Conclusions The modified segmental palmar test is a sensitive, robust and easily applicable method in diagnosing CTS.

Transcription

1

2

3 Abstract Objective To test our hypothesis that comparing the sensory nerve conduction velocity of the median nerve across the wrist with that of the forearm is more sensitive than comparing it with that of the palm in the electrodiagnostic confirmation of carpal tunnel syndrome (CTS). Methods 157 consecutive patients with clinically defined CTS were prospectively included and electrophysiologically examined. Antidromic nerve conduction velocities were measured in 3 segments of the median nerve: forearm, wrist, and palm. Differences and ratios in nerve conduction velocities were computed between the forearm and wrist and between the palm and wrist segments. Results Comparing the median nerve conduction velocities of the forearm with the wrist segment provides a greater sensitivity (79.6% and 82.8% for the second and third digit, respectively) than comparing the palm with the wrist segment (65.6% and 65.0%). Applying the ratio leads to slightly higher sensitivities for both comparisons. Conclusions The modified segmental palmar test is a sensitive, robust and easily applicable method in diagnosing CTS. Significance We recommend to use the median nerve sensory conduction velocity in the forearm as a reference in the segmental palmar test instead of that in the palm.

4 Introduction Decrease in nerve conduction velocity (NCV) of the median nerve across the carpal tunnel is the most sensitive test to confirm the clinical diagnosis of carpal tunnel syndrome (CTS). Abnormalities in more than a single test are needed to confirm the clinical diagnosis of CTS. 1 It is most common to compare sensory nerve conduction velocities of the median nerve across the wrist segment to nerve segments which are presumed to be normal, and therefore can be used as a reference. Comparison of the NCV of median nerve fibers across the carpal tunnel with those distal from the carpal tunnel (PALM test) is a sensitive test. 2,3 However, this specific test has several disadvantages. Because of the short conduction distance, the inherent small distance between stimulus electrodes and recording electrodes often introduces uncontrollable stimulus artifacts, which prevent accurate measurements of the latencies of the sensory nerve action potentials (SNAPs). On the other hand, conduction slowing over a small segment is more easily detected when applying small conduction distances. However, at the same time, this introduces larger measuring inaccuracies than in larger conduction distance. 4 More importantly, however, and not withstanding appropriately controlled skin temperature, we often find conduction slowing in segments distal from the carpal tunnel in CTS patients. As a consequence, in such cases this segment may be less suitable as a normal reference. In contrast, conduction slowing in sensory median nerve fibers over the forearm segment is rarely found. 3 The aim of the present study was to elucidate a more sophisticated variant of the traditional segmental palmar test with a potentially higher diagnostic yield. Therefore we reassessed the diagnostic value of the segmental palmar test in a prospectively conducted study to test which segment is best to be used as a reference. 3 We tested our hypothesis that comparing the NCV of the more proximal segment of the median nerve (i.e. forearm) with the palm to wrist segment is a sensitive and reliable test in the electrodiagnostic confirmation of CTS. Therefore, we determined the value and sensitivity of the segmental test of the median nerve in CTS patients, in which the NCV in the palm to wrist segment was compared to the NCV in the wrist to elbow segment. 49

5 Chapter 3 The segmental palmar test in CTS Materials and Methods Subjects In this prospective cohort study, 157 consecutive patients with clinically defined carpal tunnel syndrome were included. Carpal tunnel syndrome was considered clinically present in patients with pain and/or paresthesias in and restricted to the sensory distribution of the median nerve (involvement of the fifth finger was an exclusion criterion), and if patients met 2 or more of the following criteria: (1) nocturnal paresthesias, (2) reproduction or aggravation of paresthesias or pain by provocative tests (Tinel or Phalen s sign), (3) aggravation of paresthesias by activities such as driving a car, riding a bike, holding a book, or holding a telephone, (4) paresthesias relieved by shaking the hand. These clinical criteria have previously been used in other studies. 5 Patients were not included in this study in the following cases: clinical signs of polyneuropathy or known hereditary neuropathy with liability to pressure palsy, history of trauma or previous surgery of the symptomatic wrist, pregnancy, severe atrophy of the abductor pollicis brevis muscle, history of rheumatoid arthritis or arthrosis of the wrist, known diabetes, thyroid disease, or alcoholism. In case of bilateral complaints compatible with clinical CTS, only the most symptomatic hand was included. All candidates gave their written informed consent, and the study was approved by the local medical ethics committee. Electrodiagnostic evaluation All patients and healthy volunteers underwent standardized nerve conduction studies (NCS) in accordance with our laboratory s standard procedure in carpal tunnel syndrome. NCS were performed using a Viking Myograph type IV (Nicolet Biomedial Inc, Madison, WI, USA). Skin temperature at the site of the recording electrode was maintained at a minimum of 31.0 C, by means of hot packings, and was measured before and after each test by means of an infrared thermometer (62 Mini IR thermometer, Fluke Biomedical, Cleveland OH, USA). Electrophysiological studies were all performed by the same examiner, who was blinded to the results of the preceding physical examination. In all patients, sensory as well as motor nerve conduction studies were performed. Ring electrodes were applied for recording sensory nerve action potentials (SNAP). The proximal electrode was placed at the first interphalangeal joint and the distal recording electrode at a distance of at least 3 cm, if feasible. The ground electrode was placed between the proximal recording electrode and 50

6 the stimulation site and, if necessary, repositioned in order to reduce stimulus artifacts. The hand was manually fixed by the examiner to reduce movement artifacts. Stimuli with a duration of 0.3 ms were applied and stimulus current was adjusted in order to obtain supramaximal stimulus conditions. The optimal stimulation site was carefully determined. In case of disturbing stimulus artifacts, i.e. stimulus artifacts interfering with the observed potentials, repositioning of the anode without changing the position of the cathode by turning the stimulator was performed, and stimulus duration or current was adjusted in order to try to reduce the stimulus artifact. Signal averaging was applied on all SNAPs at least 3 times or, if necessary, several responses more in order to obtain an acceptably sharp potential take-off from baseline. Sensory nerve conduction studies comprise comparing the onset latencies between the median nerve and the ipsilateral ulnar (D4) and radial nerve (D1) at the 3 Figure 1. Schematic diagram of segments in segmental palmar test. R II & R III, recording site second and third digit; S p, stimulation palm; S w, stimulation wrist; S e, stimulation elbow; FOREARM, S e - S w ;WRIST: S w - S p ; PALM: S p R or S - R ; Black circle: cathode; White circle: anode; Hatched II p III square: ground electrode. 51

7 Chapter 3 The segmental palmar test in CTS same conduction distance, as well as segmental antidromic sensory conduction studies of the median nerve across the wrist. In the latter, digit 2 and digit 3 were used for recording with ring electrodes. Conduction distances were measured with a precision of 1 mm using a measuring tape. Distances were measured in a straight line from the recording electrode at the first interphalangeal joint to the stimulation site at the palm and subsequently the stimulation site at the wrist, which is just medial to the flexor carpi radialis tendon. The stimulation site at the palm was exactly halfway the distance from interphalangeal joint to the wrist for both digit 2 and 3. Finally, the distance to the stimulation site at the cubital fossa was measured (Figure 1). Subsequently, nerve conduction velocities of the three separate segments were computed. Motor nerve conduction studies were performed by stimulating the median nerve at the wrist and at the cubital fossa. Compound muscle action potentials (CMAPs) were recorded from the thenar eminence by means of surface electrodes, at a recording distance of 6 cm from the stimulation site at the wrist. Recording position was chosen in a way that enabled recording a CMAP as maximal as possible with a sharp initial negative deflection. A distal motor latency (DML) of > 4.0 ms is considered to be consistent with CTS. Finally, the terminal latency index (TLI) was calculated by the following equation: TLI = terminal distance/(motor conduction velocity forearm * DML). Sensory nerve conduction velocities were computed for the following segments of the median nerve: from elbow to wrist (FOREARM), from wrist to a point halfway wrist to base of the phalanx (WRIST), from halfway the wrist to the base of the phalanx to the proximal recording ring electrode (PALM), and from wrist to the proximal recording electrode (HAND). Next, the difference in NCV as well as the ratio between the tested segments were computed. According to Padua et al., 2 the distal-proximal ratio was calculated as the ratio of the NCV of the PALM and WRIST segment. Furthermore, the ratio of the WRIST and FOREARM segment, as well as that of the HAND and FOREARM segment was computed. All reference values for the nerve conduction studies were derived from a control group consisting of 47 healthy, asymptomatic, volunteers tested in the same laboratory. Statistical Analysis Data concerning clinical variables and nerve conduction studies were processed using Microsoft Office Excel and Access 2003 and all statistical analyses were 52

8 performed using SPSS Statistics Comparison between patients and controls was performed with a t-test for continuous variables or a χ 2 test for categorical variables, as appropriate. Mean differences, standard deviation, and upper and lower limits of normal (ULN and LLN, respectively) were calculated for the reference group for all nerve conduction studies. ULN and LLN are defined as the mean plus or minus twice the standard deviation, respectively. The number of patients with an abnormal electrodiagnostic test result was determined using the upper or lower limit of normal, derived from the reference group. Finally, the sensitivity of each test was calculated as the number of patients with a positive electrodiagnostic test and clinical CTS/number of patients with clinical CTS x 100%. 3 Results Study population The clinical characteristics of the patients are presented in Table 1. Forty seven hands of 47 control subjects were studied: 17 males and 30 females. Twenty-three right and 24 left hands were studied. One hundred fiftyseven patients were included in this study: 35 males and 122 females. Mean duration of symptoms was 12 months. Table 1. Clinical features in patients and reference group Patients n = 157 Reference group n = 47 Women 122 (77.7%) * 30 (63.8%) * Age (mean ± SD, years) ± 13.7 * ± 12.2 * Median symptom duration (months) 12.0 NA Wrist included left/right 71(45.2%)/ 86(54.8%) 24(51.1%)/ 23(48.9%) Atrophy of abductor pollicis brevis muscle 36 (22.9%) 0 Sensory loss 121 (77.1%) - Weakness abductor pollicis brevis muscle 46 (29.3%) - Weakness opponens pollicis muscle 10 (6.4%) - * P

9 Chapter 3 The segmental palmar test in CTS Table 2. Electrophysiological features in patients and reference group Reference group (n = 47) Patients (n = 157) Mean ± SD ULN/LLN Mean ± SD # Positive tests Sensitivity (%) D ± 0.19 * ± 0.72 * D ± 0.14 * ± 0.96 * PALM2 F-W 5.50 ± 5.73 * ± 10.8 * PALM2 P-W ± 6.21 * ± 10.3 * NCV ratio F/W ± 0.11 * ± 0.53 * NCV ratio P/W ± 0.12 * ± 0.44 * PALM3 F-W 5.64 ± 5.96 * ± 10.8 * PALM3 P-W 0.63 ± 6.32 * ± 10.3 * NCV ratio F/W ± 0.11 * ± 0.52 * NCV ratio P/W ± 0.11 * ± 0.46 * DML 3.36 ± 0.32 * ± 1.87 * TLI 0.32 ± 0.03 * ± 0.06 * * P 0.01; LLN applies to TLI only. D1: median radial latency difference; D4: median ulnar latency difference; PALM2 F-W and PALM3 F-W: NCV difference forearm-wrist recorded from second and third digit respectively; PALM2 P-W and PALM3 P-W: NCV difference palm-wrist recorded from second and third digit respectively; NCV ratio F/W2 and NCV ratio F/W3: ratio NCV forearm/wrist recorded from second and third digit respectively; NCV ratio P/W2 and NCV ratio P/W3: ratio NCV palm/wrist recorded from second and third digit respectively; DML: distal motor latency; TLI: terminal latency index. 54

10 Table 3. Distal sensory nerve conduction velocities recorded from second and third digit in patients Segment Second digit Third digit SNAP NCV * (m/s) SNAP NCV * (m/s) PALM unrecordable 22 (14.0%) - - unrecordable 25 (15.9%) - - recordable 135 (86.0%) (70.4%) recordable 132 (84.1%) (66.7%) < (29.6%) < (33.3%) WRIST unrecordable 30 (19.1%) - - unrecordable 32 (20.4%) - - recordable 127 (80.9%) (23.6%) recordable 125 (79.6%) (22.4%) < (76.4%) < (77.6%) HAND unrecordable 29 (18.5%) - - unrecordable 31 (19.7%) - - recordable 128 (81.5%) (27.3%) recordable 126 (80.3%) (25.4%) < (72.7%) < (74.6%) FOREARM unrecordable 35 (22.3%) - - unrecordable 41 (26.1%) - - recordable 122 (77.8%) (96.7%) recordable 116 (73.9%) (96.6%) < (3.3%) < (3.4%) * The cutoff value is equal to the lower limit of normal for the distal nerve conduction velocity, computed for the reference group (mean 2SD). 3 55

11 Chapter 3 The segmental palmar test in CTS Electrophysiology Details on electrophysiological features in patients and reference group are presented in Table 2. After recording from the second and third digit, 40 (25.5%) and 44 (28.0%) of the patients respectively, had a decreased NCV of the distal (PALM) segment. In 22 patients (14.0%) a SNAP was not recordable from the second digit; in 25 (15.9%) patients this was not possible from the third digit after stimulation at the palm. Four patients had a decreased NCV of the forearm segment. For further details see Table 3. Table 4. Comparison of sensitivity of different tests in patients Recording site and test ULN #Positive Tests Sens (%) Digit 2 NCV-difference PALM WRIST 12.2 m/s FOREARM WRIST 17.0 m/s FOREARM HAND 14.0 m/s NCV-ratio PALM / WRIST FOREARM / WRIST FOREARM / HAND Digit 3 NCV-difference PALM WRIST 13.3 m/s FOREARM WRIST 17.6 m/s FOREARM HAND 13.7 m/s NCV-ratio PALM / WRIST FOREARM / WRIST FOREARM / HAND

12 Comparing the nerve conduction velocities of the FOREARM with the WRIST segment provides a greater sensitivity (79.6% and 82.8%, respectively for the second and third digit) than comparison of nerve conduction velocities of the PALM and WRIST segment (65.6% and 65.0%). The same goes for comparison of nerve conduction velocities of the FOREARM with the HAND segment (77.7% and 80.9% for the second and third digit, respectively), as well as for the ratio between these two segments (85.4% and 86.6%). Furthermore, the ratio between the FOREARM and WRIST and the ratio between FOREARM and HAND are slightly more sensitive than the ratio of the PALM and WRIST segment (Table 4). 3 Discussion Conduction slowing in median nerve fibers across the carpal tunnel is the electrodiagnostic hallmark of CTS. Classically, for diagnosis, the NCV in this segment is compared to that in reasonably presumed normal nerve segments outside of the carpal tunnel. In one such test the sensory NCV of fibers across the WRIST are compared with fibers distal to the carpal tunnel. We reassessed a modification of the classic PALM test, 2 using the palm segment with the forearm as a reference. We found that the diagnostic yield is higher if one uses the forearm segment instead of the palm segment as a normal reference; this applies both to absolute difference in NCV and the ratio. Our explanation for this finding is, that in a substantial number of CTS patients conduction slowing occurs in median nerve fibers distal of the carpal tunnel. In severe cases of CTS, sensory, 3,6 mixed 7,8 or motor nerve conduction slowing 6,9,10 in the forearm segment, occurs to a much lesser extent. Likewise, comparing the NCV in the segment from wrist to fingers, i.e. in the hand to the NCV of the forearm, appeared to have no significant lower yield in confirming clinical CTS. Obviously, the latter method has several advantages, such as not having the problem of the stimulus artifacts, as was mentioned before. Moreover, since the conduction distance is larger, the measurement error is smaller. Therefore, judgment of conduction slowing of median nerve fibers across the carpal from wrist to fingers, using the forearm segment as a reference, is a sensitive, robust and easily applicable method. Comparing ratios has a slightly higher sensitivity than comparing differences in NCV. We have no explanation for the occurrence of conduction slowing in CTS patients distal to the carpal tunnel. Of course one may speculate on compression induced axonal dysfunction or diameter reduction in distal segments 11 resulting 57

13 Chapter 3 The segmental palmar test in CTS in conduction slowing, as has also been suggested by Chang et al. 9 as a cause of slowed mixed NCV of the forearm in severe cases of CTS. Conduction slowing caused by low skin temperature was ruled out by thoroughly warming the skin prior to the test procedure and measuring skin temperature of the palm and fingers after the test. Moreover, this phenomenon is encountered in our daily practice. Polyneuropathy as a cause of distal conduction slowing was ruled out anamnestically and by neurological examination. Moreover, distal sensory NCV of the ulnar nerve was normal in every patient. Since patients with polyneuropathy were not included in the present study, it was not possible to accurately calculate specificities for the nerve conduction studies. However, in patients with polyneuropathy conduction slowing in distal nerve segments usually is more profound than in proximal segments. Therefore, using the NCV in the forearm segment of the median nerve as a reference compared to NCV of distal segments of ulnar or radial nerve is likely to be more useful in CTS patients with polyneuropathy. In the present study, the number of patients with decreased NCV distal to the carpal tunnel is different from previously reported numbers. In the past, Buchthal et al. 11 and Le Quesne et al. 12 reported decreased distal NCV in 55% and 55.6% of patients, respectively. More recently, Padua et al. 2 reported decreased distal NCV in just 1 out of 50 (2%) patients with carpal tunnel syndrome. Differences may partly be explained by different techniques. For example, Buchthal et al. 11 used near-nerve needle techniques. Moreover, in all three studies, 2,11,12 nerve conduction studies were performed orthodromically by digital stimulation, instead of antidromical stimulation at the palm and wrist. According to our data, the median-ulnar sensory latency comparison test (D4) showed a high sensitivity as well (see Table 2). Some authors recommend to reserve this test for mild cases of CTS only (i.e. those cases in which other nerve conduction studies are normal). 13 However, despite of this, recently Werner et al. 6 still recommended that in all cases of suspected CTS median sensory and ulnar or radial sensory latencies in the same hand should be measured and compared. Otherwise, it is known that comparison tests like D4 can produce false-positive results. 13 By performing multiple tests one may encounter a falsepositive test result in one of the tests. 1 Therefore we applied in our criteria the presence of at least two abnormal test results to confirm the diagnosis of CTS electrophysiologically. Using at least two sensitive tests increases the likelihood of confirming CTS. Hence we believe the modified segmental palmar test to be a useful addition in diagnosing CTS. Data not obtained (Table 3) was never related to uncontrollable stimulus 58

14 artifacts. Every effort was made to eliminate the stimulus artifacts: repositioning of the anode without changing the position of the cathode, averaging of the responses, adjustment of stimulus duration or current, replacement of the ground electrode, and scrubbing and cleaning the skin at stimulation and recording sites. A higher number of SNAPs was not recordable at the second and third digit after stimulating the median nerve at the cubital fossa than after stimulating the nerve at the palm or wrist: the amplitude of the SNAP was too low to distinguish it from baseline noise. Therefore, in these cases, it was not possible to compute forearm nerve conduction velocities. The existence of rare sensory anastomoses in the forearm has to be considered as a potential pitfall when using the forearm as a reference. 14 Instead, since a functional ramus communicans from the ulnar fourth to the median third common digital nerves is far more frequent, 15 in our opinion it would make more sense to use the forearm as a reference in the segmental palmar test. Another fact worth mentioning is that, according to the data of the present study, the segmental palmar test recording from the third digit shows equal, or even higher, sensitivities than that from the second digit. We suggest that different innervation between the second and third digit might be an explanation for this finding. Unlike the third digit, the second digit is partially innervated by sensory fibers of the radial nerve. It may be that the fibers of the radial nerve are excited by stimulating at the wrist and palm due to co-stimulation. The SNAP recorded from the second digit may therefore be partially a radial nerve SNAP. The nerve conduction velocities of the radial nerve fibers are presumed to be normal in carpal tunnel syndrome. The nerve conduction velocity of the forearm remains the same. Therefore, the difference between NCV of the forearm and wrist segment recorded from the second digit may be not as large as recorded from the third digit. It remains important to mention that the present study was not designed nor powered to demonstrate a difference in sensitivity between second or third digit recording. 3 In conclusion, we would like to recommend to use the median nerve sensory NCV FOREARM as a reference instead of that of the PALM; to use the absolute NCV difference, since measurement errors may be increased by applying ratios; and, to use solely the third digit for recording. 59

15 Chapter 3 The segmental palmar test in CTS References 1. Redmond MD, Rivner MH. False positive electrodiagnostic tests in carpal tunnel syndrome. Muscle Nerve 1988; 11: Padua L, Lo Monaco M, Valente EM, Tonali PA. A useful electrophysiologic parameter for diagnosis of carpal tunnel syndrome. Muscle Nerve 1996; 19: Buchthal F, Rosenfalck A, Trojaborg W. Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 1974; 37: Van Dijk JG, Meulstee J, Zwarts MJ, Spaans F. What is the best way to assess focal slowing of the ulnar nerve? Clin Neurophysiol 2001; 112: Witt JC, Hentz JG, Stevens JC. Carpal tunnel syndrome with normal nerve conduction studies. Muscle Nerve 2004; 29: Werner RA, Andary M. Electrodiagnostic evaluation of carpal tunnel syndrome. Muscle Nerve 2011; 44: Chang MH, Wei SJ, Chiang HL, Wang HM, Hsieh PF, Huang SY. Forearm mixed nerve conduction velocity: questionable role in the evaluation of retrograde axonal atrophy in carpal tunnel syndrome. J Clin Neurophysiol 2003; 20: Chang MH, Lee YC, Hsieh PF. The role of forearm mixed nerve conduction study in the evaluation of proximal conduction slowing in carpal tunnel syndrome. Clin Neurophysiol 2008; 119: Chang MH, Lee YC, Hsieh PF. The real role of forearm mixed nerve conduction velocity in the assessment of proximal forearm conduction slowing in carpal tunnel syndrome. J Clin Neurophysiol 2008; 25: Chang MH, Liu LH, Lee YC, Hsieh PF. Alteration of proximal conduction velocity at distal nerve injury in carpal tunnel syndrome: demyelinating versus axonal change. J Clin Neurophysiol 2008; 25: 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: Le Quesne PM, Casey EB. Recovery of conduction velocity distal to a compressive lesion. J Neurol Neurosurg Psychiatry 1974; 37: Oh SJ. Nerve conduction studies. In: Clinical Electromyography. 3rd Edition. Philadelphia: Lippincott Williams & Wilkins; Hopf HC. Forearm ulnar-to-median nerve anastomosis of sensory axons. Muscle Nerve 1990; 13:

16 15. Loukas M, Bellary SS, Tubbs RS, Shoja MM, Gadol AAC. Deep palmar communications between the ulnar and median nerves. Clin Anat 2011; 24: