate NCS abnormalities and thenar muscle needle EMG.6–11 Some reviewed the combined needle EMG findings and NCS abnormalities in a cohort of patients with CTS, whereas others used statistical analysis to determine NCS cutoff values that would best predict the presence of fibrillation potentials and/or MUP changes in NCS. Kimura and Ayyar6 reviewed 639 symptomatic extremities with CTS and found significantly lower mean CMAP amplitudes (4.32 vs. 6.78 mV) in patients with compared to those without fibrillation potentials. Although they did not specifically assess patients with “mild” or “moderate” CTS, only 13.3% of 279 patients with a distal motor latency <4.7 ms demonstrated fibrillation potentials and 18.6% demonstrated reduced recruitment. Similarly, Wee retrospectively reviewed 84 patients with CTS and also found that abnormal MUPs correlated with the CMAP amplitude. Fibrillation potentials were found in only 6% of patients in whom the mean median CMAP amplitude was very low (1.9 mV). The CMAP amplitudes were normal (10.5 and 8.1 mV) in patients with no or mild MUP changes in the abductor pollicis brevis, and no information regarding needle findings in patients with mild CTS was given.7
Murga et al.8 found that the frequency of fibrillation potentials or MUP changes on needle EMG in 228 patients with CTS correlated with the degree of median motor DL prolongation, but the median motor CMAP amplitudes were not assessed. Nearly all patients had an abnormal needle EMG if the DL was >7 ms. Only 3.9% of patients had minor needle EMG abnormalities with normal DL, and of 23 patients with a mildly prolonged DL, 6 had an abnormal EMG.
Vennix et al. retrospectively reviewed 1,362 patients with CTS examined over a 3.5-year period in multiple EDX laboratories and found that median motor amplitudes and median motor DLs were independent parameters that predicted denervation in patients with CTS, although they did not analyze MUP parameters. In their cohort, 10.9% demonstrated fibrillation potentials and they found that performing needle EMG only in patients with a median CMAP amplitude <7 mV would identify more than 95% of patients with denervation. In addition, to identify 95% of patients with denervation, all cases with a median motor latency >4.5 ms would need to be examined.9
In retrospectively reviewing 480 cases of EMG confirmed CTS to determine whether the findings of axonal loss on needle EMG could be predicted from the NCS findings, Werner and Albers found needle EMG abnormalities in 48% (22% with fibrillation potentials), although they did not specifically separate out the percentage of patients with mild CTS. They found a significant difference in the median motor CMAP amplitudes and DLs in patients with fibrillation potentials compared to those without (9.0 vs. 7.5 mV and 4.4 vs. 5.1 ms, respectively). The severity of fibrillations correlated with the degree of reduction of CMAP amplitude. However, patients with borderline, mild, or moderate degrees of fibrillation potentials all had CMAP amplitudes in normal range (9.6, 8.6, and 7.9 mV, respectively), indicating that signs of denervation may occur in patients with preserved CMAP amplitudes. They calculated a relative risk of finding fibrillation potentials of 2.4 when the median motor latency was >5.1 ms but also found that estimating whether thenar needle EMG would be abnormal from NCS findings was inaccurate up to 43% of the time.10
More recently, Chang et al.11 studied 229 patients with CTS to determine whether the findings of spontaneous activity in thenar muscles could be predicted from NCS. Their CTS population was patients with moderate to severe CTS based on electrophysiologic criteria, as 96% had abnormal motor DLs and 43% had a low CMAP amplitudes. They found that the DLs and CMAP amplitudes were significantly worse in patients with fibrillation potentials than those without and that the CMAP amplitudes were the best predictor of the presence or absence of fibrillation potentials as over 95% of patients with a median CMAP amplitude less than 2.1 mV demonstrated fibrillation potentials and less than 6% of patients with CMAP amplitudes greater than 4.9 mV demonstrated fibrillation potentials.
Although the previous studies found that the median DL and CMAP amplitude values, in general, correlated with the presence of fibrillation potentials, they did not specifically assess patients with less severe CTS. Furthermore, although they identified cutoff values that predicted abnormal thenar needle EMG, the values varied by study. It would be helpful for the electromyographer to have a simple NCS measure that can help to determine whether thenar muscle needle EMG would likely demonstrate abnormalities, and therefore whether to or not to examine these muscles. In our study, we prospectively performed needle EMG in thenar muscles to specifically assess for fibrillation potentials and MUP abnormalities, but focused on only those patients with less severe CTS based on absolutely normal median CMAP amplitudes, and separated according to a simple measure of the absolute laboratory reference value of the upper limit of normal for the median motor DL. We found that patients with a normal median motor DL had essentially no significant abnormalities on needle EMG. The absence of findings in these patients is consistent with the absence of significant injury to motor axons. However, approximately one-third of CTS patients with prolonged median motor DLs but normal amplitudes demonstrated evidence of motor axonal loss and reinnervation, although to a very mild degree. Although these patients might be considered to have more severe nerve injury, the loss of just a single axon can lead to fibrillation potentials in the few denervated muscle fibers innervated by that axon. Therefore, the simple presence of fibrillation potentials should not necessarily increase the degree of severity.
This study has several limitations. First, the number of patients in this study is relatively small, which may limit interpreting clinical significance of the findings. Furthermore, although needle EMG abnormalities were found in nearly 1/3 of patients with ABN-DL, the findings were very mild and would suggest only involvement of relatively few axons. In addition, although only patients with absolutely normal median CMAP amplitudes were included in the study, an absolutely normal CMAP amplitude may actually be low for the individual patient (e.g., if their “normal” CMAP was 10 mV, a CMAP of 4.1 mV would be a reduction), and, if this was known, patients would actually be classified as having more severe CTS. Finally, the patients with ABN-DL and needle EMG abnormalities were significantly older than those without needle abnormalities. However, there is no evidence in the literature to support the presence of fibrillation potentials in hand muscles solely due to increased age, and grading of MUP abnormalities of duration and recruitment is relative to patient age. It is possible that the higher proportion of polyphasic MUPs could be, in part, due to older age.
A recent practice parameter on the EDX assessment of CTS suggested that needle EMG is not necessary if the physician is not concerned about another process (e.g., radiculopathy or plexopathy) involving the upper extremity.12 Based on our findings, we would agree that patients who have evidence of CTS with no involvement of the median motor NCS are unlikely to demonstrate prominent abnormalities on needle EMG of thenar muscles, and needle EMG of the thenar muscles is not necessary. However, in patients with CTS in which the median motor DL is prolonged, needle EMG should be considered, as it may provide value in indicating some axonal loss despite a normal median CMAP amplitude. Whether the presence or absence of axonal loss on needle EMG provides useful prognostic information or guidance, the type of intervention (e.g., surgical vs. conservative treatment) or outcome requires further studies.
1. Stevens JC. AAEM minimonograph #26. The electrodiagnosis of carpal tunnel syndrome
. Muscle Nerve 1997;20:1477–1486.
2. Sucher BM. Grading severity of carpal tunnel syndrome
in electrodiagnostic reports: why grading is recommended. Muscle Nerve 2013;48:331–333.
3. Mazur A, Wertsch JJ. Use of thenar
electromyography in the evaluation of carpal tunnel syndrome
. Muscle Nerve 1995;18:1722.
4. Mazur A. Role of thenar
electromyography in the evaluation of carpal tunnel syndrome
. Phys Med Rehabil Clin N Am 1998;9:755–764.
5. Gnatz SM, Conway RR. Issues and opinions: the role of needle electromyography in the evaluation of patients with carpal tunnel syndrome
. Muscle Nerve 1999;22:282–286.
6. Kimura I, Ayyar DR. The carpal tunnel syndrome
: electrophysiological aspects of 639 symptomatic extremities. Electromyogr Clin Neurophysiol 1985;25:151–164.
7. Wee AS. Needle electromyography in carpal tunnel syndrome
. Electromyogr Clin Neurophysiol 2002;42:253–256.
8. Murga L, Moreno JM, Menendez C, Castialla JM. The carpal tunnel syndrome
. Relationship between median distal motor latency and graded results of needle electromyography. Electromyogr Clin Neurophysiol 1994;34:377–384.
9. Vennix MJ, Hirsh DD, Chiou-Tan FY, Rossi C. Predicting acute denervation in carpal tunnel syndrome
. Arch Phys Med Rehabil 1998;79:306–312.
10. Werner RA, Albers JW. Relation between needle electromyography and nerve conduction studies in patients with carpal tunnel syndrome
. Arch Phys Med Rehabil 1995;76:246–249.
11. Chang CW, Lee WJ, Liao YC, Chang MH. Which nerve conduction parameters can predict spontaneous electromyographic activity in carpal tunnel syndrome
? Clin Neurophysiol 2013;124:2264–2268.
12. Werner RA, Andary M. Electrodiagnostic evaluation of carpal tunnel syndrome
. Muscle Nerve 2011;44:597–607.