Response of status epilepticus induced by lithium and pilocarpine to treatment with diazepam

EXPERIMENTALNEUROLOGY

101,267-275

(1988)

Response of Status Epilepticus induced by Lithium and Pilocarpine to Treatment with Diazepam NANCY

Y. WALTON

AND DAVID

M. TREIMAN’

Neurology and Research Services, Veterans Administration West Los Angeles Medical Center, LOS Angeles, California 90073; and Department of Neurology, UCLA School of Medicine, Los Angeles, California 90024 Received September 1, 1987; revision received December 18, 1987 Status epilepticus (SE) was induced in rats by administration of 3 mmol/kg lithium chloride followed 24 h later by injection of 25 m&g pilocarpine. Treatment with 20 mg/kg diazepam was initiated at the time each of four EEG patterns was seen: (i) discrete electrographic seizures; (ii) waxing and waning epileptiform activity; (iii) continuous, high-amplitude, rapid spiking; and (iv) periodic epileptiform discharges (PEDs) on a relatively flat background. Success of diazepam in stopping all seizure activity was predicted by the EEG pattern seen at the time of treatment. All rats treated while displaying discrete electrographic seizures had status stopped with diazepam, but only three of six with waxing and waning epileptiform activity and one of six each with continuous spiking and PEDS. Rats which continued to seize had a decrease in spike amplitude of 74.8 f 18.25% following diazepam injection. These data confirm the clinical impression that the longer the duration of status epilepticus, the more difficult it is to control and suggest that the EEG pattern at the time of treatment predicts the probability of success. Q 1988 Academic press, IX. INTRODUCTION

Administration of small doses of pilocarpine to rats pretreated with lithium chloride has been reported to cause status epilepticus (SE) in rats which is almost universally fatal within 24 h (3-5). Atropine administered prior to pilocarpine completely blocks the seizure activity in this model, and also Abbreviations: HPLC-high performance liquid chromatography, PEDs-periodic epileptiform discharges, SE-status epilepticus. ’ This work was supported by grants from the Epilepsy Foundation of America and the Veterans Administration. We would like to thank Mr. David L. Birken for outstanding technical assistance in the conduct of these experiments. 267 OOl4-4886/88 $3.00 Copyright 0 1988 by Academic M Inc. All rights of reproduction in any form reserved.

268

WALTON

AND

TREIMAN

stops ongoing seizures when administered shortly after SE onset (3-5). But if atropine administration is delayed until 20 min afier initial forelimb clonus, it has no effect on the ongoing SE (4), suggesting that although cholinergic stimulation is clearly responsible for initiation of SE in this model, other neurotransmitter systems must be responsible for its maintenance. Diazepam has been reported to prevent SE onset if given prior to pilocarpine and to arrest SE if given shortly after seizures begin (4-6). Olney et al. (6) report that 20 mg/kg diazepam stops status induced by lithium and pilocatpine when administered 60 min after pilocarpine. In contrast, Monisett et al. (5) report this same dose of diazepam administered at the same time to be ineffective in stopping SE. We have previously described a sequence of progressive changes in the EEG pattern seen during SE in man and three experimental models of SE in the rat, including lithium plus pilocarpine (9, 10). SE begins with discrete electrographic seizures which end abruptly and simultaneously in all channels, with a low-voltage slow postictal pattern. The EEG then shifts to a pattern in which the distinct postictal pattern is no longer seen and epileptiform activity waxes and wanes in amplitude and frequency, never disappearing altogether. This pattern is followed by one in which continuous, high-amplitude, rapid spiking occurs. Periods of isoelectric EEG may occur at any time during the continuous spiking. These flat periods gradually increase in frequency and duration until the final pattern appears-periodic epileptiform discharges (PEDs) on a relatively flat background. We have found that use of 30 mg/kg pilocarpine (as described in published reports) in this model causes the EEG progression through the first two patterns to be very rapid, usually less than 10 min. If the pilocarpine dose is decreased to 25 mg/kg, this transition time can be extended with only minimal decrease in the percentage of animals seizing. By treating SE induced by lithium and pilocarpine with diazepam at the time various EEG patterns are seen, insight may be gained as to the relationship of the EEG pattern to responsiveness of SE to treatment and to the transmitter system(s) involved in maintenance of SE in this model. METHOD Thirty adult male Sprague-Dawley rats obtained from Charles River were used as subjects in this experiment. Rats were prepared surgically under general anesthesia (87 mg/kg ketamine plus 13 mg/kg xylazine) with four epidural recording electrodes made from No. O-80 X i-in. stainless steel screws. Rats were housed individually after surgery and allowed to recover for 1 week. Lithium chloride, 3 mmol/kg, was dissolved in water and injected i.p. 24 h prior to injection of pilocarpine, 25 mg/kg, dissolved in saline and given i.p. The EEG was monitored continuously throughout the experiment.

DIAZEPAM

F4 -

TREATMENT

OF STATUS

269

EPILEPTICUS

P4

P4 - P3 P3 -

F3

F3 -

F4

(b)

(d)

P3 - F3 F3 - F4

--C_

‘-----A I

500 pv 2

set

FIG. 1. Progressive changes in EEG pattern during status epilepticus induced by lithium and pilocarpine. a-Discrete electrographic seizures, EEG recorded 1 min after onset of SE. bWaxing and waning epileptiform activity, EEG recorded 10 min after onset of SE. c-continuous, high-amplitude, rapid spiking, EEG recorded 34 min after onset of SE. d-Periodic epileptiform discharges on a relatively flat background, EEG recorded 6 h 6 min after onset of SE.

SE onset was determined by the time at which clearly identifiable epileptiform activity was first seen on the EEG. Diazepam (Elkins-Sinn, Inc.), 5 mg/ ml, was administered as a 20 mg/kg bolus i.p. at the time the following EEG patterns were seen on the EEG: (i) discrete electrographic seizures (Fig. la);

270

WALTON

AND

TREIMAN

(ii) waxing and waning epileptiform activity (Fig. lb); (iii) continuous, highamplitude, rapid spiking (Fig. lc); and (iv) periodic epileptiform discharges (PEDs) on a relatively flat background (Fig. Id). Six animals were treated at each EEG pattern. Six additional animals served as untreated controls. Visual and EEG observation continued for at least 30 min after treatment, at which time a blood sample was obtained by cardiac puncture from 12 animals. Diazepam concentration was determined by HPLC (2). The remaining animals were observed for varying periods of time beyond 30 min and were given 100 mg/kg phenytoin subcutaneously before being returned to their home cages overnight. RESULTS Data are presented as means f standard deviations unless noted otherwise. Latency from pilocarpine to onset of SE was 26.9 f 9.11 min, with a range from 13 to 46 min. Untreated controls displayed typical Class II to V limbic seizures (7) throughout the episode of SE. Motor seizure symptoms typically were more prominent at the beginning and end of the status episode, with a long period of only head bobbing and/or unilateral forelimb clonus seen during the period of continuous spiking in the middle of the episode of SE. Generalized tonic-clonic seizures were seen only rarely in this model. When seen, they occurred at the onset of SE, during the period of discrete electrographic seizures on the EEG. Four of the six untreated controls died 2.5 to 4 h after SE onset, typically following a wild “running fit.” The other two rats survived longer, up to 25.5 h before dying. Diazepam serum concentration was 146 1 + 108.14 rig/ml, with samples drawn 36.3 + 1.23 min after diazepam injection. The duration of SE at time of treatment and the response to treatment for each group are shown in Table 1. While the EEG pattern during SE is clearly related to the duration of SE, considerable overlap occurred among durations seen for the first three patterns, as can be seen in Fig. 2. Continuous rapid spiking lasted from 72 to 15 1 min before the first appearance of PEDs. All six rats treated while displaying the discrete electrographic seizure pattern stopped seizing within 2 min after diazepam injection. Only three rats whose EEG displayed the waxing and waning pattern, one of the rats with continuous spiking, and one of the rats with PEDs were successfully treated, with cessation of all behavioral and electrical seizure activity. One additional rat with PEDs ceased all seizure activity following diazepam injection, but resumed electrographic seizure activity by 20 min after the injection. Two animals from each of the waxing and waning and continuous groups and four from the PEDs group converted to SE with only subtle motor symp toms or no motor symptoms at all following treatment, but with continua-

DIAZEPAM

TREATMENT

OF STATUS

271

EPILEPTICUS

TABLE 1 Duration of SE prior to Diazepam treatment and Response to Treatment for Each of the Four EEG Pattern Groups Response to treatment with 20 mg/kg diazepam i.p.

EEG group

Duration of SE prior to treatment (min)’

Discrete seizures Waxing and waning Continuous spiking PEDs

7.3 + 16.2+ 36.7 + 127.0 f

All seizures stopb

Convert to subtle or electrographic SE

616

2.57 5.05 15.51 10.34

316

116 116

Overt SE continues

016 W 216

016 116

516

O/6

316

a Values are means k SD. b No further behavioral or electrical seizure activity.

tion of ictal discharges on the EEG (Fig. 3). Subtle motor symptoms seen included very small clonic movements of a front paw or rhythmic tremors of the mouth or facial muscles accompanied by epileptiform discharges on

1

discrete

seizures

45 Duration

of status

epileptiius

(mini

FIG. 2. Percentage of rats displaying each of the first three EEG patterns from onset of SE to 45 min SE duration. Data is from untreated controls and rats treated at continuous spiking and PEDs. (N = 18).

DIAZEPAM

TREATMENT

OF STATUS

EPILEPTICUS

273

the EEG. Rats with electrographic SE were comatose, without movement of any kind except for respiration, but with continuing epileptiform activity. One rat with a waxing and waning EEG pattern and three with continuous spiking continued to have overt seizures following treatment with diazepam. Although overt SE was stopped for all rats treated while showing PEDs, only three of these animals survived their status episode, all with continuing subtle or electrographic seizure activity. This is in marked contrast to animals treated, even unsuccessfully, at the earlier SE patterns. All of these rats survived, although they required maintenance anticonvulsant therapy to prevent recurrent seizure activity. In order to make a statistical assessment of the likelihood that the treatment outcomes we observed were due to chance, a Kruskal-Wallis one-way analysis of variance was performed (8). Treatment outcomes were ranked as to degree of success in terminating SE according to the following scheme: (i) if all seizure activity stopped, time from treatment to seizure cessation determined ranking; (ii) when seizure activity continued, electrographic SE was assigned a ranking higher than that of subtle motor SE, which ranked higher than continuing overt SE; and (iii) tied rankings within a given type of continuing SE were broken by time for cessation of other seizure types and then by percentage decrease in spike amplitude which remained. The analysis of variance revealed significant overall differences among the groups, H = 13.93, P < 0.01. The Spearman correlation coefficient of SE duration with effectiveness of treatment was 0.6 1, P < 0.0 1. Maximum spike amplitude seen during SE in this model always exceeded 2000 pV, and reached 6000 PV for some animals. All of the animals which continued to seize following treatment with diazepam showed a marked decrease in spike amplitude following treatment, 74.8 f 18.25%. The decrease in amplitude occurred over 10 to 30 min following treatment and remained stable once the new plateau amplitude was reached. DISCUSSION Our results confirm the clinical observation that SE becomes more difficult to treat the longer its duration at the time treatment begins. More interesting, FIG. 3. Effect of diazepam treatment on EEG in four different subjects. Time and sensitivity calibrations are indicated for each recording. A-recorded 1 min before (left tracing) and 5 min following (right tracing) diazepam injection. Overt SE continues. B-recorded I min before (left tracing) and 7 minutes following (right tracing) diazepam injection. Overt seizure symptoms have stopped, rat is lying down with small clonic movements of the toes in both front paws as the only remaining motor symptom of the ongoing “subtle” SE. C-recorded 1 min before (left tracing) and 10 min after (right tracing) diazepam injection. Rat is comatose, with only respiratory movements. Electrographic SE continues. D-recorded 1 min before (left tracing) and 3 min following (right tracing) diazepam injection. All seizure activity has ceased.

274

WALTON

AND TRFIMAN

however, is the fact that the EEG pattern seen at the time diazepam is administered predicts the success of treatment. SE induced by lithium and pilocarpine is arrested quickly if diazepam is administered while the EEG displays discrete electrographic seizures. The percentage of rats successfully treated is reduced by half if diazepam treatment is delayed until waxing and waning epileptiform activity is seen on the EEG. Only one of six rats treated while continuous spiking or PEDs occurred on the EEG had a complete arrest of seizure activity. The EEG pattern displayed during SE in this model is clearly a function of SE duration. There is, however, a “window,” from 10 to 40 min SE duration, when any one of the first three EEG patterns may be seen. It may be possible to manipulate this model so that the separate predictative weights of the EEG patterns and SE duration could be determined. Although these two factors are confounded in the present data, the overlap in SE duration among subjects displaying the first three SE patterns did not reduce the power of the EEG pattern to predict treatment outcome. The rather modest observed correlation between SE duration and effectiveness of treatment suggests that other factors are also important in predicting the outcome of treatment. The time course over which diazepam can be successfully used to stop SE in this model is similar to that reported for atropine (4). Maintenance of SE beyond the point where cholinergic systems are required probably involves recruitment of the excitatory amino acid transmitter systems within the cortex and its projections. While GABA is thought to be the inhibitory transmitter of cortical intemeurons, the lack of success of diazepam in stopping SE once the EEG has progressed to continuous spiking suggests that even maximal enhancement of endogenous GABA inhibition is not sufficient to shut down the excitatory overload. The very substantial decrease in spike amplitude seen following diazepam injection appears to be the most reliably produced effect of the treatment. Spike frequency is unchanged, as is the morphology of the epileptiform activity. Diazepam serum concentrations achieved following 20 mg/kg greatly exceed those reported to be effective in treating human SE (1) as well as those we have found to be effective in treating SE induced by homocysteine in cobalt-lesioned rats (10). Although we found considerable variation among rats given the same diazepam dose, even the lowest serum concentration exceeded 900 rig/ml. Our results clearly contradict those reported by Olney et al. (6), and confirm those of Morrisett et al. (5), namely, that diazepam treatment late in an episode of SE induced by lithium and pilocarpine is unlikely to stop all seizure activity. Seven of our rats were treated 50 to 70 min after pilocarpine injection. Only two of these animals stopped seizing following diazepam

DIAZEPAM

TREATMENT

OF STATUS

EPILEPTICUS

275

treatment, and two others converted to only subtle motor seizure symptoms. Diazepam, 20 mg/kg, the same dose used by Olney et al. (6), produced profound sedation in all treated rats, and it is possible that this effect might be misinterpreted as arrest of seizure activity in animals not monitored by EEG. EEG monitoring is essential in evaluation of the results of treatment, in both animal models and clinical practice. Electrographic evidence of SE may persist after all motor symptoms have ceased as can be seen in the EEGs published by Morrisett et al. (5) which are captioned “Interruption of lithium/pilocarpine status epilepticus by paraldehyde.” The EEG recorded 100 min after paraldehyde treatment shows a decrease in spike amplitude and frequency, but no cessation of continuous spikes. We would not interpret continuous spiking nearly two h after treatment to be an “interruption” of SE. SE induced by lithium and pilocarpine may prove to be an excellent model of the lo-20% of human SE cases which are not responsive to the usual anticonvulsant therapy. The appearance of continuous spiking or PEDs prior to treatment may be a predictor of drug-resistant SE, although this effect has yet to be demonstrated in human SE. REFERENCES 1. DELGADO-ESCUETA, A. V., AND F. ENRILE-BASCAL. 1983. Combination therapy for status epilepticus: Intravenous diazepam and phenytoin. Pages 477-485 in A. V. DELGADOESCUETA, C. G. WASTERLAIN, D. M. TREIMAN, AND R. J. PORTER, Eds., Advances in Neurology, Vol. 34. New York, Raven Press. 2. GUNAWAN, S., AND D. M. TREIMAN. 1988. Determination of lorazepam in plasma of patients during status epilepticus by high performance liquid chromatography. Ther. Drug. Monit. 10: 172-l 76. 3. HONCHAR, M. P., J. W. OLNEY, AND W. R. SHERMAN. 1983. Systemic cholinergic agents induce seizures and brain damage in lithium-treated rats. Science 220: 323-325. 4. JOPE,R. S., R. A. MORRISETT, AND 0. C. SNEAD. 1986. Characterization oflithium potentiation of pilocarpine-induced status epilepticus in rats. Exp. Neural. 91: 47 I-480. 5. MORRISE~, R. A., R. S. JOPE,AND 0. C. SNEAD, III. 1987. Effects of drugs on the initiation and maintenance of status epilepticus induced by administration of pilocarpine to lithium pretreated rats. Exp. Neurol. 97: 193-200. 6. OLNEY, J. W., M. P. HONCHAR, AND W. R. SHERMAN. 1983. Diazepam prevents lithiumpilocarpine neurotoxicity in rats. Sot. Neurosci. Abstr. 9: 40 1. 7. RACINE, R. J. 1972. Modification of seizure activity by electrical stimulation. II. Motor seizure. Electroencephalogr. Clin. Neurophyiol. 32: 28 l-294. 8. SIEGEL, S. 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. 9. TREIMAN, D. M., N. Y. WALTON, C. L. WICKBOLDT, AND C. M. DEGIORGIO. 1987. Predictable sequence of EEG changes during generalized convulsive status epilepticus in man and three experimental models of status epilepticus in the rat. Neurology 37 (Suppl. 1): 244-245. 10. WALTON, N. Y., AND D. M. TREIMAN. 1988. Experimental secondarily generalized convulsive status epilepticus induced by D,L-homocysteine thiolactone. Epilepsy Res. 2: 79-86.