AUTHOR(S): Morioka, Takato, M.D.; KuritaTashima, Shizuka, M.D.; Fujii, Kiyotaka, M.D.; Nakagaki, Hiroyuki, M.D.; Kato, Motohiro, M.D.; Fukui, Masashi, M.D. Departments of Neurosurgery (TM, KF, HN, MF) and Clinical Neurophysiology (TM, SK-T, MK), Neurological Institute, Faculty of Medicine, Kyushu University, Fukuoka, Japan Neurosurgery 30; 0218-222, 1992 ABSTRACT: We recorded somatosensory evoked potentials (SEPs) from the skin surface and spinal evoked potentials (SpEP) from the posterior epidural space after median nerve stimulation in 11 patients with cervical syringomyelia. We compared SEPs with SpEP to assess the possible feasibility of using these techniques to localize the offending lesion. SEP abnormalities were present on one or both sides of 7 patients (9 of 22 upper limbs; 41%). The abnormal SEP pattern was the attenuation or loss of N13, which was of little value for delineating the lesion. In patients with abnormal SEPs, three types of abnormal SpEPs were noted. In the Type A abnormality (three limbs), potentials were attenuated in all cervical segments, suggesting that the syrinx itself had enlarged to involve the posterior column. In Type B (two limbs), there was reduced amplitude or absent upper cervical potentials, probably a result of the accompanying tonsilar herniation. Finally, the Type C (four limbs) abnormality was a mixture of Type A and B abnormalities in that the attenuated cervical potentials were again affected in the most upper cervical segment. We concluded that SpEP revealed various kinds of involvement of the dorsal column pathway in the syringomyelic patients, a finding not expressed with conventional SEPs. KEY WORDS: Chiari I malformation; Median nerve stimulation; Somatosensory evoked potentials; Spinal evoked potentials; Syringomyelia Syringomyelia is characterized by dissociated sensory loss, which is caused by cavitation of the spinal gray matter. Somatosensory evoked potentials (SEPs) are thought to reflect the function of the dorsal column pathway (7,8,13,21). Consequently, SEP findings in patients with a syrinx are presumed to be normal. Recently, some abnormal SEP findings have been reported (2,12,14,23,25,29-32). Because the syringomyelic cavity often extends through many spinal segments and accompanies the caudal descent of the cerebellar tonsils, precise functional localization of the offending lesion is important in providing the proper rationale for surgical treatment. The generator sources of the subcortical SEP
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
components, however, remain controversial (4,7,8,13,14, 21) ; hence, SEPs are of little value for this purpose (2). Maglardery et al. (19) obtained spinal evoked potentials (SpEPs) from the dorsum of the spinal cord with electrodes introduced intrathecally in human volunteers; this method has many risks and is not applicable to routine clinical use. Shimoji et al. (27) have succeeded in recording SpEPs from the dorsal epidural space, on the basis of the technique of epidural anesthesia. Recently, Lüders and his colleages (5,9,17,18) developed a simple technique to record SpEPs by needle electrodes inserted intraoperatively into the intraspinal ligament. This technique allows SpEPs to be recorded by placing electrodes very close to the dura. The recorded potentials have very similar waveform and amplitude to epidurally recorded SpEPs (26). This technique offers significant advantages over the transdermal placement of epidural electrodes (9): first, the use of electrodes inserted into the intraspinal ligaments under direct visualization is both safe and rapid, adding no significant time to the operation; and, second, the method of Shimoji et al. is not applicable to procedures involving the high cervical cord, which is often affected with tonsilar herniation. SpEPs have been used to localize the offending lesion of cervical spondylotic myelopathy (10,26,28) and cervical spinal cord injury (15). Because reports of SpEP findings in patients with syringomyelia have been rare (17,31), the significance of using SpEPs to examine syringomyelia is not fully understood. We report the results of SpEPs studies in our syringomyelic patients and discuss the correlation with those of SEPs studies. PATIENTS AND METHODS This study was performed in 11 patients with cervical syringomyelia (Table 1). There were 4 men and 7 women, aged 24 to 56 (mean, 37) years. Ten patients had Chiari I malformations, and the other one a posttraumatic syringomyelia (lower lumbar and cauda equina injury). Dissociated loss of pain and temperature sensation with preservation of deep sensation was seen in nine upper limbs, and segmental loss of all modalities of cutaneous sensation was observed in seven limbs. Six limbs failed to show sensory disturbance. Hydrocephalus or evidence of intracranial hypertension was not observed in any of the patients, although defective ventricular drainage is thought to be an important factor in syringomyelia. The level and extent of the syrinx were determined by magnetic resonance imaging. Informed consent was obtained after the nature of the procedure had been fully explained. Short-latency SEPs were recorded following the guidelines set by the American Electroencephalographic Society (1). The median nerve was stimulated at the wrist through bipolar electrodes with 0.1 msec square wave pulses. The stimulus strength was adjusted to the level that induced minimal muscle contraction of the thumb at a rate of four per second. Disc electrodes were placed on the bilateral Erb's point (at the supraclavicular regions), the cervical skin over the 5th process (Cv5),
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Neurosurgery 1992-98 February 1992, Volume 30, Number 2 218 Somatosensory and Spinal Evoked Potentials in Patients with Cervical Syringomyelia Experimental and Clinical Study
RESULTS SEPs were abnormal in 7 patients (64%), 5 of whom showed abnormal SEPs only to unilateral stimulation (Table 1). Nine of 22 upper limbs (41%) showed abnormal SEP findings. Abnormalities were obtained from all of the seven limbs that showed proprioceptive sensory deficits. Two out of 9 limbs that showed only superficial sensory disturbance showed normal SEPs. Six limbs without sensory deficits showed normal SEPs. The abnormalities included absent N13/P13 and scalp N20 potential (Fig. 1, left upper traces; Fig. 2, upper traces) in four limbs, an absent N13 with a significantly reduced amplitude of N20 in two limbs, reduced N13 and N20 amplitude in two limbs, and absent N13/P13 with preserved N20 (Fig. 3, left upper traces) in one limb. No limbs showed delayed interpeak latencies of the major components.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
No correlation could be found between the location of the syrinx cavity and SEP abnormalities. There was also no apparent relation between the presence of the Chiari I malformation and SEP abnormality. Severe abnormal SEPs were seen unilaterally in the case with posttraumatic syringomyelia (without tonsilar herniation) (Fig. 1, upper traces). In patients with normal SEPs (Figs. 1 and 3, right upper traces), SpEPs showed positive-negativepositive triphasic waves with the negativity being the predominant component (Figs. 1 and 3, right lower traces). The peak latency of the negativity showed a slightly progressive increase in the caudorostral direction. SpEP abnormalities were found in all patients whose SEPs were abnormal. The SpEP abnormalities could be divided into three groups depending on the cervical level at which the abnormal waves were recorded. Type A abnormality showed reduced amplitude of the cervical potential (Fig. 1, left lower traces), suggesting that the syrinx itself had enlarged to involve the posterior column. Type A abnormality was found in three limbs. Type B abnormality showed attenuated potentials in the upper cervical level (Fig. 3, left lower traces). This may be due to compression of the medulla and upper cervical cord at the foramen magnum by the tonsilar herniation. Two of our patients had limbs that showed the Type B abnormality. Type C is a mixed type of these two, in that the attenuated cervical potentials were again affected in the uppermost cervical segments (Fig. 2, lower traces). Type C was found in four limbs. There was no relationship between the types of SEP and SpEP abnormalities (Table 3). DISCUSSION The most common SEP abnormality that we and others (12,23,25,29-31) have found in patients with syringomyelia is the marked attenuation or absence of N13 with often normal N20 potentials, although there are a small number of reports that show a prolonged central conduction time (2,12). Our results also confirm the previous reports (12,32) in that the SEPs were able to detect subclinical dysfunction of the dorsal column pathway in syringomyelic patients. It appears that the N13 may not reflect the propagation of the afferent volley in the dorsal columns, as was first thought, but may be mediated by the collateral projection to the dorsal horn (13). Therefore, N13 may be susceptible to spinal cord lesions involving the gray matter with sparing of the dorsal columns. This hypothesis could explain the dissociation between cervical and cortical SEP components, with an abnormal N13 and a normal N20, which was observed in our patients. At present, it is generally accepted that the N13 recorded with a cephalic reference, as we used, is a composite potential, reflecting "near-field" activity from the cervical cord and "far-field" activity from the lower brain stem lemniscal pathways (1,4). By using a noncephalic reference, spinal N13 negativity and P13 positivity can be recorded from the posterior and anterior neck surfaces, respectively (4,6,25,30,31). These components are thought to be generated by the
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Fz of the International 10-20 system, and the bilateral hand sensory area on the scalp (C3' or C4'; 2 cm posterior to C3 and C4, respectively). A montage consisting of C3'- or C4'-contralateral Erb's point, C3'or C4'-Fz, Cv5-Fz, and ipsilateral Erb's point-Fz was used for a four-channel system of Cadwell Quantum 84 (Cadwell Laboratories Inc., Kennewick, WA). The potentials were amplified with the filter set at 10 to 3000 Hz, and 500 sweeps were averaged with an analysis time of 30 msec. SpEPs were recorded by using Lüder's method (5,9, 17,18) intraoperatively with a hard stainless steel needle electrode, 1 mm in diameter, with a 0.25-mm bared tip (UKE-0138; Unique Medical Co., Tokyo, Japan). The electrode was inserted by the surgeon between the exposed spinous processes. The tip of the electrode was introduced up to the yellow ligament as close to the epidural space as possible, and this was confirmed with an intraoperative roentrenogram. The reference electrode was placed over Fz. Stimulus and recording conditions were identical to that used for SEPs. Between 50 and 100 responses were summated with an analysis time of 20 to 30 msec. Caudocranial sequential recording of SpEP within the operative field was made. The anesthesia used was of the nitrous oxide-fentanyl-droperidol type. Observations were obtained for a minimum of two runs to ensure that all responses were consistent. The same procedure was employed on the opposite side. The results of SEPs obtained from the patients with syringomyelia were compared with control values obtained from age-matched healthy volunteers (Table 2). The nomenclature of each recognizable component followed standards set by the American Electroencephalographic Society (1). Amplitude was measured between the estimated baseline and the major peak. A record was defined "abnormal" when a peak was focally absent or delayed with a latency value exceeding a mean + 3 standard deviations of control values. The corresponding criterion for the abnormality of each component amplitude was a reduction in size by more than 50%. SpEP were analyzed visually with regard to the waveform and amplitude of responses with reference to our previous reports (20-22).
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Chiari I malformation. Foramen magnum decompression with or without closure of the obex is a procedure currently accepted by many neurosurgeons (24). A frequently discussed issue is syrinx shunting, such as syringosubarachnoidal or syringoperitoneal shunt (11). In patients with Types A and C SpEP abnormalities, the syrinx is thought to be large enough to involve the posterior column. Therefore, syrinx shunting may be an additional recommended procedure. Furthermore, if we are able to record the positive killed end potentials along with the SpEPs, the location at which myelotomy for syrinx decompression may be determined. It is of interest to note that the unilateral abnormal SEP and SpEP findings were found in 5 out of 7 patients with abnormal SEPs and SpEP. According to pathological studies (16), the syrinx is not always situated bilaterally within the cord but has laterality at some spinal level. Tonsilar herniation is not always symmetrical, either. Our findings are consistent with these pathological findings. Received for publication, April 8, 1991; accepted, final form, August 20, 1991. Reprint requests: Takato Morioka, M.D., Department of Neurosurgery, Neurological Institute, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Fukuoka 812, Japan. ACKNOWLEDGMENTS We thank Drs. S. Tobimatsu and H. Tomoda, Department of Clinical Neurophysiology, Kyushu University, for critically reading an earlier draft of this article and A. N. Kalehua for useful comments. REFERENCES: (1-34) 1.
2.
3.
4.
5.
6.
7.
American Electroencephalographic Society: Recommended standards for short-latency somatosensory evoked potentials. J Clin Neurophysiol 1:4-53, 1984. Anderson NE, Frith RW, Synek VM: Somatosensory evoked potentials in syringomyelia. J Neurol Neurosurg Psychiatry 49:1407-1410, 1986. Deecke L, Tator CH: Neurophysiological assessment of afferent and efferent conduction in the injured spinal cord of monkeys. J Neurosurg 19:65-74, 1973. Desmedt JE: Generator sources of SEP in man, in Cracco RQ, Bodis-Wollner I (eds): Evoked Potentials. New York, Liss, 1986, pp 235-245. Dinner DS, Lüders H, Lesser RP, Morris HH: Invasive methods of somatosensory evoked potentials monitoring. J Clin Neurophysiol 3:113-130, 1986. Emerson RG, Pedley TA: Effect of cervical spinal cord lesions on early components of the median nerve somatosensory evoked potential. Neurology 36:20-26, 1986. Friedman WA: Somatosensory evoked potentials in neurosurgery. Clin Neurosurg 34:187-238, 1986.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
central gray matter in the cervical cord (4). The absence of N13/P13 with a noncephalic reference has been demonstrated in syringomyelic patients with a normal scalp far-field P14 of more rostral origin (25,30, 31) . Emerson and Pedley (6) have also shown a loss of the anterior neck P13 component in spite of a normal noncephalic N13. Bipolar recordings of SEPs have two N13 subcomponents, which are generator dipoles directed horizontally (N13a) and axially (N13b) (14). Kaji and Sumner (14) have suggested that syringomyelic patients may have absent N13a with normal N13b. Thus, by using a noncephalic reference and bipolar recording, one may be able to increase the accuracy of the lesion determination in the cervical gray matter, which deserves further investigation in syringomyelic patients (12,25). In our patients with normal SEPs, SpEPs showed positive-negative-positive triphasic waves. The progressive increase in the latency of the dominant negativity from progressively more rostral segments of the cord indicates that an ascending volley in the dorsal column was being recorded, as described earlier (17,20-22). Three types of abnormal SpEP findings were seen in our patients. This provided us with useful information regarding which underlying abnormality was responsible for the dorsal column dysfunction of the patient: the syrinx itself or the tonsilar herniation (or both). When the syrinx is large enough to involve or compress the dorsal column pathways, the cervical ascending volleys can be attenuated (Type A abnormality in our classification). When the herniated tonsil compresses the cord severely at the foramen magnum, upper cervical potentials can be attenuated (Type B abnormality). We also demonstrated a mixture of Types A and B abnormalities, in that the attenuated cervical potentials were again affected in the uppermost cervical segments (Type C abnormality). We could not find a significant confounding effect on the SEPs in syringomyelia from coexisting Chiari malformation, confirming a previous report by Jabbari et al. (12). Anderson et al. (2) reported that all 6 of their patients with abnormal central conduction time in SEPs had an associated Chiari malformation. Our present SpEP findings reveal that tonsilar compression does not develop into a conduction delay but into an attenuation or loss of the potentials. Although we detected an attenuation of the SpEP in our syringomyelic patients, positive "killed end potentials" could not be recorded. Killed end potentials occur when an impulse approaches but does not arrive at a recording electrode because the nerve fibers terminate before reaching the electrode (3,7,34) . Thus, this kind of potential is used to localize the offending lesion precisely in the patients with cervical spondylotic myelopathy (26,28), acute cervical cord injury (15), and intramedullary tumor (20,33). Because dorsal column dysfunction is not the primary lesion in the syringomyelic patient, it is conceivable that complete and abrupt disruption of the dorsal column fibers is rare. Numerous surgical interventions have been advocated for the treatment of syringomyelia with
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Greenberg JA, Kaplan PW, Erwin CW: Somatosensory evoked potentials and the dorsal column myth. J Clin Neurophysiol 4:189-196, 1990. Hahn JF, Lesser R, Klem G, Lueders H: Simple technique for monitoring intraoperative spinal cord function. Neurosurgery 9:692-695, 1981. Hattori S, Saiki K, Kawai S: Diagnosis of the level and severity of cord lesion in cervical spondylotic myelopathy: Spinal evoked potentials. Spine 6:478-485, 1979. Isu T, Iwasaki Y, Akino M, Abe H: Syringosubarachnoid shunt for syringomyelia associated with Chiari malformation (type 1) Acta Neurochir (Wien) 107:152-160, 1991. Jabbari B, Geyer C, Gunderson C, Chu A, Brophy J, McBurney JW, Jonas B: Somatosensory evoked potentials and magnetic resonance imaging in syringomyelia. Electroencephalogr Clin Neurophysiol 77:277-285, 1990. Jones SJ: Somatosensory evoked potentials: The abnormal waveform, in Halliday AM (ed): Evoked Potentials in Clinical Testing. Edinburgh, Churchill Livingstone, 1982, pp 429-470. Kaji R, Sumner AJ: Bipolar recordings of short-latency somatosensory evoked potentials after median nerve stimulation. Neurology 37:410-418, 1987. Katayama Y, Tsubokawa T, Yamamoto T, Hirayama T, Maejima S: Preoperative determination of the level of spinal cord lesions from the killed end potential. Surg Neurol 29:91-94, 1988. Larroche JC: Syringomyelia and syringobulbia, in Adams JH, Corsellis JAN, Duchen LW (eds): Greenfield's Neuropathology. London, Edward Arnold, 1984, ed 4, pp 396-402. Lesser RP, Lueders H, Hahn J, Klem G: Early somatosensory evoked potentials evoked by median nerve stimulation: Intraoperative monitoring. Neurology 31:1519-1523, 1981. Lueders H, Gurd A, Hahn J, Andrish J, Weiker G, Klem G: A new technique for intraoperative monitoring of spinal cord function. Multichannel recording of spinal cord and subcortical evoked potentials. Spine 7:110-115, 1982. Magladery JW, Porter WE, Park AM, Teasdall RD: Electrophysiological studies of nerve and reflex activity in normal man. IV. The twoneuron reflex and identification of certain action potentials from spinal roots and cord. Bull Johns Hopkins Hosp 88:499-519, 1951. Morioka T, Fujii K, Mitani M, Fukui M: Intraoperative localization of a cervicomedullary glioma from the killed end potential: Illustrative case. Neurosurgery 26:1038-1041, 1990. Morioka T, Shima F, Kato M, Fujui M: Direct recording of somatosensory evoked potentials
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32. 33. 34.
in the vicinity of dorsal column nuclei in man: Their generator mechanisms and contribution to the scalp far-field potentials. Electroencephalogr Clin Neurophysiol 80:215220, 1991. Morioka T, Tobimatsu S, Fujii K, Fukui M, Kato M, Matsubara T: Origin and distribution of brainstem somatosensory evoked potentials in humans. Electroencephalogr Clin Neurophysiol 80:221-227, 1991. Nakanishi T, Tamaki M, Ozaki Y, Arasaki K: Origins of short latency somatosensory evoked poteials to median nerve stimulation. Electroencephalogr Clin Neurophysiol 56:7485, 1983. Pillay P, Awad IA, Little JR, Hahn J: Symptomatic Chiari malformation in adults: A new classification based on magnetic resonance imaging with clinical and prognostic significance. Neurosurgery 28:639645, 1991. Restuccia D, Mauguière F: The contribution of median nerve SEPs in the functional assessment of the cervical spinal cord in syringomyelia. A study of 24 patients. Brain 114:361-379, 1991. Satomi K, Okuma T, Kenmotsu K, Nakamura Y, Hirabayashi K: Level diagnosis of cervical myelopathy using evoked spinal cord potentials. Spine 13:1217-1224, 1988. Shimoji K, Higashi H, Kano T: Epidural recording of spinal electrogram in man. Electroencephalogr Clin Neurophysiol 30:236239, 1971. Shinomiya K, Furuya K, Sato R, Okamoto A, Kurosa Y, Fuchioka M: Electrophysiological diagnosis of cervical OPLL myelopathy using evoked spinal cord potentials. Spine 13:12251233, 1988. Stöhr M, Buettner UW, Riffel B, Koletzki E: Spinal somatosensory evoked potentials in cervical cord lesions. Electroencephalogr Clin Neurophysiol 54:257-265, 1982. Urasaki E, Wada S, Kadoya C, Matsuzaki H, Yokota A, Matsuoka S: Absence of spinal N13-P13 and normal scalp far-field P14 in a patient with syringomyelia. Electroencephalogr Clin Neurophysiol 71:400404, 1988. Urasaki E, Wada S-I, Kadoya C, Tokimura T, Yokota A, Matsuoka S, Fukumura A, Hamada S: Skin and epidural recording of spinal somatosensory evoked potentials following median nerve stimulation: Correlation between the absence of spinal N13 and impaired pain sensation. J Neurol 237:410415, 1990. Veilleux M, Stevens JC: Syringomyelia: Electrophysiologic aspects. Muscle Nerve 10:449-458, 1987. Whittle IR: Intraoperative recording of killed end potential. Surg Neurol 30:162, 1988 (letter). Woodbury JW: Potentials in a volume
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
8.
conductor, in Ruch TC, Patton HD, Woodbury JW, Towe AL (eds): Neurophysiology. Philadelphia, WB Saunders, 1965, pp 83-91.
John Oro Columbia, Missouri
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
COMMENT The authors investigated the use of the spinal evoked potential (SpEP) and the somatosensory evoked potential (SEP) in 11 patients with cervical syringomyelia. The SpEP obtained by median nerve stimulation was recorded from a needle electrode placed sequentially into the cervical spinous processes. SEP was obtained by standard techniques. The abnormalities of the N13 component of the SEP were thought to be secondary to gray matter involvement by the syrinx. All patients with abnormal SEPs also had abnormal SpEPs. The SpEP abnormalities were of three types: Type A, representing posterior column involvement; Type B, representing tonsilar herniation; and Type C, representing a mixed pattern. Further study correlating the neurological and radiographic findings with the changes in SpEP will be needed to determine the clinical role of these findings.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 1. SEP and SpEP in a 44-year-old man with posttraumatic syringomyelia (Patient 11). The SEPs reveal a side-to-side difference. The SEPs to left median nerve stimulation are considered within normal limits instead of a slightly attenuated N 13. On the right side, a disappearance of N 13, scalp P 13, and N 20 are seen. Longitudinal distribution of the SpEPs after left stimulation shows the progressive increase in major negative peak latency from caudal to rostral segments. On the right side, the amplitude of the negativity component is decreased markedly at all cervical levels tested (Type A SpEP abnormality). Abbreviations of this and the subsequent figures: C3' and C4', 2 cm posterior to C3 and C4 (International 10-20 system), respectively; NC, noncephalic reference such as contralateral Erb's point; Fz, midline frontal; Cv5, cervical skin over the 5th vertebra; Erb, supraclavicula; O, occipital bone.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 2. The SEPs and SpEPs in a 24-year-old man with a syrinx extending from C2 to T6 with the Chiari I malformation (Patient 6). The SEPs show bilateral absence of N 13, scalp P 13, and N 20. The SpEPs show traceable low-amplitude potentials in positivenegative diphasic configuration on C1-C3 levels and almost no response on occipit-C1 level (Type C SpEP abnormality).
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 3. The SEPs and SpEP in a 37-year-old woman with a syrinx extending from C2 to T11 with a herniated tonsil (Patient 5). The SEPs upon right median nerve stimulation show an absence of N 13 as well as scalp P 13 with a normal N 20. Right median nerve stimulation evoked a SpEP of identical configuration and latencies to those by left stimulation, but of a smaller amplitude on upper cervical levels, compared with those of left stimulation (Type B SpEP abnormality).
Table 2. Normative Data of SEPa
Table 3. Correlation between SEP and SpEP Abnormalitiesa
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Table 1. Summary of the Clinical and Electrophysiological Findings in Our 11 Patientsa
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
components, however, remain controversial (4,7,8,13,14, 21) ; hence, SEPs are of little value for this purpose (2). Maglardery et al. (19) obtained spinal evoked potentials (SpEPs) from the dorsum of the spinal cord with electrodes introduced intrathecally in human volunteers; this method has many risks and is not applicable to routine clinical use. Shimoji et al. (27) have succeeded in recording SpEPs from the dorsal epidural space, on the basis of the technique of epidural anesthesia. Recently, Lüders and his colleages (5,9,17,18) developed a simple technique to record SpEPs by needle electrodes inserted intraoperatively into the intraspinal ligament. This technique allows SpEPs to be recorded by placing electrodes very close to the dura. The recorded potentials have very similar waveform and amplitude to epidurally recorded SpEPs (26). This technique offers significant advantages over the transdermal placement of epidural electrodes (9): first, the use of electrodes inserted into the intraspinal ligaments under direct visualization is both safe and rapid, adding no significant time to the operation; and, second, the method of Shimoji et al. is not applicable to procedures involving the high cervical cord, which is often affected with tonsilar herniation. SpEPs have been used to localize the offending lesion of cervical spondylotic myelopathy (10,26,28) and cervical spinal cord injury (15). Because reports of SpEP findings in patients with syringomyelia have been rare (17,31), the significance of using SpEPs to examine syringomyelia is not fully understood. We report the results of SpEPs studies in our syringomyelic patients and discuss the correlation with those of SEPs studies. PATIENTS AND METHODS This study was performed in 11 patients with cervical syringomyelia (Table 1). There were 4 men and 7 women, aged 24 to 56 (mean, 37) years. Ten patients had Chiari I malformations, and the other one a posttraumatic syringomyelia (lower lumbar and cauda equina injury). Dissociated loss of pain and temperature sensation with preservation of deep sensation was seen in nine upper limbs, and segmental loss of all modalities of cutaneous sensation was observed in seven limbs. Six limbs failed to show sensory disturbance. Hydrocephalus or evidence of intracranial hypertension was not observed in any of the patients, although defective ventricular drainage is thought to be an important factor in syringomyelia. The level and extent of the syrinx were determined by magnetic resonance imaging. Informed consent was obtained after the nature of the procedure had been fully explained. Short-latency SEPs were recorded following the guidelines set by the American Electroencephalographic Society (1). The median nerve was stimulated at the wrist through bipolar electrodes with 0.1 msec square wave pulses. The stimulus strength was adjusted to the level that induced minimal muscle contraction of the thumb at a rate of four per second. Disc electrodes were placed on the bilateral Erb's point (at the supraclavicular regions), the cervical skin over the 5th process (Cv5),
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Neurosurgery 1992-98 February 1992, Volume 30, Number 2 218 Somatosensory and Spinal Evoked Potentials in Patients with Cervical Syringomyelia Experimental and Clinical Study
RESULTS SEPs were abnormal in 7 patients (64%), 5 of whom showed abnormal SEPs only to unilateral stimulation (Table 1). Nine of 22 upper limbs (41%) showed abnormal SEP findings. Abnormalities were obtained from all of the seven limbs that showed proprioceptive sensory deficits. Two out of 9 limbs that showed only superficial sensory disturbance showed normal SEPs. Six limbs without sensory deficits showed normal SEPs. The abnormalities included absent N13/P13 and scalp N20 potential (Fig. 1, left upper traces; Fig. 2, upper traces) in four limbs, an absent N13 with a significantly reduced amplitude of N20 in two limbs, reduced N13 and N20 amplitude in two limbs, and absent N13/P13 with preserved N20 (Fig. 3, left upper traces) in one limb. No limbs showed delayed interpeak latencies of the major components.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
No correlation could be found between the location of the syrinx cavity and SEP abnormalities. There was also no apparent relation between the presence of the Chiari I malformation and SEP abnormality. Severe abnormal SEPs were seen unilaterally in the case with posttraumatic syringomyelia (without tonsilar herniation) (Fig. 1, upper traces). In patients with normal SEPs (Figs. 1 and 3, right upper traces), SpEPs showed positive-negativepositive triphasic waves with the negativity being the predominant component (Figs. 1 and 3, right lower traces). The peak latency of the negativity showed a slightly progressive increase in the caudorostral direction. SpEP abnormalities were found in all patients whose SEPs were abnormal. The SpEP abnormalities could be divided into three groups depending on the cervical level at which the abnormal waves were recorded. Type A abnormality showed reduced amplitude of the cervical potential (Fig. 1, left lower traces), suggesting that the syrinx itself had enlarged to involve the posterior column. Type A abnormality was found in three limbs. Type B abnormality showed attenuated potentials in the upper cervical level (Fig. 3, left lower traces). This may be due to compression of the medulla and upper cervical cord at the foramen magnum by the tonsilar herniation. Two of our patients had limbs that showed the Type B abnormality. Type C is a mixed type of these two, in that the attenuated cervical potentials were again affected in the uppermost cervical segments (Fig. 2, lower traces). Type C was found in four limbs. There was no relationship between the types of SEP and SpEP abnormalities (Table 3). DISCUSSION The most common SEP abnormality that we and others (12,23,25,29-31) have found in patients with syringomyelia is the marked attenuation or absence of N13 with often normal N20 potentials, although there are a small number of reports that show a prolonged central conduction time (2,12). Our results also confirm the previous reports (12,32) in that the SEPs were able to detect subclinical dysfunction of the dorsal column pathway in syringomyelic patients. It appears that the N13 may not reflect the propagation of the afferent volley in the dorsal columns, as was first thought, but may be mediated by the collateral projection to the dorsal horn (13). Therefore, N13 may be susceptible to spinal cord lesions involving the gray matter with sparing of the dorsal columns. This hypothesis could explain the dissociation between cervical and cortical SEP components, with an abnormal N13 and a normal N20, which was observed in our patients. At present, it is generally accepted that the N13 recorded with a cephalic reference, as we used, is a composite potential, reflecting "near-field" activity from the cervical cord and "far-field" activity from the lower brain stem lemniscal pathways (1,4). By using a noncephalic reference, spinal N13 negativity and P13 positivity can be recorded from the posterior and anterior neck surfaces, respectively (4,6,25,30,31). These components are thought to be generated by the
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Fz of the International 10-20 system, and the bilateral hand sensory area on the scalp (C3' or C4'; 2 cm posterior to C3 and C4, respectively). A montage consisting of C3'- or C4'-contralateral Erb's point, C3'or C4'-Fz, Cv5-Fz, and ipsilateral Erb's point-Fz was used for a four-channel system of Cadwell Quantum 84 (Cadwell Laboratories Inc., Kennewick, WA). The potentials were amplified with the filter set at 10 to 3000 Hz, and 500 sweeps were averaged with an analysis time of 30 msec. SpEPs were recorded by using Lüder's method (5,9, 17,18) intraoperatively with a hard stainless steel needle electrode, 1 mm in diameter, with a 0.25-mm bared tip (UKE-0138; Unique Medical Co., Tokyo, Japan). The electrode was inserted by the surgeon between the exposed spinous processes. The tip of the electrode was introduced up to the yellow ligament as close to the epidural space as possible, and this was confirmed with an intraoperative roentrenogram. The reference electrode was placed over Fz. Stimulus and recording conditions were identical to that used for SEPs. Between 50 and 100 responses were summated with an analysis time of 20 to 30 msec. Caudocranial sequential recording of SpEP within the operative field was made. The anesthesia used was of the nitrous oxide-fentanyl-droperidol type. Observations were obtained for a minimum of two runs to ensure that all responses were consistent. The same procedure was employed on the opposite side. The results of SEPs obtained from the patients with syringomyelia were compared with control values obtained from age-matched healthy volunteers (Table 2). The nomenclature of each recognizable component followed standards set by the American Electroencephalographic Society (1). Amplitude was measured between the estimated baseline and the major peak. A record was defined "abnormal" when a peak was focally absent or delayed with a latency value exceeding a mean + 3 standard deviations of control values. The corresponding criterion for the abnormality of each component amplitude was a reduction in size by more than 50%. SpEP were analyzed visually with regard to the waveform and amplitude of responses with reference to our previous reports (20-22).
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Chiari I malformation. Foramen magnum decompression with or without closure of the obex is a procedure currently accepted by many neurosurgeons (24). A frequently discussed issue is syrinx shunting, such as syringosubarachnoidal or syringoperitoneal shunt (11). In patients with Types A and C SpEP abnormalities, the syrinx is thought to be large enough to involve the posterior column. Therefore, syrinx shunting may be an additional recommended procedure. Furthermore, if we are able to record the positive killed end potentials along with the SpEPs, the location at which myelotomy for syrinx decompression may be determined. It is of interest to note that the unilateral abnormal SEP and SpEP findings were found in 5 out of 7 patients with abnormal SEPs and SpEP. According to pathological studies (16), the syrinx is not always situated bilaterally within the cord but has laterality at some spinal level. Tonsilar herniation is not always symmetrical, either. Our findings are consistent with these pathological findings. Received for publication, April 8, 1991; accepted, final form, August 20, 1991. Reprint requests: Takato Morioka, M.D., Department of Neurosurgery, Neurological Institute, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Fukuoka 812, Japan. ACKNOWLEDGMENTS We thank Drs. S. Tobimatsu and H. Tomoda, Department of Clinical Neurophysiology, Kyushu University, for critically reading an earlier draft of this article and A. N. Kalehua for useful comments. REFERENCES: (1-34) 1.
2.
3.
4.
5.
6.
7.
American Electroencephalographic Society: Recommended standards for short-latency somatosensory evoked potentials. J Clin Neurophysiol 1:4-53, 1984. Anderson NE, Frith RW, Synek VM: Somatosensory evoked potentials in syringomyelia. J Neurol Neurosurg Psychiatry 49:1407-1410, 1986. Deecke L, Tator CH: Neurophysiological assessment of afferent and efferent conduction in the injured spinal cord of monkeys. J Neurosurg 19:65-74, 1973. Desmedt JE: Generator sources of SEP in man, in Cracco RQ, Bodis-Wollner I (eds): Evoked Potentials. New York, Liss, 1986, pp 235-245. Dinner DS, Lüders H, Lesser RP, Morris HH: Invasive methods of somatosensory evoked potentials monitoring. J Clin Neurophysiol 3:113-130, 1986. Emerson RG, Pedley TA: Effect of cervical spinal cord lesions on early components of the median nerve somatosensory evoked potential. Neurology 36:20-26, 1986. Friedman WA: Somatosensory evoked potentials in neurosurgery. Clin Neurosurg 34:187-238, 1986.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
central gray matter in the cervical cord (4). The absence of N13/P13 with a noncephalic reference has been demonstrated in syringomyelic patients with a normal scalp far-field P14 of more rostral origin (25,30, 31) . Emerson and Pedley (6) have also shown a loss of the anterior neck P13 component in spite of a normal noncephalic N13. Bipolar recordings of SEPs have two N13 subcomponents, which are generator dipoles directed horizontally (N13a) and axially (N13b) (14). Kaji and Sumner (14) have suggested that syringomyelic patients may have absent N13a with normal N13b. Thus, by using a noncephalic reference and bipolar recording, one may be able to increase the accuracy of the lesion determination in the cervical gray matter, which deserves further investigation in syringomyelic patients (12,25). In our patients with normal SEPs, SpEPs showed positive-negative-positive triphasic waves. The progressive increase in the latency of the dominant negativity from progressively more rostral segments of the cord indicates that an ascending volley in the dorsal column was being recorded, as described earlier (17,20-22). Three types of abnormal SpEP findings were seen in our patients. This provided us with useful information regarding which underlying abnormality was responsible for the dorsal column dysfunction of the patient: the syrinx itself or the tonsilar herniation (or both). When the syrinx is large enough to involve or compress the dorsal column pathways, the cervical ascending volleys can be attenuated (Type A abnormality in our classification). When the herniated tonsil compresses the cord severely at the foramen magnum, upper cervical potentials can be attenuated (Type B abnormality). We also demonstrated a mixture of Types A and B abnormalities, in that the attenuated cervical potentials were again affected in the uppermost cervical segments (Type C abnormality). We could not find a significant confounding effect on the SEPs in syringomyelia from coexisting Chiari malformation, confirming a previous report by Jabbari et al. (12). Anderson et al. (2) reported that all 6 of their patients with abnormal central conduction time in SEPs had an associated Chiari malformation. Our present SpEP findings reveal that tonsilar compression does not develop into a conduction delay but into an attenuation or loss of the potentials. Although we detected an attenuation of the SpEP in our syringomyelic patients, positive "killed end potentials" could not be recorded. Killed end potentials occur when an impulse approaches but does not arrive at a recording electrode because the nerve fibers terminate before reaching the electrode (3,7,34) . Thus, this kind of potential is used to localize the offending lesion precisely in the patients with cervical spondylotic myelopathy (26,28), acute cervical cord injury (15), and intramedullary tumor (20,33). Because dorsal column dysfunction is not the primary lesion in the syringomyelic patient, it is conceivable that complete and abrupt disruption of the dorsal column fibers is rare. Numerous surgical interventions have been advocated for the treatment of syringomyelia with
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Greenberg JA, Kaplan PW, Erwin CW: Somatosensory evoked potentials and the dorsal column myth. J Clin Neurophysiol 4:189-196, 1990. Hahn JF, Lesser R, Klem G, Lueders H: Simple technique for monitoring intraoperative spinal cord function. Neurosurgery 9:692-695, 1981. Hattori S, Saiki K, Kawai S: Diagnosis of the level and severity of cord lesion in cervical spondylotic myelopathy: Spinal evoked potentials. Spine 6:478-485, 1979. Isu T, Iwasaki Y, Akino M, Abe H: Syringosubarachnoid shunt for syringomyelia associated with Chiari malformation (type 1) Acta Neurochir (Wien) 107:152-160, 1991. Jabbari B, Geyer C, Gunderson C, Chu A, Brophy J, McBurney JW, Jonas B: Somatosensory evoked potentials and magnetic resonance imaging in syringomyelia. Electroencephalogr Clin Neurophysiol 77:277-285, 1990. Jones SJ: Somatosensory evoked potentials: The abnormal waveform, in Halliday AM (ed): Evoked Potentials in Clinical Testing. Edinburgh, Churchill Livingstone, 1982, pp 429-470. Kaji R, Sumner AJ: Bipolar recordings of short-latency somatosensory evoked potentials after median nerve stimulation. Neurology 37:410-418, 1987. Katayama Y, Tsubokawa T, Yamamoto T, Hirayama T, Maejima S: Preoperative determination of the level of spinal cord lesions from the killed end potential. Surg Neurol 29:91-94, 1988. Larroche JC: Syringomyelia and syringobulbia, in Adams JH, Corsellis JAN, Duchen LW (eds): Greenfield's Neuropathology. London, Edward Arnold, 1984, ed 4, pp 396-402. Lesser RP, Lueders H, Hahn J, Klem G: Early somatosensory evoked potentials evoked by median nerve stimulation: Intraoperative monitoring. Neurology 31:1519-1523, 1981. Lueders H, Gurd A, Hahn J, Andrish J, Weiker G, Klem G: A new technique for intraoperative monitoring of spinal cord function. Multichannel recording of spinal cord and subcortical evoked potentials. Spine 7:110-115, 1982. Magladery JW, Porter WE, Park AM, Teasdall RD: Electrophysiological studies of nerve and reflex activity in normal man. IV. The twoneuron reflex and identification of certain action potentials from spinal roots and cord. Bull Johns Hopkins Hosp 88:499-519, 1951. Morioka T, Fujii K, Mitani M, Fukui M: Intraoperative localization of a cervicomedullary glioma from the killed end potential: Illustrative case. Neurosurgery 26:1038-1041, 1990. Morioka T, Shima F, Kato M, Fujui M: Direct recording of somatosensory evoked potentials
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32. 33. 34.
in the vicinity of dorsal column nuclei in man: Their generator mechanisms and contribution to the scalp far-field potentials. Electroencephalogr Clin Neurophysiol 80:215220, 1991. Morioka T, Tobimatsu S, Fujii K, Fukui M, Kato M, Matsubara T: Origin and distribution of brainstem somatosensory evoked potentials in humans. Electroencephalogr Clin Neurophysiol 80:221-227, 1991. Nakanishi T, Tamaki M, Ozaki Y, Arasaki K: Origins of short latency somatosensory evoked poteials to median nerve stimulation. Electroencephalogr Clin Neurophysiol 56:7485, 1983. Pillay P, Awad IA, Little JR, Hahn J: Symptomatic Chiari malformation in adults: A new classification based on magnetic resonance imaging with clinical and prognostic significance. Neurosurgery 28:639645, 1991. Restuccia D, Mauguière F: The contribution of median nerve SEPs in the functional assessment of the cervical spinal cord in syringomyelia. A study of 24 patients. Brain 114:361-379, 1991. Satomi K, Okuma T, Kenmotsu K, Nakamura Y, Hirabayashi K: Level diagnosis of cervical myelopathy using evoked spinal cord potentials. Spine 13:1217-1224, 1988. Shimoji K, Higashi H, Kano T: Epidural recording of spinal electrogram in man. Electroencephalogr Clin Neurophysiol 30:236239, 1971. Shinomiya K, Furuya K, Sato R, Okamoto A, Kurosa Y, Fuchioka M: Electrophysiological diagnosis of cervical OPLL myelopathy using evoked spinal cord potentials. Spine 13:12251233, 1988. Stöhr M, Buettner UW, Riffel B, Koletzki E: Spinal somatosensory evoked potentials in cervical cord lesions. Electroencephalogr Clin Neurophysiol 54:257-265, 1982. Urasaki E, Wada S, Kadoya C, Matsuzaki H, Yokota A, Matsuoka S: Absence of spinal N13-P13 and normal scalp far-field P14 in a patient with syringomyelia. Electroencephalogr Clin Neurophysiol 71:400404, 1988. Urasaki E, Wada S-I, Kadoya C, Tokimura T, Yokota A, Matsuoka S, Fukumura A, Hamada S: Skin and epidural recording of spinal somatosensory evoked potentials following median nerve stimulation: Correlation between the absence of spinal N13 and impaired pain sensation. J Neurol 237:410415, 1990. Veilleux M, Stevens JC: Syringomyelia: Electrophysiologic aspects. Muscle Nerve 10:449-458, 1987. Whittle IR: Intraoperative recording of killed end potential. Surg Neurol 30:162, 1988 (letter). Woodbury JW: Potentials in a volume
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
8.
conductor, in Ruch TC, Patton HD, Woodbury JW, Towe AL (eds): Neurophysiology. Philadelphia, WB Saunders, 1965, pp 83-91.
John Oro Columbia, Missouri
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
COMMENT The authors investigated the use of the spinal evoked potential (SpEP) and the somatosensory evoked potential (SEP) in 11 patients with cervical syringomyelia. The SpEP obtained by median nerve stimulation was recorded from a needle electrode placed sequentially into the cervical spinous processes. SEP was obtained by standard techniques. The abnormalities of the N13 component of the SEP were thought to be secondary to gray matter involvement by the syrinx. All patients with abnormal SEPs also had abnormal SpEPs. The SpEP abnormalities were of three types: Type A, representing posterior column involvement; Type B, representing tonsilar herniation; and Type C, representing a mixed pattern. Further study correlating the neurological and radiographic findings with the changes in SpEP will be needed to determine the clinical role of these findings.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 1. SEP and SpEP in a 44-year-old man with posttraumatic syringomyelia (Patient 11). The SEPs reveal a side-to-side difference. The SEPs to left median nerve stimulation are considered within normal limits instead of a slightly attenuated N 13. On the right side, a disappearance of N 13, scalp P 13, and N 20 are seen. Longitudinal distribution of the SpEPs after left stimulation shows the progressive increase in major negative peak latency from caudal to rostral segments. On the right side, the amplitude of the negativity component is decreased markedly at all cervical levels tested (Type A SpEP abnormality). Abbreviations of this and the subsequent figures: C3' and C4', 2 cm posterior to C3 and C4 (International 10-20 system), respectively; NC, noncephalic reference such as contralateral Erb's point; Fz, midline frontal; Cv5, cervical skin over the 5th vertebra; Erb, supraclavicula; O, occipital bone.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 2. The SEPs and SpEPs in a 24-year-old man with a syrinx extending from C2 to T6 with the Chiari I malformation (Patient 6). The SEPs show bilateral absence of N 13, scalp P 13, and N 20. The SpEPs show traceable low-amplitude potentials in positivenegative diphasic configuration on C1-C3 levels and almost no response on occipit-C1 level (Type C SpEP abnormality).
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 3. The SEPs and SpEP in a 37-year-old woman with a syrinx extending from C2 to T11 with a herniated tonsil (Patient 5). The SEPs upon right median nerve stimulation show an absence of N 13 as well as scalp P 13 with a normal N 20. Right median nerve stimulation evoked a SpEP of identical configuration and latencies to those by left stimulation, but of a smaller amplitude on upper cervical levels, compared with those of left stimulation (Type B SpEP abnormality).
Table 2. Normative Data of SEPa
Table 3. Correlation between SEP and SpEP Abnormalitiesa
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/30/2/218/2752974 by UB Frankfurt/Main user on 05 March 2018
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Table 1. Summary of the Clinical and Electrophysiological Findings in Our 11 Patientsa
