ORIGINAL RESEARCH
Somatosensory Evoked Potentials in Chronic Inflammatory Demyelinating Polyradiculoneuropathy Hayet Salhi,* Philippe Corcia,*† Steve Remer,* and Julien Praline*†
Purpose: Somatosensory evoked potentials (SEPs) offer complementary results to those of nerve conduction studies and contribute to the electrodiagnostic criteria of chronic inflammatory demyelinating polyradiculoneuropathy. Methods: We performed nerve conduction studies and SEPs in patients with symmetrical motor weakness, areflexia, and/or sensory disturbances lasting for at least 8 weeks. We determined two groups according to the electrodiagnostic criteria of the European Federation of Neurological Societies. Group 1 included patients who met the definite or probable electrodiagnostic criteria, and group 2 included patients who met the possible electrodiagnostic criteria. We also compared SEPs results with those of controls (group of healthy subjects). Results: Sixteen patients (14 men; mean age, 59 6 17.3 years) were included in the study. The latencies of potentials N9, N13, N7, and N22 and the intervals N9–N13 and N7–N22 were significantly increased in patients compared with controls. The N9/iP14 amplitude ratio was significantly lower in patients. There was no significant difference in the latencies of SEPs between the two groups of patients. Conclusions: We confirm the contribution of SEPs as complementary information to nerve conduction studies in chronic inflammatory demyelinating polyradiculoneuropathy diagnosis. In addition to the usual abnormalities, a decrease in the N9/iP14 amplitude ratio could potentially be used as an electrodiagnostic criterion. Key Words: Somatosensory evoked potentials, Nerve conduction study, Chronic inflammatory demyelinating polyradiculoneuropathy, Electrodiagnostic criteria. (J Clin Neurophysiol 2014;31: 241–245)
C
hronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is a chronic sensorimotor neuropathy with segmental demyelination. In fact, several variants of CIDP, such as pure motor, pure sensory or sensorimotor forms, are now well known. The diagnosis of CIDP is sometimes difficult. Several groups have proposed diagnostic criteria based on clinical, biological, and nerve conduction study (NCS) findings (Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force 1991; Joint Task Force of the EFNS and the PNS, 2010; Koski et al., 2009). Because the nerve root is frequently involved site in CIDP (Hughes et al., 2003), the sensitivity of NCSs can sometimes be low (Molenaar et al., 2002). Consequently, the study of somatosensory evoked potentials (SEPs), which examines the entire proximal pathway and so the distal extension of the ganglion cells, is particularly interesting in the exploration of radiculo-spinal diseases and offers
From the *Department of Neurology Neurophysiology, Hospital of Tours, France; and †INSERM UMR930, Tours, France. Address correspondence and reprint requests to Hayet Salhi, MD, Service de Neurologie, 2 Boulevard Tonnellé, Tours 37000, France; e-mail: salhi. [email protected]. Copyright Ó 2014 by the American Clinical Neurophysiology Society
ISSN: 0736-0258/14/3103-0241
complementary results to those of NCSs. In the European Federation of Neurological Societies (EFNS) criteria, the use of SEPs seems as supportive criteria, but their diagnostic value has not clearly been defined. Studies on acute polyradiculoneuropathy have shown slow conduction on SEPs across the roots and proximal brachial plexus when motor and sensory conduction velocities over more distal segments were normal (Brown and Feasby, 1984). Abnormalities of the N9 response and/or the spinal N13 response with a prolonged N9–N13 interval have been reported after median nerve stimulation; however, abnormal spinal N22 responses after stimulation of the lower extremities were more common (Walsh et al., 1984). Several studies have been conducted to determine the sensitivity and specificity of SEPs in CIDP (Tsukamoto et al., 2010; Viala et al., 2010; Yiannikas and Vucic, 2008). The inclusion criteria and the methods of these studies were quite different. The sensitivity of SEPs in these studies ranged from 13% to 83%. Viala et al. (2010) studied 146 patients with CIDP; however, they performed SEPs only in a subgroup of 36 patients without major electrodiagnostic criteria for demyelination. This study found proximal conduction abnormalities in 25 of the 36 patients studied with SEPs. The sensitivity of SEPs was low but not all of the patients had SEPs (Viala et al., 2010). Yiannikas and Vucic found the sensitivity of SEPs to be 62% when assessing SEPs and NCS in 47 patients, but this study was performed without using electrodiagnostic criteria to make a diagnosis of CIDP (Yiannikas and Vucic, 2008). Tsukamoto et al. compared 12 patients with CIDP to 17 patients with diabetic polyneuropathy. They recorded tibial nerve SEPs with an unusual technique, and abnormalities suggesting proximal demyelination were found in 83% of patients with CIDP and in only 13% of patients with DPN (Tsukamoto et al., 2010). The aim of our study was to evaluate the contribution of SEPs in the diagnosis of CIDP and to describe the electrophysiological abnormalities of this disease.
MATERIALS AND METHODS Subjects Between April 2009 and April 2011, patients with symmetrical proximal and distal motor weakness associated with or without areflexia, proprioceptive, and epicritic sensory disturbances or patients with pure sensory disturbance lasting for at least 8 weeks were consecutively included in the study. The exclusion criteria included evidence of Charcot–Marie–Tooth disease, the presence of immunoglobulin M monoclonal gammopathy with anti-myelin associated glycoprotein (MAG), active Lyme disease, exposure to drugs that can cause demyelinating neuropathy, or diagnosis of cancer with or without the presence of onco-neuronal antibodies. The data collected were age, gender, height, date of onset of symptoms, date of diagnosis, peak motor disability scale
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
241
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
H. Salhi et al.
TABLE 1.
Characteristics of Patients by Group
Characteristics Age of onset (average) Time to diagnosis (median) Cerebrospinal fluid protein (median)
All (n ¼ 16)
Group 1 (n ¼ 9)
Group 2 (n ¼ 7)
P
59 6 17.3 50 6 37.7 0.60 6 0.35
51 6 11.4 44.3 6 42.4 0.59 6 0.45
69.2 6 18.4 57.3 6 32.3 0.70 6 0.19
0.039 0.29 0.86
(Hughes et al., 2003), and protein level in the cerebrospinal fluid. The final diagnosis of CIDP was made after a follow-up of at least 6 months, taking into account the clinical course according to the EFNS criteria (Joint Task Force of the EFNS and the PNS, 2010).
Nerve Conduction Study All patients were initially assessed with NCS performed on a Keypoint apparatus (Alpine Biomedical, Fountain Valley, CA). The skin temperature was measured at the palm, and the plant was measured using an external thermometer. The temperature of the upper and lower extremities was maintained at or more than 338C. The motor nerves studied included at least the unilateral median, ulnar, tibial, and common peroneal nerves using supramaximal stimulation. The F-wave examination was made for all four motor nerves. The stimulus sites for the median nerve were at the wrist and elbow, and recording was performed from the abductor pollicis brevis. For the ulnar nerve, the stimulus sites were at the wrist, below the elbow, and above the elbow; and recording was performed from the abductor digit minimi. For the tibial nerve, the recording sites were at the ankle and the popliteal fossa; and recording was performed from the flexor hallucis brevis. For the common peroneal nerve, stimulation was performed at the ankle, fibular head, and knee; and recording was performed from the extensor digitorum brevis muscle. Ten or more supramaximal stimuli were delivered distally to elicit F responses from the median and common peroneal nerves. Sensory NCSs were performed using antidromic techniques on the median, ulnar, and sural nerves.
Somatosensory Evoked Potentials Somatosensory evoked potentials were obtained by transcutaneous electrical stimulation of the median nerve at the wrist and the posterior tibial nerve at the ankle, on both sides, and were alternately recorded as recommended by the International Federation of Clinical Neurophysiology (Mauguiere et al., 1999) using the Keypoint software (Alpine Biomedical). Subjects lay on an examination table, and skin temperature was controlled to be more than 328C. For the median nerve SEPs, 4 recording channels were included: ipsilateral Erb-Fz electrodes for channel 1, electrode Cv5 (spinous process of the fifth cervical vertebra); CA (supralaryngeal) for channel 2, electrodes Pc (contralateral parietal); Erb contralateral for channel 3 and Pi electrodes (ipsilateral parietal); and Erb contralateral for channel 4. For the posterior tibial nerve SEPs, the set-up consisted of 4 recording channels: built-in electrodes Pup (popliteal fossa up), Pdw (popliteal fossa down) for channel 1, electrodes L1 (spinous process of the first lumbar vertebra), umbilical for channel 2, electrodes FzCv5 for channel 3 and Pz electrodes; and Ac (contralateral ear) for channel 4. Scalp electrodes were needles, and the other recording electrodes were cup electrodes. Skin-electrode impedance should 242
be less than 5000 U. Stimulation rate was 3 to 5 Hz and the analysis time was 50 milliseconds for upper limb stimulation and 100 milliseconds for lower limb. We have used a high-pass filter at 1 Hz and a low-pass filter at 2000 Hz. Monophasic square-wave electrical pulses of 0.1 to 0.2 milliseconds were to be delivered, and the intensity of stimulation was higher than the motor threshold. Potentials were identified on the curve obtained by averaging 2 series of 1,500 acquisitions reproducible for each stimulated nerve with the target muscle minimally active. We collected the amplitude and the latency of the different components obtained in each channel. The latency was measured at the peak amplitude, and the amplitude was measured between the peak and the end of the potential. The interpretation was based on our laboratory standards that have been obtained with 47 controls using the same technique and apparatus. We also included 16 healthy subjects who were matched to the patient group according to height and age to form a control group.
Analysis The time to diagnosis was calculated for each patient. The criteria of the EFNS were applied to the results of the NCS of each patient to establish two groups. Group 1 included patients who met the definite or probable electrodiagnostic criteria, and group 2 included patients who met the possible electrodiagnostic criteria or did not meet the probable electrodiagnostic criteria. We compared the time to diagnosis, age, and gender distributions between the two groups. For the SEPs, the N9 (clavical response), N13 (cervical response), N7 (popliteal response), and N22 (lumbar response) latencies were collected to calculate the time intervals N9–N13 (clavical to cervical response) and N7–N22 (popliteal to lumbar response). We also collected the N9/iP14 amplitude ratio (P14 inverse potential was obtained on the first channel registration Erb-Fz). For each variable, the left-right difference was calculated
TABLE 2. Results of Somatosensory Evoked Potentials According to Laboratory Norms Increased/Absent SEPs Latencies N9 N13 N7 N22 Intervals N9–N13 N7–N22
Normal
Group 1
Group 2
Group 1
Group 2
6 7 5 8
6 6 5 7
3 2 4 1
1 1 2 0
4 8
6 6
5 1
1 1
SEP, Somatosensory evoked potential.
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
TABLE 3. Results of Somatosensory Evoked Potentials in Patients Compared With Controls
SEPs Latencies N9 N13 N7 N22 Intervals N7–N22 N9–N13 Amplitude ratio Ratio N9/P14i
Mean 6 SD Mean 6 SD (milliseconds) (milliseconds) Patients (Number) Controls (n ¼ 16) 6 6 6 6
P
12.7 6 1.62 (n ¼ 14) 16.3 6 1.40 (n ¼ 13) 10.1 6 2.82 (n ¼ 11) 26.8 6 3.83 (n ¼ 6)
10.3 13.6 8.5 22.8
0.55 0.62 0.50 1.14
,0.0001 ,0.0001 0.007 0.0001
15.9 6 3.66 (n ¼ 6) 3.6 6 0.69 (n ¼ 13)
13.3 6 0.82 3.4 6 0.38
0.0001 0.009
1.06 6 0.97 (n ¼ 15)
2.21 6 1.10
,0.0001
SEP, Somatosensory evoked potential.
to look for asymmetry. First, we analyzed the SEP results for each patient and classified them as normal or abnormal according to our laboratory norms. SEPs latency or interval was considered as abnormal if it was more than 2 SDs above the mean of the laboratory norms. Second, all data were compared with those of the control group and between each patient group. Left and right responses were pooled for the comparison between patients and controls and between the two groups of patients. Wilcoxon test was applied because there was a group population of fewer than 30. The significance level was set at P , 0.05.
Somatosensory Evoked Potentials in CIDP
cerebrospinal fluid protein levels (N , 0.5 g/L) with a leukocyte count ,10/mm3. After applying the electrodiagnostic criteria of the EFNS, group 1 included 9 patients (8 men, mean age, 51 6 11.4 years), and group 2 included 7 patients (1 met the possible category criteria) (6 men, mean age 69.2 6 18.9 years) (Table 1). The changes in SEP latencies are summarized in Table 2. In group 2, NCSs were normal in 28.6% who showed axonal polyneuropathy in 57% of cases, and for 1 patient, a unilateral ulnar nerve entrapment was found at the elbow. The latencies of potentials N9, N13, N7, and N22 were significantly increased in patients compared with the controls (P , 0.001), as were the N9–N13 interval (P ¼ 0.009) and the N7–N22 interval (P , 0.0001) (Table 3). Two examples of SEPs responses are shown (Figs. 1 and 2). In our patients, there was no significant right/left asymmetry for potential latencies N9, N13, N7, and N22 (P . 0.05) (data not shown). In group 1, the mean of the N9/iP14 amplitude ratio was 1, and in group 2, it was 1.41. The N9/iP14 amplitude ratio was significantly lower in the patients when compared with the controls (P , 0.0001) (Table 3 and Fig. 3). There was no significant difference in the latencies of the SEPs between the two groups of patients (Table 4). Of the 5 patients with the sensory variant of CIDP, 1 patient met the electrodiagnostic criteria (group 1) and 4 patients did not meet the electrodiagnostic criteria (group 2). The latencies of the N9 and N13 potentials were increased in all patients, and the latency of N7 was increased in 4 patients and normal in 1 patient. The latency of the N22 potential was increased in 2 patients (N22 was absent in the remaining 3 patients). The amplitude ratio, N9/iP14, was reduced for 4 of these patients.
RESULTS
DISCUSSION
A total of 18 patients fulfilled the inclusion criteria on their first visit. Two patients were excluded because of the presence of monoclonal IgM with anti-MAG. The mean age of the remaining 16 patients (14 men) was 59 6 17.3 years. The peak motor disability scale (Hughes et al., 2003) was grade 4 in 2 patients, grade 3 in 6 patients, grade 2 in 5 patients, and grade 1 in 3 patients. Sensory symptoms or signs in the legs were present in 11 patients, and sensory ataxia was evident in 4 patients. Five patients lacked motor symptoms, and thus, these patients were considered to have the pure sensory CIDP variant (Joint Task Force of the EFNS and the PNS, 2010). Thirteen patients had elevated
Our results demonstrate the contribution of SEPs in the identification of a root demyelinating process. In group 1, of the 9 patients, 78% had increased latencies of N9 and N13 or absent potentials, 56% had abnormal latencies of N7 and N22, and the mean N9/iP14 amplitude ratio was 1. For all of these patients with electrodiagnostic criteria of CIDP, we found abnormalities suggesting proximal demyelination. In group 2, of the 7 patients without these electrodiagnostic criteria, 86% had increased latencies of N9 and N13 or absent potentials, 71% had abnormal latency of N7, and 100% had abnormal N22 responses. The mean of the N9/iP14 amplitude ratio was 1.41. The abnormalities of SEPs were the same
FIG. 1. Somatosensory evoked potentials of right median nerve before (A) and after averaging (B) of 2 series of 1,500 acquisitions (52-year-old man). See the increased latencies of all potentials and decreased amplitude of N9 contrasting with relative preservation of amplitude of iP14 on the first channel (arrow). CA, supralaryngeal electrode; Cv5, electrode on spinous process of the fifth cervical vertebra; Erb c, contralateral Erb; Erb G, ipsilateral Erb; Fz, frontal electrode; Pc, contralateral parietal; Pi, ipsilateral parietal electrode. Copyright Ó 2014 by the American Clinical Neurophysiology Society
243
H. Salhi et al.
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
FIG. 2. Somatosensory evoked potentials of left posterior tibial nerve before (A) and after averaging (B) of 2 series of 1,500 acquisitions (52-year-old man). See the increased latencies of N22 and next potentials with preservation of amplitude. Ac, contralateral ear electrode; Cv5, electrode on spinous process of the fifth cervical vertebra; Fz, frontal electrode; L1, electrode on spinous process of the first lumbar vertebra; Omb, umbilical electrode; Pdw, electrode on popliteal fossa down; Pup, electrode popliteal fossa up; Pz, parietal electrode. in this group 2 than in group 1; so here, we can also say that all of these patients had signs of proximal demyelination. Interestingly, our results demonstrated a root demyelinating process in seven patients although they did not meet electrodiagnostic criteria of CIDP. Five patients lacked motor symptoms, and thus, these patients were considered to have the pure sensory CIDP variant. Of the 5 patients with the pure sensory CIDP variant, 4 were in group 2, (so without electrodiagnostic criteria); and with the realization of SEPs, we could have established the diagnosis of CIDP. For this variant, the use of SEPs is the most useful because the study of NCSs is often normal (Sinnreich et al., 2004). Thus, the collection of SEPs enabled the identification of abnormalities of proximal demyelination in our 16 patients, which included patients who did not meet the electrodiagnostic criteria for CIDP (group 2). Our results confirmed those of two previous studies that highlighted differences in the SEPs of patients who did not meet the commonly used electrodiagnostic criteria. However, in these two previous studies, the electrodiagnostic criteria of the EFNS were not applied. Tsukamoto et al. (2010) evaluated the diagnostic value of SEPs of the posterior tibial nerve in demyelinating neuropathies and axonal diabetic polyneuropathy. In addition to the potentials that are classically studied, they used an additional bypass that enabled them to study the N15 potential corresponding to the afferent volley of action potentials to the passage of the greater sciatic foramen
(Sonoo et al., 1992). The criteria used to define a demyelinating process were a normal latency of the N8 potential with a conduction time increase to a segment proximal to this potential, an increase in the ratio of the proximal/distal latency, a normal amplitude of the N8 response with a decrease in amplitude in another more proximal potential (P15, N21), or a decrease in the amplitude ratio of the proximal/distal response. These abnormalities were seen as signs of proximal damage that is specific to CIDP. The sensitivity of SEPs for the diagnosis of CIDP was 83%, and a false-positive rate of 18% was found in the control group of patients with diabetic polyneuropathy (Tsukamoto et al., 2010). The methodology of the study by Yiannikas and Vucic (2008) was quite similar to ours. Forty-seven patients with clinical signs suggestive of CIDP were investigated by NCSs and SEPs in four limbs. The SEP responses were considered abnormal and consistent with proximal demyelination if the latency exceeded the upper limit of normal, if responses were absent, or if there was an increased N9–N13 conduction time. Five homogeneous groups of patients were established according to the results of normal or abnormal NCSs and SEPs. For 7 patients (groups 1 and 4) without abnormalities of nerve conduction time in NCSs, the SEPs showed signs of demyelination with longer P39 latencies for 6 patients, no N22 response in the lower limbs for 7 patients, and longer N9 and N13 latencies in the upper limbs for 5 patients. The diagnosis of CIDP was given, despite the normality of the NCS responses (Yiannikas and Vucic, 2008). In our cohort, we found abnormalities of nerve conduction (increased N9 and N7 latency potentials) and/or root demyelination (increased latency of N13 and N22 potentials and the N9–N13 interval) in CIDP. These abnormalities correspond to those previously described (Yiannikas and Vucic, 2008). In addition, the decrease in the N9/iP14 amplitude ratio allowed identification of a decrease in
TABLE 4. Results of Somatosensory Evoked Potentials in Group one Compared to Group 2
SEPs Latencies N9 N13 N7 N22
FIG. 3. Box plots of the N9/iP14 amplitude ratio in controls and patients. 244
Median 6 SD (milliseconds) Group 1 12.7 16 9.7 26.1
6 6 6 6
2.83 2.20 3.68 5.88
(n (n (n (n
¼ ¼ ¼ ¼
16) 15) 13) 9)
Median 6 SD (milliseconds) Group 2 12.7 16.4 10.6 30.4
6 6 6 6
3.04 2.89 5.58 2.18
(n (n (n (n
¼ ¼ ¼ ¼
13) 12) 9) 4)
P 0.54 0.22 0.84 0.17
n, number of limbs (left and right); SEP, somatosensory evoked potential.
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
the amplitude of N9 in contrast with the relatively preserved amplitude of P14. Using an Fz electrode reference on the first recording channel for the upper limbs, the P14 potential was recorded on the same derivation as the N9 potential. Thus, this method allowed us to calculate the N9/iP14 amplitude ratio more rigorously than if it was calculated using potentials obtained on different channels. The decrease of the N9/iP14 amplitude ratio reflects a secondary resynchronization and allows elimination of a central impairment. The decrease of the ratio might reflect that the observed loss of amplitude was related to a desynchronization of potentials and thus to a demyelinating process. The decrease in the N9/iP14 amplitude ratio was found to be statistically significant in the control group (P , 0.0001, Tables 3 and 4), so this abnormality seems to be specific to CIDP and could be used as an electrodiagnostic criterion. The absence of right-left asymmetry in the N9, N13, N22, and N7 potential latencies confirmed the diffuse nature of the demyelinating process (Pollard and Armati, 2011). Our results also increase the importance of the investigation of all four limbs; the SEPs of the lower limbs are rarely obtained in a reliable and reproducible way. Although the EFNS criteria were developed to be very specific to select patients for therapeutic trials, some patients who do not meet the criteria lead to diagnostic difficulty. This also reinforces the need to explore supportive means of diagnosis, although the comparison to a gold standard has not been possible to date. Our study has limitations that are related to the small number of patients. The use of a recording on the elbow for the upper limb and on the bottom for the lower limb (Tsukamoto et al., 2010) would also be interesting to elucidate the segment where demyelination is most pronounced. Furthermore, the high number of missing responses in the lower limbs (no N7 response in 28.6% of cases and no N22 in 71.4% of cases) led us to carefully consider our results. Furthermore, the N22 may not be identified in control subjects, especially in aged subjects; thus loss of N22 patients in aged patients could not be abnormal. Some teams have proposed performing a spine MRI to find root hyperintensities suggestive of CIDP (Joint Task Force of the EFNS and the PNS, 2010; Tazawa et al., 2008). We have not studied the value of MRI in our groups. We might consider studying each patient with both SEPs and MRI to compare their contributions toward diagnosis and to assess any associations that may be found between the two methods. Moreover, we did not try to assess the usefulness of F-waves in our cohort. F-wave study could reflect both distal and proximal nerve conduction, especially it assess the motor nerve roots in proximal, while SEPs exploring the sensitive nerve roots. In our work, the nerve conduction studies and F-waves had been used to constitute the two groups of patients. It seemed difficult to make correlations between SEPs and F-wave abnormalities given the small number of patients and the low probability of F-wave abnormalities (Walsh et al., 1984). We elected to study the clinical benefits of SEPs in daily practice by comparing results between two groups defined according to
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Somatosensory Evoked Potentials in CIDP
the EFNS criteria. This method allowed us to study the benefits of SEPs in what can be termed “ecological conditions,” which better correspond to this technically difficult procedure. To conclude, SEP can be a very useful test for CIDP diagnosis in patients without electrodiagnostic criteria (NCSs), and we confirmed the value of SEP results as complementary to those from NCSs. An increase in N9, N13, N7, and N22 latencies and a decrease in the N9/iP14 amplitude ratio could be used as electrodiagnostic criteria for CIDP in that case. Although it is interesting to perform SEPs in the four limbs, the abnormalities of the upper limbs seem to be more reliable for the diagnosis of CIDP. It will be necessary to confirm our results on a larger and multicenter population with more generous inclusion criteria to assess the contribution of SEPs to several atypical clinical subgroups. REFERENCES Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force. Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology 1991;41:617–618. Brown WF, Feasby TE. Sensory evoked potentials in Guillain-Barre polyneuropathy. J Neurol Neurosurg Psychiatry 1984;47:288–291. Hughes RA, Wijdicks EF, Barohn R, et al. Practice parameter: immunotherapy for Guillain-Barre syndrome: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2003;61:736–740. Joint Task Force of the EFNS and the PNS. European Federation of Neurological Societies/Peripheral Nerve Society Guideline on management of chronic inflammatory demyelinating polyradiculoneuropathy: report of a joint task force of the European Federation of Neurological Societies and the Peripheral Nerve Society–First Revision. J Peripher Nerv Syst 2010;15:1–9. Koski CL, Baumgarten M, Magder LS, et al. Derivation and validation of diagnostic criteria for chronic inflammatory demyelinating polyneuropathy. J Neurol Sci 2009;277:1–8. Mauguiere F, Allison T, Babiloni C, et al. Somatosensory evoked potentials. The international federation of clinical neurophysiology. Electroencephalogr Clin Neurophysiol Suppl 1999;52:79–90. Molenaar DS, Vermeulen M, de Haan RJ. Comparison of electrodiagnostic criteria for demyelination in patients with chronic inflammatory demyelinating polyneuropathy (CIDP). J Neurol 2002;249:400–403. Pollard JD, Armati PJ. CIDPdthe relevance of recent advances in Schwann cell/ axonal neurobiology. J Peripher Nerv Syst 2011;16:15–23. Sinnreich M, Klein CJ, Daube JR, et al. Chronic immune sensory polyradiculopathy: a possibly treatable sensory ataxia. Neurology 2004;63: 1662–1669. Sonoo M, Genba K, Iwatsubo T, Mannen T. P15 in tibial nerve SEPs as an example of the junctional potential. Electroencephalogr Clin Neurophysiol 1992;84:486–491. Tazawa K, Matsuda M, Yoshida T, et al. Spinal nerve root hypertrophy on MRI: clinical significance in the diagnosis of chronic inflammatory demyelinating polyradiculoneuropathy. Intern Med 2008;47:2019–2024. Tsukamoto H, Sonoo M, Shimizu T. Segmental evaluation of the peripheral nerve using tibial nerve SEPs for the diagnosis of CIDP. Clin Neurophysiol 2010;121:77–84. Viala K, Maisonobe T, Stojkovic T, et al. A current view of the diagnosis, clinical variants, response to treatment and prognosis of chronic inflammatory demyelinating polyradiculoneuropathy. J Peripher Nerv Syst 2010;15:50–56. Walsh JC, Yiannikas C, McLeod JG. Abnormalities of proximal conduction in acute idiopathic polyneuritis: comparison of short latency evoked potentials and F-waves. J Neurol Neurosurg Psychiatry 1984;47:197–200. Yiannikas C, Vucic S. Utility of somatosensory evoked potentials in chronic acquired demyelinating neuropathy. Muscle Nerve 2008;38:1447–1454.
245
Somatosensory Evoked Potentials in Chronic Inflammatory Demyelinating Polyradiculoneuropathy Hayet Salhi,* Philippe Corcia,*† Steve Remer,* and Julien Praline*†
Purpose: Somatosensory evoked potentials (SEPs) offer complementary results to those of nerve conduction studies and contribute to the electrodiagnostic criteria of chronic inflammatory demyelinating polyradiculoneuropathy. Methods: We performed nerve conduction studies and SEPs in patients with symmetrical motor weakness, areflexia, and/or sensory disturbances lasting for at least 8 weeks. We determined two groups according to the electrodiagnostic criteria of the European Federation of Neurological Societies. Group 1 included patients who met the definite or probable electrodiagnostic criteria, and group 2 included patients who met the possible electrodiagnostic criteria. We also compared SEPs results with those of controls (group of healthy subjects). Results: Sixteen patients (14 men; mean age, 59 6 17.3 years) were included in the study. The latencies of potentials N9, N13, N7, and N22 and the intervals N9–N13 and N7–N22 were significantly increased in patients compared with controls. The N9/iP14 amplitude ratio was significantly lower in patients. There was no significant difference in the latencies of SEPs between the two groups of patients. Conclusions: We confirm the contribution of SEPs as complementary information to nerve conduction studies in chronic inflammatory demyelinating polyradiculoneuropathy diagnosis. In addition to the usual abnormalities, a decrease in the N9/iP14 amplitude ratio could potentially be used as an electrodiagnostic criterion. Key Words: Somatosensory evoked potentials, Nerve conduction study, Chronic inflammatory demyelinating polyradiculoneuropathy, Electrodiagnostic criteria. (J Clin Neurophysiol 2014;31: 241–245)
C
hronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is a chronic sensorimotor neuropathy with segmental demyelination. In fact, several variants of CIDP, such as pure motor, pure sensory or sensorimotor forms, are now well known. The diagnosis of CIDP is sometimes difficult. Several groups have proposed diagnostic criteria based on clinical, biological, and nerve conduction study (NCS) findings (Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force 1991; Joint Task Force of the EFNS and the PNS, 2010; Koski et al., 2009). Because the nerve root is frequently involved site in CIDP (Hughes et al., 2003), the sensitivity of NCSs can sometimes be low (Molenaar et al., 2002). Consequently, the study of somatosensory evoked potentials (SEPs), which examines the entire proximal pathway and so the distal extension of the ganglion cells, is particularly interesting in the exploration of radiculo-spinal diseases and offers
From the *Department of Neurology Neurophysiology, Hospital of Tours, France; and †INSERM UMR930, Tours, France. Address correspondence and reprint requests to Hayet Salhi, MD, Service de Neurologie, 2 Boulevard Tonnellé, Tours 37000, France; e-mail: salhi. [email protected]. Copyright Ó 2014 by the American Clinical Neurophysiology Society
ISSN: 0736-0258/14/3103-0241
complementary results to those of NCSs. In the European Federation of Neurological Societies (EFNS) criteria, the use of SEPs seems as supportive criteria, but their diagnostic value has not clearly been defined. Studies on acute polyradiculoneuropathy have shown slow conduction on SEPs across the roots and proximal brachial plexus when motor and sensory conduction velocities over more distal segments were normal (Brown and Feasby, 1984). Abnormalities of the N9 response and/or the spinal N13 response with a prolonged N9–N13 interval have been reported after median nerve stimulation; however, abnormal spinal N22 responses after stimulation of the lower extremities were more common (Walsh et al., 1984). Several studies have been conducted to determine the sensitivity and specificity of SEPs in CIDP (Tsukamoto et al., 2010; Viala et al., 2010; Yiannikas and Vucic, 2008). The inclusion criteria and the methods of these studies were quite different. The sensitivity of SEPs in these studies ranged from 13% to 83%. Viala et al. (2010) studied 146 patients with CIDP; however, they performed SEPs only in a subgroup of 36 patients without major electrodiagnostic criteria for demyelination. This study found proximal conduction abnormalities in 25 of the 36 patients studied with SEPs. The sensitivity of SEPs was low but not all of the patients had SEPs (Viala et al., 2010). Yiannikas and Vucic found the sensitivity of SEPs to be 62% when assessing SEPs and NCS in 47 patients, but this study was performed without using electrodiagnostic criteria to make a diagnosis of CIDP (Yiannikas and Vucic, 2008). Tsukamoto et al. compared 12 patients with CIDP to 17 patients with diabetic polyneuropathy. They recorded tibial nerve SEPs with an unusual technique, and abnormalities suggesting proximal demyelination were found in 83% of patients with CIDP and in only 13% of patients with DPN (Tsukamoto et al., 2010). The aim of our study was to evaluate the contribution of SEPs in the diagnosis of CIDP and to describe the electrophysiological abnormalities of this disease.
MATERIALS AND METHODS Subjects Between April 2009 and April 2011, patients with symmetrical proximal and distal motor weakness associated with or without areflexia, proprioceptive, and epicritic sensory disturbances or patients with pure sensory disturbance lasting for at least 8 weeks were consecutively included in the study. The exclusion criteria included evidence of Charcot–Marie–Tooth disease, the presence of immunoglobulin M monoclonal gammopathy with anti-myelin associated glycoprotein (MAG), active Lyme disease, exposure to drugs that can cause demyelinating neuropathy, or diagnosis of cancer with or without the presence of onco-neuronal antibodies. The data collected were age, gender, height, date of onset of symptoms, date of diagnosis, peak motor disability scale
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
241
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
H. Salhi et al.
TABLE 1.
Characteristics of Patients by Group
Characteristics Age of onset (average) Time to diagnosis (median) Cerebrospinal fluid protein (median)
All (n ¼ 16)
Group 1 (n ¼ 9)
Group 2 (n ¼ 7)
P
59 6 17.3 50 6 37.7 0.60 6 0.35
51 6 11.4 44.3 6 42.4 0.59 6 0.45
69.2 6 18.4 57.3 6 32.3 0.70 6 0.19
0.039 0.29 0.86
(Hughes et al., 2003), and protein level in the cerebrospinal fluid. The final diagnosis of CIDP was made after a follow-up of at least 6 months, taking into account the clinical course according to the EFNS criteria (Joint Task Force of the EFNS and the PNS, 2010).
Nerve Conduction Study All patients were initially assessed with NCS performed on a Keypoint apparatus (Alpine Biomedical, Fountain Valley, CA). The skin temperature was measured at the palm, and the plant was measured using an external thermometer. The temperature of the upper and lower extremities was maintained at or more than 338C. The motor nerves studied included at least the unilateral median, ulnar, tibial, and common peroneal nerves using supramaximal stimulation. The F-wave examination was made for all four motor nerves. The stimulus sites for the median nerve were at the wrist and elbow, and recording was performed from the abductor pollicis brevis. For the ulnar nerve, the stimulus sites were at the wrist, below the elbow, and above the elbow; and recording was performed from the abductor digit minimi. For the tibial nerve, the recording sites were at the ankle and the popliteal fossa; and recording was performed from the flexor hallucis brevis. For the common peroneal nerve, stimulation was performed at the ankle, fibular head, and knee; and recording was performed from the extensor digitorum brevis muscle. Ten or more supramaximal stimuli were delivered distally to elicit F responses from the median and common peroneal nerves. Sensory NCSs were performed using antidromic techniques on the median, ulnar, and sural nerves.
Somatosensory Evoked Potentials Somatosensory evoked potentials were obtained by transcutaneous electrical stimulation of the median nerve at the wrist and the posterior tibial nerve at the ankle, on both sides, and were alternately recorded as recommended by the International Federation of Clinical Neurophysiology (Mauguiere et al., 1999) using the Keypoint software (Alpine Biomedical). Subjects lay on an examination table, and skin temperature was controlled to be more than 328C. For the median nerve SEPs, 4 recording channels were included: ipsilateral Erb-Fz electrodes for channel 1, electrode Cv5 (spinous process of the fifth cervical vertebra); CA (supralaryngeal) for channel 2, electrodes Pc (contralateral parietal); Erb contralateral for channel 3 and Pi electrodes (ipsilateral parietal); and Erb contralateral for channel 4. For the posterior tibial nerve SEPs, the set-up consisted of 4 recording channels: built-in electrodes Pup (popliteal fossa up), Pdw (popliteal fossa down) for channel 1, electrodes L1 (spinous process of the first lumbar vertebra), umbilical for channel 2, electrodes FzCv5 for channel 3 and Pz electrodes; and Ac (contralateral ear) for channel 4. Scalp electrodes were needles, and the other recording electrodes were cup electrodes. Skin-electrode impedance should 242
be less than 5000 U. Stimulation rate was 3 to 5 Hz and the analysis time was 50 milliseconds for upper limb stimulation and 100 milliseconds for lower limb. We have used a high-pass filter at 1 Hz and a low-pass filter at 2000 Hz. Monophasic square-wave electrical pulses of 0.1 to 0.2 milliseconds were to be delivered, and the intensity of stimulation was higher than the motor threshold. Potentials were identified on the curve obtained by averaging 2 series of 1,500 acquisitions reproducible for each stimulated nerve with the target muscle minimally active. We collected the amplitude and the latency of the different components obtained in each channel. The latency was measured at the peak amplitude, and the amplitude was measured between the peak and the end of the potential. The interpretation was based on our laboratory standards that have been obtained with 47 controls using the same technique and apparatus. We also included 16 healthy subjects who were matched to the patient group according to height and age to form a control group.
Analysis The time to diagnosis was calculated for each patient. The criteria of the EFNS were applied to the results of the NCS of each patient to establish two groups. Group 1 included patients who met the definite or probable electrodiagnostic criteria, and group 2 included patients who met the possible electrodiagnostic criteria or did not meet the probable electrodiagnostic criteria. We compared the time to diagnosis, age, and gender distributions between the two groups. For the SEPs, the N9 (clavical response), N13 (cervical response), N7 (popliteal response), and N22 (lumbar response) latencies were collected to calculate the time intervals N9–N13 (clavical to cervical response) and N7–N22 (popliteal to lumbar response). We also collected the N9/iP14 amplitude ratio (P14 inverse potential was obtained on the first channel registration Erb-Fz). For each variable, the left-right difference was calculated
TABLE 2. Results of Somatosensory Evoked Potentials According to Laboratory Norms Increased/Absent SEPs Latencies N9 N13 N7 N22 Intervals N9–N13 N7–N22
Normal
Group 1
Group 2
Group 1
Group 2
6 7 5 8
6 6 5 7
3 2 4 1
1 1 2 0
4 8
6 6
5 1
1 1
SEP, Somatosensory evoked potential.
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
TABLE 3. Results of Somatosensory Evoked Potentials in Patients Compared With Controls
SEPs Latencies N9 N13 N7 N22 Intervals N7–N22 N9–N13 Amplitude ratio Ratio N9/P14i
Mean 6 SD Mean 6 SD (milliseconds) (milliseconds) Patients (Number) Controls (n ¼ 16) 6 6 6 6
P
12.7 6 1.62 (n ¼ 14) 16.3 6 1.40 (n ¼ 13) 10.1 6 2.82 (n ¼ 11) 26.8 6 3.83 (n ¼ 6)
10.3 13.6 8.5 22.8
0.55 0.62 0.50 1.14
,0.0001 ,0.0001 0.007 0.0001
15.9 6 3.66 (n ¼ 6) 3.6 6 0.69 (n ¼ 13)
13.3 6 0.82 3.4 6 0.38
0.0001 0.009
1.06 6 0.97 (n ¼ 15)
2.21 6 1.10
,0.0001
SEP, Somatosensory evoked potential.
to look for asymmetry. First, we analyzed the SEP results for each patient and classified them as normal or abnormal according to our laboratory norms. SEPs latency or interval was considered as abnormal if it was more than 2 SDs above the mean of the laboratory norms. Second, all data were compared with those of the control group and between each patient group. Left and right responses were pooled for the comparison between patients and controls and between the two groups of patients. Wilcoxon test was applied because there was a group population of fewer than 30. The significance level was set at P , 0.05.
Somatosensory Evoked Potentials in CIDP
cerebrospinal fluid protein levels (N , 0.5 g/L) with a leukocyte count ,10/mm3. After applying the electrodiagnostic criteria of the EFNS, group 1 included 9 patients (8 men, mean age, 51 6 11.4 years), and group 2 included 7 patients (1 met the possible category criteria) (6 men, mean age 69.2 6 18.9 years) (Table 1). The changes in SEP latencies are summarized in Table 2. In group 2, NCSs were normal in 28.6% who showed axonal polyneuropathy in 57% of cases, and for 1 patient, a unilateral ulnar nerve entrapment was found at the elbow. The latencies of potentials N9, N13, N7, and N22 were significantly increased in patients compared with the controls (P , 0.001), as were the N9–N13 interval (P ¼ 0.009) and the N7–N22 interval (P , 0.0001) (Table 3). Two examples of SEPs responses are shown (Figs. 1 and 2). In our patients, there was no significant right/left asymmetry for potential latencies N9, N13, N7, and N22 (P . 0.05) (data not shown). In group 1, the mean of the N9/iP14 amplitude ratio was 1, and in group 2, it was 1.41. The N9/iP14 amplitude ratio was significantly lower in the patients when compared with the controls (P , 0.0001) (Table 3 and Fig. 3). There was no significant difference in the latencies of the SEPs between the two groups of patients (Table 4). Of the 5 patients with the sensory variant of CIDP, 1 patient met the electrodiagnostic criteria (group 1) and 4 patients did not meet the electrodiagnostic criteria (group 2). The latencies of the N9 and N13 potentials were increased in all patients, and the latency of N7 was increased in 4 patients and normal in 1 patient. The latency of the N22 potential was increased in 2 patients (N22 was absent in the remaining 3 patients). The amplitude ratio, N9/iP14, was reduced for 4 of these patients.
RESULTS
DISCUSSION
A total of 18 patients fulfilled the inclusion criteria on their first visit. Two patients were excluded because of the presence of monoclonal IgM with anti-MAG. The mean age of the remaining 16 patients (14 men) was 59 6 17.3 years. The peak motor disability scale (Hughes et al., 2003) was grade 4 in 2 patients, grade 3 in 6 patients, grade 2 in 5 patients, and grade 1 in 3 patients. Sensory symptoms or signs in the legs were present in 11 patients, and sensory ataxia was evident in 4 patients. Five patients lacked motor symptoms, and thus, these patients were considered to have the pure sensory CIDP variant (Joint Task Force of the EFNS and the PNS, 2010). Thirteen patients had elevated
Our results demonstrate the contribution of SEPs in the identification of a root demyelinating process. In group 1, of the 9 patients, 78% had increased latencies of N9 and N13 or absent potentials, 56% had abnormal latencies of N7 and N22, and the mean N9/iP14 amplitude ratio was 1. For all of these patients with electrodiagnostic criteria of CIDP, we found abnormalities suggesting proximal demyelination. In group 2, of the 7 patients without these electrodiagnostic criteria, 86% had increased latencies of N9 and N13 or absent potentials, 71% had abnormal latency of N7, and 100% had abnormal N22 responses. The mean of the N9/iP14 amplitude ratio was 1.41. The abnormalities of SEPs were the same
FIG. 1. Somatosensory evoked potentials of right median nerve before (A) and after averaging (B) of 2 series of 1,500 acquisitions (52-year-old man). See the increased latencies of all potentials and decreased amplitude of N9 contrasting with relative preservation of amplitude of iP14 on the first channel (arrow). CA, supralaryngeal electrode; Cv5, electrode on spinous process of the fifth cervical vertebra; Erb c, contralateral Erb; Erb G, ipsilateral Erb; Fz, frontal electrode; Pc, contralateral parietal; Pi, ipsilateral parietal electrode. Copyright Ó 2014 by the American Clinical Neurophysiology Society
243
H. Salhi et al.
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
FIG. 2. Somatosensory evoked potentials of left posterior tibial nerve before (A) and after averaging (B) of 2 series of 1,500 acquisitions (52-year-old man). See the increased latencies of N22 and next potentials with preservation of amplitude. Ac, contralateral ear electrode; Cv5, electrode on spinous process of the fifth cervical vertebra; Fz, frontal electrode; L1, electrode on spinous process of the first lumbar vertebra; Omb, umbilical electrode; Pdw, electrode on popliteal fossa down; Pup, electrode popliteal fossa up; Pz, parietal electrode. in this group 2 than in group 1; so here, we can also say that all of these patients had signs of proximal demyelination. Interestingly, our results demonstrated a root demyelinating process in seven patients although they did not meet electrodiagnostic criteria of CIDP. Five patients lacked motor symptoms, and thus, these patients were considered to have the pure sensory CIDP variant. Of the 5 patients with the pure sensory CIDP variant, 4 were in group 2, (so without electrodiagnostic criteria); and with the realization of SEPs, we could have established the diagnosis of CIDP. For this variant, the use of SEPs is the most useful because the study of NCSs is often normal (Sinnreich et al., 2004). Thus, the collection of SEPs enabled the identification of abnormalities of proximal demyelination in our 16 patients, which included patients who did not meet the electrodiagnostic criteria for CIDP (group 2). Our results confirmed those of two previous studies that highlighted differences in the SEPs of patients who did not meet the commonly used electrodiagnostic criteria. However, in these two previous studies, the electrodiagnostic criteria of the EFNS were not applied. Tsukamoto et al. (2010) evaluated the diagnostic value of SEPs of the posterior tibial nerve in demyelinating neuropathies and axonal diabetic polyneuropathy. In addition to the potentials that are classically studied, they used an additional bypass that enabled them to study the N15 potential corresponding to the afferent volley of action potentials to the passage of the greater sciatic foramen
(Sonoo et al., 1992). The criteria used to define a demyelinating process were a normal latency of the N8 potential with a conduction time increase to a segment proximal to this potential, an increase in the ratio of the proximal/distal latency, a normal amplitude of the N8 response with a decrease in amplitude in another more proximal potential (P15, N21), or a decrease in the amplitude ratio of the proximal/distal response. These abnormalities were seen as signs of proximal damage that is specific to CIDP. The sensitivity of SEPs for the diagnosis of CIDP was 83%, and a false-positive rate of 18% was found in the control group of patients with diabetic polyneuropathy (Tsukamoto et al., 2010). The methodology of the study by Yiannikas and Vucic (2008) was quite similar to ours. Forty-seven patients with clinical signs suggestive of CIDP were investigated by NCSs and SEPs in four limbs. The SEP responses were considered abnormal and consistent with proximal demyelination if the latency exceeded the upper limit of normal, if responses were absent, or if there was an increased N9–N13 conduction time. Five homogeneous groups of patients were established according to the results of normal or abnormal NCSs and SEPs. For 7 patients (groups 1 and 4) without abnormalities of nerve conduction time in NCSs, the SEPs showed signs of demyelination with longer P39 latencies for 6 patients, no N22 response in the lower limbs for 7 patients, and longer N9 and N13 latencies in the upper limbs for 5 patients. The diagnosis of CIDP was given, despite the normality of the NCS responses (Yiannikas and Vucic, 2008). In our cohort, we found abnormalities of nerve conduction (increased N9 and N7 latency potentials) and/or root demyelination (increased latency of N13 and N22 potentials and the N9–N13 interval) in CIDP. These abnormalities correspond to those previously described (Yiannikas and Vucic, 2008). In addition, the decrease in the N9/iP14 amplitude ratio allowed identification of a decrease in
TABLE 4. Results of Somatosensory Evoked Potentials in Group one Compared to Group 2
SEPs Latencies N9 N13 N7 N22
FIG. 3. Box plots of the N9/iP14 amplitude ratio in controls and patients. 244
Median 6 SD (milliseconds) Group 1 12.7 16 9.7 26.1
6 6 6 6
2.83 2.20 3.68 5.88
(n (n (n (n
¼ ¼ ¼ ¼
16) 15) 13) 9)
Median 6 SD (milliseconds) Group 2 12.7 16.4 10.6 30.4
6 6 6 6
3.04 2.89 5.58 2.18
(n (n (n (n
¼ ¼ ¼ ¼
13) 12) 9) 4)
P 0.54 0.22 0.84 0.17
n, number of limbs (left and right); SEP, somatosensory evoked potential.
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
the amplitude of N9 in contrast with the relatively preserved amplitude of P14. Using an Fz electrode reference on the first recording channel for the upper limbs, the P14 potential was recorded on the same derivation as the N9 potential. Thus, this method allowed us to calculate the N9/iP14 amplitude ratio more rigorously than if it was calculated using potentials obtained on different channels. The decrease of the N9/iP14 amplitude ratio reflects a secondary resynchronization and allows elimination of a central impairment. The decrease of the ratio might reflect that the observed loss of amplitude was related to a desynchronization of potentials and thus to a demyelinating process. The decrease in the N9/iP14 amplitude ratio was found to be statistically significant in the control group (P , 0.0001, Tables 3 and 4), so this abnormality seems to be specific to CIDP and could be used as an electrodiagnostic criterion. The absence of right-left asymmetry in the N9, N13, N22, and N7 potential latencies confirmed the diffuse nature of the demyelinating process (Pollard and Armati, 2011). Our results also increase the importance of the investigation of all four limbs; the SEPs of the lower limbs are rarely obtained in a reliable and reproducible way. Although the EFNS criteria were developed to be very specific to select patients for therapeutic trials, some patients who do not meet the criteria lead to diagnostic difficulty. This also reinforces the need to explore supportive means of diagnosis, although the comparison to a gold standard has not been possible to date. Our study has limitations that are related to the small number of patients. The use of a recording on the elbow for the upper limb and on the bottom for the lower limb (Tsukamoto et al., 2010) would also be interesting to elucidate the segment where demyelination is most pronounced. Furthermore, the high number of missing responses in the lower limbs (no N7 response in 28.6% of cases and no N22 in 71.4% of cases) led us to carefully consider our results. Furthermore, the N22 may not be identified in control subjects, especially in aged subjects; thus loss of N22 patients in aged patients could not be abnormal. Some teams have proposed performing a spine MRI to find root hyperintensities suggestive of CIDP (Joint Task Force of the EFNS and the PNS, 2010; Tazawa et al., 2008). We have not studied the value of MRI in our groups. We might consider studying each patient with both SEPs and MRI to compare their contributions toward diagnosis and to assess any associations that may be found between the two methods. Moreover, we did not try to assess the usefulness of F-waves in our cohort. F-wave study could reflect both distal and proximal nerve conduction, especially it assess the motor nerve roots in proximal, while SEPs exploring the sensitive nerve roots. In our work, the nerve conduction studies and F-waves had been used to constitute the two groups of patients. It seemed difficult to make correlations between SEPs and F-wave abnormalities given the small number of patients and the low probability of F-wave abnormalities (Walsh et al., 1984). We elected to study the clinical benefits of SEPs in daily practice by comparing results between two groups defined according to
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Somatosensory Evoked Potentials in CIDP
the EFNS criteria. This method allowed us to study the benefits of SEPs in what can be termed “ecological conditions,” which better correspond to this technically difficult procedure. To conclude, SEP can be a very useful test for CIDP diagnosis in patients without electrodiagnostic criteria (NCSs), and we confirmed the value of SEP results as complementary to those from NCSs. An increase in N9, N13, N7, and N22 latencies and a decrease in the N9/iP14 amplitude ratio could be used as electrodiagnostic criteria for CIDP in that case. Although it is interesting to perform SEPs in the four limbs, the abnormalities of the upper limbs seem to be more reliable for the diagnosis of CIDP. It will be necessary to confirm our results on a larger and multicenter population with more generous inclusion criteria to assess the contribution of SEPs to several atypical clinical subgroups. REFERENCES Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force. Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology 1991;41:617–618. Brown WF, Feasby TE. Sensory evoked potentials in Guillain-Barre polyneuropathy. J Neurol Neurosurg Psychiatry 1984;47:288–291. Hughes RA, Wijdicks EF, Barohn R, et al. Practice parameter: immunotherapy for Guillain-Barre syndrome: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2003;61:736–740. Joint Task Force of the EFNS and the PNS. European Federation of Neurological Societies/Peripheral Nerve Society Guideline on management of chronic inflammatory demyelinating polyradiculoneuropathy: report of a joint task force of the European Federation of Neurological Societies and the Peripheral Nerve Society–First Revision. J Peripher Nerv Syst 2010;15:1–9. Koski CL, Baumgarten M, Magder LS, et al. Derivation and validation of diagnostic criteria for chronic inflammatory demyelinating polyneuropathy. J Neurol Sci 2009;277:1–8. Mauguiere F, Allison T, Babiloni C, et al. Somatosensory evoked potentials. The international federation of clinical neurophysiology. Electroencephalogr Clin Neurophysiol Suppl 1999;52:79–90. Molenaar DS, Vermeulen M, de Haan RJ. Comparison of electrodiagnostic criteria for demyelination in patients with chronic inflammatory demyelinating polyneuropathy (CIDP). J Neurol 2002;249:400–403. Pollard JD, Armati PJ. CIDPdthe relevance of recent advances in Schwann cell/ axonal neurobiology. J Peripher Nerv Syst 2011;16:15–23. Sinnreich M, Klein CJ, Daube JR, et al. Chronic immune sensory polyradiculopathy: a possibly treatable sensory ataxia. Neurology 2004;63: 1662–1669. Sonoo M, Genba K, Iwatsubo T, Mannen T. P15 in tibial nerve SEPs as an example of the junctional potential. Electroencephalogr Clin Neurophysiol 1992;84:486–491. Tazawa K, Matsuda M, Yoshida T, et al. Spinal nerve root hypertrophy on MRI: clinical significance in the diagnosis of chronic inflammatory demyelinating polyradiculoneuropathy. Intern Med 2008;47:2019–2024. Tsukamoto H, Sonoo M, Shimizu T. Segmental evaluation of the peripheral nerve using tibial nerve SEPs for the diagnosis of CIDP. Clin Neurophysiol 2010;121:77–84. Viala K, Maisonobe T, Stojkovic T, et al. A current view of the diagnosis, clinical variants, response to treatment and prognosis of chronic inflammatory demyelinating polyradiculoneuropathy. J Peripher Nerv Syst 2010;15:50–56. Walsh JC, Yiannikas C, McLeod JG. Abnormalities of proximal conduction in acute idiopathic polyneuritis: comparison of short latency evoked potentials and F-waves. J Neurol Neurosurg Psychiatry 1984;47:197–200. Yiannikas C, Vucic S. Utility of somatosensory evoked potentials in chronic acquired demyelinating neuropathy. Muscle Nerve 2008;38:1447–1454.
245
![[Chronic inflammatory demyelinating polyradiculoneuropathy revealing essential thrombocythemia].](/assets/img/hold.png)
