Terminal Latency Index, Residual Latency, and Median-Ulnar F-Wave Latency Difference in Carpal Tunnel Syndrome.

INTRODUCTION: Carpal tunnel syndrome (CTS) is the most common entrapment neuropathy, but no electrodiagnostic test alone shows sufficient sensitivity for CTS. We aimed to investigate the value of median motor terminal latency index (mTLI), median motor residual latency (mRL), and median-ulnar F-wave latency difference (FdifMU) as additional tests to nerve conduction studies which are performed traditionally in electromyography laboratories.
METHODS: This is a retrospective study. The results of electrodiagnostic studies performed on patients with CTS were examined. We divided the enrolled hands of the patients diagnosed with CTS into two groups: affected hands with abnormal electroneuromyographic parameters indicating CTS diagnosis (CTS group) and hands with normal electroneuromyographic parameters (control group). Then, we analyzed the results of these completed electrodiagnostic studies.
RESULTS: A total of 320 hands of 160 patients were studied. FdifMU and mRL were found to be significantly higher in CTS group compared with the control group (P < 0.001). mTLIs were found to be significantly higher in control group compared with the CTS group (P < 0.001). Given that, the area under the curve is more than 70% for mTLI and mRL, but not for FdifMU.
CONCLUSION: When combined with mMDL, both mTLI and mRL have excellent specificity for the diagnosis of mild and moderate CTS. However, the sensitivities for both parameters were lower. In suspected patients, FdifMU can be an additional tool for the diagnosis of CTS also, but alone it is not valuable.

INTRODUCTION

Carpal tunnel syndrome (CTS) is the most common entrapment neuropathy with an increasing rate in incidence.[1] No test alone shows sufficient sensitivity for the CTS diagnosis; due to previous studies for electrodiagnostic studies in CTS, it has been reported that the sensitivity of the conventional tests ranges from 49% to 84% with specificities of 95 or greater.[23] The severity of the CTS could also be determined electrophysiologically using nerve conduction studies (NCSs).[4] It has been reported that median motor terminal latency index (mTLI), median motor residual latency (mRL), and median-ulnar F-wave latency difference (FdifMU) give additional information to conventional electrodiagnostic studies; however, there are few studies that investigate these parameters in terms of sensitivity and specificity for the diagnosis of CTS. In this study, we aimed to investigate the value of mTLI, mRL, and FdifMU as additional tests to NCS which are performed traditionally in electromyography (EMG) laboratories for the diagnosis of CTS.

METHODS

Study population

Electrodiagnostic reports of patients with a clinical suspicion of CTS diagnosis for one of their hands were obtained from an electronic database of electrodiagnostic study results performed between January 2008 and March 2015. We retrospectively analyzed the results of these completed electrodiagnostic studies. Medical records of patients who underwent electroneuromyographic studies with an initial diagnosis of CTS were examined. We divided the enrolled hands of the patients clinically and electrophysiologically diagnosed with CTS into two groups: affected hands with abnormal electroneuromyographic parameters indicating CTS diagnosis (CTS group) and unaffected hands with normal electroneuromyographic parameters (control group). Exclusion criteria were having any suspicion of CTS diagnosis on the other hand, another mononeuropathy, polyneuropathy, severe diabetes mellitus, cervical radiculopathy or plexopathy, upper extremity trauma or fracture severe neurological disorder, a history of surgery for upper extremities, or any abnormal ulnar or radial nerve response. Also, the following information was recorded: patient age, gender, and nerve conduction parameters.

Nerve conduction studies

“Medelec Synergy” EMG device (Oxford Instruments, Surrey, England) was used for all patients. All electrophysiological examinations were conducted at temperatures above 25°C, and the distal skin temperature of each patient was measured at the hand dorsum for maintaining temperatures above 32°C. For all patients, conventional motor and sensory conduction studies of the median and ulnar nerves and also sensory conduction studies of the radial nerve were performed.

In conduction studies, filter settings were 3–10 kHz for motor conduction studies and 20–2 kHz. Using 9-mm disc surface cup (Ag/AgCl) electrodes (TECA Accessories; Medelec, Oxford Instruments, Old Woking, United Kingdom) which placed over the motor point of the abductor pollicis brevis muscle for the median nerve and over the abductor digiti minimi muscle for the ulnar nerve, the compound muscle action potentials (CMAP) for both nerves were recorded. For median motor NCS, active electrode was placed on the abductor pollicis brevis motor point with the reference electrode placed distally, stimulation has been made 8 cm proximal to the active electrode in the distal forearm, and a second stimulus was given at the antecubital fossa. Median nerve distal motor latencies (mMDL), CMAP amplitudes, and median motor nerve conduction velocities (mMNCVs) at the forearm and wrist were calculated. Sensory NCSs were performed both antidromically. For antidromic study, the median nerve was stimulated at the wrist 12 cm from the active electrode using ring electrodes from the third finger. For minimum F-wave latencies, at least 20 supramaximal stimulation periods were performed. Latencies are expressed in milliseconds (ms), CMAP amplitudes as millivolts (mV), and sensory nerve action potential amplitudes as microvolts (μV). Nerve conduction velocities (NCVs) were calculated as m/s. The same techniques were used for both hands in all patients. All of the hands in the control group demonstrated normal median nerve conduction parameters.

Terminal latency index (mTLI) was calculated as follows: mTLI = Distal nerve conduction distance (80 mm)/(proximal motor conduction velocity (MCV) × distal motor latency). A value of <0.35 is considered as abnormal.

Residual latency (mRL) was calculated as follows: mRL = Distal motor latency − (distal nerve conduction distance (80 mm)/proximal MCV). A value of >2.55 is considered as abnormal.

The difference of the minimal F-wave latencies between median and ulnar nerves (FdifMU) was calculated as follows: FdifMU = Minimal F-wave latency of median nerve (minimum mFWL) − minimal F-wave latency of ulnar nerve (minimum uFWL).

Statistical analysis

Data were analyzed using SPSS version 24.0 for Windows (IBM Corporation, Armonk, NY, USA). P < 0.05 was considered statistically significant. Continuous variables were shown as mean ± standard deviation, and categorical variables were shown as the number of cases and percentage of total (%). For analysis of continuous variables across groups, Kolmogorov–Smirnov test was used. The comparisons between the groups were calculated using Student's t-test or the nonparametric Mann–Whitney U-test. The compliance statistics of categorical data of electrophysiological diagnostic methods were analyzed by McNemar's test and Cohen's kappa test. Sensitivity and specificity of the variables were based on receiver operating characteristic (ROC) curve analysis.

RESULTS

Electrophysiological records of 320 hands of 160 patients (148 women and 12 men) were investigated. The mean age of patients was 54.44 ± 11.68 years. Table 1 shows the results of the electrophysiological examination of 160 hands from the CTS group and 160 hands from the control group.

Table 1

Comparison of electrophysiological findings between groups

	Mean±SD		P
	CTS group (n=160)	Control group (n=160)
mMDL	3.93±0.44	3.41±0.29	<0.001*
mMNCV	56.73±4.36	57.84±4.52	0.027*
mSDL	3.76±0.46	3.16±0.36	<0.001**
mSNCV (wrist-digit)	32.46±4.47	38.65±4.49	<0.001**
mCMAP	7.70±3.14	8.29±5.85	0.296**
Wrist-to-palm NCV	33.02±4.31	40.22±3.36	<0.001**
Wrist-to-palm latency	2.68±0.35	2.32±0.31	<0.001**
Minimum mFWL	25.88±1.56	24.97±1.66	<0.001**
Minimum uFWL	24.79±1.74	24.80±1.68	0.932*
Mean mFWL	26.63±1.73	25.72±1.67	<0.001*
FdifMU	1.07±1.42	0.16±1.28	<0.001*
mTLI	0.36±0.04	0.40±0.04	<0.001*
mRL	2.49±0.45	2.02±0.29	<0.001*

*Student’s t-test, **Mann–Whitney U-test. mMDL=Median motor distal latency, mMNCV=Median motor nerve conduction velocity, mSDL=Median sensory distal latency, mSNCV=Median sensory nerve conduction velocity, mCMAP=Median compound muscle action potential amplitude, NCV=Nerve conduction velocity, mFWL=Median F-wave latency, uFWL=Ulnar F-wave latency, FdifMU=Median-ulnar F-wave latency difference, mTLI=Median motor terminal latency index, mRL=Median motor residual latency, CTS=Carpal tunnel syndrome, n=Number of hands

mMDL (P < 0.001), mMNCV (P = 0.027), median sensory distal latency (mSDL) (P < 0.001), median wrist-to-palm segment latency (P < 0.001), minimum mFWL (P < 0.001), mean mFWL (P < 0.001), FdifMU (P < 0.001), and mRL (P < 0.001) were found to be significantly higher in the CTS group compared with the control group (P < 0.05) [Table 1]. On the other hand, median sensory nerve conduction velocity (P < 0.001), median wrist-to-palm segment NCV (P < 0.001), and mTLI (P < 0.001) were found to be significantly higher in the control group compared with the CTS group (P < 0.05) [Table 1].

The abnormality criteria of the diagnostic tests and normal ranges are shown in Table 2. mTLI was <0.35 (the lower limit of the normal) in 36.3% (58) hands and mRL was >2.55 (the higher limit of the normal) in 41.3% hands in the CTS group [Table 2].

Table 2

Comparison of sensitivities for different diagnostic tests

	Abnormality criteria	The number of abnormality
		CTS group (n=160), n (%)	Control group (n=160), n (%)
mTLI	<0.35	58 (36.3)	15 (9.4)
mRL	>2.55	66 (41.3)	5 (3.1)

mTLI=Median motor terminal latency index, mRL=Median motor residual latency, CTS=Carpal tunnel syndrome, n=Number of hands

When the sensitivity of all diagnostic tests was examined by McNemar's test, it was found that the results of all couplings were statistically significant (P < 0.001). The most sensitive coupling was mMDL–mSDL (93%) and the most specific coupling was mRL–mSDL (99.6%) [Table 3].

Table 3

Compliance statistics for electrophysiological methods

	PPV (%)	NPV (%)	P	Cohen’s kappa	P
mTLI - mRL	78.1	94.3	0.856*	0.731	<0.001**
mTLI - mMDL	58.9	94.3	0.023*	0.577	<0.001**
mTLI - mSDL	74.0	61.4	<0.001*	0.259	<0.001**
mRL - mMDL	78.9	99.6	0.001*	0.844	<0.001**
mRL - mSDL	85.9	64.5	<0.001*	0.362	<0.001**
mMDL - mSDL	93.0	63.2	<0.001*	0.345	<0.001**

*Mc Nemar test, **Cohen’s kappa test. PPV=Positive predictive value, NPV=Negative predictive value, mTLI=Median motor terminal latency index, mRL=Median motor residual latency, mMDL=Median motor distal latency, mSDL=Median sensory distal latency

Given that the area under the curve (AUC) is more than 70% for mTLI and mRL, both mTLI and mRL can distinguish between the groups and have diagnostic reliability. The accepted abnormality criteria for our outpatient clinic laboratory were <0.35 for mTLI and >2.55 for mRL; however, due to this study, it was found that the optimum cutoff points for distinguishing between the groups were 0.38 for mTLI and 2.22 for mRL [Table 4]. The AUC was 0.688 for FdifMU [Table 4].

Table 4

Results of receiver operating characteristic curve analysis test of median motor terminal latency index, median motor residual latency, and median-ulnar F-wave latency difference

	Diagnostic test					ROC curve		P
	Cut-off	Sensitivity	Specificity	PPV	NPV	AUC	95% CI
mTLI	<0.38	65.60	73.80	71.40	68.20	0.765	0.714-0.816	<0.001*
mRL	≥2.23	72.50	73.10	72.95	72.67	0.816	0.770-0.861	<0.001*
FdifMU	≥0.63	63.10	67.50	66.01	64.67	0.688	0.630-0.746	<0.001*

*P<0.05. PPV=Positive predictive value, NPV=Negative predictive value, AUC=Area under the Curve, CI=Confidence interval, mTLI=Median motor terminal latency index, mRL=Median motor residual latency, FdifMU=Median-ulnar F-wave latency difference, ROC=Receiver operating characteristic

The ROC curve was distinguished the diagnostic reliability between CTS and control groups for mTLI, mRL, and FdifMU [Figure 1].

Figure 1

Receiver operating characteristic curve diagram for assessing the diagnostic legitimacy of median motor terminal latency index, median motor residual latency, and median-ulnar F-wave latency difference between carpal tunnel syndrome group and control group

DISCUSSION

mTLI and mRL are electrophysiological parameters for identifying abnormalities in the distal segment of motor nerves.[56] Especially in patients with high clinical suspicion of CTS, but without abnormality in the routine NCS, mTLI can help to early diagnosis of CTS.[7] Furthermore, there will be no need to an additional electrical stimulation, so there will be advantages in the means of time consumption and patient comfort.[7] In a study by Park et al., 212 hands of 106 CTS patients were studied and it was found that mTLI and FdifMU can be used to assess the stages of CTS severity.[8] On the other hand, various investigations were made for both mTLI and the FdifMU for the diagnosis of the early stage of CTS; however, sensitivity and specificity are variable between different studies.[91011] In our study, we found that mTLI was abnormal in 36.3% (58) hands of CTS patients; however, when combined with mRL, it has a sensitivity of 78.1% and excellent specificity of 94.3%. When we combined mTLI and mMDL, we determined a sensitivity of 58.9% and an excellent specificity of 94.3%. Furthermore, AUC was more than 70% for mTLI and the difference was statistically significant between the groups (P < 0.05). Hence, it can be concluded that mTLI is an unreliable tool for the diagnosis of CTS alone, but when combined with other parameters, it can be a more useful and valuable tool.

In the literature, it has been mentioned that mRL can be used as an electrodiagnostic parameter for early diagnosis of distal peripheral neuropathies, but conflicting results have been obtained in terms of sensitivity and specificity in the diagnosis of CTS.[12] In a study by Khosrawi and Dehghan, it has been found that in mild cases of CTS in which traditional NCS shows abnormalities only in sensory studies, mRL can demonstrate the effect on median nerve motor fibers better.[12] Also, Kraft and Halvorson concluded that mRL can be useful in confirming early or mild CTS,[13] and on the other hand, Kuntzer mentioned that although mRL had a high sensitivity for the diagnosis of CTS, the specificity for these parameters was low.[14] In our study, we found that mRL was abnormal in 41.3% (66) of hands of CTS patients; however, when combined with mMDL, it has a sensitivity of 78.9% and an excellent specificity of 99.6%. Furthermore, AUC was >70% for mRL and the difference was statistically significant between the groups (P < 0.05). As mTLI, mRL is an unreliable tool for the diagnosis of CTS alone and it would be better to combine with other parameters for suspected patients.

The increase in minimum F-wave latency is a good indicator, which shows the conduction delay along a peripheral nerve.[15] The demyelinating injuries of peripheral nerves could be well evaluated with minimum F-wave latency measurement, but using F-wave studies in entrapment neuropathies remains a challenging issue.[15] In Sander et al.'s study, 79 hands of 50 CTS patients were studied and found 75%–78% sensitivity for FdifMU in the diagnosis of CTS.[16] Furthermore, similar results were demonstrated in some studies.[1718] As a conflicting result, Joshi et al. found that the FdifMU had low sensitivity and specificity for the diagnosis of CTS.[19] Alemdar conducted a study with 210 upper extremities of 114 individuals and found that FdifMU has a higher diagnostic efficacy than minimum mFWL on the diagnosis of CTS.[15] In comparison with these studies, our sample size was higher and we found that FdifMU was significantly higher in CTS group compared with the control group (P < 0.05); however, we could not calculate the sensitivity of this parameter because we had not a cutoff value. But, in our analysis, we found that the optimum cutoff for FdifMU was 0.62 (≥0.63 was considered as abnormal). AUC was not more than 70% for FdifMU, but it was nearly to this value (68.8%) and the difference was statistically significant between the groups (P < 0.05). Hence, it can be concluded that FdifMU can be a valuable tool for the diagnosis of CTS. On the other hand, we did not recorded the heights of patients; it is well known that F-wave latencies are affected by the height of the patients.[20] Because of these conflicting results, using FdifMU for the diagnosis of CTS is a challenging issue.

There are some limitations for our study. First, we did not record the height of the patients because F-wave latency values can be affected by height; there can be underestimated or overdiagnosed patients. Second, we did not determine the clinical severity due to the retrospective nature of our study; however, we included mild and moderate CTS patients due to the electrodiagnostic values. There were no severe CTS patients in our study population, so we cannot generalize our results for all CTS patients. Third, we could not compare the value of mTLI and mRL on CTS diagnosis with the other comparison techniques such as combined sensory index and F-wave persistence due to retrospective nature of our study. As a power of our study, this study which examines the mTLI, mRL, and FdifMU in the diagnosis of CTS has the largest sample size with a control group to the best of our knowledge. Furthermore, since the control group constituted the unaffected hands of the patients, NCSs were not affected by factors such as age, gender, height, and weight.

CONCLUSION

When combined with mMDL, both mTLI and mRL have excellent specificity for the diagnosis of mild and moderate CTS. However, the sensitivities for both parameters were lower. In suspected patients, FdifMU can be an additional tool for the diagnosis of CTS, but alone it is not valuable. If there is a clinical suspicious for CTS, mTLI and mRL could be good indicators for determining CTS patients, especially when combined with mMDL.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.