Patients with intractable epilepsy often undergo surgery to resect epileptogenic foci. Despite recent advances in modalities for preoperative diagnosis, the intraoperative electrocorticogram (ECoG) remains an important tool for surgery, providing definition of both the localization and extent of epileptic foci (1–4). However, ECoG findings are affected by most anesthetics (5–10). Careful attention to anesthetic management and interpretation of ECoG findings are therefore important for successful intraoperative mapping of epileptic foci under general anesthesia.
Sevoflurane is widely used as an anesthetic during neurosurgical procedures (11). However, data regarding the effects of sevoflurane on ECoG spike activities are limited with conflicting results. Hisada et al. (12) reported that sevoflurane induced a marked increase in ECoG spike activity in patients with temporal lobe epilepsy. Watts et al. (13) also indicated that spike frequency during 1.5 minimal alveolar concentration (MAC) sevoflurane anesthesia was significantly increased compared with that during 0.3 MAC isoflurane anesthesia. In contrast, Endo et al. (14) reported that 1.5 MAC sevoflurane significantly reduced ECoG spike activity in epileptic patients during fentanyl-based anesthesia.
Hyperventilation, which may be applied during neurosurgical procedures, can also affect ECoG spike activity. Tanaka et al. (15) observed that hyperventilation enhanced epileptiform activities on ECoG in epileptic patients under neuroleptoanesthesia. Yli-Hankala et al. (16) also reported that hyperventilation with large-dose sevoflurane induced epileptiform electroencephalogram (EEG) activity during anesthesia induction in nonepileptic patients. In contrast, Iijima et al. (17) reported that hyperventilation suppressed EEG spike activity during sevoflurane anesthesia. It is, therefore, controversial whether hyperventilation can enhance or suppress spike activity in epileptic patients under sevoflurane anesthesia.
In the present study, we tested the hypothesis that large-dose (1.5 MAC) sevoflurane and hyperventilation may enhance ECoG spike activity in epileptic patients. The frequency and extent of ECoG spike activity under 1.5 MAC sevoflurane, with or without hyperventilation, were compared with that under 0.5 MAC sevoflurane in 13 epileptic patients. In addition, we examined the relationship between awake and intraoperative ECoG in the subgroup of patients with chronically implanted subdural grid electrodes.
After institutional approval and informed consent, 13 patients, age ranging from 25 to 48 yr (mean, 35 yr), with intractable epilepsy were included in this study. Patient data are summarized in Table 1. All patients were scheduled for surgical resection of epileptic foci. In four of 13 patients, subdural electrodes were implanted 2 wk before the planned resection.
Patients received their regular anticonvulsant medication until the time of surgery and were premedicated with roxatidine acetate hydrochloride (histamine-2 receptor antagonist) 75 mg. Anesthesia was induced with IV injection of propofol (2.0 mg/kg), fentanyl (2 μg/kg), and vecuronium (0.15 mg/kg). After the trachea was intubated, the lungs were mechanically ventilated to maintain partial pressure of arterial carbon dioxide (Paco2) between 38 and 42 mm Hg. Anesthesia was maintained with nitrous oxide, sevoflurane, and intermittent fentanyl (total dose at the beginning of ECoG recording: range, 3.2–8.7 μg/kg; mean, 5.1 μg/kg). Vecuronium was continuously administered to maintain train-of-four response at T0 or T1. Routine monitoring for all patients included electrocardiogram, noninvasive arterial blood pressure, intraarterial catheter for arterial blood pressure measurements, pulse oximetry, end-tidal concentrations of carbon dioxide, nitrous oxide, and sevoflurane concentrations, and rectal temperature.
After craniotomy, administration of nitrous oxide was discontinued and air was added to the inspired gas mixture to maintain the fraction of inspired oxygen at 50%. After the dura was opened, grid electrodes were placed on the brain surface. When end-tidal nitrous oxide was verified to be <2%, ECoG was recorded using Neurofax EEG-1100 (Nihon Kohden, Tokyo, Japan) under the following 3 conditions: 1) 0.5 MAC (0.85%) of end-tidal sevoflurane under normo-ventilation (0.5 MAC-NV), 2) 1.5 MAC (2.5%) of end-tidal sevoflurane under normoventilation (1.5 MAC-NV), and 3) 1.5 MAC of end-tidal sevoflurane under hyperventilation (Paco2, 30–35 mm Hg) (1.5 MAC-HV). During the 10-min ECoG recordings at each condition, methoxamine was administered to maintain arterial blood pressure. To assess spike activity frequency, the number of spikes was counted in the most active lead for last 5 min during the 10-min recording interval. To assess the extent of areas with spike activity, the number of leads with more than 5 spikes per 5 min during the 10-min recording interval was counted and the percentage of leads with spikes measured was calculated. A spike was defined as a paroxysmal, isolated, high voltage, spontaneous electrical discharge with a triangular form lasting less than 70 ms that could be clearly distinguished from background ECoG. During burst suppression, spikes were defined as high-voltage, isolated discharges clearly distinguishable from the background bursts.
In 4 patients with chronically implanted subdural electrodes, the leads with ictal onset and those with interictal spikes were evaluated. The leads with ictal onset in more than 50% of seizures observed during the 2 wk-observation period were defined as the ictal onset zone. With 4 samples of interictal periods, which had typical interictal spikes in each patient, the number of spikes were counted for 5 min and averaged. Leads with more than 5 spikes per 5 min were used to define zones with interictal spikes. The leads with spikes during sevoflurane anesthesia under the conditions of 0.5 MAC-NV, 1.5 MAC-NV, and 1.5 MAC-HV were compared with those with ictal onset and interictal spikes in the awake state. Sensitivity and specificity between those variables were then calculated.
Hemodynamic, blood gas, and temperature data were compared using analysis of variance with repeated measures, followed by the Student-Newman-Keuls test for multiple comparisons. These values are expressed as mean ± sd. The number of spikes and percentage of leads with spikes were compared with the Kruskal-Wallis test, followed by the Wilcoxon's signed-rank test for multiple comparisons. These ECoG data are expressed as median (interquartile range). A probability value < 0.05 was considered statistically significant.
All 13 patients completed the study. Hemodynamic variables, temperature, and blood gas data are shown in Table 2. Mean arterial blood pressure during 1.5 MAC-NV and 1.5 MAC-HV was lower compared with 0.5 MAC-NV (P < 0.05). Heart rate during 1.5 MAC-NV and 1.5 MAC-HV were was more rapid compared with 0.5 MAC-NV (P < 0.05). Paco2 during 1.5 MAC-HV was lower compared with 0.5 MAC-NV and 1.5 MAC-NV. Arterial pH during 1.5 MAC-HV was higher than during 0.5 MAC-NV and 1.5 MAC-NV (P < 0.05). Temperature, Pao2, and hemoglobin were similar among the conditions.
Figure 1 shows the number of spikes and the percentage of leads with spikes during the three conditions. The numbers of spikes were increased during 1.5 MAC-NV and 1.5 MAC-HV compared with 0.5 MAC-NV (P = 0.0021, P = 0.0019, respectively). The number of spikes during the 1.5 MAC-HV was higher compared with 1.5 MAC-NV (P = 0.0088). The percentage of leads with spikes was larger during 1.5 MAC-NV and 1.5 MAC-HV compared with 0.5 MAC-NV (P = 0.0021, P = 0.0015, respectively). The percentage of leads with spikes during 1.5 MAC-HV was larger compared with 1.5 MAC-NV (P = 0.0077).
In the 4 patients with chronically implanted subdural electrodes, a total of 88 leads (22, 27, 16, and 23 leads in patients 1, 2, 3, and 4, respectively) could be used for both intraoperative and preoperative assessments. The frequency of seizures during the 2-wk observation period in 4 patients is shown in Table 1. In the awake state, 25 of 88 leads (28%) were defined as ictal onset. Interictal spikes in the awake state were observed in 57 of 88 leads (65%). With the same 88 leads, ECoG spike activity during 0.5 MAC-NV, 1.5 MAC-NV, and 1.5 MAC-HV was observed in 25 leads (28%), 58 leads (66%), and 66 leads (75%), respectively. The relationship (sensitivity and specificity) between the leads with epileptic discharges in the awake state and during sevoflurane anesthesia is shown in Table 3. For predicting the leads with ictal onset, sensitivity and specificity during 0.5 MAC-NV were both more than 80% (84% and 94%, respectively). For predicting the leads with interictal spikes, sensitivity and specificity during 1.5 MAC-NV were both more than 80% (91% and 81%, respectively). For prediction of leads with interictal spikes, specificity was reduced to 55% after the induction of hyperventilation. Figure 2 shows the representative ECoG findings under sevoflurane anesthesia and in the awake state in a patient (case 1) with chronically implanted subdural electrodes.
The present study showed that 1.5 MAC sevoflurane increased the frequency of spikes and extent of areas with spikes on ECoG compared with 0.5 MAC sevoflurane and that the addition of hyperventilation increased the frequency of spikes and the extent of spikes on ECoG in patients with refractory epilepsy. The areas with spikes under 0.5 MAC sevoflurane were similar to the areas of seizure onset in the awake state, whereas the areas with spikes under 1.5 MAC of sevoflurane were similar to those during the interictal periods in the awake state.
There have been several reports regarding epileptic discharges during sevoflurane anesthesia (12–14,17). Hisada et al. (12) investigated the epileptic discharges on ECoG in six patients with temporal lobe epilepsy and reported that ECoG spike activity was increased during sevoflurane (0.9%–1.0%) anesthesia with 67% nitrous oxide compared with the awake state and during isoflurane (0.85%–1.0%) anesthesia with 67% nitrous oxide. Watts et al. (13) also investigated the ECoG spike activity in 12 patients with refractory epilepsy and reported that spike activity during 1.5 MAC sevoflurane anesthesia was significantly increased compared with either the awake state or during 0.3 MAC isoflurane anesthesia. Iijima et al. (17) reported that epileptic discharges on EEG were more frequent during 2.0 MAC sevoflurane compared with during 1.0 MAC sevoflurane in epileptic patients, whereas 2.0 MAC sevoflurane did not evoke epileptic discharges in the nonepileptic patients. In contrast, Endo et al. (14) reported that the number of spikes on ECoG was significantly reduced after the induction of 1.5 MAC sevoflurane anesthesia in 10 patients with intractable temporal lobe epilepsy. The reason for these contradictory results is unknown. However, basic anesthetic status might have affected the results. In all studies that indicated increased spike activities under sevoflurane anesthesia, control data were collected in the awake state or under isoflurane or small-dose sevoflurane anesthesia with or without a small dose of fentanyl (12,13,17). In contrast, in the study by Endo et al. (14), large-dose fentanyl (mean dose, 46.7 μg/kg) was used with droperidol. Because fentanyl has been shown to be epileptogenic, it is a possibility that fentanyl propagated the spike activity during baseline recording (18).
Of importance, 1.5 MAC sevoflurane increased the extent of areas with spikes in the present study. Although the number of patients is small, the data from the patients with chronically implanted subdural electrodes clearly showed that the areas with spikes during 0.5 MAC sevoflurane were similar to those at the ictal onset, whereas 1.5 MAC sevoflurane was similar to awake interictal periods. There are several zones that we considered during epileptic surgery (4,19,20). Zones to generate interictal spikes, so-called “irritative zones,” are usually more extensive than the ictal onset zone. The ictal onset zone is the area where the seizure originates and its detection is the standard for localizing the epileptogenic zone, of which the removal or disconnection is necessary and sufficient for seizure freedom with a minimum of neurological or cognitive deficits. Using chronically implanted subdural electrodes, irritative zones with interictal spikes and ictal onset zones can be identified. Hisada et al. (12) demonstrated that areas with spikes under sevoflurane anesthesia were widely distributed but not confined to the ictal onset zone of spontaneous seizures. These are consistent with the results obtained in the present study. Although our study had a small number of patients, the results suggest that 1.5 MAC sevoflurane may be used for determination of irritative zones, whereas 0.5 MAC sevoflurane can be used for the prediction of the ictal onset zone. Such information may be effectively used for surgical planning, especially in patients in which chronic monitoring with implanted subdural electrodes was not performed.
Although hyperventilation has been used as a method to induce spike activity in awake subjects, there have been a few reports on the effects of hyperventilation on epileptic discharges under general anesthesia (15–17). Tanaka et al. (15) investigated the effects of hyperventilation on epileptiform activity in 20 children with medically intractable epilepsy under neuroleptanalgesia and demonstrated that hyperventilation enhanced or induced ECoG epileptiform activities in 17 patients and provoked EEG seizures in 10 patients. Yli-Hankala et al. (16) investigated the incidence of epileptiform EEG during mask induction of anesthesia with 8% sevoflurane in nonepileptic patients and demonstrated that the incidence of epileptiform activity was significantly more frequent under controlled hyperventilation compared with spontaneous breathing. In contrast, Iijima et al. (17) reported that the incidence of spikes was suppressed during hyperventilation under sevoflurane anesthesia in patients with epilepsy. In the present study, the induction of hyperventilation significantly increased the number of spikes and the percentage of leads with spikes in epileptic patients under 1.5 MAC sevoflurane. However, specificity for the prediction of leads with interictal spikes was decreased after the induction of hyperventilation. These data suggest that the addition of hyperventilation under large-dose sevoflurane anesthesia may overestimate electrically abnormal areas (irritative zones) with interictal spikes in the awake state. Hyperventilation should be carefully applied during the monitoring of spikes on ECoG in epileptic patients.
There are several limitations to this study. First, we only used 2 doses of sevoflurane (0.5 MAC and 1.5 MAC). Other doses may have different effects on ECoG spike activity. Further study may be required to determine the optimal doses of sevoflurane during epileptic surgery. Second, small doses of fentanyl were used. The effects of sevoflurane may vary depending on concomitant use of other drugs. Third, 5 minutes was the equilibrium period for each anesthetic condition; ECoG was then measured for the subsequent 5 minutes. Longer exposure to large-dose sevoflurane and hyperventilation might have different effects on ECoG. However, during the operation, there is usually little time for monitoring ECoG. The data in this study may be relevant to short-term ECoG assessments during epileptic surgery. Finally, arterial blood pressure was decreased after the induction of 1.5 MAC sevoflurane and hyperventilation, which may have affected the results. However, because the reduction in arterial pressure was modest, its influence on the results was likely negligible.
In summary, we investigated the effects of large-dose sevoflurane and hyperventilation on ECoG spike activity in epileptic patients. The areas with spikes on ECoG during sevoflurane anesthesia were compared with those at ictal onset and during interictal periods in the awake state. The results indicate that 1.5 MAC sevoflurane increased the frequency of spikes and extent of areas with spikes compared with 0.5 MAC sevoflurane and that the addition of hyperventilation also increased the frequency of spikes and the extent of spike activity. The areas with spikes during 0.5 MAC and 1.5 MAC sevoflurane corresponded to the areas with spikes at the seizure onset and during interictal periods, respectively, in the awake state. These data suggest that sevoflurane and hyperventilation should be carefully used during ECoG monitoring in epileptic patients. The clinical implications of ECoG spike activity during sevoflurane anesthesia may differ depending on the dosage of sevoflurane used.
1. Brainer-Lima PT, Rao S, Cukiert A, et al. Surgical treatment of refractory epilepsy associated with space occupying lesions: experience and review. Arq Neuropsiquiatr 1996;54:384–92.
2. Tran TA, Spencer SS, Javidan M, et al. Significance of spikes recorded on intraoperative electrocorticography in patients with brain tumor and epilepsy. Epilepsia 1997;38:1132–9.
3. Wennberg R, Quesney LF, Lozano A, et al. Role of electrocorticography at surgery for lesion-related frontal lobe epilepsy. Can J Neurol Sci 1999;26:33–9.
4. Kuruvilla A, Flink R. Intraoperative electrocorticography in epilepsy surgery: useful or not? Seizure 2003;12:577–84.
5. Modica PA, Tempelhoff R, White PF. Pro-and anticonvulsant effects of anesthetics (part I). Anesth Analg 1990;70:303–15.
6. Modica PA, Tempelhoff R, White PF. Pro-and anticonvulsant effects of anesthetics (part II). Anesth Analg 1990;70:433–44.
7. Fiol ME, Boening JA, Cruz-Rodrinuez R, Maxwell R. Effect of isoflurane (Forane) on intraoperative electrocorticogram. Epilepsia 1993;34:897–900.
8. Cheng MA, Tempelhoff R, Silbergeld DL, et al. Large-dose propofol alone in adult epileptic patients: electrocorticographic results. Anesth Analg 1996;83:169–74.
9. Herrick IA, Gelb AW. Anesthesia for temporal lobe epilepsy surgery. Can J Neurol Sci 2000;27:S64–7.
10. Sato Y, Sato K, Shamoto H, et al. Effect of nitrous oxide on spike activity during epilepsy surgery. Acta Neurochir (Wien) 2001;143:1213–6.
11. Duffy CM, Matta BF. Sevoflurane and anesthesia for neurosurgery: a review. J Neurosurg Anesthesiol 2000;12:128–40.
12. Hisada K, Morioka T, Fukui K, et al. Effects of sevoflurane and isoflurane on electrocorticographic activities in patients with temporal lobe epilepsy. J Neurosurg Anesth 2001;13:333–7.
13. Watts AD, Herrick IA, McLachlan RS, et al. The effects of sevoflurane and isoflurane anesthesia on interictal spike activity among patients with refractory epilepsy. Anesth Analg 1999;89:1275–81.
14. Endo T, Sato K, Shamoto H, Yoshimoto T. Effects of sevoflurane on electrocorticography in patients with intractable temporal lobe epilepsy. J Neurosurg Anesth 2002;14:59–62.
15. Tanaka T, Hashizume K, Kunimoto M, et al. Intraoperative electrocorticography in children with medically intractable epilepsy. Neurol Med Chir 1996;36:440–6.
16. Yli-Hankala A, Vakkuri A, Särkelä M, et al. Epileptiform electroencephalogram during mask induction of anesthesia with sevoflurane. Anesthesiology 1999;91:1596–603.
17. Iijima T, Nakamura Z, Iwao Y, et al. The epileptogenic properties of the volatile anesthetics sevoflurane and isoflurane in patients with epilepsy. Anesth Analg 2000;91:989–95.
18. Templelhoff R, Modica PA, Bernardo KL, Edwards I. Fentanyl-induced electrocirticographic seizures in patients with complex partial epilepsy. J Neurosurg 1992;77:201–8.
19. Englel J Jr. Multimoidal approaches in the evaluation of patients for epilepsy surgery. Clin Neurophysiol 1999;50(Suppl.):40–52.
© 2005 International Anesthesia Research Society
20. Rosenow F, Lüders H. Presurgical evaluation of epilepsy. Brain 2001;124:1683–700.