Share this article on:

Propofol Sedation During Awake Craniotomy for Seizures: Electrocorticographic and Epileptogenic Effects

Herrick, Ian A. BSc, MD, FRCPC; Craen, Rosemary A. MBBS, FANZCA; Gelb, Adrian W. MB, ChB, FRCPC; McLachlan, Richard S. MD, FRCPC; Girvin, John P. MD, FRCSC; Parrent, Andrew G. MD, FRCSC; Eliasziw, Michael PhD; Kirkby, Joyce RNA

Neurosurgical Anesthesia

This prospective study evaluated the effects of propofol sedation on the incidence of intraoperative seizures and the adequacy of electrocorticographic (ECoG) recordings during awake craniotomy performed for the management of refractory epilepsy.Thirty patients scheduled for temporal or frontal lobectomy for epilepsy under bupivacaine scalp block were randomized to receive patient-controlled propofol sedation (PCS) combined with a basal infusion of propofol (n = 15) or neurolept analgesia using an initial bolus dose of fentanyl (0.7 micro g/kg) and droperidol (0.04 mg/kg) followed by a fentanyl infusion (n = 15). Propofol administration was suspended 15 min before ECoG recording in the PCS group. The occurrence of inappropriate intraoperative seizures was noted and, based on blind review, the adequacy of ECoG recordings was compared. A higher incidence of intraoperative seizures was noted among the neurolept patients (6 vs 0, P = 0.008). Intraoperatively, ECoG recordings were adequate to proceed with resection in both groups. Evidence of low spike activity on ECoG did not correlate with the type of sedation administered. Higher frequency background ECoG activity was noted among patients who received propofol, but this did not interfere with ECoG interpretation. The use of propofol sedation does not appear to interfere with ECoG during epilepsy surgery, provided administration is suspended at least 15 min before recording.

(Anesth Analg 1997;84:1280-4)

Departments of (Herrick, Craen, Gelb, Kirkby) Anaesthesia, (McLachlan, Girvin, Parrent) Clinical Neurological Sciences, and (Eliasziw) Epidemiology and Biostatistics, London Health Sciences Centre, University of Western Ontario, J. P. Robarts Research Institute, London, Ontario, Canada.

Section Editor: Donald S. Prough.

This study was supported by a grant from the Physicians' Services Incorporated (PSI) Foundation.

Accepted for publication March 6, 1997.

Address correspondence to Ian A. Herrick, MD, Department of Anaesthesia, London Health Sciences Centre, University Campus, 339 Windermere Rd., London, Ontario N6A 5A5, Canada.

Cortical resection for the treatment of refractory seizures is often performed with the patient awake to facilitate intraoperative functional cortical mapping and to minimize interference with intraoperative electrocorticography (ECoG). Anesthesia typically includes regional blockade of the scalp combined with intravenous (IV) sedation. The need to minimize interference with ECoG recording limits the repertoire of drugs available for sedation. Traditionally, neurolept analgesia using a combination of opioid (often fentanyl) and droperidol has been a popular technique [1,2].

The use of propofol sedation during these procedures has been reported [3,4] and may be associated with several potential advantages compared with neurolept analgesia, such as ease of titration of sedation. However, the pro- and anticonvulsant effects of propofol remain controversial. Several investigators have reported the electroencephalographic (EEG) and ECoG effects of propofol. Most of these reports suggest that propofol has potent anticonvulsant effects and may depress epileptiform activity [5-8]. Thus, propofol would rarely be intentionally administered during intraoperative ECoG recording. However, because of its short duration of action, propofol administration could be suspended in advance of recording to minimize drug-induced adverse effects. Little information is available regarding the adequacy of intraoperative ECoG recordings when propofol is administered for clinical sedation during these procedures.

This prospective, randomized study was designed to evaluate the adequacy of ECoG recordings and the incidence of inappropriate intraoperative convulsions (seizure activity not associated with ECoG localization of the seizure focus) during awake seizure surgery performed under propofol sedation. This report is a component of a larger study designed to evaluate patient-controlled propofol sedation during these procedures (see accompanying article).

Back to Top | Article Outline


After institutional ethics approval and acquisition of written, informed consent, 30 adult patients (aged = 18-65 yr) scheduled for cortical resection for refractory seizures were randomized to receive either patient-controlled propofol sedation or neurolept analgesia (fentanyl and droperidol). Seven patients (from the parent study) were excluded from evaluation due to inconsistent or inadequate preoperative recordings (n = 4) or to the absence of preresection ECoG recordings (n = 3) (Table 1). One patient in this excluded group experienced an intraoperative tonicclonic convulsion. Data from this patient were included in the accompanying article but were excluded from this study because the absence of spike activity on preoperative and intraoperative ECoG recordings precluded evaluation of the ECoG effects of propofol.

Table 1

Table 1

Regional blockade of the scalp and local anesthetic infiltration of the incision lines were performed in accordance with our standard technique [9] using bupivacaine, as described in the accompanying article.

Sedation for the PCS group consisted of patient-administered propofol using a bolus dose of 0.5 mg/kg, a lockout interval of 3 min, and a basal infusion of 0.5 mg [centered dot] kg-1 [centered dot] h using a standard patient-controlled analgesia device (PCAII Baxter, McGaw Park, IL). For the neurolept group, sedation consisted of initial IV boluses of droperidol (total dose 0.04 mg/kg) and fentanyl (0.7 micro g/kg) followed by an anesthesiologist-controlled continuous infusion of fentanyl at 0.7 micro g [centered dot] kg-1 [centered dot] h-1. Both groups received supplemental anesthesiologist-administered fentanyl (25-micro g boluses) and dimenhydrinate (25-mg boluses) as needed for intraoperative pain and nausea or vomiting, respectively.

Neurolept or propofol sedation was started prior to surgery. Intraoperative monitoring consisted of electrocardiogram pulse oximetry and noninvasive automated blood pressure measurements. Each patient received supplemental oxygen via nasal prongs. Respiratory frequency was monitored using capnographic measurements obtained from a nasal prong sampling port.

ECoG recording was conducted for approximately 30 min prior to cortical resection in each patient. Recordings were conducted approximately 90-120 min after the commencement of sedation. Propofol sedation was suspended 15 min prior to ECoG recording. ECoG recordings were obtained from an electrode array (Western array) that included an arrangement of both surface and depth electrodes placed to include the site of seizure onset (based on preoperative interictal and ictal scalp EEG or subdural electrode recordings). All ECoG recordings were performed by registered EEG technicians.

Intraoperatively, the occurrence of inappropriate seizures (not associated with ECoG recording or cortical mapping) or problems with the quality of ECoG recordings were noted by an independent observer (JK). Postoperatively, ECoG records were reviewed by a neurologist (RM) who was blinded to the intraoperative sedation used. Blind ECoG evaluation was scored on the basis of spike frequency, spike localization, spike propagation, and background activity.

Data were analyzed using Fisher's exact test or the unpaired Student's t-test. A level of P <or=to 0.05 was accepted as statistically significant.

Back to Top | Article Outline


ECoG recordings from 30 patients, 15 patients in each group, were evaluated. Demographic comparisons were similar between the two groups (Table 2). Twenty-seven patients underwent temporal lobectomy, and three patients underwent frontal lobectomy. The management of preoperative anticonvulsant medications was similar between groups. Sixty percent of the patients in the neurolept group and 73% in the propofol group received a reduced dose or no dose of anticonvulsant medication on the day of surgery, having had their anticonvulsant medications tapered, partially or completely, during preoperative evaluations (Table 2). The remainder of the patients in each group received their normal dose of anticonvulsant medications preoperatively. Among the patients for whom information was available (12 in the propofol group, 14 in the neurolept group), the number of patients who experienced one or more seizures during the week preceeding surgery was also similar between groups (Table 2).

Table 2

Table 2

Intraoperative evaluation of the ECoG was considered adequate to proceed with resection for all patients. Blinded evaluation of the ECoG recordings revealed a significant difference between groups only in relation to the frequency of background activity (Table 3). The number of patients with moderate or large amounts of high-frequency background activity, versus those in whom such activity was mild or absent, was higher in the propofol group (n = 11) compared with the neurolept group (n = 1) (P = 0.0002, Fisher's exact test). Spike frequency was similar between groups comparing the number of patients with no spike activity, those with rare or infrequent spike activity, and those with frequent to extremely frequent spike activity. For one patient in each group, rare spike frequency observed on ECoG was attributed to the administration of lorazepam during the evening prior to surgery or to thiopental administered prior to ECoG for an intraoperative seizure. The location and propagation of spike activity were also similar between groups.

Among the 30 patients studied, intraoperative seizures were more common (P = 0.008, Fisher's exact test) in the neurolept group (n = 6) compared with the propofol group (n = 0). Two of these six patients experienced focal motor seizures, which required no intraoperative intervention. The remaining four patients experienced generalized convulsions, and three of these patients received IV thiopental (50- to 75-mg boluses) to terminate the seizures (mean dose 125 mg, range 50-200 mg).

Back to Top | Article Outline


Our results demonstrate a markedly reduced incidence of generalized seizures among the patients who received propofol sedation. Although ECoG activation of epileptiform activity has been reported after the administration of propofol [10,11], our findings support data from several other investigators [5-8] that demonstrate predominant anticonvulsant activity associated with propofol administration.

Among epileptic patients undergoing ECoG with depth electrodes, bolus doses of propofol have been reported to suppress interictal epileptiform activity [5,6]. Using sedative infusions, Samra et al. [7] reported a high degree of individual variability in the ECoG response to propofol, but overall they observed no significant change in epileptogenic activity. Similarly, EEG interictal epileptiform activity was either unchanged or depressed among epileptic patients administered propofol sedation during dental surgery [8]. Propofol has also been used to terminate status epilepticus [12,13].

The purpose of these operations was to identify and remove the area of the brain associated with seizure generation. In this respect, the occurrence of habitual seizures during ECoG recording may expedite localization of the seizure focus. However, the development of intraoperative seizures, particularly generalized convulsions, during other aspects of the operation is generally avoided. Generalized convulsions may lead to an obtunded patient due to either the postictal state or the pharmacological consequences of terminating the seizure. This may increase the operative risk associated with compromised airway integrity, hypoventilation, and impaired patient cooperation. In this study, all patients suffered from intractable epilepsy, and the majority of patients in each group had experienced seizures within the week preceeding the operation. Since the management of preoperative anticonvulsant medications was also similar between the two groups, the results of this study suggest that the incidence of inappropriate intraoperative seizures is lower among patients who receive propofol sedation. Further investigation is warranted to delineate whether this observation represents a protective effect associated with propofol, the absence of protection, or possibly the facilitation of seizures associated with neurolept analgesia. The pro- and anticonvulsant properties of propofol remain controversial [14]; however, at sedative doses, propofol appears to possess predominantly anticonvulsant activity [7,8]. In contrast, although the administration of droperidol has not been reported to be associated with seizure activity during anesthesia for epilepsy surgery, many neuroleptic drugs, including the butyrophenones, have been reported to lower the threshold for seizures, particularly among patients with epilepsy [15,16].

Clearly, the potential benefit associated with a reduction in the incidence of intraoperative generalized convulsions is only appropriate if it is not associated with a detrimental effect on ECoG. Evidence suggests that propofol can suppress or mask epileptiform activity recorded on the ECoG or EEG [4-6]. In the absence of long-term outcome studies comparing propofol and neurolept sedation during epilepsy surgery, such findings raise concern that ECoG suppression may detrimentally affect outcome. Although we have not addressed the issue of outcome directly, in this study, the administration of propofol was suspended 15 minutes prior to ECoG and resulted in no difficulty in demonstrating interictal spike activity or localizing the seizure focus.

Blind ECoG evaluation was based on four criteria: spike frequency, the location of spike activity, spike propagation, and the presence of background activity (Table 2). Spike frequency is probably the most important of these in terms of intraoperative ECoG localization of the seizure focus. Our results demonstrated no relationship between the sedative medication administered and the frequency of ECoG spike activity, spike location, or propagation.

High-frequency beta activity can be caused by a variety of factors, including anesthetic or sedative medications. This activity can also reflect the residual effects of anticonvulsant medications, particularly benzodiazepines, depending on the time elapsed between surgery and the tapering of anticonvulsant medications (which in this study was managed similarly in each group).

beta activity was substantially more prominent in the patients who received propofol, a finding that is consistent with reports that propofol administration is associated with an increase in beta and delta EEG activity at sedative doses [17,18]. Based on processed EEG data obtained during continuous sedative infusions of propofol (at a higher rate of administration than that used in this study), reestablishment of the normal EEG activity spectrum requires approximately 30 minutes after termination of sedation [18]. In our study, based on lower administration rates, the effects of propofol on ECoG background activity were still evident 15 minutes after the termination of administration. This may reflect an inadequate time for the effects of propofol to completely dissipate or an apparent increased sensitivity to propofol associated with ECoG versus scalp EEG recording. However, any residual effect of propofol was insufficient to suppress interictal spike activity. Based on one case report, increased beta activity associated with propofol sedation has been reported to potentially obscure spike activity during seizure surgery [4]. The mild persistent increase in ECoG background activity observed in our study did not detrimentally affect recognition of spike activity.

Although the pro- and anticonvulsant effects remain controversial [14], the use of propofol sedation has become popular during epilepsy surgery. In this study, the use of propofol was associated with a lower incidence of intraoperative generalized seizures. Based on our study protocol, which included suspending the administration of propofol 15 minutes prior to testing, propofol sedation did not adversely affect intraoperative ECoG monitoring compared with traditional neurolept analgesia. Evaluation of long-term outcome after seizure surgery performed under propofol sedation warrants further investigation.

The authors gratefully acknowledge the assistance of Ms. C. Hawke, Ms. L. Szabo (secretarial assistance), and Mr. P. Lok (data analysis) in the preparation of this manuscript.

Back to Top | Article Outline


1. Gignac E, Manninen PH, Gelb AW. Comparison of fentanyl, sufentanil and alfentanil during awake craniotomy for epilepsy. Can J Anaesth 1993;40:421-4.
2. Archer DP, McKenna JMA, Morin L, Ravussin P. Conscious-sedation analgesia during craniotomy for intractable epilepsy: a review of 354 consecutive cases. Can J Anaesth 1988;35:338-44.
3. Silbergeld DL, Mueller WM, Colley PS, et al. Use of propofol (Diprivan) for awake craniotomies: technical note. Surg Neurol 1992;38:271-2.
4. Drummond JC, Iragui-Madoz VJ, Alksne JF, Kalkman CJ. Masking of epileptiform activity by propofol during seizure surgery. Anesthesiology 1992;76:652-4.
5. Rampil IJ, Lopez CE, Laxer KD, Barbaro NM. Propofol sedation may disrupt interictal epileptiform activity from a seizure focus. Anesth Analg 1993;77:1071-3.
6. Ebrahim ZY, Schubert A, Van Ness P, et al. The effect of propofol on the electroencephalogram of patients with epilepsy. Anesth Analg 1994;78:275-9.
7. Samra SK, Sneyd JR, Ross DA, Henry TR. Effects of propofol sedation on seizures and intracranially recorded epileptiform activity in patients with partial epilepsy. Anesthesiology 1995;82:843-51.
8. Oei-Lim VLB, Kalkman CJ, Bouvy-Berends ECM, et al. A comparison of the effects of propofol and nitrous oxide on the electroencephalogram in epileptic patients during conscious sedation for dental procedures. Anesth Analg 1992;75:708-14.
9. Girvin JP. Neurosurgical considerations and general methods for craniotomy under local anesthesia. Int Anesthesiol Clin 1986;24:89-113.
10. Hodkinson BP, Frith RW, Mee EW. Propofol and the electroen-cephalogram. Lancet 1987;2:1518.
11. Smith M, Smith SJ, Scott CA, Harkness WFJ. Activation of the electrocorticogram by propofol during surgery for epilepsy. Br J Anaesth 1996;76:503-7.
12. MacKenzie SJ, Kapadia F, Grant IS. Propofol infusion for control of status epilepticus. Anaesthesia 1990;45:1043-5.
13. Yanny HF, Christmas D. Propofol infusions for status epilepticus. Anaesthesia 1988;43:514.
14. Herrick IA. Seizure activity and anesthetic agents and adjuvants. In: Albin MS, ed. Textbook of neuroanesthesia with neurosurgical and neuroscience perspectives. New York: McGraw Hill, 1997:615-42.
15. Baldessarini RJ. Drugs and the treatment of psychiatric disorders. In: Gilman AG, Goodman LS, Rall TW, Murad F, eds. The pharmacological basis of therapeutics. 7th ed. New York: MacMillan, 1985:396.
16. Canadian Pharmaceutical Association. Inapsine (droperidol) [monograph]. In: Compendium of pharmaceuticals and specialties. 31st ed. Ottawa: Canadian Pharmaceutical Association, 1996:666.
17. Veselis RA, Reinsel RA, Wronski M, et al. EEG and memory effects of low-dose infusions of propofol. Br J Anaesth 1992;69:246-54.
18. Seifert HA, Blouin RT, Conard PF, Gross JB. Sedative doses of propofol increase beta activity of the processed electroencephalogram. Anesth Analg 1993;76:976-8.
© 1997 International Anesthesia Research Society