Median motor studies are commonly “normal” in mild carpal tunnel syndrome (CTS). This reflects either the sparing of motor compared to sensory fibers, or the inability of conventional studies to detect an abnormality. A novel approach to demonstrate early motor fiber involvement in CTS is the placement of the same active electrode lateral to the third metacarpal, allowing recording from the second lumbrical or the deeper interossei, when stimulating the median or ulnar nerves at the wrist, respectively. We compared the difference between these latencies in 51 normal control hands to 107 consecutive patient hands referred with symptoms and signs suggestive of CTS, who were subsequently proven to have electrophysiologic CTS by standard nerve conduction criteria. A prolonged lumbrical-interossei latency difference (>0.4 ms) was found to be a sensitive indicator of CTS in all patient groups. It was also helpful in patients with coexistent polyneuropathy, where localization at the wrist was otherwise difficult. 0 1992 John Wiley & Sons, Inc.
Key words: carpal tunnel syndrome lubrical interossei nerve conduction studies median nerve MUSCLE & NERVE 15:1253-1257 1992
LUMBRICAL AND INTEROSSEI RECORDING I N CARPAL T U N N E L DAVID C. PRESTON, MD, and ERIC L. LOGIGIAN, MD
Nerve conduction studies are of established value in the diagnosis of carpal tunnel syndrome (CTS). However, there are patients with symptoms and signs suggestive of CTS who remain difficult to diagnosis by standard electrophysiologic tests.’ In patients with CTS, the median distal motor latency is normal in 35% to 50%.528This lack ofsensitivity of distal motor latency may be due either to sparing of motor as compared to sensory fibers, or to the inability of standard median motor studies to detect an abnormality. A variety of nerve conduction studies have been employed to demonstrate mild CTS in these patients.‘- ’() A novel technique to demonstrate CTS involves comparison of distal motor latency between the second lumbrical (2L) and interossei (INT). Using the same active recording electrode placed
From the Departments of Neurology, Tufts-New England Medical Center (Dr. Preston), and Harvard-Brigham and Women’s Hospital (Dr. Logigian), Boston, Massachusetts. Presented in part at the 38th Annual Meeting of the AAEM, Vancouver, Canada, September 25-28, 1991 Address reprint requests to David C. Preston, MD, Division of Neurology, Brigham & Women’s Hospital, 75 Francis Street, Boston, MA 02115. Accepted for publication March 10, 1992 CCC 0148-639)(192/111253-05 0 1992 John Wiley & Sons, Inc.
LUM-INT
lateral to the third metacarpal, stimulation of the median or ulnar nerves at the wrist, allows recording from the 2L or deeper I N T , respectively.“ T h e advantages of this new technique include: 1. Both muscles can be recorded from the same active electrode in the distal palm. 2. T h e axons innervating both muscles are of similar diameter size. 3. Temperature is comparable for each distal nerve segment and muscle. 4. Identical distances to each muscle are used, allowing direct coniparison of distal motor latencies. Thus, 5. T h e 2L is relatively spared in even in severe CTS when the abductor pollicis brevis (APB) muscle is completely wasted, a compound muscle action potential (CMAP) can be reliably recorded from the 2L. Thus, this technique creates an ideal internal control for median motor studies in which several variables are held constant (muscle and axon size, temperature, and distance). The only difference between the median and ulnar studies is that one is via the carpal tunnel and the other is not. v
METHODS
We studied 107 consecutive patient hands, referred with clinical symptoms and signs suggestive of CTS (i.e., intermittent wrist and arm pain, par-
MUSCLE & NERVE
November 1992
1253
esthesias in the median distribution, nocturnal symptoms, Tinel's and Phalen's signs, etc.), who were subsequently proven to have electrophysiologic CTS by standard nerve conduction criteria. All patients had one or more abnormalities of the following standard median studies: Standard median studies
Abnormal value
1. Distal motor latency to APB (DML-APB) 2. Antidromic sensory latency to digit 2 (ASL-D2) (distance 13 cm) 3. Median mixed palmar latency (MPL) (distance 8 cm) 4. Difference between median and ulnar mixed palmar latencies ( M - U PL) (distances 8 cm)
>4.4 ms
>3.5 ms (peak) >2.2 ms (peak)
>0.3 ms
'The distal median motor and sensory latencies were chosen as they are historically the oldest and most commonly used studies.' In addition, the mixed palmar study was selected among many possible median-ulnar comparisons, as this is the study felt by many to be the most sensitive test for mild CTS.',4.8," In addition, 8 patients were studied with "possible CTS." 'These patients had symptoms and signs highly suggestive of CTS, but had normal standard median studies. Each had an abnormality of one of the following additional studies which suggested CTS: (a) antidrornic median sensory latency to digit IV prolonged as compared to the ulnar sensory latency to digit IV (difference >0.4 ms); (b) median sensory conduction velocity of the wrist- palm segment slower than the palm-digit I1 segment (difference >10 m/s); or (c) EMG (reinnervation in median muscles distal to the carpal tunnel; proximal median and nonmedian C8-T1 innervated muscles normal). To record the 2L and INT, the active electrode was placed slightly lateral to the midpoint of the third metacarpal (Fig. 1). The reference electrode was placed over a bony prominence of the proximal interphalangeal joint of the second digit.',:' The motor point to the 2L was identified by an initial negative deflection with the fastest rise time. Occasionally the median mixed nerve potential was seen prior to the motor response.3 However, this potential never obscured the initial negative deflection of the 2L CMAP. The median
1254
LUM-INT
FIGURE 1. Recording and stimulation sites. G1: active electrode, G2: reference electrode; S1: Median nerve stimulation site, and S2: ulnar nerve stimulation site. Equal distances are used between G1 -S1 and G1 -S2.
and ulnar nerves were stimulated at the wrist using identical distances from the active electrode. A supramaximal CMAP was recorded from each, and the differences between their distal latencies recorded (PL-INT DIFF). Sweep speed was 2 ms/ division, and the sensitivity was either 1 or 2 mV/ division (Fig. 2). The same sensitivity was always used to compare latencies within subjects. The 2L and INT were recognized by their characteristic shape, and care was taken to avoid costimulation of both nerves at the wrist. All patients were studied on either a Dantec Counterpoint or Neuromatic 2000 EMG machine. Surface recording with a 10-mm silver disk and ring electrodes was used. Palmar skin temperature was maintained above 31°C. Although the correlation between the degree of clinical symptoms and electrical abnormalities is not absolute, we based severity on electrical criteria. Patients were thus divided into groups (severe, moderate, mild, very mild, and possible) based on the least-sensitive standard study needed to demonstrate CTS (Table 3). That is, the more sensitive the study needed to demonstrate the CTS, the more severe the CTS was graded. Some of the CTS patients were also noted to have a polyneuropathy, if they had both clinical and electrophysiologic evidence of such (i.e., distal sensory loss, depressed ankle jerks, reduced sural sensory amplitudes, denervation of distal leg muscles). Com-
MUSCLE 8, NERVE
November 1992
C.
FIGURE 2. (A) Normal control; (B) mild CTS; (C) severe CTS. Top traces: median nerve stimulation, recording second lumbrical. Bottom traces: ulnar nerve stimulation, recording interossei.
parison of the lumbrical-interossei latency difference were made between CTS and control hands. The 51 controls were either healthy volunteers or patients referred to the EMG laboratory with symptoms not referable to CTS (i.e., back pain). They had neither clinical nor electrical evidence of CTS. Means and standard deviations (SD) of the following parameters were calculated for controls: (1) distal motor latency to the 2L and INT; (2) difference between these two latencies within each control subject’s hand; and (3) CMAP amplitude of the 2L and INT (baseline to negative peak). Statistical comparison of the distal latency to the 2L and the INT was done using a two-tailed Student’s t test. RESULTS
One-hundred seven patient hands met clinical and standard electrophysiologic criteria for CTS. The average age was 49 (range 21 to 98). The female/male ratio was 3 : 1. The right hand/left hand ratio was approximately 2 : 1. Twelve patients with diabetes had a coexistent polyneuropath y.
Patient Group.
Lumbrical and interossei recording was performed in 51 control hands to obtain normative data. The average age was 31 (range 18 to 50). The mean 2L-INT DIFF was 0.08 ms. As the ulnar nerve follows a slightly curved path to reach the INT in the palm, one might have expected that the distal latency to the INT would have been longer than the distal latency to the 2L when using identical linear distances. However, this was not the case, as no statistical difference between the distal latency to the 2L and the distal latency to the INT was found in controls. The mean +2SD and the upper limit of the range were both 0.40 ms. Therefore, an ab-
normal 2L-INT DIFF was set at >0.40 ms. There was no significant correlation between age and 2L-INT DIFF (r = -0.12). The mean amplitude of the INT was higher than that of the 2L (6.55 mV and 2.81 mV, respectively). Comparison of Sensitivities of Median Nerve Conduc-
In patients with CTS, the sensitivities of standard median conduction studies and the 2L-INT DIFF study were calculated. A study was considered abnormal if the latency was prolonged or the potential absent. Whereas absent responses were occasionally noted in the standard median conduction studies, the 2L and INT CMAP responses were obtainable from all patients. The DML- APB was the least-sensitive study (54%). The ASL-D2 was abnormal in 67%. The MPL sensitivity was higher at 82%. The most sensitive standard median study was clearly the M-U PL (94%).The 6 patients who did not meet the criteria for an abnormal M-U PL each had a borderline value of 0.3 ms, and an abnormality of one of the other standard median studies to establish the electrophysiologic diagnosis of CTS. The tion Studies (Table 2).
Control Group (Table 1).
LUM- INT
Table 1. Summary of lumbrical and interossei studies in controls. Value
Mean
Mean +2 SD
Range
2L-INT DIFF (ms) 2L DL* (ms) INT DL* (ms)
0.08 3.22 3.15
0.40 3.98 3.91
(-0.30)-0.40 2.10-4.00 2 .OO-4 .OO
2L AMP (mV) INT AMP (mV)
2.81 6.55
Mean -2 SD 0.40 1.67
1 .OO-6.00 2.60- 13.60
DlFF, difference; INT, interossei; ZL, second lumbrical; DL, distal latency; AMP, amplitude; mean distance wrist to active electrode, 9 5 cm (range 8- 12 crn). *Two-tailed t test, P = 0.53. NS.
MUSCLE & NERVE
November 1992
1255
Table 2. Comparison of sensitivities of median nerve conduction studies in patients with CTS (N = 107) N
Sensitivity (YO)
absent)
58
54%
absent)
72
67%
absent)
88
82%
absent)
101
94%
Abnormal study DML-APB (>4 4 ms or ASL-D2 (>3 5 ms or MPL (>2 2 ms or M-U PL (>O 3 rns or
dard median study was the M-U PL, the 2L-INT DIFF was abnormal in 87%. In patients with possible CTS who had normal standard median studies, 75% had an abnormal 2L-INT DIFF. In a subset of patients with polyneuropathy and CTS, 100% had an abnormal 2L-INT DIFF. In 5 of these patients, localization was otherwise difficult, as the median sensory potential was absent and motor latencies to all muscles (APB, 2L, and INT) were prolonged. CONCLUSIONS
2L-INT DIFF p 0 . 4 rns)
102
95%*
_________-
DML-APB, dfsfal motor latency to abductor pollicis brews, ASLLDP, antfdromic sensory latency to dfgit 2, MPL, medfan rnfxed-nerve palmar latency, M- U PL dffference between medfan and ulnar mixed-nerve palmar latencies, PLLlNT DIFF, difference between dfstal motor latency to the second lurnbncal and interossef "Of the 5 faflures of the 2L-INT DIFF, 3 had a borderlfne value of 0 4 ms
sensitivity ofthe 2L-INT DIFF (95%)was comparable to that of the M-U PL. Of the 5 patients who did not have an abnormal 2L-INT DIFF, 3 had a borderline value of 0.4 ms. Sensitivity of the PL-IN1 Groups with CTS (Table 3).
DlFF Study in Patient
Patients were separated into groups based on the least-sensitive standard median study needed to demonstrate CTS. Patients with electrophysiologically severe CTS had a prolonged DML-APB. The 2L-INT DIFF was abnormal in 98% of such patients. In patients with moderate CTS (normal DML-APB, but prolonged ASL-DP), the 2L-INT DIFF was abnormal in 100%. In mild CTS, where the MPL was prolonged and the DML-APB and AL-D2 were normal, the 2L-INT was abnormal in 83%. Last, in very mild CTS, where the only abnormal stan-
CTS is the most common entrapment neuropathy affecting the upper extremity, and is a frequent source of referral to the EMG laboratory.7v8AS there can be other peripheral and central nervous system causes of acroparesthesia, electrophysiologic confirmation of the diagnosis is im ortant, especially before consideration of surgery. A prolonged lumbrical- interossei latency difference (>0.4 ms) was found to be a sensitive indicator of CTS in all patient groups. Other studies have also found the use of internal controls between the median and ulnar studies to be high sensitive. Comparison with contralateral median studies often fails in this regard, as CTS is frequently bilateral, either clinically or ele~trically.~ Comparisons between median and ulnar digit IV sensory latencies, and mixed-nerve palm-to-wrist latencies, have been the most studied."43839The consensus among most electromyographers is that the comparison between the median and ulnar mixed nerve palm to wrist latencies is probably the best test to detect mild CTS.'*8,'0 In our study, the sensitivity o f the lumbrical- interossei difference was similar to that of the mixed-nerve palm-towrist latency difference. Many studies have noted the low diagnostic yield of the distal motor latency to abductor polli-
F:
Table 3.Sensitivity of the 2L-INT DlFF study in patient groups with CTS Patient group Severe CTS Prolonged or absent DML-APB Moderate CTS Normal DML-APB, prolonged or absent ASL-D2 Mild CTS Normal DML-APB and ASL-D2, prolonged or absent MPL Very mild CTS Normal DML-APB, ASL-D2 MPL, prolonged M-U PL Possible CTS Normal DML-APB, ASL-D2, MPL, and M-U PL with an additional study abnormal Median-ulnar sensory to D4 Sensory CV wrist-palrn/palrn-D2 EMG VI CTS with polyneuropathy I II 111 IV V
N
Abnormal 2L-INT DIFF
58/107 (54%) 22/107 (21%) 12/107 (11%) 15/107 (14%) 8
57158 (98%) 22/22 (100%) 10112 (83%) 13115 (87%) 6 (75%)
2 5 1 12/107 (11%)
2 (100%) 3 (60%) 1 (100%) 12/12 (100%)
~
Abbrevfatfons as fn Table 2
1256
LUM-INT
MUSCLE & NERVE
November 1992
this regard, and in this study was recordable in all patients. The lumbrical-interossei latency difference is a new and sensitive measurement to aid the electromyographer in confirming the electrophysiologic diagnosis of CTS. This technique is simple and easy to perform, requiring only that the recording electrode is over the motor point of the 2L. In addition, its value is especially noted in patients with coexistent polyneuropathy, where localization at the wrist is otherwise difficult.
REFERENCES
FIGURE 3. Relative sparing of the second lumbrical (2L). Although the distal motor latency to the second lumbrical is spared compared to the abductor pollicis brevis (APB), it is still prolonged when compared to the latency to the interossei (INT) in patients with CTS. Compare normal control (left) with a patient with CTS (right).
cis brevis. This is widely believed to be the result of relative sparing of motor as compared to sensory fibers in mild CTS. However, it is apparent from the present study that electrophysiologic evidence of motor involvement is much more frequent than commonly appreciated. Indeed, in rare cases of CTS, only motor fibers are involved.’ While previous studies have shown that the 2L is spared as compared to the APB in CTS, this sparing is only relative (Fig. 3).3 It is clear from the present study that with the use of an internal control, both the lumbrical and the APB are commonly affected, the latter more than the former. During standard median studies, the absence of a sensory or motor potential may confirm the impression of a median neuropathy; however, absent responses are not of localizing value. The lumbrical- interossei recording is not limited in
LUM-INT
1. Busis NA, Logigian EL, Shahani BT: Electrophysiological ssessment of a median nerve injury in the palm. Muscle Nerve 1986;9 :208 - 2 10.
2. Dauhe JR: Percutaneous palmar median nerve stimulation for carpal tunnel syndrome. Electroencephalogr Clin Neurophysiol 1977;43:139- 140. 3 . Logigian EL, Busis NA, Berger AR, et al.: Lumbrical sparing in carpal tunnel syndrome: anatomic, physiologic, and diagnostic implications. Neurology 1987;37: 1499- 1505. 4. Macdonell R, Schwartz MS, Swash M: Carpal tunnel syndrome: which finger should he tested? An analysis of sensory conduction in digital branches of the median nerve. Muscle Nerve 1990;13:601-606. 5 . Martinez AC: Diagnostic yield of different electrophysiologic methods in carpal tunnel syndrome. Muscle Nerve 1991;2:183- 184. 6. Pease WS, Cannell CD, Johnson EW: Median to radial latency difference test in mild carpal tunnel syndrome. Muscle Nerve 1989;12:905-909. 7. Rivner MH: Carpal tunnel syndrome: a critique of “newer” nerve conduction techndiques, in Focal Perzpheral Neuropathies: Selected Topics. AAEM 14th Annual Continuing Education Course, 1991, pp 19-24. 8. Stevens JC: AAEE Minimonograph #26: the electrodiagnosis of carpal tunnel syndrome. Mwcle Nerve 1987;10:99113. 9. Uncini A, Lange DJ, Solomon M, et al: Ring finger testing in carpal tunnel syndrome: a comparative study of diagnostic utility. Muscle Nerve 1989;12:735-741. 10. White JC, Hansen SR, Johnson RK: A comparison of EMG procedures in the carpal tunnel syndrome with clinicalEMG correlations. Mwcle Nerve 1988; 11: 1177- 1182.
MUSCLE & NERVE
November 1992
1257
Key words: carpal tunnel syndrome lubrical interossei nerve conduction studies median nerve MUSCLE & NERVE 15:1253-1257 1992
LUMBRICAL AND INTEROSSEI RECORDING I N CARPAL T U N N E L DAVID C. PRESTON, MD, and ERIC L. LOGIGIAN, MD
Nerve conduction studies are of established value in the diagnosis of carpal tunnel syndrome (CTS). However, there are patients with symptoms and signs suggestive of CTS who remain difficult to diagnosis by standard electrophysiologic tests.’ In patients with CTS, the median distal motor latency is normal in 35% to 50%.528This lack ofsensitivity of distal motor latency may be due either to sparing of motor as compared to sensory fibers, or to the inability of standard median motor studies to detect an abnormality. A variety of nerve conduction studies have been employed to demonstrate mild CTS in these patients.‘- ’() A novel technique to demonstrate CTS involves comparison of distal motor latency between the second lumbrical (2L) and interossei (INT). Using the same active recording electrode placed
From the Departments of Neurology, Tufts-New England Medical Center (Dr. Preston), and Harvard-Brigham and Women’s Hospital (Dr. Logigian), Boston, Massachusetts. Presented in part at the 38th Annual Meeting of the AAEM, Vancouver, Canada, September 25-28, 1991 Address reprint requests to David C. Preston, MD, Division of Neurology, Brigham & Women’s Hospital, 75 Francis Street, Boston, MA 02115. Accepted for publication March 10, 1992 CCC 0148-639)(192/111253-05 0 1992 John Wiley & Sons, Inc.
LUM-INT
lateral to the third metacarpal, stimulation of the median or ulnar nerves at the wrist, allows recording from the 2L or deeper I N T , respectively.“ T h e advantages of this new technique include: 1. Both muscles can be recorded from the same active electrode in the distal palm. 2. T h e axons innervating both muscles are of similar diameter size. 3. Temperature is comparable for each distal nerve segment and muscle. 4. Identical distances to each muscle are used, allowing direct coniparison of distal motor latencies. Thus, 5. T h e 2L is relatively spared in even in severe CTS when the abductor pollicis brevis (APB) muscle is completely wasted, a compound muscle action potential (CMAP) can be reliably recorded from the 2L. Thus, this technique creates an ideal internal control for median motor studies in which several variables are held constant (muscle and axon size, temperature, and distance). The only difference between the median and ulnar studies is that one is via the carpal tunnel and the other is not. v
METHODS
We studied 107 consecutive patient hands, referred with clinical symptoms and signs suggestive of CTS (i.e., intermittent wrist and arm pain, par-
MUSCLE & NERVE
November 1992
1253
esthesias in the median distribution, nocturnal symptoms, Tinel's and Phalen's signs, etc.), who were subsequently proven to have electrophysiologic CTS by standard nerve conduction criteria. All patients had one or more abnormalities of the following standard median studies: Standard median studies
Abnormal value
1. Distal motor latency to APB (DML-APB) 2. Antidromic sensory latency to digit 2 (ASL-D2) (distance 13 cm) 3. Median mixed palmar latency (MPL) (distance 8 cm) 4. Difference between median and ulnar mixed palmar latencies ( M - U PL) (distances 8 cm)
>4.4 ms
>3.5 ms (peak) >2.2 ms (peak)
>0.3 ms
'The distal median motor and sensory latencies were chosen as they are historically the oldest and most commonly used studies.' In addition, the mixed palmar study was selected among many possible median-ulnar comparisons, as this is the study felt by many to be the most sensitive test for mild CTS.',4.8," In addition, 8 patients were studied with "possible CTS." 'These patients had symptoms and signs highly suggestive of CTS, but had normal standard median studies. Each had an abnormality of one of the following additional studies which suggested CTS: (a) antidrornic median sensory latency to digit IV prolonged as compared to the ulnar sensory latency to digit IV (difference >0.4 ms); (b) median sensory conduction velocity of the wrist- palm segment slower than the palm-digit I1 segment (difference >10 m/s); or (c) EMG (reinnervation in median muscles distal to the carpal tunnel; proximal median and nonmedian C8-T1 innervated muscles normal). To record the 2L and INT, the active electrode was placed slightly lateral to the midpoint of the third metacarpal (Fig. 1). The reference electrode was placed over a bony prominence of the proximal interphalangeal joint of the second digit.',:' The motor point to the 2L was identified by an initial negative deflection with the fastest rise time. Occasionally the median mixed nerve potential was seen prior to the motor response.3 However, this potential never obscured the initial negative deflection of the 2L CMAP. The median
1254
LUM-INT
FIGURE 1. Recording and stimulation sites. G1: active electrode, G2: reference electrode; S1: Median nerve stimulation site, and S2: ulnar nerve stimulation site. Equal distances are used between G1 -S1 and G1 -S2.
and ulnar nerves were stimulated at the wrist using identical distances from the active electrode. A supramaximal CMAP was recorded from each, and the differences between their distal latencies recorded (PL-INT DIFF). Sweep speed was 2 ms/ division, and the sensitivity was either 1 or 2 mV/ division (Fig. 2). The same sensitivity was always used to compare latencies within subjects. The 2L and INT were recognized by their characteristic shape, and care was taken to avoid costimulation of both nerves at the wrist. All patients were studied on either a Dantec Counterpoint or Neuromatic 2000 EMG machine. Surface recording with a 10-mm silver disk and ring electrodes was used. Palmar skin temperature was maintained above 31°C. Although the correlation between the degree of clinical symptoms and electrical abnormalities is not absolute, we based severity on electrical criteria. Patients were thus divided into groups (severe, moderate, mild, very mild, and possible) based on the least-sensitive standard study needed to demonstrate CTS (Table 3). That is, the more sensitive the study needed to demonstrate the CTS, the more severe the CTS was graded. Some of the CTS patients were also noted to have a polyneuropathy, if they had both clinical and electrophysiologic evidence of such (i.e., distal sensory loss, depressed ankle jerks, reduced sural sensory amplitudes, denervation of distal leg muscles). Com-
MUSCLE 8, NERVE
November 1992
C.
FIGURE 2. (A) Normal control; (B) mild CTS; (C) severe CTS. Top traces: median nerve stimulation, recording second lumbrical. Bottom traces: ulnar nerve stimulation, recording interossei.
parison of the lumbrical-interossei latency difference were made between CTS and control hands. The 51 controls were either healthy volunteers or patients referred to the EMG laboratory with symptoms not referable to CTS (i.e., back pain). They had neither clinical nor electrical evidence of CTS. Means and standard deviations (SD) of the following parameters were calculated for controls: (1) distal motor latency to the 2L and INT; (2) difference between these two latencies within each control subject’s hand; and (3) CMAP amplitude of the 2L and INT (baseline to negative peak). Statistical comparison of the distal latency to the 2L and the INT was done using a two-tailed Student’s t test. RESULTS
One-hundred seven patient hands met clinical and standard electrophysiologic criteria for CTS. The average age was 49 (range 21 to 98). The female/male ratio was 3 : 1. The right hand/left hand ratio was approximately 2 : 1. Twelve patients with diabetes had a coexistent polyneuropath y.
Patient Group.
Lumbrical and interossei recording was performed in 51 control hands to obtain normative data. The average age was 31 (range 18 to 50). The mean 2L-INT DIFF was 0.08 ms. As the ulnar nerve follows a slightly curved path to reach the INT in the palm, one might have expected that the distal latency to the INT would have been longer than the distal latency to the 2L when using identical linear distances. However, this was not the case, as no statistical difference between the distal latency to the 2L and the distal latency to the INT was found in controls. The mean +2SD and the upper limit of the range were both 0.40 ms. Therefore, an ab-
normal 2L-INT DIFF was set at >0.40 ms. There was no significant correlation between age and 2L-INT DIFF (r = -0.12). The mean amplitude of the INT was higher than that of the 2L (6.55 mV and 2.81 mV, respectively). Comparison of Sensitivities of Median Nerve Conduc-
In patients with CTS, the sensitivities of standard median conduction studies and the 2L-INT DIFF study were calculated. A study was considered abnormal if the latency was prolonged or the potential absent. Whereas absent responses were occasionally noted in the standard median conduction studies, the 2L and INT CMAP responses were obtainable from all patients. The DML- APB was the least-sensitive study (54%). The ASL-D2 was abnormal in 67%. The MPL sensitivity was higher at 82%. The most sensitive standard median study was clearly the M-U PL (94%).The 6 patients who did not meet the criteria for an abnormal M-U PL each had a borderline value of 0.3 ms, and an abnormality of one of the other standard median studies to establish the electrophysiologic diagnosis of CTS. The tion Studies (Table 2).
Control Group (Table 1).
LUM- INT
Table 1. Summary of lumbrical and interossei studies in controls. Value
Mean
Mean +2 SD
Range
2L-INT DIFF (ms) 2L DL* (ms) INT DL* (ms)
0.08 3.22 3.15
0.40 3.98 3.91
(-0.30)-0.40 2.10-4.00 2 .OO-4 .OO
2L AMP (mV) INT AMP (mV)
2.81 6.55
Mean -2 SD 0.40 1.67
1 .OO-6.00 2.60- 13.60
DlFF, difference; INT, interossei; ZL, second lumbrical; DL, distal latency; AMP, amplitude; mean distance wrist to active electrode, 9 5 cm (range 8- 12 crn). *Two-tailed t test, P = 0.53. NS.
MUSCLE & NERVE
November 1992
1255
Table 2. Comparison of sensitivities of median nerve conduction studies in patients with CTS (N = 107) N
Sensitivity (YO)
absent)
58
54%
absent)
72
67%
absent)
88
82%
absent)
101
94%
Abnormal study DML-APB (>4 4 ms or ASL-D2 (>3 5 ms or MPL (>2 2 ms or M-U PL (>O 3 rns or
dard median study was the M-U PL, the 2L-INT DIFF was abnormal in 87%. In patients with possible CTS who had normal standard median studies, 75% had an abnormal 2L-INT DIFF. In a subset of patients with polyneuropathy and CTS, 100% had an abnormal 2L-INT DIFF. In 5 of these patients, localization was otherwise difficult, as the median sensory potential was absent and motor latencies to all muscles (APB, 2L, and INT) were prolonged. CONCLUSIONS
2L-INT DIFF p 0 . 4 rns)
102
95%*
_________-
DML-APB, dfsfal motor latency to abductor pollicis brews, ASLLDP, antfdromic sensory latency to dfgit 2, MPL, medfan rnfxed-nerve palmar latency, M- U PL dffference between medfan and ulnar mixed-nerve palmar latencies, PLLlNT DIFF, difference between dfstal motor latency to the second lurnbncal and interossef "Of the 5 faflures of the 2L-INT DIFF, 3 had a borderlfne value of 0 4 ms
sensitivity ofthe 2L-INT DIFF (95%)was comparable to that of the M-U PL. Of the 5 patients who did not have an abnormal 2L-INT DIFF, 3 had a borderline value of 0.4 ms. Sensitivity of the PL-IN1 Groups with CTS (Table 3).
DlFF Study in Patient
Patients were separated into groups based on the least-sensitive standard median study needed to demonstrate CTS. Patients with electrophysiologically severe CTS had a prolonged DML-APB. The 2L-INT DIFF was abnormal in 98% of such patients. In patients with moderate CTS (normal DML-APB, but prolonged ASL-DP), the 2L-INT DIFF was abnormal in 100%. In mild CTS, where the MPL was prolonged and the DML-APB and AL-D2 were normal, the 2L-INT was abnormal in 83%. Last, in very mild CTS, where the only abnormal stan-
CTS is the most common entrapment neuropathy affecting the upper extremity, and is a frequent source of referral to the EMG laboratory.7v8AS there can be other peripheral and central nervous system causes of acroparesthesia, electrophysiologic confirmation of the diagnosis is im ortant, especially before consideration of surgery. A prolonged lumbrical- interossei latency difference (>0.4 ms) was found to be a sensitive indicator of CTS in all patient groups. Other studies have also found the use of internal controls between the median and ulnar studies to be high sensitive. Comparison with contralateral median studies often fails in this regard, as CTS is frequently bilateral, either clinically or ele~trically.~ Comparisons between median and ulnar digit IV sensory latencies, and mixed-nerve palm-to-wrist latencies, have been the most studied."43839The consensus among most electromyographers is that the comparison between the median and ulnar mixed nerve palm to wrist latencies is probably the best test to detect mild CTS.'*8,'0 In our study, the sensitivity o f the lumbrical- interossei difference was similar to that of the mixed-nerve palm-towrist latency difference. Many studies have noted the low diagnostic yield of the distal motor latency to abductor polli-
F:
Table 3.Sensitivity of the 2L-INT DlFF study in patient groups with CTS Patient group Severe CTS Prolonged or absent DML-APB Moderate CTS Normal DML-APB, prolonged or absent ASL-D2 Mild CTS Normal DML-APB and ASL-D2, prolonged or absent MPL Very mild CTS Normal DML-APB, ASL-D2 MPL, prolonged M-U PL Possible CTS Normal DML-APB, ASL-D2, MPL, and M-U PL with an additional study abnormal Median-ulnar sensory to D4 Sensory CV wrist-palrn/palrn-D2 EMG VI CTS with polyneuropathy I II 111 IV V
N
Abnormal 2L-INT DIFF
58/107 (54%) 22/107 (21%) 12/107 (11%) 15/107 (14%) 8
57158 (98%) 22/22 (100%) 10112 (83%) 13115 (87%) 6 (75%)
2 5 1 12/107 (11%)
2 (100%) 3 (60%) 1 (100%) 12/12 (100%)
~
Abbrevfatfons as fn Table 2
1256
LUM-INT
MUSCLE & NERVE
November 1992
this regard, and in this study was recordable in all patients. The lumbrical-interossei latency difference is a new and sensitive measurement to aid the electromyographer in confirming the electrophysiologic diagnosis of CTS. This technique is simple and easy to perform, requiring only that the recording electrode is over the motor point of the 2L. In addition, its value is especially noted in patients with coexistent polyneuropathy, where localization at the wrist is otherwise difficult.
REFERENCES
FIGURE 3. Relative sparing of the second lumbrical (2L). Although the distal motor latency to the second lumbrical is spared compared to the abductor pollicis brevis (APB), it is still prolonged when compared to the latency to the interossei (INT) in patients with CTS. Compare normal control (left) with a patient with CTS (right).
cis brevis. This is widely believed to be the result of relative sparing of motor as compared to sensory fibers in mild CTS. However, it is apparent from the present study that electrophysiologic evidence of motor involvement is much more frequent than commonly appreciated. Indeed, in rare cases of CTS, only motor fibers are involved.’ While previous studies have shown that the 2L is spared as compared to the APB in CTS, this sparing is only relative (Fig. 3).3 It is clear from the present study that with the use of an internal control, both the lumbrical and the APB are commonly affected, the latter more than the former. During standard median studies, the absence of a sensory or motor potential may confirm the impression of a median neuropathy; however, absent responses are not of localizing value. The lumbrical- interossei recording is not limited in
LUM-INT
1. Busis NA, Logigian EL, Shahani BT: Electrophysiological ssessment of a median nerve injury in the palm. Muscle Nerve 1986;9 :208 - 2 10.
2. Dauhe JR: Percutaneous palmar median nerve stimulation for carpal tunnel syndrome. Electroencephalogr Clin Neurophysiol 1977;43:139- 140. 3 . Logigian EL, Busis NA, Berger AR, et al.: Lumbrical sparing in carpal tunnel syndrome: anatomic, physiologic, and diagnostic implications. Neurology 1987;37: 1499- 1505. 4. Macdonell R, Schwartz MS, Swash M: Carpal tunnel syndrome: which finger should he tested? An analysis of sensory conduction in digital branches of the median nerve. Muscle Nerve 1990;13:601-606. 5 . Martinez AC: Diagnostic yield of different electrophysiologic methods in carpal tunnel syndrome. Muscle Nerve 1991;2:183- 184. 6. Pease WS, Cannell CD, Johnson EW: Median to radial latency difference test in mild carpal tunnel syndrome. Muscle Nerve 1989;12:905-909. 7. Rivner MH: Carpal tunnel syndrome: a critique of “newer” nerve conduction techndiques, in Focal Perzpheral Neuropathies: Selected Topics. AAEM 14th Annual Continuing Education Course, 1991, pp 19-24. 8. Stevens JC: AAEE Minimonograph #26: the electrodiagnosis of carpal tunnel syndrome. Mwcle Nerve 1987;10:99113. 9. Uncini A, Lange DJ, Solomon M, et al: Ring finger testing in carpal tunnel syndrome: a comparative study of diagnostic utility. Muscle Nerve 1989;12:735-741. 10. White JC, Hansen SR, Johnson RK: A comparison of EMG procedures in the carpal tunnel syndrome with clinicalEMG correlations. Mwcle Nerve 1988; 11: 1177- 1182.
MUSCLE & NERVE
November 1992
1257
