Electrophysiologic diagnosis of unilateral meralgia paresthetica is usually assessed by side-to-side comparison of SNAP amplitudes, SNCVs, and SEP latencies following stimulation of lateral femoral cutaneous nerves. To determine the relevance for diagnosis of these tests and side-to-side comparison, the results were compared in patients with unilateral meralgia paresthetica and normal subjects. The long-term outcome was also considered, in order to determine whether electrophysiologicfindings contribute to the prognosis. In our study, SNAP amplitude comparison was found to be more useful for diagnosis than SNCV and SEP latency comparisons. However the value of the SNAP amplitude on the affected side, just as the results of the other tests, was not found to be predictive of the outcome. Also the results of the tests depend on the methods used and on the nerve's route. Key words: Meralgia paresthetica electrophysiology diagnosis prognosis MUSCLE & NERVE 14:51-56 1991
DIAGNOSTIC AND PROGNOSTIC VALUE OF ELECTROPHYSIOLOGIC TESTS IN MERALGIA PARESTHETICA ALAIN LAGUENY, MD, MARIE M. DELIAC, MD, PHlLlPPE DELIAC, MD, and ALAIN DURANDEAU, MD
Meralgia paresthetica (MP), the mononeuropathy demonstrated by NCV studies on additional of the lateral femoral cutaneous nerve (LFCN), is nerves and by needle EMG examination. characterized by the distribution of paresthesia, However the usefulness of electrophysiology in pain, and objective sensory changes on the outer MP has been debated,'" so various methods have side of the thigh. Many etiologies have been been developed for picking up the sensory nerve suggested.:4.14.16- 19.23 T h e commonest cause is the action potential (SNAP) with the aim to improve entrapment of the LFCN as it passes through the sensitivity and therefore diagnosis. inguinal ligament, just medial and inferior to T h e purposes of our study were the anterior superior iliac spine (ASIS). T h e elecTo determine to what degree affected to unaftrophysiologic tests routinely performed to confected side comparison of SNAP amplitudes, of firm the diagnosis of LFCN neuropathy, include SNCVs and of SEP latencies, contributes to the sensory nerve conduction velocity (SNCV) and diagnosis of LFCN neuropathy. We thus comsomatosensory-evoked potentials (SEP) recorded pared the results of the tests carried out on 22 on the scalp following LFCN stimulation. Nondepatients with unilateral MP and on 20 normal tectable or low SNAPS, diminished SNCVs, desubjects. layed and attenuated SEPs are the abnormalities T o determine retrospectively whether the electhought consistent with this d i a g n o ~ i s . ~ ~ ~ ~ ' " trophysiologic ~ ' ~ ~ ~ ~ ~ ~findings and the treatments conMoreover multiplex mononeuritis or lumbar plextribute to the prognosis, we sent out in 1988, a questionnaire concerning their long-term outopathy clinically masquerading as MY, may be come, to 2 3 other patients with unilateral M P examined between 1982 and 1984. Seventeen replied. The questionnaire was not sent to feFrom the Department de Neurologie (Drs. A Lagueny and M M. Deliac), Department de Physiologie (Dr. Ph. Deliac), and Department de Chirurmales with MY appearing during the course of gie Orthopedique (Dr Durandeau). Centre Hospitalier Universitaire Borpregnancy because the symptoms usually disapdeaux, France. pear after delivery. Address reprint requests to Dr. A. Lagueny. Department de Neurologie, Hopital du Haut-Leveque, Avenue de Magellan, 33600 Pessac. France Accepted for publication January 13, 1990
CCC 0148-639X191101051-06 $04.00 0 1991 John Wiley & Sons, Inc.
Meralgia Paresthetica. Electrophysiology
SUBJECTS AND METHODS
Controls had not suffered from any previous neuromuscular disease and none displayed abnormal-
MUSCLE & NERVE
January 1991
51
ities on neurological examination at the time of the study. ‘I‘here were 11 wonien and 9 men aged from 20 to 60 years (mean: 40.6). Patients with unilateral MP were selected on the following electrophysiologic criteria: normal coaxial needle examination and normal NCVs except on the LFC:N. In the comparative study there were 10 women and 12 men aged from 18 to 68 years (mean: 43.3): group 1 (patients). In the retrospective study of the long-term outcome there were 7 women and 10 men aged from 22 to 67 (mean 45.5): group 2 (patients). In controls and group 1, electrophysiologic tests were performed using Viking (Nicolet) equipment, and in group 2, using MS6 (Medelec). ‘The method was the same for both controls and patients. Small platinum needles (1 3L70 Dantec) were used to stimulate the LFCN and record the SNAPS and SEPs. Antidromic and orthodroniic SNCVs were performed in each case. Aritidromic: A stimulating cathode was located 1.5 cm distal and medial to the ASIS with the anode 3 cni proximal to the cathode. Kecording electrodes were located on a straight line between the ASIS and the lateral border of the patella, with an interelectrode distance of 3 cm, the proximal being 12 cm distal to the ASIS. T h e stimulating cathode was sometimes relocated 6 o r 7 times along the inferior border of the outer third of the inguinal ligament (IL) and also in the vincinity of the ASIS, to deterruine the best location at which the lower threshold of the stimulation evoked a sensation in the territory of the anterior branch, and in patients, evoked a sensation similar to their symptoms on their affected side. Subjects considered this relocation of the cathode as painless. In controls, stimulation at a threshold intensity of about 1 mA with a current duration of 0.1 msec was usually enough to evoke a sensation in the territory of the anterior branch of the LFCN. At 3 times the threshold, the maximal SNAP amplitude was easily obtained. Orthodromic: Electrodes were left in place b u t their use inversed. When orthodromic SNAPs could not be picked u p using the small platinum needle, especially in the case o f overweight people, a Teflon-coated needle (13L.61 Dantec) was used. T h e SNAP value is the average of 20 sweeps. If necessary 2 o r 3 averaged SNAPS were superimposed to get a more reliable onset of the potential for SNCV measurement. Usually the orthodromic stimulation artifact was greater than the antidroniic, and the onset, o r the first positive
52
Meralgia Paresthetica. Electrophysiology
peak, of the orthodromic SNAP was less easily identified than for the antidromic (Fig. 1A). T h e ground electrode was located between stimulation and recording points to reduce the stimulus artifact. Measurement of amplitude was made between the highest negative and positive peaks of the SNAP (Fig. 1A). T h e distance between the stimulating a n d recording electrodes was measured with the patient lying with legs straight. T h e skin temperature was maintained above 32°C. T h e SEPs were recorded following stirnulation of the LFCN, o r its anterior branch, 12 cni distal to the ASIS. T h e active electrode was inserted 20 nini posterior to CZ and the reference at FZ. ’The latency of the first positive peak was measured and called P30 (Figs. 1B and 1C). I n addition, SEPs were also recorded following sural nerves stimulation and P40 latency measured in the controls. P40 latency was used to discard, as far as possible, subjects with infraclinical sensory abnormalities. T h u s the results of P30 latency were not taken into consideration in 2 controls in whom P40 was delayed. All the tests were performed bilaterally. Stimuli were delivered at 2 Hz. T h e bandwith was 3.2 Hz-3.2 KHz with MS6 for both SNAPs and SEPs. With Viking it was 2 Hz-2 KHz for SNAPs and 20 Hz- 1 K H z for SEPs. For each test, mean values, standard deviations o r ranges, and side-to-side comparisons were performed and values for group 1 compared with controls. Statistical analysis was made using Student’s t-test. RESULTS
Kesults of the SNAPs were retained only when antidromic stimulation evoked a sensation in the territory of the anterior branch o r a sensation similar to the symptoms of meralgia paresthetica on the af-fected side. Absence of SNAP in 3 controls and 3 patients in group 1 was therefore not considered significant. T h e 3 rejected controls felt a sensation at the cathode o r below the greater trochanter on both sides in one, and on one side in the two others. T h e three rejected patients felt a sensation at the location of the cathode only. In Table 1 are given the results of the tests and of the side-to-side comparison in the 15 retained controls. SNAPS were more easily obtained antidromically. Orthodroinic SNAP often required use of a Teflon-coated needle. ’The orthodroniic artefact was larger than the antidromic. However, there is no significant difference between antidromic and orthodromic SNAP amplitudes and between antidromic and orthodromic SNCVs ( P >
MUSCLE & NERVE
January 1991
B
C
FIGURE 1. SNAPs and SEPs after LFCN stimulation (A) Normal averaged SNAP. Upper trace: antidromic; lower: orthodromic. p-p is highest peak-to-peak amplitude. Orthodromic is larger than antidromic stimulation artifact (S). Horizontal bar: 1 rnsec; vertical bar: 10 pV. (B) Normal SEP. Arrow at P30, first positive peak. Horizontal bar: 10 msec; vertical bar: 1pV. (C) SEPs in a case of rneralgia paresthetica. Upper trace: unaffected; lower: affected side. On affected side P30 is attenuated but not delayed. Horizontal bar: 10 msec; vertical bar: 1 pV.
0.05). 'The lower limit of the normal is the same for anti- and orthodrornic amplitudes and SNCVs ('l'able 1). Aniong the 19 patients retained, S N A P aniplitudes on the unaffected side are in the range of the control values. Distribution of abnormalities on the affected side are given in Table 2. In 9 patients the S N A P was obtained neither anti- nor orthodromically even on replacing the small platinum needle by the Teflon-coated needle at 1L. In the 10 others, 4 have both low amplitudes and abnormal side-to-side amplitude ratios, and 3 have
normal amplitudes but abnormal side-to-side amplitude ratios. SNCVs on the unaffected sick are in the range of the control values although the mean value (54 m/sec) is lower than for controls. On the affected side, in the 1 0 patients where SNCVs were obtainable 2 are low. In one of these, the affected to unaffected side C V difference is abnormal. In both groups, SEPs were always detected even when SNAPs were not obtained. Of the 19 patients, 4 had abnormal side-to-side latency differences, 2 of which also had delayed P30 laten-
Table 1 . Results of tests and of side-to-side comparison in the 15 retained controls.*
Results
Side-to-side comparison
Mean value
(Range)+ SD*
Limit of normal
Mean value
SNAP amplitude p V Antidromic Orthodromic
12 10
(4 02-28 10) (340-27)
3 PV
70 1 68
SNCV mis Antidromic Orthodromic
60 9
45 mis
57 3
6 20 5 03
51 71
SEP latency msec
30.46
2.44
35 msec
Tests
10 59
(Range) SD
Ratio % (45 5-93 6) (45 2-88 3)
Limit of normal 45%
Difference mis (0- 12) (0- 15)
Difference msec 0 85
1 04
15 m/s
3 msec
'The results of the 3 controls with nonrecordable SNAPs and of the 2 controls with prolonged P40 are not included in the tabie ' S D IS given /or results with Gaussian distribution *Range is given for resuits with non-Gaussian distribution
Meralgia Paresthetica. Electrophysiology
MUSCLE & NERVE
January 1991
53
Table 2. Distributlon of abnormalities on the affected side in the 19 retained patients SNAP amplitude
N*
SNCV
N
SEP latency
N
Normal Low Absent SNAP Abnormal side-to-side ratio
6 4 9 7
Normal Low Nonmeasurable (absent SNAP) Abnormal side-to-side difference
8
Normal Delayed Absent P30 Abnormal side-to-side difference
17 2 0 4
*Number of patients fAbsent SNAPS and nonmeasurable SNCVs are not included
in
2 9 1
the side-towde companson
cies on the affected side. Thus the results of the SEP latencies on the affected side is conclusive in only 4 patients. Amplitude asymmetry alone is not usually considered abnormal. However amplitude for the affected side is more than 50% lower than the unaffected in 6 patients in whom the P30 latency is considered normal. In controls the amplitude on one side is never more than 50% lower than the other.
Retrospective Study. T h e results of the long-term outcome of the 17 patients in group 2 are summarized in Table 3 . Eleven patients recovered: 7, in whom local injection of a corticoid at the 1L only teniporarily relieved the symptoms, recovered after neurolysis of the LFCN, the 4 others recovered after neither treatment. Of the 6 displaying no long-term improvement, 2 answered that they had received neither treatment, 2 had had local injection of corticoid, and the 2 others, after only temporary improvement following corticoid in-jection, had undergone neurolysis. Two of the nonimproved patients were diabetics. SNAP amplitudes had the most useful diagnostic values in our comparative study, but were not
found to be predictive of the long-term outcome (Table 3) neither were SNCVs or SEPs.
DISCUSSION
Electrophysiologic diagnosis of unilateral meralgia paresthetica is usually assessed by side-to-side comparison of SNAP amplitudes, SNCVs, and SEP latencies. In our study SNAP amplitude comparison was found to be more useful for diagnosis than SNCV and SEP latency comparisons. However, the value of the amplitude on the affected side, just as the results of the other tests, was not found to be predictive of outcome. In both patients and controls, orthodromic and antidromic SNAP amplitudes were not significantly different. However, the antidromic SNAP was more easily obtained than the orthodromic. The precise location of needle electrodes on the nerve pathway is necessary to obtain the highest SNAP amplitude. This is especially true at the IL, where the usual route of the LFCN is just medial and inferior to the ASIS, through an opening which is the classical site of LFCN entrapment. In the 15 retained controls, only a very low stimulation intensity at the IL was needed to pro-
Table 3. Long-term outcome in 17 patients Contribution of SNAP amplitude and treatment to the prognosis of meralgia paresthetica
Outcome
N*
Recovery
11
No improvement
6
Treatment
N
+Neurolysis
7
No treatment
4
'Neurolysis Local injection with corticoids No treatment
2 2
2
SNAP amplitude
N
Low Nondetectable Low Nondetectable
3 4 2 2
Nondetectable Low Nondetectable Low Nondetectable
2 1 1 1 1
*Number of patients 'Performed after only transient success of local corticoid injection at IL
54
Meralgia Paresthetica. Electrophysiology
MUSCLE & NERVE
January 1991
voke a sensation in the area of the anterior branch, and to evoke an antidromic SNAP of maximal amplitude. This was usually achieved with the stimulation needle at the standard location, or in some, with slight displacements. However at IL,, variations in the nerve's route have been reported.".".'" In the 3 rejected controls, antidromic stimulation evoked neither sensation in the area of the anterior branch nor SNAPs, even when using a longer stimulation needle and with tentative relocation. 71hey did feel sensation at the stimulation point or below the greater trochanter, suggesting unusual routes of the nerve, or its anterior branch, at the IL. On the other hand, tentative relocation of electrodes around the site 12 cm below the greater trochanter was found to be unnecessary, suggesting uniformity of routes here. In some of our controls, SNAP amplitudes were low. This could be the consequence of subclinical neuropathy. Subclinical entrapment of the I,FCN at IL has been found at autopsy,5 and in one study pathologic changes in niyelinated fibers have been reported. .' I t is generally considered that the degree of reduction of SNAP amplitudes is related to the degree of involvement of the large myelinated fibers. However, in our patients, the value of the SNAP amplitude did not indicate the severity of the symptoms. In 3 patients with intense pain, SNAP amplitudes were less diminished than in others who had only light paresthesia. Pain could result from the involvement of the small fibers which are perhaps more frequently involved than usually thought in MP."' These 3 patients had suffered from what could be considered as repeated traumas or microtramas of the LFCN (repetaed parachuting or repeated prolonged walking). Trauma as a cause of intense pain in entrapment neuropathy has been considered.2.' Repeated traumas could be responsible for involvement of the unmyelinated fibers, or for a stimulation of their activity, or also for an interference with the process of regeneration, thus increasing the number of unmyelinated hyperexcitable sprouts. Yet, whatever the mechanism, 2 of these 3 patients dramatically improved following neurolysis. SNCVs on the affected side are low only in a few of our patients and, even then, only moderately. In other studies, some SNCVs in MP have been found to be lower than ours.23 This discrepancy could be the consequence of the methods employed. Entrapment across an opening at IL is the main cause of LFCN neuropathy. If the SNAP is recorded, or the LFCN stimulated, proximal to
Meralgia Paresthetica. Electrophysiology
the IL, and thus to the site of entrapment, the SNCV could be lower than when measured distal to the ligament. However, picking u p the SNAP proximally may be difficult because of the often deep location of the nerve here. Also the latency at the negative peak reflects the dispersion of the conduction along the fast fibers and not the maximal velocity reflected by the onset, or first positive peak. In meralgia paresthetica, recordings of SEPs at the cortex following LFCN stimulation have been considered of value in demonstrating involvement of the nerve when SNAPs are not obtained. Ability to record a sizable SEP with absence of a recordable SNAP would result from a central amplifying p r o c e ~ s T . ~h e value of the first positive peak latency depends on the site of stimulation, either at IL or more distaL7 In our study P30 was delayed, on stimulation of the affected side, in only 2 patients. It may be difficult to discriminate a prolonged early component from a normal late component,12 and in our patients averaging of at least 1000 sweeps were sometimes necessary to identify P30. Prolonged SEP latencies could result, in some cases, from another cause than inguinal entrapment, such as an intrapelvic or an intra-abdominal compression of the LFCN, explaining lack of improvement after local treatment at IL.25 In our patients, attenuated and not well-formed SEPs were often found on the affected side. Amplitude asymmetry alone is not usually considered a b n ~ r m a l . 'However, ~ our results suggest that an amplitude ratio below 50% could be significant even if the latency is not significantly delayed. In the 3 rejected controls and in the 3 rejected patients without SNAPs, in whom antidromic stimulation did not evoke sensation in the area of the anterior branch or a sensation similar to the symptoms, SEPs were obtained. 'These SEPs could result from the stimulation of another nerve than the LFCN. However, the location of the stimulating electrodes (see Methods), and an absence of SNAPs on the femoral nerve at IL following stimulation, suggest in these individuals that it is a stimulation of the LFCN with an unusual route at I L and absence of SNAP, which produces the SEPs and not a stimulation of a sensory branch of the femoral nerve. Various treatments have been reported as successful in MY: local injection of corticoids, neurolysis when conservative treatments have failed to improve the symptoms,"' and epidural injection of dexamethasone" or transcutaneous electrical
MUSCLE & NERVE
January 1991
55
stimulation." However, the appraisal of therapeutic efficiency is not easy, because the course of the neuropathy seems to be different from case to case. T h e difference in the long-term outcome was obvious in o u r study (Table 3 ) even though o u r population sample is limited. SNAP amplitudes were not predictive of the outcome of o u r patients. In neuropathies, low SNAPS with normal o r slightly diminished SNCVs are usually considered as a sign of axonal loss of sensory myelinated fibers. However, it may be dif-
ficult to determine whether the reduction o f the SNAP amplitudes results from a reduction of' the number of the myelinated fibers, or from conduction blocks o r temporal dispersion along them. This could be the case in the entrapment of the LFCN at IL, and the reason why the degree of reduction of SNAP amplitudes is unreliable as a factor for predicting the outcome. Finally, in spite of possible failure, neurolysis at IL seems to be the most effective treatment especially in hyperalgic forms of meralgia paresthetica.
'
REFERENCES I. Ruchthal F, Kosenfalk A: Sensory potentials in polyncuH ) i i i ) r 1!)71;94:24l-202. I',,lolitison EW, Kaye %A:Normal conduction vclie 1atcr;il fknioral cutaneous nerve. Arch L"/z~.I M d K o / i a / ) i / 1!)74;55:3 1 -32. 3. (:Iihitttani I'N, Cliawla LS, Sharma 'I'D: Meralgia paraesetica. A d o NPUUJ/ Srcciid 1966;42:483-490. , Sut-r,cs S, I loircli M: Central nervous systcni ani) t i : its potential in the diagnosis o t early multiple , Nr ti t d ~ (Nl') q I 982 ;:32 :359- 364. 5. E:tlclson ,I(;, Naih;rri H: Mcl-algia paresthetica- A n ;Ina to niiciil i 11tcrp t.etiitio ti. Cliit Or//iop I W i u h l i p s l!)75; 122:25.5-2(i2. 6 . I+het. Al': '1.1-~iiiscittancoiiselectrical nerve stitnulation i n nieralgia p;ii.arstlieiica of p~-rgnancy.HI- J OO.s/d (;ytia(wd I(JX7;04:603- w 5 . 7. Fldgel K A , Stcii-in I],Skil);i N: Somatosensi1,el cvoiicrte I'otenti;ilc n i i ( . l i Stiniiilation tles N. cutiineus fcmoi-is I:itet-alis I x i N o r n i ~ t l ~ ~ c r s o r iuntl ~ i i 1';iticnten mit Mer;ilgi;i p t - x s ther ica, Z k;b;
DIAGNOSTIC AND PROGNOSTIC VALUE OF ELECTROPHYSIOLOGIC TESTS IN MERALGIA PARESTHETICA ALAIN LAGUENY, MD, MARIE M. DELIAC, MD, PHlLlPPE DELIAC, MD, and ALAIN DURANDEAU, MD
Meralgia paresthetica (MP), the mononeuropathy demonstrated by NCV studies on additional of the lateral femoral cutaneous nerve (LFCN), is nerves and by needle EMG examination. characterized by the distribution of paresthesia, However the usefulness of electrophysiology in pain, and objective sensory changes on the outer MP has been debated,'" so various methods have side of the thigh. Many etiologies have been been developed for picking up the sensory nerve suggested.:4.14.16- 19.23 T h e commonest cause is the action potential (SNAP) with the aim to improve entrapment of the LFCN as it passes through the sensitivity and therefore diagnosis. inguinal ligament, just medial and inferior to T h e purposes of our study were the anterior superior iliac spine (ASIS). T h e elecTo determine to what degree affected to unaftrophysiologic tests routinely performed to confected side comparison of SNAP amplitudes, of firm the diagnosis of LFCN neuropathy, include SNCVs and of SEP latencies, contributes to the sensory nerve conduction velocity (SNCV) and diagnosis of LFCN neuropathy. We thus comsomatosensory-evoked potentials (SEP) recorded pared the results of the tests carried out on 22 on the scalp following LFCN stimulation. Nondepatients with unilateral MP and on 20 normal tectable or low SNAPS, diminished SNCVs, desubjects. layed and attenuated SEPs are the abnormalities T o determine retrospectively whether the electhought consistent with this d i a g n o ~ i s . ~ ~ ~ ~ ' " trophysiologic ~ ' ~ ~ ~ ~ ~ ~findings and the treatments conMoreover multiplex mononeuritis or lumbar plextribute to the prognosis, we sent out in 1988, a questionnaire concerning their long-term outopathy clinically masquerading as MY, may be come, to 2 3 other patients with unilateral M P examined between 1982 and 1984. Seventeen replied. The questionnaire was not sent to feFrom the Department de Neurologie (Drs. A Lagueny and M M. Deliac), Department de Physiologie (Dr. Ph. Deliac), and Department de Chirurmales with MY appearing during the course of gie Orthopedique (Dr Durandeau). Centre Hospitalier Universitaire Borpregnancy because the symptoms usually disapdeaux, France. pear after delivery. Address reprint requests to Dr. A. Lagueny. Department de Neurologie, Hopital du Haut-Leveque, Avenue de Magellan, 33600 Pessac. France Accepted for publication January 13, 1990
CCC 0148-639X191101051-06 $04.00 0 1991 John Wiley & Sons, Inc.
Meralgia Paresthetica. Electrophysiology
SUBJECTS AND METHODS
Controls had not suffered from any previous neuromuscular disease and none displayed abnormal-
MUSCLE & NERVE
January 1991
51
ities on neurological examination at the time of the study. ‘I‘here were 11 wonien and 9 men aged from 20 to 60 years (mean: 40.6). Patients with unilateral MP were selected on the following electrophysiologic criteria: normal coaxial needle examination and normal NCVs except on the LFC:N. In the comparative study there were 10 women and 12 men aged from 18 to 68 years (mean: 43.3): group 1 (patients). In the retrospective study of the long-term outcome there were 7 women and 10 men aged from 22 to 67 (mean 45.5): group 2 (patients). In controls and group 1, electrophysiologic tests were performed using Viking (Nicolet) equipment, and in group 2, using MS6 (Medelec). ‘The method was the same for both controls and patients. Small platinum needles (1 3L70 Dantec) were used to stimulate the LFCN and record the SNAPS and SEPs. Antidromic and orthodroniic SNCVs were performed in each case. Aritidromic: A stimulating cathode was located 1.5 cm distal and medial to the ASIS with the anode 3 cni proximal to the cathode. Kecording electrodes were located on a straight line between the ASIS and the lateral border of the patella, with an interelectrode distance of 3 cm, the proximal being 12 cm distal to the ASIS. T h e stimulating cathode was sometimes relocated 6 o r 7 times along the inferior border of the outer third of the inguinal ligament (IL) and also in the vincinity of the ASIS, to deterruine the best location at which the lower threshold of the stimulation evoked a sensation in the territory of the anterior branch, and in patients, evoked a sensation similar to their symptoms on their affected side. Subjects considered this relocation of the cathode as painless. In controls, stimulation at a threshold intensity of about 1 mA with a current duration of 0.1 msec was usually enough to evoke a sensation in the territory of the anterior branch of the LFCN. At 3 times the threshold, the maximal SNAP amplitude was easily obtained. Orthodromic: Electrodes were left in place b u t their use inversed. When orthodromic SNAPs could not be picked u p using the small platinum needle, especially in the case o f overweight people, a Teflon-coated needle (13L.61 Dantec) was used. T h e SNAP value is the average of 20 sweeps. If necessary 2 o r 3 averaged SNAPS were superimposed to get a more reliable onset of the potential for SNCV measurement. Usually the orthodromic stimulation artifact was greater than the antidroniic, and the onset, o r the first positive
52
Meralgia Paresthetica. Electrophysiology
peak, of the orthodromic SNAP was less easily identified than for the antidromic (Fig. 1A). T h e ground electrode was located between stimulation and recording points to reduce the stimulus artifact. Measurement of amplitude was made between the highest negative and positive peaks of the SNAP (Fig. 1A). T h e distance between the stimulating a n d recording electrodes was measured with the patient lying with legs straight. T h e skin temperature was maintained above 32°C. T h e SEPs were recorded following stirnulation of the LFCN, o r its anterior branch, 12 cni distal to the ASIS. T h e active electrode was inserted 20 nini posterior to CZ and the reference at FZ. ’The latency of the first positive peak was measured and called P30 (Figs. 1B and 1C). I n addition, SEPs were also recorded following sural nerves stimulation and P40 latency measured in the controls. P40 latency was used to discard, as far as possible, subjects with infraclinical sensory abnormalities. T h u s the results of P30 latency were not taken into consideration in 2 controls in whom P40 was delayed. All the tests were performed bilaterally. Stimuli were delivered at 2 Hz. T h e bandwith was 3.2 Hz-3.2 KHz with MS6 for both SNAPs and SEPs. With Viking it was 2 Hz-2 KHz for SNAPs and 20 Hz- 1 K H z for SEPs. For each test, mean values, standard deviations o r ranges, and side-to-side comparisons were performed and values for group 1 compared with controls. Statistical analysis was made using Student’s t-test. RESULTS
Kesults of the SNAPs were retained only when antidromic stimulation evoked a sensation in the territory of the anterior branch o r a sensation similar to the symptoms of meralgia paresthetica on the af-fected side. Absence of SNAP in 3 controls and 3 patients in group 1 was therefore not considered significant. T h e 3 rejected controls felt a sensation at the cathode o r below the greater trochanter on both sides in one, and on one side in the two others. T h e three rejected patients felt a sensation at the location of the cathode only. In Table 1 are given the results of the tests and of the side-to-side comparison in the 15 retained controls. SNAPS were more easily obtained antidromically. Orthodroinic SNAP often required use of a Teflon-coated needle. ’The orthodroniic artefact was larger than the antidromic. However, there is no significant difference between antidromic and orthodromic SNAP amplitudes and between antidromic and orthodromic SNCVs ( P >
MUSCLE & NERVE
January 1991
B
C
FIGURE 1. SNAPs and SEPs after LFCN stimulation (A) Normal averaged SNAP. Upper trace: antidromic; lower: orthodromic. p-p is highest peak-to-peak amplitude. Orthodromic is larger than antidromic stimulation artifact (S). Horizontal bar: 1 rnsec; vertical bar: 10 pV. (B) Normal SEP. Arrow at P30, first positive peak. Horizontal bar: 10 msec; vertical bar: 1pV. (C) SEPs in a case of rneralgia paresthetica. Upper trace: unaffected; lower: affected side. On affected side P30 is attenuated but not delayed. Horizontal bar: 10 msec; vertical bar: 1 pV.
0.05). 'The lower limit of the normal is the same for anti- and orthodrornic amplitudes and SNCVs ('l'able 1). Aniong the 19 patients retained, S N A P aniplitudes on the unaffected side are in the range of the control values. Distribution of abnormalities on the affected side are given in Table 2. In 9 patients the S N A P was obtained neither anti- nor orthodromically even on replacing the small platinum needle by the Teflon-coated needle at 1L. In the 10 others, 4 have both low amplitudes and abnormal side-to-side amplitude ratios, and 3 have
normal amplitudes but abnormal side-to-side amplitude ratios. SNCVs on the unaffected sick are in the range of the control values although the mean value (54 m/sec) is lower than for controls. On the affected side, in the 1 0 patients where SNCVs were obtainable 2 are low. In one of these, the affected to unaffected side C V difference is abnormal. In both groups, SEPs were always detected even when SNAPs were not obtained. Of the 19 patients, 4 had abnormal side-to-side latency differences, 2 of which also had delayed P30 laten-
Table 1 . Results of tests and of side-to-side comparison in the 15 retained controls.*
Results
Side-to-side comparison
Mean value
(Range)+ SD*
Limit of normal
Mean value
SNAP amplitude p V Antidromic Orthodromic
12 10
(4 02-28 10) (340-27)
3 PV
70 1 68
SNCV mis Antidromic Orthodromic
60 9
45 mis
57 3
6 20 5 03
51 71
SEP latency msec
30.46
2.44
35 msec
Tests
10 59
(Range) SD
Ratio % (45 5-93 6) (45 2-88 3)
Limit of normal 45%
Difference mis (0- 12) (0- 15)
Difference msec 0 85
1 04
15 m/s
3 msec
'The results of the 3 controls with nonrecordable SNAPs and of the 2 controls with prolonged P40 are not included in the tabie ' S D IS given /or results with Gaussian distribution *Range is given for resuits with non-Gaussian distribution
Meralgia Paresthetica. Electrophysiology
MUSCLE & NERVE
January 1991
53
Table 2. Distributlon of abnormalities on the affected side in the 19 retained patients SNAP amplitude
N*
SNCV
N
SEP latency
N
Normal Low Absent SNAP Abnormal side-to-side ratio
6 4 9 7
Normal Low Nonmeasurable (absent SNAP) Abnormal side-to-side difference
8
Normal Delayed Absent P30 Abnormal side-to-side difference
17 2 0 4
*Number of patients fAbsent SNAPS and nonmeasurable SNCVs are not included
in
2 9 1
the side-towde companson
cies on the affected side. Thus the results of the SEP latencies on the affected side is conclusive in only 4 patients. Amplitude asymmetry alone is not usually considered abnormal. However amplitude for the affected side is more than 50% lower than the unaffected in 6 patients in whom the P30 latency is considered normal. In controls the amplitude on one side is never more than 50% lower than the other.
Retrospective Study. T h e results of the long-term outcome of the 17 patients in group 2 are summarized in Table 3 . Eleven patients recovered: 7, in whom local injection of a corticoid at the 1L only teniporarily relieved the symptoms, recovered after neurolysis of the LFCN, the 4 others recovered after neither treatment. Of the 6 displaying no long-term improvement, 2 answered that they had received neither treatment, 2 had had local injection of corticoid, and the 2 others, after only temporary improvement following corticoid in-jection, had undergone neurolysis. Two of the nonimproved patients were diabetics. SNAP amplitudes had the most useful diagnostic values in our comparative study, but were not
found to be predictive of the long-term outcome (Table 3) neither were SNCVs or SEPs.
DISCUSSION
Electrophysiologic diagnosis of unilateral meralgia paresthetica is usually assessed by side-to-side comparison of SNAP amplitudes, SNCVs, and SEP latencies. In our study SNAP amplitude comparison was found to be more useful for diagnosis than SNCV and SEP latency comparisons. However, the value of the amplitude on the affected side, just as the results of the other tests, was not found to be predictive of outcome. In both patients and controls, orthodromic and antidromic SNAP amplitudes were not significantly different. However, the antidromic SNAP was more easily obtained than the orthodromic. The precise location of needle electrodes on the nerve pathway is necessary to obtain the highest SNAP amplitude. This is especially true at the IL, where the usual route of the LFCN is just medial and inferior to the ASIS, through an opening which is the classical site of LFCN entrapment. In the 15 retained controls, only a very low stimulation intensity at the IL was needed to pro-
Table 3. Long-term outcome in 17 patients Contribution of SNAP amplitude and treatment to the prognosis of meralgia paresthetica
Outcome
N*
Recovery
11
No improvement
6
Treatment
N
+Neurolysis
7
No treatment
4
'Neurolysis Local injection with corticoids No treatment
2 2
2
SNAP amplitude
N
Low Nondetectable Low Nondetectable
3 4 2 2
Nondetectable Low Nondetectable Low Nondetectable
2 1 1 1 1
*Number of patients 'Performed after only transient success of local corticoid injection at IL
54
Meralgia Paresthetica. Electrophysiology
MUSCLE & NERVE
January 1991
voke a sensation in the area of the anterior branch, and to evoke an antidromic SNAP of maximal amplitude. This was usually achieved with the stimulation needle at the standard location, or in some, with slight displacements. However at IL,, variations in the nerve's route have been reported.".".'" In the 3 rejected controls, antidromic stimulation evoked neither sensation in the area of the anterior branch nor SNAPs, even when using a longer stimulation needle and with tentative relocation. 71hey did feel sensation at the stimulation point or below the greater trochanter, suggesting unusual routes of the nerve, or its anterior branch, at the IL. On the other hand, tentative relocation of electrodes around the site 12 cm below the greater trochanter was found to be unnecessary, suggesting uniformity of routes here. In some of our controls, SNAP amplitudes were low. This could be the consequence of subclinical neuropathy. Subclinical entrapment of the I,FCN at IL has been found at autopsy,5 and in one study pathologic changes in niyelinated fibers have been reported. .' I t is generally considered that the degree of reduction of SNAP amplitudes is related to the degree of involvement of the large myelinated fibers. However, in our patients, the value of the SNAP amplitude did not indicate the severity of the symptoms. In 3 patients with intense pain, SNAP amplitudes were less diminished than in others who had only light paresthesia. Pain could result from the involvement of the small fibers which are perhaps more frequently involved than usually thought in MP."' These 3 patients had suffered from what could be considered as repeated traumas or microtramas of the LFCN (repetaed parachuting or repeated prolonged walking). Trauma as a cause of intense pain in entrapment neuropathy has been considered.2.' Repeated traumas could be responsible for involvement of the unmyelinated fibers, or for a stimulation of their activity, or also for an interference with the process of regeneration, thus increasing the number of unmyelinated hyperexcitable sprouts. Yet, whatever the mechanism, 2 of these 3 patients dramatically improved following neurolysis. SNCVs on the affected side are low only in a few of our patients and, even then, only moderately. In other studies, some SNCVs in MP have been found to be lower than ours.23 This discrepancy could be the consequence of the methods employed. Entrapment across an opening at IL is the main cause of LFCN neuropathy. If the SNAP is recorded, or the LFCN stimulated, proximal to
Meralgia Paresthetica. Electrophysiology
the IL, and thus to the site of entrapment, the SNCV could be lower than when measured distal to the ligament. However, picking u p the SNAP proximally may be difficult because of the often deep location of the nerve here. Also the latency at the negative peak reflects the dispersion of the conduction along the fast fibers and not the maximal velocity reflected by the onset, or first positive peak. In meralgia paresthetica, recordings of SEPs at the cortex following LFCN stimulation have been considered of value in demonstrating involvement of the nerve when SNAPs are not obtained. Ability to record a sizable SEP with absence of a recordable SNAP would result from a central amplifying p r o c e ~ s T . ~h e value of the first positive peak latency depends on the site of stimulation, either at IL or more distaL7 In our study P30 was delayed, on stimulation of the affected side, in only 2 patients. It may be difficult to discriminate a prolonged early component from a normal late component,12 and in our patients averaging of at least 1000 sweeps were sometimes necessary to identify P30. Prolonged SEP latencies could result, in some cases, from another cause than inguinal entrapment, such as an intrapelvic or an intra-abdominal compression of the LFCN, explaining lack of improvement after local treatment at IL.25 In our patients, attenuated and not well-formed SEPs were often found on the affected side. Amplitude asymmetry alone is not usually considered a b n ~ r m a l . 'However, ~ our results suggest that an amplitude ratio below 50% could be significant even if the latency is not significantly delayed. In the 3 rejected controls and in the 3 rejected patients without SNAPs, in whom antidromic stimulation did not evoke sensation in the area of the anterior branch or a sensation similar to the symptoms, SEPs were obtained. 'These SEPs could result from the stimulation of another nerve than the LFCN. However, the location of the stimulating electrodes (see Methods), and an absence of SNAPs on the femoral nerve at IL following stimulation, suggest in these individuals that it is a stimulation of the LFCN with an unusual route at I L and absence of SNAP, which produces the SEPs and not a stimulation of a sensory branch of the femoral nerve. Various treatments have been reported as successful in MY: local injection of corticoids, neurolysis when conservative treatments have failed to improve the symptoms,"' and epidural injection of dexamethasone" or transcutaneous electrical
MUSCLE & NERVE
January 1991
55
stimulation." However, the appraisal of therapeutic efficiency is not easy, because the course of the neuropathy seems to be different from case to case. T h e difference in the long-term outcome was obvious in o u r study (Table 3 ) even though o u r population sample is limited. SNAP amplitudes were not predictive of the outcome of o u r patients. In neuropathies, low SNAPS with normal o r slightly diminished SNCVs are usually considered as a sign of axonal loss of sensory myelinated fibers. However, it may be dif-
ficult to determine whether the reduction o f the SNAP amplitudes results from a reduction of' the number of the myelinated fibers, or from conduction blocks o r temporal dispersion along them. This could be the case in the entrapment of the LFCN at IL, and the reason why the degree of reduction of SNAP amplitudes is unreliable as a factor for predicting the outcome. Finally, in spite of possible failure, neurolysis at IL seems to be the most effective treatment especially in hyperalgic forms of meralgia paresthetica.
'
REFERENCES I. Ruchthal F, Kosenfalk A: Sensory potentials in polyncuH ) i i i ) r 1!)71;94:24l-202. I',,lolitison EW, Kaye %A:Normal conduction vclie 1atcr;il fknioral cutaneous nerve. Arch L"/z~.I M d K o / i a / ) i / 1!)74;55:3 1 -32. 3. (:Iihitttani I'N, Cliawla LS, Sharma 'I'D: Meralgia paraesetica. A d o NPUUJ/ Srcciid 1966;42:483-490. , Sut-r,cs S, I loircli M: Central nervous systcni ani) t i : its potential in the diagnosis o t early multiple , Nr ti t d ~ (Nl') q I 982 ;:32 :359- 364. 5. E:tlclson ,I(;, Naih;rri H: Mcl-algia paresthetica- A n ;Ina to niiciil i 11tcrp t.etiitio ti. Cliit Or//iop I W i u h l i p s l!)75; 122:25.5-2(i2. 6 . I+het. Al': '1.1-~iiiscittancoiiselectrical nerve stitnulation i n nieralgia p;ii.arstlieiica of p~-rgnancy.HI- J OO.s/d (;ytia(wd I(JX7;04:603- w 5 . 7. Fldgel K A , Stcii-in I],Skil);i N: Somatosensi1,el cvoiicrte I'otenti;ilc n i i ( . l i Stiniiilation tles N. cutiineus fcmoi-is I:itet-alis I x i N o r n i ~ t l ~ ~ c r s o r iuntl ~ i i 1';iticnten mit Mer;ilgi;i p t - x s ther ica, Z k;b;
