Reduction of seizures with low-dose clonazepam in children with epilepsy

Introduction 

Clonazepam has been used routinely in the treatment of epilepsy in children for several years. Controlled and uncontrolled studies have confirmed its efficacy in reducing seizure frequency in epilepsies with different types of seizures and epileptic syndromes in children: absence, myoclonic, atonic, and partial epilepsy [1], [2], [3], [4], [5], [6], [7], [8], [9]. No correlation between dose and clinical effect has been demonstrated. In studies in children, responders and nonresponders were found among those with high doses and plasma levels of clonazepam and those with low ones. A clear-cut minimal therapeutic plasma level has not been established [1], [10].

The main limitations in the use of clonazepam have been side effects and the development of tolerance to its antiepileptic efficacy [3]. Many authors consider the side effects to be dose-dependent [1], [2] and found them to be diminished when the dose was reduced [3], [9].

In our clinical experience in the treatment of children with epilepsy with clonazepam, seizure reduction has often been achieved even at low doses in the early phase of the titration period. If the low doses are kept at approximately 0.02 mg/kg/day, the side effects have been markedly diminished compared with the usually recommended higher doses [10], [11] and the clinical response often persists. In an earlier report, quantified epileptiform activity in children with epilepsy could be significantly reduced at this dose level [12]. To our knowledge, no studies of the effects on seizures with consistent use of this dose level are available. The aim of this study was to examine whether these low doses of clonazepam affect seizure frequency. It was conducted as an open study of the acute effects of clonazepam in children with generalized and partial epilepsy.

Patients and methods 

Patients 

The children, who were patients in the Pediatric Epilepsy Unit of Karolinska Hospital, were enrolled because of a need to revise antiepileptic drug treatment due to unsatisfactory seizure control or naive patients begun on antiepileptic drugs. Patients were included in the study if they were 6 months-18 years of age, had a diagnosis of epilepsy, had two or more seizures per day, received concomitant antiepileptic drug therapy in appropriate doses with stable plasma levels, had no concomitant treatment with benzodiazepines, and received no emergency treatment with diazepam within 8 days before entering the study. The study was approved by the Karolinska Hospital Ethics Committee. The informed consent of all parents was obtained and, when possible, also of the patients.

The study included 19 children, including 11 males and 8 females (Table 1). The median age was 3.2 years (range = 0.7-18.3 years). The onset of epilepsy occurred at a median of 0.8 years (range = 0.1-3.6 years). The types of epilepsies and seizures and epileptic syndromes were defined according to the classification of the International League Against Epilepsy including seizure semiology and electroencephalogram characteristics before study entry [13], [14]. Generalized epilepsy was found in 15 of the children, whereas four had partial epilepsy. A computed tomography scan and/or magnetic resonance imaging of the brain was performed in all children. Tests for neurometabolic disorders, infectious diseases, and chromosomal aberrations were performed in a large majority of children. An epileptic syndrome could be defined in 11 children and an etiology could be identified in six. Children in which an etiology could not be detected, although they had other signs of central nervous system pathology, were considered to be cryptogenic. Before entering the study the patients had been on long-term treatment with a mean of 2.8 antiepileptic drugs (range = 0-9). At the time of study the children were on maintenance treatment with a mean of 1.1 antiepileptic drugs (range = 0-2). Fourteen of the children had mental retardation (mild, n = 5; severe, n = 9). Other associated neurologic disorders, such as autism spectrum disorder, were found in five children and motor handicaps were demonstrated in five.

Table 1. Demographic data on the children in the study group

	Patient No.	Age (yr)	Sex	Epilepsy Onset (yr)	Epilepsy Type	Seizure Type(s)	Epileptic Syndrome/Etiology	Previous AEDs (No.)	Concomitant AED(s)
	1	0.7	M	0.3	Gen	Infantile spasms	Infantile spasms; idiop	2	VPA
	2	0.7	M	0.5	Gen	Infantile spasms	Infantile spasms; HIE	1	VPA
	3	2.8	M	0.2	Gen	Tonic, Myocl, Infantile spasms	Lennox-Gastaut syndrome	6	LTG, VPA
	4	3.2	F	0.3	Gen	Aton, Tonic, CP, 2°GTC	Lennox-Gastaut syndrome	7	VGB
	5	3.2	M	2.7	Gen	Myocl-astat	Myoclonic-astatic; idiop	0
	6	3.8	F	3.6	Gen	Myocl-astat, GTC	Myoclonic-astatic; idiop	2	VPA, CBZ
	7	0.8	M	0.8	Gen	Myocl-astat	Myoclonic-astatic; crypt	0
	8	2.3	M	2	Gen	Myocl-astat	Myoclonic-astatic; crypt	2	VPA
	9	3.6	F	0.8	Gen	Myocl-astat	Myoclonic-astatic; crypt	7	PRM, ESM
	10	6.9	M	1.1	Gen	Myocl-astat	Myoclonic-astatic; crypt	2	LTG, VPA
	11	18.3	F	3	Gen	Ab, GTC	CAE	3	ESM, CBZ
	12	1.5	F	0.8	Gen	Aton	Angelman syndrome	1	VPA
	13	0.7	F	0.6	Gen	Tonic	Contusio cerebri	0
	14	1.3	M	0.9	Gen	Tonic	Cryptogenic	1	CBZ
	15	4.9	M	0.8	Gen	Tonic	Cryptogenic	6
	16	1.3	F	0.4	Partial	Atyp Ab, CP, 2°GTC	Tuberous sclerosis	2	VGB, CBZ
	17	3.6	M	2.8	Partial	Tonic, CP, 2°GTC	HIE	2	CBZ
	18	9.6	M	0.3	Partial	Tonic, 2°GTC	Postencephalitic	9	VPA
	19	12.8	F	0.1	Partial	SP, CP	Cryptogenic	1	CBZ

*Epilepsy and seizure types according to the International Classification of the International League Against Epilepsy. Abbreviations Tb  Absence   AEDs  Antiepileptic drugs   Aton  Atonic   AtypAb  Atypical absence   CAE  Childhood absence epilepsy   CBZ  Carbamazepine   CP  Complex partial   Crypt  Cryptogenic   ESM  Ethosuximide   Gen  Generalized   GTC  Generalized tonic-clonic   2°GTC  Secondary generalized tonic-clonic   HIE  Hypoxic-ischemic encephalopathy   Idiop  Idiopathic   LTG  Lamotrigine   Myocl  Myoclonic   Myocl-astat  Myoclonic-astatic   PRM  Primidone   SP  Simple partial   VGB  Vigabatrin   VPA  Valproic acid

Study design 

The children were inpatients in the epilepsy unit of the neuropediatric ward throughout the study, which consisted of two consecutive 24-hour observation periods with seizure recording. The first 24-hour period was regarded as baseline: the control day. In the morning of the second day an intramuscular injection of low-dose clonazepam was administered and was followed by another 24-hour period of observation. During these 2 days the children were observed by nurses belonging to a group especially trained in the detection of seizures. The types and numbers of seizures were recorded on a seizure calendar. For practical reasons, observations with seizure recordings could not be made during the complete 24-hour period in all patients (Table 2). When the observation had to be restricted in time, the total number of hours, as well as the real time of day used to record seizures during the two examination days was exactly the same for each individual child. All time-restricted observation periods (less than 24 hours) took place within 12 hours after the injection. Play activities and the timing of meals and sleep were kept as similar as possible.

Table 2. Total number of seizures and their relative changes determined during corresponding periods before (control) and after administration of low-dose clonazepam. The maximal plasma levels of clonazepam and the total time of seizure recording during each day are given

	Patient No.	Control	Clonazepam	Relative Changes (%)	Maximal Levels of Clonazepam (nmol/L)	Total Recording Time (hr)
	1	113	7	−94	16	24
	2	51	73	+43	17	24
	3	22	8	−64	20	24
	4	15	10	−33	23	24
	5	23	4	−83	26	24
	6	33	10	−70	22	4
	7	20	0	−100	32	12
	8	180	0	−100	41	8
	9	70	53	−24	27	24
	10	15	18	+20	30	10
	11	2	0	−100	24	4
	12	27	1	−96	13	4
	13	14	6	−57	11	4
	14	90	59	−34	25	24
	15	25	17	−32	30	24
	16	12	5	−58	20	24
	17	7	2	−71	23	24
	18*	1	0	(−100)	19	24
	19*	1	0	(−100)	27	24

* Results regarding seizure frequency are not included in the statistical analysis.

On the morning of the second examination day, blood samples were drawn to determine the plasma concentrations of the concomitant antiepileptic drugs before intake of the morning dose. After this an injection of clonazepam was administered intramuscularly. The dosages of clonazepam were 0.01 (n = 2; Patients 12,13), 0.02 (n = 14), 0.03 (n = 1; Patient 8), and 0.04 (n = 2; Patients 1,7) mg/kg. Blood samples for the determination of plasma levels of clonazepam were drawn before and 0.5, 1, 1.5, 2, 4, 8, 12, and 24 hours after administration of clonazepam. The concentrations of clonazepam were determined by means of high-performance liquid chromatography, using methods modified described by Rovei and Boukhabza [15], [16]. The internal standard was N-methylnitrazepam. The coefficient of variance within the run was 7.5%. The between-day coefficient of variance was 13% at 15 nmol/L and 7% at 150 nmol/L. The limit of detection of clonazepam was 2.5 nmol/L and the limit of quantitation, 10 nmol/L.

Results 

The total number and the relative changes of seizures recorded during control days and days after clonazepam injection are given for all patients in Table 2. There was an increased number of seizures in two children after clonazepam administration (Patients 2 and 10), whereas the number was decreased in 17 children. The median number of seizures on control days was 22 (range = 1-180) and, on days after clonazepam injection, the median was 6 (range = 0-73). The relative changes demonstrated a median of −70% with a range from −100% to + 43%. Although one of the inclusion criteria included two or more seizures per day, two children had only one seizure per day on control days (Patients 18 and 19). Therefore these two children were not included in the statistical analyses. A nonparametric statistical test, the Wilcoxon test for paired data, was used because the data did not demonstrate a normal distribution. A highly significant reduction of seizures was found after clonazepam, P = 0.0031. The Wilcoxon test also provided significant reductions in the patients on the dose of 0.02 mg/kg (P = 0.0342).

Figure 1 shows the plasma concentrations of clonazepam taken during the 24-hour period after the injection for all 19 children. The median value of the maximal plasma levels of clonazepam was 23 nmol/L, with a range of 11-41 nmol/L. The maximal plasma level was reached with a mean delay of 1.7 hours (0.5-4 hour). Table 2 presents the maximal plasma concentrations of clonazepam for each child. To study the relationship between the maximal plasma levels of clonazepam and the relative changes in seizure frequency, the regression line, and correlation coefficient of the two variables were analyzed. No correlation was found, however, between the maximal plasma level and effect.

View Large Image Download to PowerPoint

Figure 1. Plasma concentrations of clonazepam from blood samples drawn at 0.5, 1, 1.5, 2, 4, 8, 12, and 24 hours after a single intramuscular injection of clonazepam in a low dose. The box and whisker plots demonstrate the median, interquartile range, and complete range for the 19 children in the study. The number of blood samples included on each occasion and the number of samples below the detection limit of 10 nmol/L (omitted in the figure) are indicated.

The total time of seizure recording during the examination days are provided in Table 2 for each child. A 24-hour seizure-recording period was achieved each day in 12 children, and a period of 4-12 hours in the remaining ones. The mean observation time for all children was 17.6 hours per day.

The plasma levels of all the concomitant antiepileptic drugs except vigabatrin were determined. The concentrations found were within the usually recommended therapeutic ranges in all children except Patient 3, who had a slightly elevated level of valproate (952 micromol/L, the therapeutic range used was 350-700 micromol/L; dosage 36 mg/kg body weight).

Some children experienced mild sedation after clonazepam, although no other adverse effects were evident.

Discussion 

In this study of children with epilepsy on the acute effects of a single dose of clonazepam in the low-dose range on seizure frequency, we found a significant reduction. The concomitant plasma concentrations of clonazepam were low. Thus inhibition of seizure activity seems to be achieved even at low plasma levels of the drug. In the present study the results pertaining to seizure reduction were in agreement with those of previous studies, although they were achieved at consistently lower doses and plasma levels of the drug [4], [5], [6], [7], [8], [9].

A good correlation between the dose and plasma level of clonazepam has been demonstrated [4], [9] but not between the dose or plasma level and the clinical effect [10]. No studies evaluating consistently low doses are available. In most long-term studies on oral clonazepam, the dose has been increased rapidly according to a titration schedule and longer planned evaluation periods at different dose levels have not been examined. Clinical efficacy at lower doses may therefore not have been detected [8]. Responders and nonresponders are found among both those who received a high-dose and those with a lower one. Even though previous studies include patients with a good response to clonazepam in the low-dose range, similar to the doses we used, these studies have not been designed to determine if the responses obtained at higher doses could have already been obtained at low doses. The proposed therapeutic dose regimen has been to start with 0.01-0.03 mg/kg with increments approximately every third day up to 0.1-0.2 mg/kg or until clinical response is obtained [1], [10], [11]. The therapeutic plasma concentration range has been suggested to be 48-190 nmol/L [17], 16-222 nmol/L [1], [10], 65-230 nmol/L [11], 80-240 nmol/L [18], and 63-254 nmol/L [19]. The results of the present study suggest to start treatment at low doses of clonazepam and evaluate the individual effect carefully during dose escalation aiming at lowest possible dose with therapeutic effect.

Studies have indicated that the protection of seizures usually lasts for up to 24 hours after an intravenous dose of clonazepam [20]. Thus our expectations were that the 24-hour seizure evaluation after the single intramuscular dose of clonazepam would mirror the time frame of the effect of clonazepam. In many children the plasma levels were low at 12 and 24 hours, but the effect seemed to persist, which likely reflects longer pharmacodynamic effects at the receptor level than would be indicated by the plasma levels.

At the start of the present study the doses of 0.01-0.04 mg/kg were titrated (see Patients and Methods). The dose of 0.02 mg/kg was found to provide plasma levels of 30 nmol/L or less, which, in our clinical experience, usually gives few side effects, and this dose was used subsequently in 14 of the patients. The three children included on a dose of 0.03-0.04 mg/kg had plasma levels below the usually recommended range [10], [11], [17], [18], [19]. Within the narrow plasma level range used in this study, no correlation could be found between plasma concentration and clinical effect.

Changes or interactions of the concomitant antiepileptic drugs could be a source of error. The same dose of the concomitant drugs was taken on each of the study days and was administered by a nurse. The plasma levels of the antiepileptic drugs were within the generally recommended therapeutic range except in one patient who had moderately elevated valproate levels. Pharmacokinetic interactions of clonazepam with the other antiepileptic drugs cannot be excluded, although this has not been established [3], [21].

Diurnal variation in seizure frequency among epilepsy patients is a well known phenomenon [22]. Therefore in this study, the observations of each individual were made at the same real time of day.

The two children in this study who demonstrated an increase in seizure frequency did not differ from the others regarding age, seizure type, or type of epileptic syndrome (infantile spasms and myoclonic-astatic epilepsy). However, it is well known, that clonazepam may provide paradoxic responses in some children [3].