SEARCH

SEARCH BY CITATION

Keywords:

  • Seizures;
  • Anterior nucleus of the thalamus;
  • Deep brain stimulation;
  • Rat;
  • Electrical stimulation;
  • Kainate;
  • Kainic acid;
  • Epilepsy;
  • Prolonged stimulation

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgments
  7. REFERENCES

Summary: Purpose: Electrical stimulation of the anterior nucleus of the thalamus (ANT) is receiving increased attention as a novel means of controlling intractable epilepsy, and has entered human clinical trial. Animal data supporting the anticonvulsant benefit of ANT stimulation, however, has been obtained from acute chemoconvulsant models of epilepsy rather than models of chronic epilepsy with spontaneous seizures. It is unknown whether ANT stimulation is effective in models of chronic epilepsy.

Methods: Bilateral ANT stimulation was evaluated in rats with chronic epilepsy following acute status epilepticus (SE) produced by systemic kainic acid (KA) administration. The evolution of epilepsy following KA SE and the effects of ANT stimulation were monitored by continuous video-EEG.

Results: Following KA SE, most rats have 2–8 seizures per day, and the average seizure rate increases over time, doubling over the course of 14 weeks. Behavioral seizure severity, after the initial development of epilepsy, remains stable. Seizure frequency during ANT stimulation was 2.5 times the baseline seizure frequency. In some cases stimulation triggered seizures were observed. The effects of stimulation were specific to the ANT. Stimulation applied to electrodes placed outside the ANT did not significantly worsen seizure frequency.

Conclusions: ANT stimulation exacerbated seizure frequency in rats with chronic epilepsy following kainate status epilepticus.

Electrical stimulation of the anterior nucleus of the thalamus (ANT) is currently undergoing clinical trial in humans with medically intractable seizures (1). The data supporting efficacy of ANT stimulation are based on the effects of stimulation on delaying seizures produced immediately by chemoconvulsants in rats (2,3), and on small case series indicating improved seizure control in some human patients after implantation of ANT-stimulation electrodes (4,5). One of these case series found that electrical stimulation did not improve seizure control compared with the effect of inserting electrodes in the ANT (4). ANT stimulation has not been tested against any animal models of chronic epilepsy. Accordingly, ANT stimulation was applied to rats made epileptic after kainic acid (KA) status epilepticus (SE). Unexpectedly, it was found that ANT stimulation with different stimulation paradigms increased seizure frequency.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgments
  7. REFERENCES

Male Sprague–Dawley rats (Taconic Farms, NY, U.S.A.) weighing 200–225 g were used in these experime-nts and were housed in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-approved animal facility in an environmentally controlled room (20–23°C, 12-h light/12-h dark cycle, lights on 07:00). Animals were provided water and rat chow ad libitum. Experimental methods were approved by an institutional animal use committee. Experiments were conducted in accord with the National Institutes of Health guidelines for the care and use of laboratory animals.

Surgical anesthesia consisted of a mixture of ketamine (70 mg/kg) and xylazine (6 mg/kg) intraperitoneally. After anesthesia, concentric bipolar electrodes (MS306; PlasticsOne, Roanoke, VA, U.S.A.), with the inner conductor tip extending 1 mm beyond the outer conductor, were sterotaxically implanted in the ANT bilaterally with the incisor bar set at −3.5 mm, and inserted 20 degrees off the sagittal plane. Stereotaxic coordinates for ANT electrodes were anteroposterior (AP), −2.0 mm (posterior to bregma is negative), ±3.5 mm lateral to the sagittal suture, and −6.4 mm ventral to the skull surface (6). In addition, bilateral wire electrodes (E363/3; PlasticsOne) were placed in or near the hilus of the dentate gyrus bilaterally by using coordinates: −4.2 mm (AP), ±2.6 mm (lateral), −3.6 mm (depth), and bilateral cortical screw electrodes were implanted at +2 mm (AP), ±3.0 mm (lateral). A reference screw electrode was placed in the right frontal bone. A ground electrode was placed in the midline anterior to the occipital suture. Electrodes were fixed to the skull by using additional screws and dental acrylic cement. After surgery, rats were placed under a warming lamp until they were able to move around the cage and were then transferred back to the animal housing facility to recover for at least 5 days.

After recovery, rats received KA intraperitoneally to induce SE. Rats were treated with 10 mg/kg KA dissolved in normal saline supplemented by a 5-mg/kg dose after 60 min if seizures with rearing and falling were not observed. Seizures were allowed to continue without intervention except that 3–4 ml saline or lactated Ringers was administered subcutaneously after resolution of the continuous convulsive seizures, usually approximately 3 h. The next day after SE, rats were offered moistened rat chow and then a regular diet.

Rat EEG monitoring was performed in a custom-built eight-cage recording setup similar to the design described by Bertram et al. (7). Rats were housed in cylindrical 10-inch-diameter acrylic cages with acrylic lids. Rats were connected via headcap to a custom connector (PlasticsOne) to a unit-gain preamplifier (HS-27; Neuralynx, Tucson, AZ, U.S.A.), and the buffered signals were passed through a 12-channel commutator (SL-12C; PlasticsOne) to a bank of four eight-channel amplifiers (Lynx-8; Neuralynx). Signals were digitized (NI AT-MIO-64E-3; National Instruments, Austin, TX, U.S.A.) at 200 Hz per channel and acquired and displayed by using commercial EEG software (Harmonie; Stellate Systems, Montreal, Quebec, Canada). Video recordings were made by using a black-and-white video camera and 24-h video recorder. Synchronization of video and EEG was assured by time-stamping the video image (TC-100; BCD Associates, Oklahoma City, OK, U.S.A.).

The first cohorts of rats prepared at the outset of the experiment were placed into monitoring cages at the completion of SE to determine the timing of seizure development. After the initial cohorts, rats were placed in monitoring cages 4 to 8 weeks after KA exposure. Monitoring consisted of continuous EEG and video recordings for periods from 4 to 14 days to characterize baseline seizure frequency, seizure clustering, seizure severity, and changes in seizure frequency during ANT stimulation. Rats were remonitored intermittently over a period of several months to characterize gradual shifts in baseline seizure rate. Seizure severity was scored from video by using a simplified version of the Racine scale (8), because animal position or video resolution did not always permit characterization of subtle limbic seizures consisting of mouth or eye clonus or head bobbing. Seizures were rated as ≤3, no consistently detectable behavior on video; 4, rearing without falling; 5, rearing with falling; or 6, wild running and jumping.

Rats were selected for ANT stimulation and subsequent analysis after a minimum of 5 days of baseline monitoring. Average seizure frequency was determined by dividing the total number of seizures recorded by the monitoring duration in days. Baseline seizure rates were examined for evidence of seizure clustering, defined as one or more seizures daily for 2 or more days followed by 2 or more days without seizures. In rats appearing to have clustered seizures, baseline monitoring was extended to determine the intercluster period. Rats with clustered seizures were included when baseline and ANT stimulation monitoring periods were sufficiently long to include a complete cluster and the intercluster period. Rats developing complications during baseline monitoring or stimulation, such as cap loss, or detected histologically, such as intracerebral infection, were excluded from analysis.

Rats were stimulated bilaterally via concentric bipolar electrodes directed at each ANT. Stimulation pulse trains consisted of 100-μs biphasic rectangular pulses at a rate of 100 Hz (model 2100 Pulse Stimulator; A-M Systems, Sequim, WA, U.S.A.) and were applied either continuously or intermittently by using different stimulation durations and interstimulation intervals. Stimulation parameters, including stimulation frequency, intensity, duration, and interstimulation interval were similar to values used in prior rat or human studies of ANT stimulation (2–5). Constant-current stimulation was used in the initial stimulation trials, but frequently resulted in complications, notably intracerebral abscesses. Voltage stimulation was not associated with these complications, and a majority of rats were stimulated by using voltage. After completion of experiments, it was determined that the increased complications seen with current stimulation were due to incorrect stimulator calibration, which allowed a bias current and asymmetrical positive and negative pulses that were not present with voltage stimulation. The typical impedance of a stimulation electrode was ∼7 kOhm. Stimulation intensity was in the range of either 100 to 300 μA per electrode or 1 to 4 V. This range of voltage intensity corresponded to currents of 140 to 550 μA per electrode. The stimulation intensity was set below the threshold for behavioral or motor effects and was comparable to stimulation intensities reported effective against acute pentylenetetrazol-induced seizures (3). To minimize animal and data loss, as occurred initially because of stimulation-related complications, most rats received only a single stimulation trial, although two rats received two stimulation trials using different parameters. Results were analyzed by using analysis of variance of repeated measures, and each animal served as its own control. The level of significance was preset at p < 0.05.

After completion of seizure monitoring and stimulation testing, rats were deeply anesthetized with urethane (1.5 mg/kg) intraperitoneally. Brains were rapidly removed and frozen in methylbutane cooled on dry ice to −35°C. Coronal slices (40 μm thick) containing the STN and electrode tracts were cut on a cryostat, mounted on glass slides, stained with thionin, and coverslipped. The location of the stimulating electrode tips was identified by light microscopy. The histology of each rat was used to characterize the location of stimulation electrodes with respect to the ANT as a bilateral hit or as a unilateral or bilateral miss. An ANT hit was defined as the placement of either the distal or proximal contact of the concentric bipolar stimulation electrode within the boundary of the ANT.

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgments
  7. REFERENCES

Characterization of chronic seizures after kainic acid status epilepticus

Average daily seizure-frequency and behavioral seizure-severity scores were calculated from cohorts of rats monitored at the same latency after KA SE. Individual rats were not monitored throughout the entire 14-week period after KA administration because prolonged monitoring of a few rats would have precluded monitoring a larger number of rats. Average seizure frequency and severity was calculated only for those latencies after KA SE for which observations from five or more rats were available.

On average, the daily seizure rate after KA administration increased gradually more than twofold over the 14-week period after KA (Fig. 1) from 3.2 seizures per day in the first 2 weeks after KA SE to 7.1 seizures per day in weeks 13 to 14. In individual rats, though, the day-to-day variability in daily seizure rate obscured the gradual changes seen in the average. Beyond 14 weeks, most rats had either received ANT stimulation or had not met criteria to receive ANT stimulation and were no longer monitored. Almost all rats undergoing KA SE developed spontaneous seizures during the monitoring period. Of rats monitored in the first 2 weeks after KA SE, 60% had spontaneous seizures. Most rats had several seizures daily (average, 4.4 seizures/day; median, 2.6 seizures/day). Seizure onset, detected by bilateral intrahippocampal electrodes and bilateral frontal epidural electrodes, was consistently bilateral and seen earliest in the hippocampal leads. Seizure evolution was electrographically stereotyped for each rat, so that the pattern of waxing and waning of the seizure discharge was repeated in each seizure.

image

Figure 1. The average rate and the standard error of seizures per day of a total of 21 rats after kainic acid (KA) status epilepticus (SE) is shown in 2-week intervals. Individual rats were monitored intermittently during the 14-week period after KA SE. Average seizure rates were calculated before anterior nucleus of the thalamus (ANT) stimulation from cohorts of five or more rats, with a large variability between rats. Initial average seizure frequency is three seizures per day in the first 2 weeks after KA administration. Average seizure frequency increases to more than seven seizures per day by 13 to 14 weeks after KA administration.

Download figure to PowerPoint

Initial seizures in the first weeks after SE were usually nonconvulsive, with convulsive seizures occurring after the initial 3 to 4 weeks. Systematic observations of behavioral seizure stage began during baseline monitoring before ANT stimulation. Seizure staging from review of video was performed in 23 rats 3 to 16 weeks after KA SE. Averaged behavioral seizure stages were calculated for weeks 5 to 15 for which data from five or more monitored rats were available. Over the interval of baseline seizures scoring between weeks 5 and 15, seizure severity remained stable, in the range of stages 3 to 4. Once convulsive seizures appeared (stage 3), seizures did not significantly worsen over time. (data not shown)

Effects of ANT stimulation

Of rats in which electrode location could be determined histologically, electrodes were in the ANT bilaterally in nearly 60% of rats. Rats with bilateral ANT hits by histology were included in the analysis of the effects of stimulation on seizure frequency. Rats with stimulation electrodes lying outside the ANT bilaterally (i.e., ANT misses) were used to determine whether the effects of stimulation were specific to the ANT. ANT misses were located mainly in white matter tracts anterior to the ANT and did not include other potentially anticonvulsant thalamic targets, such as the mediodorsal nucleus (MD) (Fig. 2).

image

Figure 2. A: Location of electrode tips within the anterior nucleus of the thalamus (ANT) shown schematically (modified from atlas of Paxinos and Watson (6). B: Location of electrode tips outside of the ANT; for the most part, electrode “misses” were anterior to the ANT, although one rat had electrodes deep to the ANT.

Download figure to PowerPoint

Onset of electrical stimulation of the ANT produced variable acute changes in a few rats. In some rats, stimulation onset produced immediate suppression of interictal spike activity in the hippocampal EEG and a change in the background activity (Fig. 3A). In others, stimulation onset triggered seizure discharges (Fig. 3B). In the majority of rats, however, stimulation onset was not clearly associated with changes in hippocampal EEG.

image

Figure 3. A: Acute EEG effects of anterior nucleus of the thalamus (ANT) stimulation. Stimulation artifact in cortex (CTX) tracing indicates timing and duration of stimulation. At onset of stimulation (left arrowhead), interictal spike activity in the hippocampus was attenuated for the duration of the stimulation. The effect was reproducible and is present in the subsequent stimulation as well (right arrowhead). B: ANT stimulation triggers a seizure. EEG traces from right hippocampus (RHC) and left hippocampus (LHC) and cortex (CTX) show that interictal background and interictal spikes (arrows) disappear at onset of stimulation (arrowhead), replaced by an evolving seizure discharge. ANT stimulation produces a stimulation artifact, seen in the CTX tracing as a thickening of the background and period activity. Stimulation also results in a high-frequency discharge that evolves into a seizure discharge seen best in the hippocampal leads, RHC and LHC.

Download figure to PowerPoint

Prolonged stimulation of the ANT more than doubled the seizure rate in five of eight rats and did not reduce the seizure rate in any rat (Table 1). The increase in seizure rates was specific to stimulation of the ANT, because stimulation applied to rats with electrodes outside of the ANT (Table 2) had a seizure rate comparable to the baseline seizure rate. Because a large number of rats and stimulation trials were excluded from analysis because of complications of stimulation and cap loss, the final number of stimulation trials conducted with a particular stimulation paradigm was small. Nevertheless, no stimulation trial showed a benefit of either intermittent or continuous stimulation. ANT stimulation did not affect behavioral seizure severity (average change in score was −0.2 ± 0.3) or alter the stereotyped electrographic pattern of seizure evolution for a particular rat (data not shown).

Table 1. Stimulation outcome for ANT “hits”
RatBaseline (sz/day)Stimulation (sz/day)Stimulation duration (days)Stimulation parameters% Change
m609.518.3 230 s/2 min191
m915.812.2 720 s/5 min212
m1093.6 4.7 620 s/5 min106
m570.6 1.8 430 s/15 min299
m670.1 0.6 530 s/15 min540
m1063.6 4.7 630 s/15 min287
m995.918.1 6Continuous stim306
m1013.6 4.7 6Continuous stim133
AVG4.1 8.1   5.3 259 ± 51% (SEM)
Table 2. Stimulation outcome for ANT “misses”
RatBaseline (sz/day)Stimulation (sz/day)Stimulation duration (days)Stimulation parameters% Change
m56 2.5 7.0 930 s/5 min276
m76 1.9 0.3 920 s/5 min 18
m9510.612.6 620 s/5 min118
m95 6.910.9 620 s/5 min158
m96 8.5 9.4 620 s/15 min111
m9613.614.2 420 s/15 min104
m50 9.514.4 730 s/15 min152
AVG 7.6 9.8   6.7 134%± 32% (SEM)

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgments
  7. REFERENCES

The results indicate that (a) spontaneous seizures begin in the first weeks after KA SE, becoming more frequent over time, but not more severe; and (b) 100-Hz ANT stimulation exacerbates chronic seizures by increasing seizure frequency. Early seizures after SE have been reported after focal KA injection in hippocampus and amygdala, and after intravenous KA administration (9–11). Hellier et al. (12), however, observed few seizures in the first months after SE, with the first seizure—detected by visual observation—occurring on average after 3 months. Seizure detection using visual observation may have undercounted seizures that occurred rarely or without convulsive symptoms. The milder induction of SE by using repeated small doses of KA, used by Hellier et al., may also have resulted in a longer latent period. Once seizures began, the increase in seizure frequency over time was comparable to the increase observed in this study (12).

Contrary to initial expectations, long-term stimulation of the ANT worsened seizure frequency rather than improving it. Stimulation did not affect the severity of individual seizures, in contrast to the improvement in severity reported in a human case series of five patients (5). The exacerbation of seizure frequency may be the result of numerous factors. First, the extent of similarity between human limbic epilepsy and spontaneous seizures after KA SE in rats is largely uncharacterized. Comparisons of human and rat epilepsy after KA SE have focused on histopathologic similarities in the hippocampus. However, the neuropathologic injury produced by intraperitoneal KA administration is widespread, affecting multiple cortical and subcortical targets (13–15). Second, ANT stimulation may have different effects in a largely normal brain with a single epileptic focus, as perhaps in human temporal lobe epilepsy (TLE), compared with a diffusely injured brain with multifocal or diffuse regions of seizure onset, as exists after KA SE in the rat. Because of widespread neuronal loss and injury in the systemic KA SE model, it is possible that potential anticonvulsant circuits are injured and unable to mediate an anticonvulsant effect of ANT stimulation. After KA SE, the neurons in the lateral ANT are largely preserved, but medial thalamic nuclei, including medial ANT and the adjacent MD nucleus, show significant neuronal loss (13–17). It is not known whether the anticonvulsant activity of the ANT varies by subregion, or whether seizure control may be influenced by adjacent thalamic nuclei that receive current spread from the ANT. Inhibition of the MD, for example, attenuates limbic seizures in rats (18). Extensive neuronal loss in the medial thalamus after kainate SE could potentially disconnect circuits mediating the effectiveness of ANT stimulation. Third, seizure onset in the KA SE rat model appears to be rapidly synchronized bilaterally throughout the limbic circuit, as observed in the current study. Rapid multifocal synchrony of limbic seizure onset has been described in another rat model of chronic limbic epilepsy in rats after SE (19) and may be a characteristic of the species. The relative prominence of the hippocampus, hippocampal commissures, and limbic cortex in the rat brain, relative to the thalamus—the ANT in particular—may facilitate bilateral seizure spread and decrease the effectiveness of ANT stimulation in suppressing or containing limbic seizures. Human limbic epilepsy, in contrast, often develops unilaterally and manifests with focal seizure discharges that evolve ipsilaterally before spreading to the contralaterally. The relative increase of thalamic size and nonlimbic cortex in human brain, and the relative decrease in the size of the hippocampi and hippocampal commissures may increase the effect of the ANT stimulation on limbic activity.

In the current experiments, the increase in seizure frequency was specific to the ANT. Stimulation outside—predominantly anterior—to the ANT did not result in a significant change in seizure rate (see Tables 1 and 2). This comparison reflects the difference between stimulating gray matter of the thalamus and stimulating either white matter tracts anterior to the thalamus, or electrode placement in the ventricles. However, these results do not compare the relative effects of stimulating different thalamic regions.

Sufficient important differences occur between the KA SE rat model of limbic epilepsy and human TLE to temper predictions of the effects of ANT stimulation in humans based on the results described here. However, it is important to note that the difference between detrimental and beneficial effects of ANT stimulation may rest on subtle variables that differ between humans and rats in the present instance but may also vary between different forms of human epilepsy. The extent of underlying neuronal injury, for example, may vary widely between epilepsy subtypes and between individuals with the same type of epilepsy. The observation of detrimental effects in the KA rat model of limbic epilepsy raises the question whether some forms of human epilepsy will be exacerbated by ANT deep-brain stimulation. It is possible that the beneficial or possibly detrimental effects of ANT stimulation will vary depending on epileptic phenotype and on underlying brain injury and may not have the near “universal” applicability seen, for example, with vagal nerve stimulation. If so, identification of appropriate patients will be a critical step in achieving meaningful benefits from ANT stimulation.

Acknowledgments

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgments
  7. REFERENCES

Acknowledgment:  This study was supported by NIH grants NS41340 and NS20253, and a grant from Medtronic. I am grateful to Ms. Hong Wang for excellent technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgments
  7. REFERENCES
  • 1
    SANTE: Stimulation of the Anterior Nucleus of the Thalamus for Epilepsy. http://www.clinicaltrials.gov/ct/show/NCT00101933; 2003.
  • 2
    Hamani C, Ewerton FI, Bonilha SM, et al. Bilateral anterior thalamic nucleus lesions and high-frequency stimulation are protective against pilocarpine-induced seizures and status epilepticus. Neurosurgery 2004;54: 91195, discussion 195–197.
  • 3
    Mirski MA, Rossell LA, Terry JB, et al. Anticonvulsant effect of anterior thalamic high frequency electrical stimulation in the rat. Epilepsy Res 1997;28: 89100.
  • 4
    Hodaie M, Wennberg RA, Dostrovsky JO, et al. Chronic anterior thalamus stimulation for intractable epilepsy. Epilepsia 2002;43: 603608.
  • 5
    Kerrigan JF, Litt B, Fisher RS, et al. Electrical stimulation of the anterior nucleus of the thalamus for the treatment of intractable epilepsy. Epilepsia 2004;45: 346354.
  • 6
    Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates, 2nd ed. New York : Academic Press, 1986.
  • 7
    Bertram EH, Williamson JM, Cornett JF, et al. Design and construction of a long-term continuous video-EEG monitoring unit for simultaneous recording of multiple small animals. Brain Res Brain Res Protoc 1997;2: 8597.
  • 8
    Racine RJ. Modification of seizure activity by electrical stimulation, II: motor seizure. Electroencephalogr Clin Neurophysiol 1972;32: 281294.
  • 9
    Cavalheiro EA, Riche DA, Le Gal La Salle G, et al. Long-term effects of intrahippocampal kainic acid injection in rats: a method for inducing spontaneous recurrent seizures. Electroencephalogr Clin Neurophysiol 1982;53: 581589.
  • 10
    Mascott CR, Gotman J, Beaudet A. Automated EEG monitoring in defining a chronic epilepsy model. Epilepsia 1994;35: 895902.
  • 11
    Cronin J, Dudek FE. Chronic seizures and collateral sprouting of dentate mossy fibers after kainic acid treatment in rats. Brain Res 1988;474: 181184.
  • 12
    Hellier JL, Patrylo PR, Buckmaster PS, et al. Recurrent spontaneous motor seizures after repeated low-dose systemic treatment with kainate: assessment of a rat model of temporal lobe epilepsy. Epilepsy Res 1998;31: 7384.
  • 13
    Ben-Ari Y, Tremblay E, Riche D, et al. Electrographic, clinical and pathological alterations following systemic administration of kainic acid, bicuculline or pentetrazole: metabolic mapping using the deoxyglucose method with special reference to the pathology of epilepsy. Neuroscience 1981;6: 13611391.
  • 14
    Covolan L, Mello LE. Temporal profile of neuronal injury following pilocarpine or kainic acid-induced status epilepticus. Epilepsy Res 2000;39: 133152.
  • 15
    Schwob JE, Fuller T, Price JL, et al. Widespread patterns of neuronal damage following systemic or intracerebral injections of kainic acid: a histological study. Neuroscience 1980;5: 9911014.
  • 16
    Brandt C, Potschka H, Loscher W, et al. N-Methyl-d-aspartate receptor blockade after status epilepticus protects against limbic brain damage but not against epilepsy in the kainate model of temporal lobe epilepsy. Neuroscience 2003;118: 727740.
  • 17
    Vass K, Berger ML, Nowak TS Jr, et al. Induction of stress protein HSP70 in nerve cells after status epilepticus in the rat. Neurosci Lett 1989;100: 259264.
  • 18
    Bertram EH, Mangan PS, Zhang D, et al. The midline thalamus: alterations and a potential role in limbic epilepsy. Epilepsia 2001;42: 967978.
  • 19
    Bertram EH. Functional anatomy of spontaneous seizures in a rat model of limbic epilepsy. Epilepsia 1997;38: 95105.