Epileptiform and non-epileptiform paroxysmal activity from isolated cortex after functional hemispherectomy

ELSEVIER

Epileptiform

Electroencephalography

and clinical Neurophysiology

102

( 1997) 437-441

and non-epileptiform paroxysmal activity from isolated cortex after functional hemispherectomy ’ R.A. Wennberg, Montrwl

Neuro/o,~icctl

lmtitute

L.F. Quesney*,

md Hospitrtl.

Accepted

3X01 Clnirvr,sitv

for publication:

J.-G. Villemure St.. Morrtrunl.

9 January

Qwtw~~. H3A IH4.

(;rmtd~r

1997

Abstract The relationship of acute complete cortical isolation to paroxysmal cerebral activity was examined in 16 patients with electrocorticography (ECOG) before and after functional hemispherectomy (FH). Burst-suppression activity appeared over isolated cortex in all cases, the severity of which could be increased by systemic administration of propofol or methohexital. Interictal epileptiform activity (EA) recorded from frontal or parietal-occipital cortex before FH invariably persisted after FH ( I I cases). No EA was recorded before or after FH in 3 cases while in 2 cases EA appeared following FH which had not been present before FH. The intensity of burst-suppression activity was not related to the presence or absence of post-excision EA. In total, 30 disconnected cortices were recorded from; relative abundance of EA was increased in IO, unchanged in 17, and decreased in 3 cases. Unrelated to the induction of burst-suppression activity, cortical isolation may decrease the threshold for expression of interictal EA. 0 1997 Elsevier Science Ireland Ltd. Kq\‘\z~o&: Burst-suppression:

EJectrocorticography

(ECOG); Epilepsy:

1. Introduction Electrocorticographic recordings from isolated cortex in patients undergoing different surgical procedures have been reported to show alterations in cerebral electrical activity following disconnection. Reports have described burst-suppression activity in non-epileptic patients (Echlin et al.. 1952; Henry and Scoville, 1952) and increased interictal epileptiform activity (EA) in epileptic patients (Niemeyer. 1958; Polkey et al., 1989; Cendes et al., 1993). Differentiation between the two forms of paroxysmal activity was not described in the latter papers though a recent report suggests that the two exist independently of each other (Wennberg et al., 1995). The surgical technique of functional hemispherectomy (FH) was introduced by Rasmussen in 1974 to avoid the late complication of superficial cerebral hemosiderosis seen to follow earlier ‘anatomical’ hemispherectomies (Rasmussen, 1983; Tinuper et al.. 1988: Villemure. 1992).

* Corresponding author. Tel.: +I 514 3981995; fax: +I 514 3988540. ’This work was presented in part at the 31st meeting of the Canadian Congress

of Neurological

Sciences.

London. ON, June 1996.

00 13-1694/97/X 17.00 0 I Y97 Elm ier Science Ireland Ltd. All right\ reserved PI/ SOY? I -XXIS(Y7 lY60~7m 1

Epileptiform

activity; Functional

hemispherectomj

The technique leaves disconnected frontal and parietal-occipita1 lobes in place after resection of the central region and temporal lobectomy (Villemure. 1992). Clinical indications for FH (comprising intractable partial epilepsy and contralateral hemiparesis) and post-operative results have been discussed previously (Tinuper et al.. 1988: Smith et al., 1991; Andermann, 1992; Villemure et al., 1993) as have the pre- and post-operative EEG findings (Smith et al., 1991: Pierre-Louis and Morrell. 1992: Carment et al., 1995). In this report. the results of electrocorticography (ECOG) performed before and after FH in 16 patients are presented and discussed with respect to the effects of acute complete cortical isolation upon epileptiform and non-epileptiform cerebral potentials.

2. Materials

and methods

From 1987 through 1993 pre- and post-excision ECOG was performed on 16 patients who underwent FH. All patients had intractable epilepsy and contralateral hemiparesis prior to surgery. Etiologies were remote vascular lesion

El% Y60-17

(9), chronic (Rasmussen’s) encephalitis (5) Sturge-Weber syndrome (1) and remote traumatic lesion (1). Two patients had undergone previous surgery for epilepsy (anterior callosotomy in one patient (patient 3) and anterior callosotomy + anterior temporal and frontal resection in the other (patient 4)). A summary of patients is given in Table 1. ECOG technique has been described previously (Gloor, 1975). Recordings were obtained under light general anesthesia (nitrous oxide supplemented with fentanyl and droperidol) and averaged 1O-l 5 min in duration. Methohexital (cases 2, 6, 10 and 11) or propofol (cases 4, 9 and 12) activation was performed inconstantly and did not induce the presence of EA in any of these cases. Only the ECOG record obtained prior to activation was used for quantitative analysis of EA. All patients had pre-excision recordings over parietaloccipital cortex; 14/16 patients had pre-excision recordings over frontal cortex. EA (sharp waves, spikes and multiple spikes) was defined using the same electrographic criteria outlined for scalp EEG by Gloor (1975). Presence or absence of post-excision EA recorded from isolated parietal-occipital and frontal cortices was documented. Relative abundance of EA was arbitrarily classified into four groups of increasing spike discharge frequency based on visual analysis of representative 60 s pre- and post-excision epochs (Group A. 16 spikes mini’; Group B, >6-12 spikes mini’; Group C, > 12-24 spikes mini’; Group D, >24 spikes mini’ (51. >I-2, >2-4, and >4 ‘spikes-per-page’. respectively)). EA recorded from central and temporal cortex before FH and from insular cortex after FH was noted but not quantitated for the purposes of this study. Burst-suppression was identified in the post-excision tracings as a paroxysmal pattern of burst periods (l-6 s of

Table 7 ECOG tindings bet’orc and after FH Patient

I 2 3 3 5 6 7 8 9

Pre-FH EA

Poht-FH burstwpprc\ion

Post-FH EA

Frontal

Parielaloccipital

Frontal

Parietaloccipital

D A C D D D

D _

A C D C C D

D

c A A D _ _

C B B D _ _

B D

C

I0

C _

II I.! 13 1-I IS 16

C D _

No record No record A _ _

D B B _ _

No record No record C c

D D B c c

Sevcrc Mild Moderate Scwrc Scwrc Mild Modcratc Modcrate Modcrate Mild MI111 Mild Mild Mild :Mild Sc\crc

EA classified by spike discharge frequency: A. 56 aplkes mln ‘: B. ~6-12 spikes min-‘; C, >I?-?4 spikes min-‘: D. >24 spikes min-‘: -. no EA.

mixed frequency medium-high amplitude slow activity and low-medium amplitude fast activity) alternating with l-8 s periods of suppression of cerebral activity. It was interpreted according to the electrical activity present during the suppressed periods as mild (sustained lowmedium amplitude), moderate (unsustained low amplitude) or severe (absence of discernable cerebral electrical activity).

3. Results Table I Summary

of patients

Patient

Age

MRIKT

imaging

I 2 3 .I 5 6 7

19 24 6 12 8 5 7

8 9 10 II 12

11 17 10 36 33

13 14

IS 34

15 16

4 19

Left hemisphere atrophy Left perisylvian infarct Right hemisphere atrophy Right frontal/temporal excision Left hemisphere atrophy Right hemisphere atrophy Right hemisphere atrophy + right occipital calcification Left encephalomalacia/gliosis Right hemisphere atrophy Left hemisphere atrophy/gliosis Right hemisphere atrophy Right hemisphere atrophy + frontotemporal porencephaly Left perisylvian infarct Right hemisphere atrophy + frontocentral porencephaly Right hemisphere atrophy Left perisylvian porencephaly

Etiology

Hemiparesis

Vascular Vascular Encephalitis Encephalitis Encephalitis Vascular Sturge-Weber

Right Right Left Left Right Left Left

Traumatic Vascular Encephalitis Vascular Vascular

Right Left Right Left Left

Vascular Vascular

Right Left

Encephalitis Vascular

Left Right

Table 2 summarizes the pre- and post-excision EA recorded from frontal and parietal-occipital cortices before and after FH. Seven patients had pre-excision EA recorded independently from frontal cortex and parietal-occipital cortex; all 7 of these had EA recorded from isolated frontal and parietal-occipital cortices following FH (Fig. 1). Two cases had EA restricted to frontal cortex before and after FH. Two cases with no pre-excision recording from frontal cortex had EA recorded from parietal-occipital cortex before and after FH. Five patients showed no frontal or parietal-occipital EA before FH; 3 of these showed no EA following FH, I showed EA restricted to isolated parietal-occipital cortex, and 1 had EA recorded from isolated frontal and parietaloccipital cortices. Relative abundance of pre- and post-excision EA recorded from frontal and parietal-occipital cortices was classified as described in Section 2. In total, 30 disconnected cortices were recorded from; relative abundance of EA was increased in 10. unchanged in 17, and decreased in 3 cases

Fig.

I. Patwnt

9. Top: pre-excision

showed EA recorded

l’rom i\olatcd

ECOG frontal

showed

independent

and parirtal-occipital

EA recorded cortices

(Table 2). There was no evident correlation between etiology (i.e. remote vascular lesion or chronic encephalitis) and the ECOG response to cortical disconnection (cf. Tables 1 and 2). In all cases a burst-suppression pattern (not present before FH) was recorded from isolated cortex in the post-excision ECOG. This was interpreted as mild, moderate or severe (Table 2; Fig. 2). There was no correlation between the intensity of burst-suppression activity and the presence 01 absence of post-excision EA The administration of 120-- 130 mg propofol to 3 patients (cases 4. 9 and 12) during the post-excision recording clearly increased the severity of burst-suppression recorded from the isolated cortices (Fiy. 3). Administration of 20-25

Fig. 2. Patlent J (lop) and patient bur+suppression

pattern.

5 (bottom):

post-excision

ECOG

from

trontal

and a moderate

and parietal-occipital

bur\t-\uppres\ion

corticeh.

Bottom:

post-excision

ECOG

pattern.

mg methohexital to 2 patients (cases 2 and 10) following FH induced a minimal increase in burst-suppression activity recorded from isolated cortex. Though not a focus of this study, EA recorded from central and/or temporal cortex before FH and from insular cortex after FH was noted. Ten of sixteen patients had preexcision EA recorded from (subsequently resected) central and/or temporal cortex before FH. Post-excision recording from insular cortex was performed in IX16 patients; EA was recorded in S/13 and burst-suppression (usually mild) in 9/l 3 patients. In the 2 patients with post-excision insular recordings who were given propofol, burst-suppression was induced in 1 (case 4) and increased in the other (case 9) over the insult.

showed EA recorded from isolated frontal

and parietnl-occipital

cortu\

and ;L severe

440

Fig, 3. Patient 9 (top) and p&ient 4 (bottom): post-excision f&n

isolated co&es

(cf.. Figs.

ECOG following

propofol administration

showed incrcasrd

severity of burst-suppression

recorded

I and 3).

4. Discussion Reports of paroxysmal cerebral activity recorded from acutely isolated cortex in humans have appeared sporadically in the literature over the past half-century. Henry and Scoville (1952) demonstrated the induction of ‘suppressionburst’ activity immediately following partial isolation of cortex during various lobotomy or leucotomy procedures. Other studies performed on animals and human patients with and without epilepsy corroborated these findings and suggested a possible relationship between cortical isolation and partial epilepsy (Dusser de Barenne and Marshall, 193 1; Burns, 1951; Echlin et al., 1952; Echlin, 1959: Echlin and Battista, 1963; Fisher-Williams, 1963: Rutledge et al., 1967) although induction of spontaneous EA appeared dependent on chronic cortical isolation, with acute cortical disconnection producing only burst-suppression activity (Bremer, 1935; Bremer. 1938; Kristiansen and Courtois, 1949; Echlin et al., 1952; Henry and Scoville, 1952; Grafstein and Sastry, 1957; Sharpless and Halpern, 1962; Spehlmann et al., 1971). Acutely increased EA has been described to follow surgical procedures for temporal lobe epilepsy that result in partially disconnected temporal neocortex (Niemeyer, 1958: Echlin, 1959: Polkey et al., 1989: Cendes et al., 1993; Wennberg et al., 1995). However. the described activation of EA is not seen in all cases (Polkey et al., 1989; Wennberg et al., 1995) and appears limited to patients with pre-excision neocortical EA (Wennberg et al., 1995). The specificity of this activation phenomenon for epileptogenic cortex parallels that seen following the administration of methohexital (Hufnagel et al., 1992) and is in direct contrast to the non-specific induction of burst-suppression activity seen in both epileptogenic and non-epileptogenic cortex

after either acute partial cortical disconnection or methohexital administration (Hufnagel et al., 1992; Wennberg et al., 1995). The concurrent appearance of both forms of paroxysmal activity in epileptogenic cortex was termed ‘spikeburst-suppression’ by Hufnagel et al. (1992) and discussed further by Jantti et al. ( 1994). That either type of paroxysmal activity recorded after surgical isolation be due to ‘injury potentials’ or ischemia related to the various disconnection techniques rather than to the cortical isolation per se has been adequately discounted in the past (Echlin et al., 1952; Henry and Scoville. 1952; Echlin. 1959). FH results in complete surgical disconnection of thalamocortical and basogangliocortical brain circuitry and ECOG records obtained from the isolated frontal and parietal-occipital lobes are free from all ascending (or reverberating) subcortical influence. Significant corticocortical connectivity is presumably left intact in the isolated blocs after FH. lnsular recordings post-FH are from partially isolated cortex, i.e. isolated from corticocortical connections but with retained subcortical connectivity (Villemure and Mascott, 1995). In this study, the acute electrographic effects of complete cortical isolation were examined by comparing the paroxysmal activity recorded before FH (i.e. EA) with the paroxysmal activity recorded from isolated cortex after FH, both epileptiform and non-epileptiform (i.e. burst-suppression). Burst-suppression activity was recorded from all isolated cortical blocs following FH, though the severity ranged widely between patients. from very mild (requiring comparison with the pre-excision tracing to identify) to severe. unrelated to the pathologic etiology. Depth of anesthesia was similar in all patients and thus unlikely to explain the differing severities of burst-suppression, a point further supported by the absence of burst-suppression over the insula

(prior to propofol bolus) in the presence of severe burstsuppression over isolated frontal and parietal-occipital cortices in patient 4 (see Section 3). One hypothesis would be that the severity of burst-suppression recorded over isolated cortex is an indicator of the degree of persistent corticocortical connectivity in the isolated bloc. It has been proposed, based on experimental results of intra-cellular neuronal recordings in anesthetized cats, that cortical burst-suppression arises from complete disconnection of thalamocortical circuitry (Steriade et al., 1994). a condition replicated by FH. However, the observation in this study that burst-suppression can be increased in isolated cortical blocs hy the systemi; administration of anesthetic agents (propofol and methohexital) known to induce burstsuppression in the intact cerebrum (Musella et al., 1971: Ebrahim et al.. 1994) clearly implicates a cortical level mechanism in the modulation of burst-suppression. Interictal EA recorded from frontal or parietal-occipital cortex before FH invariably persisted after FH ( 1 l/l6 cases). In 2 cases EA appeared following FH which had not been present before FH. Previous work has demonstrated the ability of pre-excision chemical activation with methohexital to predict the post-excision induction of EA over disconnected epileptopenic cortex (Wennberg et al., I995 1: unfortunately chemical activation was not performed prior to FH in the latter 2 cases (patients 15 and 16). There was no relationship noted between the intensity of burstsuppression and the present-_ or absence of post-excision EA. The classitication system used to quantitate pre- and postexcision EA was devised as an attempt to provide some quantitation to a phenomenon most readily appreciated by the electroencephalographer viewing the ECOG record. The schema facilitates conceptualization of EA abundance on a ‘spike-per-page’ basis. Though the representative sample epochs were necessarily short in duration it proved quite simple to classify pre- and post-excision EA into one of the four groups. The classitication may have underestimated post-excision increases in EA by including the suppressed periods of burst-suppression activity, which are probabl) anti-epileptic in nature (Yli-Hankala et al., 1993; Jsntti et al., 1994 ). In this regard. it ia of note that the three isolated cortices with decreased EA post-FH all showed a sebert: burst-suppression pattern. The specitic changes in cortical excitatory/inhibitory bal-ante induced by acute disconnection (i.e. following release from ascending brainstem d’ssynchronizing influences and interruption 01‘ subcorticocortical reverberating circuitry) and the mean\ by which they may activate pre-existent EA (e.g. increased synchronl/ation through increased inhibition versus tncreased local or recurrent excitation, either directly or via decreased inhibition) remain to be elucidated at the cellular level. The demonstration that such activation of EA ma) bc temporary (Niemeyer, 1958) suggests that acute changes in cortical inhibitory/excitatory balance following disconnection may IX compensated for over a rela-

tively short time period (e.g. days), perhaps through reorganization of intracortical connectivity or changes in neurochemical regulation. Such a pattern of acute facilitation of EA after disconnection followed by a period of disfacilitating intracortical reorganization could explain the gradual diminution of. and lack of clinical relevance of, EA recorded acutely from partially disconnected temporal and insular cortex (Niemeyer, 1958: Silfvenius et al.. 1964: Polkey et al.. 1989; Villemure et al.. 1989; Mascott ct al., 1990: Cendes et al., 1993) as well as bear relevance to the significance of post-excision ECOG \pikea in general. In conclusion, the ECOG results of this study corroborate previous findings of induction of burst-suppression as a nonspecifc cortical response to acute disconnection. It is demonstrated that thalamocortical disconnection is sufficient to induce burst-suppression but that it is not necessary to modulate burst-suppression. which can be augmented intracortically (in isolated cortical blocs) via systemic anesthetic administration. Unrelated to burst-suppression activity. an activation of EA is demonstrated in many patients following acute complete cortical isolation. in accordance with previous reports following acute partial cortical isolation. Such activation of EA is specific for cpileptogenic cortex and may represent a decreased threshold for the expression of EA related to alterations of excitatory/inhibitory balance in acutely disconnected cortex. Finally, it should be noted that post-FH patients also provide a model for examination of the effects of chronic complete cortical isolation in humans (Pierre-Louis and Morrell. 1992). Llnfortunately scalp EEG is insensitive to both the burst-suppression activity recorded directly from isolated cortex (Henry and Scoville. 1952: Echlin. 1959) and to a significant proportion of intracranially-recorded EA (Floor. 1975; Liiders et al., 1987: Mark\ et al.. 1992). Future developments in functional brain-imaging techniques (e.g. posiand/or magnetoencephalotron emission tomography) graphy may enable a more detailed and non-invasive study of paroxysmal activity from chronically isolated cortex.

Acknowledgements We would like to acknowledge the work of Drs F. Dubeau and E. Andermann who performed some of the electrocorticograms used in this study.

References Andermann. F. Chmcal indications for hcmispherectomy :md calloso~omy. Epilepsy Rcs.. 1991. Suppl. 5: IXY-1Y9. Brenwr. F. Cervrau

‘isolk’ et phyklopie

du uxnmcil.

C. R. SW. Biol..

1035. I IX: 1’35-1211. Bremer. F. El’fets de la d6aft’kentation

complkte d’unc n&on

cdrCbrali: SW bon activitP Clectrique apontank. 127: 355-35’).

dc I‘korce

C. R. lioc. Bml

.

193X.

Bums, D. Some properties of isolated cerebral cortex in the unanaesthetiaed cat. J. Physiol.. 195 1, 1 12: I56- 175. Carment. L., Kramer, U., Riviello, J., Helmers. S., Mikati, M., Madsen. J.. Black. P., Lombroso, C. and Holmes. G. EEG prior to hemispherectomy: correlation with outcome and pathology. Elrctroenceph. clin. Neurophysiol., 1995, 94: 265-270. Cendes, F.. Dubeau. F., Olivier. A., Cukiert. A.. Andermann. E.. Quesney. F. and Andermann, F. Increased neocortical spiking and surgical outcome after selective amygdalo-hippocampectomy. Epilepsy Res.. 1993, 16: 195-206. Dusser de Barennr, J. and Marshall, C. On a release phenomenon in electrical stimulation of the ‘motor’ cerebral cortex. Science. 1931. 73: 313-214. Ebrahim. Z., Schubert, A., Van Ness, P.. Wolgamuth, B. and Awad. I. The effect of propofol on the electroencephalogram of patients with epilepsy. Anesth. Analg.. 1994, 78: 275-279. Echlin, F.. Amett, V. and Zoll. .I. Paroxysmal high voltage discharges from isolated and partially isolated human and animal cerebral cortex. Electroenceph. clin. Neurophysiol.. 1952, 4: 147- 164. Echlin, F. The supersensitivity of chronically ‘isolated’ cerebral cortex as a mechanism in focal epilepsy. Electroenceph. clin. Neurophysiol., 1959. I 1: 697-7’22. Echlin. F. and Battista, A. Epileptiform seizures from chronic isolated cortex. Arch. Neural., 1963. 9: 154-170. Fisher-Williams. M. Burst-suppression activity as an indication of undercut cortex. Electroenceph. clin. Neurophysiol., 1963, 15: 723-724. Gloor, P. Contributions of electroencephalography and electrocorticography to the neurosurgical treatment of the epilepsies. In: D. Purpura. J. Penry and R. Walter (Eds.). Neurosurgical Management of the Epilepsies. Raven Press. New York, 1975. pp. 59-105. Grafstein. B. and Sastry, B. Some preliminary electrophysiological studies on chronic neuronally isolated cerebral cortex. Electroenceph. clin. Neurophysiol.. 1957, 9: 723-725. Henry. C. and Scoville. W. Suppression-burst activity from isolated cercbra1 cortex in man. Electroenceph. clin. Neurophysiol., 1952. 4: l-22. Hufnagel. A., Burr, W., Elger, C.. Nadstawek, J. and Hefner, G. Localiration of the epileptic focus during methohexital-induced anesthesia. Epilepsia, 1992, 33: 271-274. JBntti. V., Eriksson, K., Hartikainen. K. and Baer, G. Epileptic EEG discharges during burst suppression. Neuropediatrics, 1994. 25: 371-273. Kristianscn, K. and Courtois, G. Rhythmic electrical activity from isolated cerebral cortex. Electroenceph. clin. Neurophysiol., 1949, 1: 265-272. Liiders. H., Lesser, R., Dinner, D., Morris. H.. Hahn, J.. Friedman, L.. Skipper, G., Wyllie. E. and Friedman, D. Commentary: chronic intracranial recording and stimulation with subdural electrodes. In: J. Engel (Ed.). Surgical Treatment of the Epilepsies. Raven Press. New York. 1987, pp. 297-321. Marks, D., Katz, A.. Bookc. J.. Spencer, D. and Spencer. S. Comparison and correlation of surface and sphenoidal electrodes with simultaneous intracranial recording: an interictal study. Electroenceph. clin. Neurophysiol.. 1992. 82: 23-29. Mascott. C.. Villemure. J.-G., Andermann, F. and Rasmussen. T. Hemispherectomy and the insula. Can. J. Neural. Sci.. 1990, 17: 236. Musella. L.. Wilder. B. and Schmidt. R. Electroencephalographic activa-

tion with intravrnou? methohexital in psychomotor epilepsy. Neurology. 1971, 21: s9+602. Nierneyer, P. The transvrntricular amygdalo-hippocampectomy in tcnporal lobe epilepsy. In: M. Baldwin and P. Bailey (Eds.). Tcmpond Lobe Epilepsy. C.C. Thomas. Springfield. IL. 1958. pp. -161-381. P&e-Louis. S. and Morrell, F. Epilcptiform acti\dity in chronically istllated cerebral cortex in humans. Electroenccph. clin. Neurophysiol.. 1992. 57: 238-253. Polkey, C.. Binnie. C. and Janata. I. Acute hippocampal recording and pathology at temporal lobe rcaection and amygdalo-hippocampcctom) for epilepsy. J. Neural. Neurosurg. Psychiat.. 1989, 52: IOX& 1057. Rasmussen. T. Hemispherrctomy for seizures r&sited. Can. J. Neural. sci.. 1983. IO: 71-78. Rutledge. L.. Ranck. J. and Duncan. J. Prevention of supersensitivity in partially isolated cerebral cortex. Electroenceph. clin. Ncurophysiol.. 1967, 2.3: 756-267. Sharpless. S. and Halpern. L. The electrical excitability of chronically isolated cortex studied by means of permanently implanted elcctrodea. Electroenceph. clin. Neurophysiol., 1962. 14: 244-255. Silfvenius. H.. Gloor, P. and Rasmussen. T. Evaluation of insular ablation in surgical treatment of temporal lobe epilepsy. Epilepsia. 1964. 5: 307320. Smith. S.. Andermann. F.. Villemure. J.-G.. Rasmussen, T. and Qursney. F. Functional hrmispherectomy: EEG findings, spiking from isolated brain postoperatively, and prediction of outcome. Neurology. 1991, 41: 1790-1794. Spehlmann. R.. Daniel% J. and Chang, C. Acetylcholine and the epileptiform activity oichronically isolated cortex. I. Macroelectrode studies. Arch. Neural.. lY7l. 24: 40-408. Steriade, M.. Amzica, F. and Contreras. D. Cortical and thalamic cellular correlates of electroencephalographic burst-supprersion. Electroenceph. clin. Neurophysiol.. 1994. 90: I - 16. Tinupcr. P.. Andcrmann. F.. Villemurc, J.-G., Rasmussen. T. and Quesney. F. Functional hemispherectomy for treatment of epilepsy associated with hemiplcgia: rationale, indications. results. and comparison with callosotomy. Ann. Neural., 1988. 24: 27-34. Villemure. J.-G.. Mascott. C., Andermann, F. and Rasmushen. T. Is removal of the insular cortex in hemispherectomy necessary? Epilepsia. 1989. 30: 778. Villemure, J.-G. Anatomical to functional hemispherectomy from Krynauw to Rasmussen. Epilepsy Res., 1992, Suppl. 5: 209-215. Villemurc. J.-G.. Adams. C.. Hoffman, H. and Peacock, W. Hcmispherectomy. In: J. Engel (Ed.). Surgical Treatment of the Epilepsies. Raven Press. New York. 1993. pp. 5 I I-518. Villemure. J.-G. and Mascott, C. Pcri-insular hcmispherectomy: surgical principle< and anatomy. Neurosurgery, 1995. 37: 975-981. Wennberg. R.. Quesney. F. and Olivier. A. Methohexital and aelective amygdalo-hippocampectomy: activation of ncocortical cplleptiform activity hy a common mechanism’? Electroenceph. clin. Neumphysiol.. 1995. 97: 23X-239. Yli-Hankala. A.. Jantti. V., PyykhG, 1. and Lindgren. L. Vibration stimulus induced EEG bur\ta in isoiluranc anaestheaia. Electroenceph. clin. Neurophysiol.. I Y93, X7: 2 15-220.