This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Balakrishnan, G. Articles by Bowyer, S. M. Search for Related Content PubMed Articles by Balakrishnan, G. Articles by Bowyer, S. M. Related Collections Neuroanesthesia Monitoring (Non-cardiac) Pharmacology

---

NEUROSURGICAL ANESTHESIA

A Retrospective Analysis of the Effect of General Anesthetics on the Successful Detection of Interictal Epileptiform Activity in Magnetoencephalography

Guruswamy Balakrishnan, MD^*, Kavita M. Grover, MD, Karen Mason, BS T, Brien Smith, MD, Gregory L. Barkley, MD, Norman Tepley, PhD, and Susan M. Bowyer, PhD

From the Departments of *Anesthesiology, and Neurology, Henry Ford Medical Group, Detroit, Michigan; Department of Neurology, Wayne State University, Detroit, Michigan; and Department of Physics, Oakland University, Rochester, Michigan.

Address correspondence and reprint requests to S. M. Bowyer, PhD, Henry Ford Hospital, Neuromagnetism Lab CFP 78/79, 2799 West Grand Boulevard, Detroit, MI 48202. Address e-mail to sbowyer1{at}hfhs.org.

	Abstract

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
BACKGROUND: A magnetoencephalography (MEG) study requires the patient to lie still for a prolonged period of time. In children and uncooperative adults with epilepsy, general anesthesia or sedation may be required to insure a good quality study. As general anesthetics have anticonvulsant and proconvulsant properties, we investigated whether the use of anesthesia reduced the successful detection of interictal epilepsy activity.

METHODS: MEG testing was performed on 41 epilepsy patients (10 women, 31 men; 1–48 yr) while anesthetized. To determine the impact of anesthesia on the identification of epileptiform activity, the anesthesia group of patients was compared with all other patients with epilepsy who were recorded in our laboratory without anesthesia, as well as with a subgroup of children with epilepsy who were able to be recorded without the need for anesthesia.

RESULTS: Propofol was used in 38 patients, etomidate in two, and one received sevoflurane. Twenty-nine (71%) were found to have interictal epileptiform activity in their MEG results. The percentage of MEG studies with a positive yield for interictal epileptiform activity is comparable with the percentage (63%) found in the patients with epilepsy undergoing MEG without anesthesia. In the 38 children younger than 18 yr, 28 (74%) had interictal epileptiform activity compared with 80% done without anesthesia.

CONCLUSION: We conclude that levels of anesthesia needed to provide unconsciousness and immobility during MEG studies do not significantly alter the likelihood of recording interictal epileptiform spike activity with MEG.

	Introduction

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
Magnetoencephalography (MEG) (1,2) is a noninvasive functional imaging technique currently used clinically for presurgical localization of epileptic tissue. MEG epilepsy source localization of cortical regions helps neurologists and neurosurgeons to determine whether the source of epileptic activity can be surgically removed safely, and to plan the safest surgical approach to the epileptogenic area (3). MEG is also used to estimate the location of brain lesions in relation to the brain’s motor and sensory centers and to assess the possibility of postoperative neurologic deficits. The MEG technique can detect interictal spikes and, by means of a single equivalent current dipole (ECD) modeling algorithm, can map the sources of spike activity onto the patient’s magnetic resonance imaging (MRI) scan (3–6).

A MEG study requires the patient to lie still for a prolonged period (30–60 min) in a magnetically shielded room (MSR). The MSR housing the biomagnetometer is designed to attenuate ambient magnetic field and radiofrequency interference. In children and uncooperative adults with epilepsy, general anesthesia or sedation may be required to reduce movements and insure a good quality MEG study. General anesthetics have both anticonvulsant and proconvulsant properties (7,8). Thiopental and midazolam are used as potent anticonvulsants to abort status epilepticus. Methohexital (Brevital) and etomidate in small doses are useful to activate epileptiform foci during epilepsy surgery (9,10). Inhaled anesthetics are predominantly anticonvulsants with low potential for proconvulsant properties except for enflurane, which causes epileptiform activity on the electroencephalogram (EEG) at high alveolar concentrations and in hypocapnic conditions (11).

When used as anesthetics during MEG studies of epileptic patients, these drugs must not excessively suppress interictal epileptiform activity, induce epileptiform activity, or expand the region of existing epileptiform foci. This retrospective study tested the effect of anesthetics on 41 patients who underwent MEG studies while anesthetized to determine if use of anesthesia in epilepsy patients alters the detection of interictal epileptic activity.

	METHODS

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
Subjects
In this IRB approved study, 41 patients (31 men; 10 women) with medically refractory partial epilepsy (age range 1–48 yr; average age 9.9 ± 9.6 yr) were monitored with 148 MEG channels (Magnes WH2500, 4D Neuroimaging), and 32 EEG channels using the 10–20 system of electrode placement, as well as one channel of electrocardiogram (ECG) (Synamps, Neuroscan). These patients were selected for anesthesia because they were unable to lie still for 10 min. All patients continued to receive their anticonvulsant medications at the time of the study.

Patients scheduled for a MEG study received routine preanesthetic evaluation and pretreatment of the skin on the dorsum of both hands with lidocaine–prilocaine cream (EMLA cream 5%) to minimize pain on IV catheter placement. Many of these patients were also scheduled to have brain MRI scans performed under sedation before MEG study. If the patient was assessed to be anxious and uncooperative, oral midazolam 0.25–0.5 mg/kg was administered to a maximum of 15 mg regardless of total body weight. IV access was secured and atropine sulfate 0.02 mg/kg IV was administered as an antisialogogue to decrease airway secretions. Each subject arrived at the MEG lab in a hospital gown; all metal articles removed from his/her body, except for dental work, which was demagnetized with a commercial videotape eraser after sedation. The time between midazolam and MEG recordings was well over 1 h and increased to 2–3 h if an MRI was performed before coming to the MEG lab.

General anesthesia was induced with ketamine 0.5–1 mg/kg (to avoid significant pain associated with injection of propofol into small peripheral veins in awake children) and maintained by using propofol infusion titrated at 50–150 mg kg^–1 min^–1 in 38 patients. Oxygen administration and capnography monitoring were accomplished by using nasal cannula. A noninvasive automatic blood pressure cuff and pulse oximeter probe were placed on a lower limb, away from the MEG sensor. All patients maintained spontaneous respiration. Two patients needed laryngeal masks for airway support. Sevoflurane was used in one patient with a laryngeal mask for airway maintenance. Etomidate (0.3 mg IV bolus followed with 5–10 mcg kg^–1 min^–1 for maintenance) was used in two patients. After a prolonged recording of normal MEG activity, a bolus of methohexital 0.25 mg/kg was administered in one patient (Patient #32) who was receiving propofol maintenance in an attempt to activate interictal epileptiform activity during methohexital washout.

After the patients were sedated, gold EEG electrodes were attached to the scalp and held in place with collodion. An ECG electrode was also attached to each patient’s left clavicle. Three small coils, used to transmit subject location information to the neuromagnetometer probe, were taped to the patient’s forehead with two-sided tape. One coil was located in the center of the forehead and one coil on each side, approximately 2 cm apart. Disposable ear molds of the correct size were placed in the ears and an additional localization coil was attached to each ear mold. The patient was moved into the MSR and transferred onto the bed. The patient then lay comfortably on the bed inside of the MSR, and automatic probe position routines were used to locate the head with respect to the neuromagnetometer detector coils. The head shape was also digitized for coregistration to the patient’s MRI. The neuromagnetometer helmet containing the detector array was placed over the patient’s head in close proximity to most of the cortical surface. Changes in the patient’s position during the study were detected by changes in magnetic field locations arising from the coils on the forehead and ears. Runs during which the subject moved more than 0.8 cm were repeated. During three or more 10-min data acquisition periods the anesthesiologist titrated the anesthetic drugs and monitored the patient from outside the MSR.

Standard anesthesia machine and monitoring equipment were placed outside the MSR and extended breathing circuits, IV lines and monitoring equipment were passed through a porthole in the MSR wall to the patient. After the patient’s vital signs were stable, the door to the MSR was closed and data were collected in 10 min segments. The patient’s EEG, ECG, arterial blood pressure, end-tidal CO₂, and temperature were continuously monitored during the MEG testing.

Data were digitized at 508 samples per s from 0.01 Hz to 100 Hz. Three or more 10-min continuous acquisition segments were collected. During these data collection segments the anesthesiologist monitored the patient’s vital signs outside the MSR. The anesthesiologist entered the room between data collection runs as needed to check the patient. The patient was also monitored by two-way intercom and a video monitor. In some cases, the anesthesiologist or parent, after removing all metal objects, sat quietly in the corner of the MSR during the 10-min data collection segments. There were no significant hemodynamic or respiratory complications during these procedures. No patient experienced status epilepticus or convulsive seizure. At the conclusion of the MEG study, patients were monitored in the postanesthesia care unit and were discharged on the same day after meeting criteria for outpatient discharge.

MEG Data Analysis
MEG data were forward and backward filtered using a 3–100 Hz bandpass with a 60 Hz notch filter. All data were visually inspected for epileptic spikes. Two neurophysiologists (BJS, GLB) and a Registered EEG Technologist (KM) reviewed the raw data. For each spike, a small interval of time [<500 ms] encompassing an epileptic spike was selected and a single ECD fit (6) was performed using the Magnetic Source Imaging algorithm (12). The study was considered a success if a single ECD met our selection criteria [Correlation Coefficient 0.98, Goodness of Fit 0.95, Root mean square >400 fT, and Dipole moment >200 nAm, and Confidence volume <1.0 cm³] (6). ECD fits were displayed on the patient’s MRI data set.

Statistical Methods
Comparisons of detection of epileptiform activity within the patients given anesthesia were done using two sample t-tests and ² tests. Similar analyses were done to compare patients with and without anesthesia.

	RESULTS

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
Of all 41 patients who underwent MEG studies while anesthetized, 29 (71%) had epileptiform activity detected and coregistered onto their MRI scans. Of the 37 patients who received propofol for maintenance of anesthesia and then underwent MEG studies, 26 patients (70%) had epileptiform activity present in their MEG recordings. Table 1 shows the patient population characteristics and the number of epileptiform interictal spikes that were found. Epileptiform activity was detected in the MEG studies in the two patients in whom etomidate was used and the one in whom sevoflurane was used. Of 10 female patients, 9 (90%) had epileptiform activity detected. Epileptiform activity was detected in 20 of the 31 male patients (65%). The mean age for patients with epileptiform activity was 10.9 yr (sd = 11.0, range 1–48) and for patients without epileptiform activity the mean age was 7.4 yr (sd = 4.0, range = 1–16). Our results show no statistical differences based on age (P = 0.144) or gender (P = 0.123). Also, of 38 patients younger than 18 yr, 28 (74%) had epileptiform activity identified.

Twenty-two patients did not have unanesthetized EEG recorded in our institution because they were referred for the MEG study from other epilepsy centers across the country. For patients who had prior EEGs performed without anesthesia in our institution, 5 of 19 patients had interictal epileptiform activity detected with the unanesthetized EEG but not with the anesthetized EEG recorded simultaneously with the MEG. This may, in part, be due to the fact that interictal epileptiform discharges occur intermittently, and may not be detected in every recording session. In addition, in some of the patients, the recording time for the unanesthetized EEG was far longer as these patients were admitted in the epilepsy monitoring unit and video EEG recordings were performed continuously for several days. In contrast, in two patients, interictal epileptiform activity was seen only in the EEG performed under anesthesia and not in the previous EEG recordings performed without anesthesia. For technical reasons, an EEG tracing under anesthesia could not be obtained along with the MEG inpatient # 7.

There were three patients in whom interictal epileptiform discharges were seen in the MEG recording but not in the simultaneously recorded EEG and one patient in whom interictal epileptiform discharges were seen in the EEG recordings but not in the MEG recordings. This discrepancy may have been due to MEG’s sensitivity to tangential dipoles and insensitivity to radial dipoles. In addition, MEG fields fall off with distance from the source. Thus the EEG may be a positive when only radial sources are present and/or if the epileptiform discharges arise from a deep source such as the mesial temporal lobe. MEG is more sensitive than EEG for superficial cortical sources and discharges arising in the insular region. As is the case, in our series, most patients have epileptiform discharges detected simultaneously by both methodologies.

To determine the impact of anesthesia on the identification of epileptiform activity, the anesthesia group of patients was compared with all other patients with epilepsy who were recorded in our lab without anesthesia, as well as with a subgroup of children younger than 18 yr with epilepsy who were able to be recorded without the need for anesthesia. This subgroup more closely matched the group of patients recorded under anesthesia than the largely adult population in the overall group. In our laboratory at Henry Ford Hospital, of 382 patients with epilepsy who have had MEG studies not requiring anesthesia, 240 (63%) had epileptiform activity detected by MEG. Considering only the patients with epilepsy younger than 18 yr, of the 60 (31 women, 29 men) recorded without anesthesia, 48 patients (25 women and 23 men) (80%) had epileptiform activity detected. Thus, no difference in the detection of epileptiform activity was observed between patients younger than 18 yr when recorded with and without anesthesia (74% vs 80%. P = 0.465).

	DISCUSSION

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
In our study, the comparison of positive MEG results between patients who received general anesthesia (71%) with those in whom no anesthesia (63%) was administered showed no statistical difference (P = 0.317). The percentage of patients with identifiable epileptiform activity in MEG recordings performed with anesthesia was similar to the percentage of patients with epileptiform activity in MEG recordings of a group of nonanesthetized children as well as a group of nonanesthetized adults. Therefore, we believe anesthesia did not significantly alter the number of patients who demonstrated interictal epileptiform activity. These results are similar to those reported by Szmuk et al. (13), and Stefan et al. (14). Szmuk et al. (13) showed the average sensitivity of MEG to detect epileptiform activity with midazolam, sevoflurane, and propofol anesthesia was 75% in young children (5.7 ± 4.3 yr). Stefan et al. (14) found a 76% success rate of detecting epileptiform activity on MEG in a group of 300 unanesthetized patients ranging in age from 7–79 yr.

This study does not conclusively answer the question of whether anesthesia increases or decreases the detection of interictal epileptiform discharges with MEG, since some patients only had MEG recordings performed under anesthesia. The best way to determine the effects of anesthesia would be to do two studies on the same patient; one study with anesthesia and one without. Since this was not feasible, we compared patients recorded under anesthesia to a larger population of patients with epilepsy of all ages, as well as a subgroup of children with epilepsy, and found that the results in each group were similar. We found successful detection of interictal epileptiform activity in 71% of the patients who received general anesthesia compared to 80% success where no anesthesia was used.

Though the medical literature on the effects of anesthetics on neurophysiological monitoring, EEG, and interictal spikes is extensive, it is at times confusing and contradictory, since the action of these drugs depends on the dose and the physiological state of the subject. Most general anesthetics have anticonvulsant properties and, under some conditions, many are also proconvulsants. Unlike their anticonvulsant effects, which are predictable, their proconvulsant neuroexcitatory effects are not, as evidenced by sporadic case reports of their occurrence in patients not known to have epilepsy, as well as in patients with preexisting epilepsy (8,15–19).

Barbiturates are clinically proven potent anticonvulsants and are often used in anesthetic doses to treat patients with refractory status epilepticus. Epileptiform activity is often increased during abrupt withdrawal of barbiturates. This phenomenon has been used clinically to enhance epileptiform activity in presurgical localization of epileptic foci by administering short-acting barbiturates. The short-acting barbiturate usually used in this manner is methohexital for a procedure known as the "methohexital suppression test." However, methohexital, an oxybarbituate, causes epileptiform EEG activity in susceptible patients when given in doses of <0.5 mg/kg. Methohexital has been used to activate interictal epileptiform activity on intraoperative electrocorticography during epilepsy surgery (9). In our study, we attempted to see if methohexital would be effective in causing epileptiform activity in one patient. Though it was unsuccessful in this patient we are contemplating adding this to our procedure if there has been no interictal activity after 1 h.

Propofol is a non-barbiturate induction drug with both neuroexcitatory and anticonvulsant properties. Propofol has been reported to cause tremors, myoclonus, dystonic posturing and tonic-clonic convulsions in both epileptic and nonepileptic patients (17,20). A sedative infusion of propofol has been successfully used to record interictal epileptiform spikes during intraoperative electrocorticography for epilepsy surgery (21,22). Continuous infusion of propofol has been shown to be effective in the treatment of refractory status epilepticus (23). In our study group, continuous IV infusion of propofol at 50–150 mg kg^–1 min^–1 was administered for maintenance of anesthesia during MEG recordings. To avoid significant pain associated with injection of propofol into small peripheral veins in awake children, ketamine was used to induce anesthesia before infusion of propofol began.

Etomidate (non-barbiturate) has more neuroexcitatory properties than propofol. It has been reported to cause EEG-documented seizure activity and also produces myoclonic movements that could be mistaken for seizures in a significant percentage of patients (17). Etomidate is also used intraoperatively to activate epileptic foci and to augment somotosensory evoked potentials (19). In the two patients in this study in whom etomidate sedation was used, epileptiform activity was detected in both the MEG and the EEG.

Sevoflurane, like other inhaled anesthetics is a potent anticonvulsant at anesthetic concentrations. Unlike enflurane, it does not have significant proconvulsant effects in epileptic patients (8). In one patient who was anesthetized with sevoflurane because he had soy allergies [Propofol preparations contain soybean oil and egg lecithin (24)], epileptiform activity was detected.

In this report, we retrospectively analyzed the feasibility of using anesthetics in children and uncooperative adults undergoing MEG study. Deep levels of anesthesia suppress seizures. The concern about the use of anesthesia has been whether an anesthetic to provide immobility might also inhibit epileptiform discharges in epileptic patients. We have shown in this study that we can record epileptiform discharges in patients under anesthesia with detection rates similar to that in patients recorded without anesthesia. We conclude that the use of anesthesia to provide unconsciousness and immobility during MEG does not significantly alter the success of detecting interictal epileptiform activity in patients with medically intractable epilepsy.

	Footnotes

 
Accepted for publication March 7, 2007.

Supported in part by NIH/NINDS Grant R01 NS30914.

	REFERENCES

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 

1. Hamalainen M, Hari R, Ilmoniemi RJ, et al. Magnetoencephalography-theory, instrumentation, and applications to noninvasive studies of the working human brain. Rev Modern Phys 1993;65:413–97.[Web of Science]
2. Hari R. Magnetoencephalography as a tool of clinical neurophysiology. In: Niedermeyer E, Lopes da Silva F, eds. Electroencephalography: basic principles, clinical applications and related fields. Baltimore, Maryland: Williams & Wilkins, 1998;1107–34.
3. Ebersole JS. Magnetoencephalography/magnetic source imaging in the assessment of patients with epilepsy. Epilepsia 1997;38:S1–S5.
4. Ebersole JS. EEG/MEG dipole modeling. In: Engle J, Pedley T, eds. Comprehensive epileptology. Philadelphia: Lippincott-Raven, 1997:919–36.
5. Barkley GL, Baumgartner C. MEG and EEG in epilepsy. J Clin Neurophysiol 2003;20:163–78.[Web of Science][Medline]
6. Bowyer SM, Mason KM, Tepley N, et al. MEG validation parameters for clinical evaluation of interictal epileptic activity. J Clin Neurophysiol 2003;20:87–93.[Web of Science][Medline]
7. Modica PA, Tempelhoff R, White PF. Pro- and anticonvulsant effects of anesthetics (Part II). Anesth Analg 1990;70:433–44.[Free Full Text]
8. Modica PA, Tempelhoff R, White PF. Pro- and anticonvulsant effects of anesthetics (Part I). Anesth Analg 1990;70:303–15.[Free Full Text]
9. Gumpert J, Paul R. Activation of the electroencephalogram with intravenous Brietal (methohexitone): the findings in 100 cases. J Neurol Neurosurg Psychiatry 1971;34:646–8.[Abstract/Free Full Text]
10. Gancher S, Laxer KD, Krieger W. Activation of epileptogenic activity by etomidate. Anesthesiology 1984;61:616–18.[Web of Science][Medline]
11. Lebowitz MH, Blitt CD, Dillon JB. Enflurane-induced central nervous system excitation and its relation to carbon dioxide tension. Anesth Analg 1972;51:355–63.[Free Full Text]
12. Magnes software, 4-D Neuroimaging, San Diego, CA.
13. Szmuk P, Kee S, Pivalizza EG, et al. Anaesthesia for magnetoencephalography in children with intractable seizures. Paediatr Anaesth 2003;13:811–17.[Web of Science][Medline]
14. Stefan H, Hummel C, Genow A, et al. Magnetic brain source imaging of focal epileptic activity: a synopsis of 455 cases. Brain 2003;126:2396–405.[Abstract/Free Full Text]
15. Schuh LA, Drury I. Intraoperative electrocorticography and direct cortical electrical stimulation. Semin Anesth 1997;16:46–55.[Medline]
16. Thompson GE. Ketamine induced convulsions. Anesthesiology 1972;37:662–3.[Medline]
17. DeFriez CB, Wong HC. Seizures and opisthotonus after propofol anesthesia. Anesth Analg 1992;75:630–2.[Free Full Text]
18. Hansen HC, Drenck NE. Generalized seizures after etomidate. Anaesthesia 1988;43:805–6.[Web of Science][Medline]
19. Gancher S, Laxen KD, Krieger W. Activation of epileptogenic activity by etomidate. Anesthesiology 1984;61:616–18.[Web of Science][Medline]
20. Collier C, Kelly K. Propofol and convulsions—The evidence mounts. Anaesth Intensive Care 1991;19:573–5.[Web of Science][Medline]
21. Herrick IA, Craen RA, Gelb AW, et al. Propofol sedation during awake craniotomy for seizures. Anesth Analg 1997;84:1280–4.[Abstract]
22. Ebrahim ZY, Schubert A, Van Ness P, et al. The effect of propofol on electroencephalogram of patients with epilepsy. Anesth Analg 1994;78:275–9.[Web of Science][Medline]
23. Claassen J, Hvisch LJ, Emerson RG, Mayer SA. Treatment of refractory status epilepticus with pentobarbital, propofol or midazolam. A systemic review. Epilepsia 2002;43:146–53.[Web of Science][Medline]
24. Marik PE. Propofol: therapeutic indications and side-effects. Curr Pharm Des 2004;10:3639–49.[Web of Science][Medline]

This article has been cited by other articles:

				E.S. Schwartz, D.J. Dlugos, P.B. Storm, J. Dell, R. Magee, T.P. Flynn, D.M. Zarnow, R.A. Zimmerman, and T.P.L. Roberts Magnetoencephalography for Pediatric Epilepsy: How We Do It AJNR Am. J. Neuroradiol., May 1, 2008; 29(5): 832 - 837. [Abstract] [Full Text] [PDF]

				G. Balakrishnan, K. M. Grover, K. Mason, B. Smith, G. L. Barkley, S. M. Bowyer, and N. Tepley Effect of Propofol on Interictal Epileptiform Activity Anesth. Analg., November 1, 2007; 105(5): 1510 - 1510. [Full Text] [PDF]

				S. Meyer, U. Grundmann, G. Shamdeen, S. Gottschling, and L. Gortner Effect of Propofol on Interictal Epileptiform Activity Anesth. Analg., November 1, 2007; 105(5): 1509 - 1510. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Balakrishnan, G. Articles by Bowyer, S. M. Search for Related Content PubMed Articles by Balakrishnan, G. Articles by Bowyer, S. M. Related Collections Neuroanesthesia Monitoring (Non-cardiac) Pharmacology