Evidence-based guideline update: intraoperative spinal monitoring with somatosensory and transcranial electrical motor evoked potentials: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology and the American Clinical Neurophysiology Society.

OBJECTIVE: To evaluate whether spinal cord intraoperative monitoring (IOM) with somatosensory and transcranial electrical motor evoked potentials (EPs) predicts adverse surgical outcomes. METHODS: A panel of experts reviewed the results of a comprehensive literature search and identified published studies relevant to the clinical question. These studies were classified according to the evidence-based methodology of the American Academy of Neurology. Objective outcomes of postoperative onset of paraparesis, paraplegia, and quadriplegia were used because no randomized or masked studies were available. RESULTS AND RECOMMENDATIONS: Four Class I and 8 Class II studies met inclusion criteria for analysis. The 4 Class I studies and 7 of the 8 Class II studies reached significance in showing that paraparesis, paraplegia, and quadriplegia occurred in the IOM patients with EP changes compared with the IOM group without EP changes. All studies were consistent in showing all occurrences of paraparesis, paraplegia, and quadriplegia in the IOM patients with EP changes, with no occurrences of paraparesis, paraplegia, and quadriplegia in patients without EP changes. In the Class I studies, 16%-40% of the IOM patients with EP changes developed postoperative-onset paraparesis, paraplegia, or quadriplegia. IOM is established as effective to predict an increased risk of the adverse outcomes of paraparesis, paraplegia, and quadriplegia in spinal surgery (4 Class I and 7 Class II studies). Surgeons and other members of the operating team should be alerted to the increased risk of severe adverse neurologic outcomes in patients with important IOM changes (Level A).

Alteration of the cortical motor map in a patient with intractable focal seizures.

Patients with epilepsia partialis continua may develop progressive neurological deficits of unclear origin. It is possible that repetitive epileptic spikes induce plastic changes in the cortex analogous to the changes observed following direct microstimulation. A child is reported with focal cortical dysplasia, intractable focal seizures, worsening hemiparesis, and alteration of the cortical motor map over time. At age 7, he underwent cortical motor mapping before partial resection of a seizure focus within the right postcentral gyrus. No deficits were present after surgery, and seizure frequency declined by more than 90%. Seizures subsequently worsened and a progressive left hemiparesis developed. Cortical remapping at age 12 showed motor centres for left arm, face, and eye movements in an unusual configuration. The location of the motor representation of the face differed from the location obtained at age 7. This case provides direct electrophysiological evidence of reorganisation of the cortical motor map in the human brain.

Topography of the initial cortical component of the median nerve somatosensory evoked potential. Relationship to central sulcus anatomy.

The initial cortical component of the median nerve somatosensory evoked potential (SSEP), the parietal N20, is generated in the posterior bank of the central sulcus and inverts in polarity across the sulcus. The inversion is used to identify the central sulcus. The precentral P20 is sometimes not identifiable in scalp recordings, and this has been attributed to a dipole orientation that directs the maximum positivity downward, into the brain. The authors mapped cortical SSEPs during resection of an arteriovenous malformation in the left sensorimotor area. Preoperative scalp SSEPs over the right hemisphere were normal with a frontal P20, but those over the left hemisphere had an unusual topography with a frontal N20. Intraoperative cortical surface recordings demonstrated an N20-P20 inversion in the inferior-superior rather than the usual posterior-anterior direction. This was a result of the trajectory of the central sulcus over the surface of the brain. The section containing the hand representation was coursing in an anterior-posterior direction. This anatomic variant is an additional cause of absent frontal P20 in scalp recordings. Variations in central sulcus anatomy may cause unusual SSEP topographies, but two-dimensional SSEP mapping and correlation with the sulcal anatomy can still permit localization of the central sulcus in such cases.

The association between seizure clustering and convulsive status epilepticus in patients with intractable complex partial seizures.

PURPOSE: We examined the association between seizure clustering and convulsive status epilepticus (SE) in patients with intractable complex partial seizures, to identify whether patients whose seizures typically cluster are at high risk for convulsive SE (CSE). METHODS: Seventy-six patients with intractable complex partial epilepsy who underwent presurgical evaluation in the Montefiore Epilepsy Management Unit from 1993 to 1997 were contacted and interviewed about typical seizure frequency and distribution and history of CSE. Seizure clustering was defined as three or more complex partial seizures within a 24-h period, with return to baseline between seizures. RESULTS: Of the 76 patients contacted, 21 (28%) had experienced at least one episode of CSE, and 36 (47%) typically experienced clustered seizures. SE occurred in 16 (44%) of 36 patients with clustered seizures, and in five (12.5%) of 40 patients with nonclustered seizures (p < 0.002). Of 53 patients with temporal lobe epilepsy, CSE occurred in 13 (50%) of 26 patients with clustered seizures, and four (14.8%) of 27 patients with nonclustered seizures (p < 0.006). CONCLUSIONS: Patients with intractable complex partial or localization-related epilepsy who typically experience seizure clustering are at a significantly higher risk for CSE than are patients with nonclustered seizures.

Angiographic criteria reliably predict when carotid endarterectomy can be safely performed without a shunt.

BACKGROUND: Selective shunting during carotid endarterectomy is widely performed, but the optimal approach for predicting when a shunt is unnecessary remains uncertain. We evaluated the ability of preoperative cerebral angiography to predict when carotid endarterectomy could be safely performed without a shunt. STUDY DESIGN: Eighty-seven patients undergoing carotid endarterectomy between August 1991 and December 1997 had preoperative cerebral angiograms. The angiograms were evaluated for the presence of collateral flow from the contralateral carotid through the anterior communicating artery and from the posterior circulation through the posterior communicating artery. Patients then underwent endarterectomy and were selectively shunted based on somatosensory evoked potential changes. Internal carotid artery stump pressure was routinely measured in all patients. RESULTS: Nine patients (10%) had a shunt placed based on somatosensory evoked potential changes and none of the 87 patients had a perioperative (30 days) stroke. Angiography revealed that 36 patients (41%) had no cross-filling from the contralateral carotid through the anterior communicating artery. Nine of these patients (25%) required a shunt; none of the 51 patients with adequate cross-filling (p < 0.001) did. Furthermore, 94% of the patients without cross-filling but with a patent ipsilateral posterior communicating artery did not require a shunt using somatosensory evoked potential changes as the standard for shunt insertion. Stump pressure measurements (> or = 25 mmHg) or (> or = 50 mmHg) did not reliably exclude the need for a shunt. Only 2 of 15 patients with contralateral carotid occlusion and 1 of 16 patients with a prior ipsilateral stroke required shunts. CONCLUSIONS: In the presence of cross-filling from the contralateral carotid artery, shunt insertion was uniformly unnecessary. In addition, routine shunting of patients with previous ipsilateral strokes or contralateral carotid occlusion was not always necessary. Stump pressures were less sensitive than angiographic criteria in determining when a shunt was unnecessary. Evaluation of cross-filling from the contralateral carotid artery on preoperative angiography can predict with certainty which patients will not require a shunt.

Seizure lateralization during EEG monitoring in patients with bilateral foci: the cluster effect.

PURPOSE: To determine whether seizures that occur in clusters are more likely to reflect activity of the same focus than are seizures that are widely separated in time. METHODS: EEG monitoring data from 14 patients with bilateral independent seizure onsets were analyzed. Twelve of the 14 patients had surface recordings only, and two had implanted electrodes. Interseizure intervals (ISIs) for 151 seizure pairs were measured. Seizure onsets were classified as right hemispheric, left hemispheric, or indeterminate. Seizure pairs were classified as concordant for hemisphere of onset, discordant, or indeterminate. The relation between seizure-pair concordance and ISI was examined by using univariate analysis and analysis of variance (ANOVA). RESULTS: Both seizures originated from the same hemisphere in 61 (75%) of 81 seizure pairs with ISIs <8 h, compared with 28 (55%) of 51 seizure pairs with ISIs >8 h (p < 0.015). The cluster effect was not more pronounced for ISIs <2 h. ANOVA demonstrated that the relation between ISI and seizure concordance was not a result of the variability in seizure rates among patients. In three patients, the presence of bilateral foci was not demonstrated until more than five seizures were recorded. CONCLUSIONS: Seizures that occur after an ISI of <8 h are more likely to come from the same side as the previous seizure than are those with longer ISIs. Thus clustered seizures should not be given the same weight as seizures widely separated in time. In addition, more than five seizures may sometimes be needed to adequately assess patients being evaluated for epilepsy surgery.

Surface mapping of spike potential fields: experienced EEGers vs. computerized analysis.

An EEG epileptiform spike focus recorded with scalp electrodes is clinically localized by visual estimation of the point of maximal voltage and the distribution of its surrounding voltages. We compared such estimated voltage maps, drawn by experienced electroencephalographers (EEGers), with a computerized spline interpolation technique employed in the commercially available software package FOCUS. Twenty-two spikes were recorded from 15 patients during long-term continuous EEG monitoring. Maps of voltage distribution from the 28 electrodes surrounding the points of maximum change in slope (the spike maximum) were constructed by the EEGer. The same points of maximum spike and voltage distributions at the 29 electrodes were mapped by computerized spline interpolation and a comparison between the two methods was made. The findings indicate that the computerized spline mapping techniques employed in FOCUS construct voltage maps with similar maxima and distributions as the maps created by experienced EEGers. The dynamics of spike activity, including correlations, are better visualized using the computerized technique than by manual interpretation alone. Its use as a technique for spike localization is accurate and adds information of potential clinical value.

Somatosensory evoked potential monitoring during carotid surgery.

Controversy exists over the value of intraoperative monitoring and shunting in patients undergoing carotid endarterectomy. Although it is widely believed that contralateral carotid occlusion and previous stroke mandate intraoperative shunting, the susceptibility of these two groups of patients to cerebral ischemia during carotid artery endarterectomy is not well defined. Somatosensory evoked potentials (SSEPs) were monitored in 113 carotid artery endarterectomy patients. Of these, 32 (28.3%) had a previous stroke, 24 (21.2%) had a contralateral carotid occlusion and 33 (29.2%) were diabetic. There were no deaths and only one perioperative stroke (0.9%). Cerebral ischemia occurred in 14 patients (12.4%). Six of these patients had a contralateral carotid occlusion. Some 29 patients (25.7%) were shunted, including 10 with contralateral carotid occlusions that did not have major SSEP changes. In the latter half of the study, 14 patients with contralateral carotid occlusions were selectively shunted (six shunted, eight not shunted) with no neurological complications. Thirty-two patients with prior strokes were selectively shunted (nine shunted, 23 not shunted); of these, one shunted patient undergoing combined carotid artery endarterectomy and coronary artery bypass grafting had a perioperative stroke. Intraoperative monitoring with SSEPs accurately identifies cerebral ischemia secondary to carotid clamping as well as patients requiring shunts. With use of intraoperative SSEP monitoring, selective shunting may be safely performed in patients with a contralateral carotid occlusion or a previous stroke.

Sacral root stimulation for controlled micturition: prevention of detrusor-external sphincter dyssynergia by intraoperative identification and selective section of sacral nerve branches.

Electrical stimulation of the S2 nerve root can be used to produce detrusor contraction and voiding in patients with spinal cord injury, but concurrent stimulation of the external urethral sphincter causes detrusor-sphincter dyssynergia. This has been managed with a second surgical procedure, peripheral transection of the pudendal nerve. In this study, performed in dogs after spinal cord transection, laminectomy and ventral foraminotomy permitted tracing of the S2 root into the pelvis, where its branches were identified by electrical stimulation and urodynamic recording. The pudendal (somatic) branch was sectioned; the autonomic branch innervating the detrusor was preserved. Electrical stimulation of the proximal S2 root then produced detrusor contraction without contraction of the external urethral sphincter. This approach, which requires a single operation and spares pudendal nerve functions mediated by nerve roots other than S2, may enable a neurostimulator to provide effective voiding, without detrusor-external sphincter dyssynergia, in man.

Electrical stimulation and multichannel EMG recording for identification of functional neural tissue during cauda equina surgery.

Electrical stimulation of structures within the surgical field was used to identify functional neural elements during 25 cauda equina operations. EMG responses from anterior thigh, posterior thigh, and anal sphincter muscles were recorded simultaneously using a multichannel signal averager. During nine operations, stimulation of a presumed filum terminale or other tissue produced clear EMG responses, prompting modification of surgical procedures. In one patient, this resulted in preservation of a flattened spinal cord which resembled a band of scar tissue. Some EMG responses were restricted to a single muscle group; these neural structures would probably not have been identified if only a single-channel EMG recording was used. Visual examination alone was not adequate for identifying functional neural elements, or for determining whether atretic-appearing nerve roots were functional. Electrical stimulation with multichannel EMG recording facilitates the preservation of functional neural elements and the optimization of surgical results in cauda equina surgery.

Normal brain-stem auditory evoked potentials with abnormal latency-intensity studies in patients with acoustic neuromas.

Brain-stem auditory evoked potentials (BAEPs) are highly sensitive for detecting acoustic neuromas but false-negative results occur. We studied BAEPs preoperatively in 39 cases of acoustic neuroma. Absolute and interpeak latencies ipsilateral to the tumor, and interaural latency differences, were normal in four patients with small tumors. In three of these, however, results of latency-intensity studies were abnormal. In one patient, the latency-intensity result became normal postoperatively. If acoustic neuroma is suspected, and BAEPs are normal by usual criteria, latency-intensity functions should be examined to maximize chances of detecting a small tumor.

The anatomic and physiologic bases of brain stem auditory evoked potentials.

The generators of human BAEPs, as summarized in Table 1 and Figure 12, are more complex than has often been presumed. Most peaks reflect activity in several structures, though patterns of BAEP abnormality can be correlated with abnormalities in general areas of the brain stem. The phasic peaks originate in action potentials, whereas the SN reflects postsynaptic potentials within brain-stem auditory nuclei.

Global aphasia without hemiparesis: multiple etiologies.

Acute global aphasia without hemiparesis has been considered pathognomonic of embolic stroke. During 1 year, we encountered six patients with this syndrome. Two had multiple strokes, probably embolic. One had atrial fibrillation; at autopsy, there were metastases as well as multiple infarcts in the left hemisphere. One had a single large infarct in the territory of an anterior branch of the middle cerebral artery (MCA), one had subarachnoid hemorrhage of unknown origin, and one had a sylvian fissure hematoma with intraparenchymal extension from a ruptured MCA aneurysm. Nonembolic etiologies are therefore also possible and include conditions that bar anticoagulation.

Short-latency auditory evoked potentials in the monkey. I. Wave shape and surface topography.

The components and topography of the short-latency auditory evoked potentials (SLAEPs) in the monkey were analyzed using a combination of surface and epidural electrodes. Nine positive peaks, and a slow negativity following wave 7, are identified in the SLAEP recorded at the vertex referred to the mastoid ipsilateral to a monaural stimulus. Subcomponents of waves 1 and 3 are consistently observed, and other waves are probably compound as well. Barbiturate anesthesia results in latency prolongations which are cumulative across components but does not alter the component sequence. With binaural stimulation there is simple summation of the monaural responses for the components preceding wave 7, but a binaural interaction for wave 7. The polarity and clarity of the various components depend on the recording configuration selected. No single configuration is optimal for the demonstration of all components, and the use of montages with multiple electrodes may increase the information gained from clinical SLAEP recordings in man. The monkey SLAEP components can be grouped based on similarities of their surface topographies, which suggest similar anatomic orientations of their underlying generators. Probable homologues between the components of the simian and human SLAEP, based on temporal sequence, topography, and the effects of variation of the stimulus parameters, are: Jewett's waves I through VII correspond to monkey waves 1, 3, and 5 through 9. Due to differences in the size and shape of the heads, the monkey waves 2 and 4 correspond to negativities in the human SLAEP, labeled IN and IIN after the components which they follow. Another peak, with latency between waves 7 and 8 in the monkey and V and VI in man, is only clearly seen in the mastoid-to-mastoid linkage, contralateral negative.

Short-latency auditory evoked potentials in the monkey. II. Intracranial generators.

The generators of the short-latency auditory evoked potentials (SLAEPs) in the monkey have been defined by intracranial mapping from cochlea to auditory cortex. SLAEP components other than 1a and the slow negativity (SN) following wave 7 derive from compound action potentials propagated in subcortical auditory pathways. The component generators are complex due to the presence of two bursts of activity in the eighth nerve, to the fact that the ascending auditory fibers both synapse on and bypass specific relay nuclei, and to the differences in orientation of segments of the auditory pathways. Most SLAEP components recorded at the surface reflect the summation of activity from multiple generators. However, much of the activity seen within subcortical structures cannot be traced to the surface of the brain. Component 1a is identified with the cochlear summating potential, while 1b reflects the initial afferent volley in the distal portion of the eighth nerve. Component 2 represents the initial depolarization of the eighth nerve terminals within the ipsilateral cochlear nucleus. Component 3h reflects the second volley of activity in the distal portion of the eighth nerve and the outflow of the cochlear nucleus which decussates in the trapezoid body. Component 3v represents the initial cochlear nucleus outflow volley ascending the lateral lemniscus. Component 4 principally reflects the second volley of activity within the eighth nerve terminals, and outflow from the ipsilateral superior olivary complex ascending in that lateral lemniscus, with a possible contribution from activity in the contralateral CNC. Component 5 represents the outflow of the contralateral superior olivary complex ascending in that lateral lemniscus. Component 6 reflects another volley from the ipsilateral superior olivary complex ascending in that lateral lemniscus, as well as outflow from both inferior colliculi propagating in their brachii. The generators of component 7 are the most complex encountered, representing volleys in both lateral lemnisci, activity of the contralateral inferior colliculus, and activity in both auditory radiations. A component that follows wave 7, seen best in mastoid-to-mastoid recording linkage, represents outflow from both inferior colliculi propagating in their brachia. Components 8 and 9 principally reflect propagated action potentials in the auditory radiations bilaterally, with an additional contribution from activity of both inferior colliculi. The SN mainly represents volume-conducted postsynaptic potentials from both inferior colliculi and cochlear nuclei.

Topography and intracranial sources of somatosensory evoked potentials in the monkey. II. Cortical components.

Averaged somatosensory evoked potentials (SEPs) and associated multiple unit activity (MUA) were recorded from a series of epidural and intracortical locations following stimulation of the contralateral median nerve in the monkey. Cortical components were differentiated from the earlier subcortical activity and the intracerebral distribution and sources of each cortical potential were determined. Under barbiturate anesthesia the SEP wave form is simplified and can be wholly attributed to two sources. The earliest cortical activity consists of a biphasic P10-N20 wave which is generated in the posterior bank of the central sulcus. A second wave form, P12-N25, originates in the crown of the postcentral gyrus. No other cortical areas are active. In the alert state the morphology of the surface SEP is complex and reflects the interaction of volume conducted activity from several adjacent cortical sources. The wave form overlying the hand area of the postcentral gyrus consists of P12, P20, P40, N45 and P110. Precentral recordings exhibit P10, P13, N13, N20, P24, N45 and P110. Six anatomical sources have been identified. P10 and N20 originate in the posterior bank of the central sulcus including areas 3a and 3b and are volume conducted in an anteroposterior direction. P12 originates in area 1 as well as the anterior portion of area 2. P20 is generated in the medial portion of the postcentral gyrus including area 5. The source of P40 lies within the lateral portion of the parietal lobe including area 7b. Two components were generated in precentral cortex: P13/N13 originates principally in area 4 within the anterior bank of the central sulcus and P24 reflects activity in the anteromedial portion of the precentral gyrus including area 6. The long latency SEP components, N45 and P110, are generated widely within the somesthetic areas of postcentral cortex. The early cortical SEP components recorded in the monkey closely resemble in configuration and topography those recorded from man although the latter are longer in latency, reflecting interspecies differences in the length of conduction pathways as well as in cortical processing time.

Averaged multiple unit activity as an estimate of phasic changes in local neuronal activity: effects of volume-conducted potentials.

A technique for the derivation of digitally-averaged multiple unit activity (MUA) is described. The use of signal averaging instead of analog integration improves the temporal resolution and thus provides a clearer picture of the instantaneous MUA level. MUA recordings have been used in the identification of regions active in the generation of event-related potentials, based in part on the limited volume within which a semi-microelectrode 'sees' action potentials. However, averaged MUA waveforms may be affected by time-locked activity volume-conducted to the electrode site. A theoretical analysis of the magnitude of this effect is presented, along with experimental data in support of its assumptions and predictions. The most important factor is not the absolute size of the volume-conducted potentials, but their magnitude relative to that of the locally-generated MUA. When full-wave rectification is used, volume-conducted activity which is a considerable fraction of the MUA level will not significantly affect the averaged MUA waveform. Half-wave rectification should not be used, as it leads to a much larger effect from small far-field potentials.