Mutually suppressive interrelations of symmetric epileptic foci in bitemporal epilepsy and their inhibitory stimulation

Clinical Neurology and Neurosurgery 109 (2007) 7–22

Mutually suppressive interrelations of symmetric epileptic foci in bitemporal epilepsy and their inhibitory stimulation Sozari A. Chkhenkeli a,b,∗ , Vernon L. Towle a,c , George S. Lortkipanidze b , Jean-Paul Spire a , Eteri Sh. Bregvadze b , John D. Hunter c , Michael Kohrman c , David M. Frim d b

a Department of Neurology, The University of Chicago, USA Department of Functional Neurosurgery, Center of Epilepsy Surgery, Institute of Clinical & Experimental Neurology, Tbilisi, Georgia c Department of Pediatrics, The University of Chicago, USA d Department of Neurosurgery, The University of Chicago, Chicago, IL, USA

Received 10 November 2005; received in revised form 27 March 2006; accepted 31 March 2006

Abstract Objectives: The goal of this study is to analyze the suppressive interaction of symmetric temporal lobe epileptic foci, assess some failures of epilepsy surgery, and evaluate the possibility of terminating focal seizures with stimulation of symmetric epileptic foci. Materials and methods: One hundred and twenty-nine intractable epilepsy patients (age range 6–53 years) with bitemporal epileptiform abnormalities in multiple scalp EEGs were evaluated with chronically implanted depth and subdural electrodes. Interelectrode coherence and power spectra were studied using internally developed software. Results: Bitemporal epileptic foci were found in 85/129 (66%) patients with reciprocal relations between these foci in 57/85 (67%) patients. Temporal lobectomy was performed for 67/85 patients. 12/67 patients became free of seizures (Engel’s Class I), 32/67 improved (Classes II and III), and 23/67 did not improve. 14/23 patients demonstrated post-surgical activation of the contralateral temporal lobe epileptic focus. For 8/14 of these patients, the stereotactic cryoamygdalatomy was performed in the temporal lobe contralateral to the first surgery. 5/8 patients became free of seizures. It was found that stimulation of temporal lobe deep epileptic focus may terminate focal seizures in the contralateral symmetric structures. Conclusion: A mutually suppressive relationship is one of variants of the interaction of symmetric epileptic foci. Some epilepsy surgery failures may be a result of post-surgical activation of the intact focus. The increase of coherence between both temporal lobes before the seizure onset of the seizure suggests the establishment of functional interrelations between two epileptic foci at an early, “hidden” phase of seizures, and may predict the direction of seizure spread. Mutually suppressive interrelations of symmetric epileptic foci might be employed for chronic therapeutic stimulation. © 2006 Elsevier B.V. All rights reserved. Keywords: Bitemporal epilepsy; Deep epileptic foci; ECoG; SEEG; Coherence; Mutual suppression; Surgery failures; Therapeutic brain stimulation

1. Introduction Invasive recordings in intractable epilepsy patients with synchronous and independent epileptiform abnormalities involving both temporal regions in surface EEG have been ∗ Corresponding author. Department of Neurology MC-2030, The University of Chicago, 5841 South Maryland, Chicago, IL 60637, USA. Tel.: +1 773 834 4704/702 3271; fax: +1 773 702 4066. E-mail address: [email protected] (S.A. Chkhenkeli).

0303-8467/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.clineuro.2006.03.011

obtained by many epilepsy surgery centers in order to reliable lateralize the epileptogenic zone(s) and facilitate decisions about surgical treatment [1–13]. Experimental and clinical studies of the interrelation of mesiobasal temporal lobe structures have mostly focused upon interhemispheric connectivity, the spread of epileptiform activity from the one hemisphere to the other and evaluation of the pathways of this spread. However, the relationships between symmetric temporal lobe mesiobasal epileptic foci are still not clear, mostly being limited to animal experiments. Racine [14] described an increase of the afterdischarge threshold in rat hippocampus

8

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

after repeated stimulation of the contralateral homologous site. Mutani et al. [15] demonstrated mutually suppressive relations of asymmetric epileptic foci in cats. McIntyre and Goddard [16] have reported that during the experiments with kindling “transfer” between amygdalae, seizure development at the transfer site blocked the development of seizures from the original primary site, a phenomenon they termed “interference.” This phenomenon has also been shown to accompany transfer to the amygdala from other limbic sites [17]. McCaughran et al. [18] reported that dissection of corpus callosum, hippocampal commissure, or anterior commissure in rats resulted in a significant facilitation of primary-site kindling. In contrast to the animal studies, the clinical experience is limited. The activation of the contralateral epileptic focus after the surgical removal of an epileptic focus in the opposite temporal lobe has been reported and considered as a possible reason of epilepsy surgery failures [19,20]. Spencer et al. [21] described seizures developing after callosotomy in patients with bilateral frontal epileptic foci, suggesting the lack of a transcallosal inhibitory effect of epileptic foci. These data support the hypothesis of “disinhibitory” seizures [22,23]. The purpose of this study is to demonstrate mutually suppressive interactions of mesiobasal epileptic foci based on the analysis of pre-surgical EEG evaluations, studies using intraoperative and chronic SEEG and ECoG recordings, and to evaluate the possibility of artificially enhancing these interactions to terminate epileptic seizures.

2. Materials and methods 2.1. Patients The interrelations of symmetric temporal lobe mesiobasal structures were studied in 129 intractable epilepsy patients (88 men, 41 women, mean age 23 years (6–53 years), mean duration of illness 13 years (3–32 years), frequency of seizures 4–70/month) with bitemporal independent, as well as, synchronous interictal and ictal epileptiform abnormalities identified in multiple scalp EEGs. The intractability of seizures was determined on the basis of failure of 2–3 years of unsuccessful antiepileptic medication with phenobarbital, phenytoine, carbamazepine, sodium valproate used individually or in combinations, and a continuation or increase epileptic seizures frequency and severity, along with concomitant EEG and neuropsychological changes. Most of these patients were defined as patients with temporal lobe epilepsy, and had complex partial seizures with secondary generalization. The results of neurologic, CT, and MRI evaluation were inconclusive about the site of seizure origin. The pre-surgical diagnosis of bitemporal epilepsy was suggested on the basis of intermittent independent interictal epileptic activity and spontaneous independent subclinical focal discharges, as well as clinical seizures, probably originat-

ing from the temporal lobe mesiobasal structures of both hemispheres. Such clinico-EEG events as (a) independent bitemporal seizure onset, (b) non-lateralizing seizure onset, (c) bitemporal spiking, (d) spiking contralateral to the EEG seizure onset, (e) clinical manifestation clearly preceding the EEG onset, and (f) EEG onset being synchronously bitemporal, served as the indications for bilateral invasive evaluations for the possible lateralization of the dominant epileptic focus with SEEG and/or intracranial grid monitoring. Of the 129 patients, both temporal lobes were evaluated in 117 patients (1972–1996) intraoperatively and with chronically implanted depth and subdural electrodes at the Institute of Clinical and Experimental Neurology (Tbilisi, Georgia) and 12 were studied (1997–2005) at the University of Chicago with subdural electrode grids implanted bilaterally subtemporally and over the convexital surface of temporal, frontal, and parietal lobes. The experimental protocol was approved by the Institutional Review Boards of The Institute of Clinical and Experimental Neurology (Tbilisi) and The University of Chicago. Written informed consent was obtained from all patients or their guardians. 2.2. Surgery, implanted electrodes, functional tests Stereotactic operations were performed using Talairach’s stereotactic frame. The initial stage of surgery was usually performed under local and neurolept anesthesia with N2 O + O2 ventilation. The diagnostic electrodes for evaluation of temporal lobe mesiobasal structures were inserted using the lateral approach [24]. Subsequent intrasurgical diagnostic studies (usually 1–4 h) were performed in extubated awake patients. The coordinates of the intracerebral structures were calculated in the proportional system [25], with the exact location of the electrodes verified by intrasurgical orthogonal televentriculography with water-soluble contrast agents. Intracerebral electrodes for intrasurgical SEEG evaluation (Alexander & Co., Paris) had five 2 mm contact surfaces, contact-to-contact spacing of 2 mm, with a total span of 18 mm. The electrodes for long-term video-SEEG monitoring studies lasting 2–6 weeks consisted of flexible arrays of five to six platinum or gold wires with five 1–1.5 mm contact surfaces, with contact-to-contact spacing of 2–3 mm and a total span of 16–19.5 mm. ECoGs were recorded from the subdural 6–64 electrode arrays with interelectrode center-to-center spacing of 7.5 or 10 mm (Adtech Corporation, Milwaukee, WI). The temporal lobe resections included two modifications of the surgery, depending on the hemispheric dominance. The “standard temporal lobe resection” [25,26] in “en block” [27] modification was performed in the non-dominant hemispheres (29/67 patients). This resection usually included 6.0–6.5 cm of lateral cortex, uncus, amygdala, and 2–4 cm of the anterior hippocampus. In the dominant hemisphere (38/67 patients) the extension of cortical resection was reduced to 3–4 cm, and usually performed as a modification named “anterior medial temporal lobectomy” [28]. This modifica-

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

tion includes incision of the temporal lobe cortex for 3–3.5 cm from the temporal pole along the inferior surface of the superior temporal gyrus. During this surgery, the volume of resected hippocampus was less than during the standard temporal lobe resection. The decision about the site of surgery in patients with bitemporal pathology in SEEG and/or ECoG was made when 65–70% or more of electrographic seizures onset accompanying with clinical seizures events emanated from the site designated for surgery. The duration of electrographic seizures, and their intensity were also taken into consideration during the independent assessment (SCh, GL, and ShB) of electrographic data. The temporal lobe resections were not performed for patients with a frequency of unilateral independent clinical (viz. behavioral) psychomotor or secondary generalized epileptic seizures less than 80–85%. Additional assessment of the functional interrelations of symmetric temporal lobe mesiobasal structures during stereotactic operations was based on the results of local reversible cooling of deep brain structures using a special cryosurgical apparatus [29], and local diagnostic electrostimulation performed during both intrasurgical evaluation and long-term EEG/SEEG monitoring. This cryosurgical apparatus is a portable, lightweight (80 g) instrument consisting of two parts. The first is a 22 cm long thin-walled stainless steel cryocannula with an external diameter of 2 mm. Inside the cannula, there is a thin (1 mm diameter) hollow cold duct through which liquid nitrogen reaches the active (non-insulated) tip of the cannula. The body of the cannula is separated from the cold duct inside it (the space between them being only 200 ␮m) by a vacuum layer (degree of vacuum 10−6 ), ensuring a normal temperature of the entire external surface of the cannula with the exception of its 2 mm × 2 mm silver active tip. The second part of the instrument is a removable thermostable 100 ml foam plastic reservoir that fits tightly onto the thicker peripheral end of the cannula. For cooling or freezing, liquid nitrogen (−196 ◦ C) is poured into the reservoir, passes through the cold duct to enter the active tip, cools it, evaporates, and is aspirated. For the temporary cooling of the subcortical structure, 5–10 ml of liquid nitrogen is poured into the reservoir to lower the temperature at the active tip of the probe to 5–22 ◦ C (the degree of cooling depends on the volume of liquid nitrogen). This inactivated a 6–12 mm zone of cerebral tissue (2–5 mm of tissue around the 2 mm cannula). The cryolesion (freezing) could be achieved by filling the reservoir with 25 ml of liquid nitrogen (produced an ice sphere with a diameter of 5–6 mm), 40 ml produced an ice sphere of 8–9 mm, and 60 ml an ice sphere 13–14 mm [29,30]. The temperature at the tip of cannule and volume of lesion were preoperatively calibrated in the gelatin. Local diagnostic bipolar stimulations were performed using Nihon Kohden (Japan) and Grass S12 (Grass Med. Instruments, Quincy, MA, U.S.A.) stimulators and constant-current square pulses of alternate polarity with parameters chosen to avoid tissue damage [31].

9

2.3. Recordings and analysis EEG/SEEG recordings were obtained with 20-channel Alvar (Alvar-Electronic, France), Neurofax (Nihon-Kohden, Japan) and BMSI-6000 (Nicolet Biomedical, Madison, WI, USA) digital EEG machines. The data were recorded using a bandpass digital filter of 0.5–100 Hz and sampling rates of 400 or 500 Hz. ECoG inter-electrode coherence was calculated with the Scan 4.3 Edit Module (Compumedics Ltd., Australia) for all possible electrode pairs for each of the frequency bands used to describe surface EEG recordings: 1–4 Hz (delta), 4–7 Hz (theta), 7–13 Hz (alpha), 13–30 Hz (beta), 30–50 Hz (low gamma). We have developed a software package for determining the location of subdural electrodes, which displays lateral coherence patterns, power, and other EEG scalars [32]. The program displays EEG waveforms in a conventional manner along with spectrograms, allowing flexibility to choose channels of interest, and identification of various stages of seizure activity. Mean power spectra between 1 and 50 Hz were obtained [33] by averaging the spectra computed from each 512 point segment of the ECoG during different stages of the epileptic process.

3. Results Clinical investigations revealed that for patients suffering from intractable epilepsy with bitemporal epileptiform abnormalities in conventional EEGs, independent interictal intermittent spiking, with spontaneous focal electrographic and clinico-electrographic seizures in SEEGs and/or ECoGs were observed in both temporal lobes in 85/129 (66%) of patients (Fig. 1). For patients who underwent SEEG exploration with depth electrodes and subdural grids (117/129), independent bilateral deep temporal lobe foci were observed in 78/117 (66%) patients. For patients studied with subdural grids (12/129), bilateral independent epileptic activity was found in the ECoG of 7/12. A strongly unilateral clinico-electrographic onset of focal temporal lobe seizures was found in the remaining 44/129 (34%) of patients. For 7/44 patients, unilateral deep focal discharges altered the background EEG and even elicited epileptiform activity in the contralateral hemisphere without the involvement of symmetric deep temporal lobe structures. This group of patients was not included in the analysis. Reciprocal interrelations of symmetric temporal lobe structures were observed in the 57/85 (67%) of patients with independent active bitemporal foci. 10/85 patients with bitemporal epileptic electrical activity demonstrated a strongly unilateral onset of clinico-electrographic seizures, and did not reveal signs of reciprocal interrelations (Table 1). Unilateral temporal lobe resections was performed on 67/85 patients with bitemporal epileptiform abnormalities. Surgery was not performed for the remaining 18/85 patients because

10

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

Fig. 1. Groups of patients underwent preoperative EEG, intraoperative and chronic SEEG, chronic ECoG evaluations and surgery for temporal lobe epilepsy. Table 1 Electrophysiological findings in the 129 intractable epilepsy patients with bitemporal abnormalities on the EEG Findings

Bitemporal independent seizure activity Reciprocal interrelation between symmetric focia Without reciprocal interrelations between symmetric focia Indefinite prevalence of one of focia Strongly unilateral epileptic focusb a b

Study methods SEEG

SEEG and subdural grids

Subdural grids

Total

70 48 8 12 39

8 5 – 3 –

7 4 2 3 5

85 57 10 18 44

These three groups of patients composed the group of 85/129 patients with bitemporal independent SEEG/ECoG abnormalities. This group of patients was not included in analysis, because the absence of bitemporal SEEG/ECoG abnormalities.

of difficulty detecting the dominance of one focus in very active bitemporal epileptic systems. After 2 years, 12/67 patients were seizure free (Engel’s [34] Class I), 13/67 had rare seizures (Engel’s Class II), 19/67 were worthwhile improved (Engel’s Class III), and 23/67 did not improve (Table 2). None of the patients with a strongly unilateral onset of clinico-electrographic seizures and without signs of reciprocal interrelations of electrographic foci

(10/85) had an activation of the contralateral epileptic focus in the immediate post-surgical period, as well as during the 2 years of follow-up. Four of these 10 patients were in Engel’s Class I group (12/67), three patients were in the Class II group (13/67), and three were in the Class III group (19/67). 14/23 non-improved patients demonstrated an EEG activation of the intact contralateral epileptic foci. The evident

Table 2 Results of temporal lobectomy in 67/85 patients with independent bitemporal epileptic foci verified with SEEG and subdural electrode studies Patients

Results Class I

With reciprocal interrelation between symmetric foci Without reciprocal interrelations between symmetric foci Total

Class II

Class III

Class IV

Total

8 4

10 3

16 3

23 –

57 10

12

13

19

23

67

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

reciprocal interrelation between symmetric temporal lobe foci was detected in all of these 14 patients pre-surgically using the deep and subdural electrode studies. 6/14 patient were improved with medication. 8/14 patients underwent SEEG evaluation and stereotactic cryoamygdalatomy on the other side after unsuccessful attempts to achieve improvement with a correction of the medication. The indications for the second limited stereotactic amygdalatomy were, continued frequent disabling seizures, absence of the EEGevidences of incomplete resection during the first surgery, emergence of epileptic activity on EEG in the contralateral temporal lobe, SEEG verification of the active epileptic focus in the deep temporal lobe structures, and absence of evident memory problems. No additional neuropsychological deficit, especially memory deficit, was detected prior to the second surgery, and no surgical manipulations on the contralateral temporal lobe neocortex or hippocampal formation were performed in this group of patients. After their second surgery, 5/8 patients were seizure free (Engel’s Class I) and 3/8 had rare seizures (Engel’s Class II). The indications and detailed results of the bilateral surgery for bitemporal epilepsy is beyond of slope of our article, but we would state that no additional neuropsychological deficit was detected with a battery of neuropsychological tests (Wechsler, MMPI) in this group of patients. 3.1. Post-surgical activation of contralateral epileptic foci Fig. 2 depicts the process of activation of a right temporal lobe epileptic focus after a left anterior temporal lobectomy, because of a steady interictal and ictal EEG epileptic

11

focus in the left temporal region and intractable psychomotor seizures with frequent secondary generalization. After surgery, the frequency and severity of seizures considerably decreased. Ten months after surgery, less intense, but frequent and disabling psychomotor seizures reappeared. The EEGs obtained during this period revealed a progressive increase of interictal and ictal spike activity in the right anterior temporal region that was not recorded before surgery. The intraoperative SEEG study during his second surgery revealed active interictal epileptic activity and spontaneous focal epileptic electrographic and clinico-electrographic seizures emanating from the right amygdala. A right stereotactic cryoamygdalatomy was performed, and the patient has been without clinical seizures or aura and marked EEG abnormalities during the 4-year period of follow-up. This illustrates that the activation of a right temporal lobe dormant epileptic focus after removal of the left temporal lobe dominant epileptic focus in the epileptic system can be one of the reasons for the failure of surgical treatment, and the postoperative reappearance of seizures [20]. The discovery of inhibitory relationships between symmetric temporal lobe epileptic foci does not always need depth electrode investigations. Fig. 3 depicts the EEG of a patient with intractable epilepsy whose multiple EEG studies demonstrated a stable focus of epileptic activity in the left temporal region (Fig. 3A). During a Wada study, when the suppression of background activity developed over the left hemisphere after the injection of the left carotid artery, a clear focus of spike activity emanated from the right temporal region (Fig. 3B). Conjecture about the existence of a less active symmetric epileptic focus in the right temporal lobe was confirmed by clinical and elec-

Fig. 2. EEG and SEEG of a patient with bitemporal epilepsy in the process of activating of the dormant deep right temporal lobe epileptic focus after a left anterior temporal lobectomy. 1-A: Interictal EEG before the first surgery; 1-B: telemetric EEG during the short psychomotor seizure; 2-A: interictal EEG 10 months after surgery; 2-B: SEEG and EEG during the second surgery: right cryoamygdalatomy contralateral to the previous left lobectomy. Spontaneous epileptic activity is observed in the right amygdala.

12

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

Fig. 3. Activation of symmetric temporal lobe epileptic focus. (A) Routine pre-operative EEG—upper eight channels: left hemisphere; lower eight channels: right hemisphere; (B) EEG during the Wada test (left internal carotid artery is injected)—upper four channels: left hemisphere; lower four channels: right hemisphere; (C and D) EEG 3 weeks after a left anterior temporal lobectomy; (D) anterior temporal (AT) montage—fourth channel (T1 − T2 ) is a standard AT (anterior temporal) derivation.

trophysiological data obtained during the first 3 weeks after the left anterior temporal lobectomy. During that period the patient had a few complex partial seizures, which were much shorter than preoperative ones, without postictal total amnesia. Postoperative EEGs (Fig. 3C and D) revealed epileptic activity over the right temporal lobe, which was not observed before surgery. A correction of current medication reduced, and then stopped the clinical seizures and tapered the EEG abnormalities.

3.2. Inhibitory (suppressive) interrelations of deep temporal lobe epileptic foci confirmed by spontaneous seizures and cryoprobe Reciprocal inhibitory interrelations of symmetric mesial temporal lobe structures can be seen not only after the ablation of deep epileptic foci, but also during development of deep temporal lobe discharges in patients with bitemporal epilepsy (Fig. 4).

Fig. 4. Suppression of the background activity in the right hemisphere during the “tonic” phase of the left depth discharge. (A) Afterdischarge in the left amygdala–hippocampal complex (3 and 4) evoked by electrostimulation (6 mA, 0.2 ms, 5 s, 50 Hz) of the left amygdala. (B) The development of spontaneous focal discharges in the right amygdala and hippocampus (1 and 2) immediately after the cessation of left depth discharges during the “inactivation” phase in the initiating focus.

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

13

Fig. 5. Alternate facilitation of focal mesial temporal lobe discharges after the cessation of contralateral spontaneous focal electrographic seizures. (A) Development of spontaneous epileptic discharge in the right anterior (5) and posterior (6) hippocampus with the involvement of the ipsilateral amygdala (4); (B) development of spontaneous focal discharges in the left anterior (2) and posterior (3) hippocampus during the cessation (“clonic” phase) of the right depth discharges; (C) repeated spontaneous focal seizure emanated from the right anterior (5) hippocampus and spread to the posterior hippocampus (6) and right amygdala (4). This discharge spontaneously developed again after the cessation of a contralateral focal seizure.

Contralateral spontaneous epileptic discharge in the right hippocampus (Fig. 4B) can develop after the cessation of a focal electrographic seizure in the left amygdalahippocampal complex evoked by electrostimulation (Fig. 4A). Relative suppression of background activity in right deep structures was observed during the discharge in the left temporal lobe. The alternate outbursts of epileptic discharges in contralateral symmetrical temporal lobe structures can be observed both after electrically provoked discharges and spontaneously (Fig. 5). A spontaneous focal electrographic seizure in the right hippocampus with involvement of ipsilateral amygdala lead to the appearance of a focal epileptic discharges in symmetric structures of the contralateral temporal lobe. This event could be considered as an “activation” of the dormant epileptic focus by the discharge in the dominant focus, or as a simple contralateral spread of epileptic activity and involvement of other temporal lobe structures. However, it is critical to note that the contralateral discharge developed only after the cessation of activity in the dominant focus, and that the reappearance of a second focal seizure in the same right temporal lobe developed again after termination of the left temporal lobe discharge. Such reciprocal dynamics of epileptic discharges in symmetric deep temporal lobe epileptic foci were observed in 57/85 bitemporal epilepsy patients. Fig. 6 demonstrates the existence of a suppressive relationship between two symmetric amygdalar complexes in the absence of previous epileptic discharges, confirmed by the development of focal epileptic discharge in the right amygdala during the process of the diagnostic reversible local cooling of the left amygdala. Similar events were observed in 32 of 78 patients with independent bilateral deep temporal lobe epileptic foci after diagnostic reversible local cooling of the one deep epileptic focus.

3.3. Inhibitory (suppressive) interrelations of cortical temporal lobe epileptic foci confirmed by ECoG, spectrogram, power spectra and coherence Fig. 7 depicts the dynamics of the raw ECoG, spectrogram and power spectra during development of a spontaneous seizure involving both temporal lobes in a patient with independent epileptic foci evident from multiple pre-surgical EEGs. The asymmetry of the electrographic events over the right and left temporal lobe is evident, starting from the electrodecremental phase of the seizure and much more evident over the right anterior temporal lobe (Fig. 7B). The spectrogram of the ECoG channels primarily involved in the seizure reveals that the onset of the seizure in channel 4 (Fig. 7C) did not alter the activity over the left temporal lobe during this period. The ECoG activity over the left temporal lobe decreased with the progress of seizure development in the right temporal lobe (Fig. 7D for channels 1 and 2, and Fig. 7D–F for channel 3). In contrast, the progress and spread of the seizure over the left temporal lobe developed in parallel with a gradual decrement of seizure activity over the right temporal lobe (Fig. 7G and H). This concurrent interrelation of spontaneous epileptic seizures in both temporal lobes is also evident from the power spectra diagrams presented in the lower part of Fig. 7. The coherence started to increase in both epileptic foci in all frequency bands during the electrodecremental event (Fig. 8, block B) and after peaking at 0.6 started to drop after seizure initiation in channel 4 (block C). There are differences in the coherence dynamics in the different frequency bands between the initiating epileptic focus in the right temporal lobe (channels 3 and 4) and the left temporal lobe (channels 1 and 2), which is involved later. The coherence in the right temporal lobe increased predominantly in the theta- and gamma-bands, and less in the alpha-band, whereas coherence

14

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

Fig. 6. Inhibitory relationships between symmetric amygdalar epileptic foci. Development of a epileptic discharge in the right amygdala (1) in a patient with bitemporal epilepsy during the reversible cooling of the left amygdala (2). Photo (of a different case) demonstrates the process of cryoprobe. The cryocannule (black arrow) is inserted into the amygdala.

in the left temporal lobe increased in the alpha- and deltabands with some increase of gamma-band coherence. This first wave of coherence increase preceded the increase in the power spectra, which started to rise in the seizure initiating right epileptic focus 5–10 s later (block D), without a power increase in the left temporal lobe. This increase of power was accompanied by a drop of coherence in the initiating epileptic focus in the few seconds after seizure onset (block C). In contrast to the right temporal lobe, coherence between two channels representing the activity of the left temporal lobe epileptic focus in delta and theta bands remained high (0.7–0.75). Two peaks of coherence were observed in the left temporal lobe. The first was associated with a decrease in coherence in the right temporal lobe epileptic focus and seizure onset on channels 1 and 2. The second was associated with a decrease in both coherence and power in the right temporal lobe focus (block H) and maximum power level in the left temporal lobe focus. The dynamics of the power spectra and coherence in the right and left temporal lobes suggests the existence of concurrent interrelations between the epileptic foci in bitemporal epilepsy. The comparison of interchannel coherence between left and right temporal lobes epileptic foci (channel 1 to channels 3 and 4, channel 2 to channels 3 and 4), and coherence between channels within epileptic foci (channels 1 and 2, and channels 3 and 4) demonstrated that the interchannel coherence changes were more synchronous than coherence within the foci where it became synchronous only for a short period of time during the generation of the onset of seizures and concomitant increase in power. It was found that during the fully developed seizure in the right temporal lobe and the initial phase of the seizure in the left temporal

lobe (Fig. 7, block E) the alpha-band coherence in these areas changed in opposite directions. Alpha-band coherence in the seizure-initiating focus was maximal (0.7) in block E, but at the same point alpha-band coherence in the left temporal lobe dropped from 0.55 to 0.1, even lower than it was during the electrodecremental event. But there is a particular pattern: during the second wave of coherence increase in the initiating focus (block G), this increase developed in all frequency bands except the thetaband, which continued to drop, whereas in the left temporal lobe all frequency bands without exception increased in coherence and participated in the seizure build-up. Such coherence dynamics in different frequency bands may be a pattern distinguishing the gradually ceasing seizure in the right temporal lobe from the progressing seizure in the left temporal lobe. 3.4. Suppression of mesiobasal epileptic discharges with stimulation of contralateral temporal lobe mesiobasal structures Diagnostic electrostimulation of deep temporal lobe structures in patients with bilaterally implanted electrodes revealed complete suppression of the evoked mesiobasal temporal lobe epileptic discharge by stimulation of the contralateral symmetric structure. Fig. 9 demonstrates that high frequency stimulation (50 Hz, 4 mA, 0.2 ms, 1–1.5 s) of the right amygdala applied during epileptic discharge in the left amygdalahippocampal complex evoked an afterdischarge in the right amygdala with the concomitant complete cessation of the left temporal lobe seizure.

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

15

Fig. 7. ECoG, spectrogram (0–40 Hz), and power spectra (in ␮V2 /s) dynamics of the left (LTL) and right (RTL) temporal lobes during the development of spontaneous epileptic seizure. (A–J) The different stages of seizure chosen for the analysis. Channels 1–4 are chosen as the most active channels exhibiting the seizure activity. In the spectrograms, the frequency band for each channel is 0–40 Hz, the intensity of the red color depicts the highest power, and the intensity of the green color depicts the lowest power.

Fig. 8. Coherence dynamics during the epileptic seizure involved the right and left temporal lobes. Channels 1–4 and seizure stages A–J are consistent with channels 1–4 and A–J seizure stages in Fig. 6.

16

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

Fig. 9. Suppression of the epileptic discharge with the epileptogenic stimulation of the symmetric temporal lobe structures. The amygdalohippocampal discharge (1 and 2) and its complete suppression with the epileptogenic stimulation (5 mA, 0.2 ms, 2 s, 50 Hz) of the right amygdala (3); 2 and 4 SEEG of the left and right hippocampi, respectively. The rectangle placed over the right amygdala channels marks the moment of stimulation.

Fig. 10 illustrates that mutually suppressive interaction of the symmetrical temporal lobe mesiobasal structures can be realized in both directions. Stimulation of the right amygdala evoked an epileptiform afterdischarge (Fig. 10A), which was terminated by the amygdalahippocampal afterdischarge caused by short high frequency stimulation of the left amyg-

dala (Fig. 10B). Repeated stimulation of the right amygdala did not evoke an epileptiform afterdischarge, but rather terminated epileptiform activity in both hemispheres, with a generalized suppression of background activity (Fig. 10C). It was found that symmetric mesiobasal structure stimulation can have a more powerful suppressive impact on seizure

Fig. 10. The complete inhibition of the mesiobasal discharges in the both temporal lobes by the alternate stimulation and alternate development of convulsive discharges. (A) Epileptogenic stimulation (5 mA, 0.2 ms, 10 s, 50 Hz) of the right amygdala (3); (B) development of the focal discharge in the right amygdala; (C) epileptogenic stimulation (5 mA, 0.2 ms, 4 s, 50 Hz) of the left amygdala (1); (D) development of the discharge in left amygdala and hippocampus (1 and 2) and cessation of the discharge in the right amygdala; (E) stimulation of the right amygdale (3) did not evoke afterdischarge, but terminated the left mediobasal evoked discharge, and (F) suppression of background activity in both hemispheres developed after the last stimulation of right amygdala.

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

17

Fig. 11. The excitatory effect of the epileptic focus stimulation during a focal seizure and the inhibitory effect of the symmetric epileptic focus stimulation (5mA, 0.2 ms, 2 s, 50 Hz). (A) Epileptogenic stimulation of the right amygdala; (B) stimulation of the right amygdala during the afterdischarge; (C) increase of the right amygdala discharge intensity; (D) stimulation of the contralateral (left) amygdala; (E) termination of the afterdisharge.

activity than stimulation of the epileptic focus (Fig. 10). Low frequency stimulation (4 Hz, 4 mA, 0.2 ms, 1–1.5 s) of the epileptic focus in right amygdala during focal epileptic discharge (Fig. 11A) enhanced the intensity of the existing seizure, whereas high frequency stimulation (50 Hz, 4 mA, 0.2 ms, 5 s) of the symmetric amygdala terminated the contralateral seizure without epileptiform afterdischarge in the left amygdala. The described phenomena were observed in 44/57 patients (78%) with bilateral independent epileptiform activity and were reproducible in 85% of tests performed for one individual patient.

4. Discussion These findings reveal that in intractable bitemporal epilepsy patients symmetric deep temporal lobe structures cannot only produce epileptic discharges independently, or take part in the spread of seizure activity, but also interact with each other in an inhibitory (or suppressive) way. This interaction includes: (a) suppression of the activity of a contralateral epileptic focus at the onset of focal seizures; (b) suppression of interictal activity in the contralateral epileptic focus when the focal seizure intensity increases; (c) the development of spontaneous epileptic focal discharges in the contralateral temporal lobe immediately after termination of the first seizure; and (d) increase of the interictal and ictal activity in the dormant symmetric epileptic focus after surgical removal of the dominant epileptic focus. The last two phenomena suggest a release of the contralateral epileptic focus from the inhibitory influence of the symmetric epileptic

focus. There is also another, transitional variant, when during a seizure involving both temporal lobes, activity in temporal lobe contralateral to the seizure initiating epileptic focus significantly increases and reaches maximum only when the activity of initiating focus starts to slow and terminate. In our group of patients, independent epileptiform interictal intermittent spiking and focal electrographic and clinicoelectrographic seizures were observed in deep structures of both temporal lobes in two-thirds of patients. Strongly unilateral interictal epileptic activity and unilateral clinicoelectrographic focal seizures restricted to the mesial structures of a single temporal lobe were found in the remaining patients. In seven patients in that last group unilateral spontaneous deep focal discharges altered the background EEG and even elicited epileptiform activity in the contralateral hemisphere without the involvement of the symmetric deep temporal lobe structures, as has been reported earlier [6,20,35]. Bilateral ECoG independent interictal epileptic activity was found in about two-thirds of patients studied with subdural grids. Four of seven patients exhibited the independent onset of seizures in one temporal lobe with subsequent involvement of the contralateral temporal lobe. These findings differ from the findings reported by So et al. [5], who found that a majority of seizures originated exclusively or with a strong predominance in one temporal lobe in 44/57 (77%) patients. In the remaining 13 patients (23%), 8 (14%) had seizures originating independently in either temporal lobe without significant lateralized predominance. This difference, as well as considerably high percentage of the independent bitemporal epilepsy cases in our group of patients may be explained by differences in the patients’ cohorts, and the degree and

18

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

duration of withdrawal of medication. Different variants of the reciprocal interrelation of symmetric temporal lobe structures were observed in 67% patients studied with implanted depth electrodes and in 4 of 7 patients who were studied with subdural electrodes. The existence of the mutually suppressive interactions between symmetric temporal lobe structures has been observed in numerous experimental studies. Meglio et al. [36,37] evaluated the inhibitory and facilitatory interaction between two symmetrical amygdalar penicillin epileptic foci, and found that these reciprocal inhibitory interrelations could be prevented by lesioning the head of the caudate nucleus as well as by lesioning the anterior commissure. The authors suggested that these interactions could not be attributed to a single cerebral structure and involve several structures at different encephalic levels. Duchowny and Burchfiel [38] and Burchfiel et al. [39] described a phenomenon they termed “kindling antagonism”-suppression of seizures from one or both sides by administering alternating stimulation of two hippocampi and other limbic foci in rats. Fernandes de Lima et al. [40], based on experiments on rats, concluded that to eliminate interhemispheric spread, it is necessary and sufficient to section the specific commissural fiber system, and in the case of hippocampal epileptic foci, it is not the corpus callosum. Visually independent epileptic activity in human deep brain structures was reported at the beginning of the depth electrode era [41–43], which immediately raised questions about the interhemispheric spread of this activity and the pathways mediating interstructural interactions. Pagni [44] mentioned that deep temporal lobe epileptic discharges never spread to the contralateral symmetric structures. Later on, Pagni and Marossero [45], Angeleri et al. [46] and Buser and Bancaud [47] considered the contralateral spread of the rhinencephalic discharges as an exception, rather than a rule. Even using evoked responses and cross-correlation analysis, Brazier [48–50] and Wilson et al. [51,52] did not find connections between hippocampal symmetric points in humans as well as high interhemispheric coherence between amygdalae and hippocampi. Experimental morphological data supporting these electrophysiologic findings were reported earlier by Amaral et al. [53] and Demeter et al. [54], who observed a reduction of direct commissural connections in the phylogenetic row. Wilson et al. [55] suggested that the differences in degree of anatomical and functional communications between hippocampi in humans or primates and sub-primates might be reflected in interhemispheric differences and lateralization of function. Engel [56] reported that contralateral seizures appear after a series of habitual seizures, and recognized them as atypical events. Our data suggest that at least for a part of such seizures the activation of contralateral epileptogenic region can be prognostically important. This is especially important when this activation develops not for a short time after the seizure (Figs. 4 and 5), but rather as a constant and incrementing phenomenon having an effect on the intensity of the clinical and electrophysiological events after

resection (Figs. 2 and 3) of the primary (dominant) “epileptogenic region” and makes surgery less efficient. Activation of contralateral temporal lobe epileptic focus after temporal lobectomy, or temporary cooling of a deep epileptic focus (Figs. 2 and 6), emanation of epileptic activity in the temporal lobe contralateral to the carotid artery injected during the Wada test (Fig. 3), and dynamics of epileptic seizures in both temporal lobes (Figs. 4, 5 and 7) may be explained with mutually suppressive interactions of symmetric epileptic foci. Disinhibition of one of them occurs when the second focus is removed or lowers its activity. The problem of which neural pathways are really involved in the interhemispheric propagation of epileptic discharges originating in mesial temporal lobe structures in humans has been discussed for a long time. Wilson et al. [52], Buser et al. [57], and Catenoix et al. [58] failed to obtain evoked responses with stimulation of contralateral homotopic sites. Lieb et al. [59,60], Bertashius [61], and Wilson and Engel [62] found low interhippocampal coherence of epileptiform activity, and concluded that hippocampal and forebrain commissures play a minor role in interhemispheric propagation of the seizures emanating from deep temporal lobe structures in humans. This suggestion was supported later by coherence/phase analysis of seizure activity [60]. Wilson and Engel [63] and Lieb et al. [62] emphasized the role of orbito-frontal cortex in the interhemipheric propagation of temporal lobe seizures. Clinical investigations in patients with multicontact electrodes revealed strong evidences that seizure discharges originating in the deep structures of one temporal lobe can spread to contralateral structures without prior involvement of thalamic nuclei or ipsi- and contralateral neocortex [6,64–66]. Spencer et al. [64] suggested the important role of the ventral hippocampal commissure. Gloor and coworkers [67,68] found that the dorsal hippocampal commissure appears to be a functional pathway in humans and made the very important suggestion that repetitive stimulation might be necessary to effectively activate this pathway. That may explain the failures to detect interhippocampal connection using the relatively rare stimulating stimuli [48–52], whereas high frequency epileptic discharges rapidly spread to the contralateral hippocampus. Animal experiments [20] and coherence/phase analyze of electrical activity in patients with implanted electrodes [7,60] indicate that the anterior commissure may be relatively unimportant for interhemispheric propagation of mesial temporal lobe discharges. In contrast, Tomasch [69] made quantitative investigations of the anterior commissure in humans and found that altogether 2.4–4.2 million fibers pass through the sagittal 3–5 mm2 section of the anterior commissure. Iwai and Yukie [70] reported that if commissural afferents to amygdala exist, they connect not the symmetric structures, but the amygdala receives fibers from the contralateral inferior temporal cortex. These last findings suggest that the reciprocal interrelations transferred through the anterior commissure might be realized not only between strictly symmetric temporal lobe structures but also after the spread of the deep

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

discharges onto the homolateral inferior temporal lobe cortex, and than to the contralateral temporal lobe. Thus, the participation of the anterior commissure in the direct connections between the temporal lobes cannot be ruled out. Fernandes de Lima et al. [40] suggested that the commissural system might be “strong enough to ensure the formatting of the one bilateral oscillating system.” These discussions had clinical (surgical) sense when the surgical strategy of the strictly localized stereotactic lesion of only the pathways funneling the epileptic discharge was accepted as a promising method of epilepsy surgery. But many years experience obtained around the world have incontrovertibly demonstrated that such a strategy has failed in the treatment of patients with intractable epilepsy. The question is not about the spread of discharges throughout the anterior or posterior commissure or other pathways, but how symmetrical structures of both temporal lobes become involved in the process of epileptogenesis, what surgical tactic (resective surgery or therapeutic stimulation, or both) should be chosen for the successful treatment of that cohort of patients, and what an impact the interstructural interrelations may have on the results of surgery. Our data demonstrate (a) suppression of background activity in the symmetric areas of the contralateral temporal lobe during epileptic discharges in the other temporal lobe, (b) an increase of power of the contralateral discharge only when the initiating discharge started to cease, and (c) emanation of epileptic discharges immediately after the cessation of discharges in the contralateral seizure initiating focus. The increase in coherence of two epileptic foci even before obvious seizure onset and at the onset of the seizure, and suppression of the focal epileptic discharges with stimulation of the symmetric structure suggest the close interrelations between the both temporal lobes in bitemporal epilepsy howsoever that interaction is realized. Gotman [4] found that coherence between seizure activities in both temporal lobe structures was generally low, except the patients with independent bitemporal seizure onset. Our data may indicate that the high coherence between two temporal lobe epileptic foci immediately before and at the onset of seizure is a specific pattern of bitemporal epilepsy. Analysis of coherence and power spectra dynamics suggests that the theta- and alpha-bands might have a particular importance for seizure build-up. The opposite directions of alpha-band coherence dynamics in symmetric epileptic foci (Fig. 7, channels 1–4 coherence, block E) was observed only for alpha-band coherence, and may represent a pattern of mutual suppression interrelation between two symmetric temporal lobe epileptic foci. The increase of coherence between both temporal lobes before the visually detectable electrographic onset of the seizure (Fig. 7, block B) and at the onset of the seizure (Fig. 7, block C) without the concomitant increase of power in the left temporal lobe suggests the establishment of functional interrelations between two epileptic foci at an early, “hidden” phase of seizures, and may predict the direction of seizure spread. Perhaps, the initial increase of coherence (Fig. 7, block B) heralding imminent seizure

19

development may be used to trigger a stimulating device to preclude further development of a seizure. A comparison of the coherence between channels recording the activity of both temporal lobes (channel 1 to channels 3 and 4, channel 2 to channels 3 and 4) coherence and coherence between epileptic foci channels (channels 1 and 2, and channels 3 and 4) demonstrated that the coherence changes in these channels are more synchronous than coherence within the foci where it became synchronous only for short period of time during the generation of the onset of seizures and increase of power. These findings may reflect the establishment of functional connectivity between two mesiobasal epileptic foci during the seizure. The generalized suppression of background activity after focal seizure termination with contralateral structure stimulation (Fig. 10) raises a question about activation of, an additional to direct interstructural influences, universal and powerful mechanism having an inhibitory effect on both hemispheres [71]. The inhibitory interrelation between epileptic foci may not be a “privilege” of mesiobasal temporal lobe structures and symmetric epileptic foci only. Spencer et al. [21] described that more intense and newly patterned focal seizures developed after callosotomy in 5/7 patients with bifrontal independent asymmetric EEG epileptic foci. They suggested the existence of an inhibitory interrelation between these active asymmetric foci preoperatively, and hypothesized that the homotopic foci may have facilitating interrelations. Our data demonstrated that the mutually inhibitory interrelation at least exist between deep temporal lobe symmetric structures. Recently, attempts were undertaken to stimulate deep temporal lobe epileptic foci for therapeutic purposes and explain the effects of stimulation with different mechanisms. Velasco et al. [72,73] found that in 7/10 intractable temporal lobe epilepsy patients subacute electrical stimulation of the hippocampal formation or gyrus abolished clinical seizures and decreased the number of interictal spikes. Cuellar-Herrera et al. [74] concluded that electrical stimulation of deep temporal lobe foci is associated with a high GABA tissue content and a low rate of cell loss in patients with less severe epilepsy. Vonck et al. [75] found that amygdalahippocampal stimulation lead to a greater than 50% reduction in seizure frequency. Vonck et al. [76] considered desynchronizations of abnormal synchronous epileptic activity as one of the hypotheses of the mode of action that might primarily be responsible for the anti-seizure effects. Our data about the high incidence of bitemporal epilepsy among the intractable epilepsy patients allows us suggest that at least in the some of above mentioned cases the beneficial results with temporal epileptic focus stimulation could be on the account of an activation of the mutually suppressive mechanisms.

5. Conclusion Despite the positive results of surgery (2/3 of patients became seizure free or improved), this article is not designed

20

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22

to advocate surgical treatment of all bitemporal intractable epilepsy cases. Besides shedding light on the neurophysiologic mechanisms of interaction of epileptic foci, the results obtained demonstrate that even in cases with obvious dominance of one epileptic focus over another this subjugated focus might activate and become responsible for continued clinical seizures. These data should be considered and additional precautions should be taken during the indications for surgery in patients with seemingly unilateral temporal lobe foci, or an “obvious” predominance of one of temporal lobe foci on scalp EEG. EEG/ECoG/SEEG studies and analysis of the surgical results in bitemporal epilepsy suggest the existence of a mutually suppressive interrelation between symmetrical epileptic foci. The results of surgery demonstrate that some failures are results of activation of the contralateral focus verified with EEG/ECoG/SEEG evaluation. The likelihood of development of the described phenomenon of post-surgical activation of the remaining intact epileptic focus requires the meticulous assessment of pre- and post-surgical clinico-electrophysiologic data in order to not to loose valuable EEG information, which can be a clue to the activation of the dormant, or subdominant epileptic focus in the contralateral temporal lobe. Such an assessment is important for distinguishing real failures of surgery because of incorrect choice of type of surgery, or incomplete removal of epileptogenic region from insufficient pre-surgical knowledge about the real organization of the individual for patient’s epileptic system. The existence of these suppressive (inhibitory) reciprocal interrelations could be detected during the pre-surgical evaluation (i.e. during the Wada test), by the suppression of EEG/ECoG/SEEG background activity in the contralateral focus during focal epileptic discharges in the active epileptic focus, and development of a focal seizure in the contralateral structure during the postictal depression in the initiating focus. The mutually suppressive interaction of bilateral seizures could be confirmed by the decrease of coherence, power spectra and spectrogram parameters in one epileptic focus during the increase of seizure intensity in the other. Intrasurgically, these interrelations could be verified with epileptogenic stimulation of one epileptic focus and suppression of baseline activity in the symmetric structure, as well as, with the development of the epileptic discharges during reversible cooling of the one focus, or after its’ cryogenic lesion. Finally, the existence of the active inhibitory interrelations is verified with the termination of developing seizures in one temporal lobe with high frequency stimulation of the symmetric temporal lobe structures. Our data suggest that these mutually suppressive interrelations are bi-directional, and may be realized through direct commissural mono- or oligosynaptic connections, as well as, through polysynaptic pathways activating diffuse brain inhibitory system(s). Our findings allow us to suggest that these mutually suppressive interrelations might be employed for direct chronic therapeutic stimulation of the epileptic focus in bitemporal epilepsy, a condition for which the indica-

tions for the ablative surgical treatment remains strictly limited.

References [1] Bancaud J, Talairach J, Bonis A, et al. La Stereoelectroencephalograpie dans L’epilepsie. Paris: Masson et Gie; 1965. 321 pp. [2] Nashold Jr BS, Flanigin Jr H, Wilson WP, Stewart B. Stereotactic evaluation of bitemporal epilepsy with electrodes and lesions. Confin Neurol 1973;35:94–100. [3] Olivier A, Gloor P, Andermann F, Quesney LF. The place of stereotactic depth electrode recording in epilepsy. Appl Neurophysiol 1985;48:395–9. [4] Gotman J. Interhemispheric interaction in seizures of focal onset: data from human intracranial recordings. Electroencephalogr Clin Neurophysiol 1987;67:120–33. [5] So N, Gloor P, Quesney LF, Jones-Gotman M, Olivier A, Andermann F. Depth electrode investigations in patients with bitemporal epileptiform abnormalities. Ann Neurol 1989;25:423–31. ˇ [6] Chkhenkeli SA, Sramka M, Epilepsy and its surgical treatment, vol. I, Topical Diagnosis. Acad. Romodanov AP. (Ed.), Bratislava: Slovak Acad. of Sci. “VEDA” Publ. 1990. 274 pp. [7] Hirsch LJ, Spencer SS, Spencer DD, Williamson PD, Mattson RH. Temporal lobectomy in patients with bitemporal epilepsy defined by depth electroencephalography. Ann Neurol 1991;30:347– 56. [8] Kuzniecky R, Faught E, Morawetz R. Electroencephalographic correlations of extracranial and epidural electrodes in temporal lobe epilepsy. Epilepsia 1991;32:335–40. [9] Cascino GD, Trenerry MR, Sharbrough FW, So EL, Marsh WR, Strelow DC. Depth electrode studies in temporal lobe epilepsy: relation to quantitative magnetic resonance imaging and operative outcome. Epilepsia 1995;36:230–5. [10] Blatt DR, Roper SN, Friedman WA. Invasive monitoring of limbic epilepsy using stereotactic depth and subdural strip electrodes: surgical technique. Surg Neurol 1997;48:74–9. [11] Spanedda F, Cendes F, Gotman J. Relations between EEG seizure morphology, interhemispheric spread and mesial temporal atrophy in bitemporal epilepsy. Epilepsia 1997;38:1300–14. [12] Diehl B, L¨uders HO. Temporal lobe epilepsy: when are invasive recordings needed? Epilepsia 2000;41:S61–74. [13] Holmes MD, Miles AN, Dodrill CB, Ojemann GA, Wilensky AJ. Identifying potential surgical candidates in patients with evidence of bitemporal epilepsy. Epilepsia 2003;44:1075–9. [14] Racine RJ. Modification of seizure activity by electrical stimulation. I. After-discharge threshold. Electroencephalogr Clin Neurophysiol 1972;32:269–79. [15] Mutani R, Bergamini L, Fariello R, Quattrocolo G. An experimental investigation on the mechanisms of interaction of asymmetrical acute epileptic foci. Epilepsia 1972;13:597–608. [16] McIntyre DC, Goddard CV. Transfer, interference and spontaneous recovery of convulsions kindled from the rat amygdala. Electroencephalogr Clin Neurophysiol 1973;35:533–43. [17] Burnham WM. Primary and “transfer” seizure development in the kindled rat. Can J Neurol Sci 1975;2(4):417–28. [18] McCaughran Jr JA, Corcoran ME, Wada JA. Role of the forebrain commissures in amygdaloid kindling in rats. Epilepsia 1978;19:19– 33. [19] Chkhenkeli SA. Analysis of several causes of inadequate surgical treatment of epilepsy. Zurnal Voprosy Neirokhirurgii Imeni N-N-Burdenko 1981;(3):47–55. [20] Chkhenkeli SA. Neurophysiological basis and results of combined stereotactic surgical treatment of complex forms of epileptic seizures. Zhurnal Nevrpatologii I Psikhiarii Imeni S-S-Korsakova 1982;82:891–901.

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22 [21] Spencer SS, Spencer DD, Glaser GH, Williamson PD, Mattson RH. More intense focal seizure types after callosal section: the role of inhibition. Ann Neurol 1984;16:686–93. [22] Engel Jr J. The Hans Berger lecture. Functional explorations of the human epileptic brain and their therapeutic implications. Electroencephalogr Clin Neurophysiol 1990;76:296–316. [23] Engel Jr J. Clinical aspects of epilepsy. Epilepsy Res 1991;10:9–17. [24] Talairach J, Szikla G, Tournoux P, Corredor H, Kvasina T. Atlas d’anatomie st´er´eotaxique du t´el´enc´ephale. Paris: Masson et Gie; 1967, 294 pp. [25] Morris AA. Temporal lobectomy with removal of uncus, hippocampus, and amygdala. Arch Neurol Psychiat 1956;79:479–96. [26] Walker AE. Surgery for epilepsy. In: Magnus O, Lorentz de Haas AM. (Eds.), Handbook of clinical neurology, vol. 15. Amsterdam, NorthHolland Pub. Co.; 1974. p. 39–757. [27] Falconer MA, Cavanagh JB. Clinicopathological considerations of temporal lobe epilepsy due to small focal lesions. Brain 1956;82: 483–504. [28] Spencer DD, Doyle WK. Temporal lobe operations for epilepsy: radical hippocampectomy. In: Schmidek HH, Sweet WH, editors. Operative neurosurgical techniques, vol. 2, 3rd ed. 1988. p. 1305–16. [29] Kandel EI. In: Walker AE, editor. Functional and stereotactic surgery. New York/London: Plenum Press; 1989. p. 695. [30] Chkhenkeli SA. Cryostereoencephalotomy in the temporal lobe epilepsy. Bull Georgian Acad Sci 1978;90:721–4. [31] Gordon B, Lesser RP, Rance NE, et al. Parameters for direct cortical electrical stimulation in the human: histopathologic confirmation. Electroencephalogr Clin Neurophysiol 1990;75:371–7. [32] Hunter JD, Hanan DM, Singer BF, et al. Locating chronically implanted subdural electrodes using 3-D rendering. Clin Neurophysiol 2005;116(8):1984–7. [33] Press WH, Teukolsky SA, Vetterling SA, Flannery BP. Numerical recipes in C: the art of scientific computing. 2nd ed. Cambridge: Cambridge University Press; 1992. [34] Engel Jr J. Outcome with respect to epileptic seizures. In: Engel Jr J, editor. Surgical treatment of the epilepsies. New York: Raven Press; 1987. p. 553–71. [35] Mintzer S, Cendes F, Soss J, et al. Unilateral hippocampal sclerosis with contralateral temporal scalp ictal onset. Epilepsia 2004;45:792–802. [36] Meglio M, Ianelli A, Anile C, Rossi GF. Experimental bilateral deep temporal epilepsy. Effects of ablation of one focus and of different brain lesions. Epilepsia 1976;17:449–59. [37] Meglio M, Iannelli A, Anile C, Rossi GF. Interaction of epileptic activities of bilateral deep temporal origin. An experimental study. Epilepsia 1976;17:437–48. [38] Duchowny MS, Burchfiel JL. Facilitation and antagonism of kindled seizure development in the limbic system of the cat. Electroencephalogr Clin Neurophysiol 1981;51:403–16. [39] Burchfiel JL, Serpa KA, Duffy FH. Further studies of antagonism of seizure development between concurrently developing kindled limbic foci in the rat. Exp Neurol 1982;75:476–89. [40] Fernandes de Lima VM, Pijn JP, Nunes FC, Lopes da Silva F. The role of hippocampal commissures in the interhemispheric transfer of epileptiform afterdischarges in the rat: a study using linear and non-linear regression analysis. Electroencephalogr Clin Neurophysiol 1990;76:520–39. [41] Bickford RG. The application of depth electroencephalography in some varieties of epilepsy. Electroencephalogr Clin Neurophysiol 1956;8:526–7. [42] Ajmone-Marsan C, Van Buren JM. Epileptiform activity in cortical and subcortical structures in the temporal lobe of man. In: Baldwin M, Bailey P, editors. Temporal lobe epilepsy. Springfield, Ill. Ch.C.Thomas; 1958. p. 78–108. [43] Fisher-Williams M, Cooper RA. Depth recording from the human brain in epilepsy. Electroencephalogr Clin Neurophysiol 1963;15:568–87. [44] Pagni CA. Etude electro-clinique des post-decharges amygdalohippocampiques chez l’homme par moyen d’electrodes de pro-

[45]

[46]

[47]

[48] [49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

21

fondeur placees avec methode stereotaxique. Confin Neurol 1963;23: 477–99. Pagni CA, Marossero F. Some observations on the human rhinencephalon—a stereo-electroencephalographic study. Electroencephalogr Clin Neurophysiol 1965;18:260–71. Angeleri F, Ferro-Milone F, Parigi S. Electrical activity and reactivity of the rhinencephalic, pararhinencephalic and thalamic structures: prolonged implantation of electrodes in man. Electroencephalogr Clin Neurophysiol 1964;16:100–29. Buser P, Bancaud J. Bases techniqueset methodologiyes del’exploration fonctionnelle stereotaxique du telencephale. In: Talairach J, Szikla G, Tournoux P, Corredor H, Kvasina T, editors. Atlas d’anatomie st´er´eotaxique du t´el´enc´ephale. Paris: Masson et Gie; 1967. p. 251–320. Brazier MAB. Evoked responses from the depths of the human brain. Ann N Y Acad Sci 1964;112:33–59. Brazier MAB. Spread of seizure discharges in epilepsy: anatomical and electrophysiological considerations. Exp Neurol 1972;36: 263–72. Brazier MAB. Electrical seizure discharges within the human brain: the problem of spread. In: Brazier MAB, editor. Epilepsy, its phenomena in man. New York: Academic Press; 1973. p. 153–70. Wilson CL, Isokawa-Akesson M, Wang ML, Babb TL, Engel Jr J. Lack of functional evidence for a human hippocampal commissure. Electroencephalogr Clin Neurophysiol 1985;61:S54–5. Wilson C, Isokawa M, Babb T, Crandall P. Functional connections in the human temporal lobe. I. Analysis of limbic system pathways using neuronal responses evoked by electrical stimulation. Exp Brain Res 1990;82:279–92. Amaral DG, Insausti R, Cowan WM. The commissural connections of the monkey hippocampal formation. J Comp Neurol 1984;224:307– 36. Demeter S, Rosene DL, Van Hoesen GW. Interhemispheric pathways of the hippocampal formation, presubiculum, and entorinal and posterior parahippocampal cortices in the rhesus monkey: the structure and organization of the hippocampal commissures. J Comp Neurol 1985;233:30–47. Wilson CL, Isokawa-Akesson M, Babb TL, Engel Jr J, Cahan LD, Crandall PH. A comparative view of local and interhemispheric limbic pathways in humans: an evoked potential analysis. In: Engel Jr J, Ojemann GA, L¨uders HO, Williamson PD, editors. Fundamental mechanisms of human brain function. New York: Raven Press; 1987. p. 27–38. Engel Jr J. Bilateral temporal lobe epilepsy. In: Wolf P, editor. Epileptic seizures and syndromes. John Libbey and Company Ltd.; 1994. p. 359–68. Buser P, Bancaud J, Talairach J, Szikla G. Interconnections amygdalo-hippocampiques chezl’homme. Etude physiologiue u cours d’explorations st´er´eotaxiques. Revue Neurol 1968;119:283–8. Catenoix H, Magnin M, Guenot M, et al. Hippocampal connectivity in human: an electrical stimulation study. Epilepsia 2004;45(Suppl 7):58–9. Lieb J, Babb T, Engel J, Darcey T. Propagation pathways of interhemispheric seizure discharge compared in human and animal hippocampal epilepsy. In: Engel Jr J, editor. Fundamental mechanisms of human brain function. New York: Raven Press; 1987. p. 165–70. Lieb J, Hoque K, Skomer C, Song X. Inter-hemispheric propagation of human mesial temporal lobe seizures: a coherence/phase analysis. Electroencephalogr Clin Neurophysiol 1987;67:101–19. Bertashius KM. Propagation of human complex-partial seizures: a correlation analysis. Electroencephalogr Clin Neurophysiol 1991;78:333–40. Lieb J, Dasheiff R, Engel J. Role of the frontal lobe in the propagation of human mesial temporal lobe seizures. Epilepsia 1991;32: 822–37. Wilson C, Engel J. Electrical stimulation of the human epileptic limbic cortex. In: Devinsky O, Beric A, Dogali M, editors. Electrical and

22

[64]

[65]

[66]

[67]

[68] [69] [70]

S.A. Chkhenkeli et al. / Clinical Neurology and Neurosurgery 109 (2007) 7–22 magnetic stimulation of the brain and spinal cord, vol. 8. New York: Raven Press; 1993. p. 103–13. Spencer SS, Williamson PD, Spencer DD, Mattson RH. Human hippocampal seizure spread studied by depth and subdural recording: the hippocampal commissure. Epilepsia 1987;28:479–89. Wieser HG. Human limbic seizures: EEG studies, origin, and patterns of spread. In: Mekdrum BS, Ferrendelli JA, Wieser HG, editors. Anatomy of epileptogenesis. London and Paris: John Libbey; 1988. p. 127–38. ˇ Sramka M, Chkhenkeli SA. Epilepsy and its surgical treatment, vol. 2. Surgical Treatment. Acad. Romodanov, Pogady I, Drobny M. (Eds.), Bratislava: Slovak Acad. of Sci. “VEDA” Publ; 1993. 290 pp. Gloor P, Salanova V, Olivier A, Quesney LF. The human dorsal hippocampal commissure. An anatomically identifiable and functional pathway. Brain 1993;116:1249–73. Gloor P. The temporal lobe and limbic system. New York: Oxford University Press; 1997, 865 pp. Tomasch J. A quantitative analysis of the human anterior commissure. Acta Anat 1957;30:902–6. Iwai E, Yukie M. Amygdalofugal and amygdalopetal connections with modality-specific visual cortical areas in macaques (Macaca fus-

[71]

[72]

[73]

[74]

[75]

[76]

cata, M. mulata, and M. fascicularis). J Comp Neurol 1987;261:362– 87. ˇ Chkhenkeli SA, Sramka M, Lortkipanidze GS, et al. Electrophysiological effects and clinical results of direct brain stimulation for intractable epilepsy. Clin Neurol Neurosurg 2004;106:318–29. Velasco M, Velasco F, Velasco AL, et al. Subacute electrical stimulation of the hippocampus blocks intractable temporal lobe seizures and paroxysmal EEG activities. Epilepsia 2000;41:158–69. Velasco F, Velasco M, Velasco AL, Menez D, Rocha L. Electrical stimulation for epilepsy: stimulation of hippocampal foci. Stereotact Funct Neurosurg 2001;77:223–7. Cuellar-Herrera M, Velasco M, Velasco F, et al. Evaluation of GABA system and cell damage in parahippocampus of patients with temporal lobe epilepsy showing antiepileptic effects after subacute electrical stimulation. Epilepsia 2004;45:459–66. Vonck K, Boon P, Achten E, De ReuckJ, Caemaert J. Long-term amygdalohippocampal stimulation for refractory temporal lobe epilepsy. Ann Neurol 2002;52:556–65. Vonck K, Boon P, Goossens L, et al. Neurostimulation for refractory epilepsy. Acta Neurol Belg 2003;103:213–7.