- Email: [email protected]
Spinally elicited peripheral nerve responses are sensory rather than motor
Clinical Neurophysiology 111 (2000) 736±742 www.elsevier.com/locate/clinph
Spinally elicited peripheral nerve responses are sensory rather than motor J.R. Toleikis*, J.P. Skelly, A.O. Carlvin, J.K. Burkus Hughston Sports Medicine Hospital, Columbus, GA, USA Accepted 29 November 1999
Abstract Objectives: Spinally elicited peripheral nerve responses, commonly called neurogenic motor evoked potentials (NMEPs), are widely used to monitor spinal cord motor function during surgery. However, numerous evidence suggests that these responses are primarily sensory rather than motor. The collision technique was utilized to address this issue. Methods: Collision studies were performed in 7 patients during surgery. An ascending volley of sensory (AS) and motor activity (AM) was elicited by posterior tibial nerve stimulation at the popliteal fossa. After a short time delay, high cervical spinal stimulation produced a descending volley of sensory (DS) and motor (DM) activity. The AM volley ascended only to the anterior horn cells whereas the AS and DS volleys collided in the spinal cord. The inter-stimulus delays were varied so as to affect the degree of spinal cord collision. The DS and DM activity which remained after collision was recorded from the posterior tibial nerves at the ankle. Results: Inter-stimulus delays of 18 ms or less resulted in no apparent peripheral descending volleys. These ®ndings were consistent for all the patients studied. Conclusions: Spinally elicited peripheral nerve responses are primarily sensory rather than motor and are mediated by the same neural pathways as SEPs. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Motor evoked potentials; Intraoperative monitoring; Spinal cord stimulation
1. Introduction Somatosensory evoked potentials (SEPs) have been used for many years to assess spinal cord function during surgery (Nash et al., 1977; Engler et al., 1978; Nuwer and Dawson, 1984) and their use is now commonplace. These responses are known to be mediated by the dorsal column pathways which are located in the posterior portion of the spinal cord and receive their blood supply primarily from the posterior spinal arteries. However, because the motor pathways are located in the anterior and lateral portions of the spinal cord which receive their blood supply primarily from the anterior spinal artery, it has been a concern that SEPs may be insensitive to mechanical or ischemic insults which result in a profound loss of motor function. Therefore, optimal monitoring of the physiological integrity of the spinal cord requires the testing of both sensory and motor pathways. Theoretically, motor pathway function can be monitored using either spinal cord or transcranial stimulation to elicit descending volleys of motor activity which can then be * Corresponding author. RSOC, Inc., 6298 Veterans Parkway, Suite 5A, Columbus, GA 31908-8068, USA. Tel.: 11-706-576-2463; fax: 11-706576-2169. E-mail address: [email protected] (J.R. Toleikis)
recorded either centrally from the spinal cord or peripherally from nerve or muscle. However, transcranial stimulation has not received widespread acceptance as a monitoring tool largely because of the need for special equipment to elicit these responses, the sensitivity of the responses to commonly used anesthetics, and concerns about patient safety. The use of spinal stimulation to elicit and record muscle activity as a monitoring tool is generally associated with undesirable patient movement because of the requirement that the patient's muscles not be paralyzed. On the other hand, the use of spinal stimulation to elicit and record responses from peripheral nerves has been reported not to have the limitations of the previous monitoring techniques (Owen, 1997). This monitoring technique was introduced by Owen et al. (1988) who named these peripheral nerve responses neurogenic motor evoked potentials (NMEPs). Although they are called motor responses, it is generally agreed that because of the placement of the stimulation electrodes, they are composed of both orthodromic motor and antidromic sensory components (Machida et al., 1985; Yokoyama et al., 1991; Su et al., 1992; Deletis, 1993; Kai et al., 1993; Owen, 1993, 1997; Haghighi et al., 1994; Machida, 1994; Burke and Hicks, 1998; Padberg et al., 1998; Rose, 1998; Toleikis et al., 1999). NMEPs consist
1388-2457/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88-2457(99)0031 7-X
CLINPH 98130
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
of a prominent early biphasic component followed by polyphasic activity (Fig. 1). It has been reported that the large early biphasic component is composed of motor activity and the following polyphasic activity is sensory (Kai et al., 1993; Owen, 1993, 1997). However, this description of the functional composition of NMEPs is controversial. Numerous theoretical (Deletis, 1993; Rose, 1996, 1998), animal (Yokoyama et al., 1991; Su et al., 1992; Haghighi et al., 1994), and human (Machida et al., 1985; Machida, 1994; Bernard et al., 1996; Leppanen et al., 1999) studies have provided evidence which supports the opinion that NMEPs are primarily composed of sensory rather than motor activity. The purpose of this study was to utilize the collision technique to de®ne the composition of these responses so as to substantiate or refute the basis for their utilization. 2. Materials and methods Data was acquired from 7 patients (3 males and 4 females) ranging in age from 10 to 15 years. All were diagnosed as having idiopathic scoliosis and underwent surgery for scoliosis repair at the Hughston Sports Medicine Hospital. Their preoperative neurologic examinations were normal. As part of our standard monitoring protocol, the neurological function of all of the patients was monitored throughout surgery using (1) somatosensory evoked potentials (SEPs) elicited by bilateral posterior tibial and ulnar nerve stimulation and (2) NMEPs elicited by percutaneous spinal stimulation. Collision studies were performed prior to incision and during surgery. These were performed using a combination of the two standard monitoring techniques and therefore informed consent was not obtained. All monitoring and collision study data were acquired on a Nicolet Viking IV. Standard stimulation and recording techniques were used to acquire the SEP responses. The NMEPs were an averaged response consisting of 100 separate responses elicited by means of a pair of 70 mm needle electrodes placed super-
737
®cial to the spinal lamina of the third and fourth cervical vertebrae for the purpose of electrically stimulating the underlying spinal cord. The stimulating electrodes were insulated except at their tips. Stimulation consisted of 4.7 Hz, 0.3 ms duration constant voltage (150±400 V) pulses. The NMEPs were recorded with subdermal needle electrodes from posterior tibial nerves at the ankle using a ®lter bandpass of 30±2000 Hz and the same shoulder ground used to record the SEPs. The ability to elicit well-de®ned, repeatable NMEPs determined the spinal stimulation intensity that was normally utilized for each patient. Occasionally, maximal spinal stimulation intensities of 400 V were used. Anesthetic management consisted of nitrous oxide (50%) and either des¯urane (3±5%) or iso¯urane (0.5±1%), fentanyl, and muscle relaxants. The concept and methodology for performing collision studies has previously been described (Kai et al., 1993; Owen, 1993). Stimulation of a peripheral nerve produces a volley of ascending sensory (AS) and ascending motor (AM) activity. The AM activity ascends to the anterior horn cells where it stops, whereas the AS activity continues to ascend in the dorsal column pathways of the spine (Fig. 2a). After a short delay, spinal stimulation follows and produces a descending volley believed to consist of both descending antidromic sensory (DS) and descending orthodromic motor (DM) activity. The time delay between the peripheral and spinal stimuli can be varied so that the resulting AS and DS volleys either (1) do not collide at all or (2) they collide partially or totally in the spinal cord or periphery. The portions of the AS and DS volleys which collide in the spinal cord will cancel and the remaining DS and DM activity can be peripherally recorded from the posterior tibial nerve ipsilateral to the side of peripheral stimulation (Fig. 2b). Spinal stimulation consisted of 4.7 Hz, 0.3 ms duration constant voltage (150±400 V) pulses. Because of large artifacts which result from the stimulation currents, the peripheral stimulation and recording sites could not be the same. Therefore, peripheral stimulation of the posterior tibial nerve stimulation occurred at the popliteal fossa using subdermal needle electrodes (4.7 Hz, 0.3 ms duration, 40 mA pulses) and the peripheral responses, like the NMEPs, were recorded with subdermal needle electrodes from the posterior tibial nerves at the ankle using a ®lter bandpass of 30±2000 Hz. 3. Results
Fig. 1. Typical peripherally recorded neurogenic motor evoked potential (NMEP) response elicited by spinal stimulation. The response consists of a prominent early biphasic component followed by small amplitude polyphasic components.
The amount of collision between the AS and DS volleys is dependent upon the delay time between the peripheral and spinal stimuli.. Total collision will occur when this delay time is less than the conduction time between the peripheral and spinal stimulation site for the entire AS volley. A study was conducted in one patient to determine the shortest conduction time of such a volley. The posterior tibial nerve was stimulated at the ankle and AS volleys were
738
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
electrodes) was found to be about 17 ms (Fig. 3a). Therefore, time delays of 17 ms or less should result in total collisions between the AS and DS volleys. This would leave just the DM volleys to be recorded at the ankle. Longer time delays should result in collisions between only the slower components of the AS and DS volleys. There would be no collisions between the faster AS and DS components of the volleys because the faster AS components would have already ascended beyond the spinal stimulation site. Therefore, the DM activity and only the faster conducting DS activity would be recorded at the ankle. In Fig. 3b, both the top and bottom traces illustrate the responses that result when no collisions take place (except for possibly the late components in the top trace). The bottom trace resulted from spinal stimulation and peripheral stimuli of such low intensity that no ascending activity was produced. The top trace resulted from spinal and peripheral stimulation with an inter-stimulus delay so large that the entire AS volley may have passed the spinal stimulation site before spinal stimulation occurred. Progressively shorter inter-stimulus delays resulted in collisions between more of the slower components of the AS and DS volleys and the amount of resulting descending activity recorded at the ankle became smaller. When the inter-stimulus interval was shortened to 17 ms, all the components of the AS and DS volleys collided leaving just the DM components to be recorded at the ankle. However, the result was a total absence of a peripherally recorded response. An example of comparable ®ndings from another patient is illustrated in Fig. 4. The lack of any remaining activity suggests that the descending volleys resulting from spinal stimulation contain only sensory activity and that these volleys and the AS volleys resulting from peripheral stimulation are mediated by the same sensory pathways. Such ®ndings were consistent for all 7 patients in the study.
Fig. 2. Collision technique. (a) Stimulation of the posterior tibial nerve (PTn) at the popliteal fossa occurs at t1. This results in ascending orthodromic sensory (AS) and ascending antidromic motor (AM) activity. The AM activity is blocked at the anterior horn cell synapse. The AS activity enters the spinal cord and is conveyed by the dorsal column pathways. For collision to occur in the spinal cord, spinal stimulation occurs at t2, approximately 15 ms after t1. It results in descending antidromic sensory (DM) and descending orthodromic motor (DM) activity. The AS and DS activity collide and proceed no further. The DM activity proceeds down the spinal cord. (b) The post-collision DM activity exits the spinal cord and proceeds down the PTn to near the medial malleolus where it is recorded at t3, approximately 25 ms after the onset of the spinal stimulus.
recorded with subdermal needle electrodes from the popliteal fossa, from the spinal stimulation electrodes used to produce the descending volleys, and from subdermal needle electrodes placed in the high cervical region and on the scalp. The conduction time for the fastest conducting ®bers (i.e. from the onset of the popliteal fossa responses to the onset of the responses recorded from the spinal stimulation
4. Discussion The descending volleys of activity, known as NMEPs, that result from percutaneous spinal stimulation appear to be primarily composed of sensory activity components mediated by the same neural pathways as SEPs. These ®ndings are consistent with those of Leppanen et al. (1999) and contradict those of the Owen group (Kai et al., 1993; Owen, 1993) who also utilized the collision technique to assess the composition of NMEPs. Leppanen et al. (1999) reported that collisions resulted in the early biphasic components of the NMEP being mostly or totally blocked, whereas the Owen group (Kai et al., 1993; Owen, 1993) reported that collisions resulted in a loss of only the small, polyphasic late components of the NMEPs but the large, early biphasic components were unaffected. As a result, the ®ndings of the Leppanen group suggest that the early large biphasic component of NMEPs consists of primarily sensory activity whereas the Owen group has reported that this component
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
739
consists of motor activity. It is dif®cult to explain this large discrepancy since the Leppanen and Owen groups used similar methodologies. A ¯aw in the collision technique itself does not explain
Fig. 4. Collision study in patient #2 (J.R. ± 14-year-old male). Peripheral responses acquired using spinal stimuli (4.7 Hz, 0.3 ms, 320 V pulses) with no peripheral stimuli (bottom trace) or supramaximal peripheral stimuli (4.7 Hz, 0.3 ms, 40 mA constant current pulses, top 3 traces) at the popliteal fossa. There is no difference between the early biphasic portions of the response with no collisions and the response with a 25 ms inter-stimulus delay. As the delay between the peripheral and spinal stimuli decreased, the post-collision responses also decreased. An 18 ms inter-stimulus delay resulted in total obliteration of the descending NMEP response (third trace).
this discrepancy but is one possible explanation for the apparent lack of any signi®cant descending motor (DM) activity in our results and those of Leppanen et al. (1999). The ascending volleys could cause a blockage of DM activity by affecting motor neuron or inter-neuron excitability through recurrent inhibition or other mechanisms (Eccles, 1964). However, based on our data, this explanation seems unlikely. In Fig. 3b, there is no difference between the response in the bottom trace that results from spinal stimulation and minimal peripheral stimulation (and therefore no
Fig. 3. Collision study in patient #1 (K.B. ± 15-year-old female). (a) SEPs acquired using constant current stimulation (4.7 Hz, 0.3 ms, 40 mA pulses) of the posterior tibial nerve at the medial malleolus. Note that the fastest ascending sensory activity takes approximately 17 ms to travel from the popliteal fossa to the spinal stimulation electrodes. (b) Peripheral responses acquired using spinal stimuli (4.7 Hz, 0.3 ms, 400 V pulses) and different intensities of peripheral stimuli (4.7 Hz, 0.3 ms, constant current pulses) (bottom trace, subthreshold 1 mA; top 3 traces, 40 mA). As the delay between the peripheral and spinal stimuli decreased, the responses that remained after collisions also decreased in size. With a 17 ms delay (the observed SEP conduction time from the popliteal fossa to the stimulation electrodes ± (a)), the descending peripheral response was totally obliterated (second trace from the bottom). The large stimulus artifact is due to the use of a shoulder recording ground.
740
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
ascending inhibitory in¯uences) and the response in the top trace that results from spinal stimulation and supramaximal peripheral stimulation. If ascending inhibitory in¯uences are present and the responses contain motor activity, there should be some difference in these traces but none are evident except in possibly the late polyphasic components. It is also possible that these in¯uences could be present but have a shorter duration of action than the inter-stimulus delay period. However, if this were the case and the early biphasic component contained motor activity, then as the inter-stimulus delay became shorter, these in¯uences should affect the initial portions of the biphasic component. However, no changes were observed in this component except for those explained by the collision process itself. Therefore, regardless of whether ascending inhibitory in¯uences are present or not, the large biphasic component appears to contain only sensory activity and the late polyphasic components may consist of sensory and/or motor activity. Another possible explanation for the lack of motor activity in our responses could be the anesthetic management used during our studies. Our data were acquired using concentrations of volatile anesthetic agents common to spinal surgeries. Recent studies have reported that volatile anesthetics signi®cantly depress spinal cord motor neuron excitability (King and Rampil, 1994; Zhou et al., 1997). Therefore, it is likely that the anesthetic management used in our studies could have some effect on our responses if motor activity was present. However, Bernard et al. (1996) have shown that increasing concentrations of iso¯urane have no effect on the early NMEP components, thus supporting our ®ndings that these components contain sensory rather than motor activity. We also considered the possible effects that stimulation intensity might have on the composition of the descending volley. Because of the placement of the spinal stimulation electrodes, descending sensory and descending motor activity would probably be mediated by the dorsal columns and the lateral corticospinal tracts, respectively. Because the lateral corticospinal tracts would probably be located more distal to the stimulation electrodes than the dorsal columns, higher stimulation intensities might be required to excite these tracts. However, maximal spinal stimulation intensities (400 V) did not affect our ®ndings. Therefore, within the practical stimulation limits of the equipment we were using, the early components of the percutaneously elicited peripheral responses appear to contain only sensory activity. Animal studies have also indicated that the responses elicited by low intensity spinal cord stimulation contain primarily sensory activity (Su et al., 1992; Haghighi et al., 1994). Epidural stimulation may be more effective than percutaneous translaminar stimulation in eliciting motor activity because of current spread to the lateral corticospinal tracts. However, in other animal studies, when higher stimulation intensities were used and motor activity was present, the motor activity had slower conduction velo-
cities than the sensory activity (Yokoyama et al., 1991), thus supporting the idea that if motor activity is present in the peripheral nerve responses, it probably occurs in the late polyphasic components or even later or may go unnoticed because it is small in amplitude, temporally dispersed or delayed (Machida et al., 1985) relative to the early large biphasic component. In support of this idea, it has been reported that myogenic activity from leg muscles can be present even in the absence of any peripheral nerve responses, thus demonstrating that the peripheral nerve responses resulting from percutaneous spinal stimulation do not contain a major motor component (Leppanen et al., 1999). Therefore, monitoring the myogenic activity from leg muscles in response to spinal stimulation appears to be a better way to assess spinal motor function than recording peripheral nerve responses. However, it is possible that such myogenic activity can also result from re¯ex activation of alpha-motoneurons by antidromically activated dorsal column pathways (Deletis, 1993; Rose, 1996, 1998). Although the peripheral nerve responses elicited by spinal stimulation appear to be mediated by the same pathways as SEPs, the acquisition of both monitoring modalities can serve as useful adjuncts to one another. Although SEPs can normally be acquired very easily, the peripheral nerve responses can typically be acquired more quickly than SEPs and may be more robust as well. In addition, there are instances when useful SEPs cannot be acquired either due to anesthetic management or pathology and the acquisition of these peripheral nerve responses may be the only means available for assessing spinal cord function. In addition, these peripheral nerve responses have been reported to be a more rapid or more sensitive indicator of changes in spinal cord function than traditional cortical SEPs (Mustain and Kendig, 1991; Owen et al., 1991; Owen, 1993, 1997; Owen and Tamaki, 1997; Padberg et al., 1998; Siddiqui et al., 1998). This ®nding is questionable considering that both monitoring modalities appear to be mediated by the same neural pathways, however, relative sensitivity may depend on the recording sites that are used to acquire responses rather than the monitoring modality itself. The SEPs that are recorded over the cervical spine are probably preferable to cortical SEPs because the amplitude of the dorsal column-mediated ascending traveling wave which produces these subcortical responses is directly related to the size of the incoming afferent volley (Burke and Hicks, 1998), whereas it has been reported that a greater than 80% reduction in the amplitude of the dorsal column-mediated incoming volley is required before the amplitude of the cortical SEPs is reduced by 50% (Eisen et al., 1982; Gandevia and Burke, 1984). Similarly, the peripheral nerve responses which result from spinal stimulation should be directly related to the descending afferent volley and may be more sensitive to changes in spinal function than cortical SEPs. Nevertheless, the use of only SEPs and/or the spinally elicited peripheral nerve responses for monitoring purposes may be insensitive to spinal cord insults which produce only
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
changes in motor pathway function. Therefore, despite the reports of others (Padberg et al., 1998), it may be unwise to abandon the use of the wake-up test when neither of these monitoring modalities demonstrates signi®cant changes. Transcranial rather than spinal stimulation techniques are now available for directly assessing motor pathway function (Edmonds et al., 1989; Zentner, 1989; Zentner and Ebner, 1989; Haghighi et al., 1990a,b; Calancie et al., 1991, 1998; Burke et al., 1992; Schmid et al., 1992; Zentner et al., 1992, 1997; Deletis, 1993; Jones et al., 1996; Linden et al., 1997; Morota et al., 1997; Burke and Hicks, 1998). These techniques have not yet received FDA approval and therefore their use for monitoring purposes currently requires institutional review board approval and informed consent. However, the special equipment needed to utilize these techniques is relatively inexpensive, responses can be acquired using common anesthetic management (Jones et al., 1996; Calancie et al., 1998), and there have been no reported safety problems associated with their use. Therefore, there appears to be very little reason for not considering their use when an accurate assessment of spinal cord motor function is required.
References Bernard JM, Pereon Y, Fayet G, Guiheneuc P. Effects of iso¯urane and des¯urane on neurogenic motor- and somatosensory-evoked potential monitoring for scoliosis surgery. Anesthesiology 1996;85:1013±1019. Burke D, Hicks RG. Surgical monitoring of motor pathways. J Clin Neurophysiol 1998;15:194±205. Burke D, Hicks R, Stephen J, Woodworth I, Crawford M. Assessment of corticospinal and somatosensory conduction simultaneously during scoliosis surgery. Electroenceph clin Neurophysiol 1992;85:388±396. Calancie B, Klose KJ, Baier S, Green BA. Iso¯urane induced attenuation of motor evoked potentials caused by electrical motor cortex stimulation during surgery. J Neurosurg 1991;74:897±904. Calancie B, Harris W, Broton JG, Alexeeva N, Green BA. ``Thresholdlevel'' multipulse transcranial electrical stimulation of motor cortex for intraoperative monitoring of spinal motor tracts: description of method and comparison to somatosensory evoked potential monitoring. J Neurosurg 1998;88:457±470. Deletis V. Intraoperative monitoring of the functional integrity of the motor pathways. In: Devinsky O, Beric A, Dogali M, editors. Electrical and magnetic stimulation of the brain and spinal cord, New York: Raven Press, 1993. pp. 201±214. Eccles JC. The physiology of synapses, New York: Academic Press, 1964. Edmonds HL, Paloheimo MPJ, Backman MH, Johnson JR, Holt RT, Shields CB. Transcranial magnetic motor evoked potentials (tcMMEP) for functional monitoring of motor pathways during scoliosis surgery. Spine 1989;14:683±686. Eisen A, Purves S, Hoirch M. Central nervous system ampli®cation: its potential in the diagnosis of early multiple sclerosis. Neurology 1982;32:359±364. Engler GL, Spielholtz NI, Bernard WN, Danziger F, Merkin H, Wolff T. Somatosensory evoked potentials during Harrington instrumentation for scoliosis. J Bone Joint Surg 1978;60A:528±532. Gandevia SC, Burke D. Saturation in human somatosensory pathways. Exp Brain Res 1984;54:582±585. Haghighi SS, Madsen R, Green KD, Oro JJ, Kracke GR. Suppression of motor evoked potentials by inhalational anesthetics. J Neurosurg Anesthesiol 1990a;2:73±78. Haghighi SS, Green KD, Oro JJ, Drake RK, Kracke GR. Depressive effect
741
of iso¯urane anesthesia on motor evoked potentials. Neurosurgery 1990b;26:993±997. Haghighi SS, York DH, Gaines RW, Oro JJ. Monitoring of motor tracts with spinal cord stimulation. Spine 1994;19:1518±1524. Jones SJ, Harrison R, Koh KF, Mendoza N, Crockard HA. Motor evoked potential monitoring during spinal surgery: responses of distal limb muscles to transcranial cortical stimulation with pulse trains. Electroenceph clin Neurophysiol 1996;100:375±383. Kai Y, Owen JH, Lenke LG, Bridwell KH, Oakley DM, Sugioka Y. Use of sciatic neurogenic motor evoked potentials versus spinal potentials to predict early-onset neurologic de®cits when intervention is still possible during overdistraction. Spine 1993;18:1134±1139. King BS, Rampil IJ. Anesthetic depression of spinal motor neurons may contribute to lack of movement in response to noxious stimuli. Anesthesiology 1994;81:1484±1492. Leppanen R, Madigan R, Sears C, Maguire J, Wallace S, Captain J. Intraoperative collision studies demonstrate descending spinal cord stimulated evoked potentials and ascending somatosensory evoked potentials are mediated through common pathways. J Clin Neurophysiol 1999;16:170. Linden RD, Johnson JR, Shields CB, Edmonds HL, Glassman SD. Intraoperative spinal cord monitoring with motor evoked potentials elicited by transcranial magnetic stimulation. In: Bridwell KH, Dewald RL, editors. The textbook of spinal surgery, 2nd ed. Philadelphia, PA: Lippincott-Raven, 1997. pp. 77±83. Machida M. Point of view. Spine 1994;19:1167±1168. Machida M, Weinstein SL, Yamada T, Kimura J. Spinal cord monitoring. Electrophysiological measures of sensory and motor function during spinal surgery. Spine 1985;10:407±413. Morota N, Deletis V, Constantini S, Ko¯er M, Cohen H, Epstein FJ. The role of motor evoked potentials during surgery for intramedullary spinal cord tumors. Neurosurgery 1997;41:1327±1336. Mustain WD, Kendig RJ. Dissociation of neurogenic motor and somatosensory evoked potentials: a case report. Spine 1991;16:851±853. Nash CL, Lorig RA, Schatzinger LA, Brown RH. Spinal cord monitoring during operative treatment of the spine. Clin Orthop 1977;126:100±105. Nuwer MR, Dawson E. Intraoperative evoked potential monitoring of the spinal cord: enhanced stability of cortical recordings. Electroenceph clin Neurophysiol 1984;59:318±327. Owen JH. Intraoperative stimulation of the spinal cord for prevention of spinal cord injury. In: Devinsky O, Beric A, Dogali M, editors. Electrical and magnetic stimulation of the brain and spinal cord, New York: Raven Press, 1993. pp. 271±288. Owen JH. Monitoring during surgery for spinal deformities. In: Bridwell KH, Dewald RL, editors. The textbook of spinal surgery, 2nd ed. Philadelphia, PA: Lippincott-Raven, 1997. pp. 39±60. Owen JH, Tamaki T. Applications of neurophysiological measures during surgery of the spine. In: Frymoyer JW, editor. The adult spine: principles and practice, 2nd ed. Philadelphia, PA: Lippincott-Raven, 1997. pp. 673±702. Owen JH, Laschinger J, Bridwell K, et al. Sensitivity and speci®city of somatosensory and neurogenic-motor evoked potentials in animals and humans. Spine 1988;13:1111±1118. Owen JH, Bridwell KH, Grubb R, et al. The clinical application of neurogenic motor evoked potentials to monitor spinal cord function during surgery. Spine 1991;16:S385±S390. Padberg AM, Wilson-Holden TJ, Lenke LG, Bridwell KH. Somatosensoryand motor-evoked potential monitoring without a wake-up test during idiopathic scoliosis surgery. Spine 1998;23:1392±1400. Rose RD. Sensory component of cervically evoked motor potentials. Med Hypoth 1996;46:577±579. Rose RD. Removing the antidromically driven sensory component from cervically evoked motor potentials. Med Hypoth 1998;50:147±154. Schmid UD, Boll J, Liechti S, Schmid J, Hess CW. In¯uence of some anesthetic agents on muscle responses to transcranial magnetic stimulation: a pilot study in humans. Neurosurgery 1992;30:85±92. Siddiqui A, Blackburn TW, Page RB, Grubb MR, Russell GB. Spinal cord
742
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
stimulation reliably elicits motor evoked potentials. Presented at the Ninth Annual Meeting of the American Society of Neurophysiological Monitoring. Philadelphia, PA, 1998. Su CF, Haghighi SS, Oro JJ, Gaines RW. ``Back®ring'' in spinal cord monitoring. High thoracic spinal cord stimulation evokes sciatic response by antidromic sensory pathway conduction, not motor tract conduction. Spine 1992;17:504±508. Toleikis JR, Skelly JP, Carlvin AO, Burkus JK. Spinally elicited neurogenic ``motor'' evoked potentials (NMEPs) assess sensory rather than motor function. J Clin Neurophysiol 1999;16:169. Yokoyama M, Shinomiya K, Sato H, Komori H, Furuya K. Monitoring motor function of the spinal cord: the descending segmental evoked spinal cord potential. In: Shomoji K, Kurokawa T, Tamaki T, Willis WD, editors. Spinal cord monitoring and electrodiagnosis, Berlin: Springer, 1991. pp. 27±35.
Zentner J. Noninvasive motor evoked potential monitoring during neurosurgical operations on the spinal cord. Neurosurgery 1989;24:709± 712. Zentner J, Ebner A. Nitrous oxide suppresses the electromyographic response evoked by electrical stimulation of the motor cortex. Neurosurgery 1989;24:60±62. Zentner J, Albrecht T, Heuser D. In¯uence of halothane, en¯urane, and iso¯urane on motor evoked potentials. Neurosurgery 1992;31:298± 305. Zentner J, Thees C, Pechstein U, Scheu¯er K, Wurker J, Nadstawek J. In¯uence of nitrous oxide on motor-evoked potentials. Spine 1997;22:1002±1006. Zhou HH, Mehta M, Leis AA. Spinal cord motoneuron excitability during iso¯urane and nitrous oxide anesthesia. Anesthesiology 1997;86:302± 307.
Spinally elicited peripheral nerve responses are sensory rather than motor J.R. Toleikis*, J.P. Skelly, A.O. Carlvin, J.K. Burkus Hughston Sports Medicine Hospital, Columbus, GA, USA Accepted 29 November 1999
Abstract Objectives: Spinally elicited peripheral nerve responses, commonly called neurogenic motor evoked potentials (NMEPs), are widely used to monitor spinal cord motor function during surgery. However, numerous evidence suggests that these responses are primarily sensory rather than motor. The collision technique was utilized to address this issue. Methods: Collision studies were performed in 7 patients during surgery. An ascending volley of sensory (AS) and motor activity (AM) was elicited by posterior tibial nerve stimulation at the popliteal fossa. After a short time delay, high cervical spinal stimulation produced a descending volley of sensory (DS) and motor (DM) activity. The AM volley ascended only to the anterior horn cells whereas the AS and DS volleys collided in the spinal cord. The inter-stimulus delays were varied so as to affect the degree of spinal cord collision. The DS and DM activity which remained after collision was recorded from the posterior tibial nerves at the ankle. Results: Inter-stimulus delays of 18 ms or less resulted in no apparent peripheral descending volleys. These ®ndings were consistent for all the patients studied. Conclusions: Spinally elicited peripheral nerve responses are primarily sensory rather than motor and are mediated by the same neural pathways as SEPs. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Motor evoked potentials; Intraoperative monitoring; Spinal cord stimulation
1. Introduction Somatosensory evoked potentials (SEPs) have been used for many years to assess spinal cord function during surgery (Nash et al., 1977; Engler et al., 1978; Nuwer and Dawson, 1984) and their use is now commonplace. These responses are known to be mediated by the dorsal column pathways which are located in the posterior portion of the spinal cord and receive their blood supply primarily from the posterior spinal arteries. However, because the motor pathways are located in the anterior and lateral portions of the spinal cord which receive their blood supply primarily from the anterior spinal artery, it has been a concern that SEPs may be insensitive to mechanical or ischemic insults which result in a profound loss of motor function. Therefore, optimal monitoring of the physiological integrity of the spinal cord requires the testing of both sensory and motor pathways. Theoretically, motor pathway function can be monitored using either spinal cord or transcranial stimulation to elicit descending volleys of motor activity which can then be * Corresponding author. RSOC, Inc., 6298 Veterans Parkway, Suite 5A, Columbus, GA 31908-8068, USA. Tel.: 11-706-576-2463; fax: 11-706576-2169. E-mail address: [email protected] (J.R. Toleikis)
recorded either centrally from the spinal cord or peripherally from nerve or muscle. However, transcranial stimulation has not received widespread acceptance as a monitoring tool largely because of the need for special equipment to elicit these responses, the sensitivity of the responses to commonly used anesthetics, and concerns about patient safety. The use of spinal stimulation to elicit and record muscle activity as a monitoring tool is generally associated with undesirable patient movement because of the requirement that the patient's muscles not be paralyzed. On the other hand, the use of spinal stimulation to elicit and record responses from peripheral nerves has been reported not to have the limitations of the previous monitoring techniques (Owen, 1997). This monitoring technique was introduced by Owen et al. (1988) who named these peripheral nerve responses neurogenic motor evoked potentials (NMEPs). Although they are called motor responses, it is generally agreed that because of the placement of the stimulation electrodes, they are composed of both orthodromic motor and antidromic sensory components (Machida et al., 1985; Yokoyama et al., 1991; Su et al., 1992; Deletis, 1993; Kai et al., 1993; Owen, 1993, 1997; Haghighi et al., 1994; Machida, 1994; Burke and Hicks, 1998; Padberg et al., 1998; Rose, 1998; Toleikis et al., 1999). NMEPs consist
1388-2457/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88-2457(99)0031 7-X
CLINPH 98130
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
of a prominent early biphasic component followed by polyphasic activity (Fig. 1). It has been reported that the large early biphasic component is composed of motor activity and the following polyphasic activity is sensory (Kai et al., 1993; Owen, 1993, 1997). However, this description of the functional composition of NMEPs is controversial. Numerous theoretical (Deletis, 1993; Rose, 1996, 1998), animal (Yokoyama et al., 1991; Su et al., 1992; Haghighi et al., 1994), and human (Machida et al., 1985; Machida, 1994; Bernard et al., 1996; Leppanen et al., 1999) studies have provided evidence which supports the opinion that NMEPs are primarily composed of sensory rather than motor activity. The purpose of this study was to utilize the collision technique to de®ne the composition of these responses so as to substantiate or refute the basis for their utilization. 2. Materials and methods Data was acquired from 7 patients (3 males and 4 females) ranging in age from 10 to 15 years. All were diagnosed as having idiopathic scoliosis and underwent surgery for scoliosis repair at the Hughston Sports Medicine Hospital. Their preoperative neurologic examinations were normal. As part of our standard monitoring protocol, the neurological function of all of the patients was monitored throughout surgery using (1) somatosensory evoked potentials (SEPs) elicited by bilateral posterior tibial and ulnar nerve stimulation and (2) NMEPs elicited by percutaneous spinal stimulation. Collision studies were performed prior to incision and during surgery. These were performed using a combination of the two standard monitoring techniques and therefore informed consent was not obtained. All monitoring and collision study data were acquired on a Nicolet Viking IV. Standard stimulation and recording techniques were used to acquire the SEP responses. The NMEPs were an averaged response consisting of 100 separate responses elicited by means of a pair of 70 mm needle electrodes placed super-
737
®cial to the spinal lamina of the third and fourth cervical vertebrae for the purpose of electrically stimulating the underlying spinal cord. The stimulating electrodes were insulated except at their tips. Stimulation consisted of 4.7 Hz, 0.3 ms duration constant voltage (150±400 V) pulses. The NMEPs were recorded with subdermal needle electrodes from posterior tibial nerves at the ankle using a ®lter bandpass of 30±2000 Hz and the same shoulder ground used to record the SEPs. The ability to elicit well-de®ned, repeatable NMEPs determined the spinal stimulation intensity that was normally utilized for each patient. Occasionally, maximal spinal stimulation intensities of 400 V were used. Anesthetic management consisted of nitrous oxide (50%) and either des¯urane (3±5%) or iso¯urane (0.5±1%), fentanyl, and muscle relaxants. The concept and methodology for performing collision studies has previously been described (Kai et al., 1993; Owen, 1993). Stimulation of a peripheral nerve produces a volley of ascending sensory (AS) and ascending motor (AM) activity. The AM activity ascends to the anterior horn cells where it stops, whereas the AS activity continues to ascend in the dorsal column pathways of the spine (Fig. 2a). After a short delay, spinal stimulation follows and produces a descending volley believed to consist of both descending antidromic sensory (DS) and descending orthodromic motor (DM) activity. The time delay between the peripheral and spinal stimuli can be varied so that the resulting AS and DS volleys either (1) do not collide at all or (2) they collide partially or totally in the spinal cord or periphery. The portions of the AS and DS volleys which collide in the spinal cord will cancel and the remaining DS and DM activity can be peripherally recorded from the posterior tibial nerve ipsilateral to the side of peripheral stimulation (Fig. 2b). Spinal stimulation consisted of 4.7 Hz, 0.3 ms duration constant voltage (150±400 V) pulses. Because of large artifacts which result from the stimulation currents, the peripheral stimulation and recording sites could not be the same. Therefore, peripheral stimulation of the posterior tibial nerve stimulation occurred at the popliteal fossa using subdermal needle electrodes (4.7 Hz, 0.3 ms duration, 40 mA pulses) and the peripheral responses, like the NMEPs, were recorded with subdermal needle electrodes from the posterior tibial nerves at the ankle using a ®lter bandpass of 30±2000 Hz. 3. Results
Fig. 1. Typical peripherally recorded neurogenic motor evoked potential (NMEP) response elicited by spinal stimulation. The response consists of a prominent early biphasic component followed by small amplitude polyphasic components.
The amount of collision between the AS and DS volleys is dependent upon the delay time between the peripheral and spinal stimuli.. Total collision will occur when this delay time is less than the conduction time between the peripheral and spinal stimulation site for the entire AS volley. A study was conducted in one patient to determine the shortest conduction time of such a volley. The posterior tibial nerve was stimulated at the ankle and AS volleys were
738
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
electrodes) was found to be about 17 ms (Fig. 3a). Therefore, time delays of 17 ms or less should result in total collisions between the AS and DS volleys. This would leave just the DM volleys to be recorded at the ankle. Longer time delays should result in collisions between only the slower components of the AS and DS volleys. There would be no collisions between the faster AS and DS components of the volleys because the faster AS components would have already ascended beyond the spinal stimulation site. Therefore, the DM activity and only the faster conducting DS activity would be recorded at the ankle. In Fig. 3b, both the top and bottom traces illustrate the responses that result when no collisions take place (except for possibly the late components in the top trace). The bottom trace resulted from spinal stimulation and peripheral stimuli of such low intensity that no ascending activity was produced. The top trace resulted from spinal and peripheral stimulation with an inter-stimulus delay so large that the entire AS volley may have passed the spinal stimulation site before spinal stimulation occurred. Progressively shorter inter-stimulus delays resulted in collisions between more of the slower components of the AS and DS volleys and the amount of resulting descending activity recorded at the ankle became smaller. When the inter-stimulus interval was shortened to 17 ms, all the components of the AS and DS volleys collided leaving just the DM components to be recorded at the ankle. However, the result was a total absence of a peripherally recorded response. An example of comparable ®ndings from another patient is illustrated in Fig. 4. The lack of any remaining activity suggests that the descending volleys resulting from spinal stimulation contain only sensory activity and that these volleys and the AS volleys resulting from peripheral stimulation are mediated by the same sensory pathways. Such ®ndings were consistent for all 7 patients in the study.
Fig. 2. Collision technique. (a) Stimulation of the posterior tibial nerve (PTn) at the popliteal fossa occurs at t1. This results in ascending orthodromic sensory (AS) and ascending antidromic motor (AM) activity. The AM activity is blocked at the anterior horn cell synapse. The AS activity enters the spinal cord and is conveyed by the dorsal column pathways. For collision to occur in the spinal cord, spinal stimulation occurs at t2, approximately 15 ms after t1. It results in descending antidromic sensory (DM) and descending orthodromic motor (DM) activity. The AS and DS activity collide and proceed no further. The DM activity proceeds down the spinal cord. (b) The post-collision DM activity exits the spinal cord and proceeds down the PTn to near the medial malleolus where it is recorded at t3, approximately 25 ms after the onset of the spinal stimulus.
recorded with subdermal needle electrodes from the popliteal fossa, from the spinal stimulation electrodes used to produce the descending volleys, and from subdermal needle electrodes placed in the high cervical region and on the scalp. The conduction time for the fastest conducting ®bers (i.e. from the onset of the popliteal fossa responses to the onset of the responses recorded from the spinal stimulation
4. Discussion The descending volleys of activity, known as NMEPs, that result from percutaneous spinal stimulation appear to be primarily composed of sensory activity components mediated by the same neural pathways as SEPs. These ®ndings are consistent with those of Leppanen et al. (1999) and contradict those of the Owen group (Kai et al., 1993; Owen, 1993) who also utilized the collision technique to assess the composition of NMEPs. Leppanen et al. (1999) reported that collisions resulted in the early biphasic components of the NMEP being mostly or totally blocked, whereas the Owen group (Kai et al., 1993; Owen, 1993) reported that collisions resulted in a loss of only the small, polyphasic late components of the NMEPs but the large, early biphasic components were unaffected. As a result, the ®ndings of the Leppanen group suggest that the early large biphasic component of NMEPs consists of primarily sensory activity whereas the Owen group has reported that this component
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
739
consists of motor activity. It is dif®cult to explain this large discrepancy since the Leppanen and Owen groups used similar methodologies. A ¯aw in the collision technique itself does not explain
Fig. 4. Collision study in patient #2 (J.R. ± 14-year-old male). Peripheral responses acquired using spinal stimuli (4.7 Hz, 0.3 ms, 320 V pulses) with no peripheral stimuli (bottom trace) or supramaximal peripheral stimuli (4.7 Hz, 0.3 ms, 40 mA constant current pulses, top 3 traces) at the popliteal fossa. There is no difference between the early biphasic portions of the response with no collisions and the response with a 25 ms inter-stimulus delay. As the delay between the peripheral and spinal stimuli decreased, the post-collision responses also decreased. An 18 ms inter-stimulus delay resulted in total obliteration of the descending NMEP response (third trace).
this discrepancy but is one possible explanation for the apparent lack of any signi®cant descending motor (DM) activity in our results and those of Leppanen et al. (1999). The ascending volleys could cause a blockage of DM activity by affecting motor neuron or inter-neuron excitability through recurrent inhibition or other mechanisms (Eccles, 1964). However, based on our data, this explanation seems unlikely. In Fig. 3b, there is no difference between the response in the bottom trace that results from spinal stimulation and minimal peripheral stimulation (and therefore no
Fig. 3. Collision study in patient #1 (K.B. ± 15-year-old female). (a) SEPs acquired using constant current stimulation (4.7 Hz, 0.3 ms, 40 mA pulses) of the posterior tibial nerve at the medial malleolus. Note that the fastest ascending sensory activity takes approximately 17 ms to travel from the popliteal fossa to the spinal stimulation electrodes. (b) Peripheral responses acquired using spinal stimuli (4.7 Hz, 0.3 ms, 400 V pulses) and different intensities of peripheral stimuli (4.7 Hz, 0.3 ms, constant current pulses) (bottom trace, subthreshold 1 mA; top 3 traces, 40 mA). As the delay between the peripheral and spinal stimuli decreased, the responses that remained after collisions also decreased in size. With a 17 ms delay (the observed SEP conduction time from the popliteal fossa to the stimulation electrodes ± (a)), the descending peripheral response was totally obliterated (second trace from the bottom). The large stimulus artifact is due to the use of a shoulder recording ground.
740
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
ascending inhibitory in¯uences) and the response in the top trace that results from spinal stimulation and supramaximal peripheral stimulation. If ascending inhibitory in¯uences are present and the responses contain motor activity, there should be some difference in these traces but none are evident except in possibly the late polyphasic components. It is also possible that these in¯uences could be present but have a shorter duration of action than the inter-stimulus delay period. However, if this were the case and the early biphasic component contained motor activity, then as the inter-stimulus delay became shorter, these in¯uences should affect the initial portions of the biphasic component. However, no changes were observed in this component except for those explained by the collision process itself. Therefore, regardless of whether ascending inhibitory in¯uences are present or not, the large biphasic component appears to contain only sensory activity and the late polyphasic components may consist of sensory and/or motor activity. Another possible explanation for the lack of motor activity in our responses could be the anesthetic management used during our studies. Our data were acquired using concentrations of volatile anesthetic agents common to spinal surgeries. Recent studies have reported that volatile anesthetics signi®cantly depress spinal cord motor neuron excitability (King and Rampil, 1994; Zhou et al., 1997). Therefore, it is likely that the anesthetic management used in our studies could have some effect on our responses if motor activity was present. However, Bernard et al. (1996) have shown that increasing concentrations of iso¯urane have no effect on the early NMEP components, thus supporting our ®ndings that these components contain sensory rather than motor activity. We also considered the possible effects that stimulation intensity might have on the composition of the descending volley. Because of the placement of the spinal stimulation electrodes, descending sensory and descending motor activity would probably be mediated by the dorsal columns and the lateral corticospinal tracts, respectively. Because the lateral corticospinal tracts would probably be located more distal to the stimulation electrodes than the dorsal columns, higher stimulation intensities might be required to excite these tracts. However, maximal spinal stimulation intensities (400 V) did not affect our ®ndings. Therefore, within the practical stimulation limits of the equipment we were using, the early components of the percutaneously elicited peripheral responses appear to contain only sensory activity. Animal studies have also indicated that the responses elicited by low intensity spinal cord stimulation contain primarily sensory activity (Su et al., 1992; Haghighi et al., 1994). Epidural stimulation may be more effective than percutaneous translaminar stimulation in eliciting motor activity because of current spread to the lateral corticospinal tracts. However, in other animal studies, when higher stimulation intensities were used and motor activity was present, the motor activity had slower conduction velo-
cities than the sensory activity (Yokoyama et al., 1991), thus supporting the idea that if motor activity is present in the peripheral nerve responses, it probably occurs in the late polyphasic components or even later or may go unnoticed because it is small in amplitude, temporally dispersed or delayed (Machida et al., 1985) relative to the early large biphasic component. In support of this idea, it has been reported that myogenic activity from leg muscles can be present even in the absence of any peripheral nerve responses, thus demonstrating that the peripheral nerve responses resulting from percutaneous spinal stimulation do not contain a major motor component (Leppanen et al., 1999). Therefore, monitoring the myogenic activity from leg muscles in response to spinal stimulation appears to be a better way to assess spinal motor function than recording peripheral nerve responses. However, it is possible that such myogenic activity can also result from re¯ex activation of alpha-motoneurons by antidromically activated dorsal column pathways (Deletis, 1993; Rose, 1996, 1998). Although the peripheral nerve responses elicited by spinal stimulation appear to be mediated by the same pathways as SEPs, the acquisition of both monitoring modalities can serve as useful adjuncts to one another. Although SEPs can normally be acquired very easily, the peripheral nerve responses can typically be acquired more quickly than SEPs and may be more robust as well. In addition, there are instances when useful SEPs cannot be acquired either due to anesthetic management or pathology and the acquisition of these peripheral nerve responses may be the only means available for assessing spinal cord function. In addition, these peripheral nerve responses have been reported to be a more rapid or more sensitive indicator of changes in spinal cord function than traditional cortical SEPs (Mustain and Kendig, 1991; Owen et al., 1991; Owen, 1993, 1997; Owen and Tamaki, 1997; Padberg et al., 1998; Siddiqui et al., 1998). This ®nding is questionable considering that both monitoring modalities appear to be mediated by the same neural pathways, however, relative sensitivity may depend on the recording sites that are used to acquire responses rather than the monitoring modality itself. The SEPs that are recorded over the cervical spine are probably preferable to cortical SEPs because the amplitude of the dorsal column-mediated ascending traveling wave which produces these subcortical responses is directly related to the size of the incoming afferent volley (Burke and Hicks, 1998), whereas it has been reported that a greater than 80% reduction in the amplitude of the dorsal column-mediated incoming volley is required before the amplitude of the cortical SEPs is reduced by 50% (Eisen et al., 1982; Gandevia and Burke, 1984). Similarly, the peripheral nerve responses which result from spinal stimulation should be directly related to the descending afferent volley and may be more sensitive to changes in spinal function than cortical SEPs. Nevertheless, the use of only SEPs and/or the spinally elicited peripheral nerve responses for monitoring purposes may be insensitive to spinal cord insults which produce only
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
changes in motor pathway function. Therefore, despite the reports of others (Padberg et al., 1998), it may be unwise to abandon the use of the wake-up test when neither of these monitoring modalities demonstrates signi®cant changes. Transcranial rather than spinal stimulation techniques are now available for directly assessing motor pathway function (Edmonds et al., 1989; Zentner, 1989; Zentner and Ebner, 1989; Haghighi et al., 1990a,b; Calancie et al., 1991, 1998; Burke et al., 1992; Schmid et al., 1992; Zentner et al., 1992, 1997; Deletis, 1993; Jones et al., 1996; Linden et al., 1997; Morota et al., 1997; Burke and Hicks, 1998). These techniques have not yet received FDA approval and therefore their use for monitoring purposes currently requires institutional review board approval and informed consent. However, the special equipment needed to utilize these techniques is relatively inexpensive, responses can be acquired using common anesthetic management (Jones et al., 1996; Calancie et al., 1998), and there have been no reported safety problems associated with their use. Therefore, there appears to be very little reason for not considering their use when an accurate assessment of spinal cord motor function is required.
References Bernard JM, Pereon Y, Fayet G, Guiheneuc P. Effects of iso¯urane and des¯urane on neurogenic motor- and somatosensory-evoked potential monitoring for scoliosis surgery. Anesthesiology 1996;85:1013±1019. Burke D, Hicks RG. Surgical monitoring of motor pathways. J Clin Neurophysiol 1998;15:194±205. Burke D, Hicks R, Stephen J, Woodworth I, Crawford M. Assessment of corticospinal and somatosensory conduction simultaneously during scoliosis surgery. Electroenceph clin Neurophysiol 1992;85:388±396. Calancie B, Klose KJ, Baier S, Green BA. Iso¯urane induced attenuation of motor evoked potentials caused by electrical motor cortex stimulation during surgery. J Neurosurg 1991;74:897±904. Calancie B, Harris W, Broton JG, Alexeeva N, Green BA. ``Thresholdlevel'' multipulse transcranial electrical stimulation of motor cortex for intraoperative monitoring of spinal motor tracts: description of method and comparison to somatosensory evoked potential monitoring. J Neurosurg 1998;88:457±470. Deletis V. Intraoperative monitoring of the functional integrity of the motor pathways. In: Devinsky O, Beric A, Dogali M, editors. Electrical and magnetic stimulation of the brain and spinal cord, New York: Raven Press, 1993. pp. 201±214. Eccles JC. The physiology of synapses, New York: Academic Press, 1964. Edmonds HL, Paloheimo MPJ, Backman MH, Johnson JR, Holt RT, Shields CB. Transcranial magnetic motor evoked potentials (tcMMEP) for functional monitoring of motor pathways during scoliosis surgery. Spine 1989;14:683±686. Eisen A, Purves S, Hoirch M. Central nervous system ampli®cation: its potential in the diagnosis of early multiple sclerosis. Neurology 1982;32:359±364. Engler GL, Spielholtz NI, Bernard WN, Danziger F, Merkin H, Wolff T. Somatosensory evoked potentials during Harrington instrumentation for scoliosis. J Bone Joint Surg 1978;60A:528±532. Gandevia SC, Burke D. Saturation in human somatosensory pathways. Exp Brain Res 1984;54:582±585. Haghighi SS, Madsen R, Green KD, Oro JJ, Kracke GR. Suppression of motor evoked potentials by inhalational anesthetics. J Neurosurg Anesthesiol 1990a;2:73±78. Haghighi SS, Green KD, Oro JJ, Drake RK, Kracke GR. Depressive effect
741
of iso¯urane anesthesia on motor evoked potentials. Neurosurgery 1990b;26:993±997. Haghighi SS, York DH, Gaines RW, Oro JJ. Monitoring of motor tracts with spinal cord stimulation. Spine 1994;19:1518±1524. Jones SJ, Harrison R, Koh KF, Mendoza N, Crockard HA. Motor evoked potential monitoring during spinal surgery: responses of distal limb muscles to transcranial cortical stimulation with pulse trains. Electroenceph clin Neurophysiol 1996;100:375±383. Kai Y, Owen JH, Lenke LG, Bridwell KH, Oakley DM, Sugioka Y. Use of sciatic neurogenic motor evoked potentials versus spinal potentials to predict early-onset neurologic de®cits when intervention is still possible during overdistraction. Spine 1993;18:1134±1139. King BS, Rampil IJ. Anesthetic depression of spinal motor neurons may contribute to lack of movement in response to noxious stimuli. Anesthesiology 1994;81:1484±1492. Leppanen R, Madigan R, Sears C, Maguire J, Wallace S, Captain J. Intraoperative collision studies demonstrate descending spinal cord stimulated evoked potentials and ascending somatosensory evoked potentials are mediated through common pathways. J Clin Neurophysiol 1999;16:170. Linden RD, Johnson JR, Shields CB, Edmonds HL, Glassman SD. Intraoperative spinal cord monitoring with motor evoked potentials elicited by transcranial magnetic stimulation. In: Bridwell KH, Dewald RL, editors. The textbook of spinal surgery, 2nd ed. Philadelphia, PA: Lippincott-Raven, 1997. pp. 77±83. Machida M. Point of view. Spine 1994;19:1167±1168. Machida M, Weinstein SL, Yamada T, Kimura J. Spinal cord monitoring. Electrophysiological measures of sensory and motor function during spinal surgery. Spine 1985;10:407±413. Morota N, Deletis V, Constantini S, Ko¯er M, Cohen H, Epstein FJ. The role of motor evoked potentials during surgery for intramedullary spinal cord tumors. Neurosurgery 1997;41:1327±1336. Mustain WD, Kendig RJ. Dissociation of neurogenic motor and somatosensory evoked potentials: a case report. Spine 1991;16:851±853. Nash CL, Lorig RA, Schatzinger LA, Brown RH. Spinal cord monitoring during operative treatment of the spine. Clin Orthop 1977;126:100±105. Nuwer MR, Dawson E. Intraoperative evoked potential monitoring of the spinal cord: enhanced stability of cortical recordings. Electroenceph clin Neurophysiol 1984;59:318±327. Owen JH. Intraoperative stimulation of the spinal cord for prevention of spinal cord injury. In: Devinsky O, Beric A, Dogali M, editors. Electrical and magnetic stimulation of the brain and spinal cord, New York: Raven Press, 1993. pp. 271±288. Owen JH. Monitoring during surgery for spinal deformities. In: Bridwell KH, Dewald RL, editors. The textbook of spinal surgery, 2nd ed. Philadelphia, PA: Lippincott-Raven, 1997. pp. 39±60. Owen JH, Tamaki T. Applications of neurophysiological measures during surgery of the spine. In: Frymoyer JW, editor. The adult spine: principles and practice, 2nd ed. Philadelphia, PA: Lippincott-Raven, 1997. pp. 673±702. Owen JH, Laschinger J, Bridwell K, et al. Sensitivity and speci®city of somatosensory and neurogenic-motor evoked potentials in animals and humans. Spine 1988;13:1111±1118. Owen JH, Bridwell KH, Grubb R, et al. The clinical application of neurogenic motor evoked potentials to monitor spinal cord function during surgery. Spine 1991;16:S385±S390. Padberg AM, Wilson-Holden TJ, Lenke LG, Bridwell KH. Somatosensoryand motor-evoked potential monitoring without a wake-up test during idiopathic scoliosis surgery. Spine 1998;23:1392±1400. Rose RD. Sensory component of cervically evoked motor potentials. Med Hypoth 1996;46:577±579. Rose RD. Removing the antidromically driven sensory component from cervically evoked motor potentials. Med Hypoth 1998;50:147±154. Schmid UD, Boll J, Liechti S, Schmid J, Hess CW. In¯uence of some anesthetic agents on muscle responses to transcranial magnetic stimulation: a pilot study in humans. Neurosurgery 1992;30:85±92. Siddiqui A, Blackburn TW, Page RB, Grubb MR, Russell GB. Spinal cord
742
J.R. Toleikis et al. / Clinical Neurophysiology 111 (2000) 736±742
stimulation reliably elicits motor evoked potentials. Presented at the Ninth Annual Meeting of the American Society of Neurophysiological Monitoring. Philadelphia, PA, 1998. Su CF, Haghighi SS, Oro JJ, Gaines RW. ``Back®ring'' in spinal cord monitoring. High thoracic spinal cord stimulation evokes sciatic response by antidromic sensory pathway conduction, not motor tract conduction. Spine 1992;17:504±508. Toleikis JR, Skelly JP, Carlvin AO, Burkus JK. Spinally elicited neurogenic ``motor'' evoked potentials (NMEPs) assess sensory rather than motor function. J Clin Neurophysiol 1999;16:169. Yokoyama M, Shinomiya K, Sato H, Komori H, Furuya K. Monitoring motor function of the spinal cord: the descending segmental evoked spinal cord potential. In: Shomoji K, Kurokawa T, Tamaki T, Willis WD, editors. Spinal cord monitoring and electrodiagnosis, Berlin: Springer, 1991. pp. 27±35.
Zentner J. Noninvasive motor evoked potential monitoring during neurosurgical operations on the spinal cord. Neurosurgery 1989;24:709± 712. Zentner J, Ebner A. Nitrous oxide suppresses the electromyographic response evoked by electrical stimulation of the motor cortex. Neurosurgery 1989;24:60±62. Zentner J, Albrecht T, Heuser D. In¯uence of halothane, en¯urane, and iso¯urane on motor evoked potentials. Neurosurgery 1992;31:298± 305. Zentner J, Thees C, Pechstein U, Scheu¯er K, Wurker J, Nadstawek J. In¯uence of nitrous oxide on motor-evoked potentials. Spine 1997;22:1002±1006. Zhou HH, Mehta M, Leis AA. Spinal cord motoneuron excitability during iso¯urane and nitrous oxide anesthesia. Anesthesiology 1997;86:302± 307.