Clinical pharmacokinetics of antiepileptic drugs in paediatric patients. Part I: Phenobarbital, primidone, valproic acid, ethosuximide and mesuximide.

This article reviews 119 papers published since 1964 on the pharmacokinetics of phenobarbital, primidone, valproic acid, ethosuximide and mesuximide (methsuximide) in paediatric patients. Particular attention has been paid to the role of age in determining the variability of pharmacokinetic parameters, but the effect of other factors, such as different formulations and routes of administration, concomitant treatments, gender and pathological conditions other than epilepsy, have also been considered. Mean phenobarbital terminal half-life (t1/2z) is very long in neonates (45 to 409 hours) and decreases with age. Therefore, a low dose per kilogram (dose/kg) is recommended during the neonatal period. The dose requirement decreases with increasing age, especially in children also taking valproic acid, which inhibits phenobarbital metabolism. Primidone is metabolised to phenobarbital and phenylethylmalonilamide; the metabolic conversion rate is increased by enzyme-inducing drugs and inversely correlated with age, being virtually absent in neonates. Valproic acid is extensively bound to plasma proteins, but there is a high interindividual and intraindividual diurnal variability in the binding, which depends on the concentration of binding proteins (i.e. albumin) and binding modulators (e.g. free fatty acids) but not on age (at least in those patients aged between 3 months and 65 years). The clearance (CL/F) of valproic acid positively correlates with the unbound concentrations and is strongly age-dependent, being low in neonates and high at the end of the first postnatal month, and progressively decreasing from 2 months to 14 years. The combination of these factors leads to a very poor correlation between plasma concentrations and dose/kg (C/D) and between plasma concentrations of total valproic acid and efficacy. Children also taking enzyme-inducing antiepileptic drugs require a larger valproic acid dose/kg, whereas the coadministration of aspirin (acetylsalicylic acid) may decrease the clearance of unbound drug (CLu/F), and thus require a decrease in the daily dose of valproic acid. Ethosuximide is well absorbed, minimally protein bound and slowly eliminated. Lower C/D ratios are reported in children younger than 10 years old than in older children and in individuals also taking enzyme-inducing drugs (i.e. primidone). According to the only available paper on mesuximide in paediatric patients, the C/D ratio is less sensitive to both age and associated therapy.

Clinical pharmacokinetics of antiepileptic drugs in paediatric patients. Part II. Phenytoin, carbamazepine, sulthiame, lamotrigine, vigabatrin, oxcarbazepine and felbamate.

This article is the second part of a review of the pharmacokinetics of antiepileptic drugs (AEDs) in paediatric patients. It reviews 139 papers published since 1969 on the pharmacokinetics of phenytoin, carbamazepine, sulthiame, lamotrigine (phenyltriazine), vigabatrin, oxcarbazepine and felbamate in this population. The pharmacokinetics of phenytoin are significantly affected by age. The terminal elimination half-life (t1/2z) is relatively long in neonates; it then decreases during the first postnatal month to lower values than in adults, and then progressively increases with age due to an age-dependent decrease in the metabolic rate. Rate of elimination is strongly dose-dependent at all ages. The combination of these factors makes it difficult to predict what plasma concentrations would result from dose per kilogram (dose/kg) adjustments in neonates and children, especially when phenytoin is coadministered with other liver enzyme-inducing drugs, such as phenobarbital and carbamazepine. The concentration of phenytoin in brain and other tissues depends on the unbound/total concentration ratio. For neonates this ratio is higher than that found in adults; it then decreases over the first 3 postnatal months to approach adult values. The fraction of unbound phenytoin is significantly higher in patients also receiving valproic acid. Carbamazepine is almost completely epoxidised to the active metabolite carbamazepine epoxide, which is in turn converted to carbamazepine diol. Metabolic conversion of carbamazepine and renal clearance of carbamazepine diol are much higher in children than in adults; t1/2z of carbamazepine is thus very short in young children, increasing with age. No data are available on the neonatal period. The carbamazepine epoxide/carbamazepine ratio may be significantly increased by metabolic inducers (e.g. phenytoin, phenobarbital and primidone) or by inhibitors of the carbamazepine epoxide to carbamazepine diol conversion (e.g. valproic acid). Macrolides inhibit carbamazepine metabolism, thus increasing carbamazepine plasma concentrations. Drug-induced changes in carbamazepine kinetics are particularly pronounced in children. In children, a higher dose/kg of sulthiame, lamotrigine, oxcarbazepine and felbamate than in adults is required to obtain an effective plasma concentration. The published data do not support the use of a different dose/kg of vigabatrin in children age between 1 month and 15 years. The pharmacokinetic information in the paediatric literature may help in assessing AED prescriptions in childhood to prevent seizures and AED-related adverse effects on the ongoing maturational processes of the brain.

Ethosuximide plasma concentrations: influence of age and associated concomitant therapy.

The relationship between oral dose and plasma concentration of ethosuximide was evaluated retrospectively in 198 epileptic patients aged 2.5 to 34 years. Age appears to be a major factor in determining the ethosuximide plasma level/dose (L/D) ratio. Children younger than 10 years had men L/D ratios significantly lower (p less than 0.0003) than adolescents (10 to 15 years of age) and adults (16 to 34 years of age). Associated antiepileptic therapy reduced the ethosuximide L/D ratio: mean ethosuximide L/D ratios were significantly lower in patients also taking primidone (p less than 0.0005) or valproic acid (p less than 0.02). The correlation between the dose of ethosuximide administered and the plasma concentration was significant in the 3 age groups considered (p less than 0.0004), but the wide scattering of individual plasma concentrations makes it impossible to predict what plasma concentration of ethosuximide will be obtained after a given dose. For this reason, routine monitoring of ethosuximide plasma concentrations still appears to be necessary, especially in children and patients on polytherapy.

Changes in primidone/phenobarbitone ratio during pregnancy and the puerperium.

Plasma concentrations of primidone and its metabolite phenobarbitone were monitored in 9 pregnant epileptic patients treated with primidone (and in 3 cases other antiepileptic drugs) given at constant doses throughout pregnancy and the puerperium. Phenobarbitone plasma concentrations were monitored in another 6 patients given phenobarbitone itself. A trend towards increasing primidone plasma concentrations during the second quarter of pregnancy was evident in all patients, with a concomitant significant decrease in primidone-derived phenobarbitone plasma concentrations. A trend towards a lowering of plasma concentrations of phenobarbitone administered as such was confirmed. These results suggest the usefulness of a careful monitoring of primidone and primadone-derived phenobarbitone during pregnancy and the puerperium. Discrepancies of findings with primidone and phenobarbitone are discussed in view of the possible mechanism involved.

Plasma concentrations of carbamazepine and carbamazepine 10,11-epoxide during pregnancy and after delivery.

Plasma concentrations of carbamazepine were monitored in 9 pregnant epileptic patients treated with the drug alone at constant doses during pregnancy and for at least 3 months after delivery. In addition, plasma concentrations of the metabolite, carbamazepine 10,11-epoxide were measured in 6 of the 9 patients. Plasma carbamazepine concentrations were fairly stable during pregnancy, and carbamazepine relative plasma clearances were significantly higher in weeks 4 to 24 than in weeks 25 to 32. After the end of the second trimester, there were no variations in plasma carbamazepine 10,11-epoxide concentrations and carbamazepine 10,11-epoxide:carbamazepine ratios. Both parameters were significantly higher in weeks 4 to 24 than in weeks 25 to 32 of pregnancy.

Trials in women.

Women of childbearing age have been viewed as a 'vulnerable population' and have been systematically excluded from early clinical trials. A change in attitude and policies occurred in the last decade, with a consequent increase of women participating in clinical trials. The implications are increasing respect for the woman's capacity to make her own risk-benefit choices, early evaluation of patients that represent the ultimate user of a drug and equal opportunities for women to benefit from the therapeutic potential of new drugs. Drug trials should be designed to identify sex-related effects and to analyse the efficacy and tolerability of antiepileptic drugs (AEDs) by gender. Further aspects should be considered, including changes in response in relation to the menstrual cycle and over the various stages of reproductive life; interactions between hormonal therapies and AEDs; the effect of AEDs on reproductive function and possible consequences of prenatal exposure to AEDs. These considerations become even more critical when pregnant and lactating women are considered, since any risk for the offspring is unacceptable unless drug administration is likely to have major medical benefits for the mother.

Assessment of teratogenic risk.

Although the association between prenatal exposure to old generation antiepileptic drugs (AEDs) and major congenital malformations has been studied for many years, it is not clear whether any specific AED, or AED combination, is more harmful than others, or whether any pattern of malformations can be considered specific for any given drug. Relationships between dosage and plasma concentrations of AEDs and the risk of malformations also need to be clarified. The greatest limitation of all studies performed to date is the fact that none included a sufficiently large number of pregnancies. For newer generation AEDs the teratogenic risk, if any, is unknown. Large prospective studies are needed. The best approach is the establishment of registries through international collaboration. Inclusion of non-epileptic controls and untreated women is not strictly necessary to evaluate the comparative teratogenic risk of AEDs. The modalities of data collection should be pre-defined; common protocols, sufficiently exhaustive but at the same time easy to perform, should be shared from the beginning. Information on the presence or absence of major malformations or prenatal growth retardation, and on all major factors that may affect the teratogenic endpoints should be obtained. Study designs should ensure high quality data recording, and adequate quality assurance and auditing procedures. Further requisites are a clear definition of congenital malformation and prolonged follow-up to detect the occurrence of congenital malformations not detected at birth.

Congenital malformations due to antiepileptic drugs.

To identify the major risk factors for the increased incidence of congenital malformations in offspring of mothers being treated for epilepsy with antiepileptic drugs (AEDs) during pregnancy and, to determine the relative teratogenic risk of AEDs, we prospectively analyzed 983 offspring born in Japan, Italy, and Canada. The incidence of congenital malformations in offspring without drug exposure was 3.1%, versus an incidence with drug exposure of 9.0%. The highest incidence in offspring exposed to a single AED occurred with primidone (PRM; 14.3%), which was followed by valproate (VPA; 11.1%), phenytoin (PHT; 9.1%), carbamazepine (CBZ; 5.7%), and phenobarbital (PB; 5.1%). The VPA dose and level positively correlated with the incidence of malformations. This study first determined a cut-off value of VPA dose and level at 1000 mg/day and 70 microg/ml, respectively, to avoid the occurrence of malformations. The incidence of malformations increases as the number of drugs increases, and as the total daily dose increases. Specific combinations of AEDs such as VPA + CBZ and PHT + PRM + PB produced a higher incidence of congenital malformations. The incidence of malformations was not associated with any background factors studied except for the presence of malformations in siblings. These results indicate that the increased incidence of congenital malformations was caused primarily by AEDs, suggesting that malformations can be prevented by improvements in drug regimen, and by avoiding polypharmacy and high levels of VPA (more than 70 microg/ml) in the treatment of epileptic women of childbearimg age.

Intrauterine growth in the offspring of epileptic women: a prospective multicenter study.

The aim of the present study was to evaluate the risk of intrauterine growth delay in the offspring of epileptic mothers and to quantify the risks of intrauterine exposure to antiepileptic drugs (AEDs). Data concerning 870 newborns, prospectively collected in Canada, Japan and Italy, using the same study design, were pooled and analyzed. The overall proportion of newborns whose body weight (7.8%) or head circumference (11.1%) at birth were below the 10th percentile was not increased. However, logistic regression analysis showed that the risk of small head circumference was significantly higher in Italian than in Japanese (RR 4.2; 95% CI: 2.2-8.0) or Canadian children (RR 2.6; 95% CI: 1.1-6.5), and in children exposed to polytherapy (RR 2.7; 95% CI: 1.2-6.3), phenobarbital (PB) (RR 3.6; 95% CI: 1.4-9.4) and primidone (PRM) (RR 4.5; 95% CI: 1.5-13.8). Country was also the only factor affecting low body weight, with Italian children having a higher risk than Japanese (RR 5.2; 95% CI: 2.6-10.4) or Canadian (RR 8.8; 95% CI: 2.0-38.1) children. Due to the small categories, the influence of AED doses and plasma concentrations was studied for each individual AED, without adjustment for the other potential confounding factors. A clear dose-dependent effect was found for PB and PRM in terms of both small head circumference and low body weight, and a concentration-dependent effect for PB in terms of small head circumferences. The size of the difference between the Italian and the other two populations, which is only partially explained by differences in therapeutic regimens, suggests that genetic, environmental and ethnic factors also need to be taken into account when considering possible explanations.

[Malformations in the offspring of pregnant women with epilepsy. Presentation of an international registry of antiepileptic drugs and pregnancy (EURAP)]. / Malformaciones en los hijos de embarazadas con epilepsia. Presentación de un registro internacional de fármacos antiepilépticos y embarazo (EURAP).

The interaction between epilepsy and pregnancy has been studied for many years; nonetheless the risk associated with individual antiepileptic drug has not been adequately characterized up to date. Moreover, virtually nothing is known about the possible human teratogenicity of the newer antiepileptic drugs. Because of the complexity of the mechanisms involved, the crucial evidence needed can only come from very large population based studies, and a collaborative European multicentre investigation has been set up to this purpose. Specific objectives include the evaluation of the risk of major foetal malformations and of delay in prenatal growth following exposure to antiepileptic drugs, assessment of the pattern of congenital abnormalities associated with older and newer antiepileptic drugs and their combinations, and identification of possible relationships with dosage and with a variety of other risk factors. All women exposed to antiepileptic drugs at the time of conception are eligible for entry. The protocol is purely observational and does not entail any change in prescribing pattern or management policies, which are left to the discretion of the treating physician. Data obtained during prospective monitoring for up to 1 year after birth are regularly collected in especially designed forms and entered into Regional Registries prior to transfer to a Central European Registry of Antiepileptic Drugs and Pregnancy (EURAP). Evaluations of incidence and prevalence of teratogenic endpoints will be based exclusively on cases enrolled before foetal outcome is known and in any case not after the 16th week of pregnancy. Cases enrolled after birth, after the 16th week of pregnancy or after prenatal diagnosis will only be reported descriptively. The study is being implemented gradually in 19 countries in Western and Eastern Europe. Wide participation from interested physicians is essential for the achievement of the study objectives, which are expected to lead to important advances in pre pregnancy counselling and overall clinical management of women with epilepsy.

Changes in unbound and total valproic acid concentrations after replacement of carbamazepine with oxcarbazepine.

Total and free valproic acid (VPA) concentrations were measured in patients switched from a combined VPA and carbamazepine (CBZ) to a combined VPA and oxcarbazepine (OXC) therapy, both administered in individualized daily doses. Four young epileptic patients (13-17 years old) were studied for a 12-week period, total and free VPA concentrations being analyzed just before and 4 h after the morning dose, at the end of VPA and CBZ therapy, and 2 and 10 weeks after CBZ was replaced by OXC. The expected increase in the level/dose (L/D) ratio of total VPA observed at the end of the study was preceded by a clear-cut increase in the L/D ratio of free VPA, which led to VPA-related side effects and required the retitration of VPA daily doses.

Single-dose pharmacokinetics of lamotrigine in children: influence of age and antiepileptic comedication.

To evaluate the influence of pediatric age and antiepileptic comedication on the single-dose pharmacokinetics of lamotrigine, 19 patients with epilepsy (10 comedicated with enzyme inducers and 9 comedicated with valproic acid) aged 8 months to 30 years received a single oral dose of lamotrigine (0.6 to 2.2 mg/kg) after an overnight fast. Blood samples were collected for at least 36 hours and plasma lamotrigine concentrations were determined by high-performance liquid chromatography. Pharmacokinetic parameters were calculated by noncompartmental analysis. Lamotrigine half-life (T1/2) and oral clearance (Cl/F) values were significantly lower and significantly higher, respectively, in patients comedicated with enzyme inducers than in those receiving valproic acid (T1/2 = 8.1 vs. 41.7 hours respectively, P < 0.001; Cl/F = 0.11 vs. 0.04 L/h per kg respectively, P < 0.005, geometric means), whereas Cmax and Tmax values were comparable in the two groups. The differences in pharmacokinetic parameters persisted when comparisons were made within subgroups stratified according to age. Within groups of patients homogeneous for type of comedication, Cmax and AUC values tended to be lower in children aged less than 12 years than in older patients. There was no significant relationship between half-life values and age. The authors conclude that both age and type of comedication influence lamotrigine pharmacokinetics. The reduction in lamotrigine concentrations caused by enzyme inducers and the elevation caused by valproic acid can be explained by stimulation and inhibition, respectively, of lamotrigine glucuronidation. On the other hand, the lower plasma lamotrigine levels in children than in adolescents and older patients may not be explainable solely by differences in metabolic rate.

Lamotrigine plasma concentrations in children and adults: influence of age and associated therapy.

The effects of age and concomitant treatment on plasma lamotrigine (LTG) concentration/dose (C/D) ratios were retrospectively evaluated on 482 consecutive routine LTG determinations from 106 chronically-treated patients with epilepsy (40 children and adolescents aged younger than 16 years, and 66 adults aged 17 to 62 years). A linear dose/level relationship was observed in individual patients but not in the cumulative analysis, which failed to show any correlation between the administered LTG dose and plasma concentrations. In the adult group, there were no correlations between the administered LTG dose and plasma concentrations. Associated antiepileptic therapy affected the LTG concentration/dose ratio, which was significantly higher in the patients receiving valproic acid (3.4 +/- 2.0, n = 23) and significantly lower in those treated with enzyme-inducing antiepileptic drugs (0.6 +/- 0.5, n = 57) than in the patients receiving valproic acid in combination with enzyme-inducing antiepileptic drugs, ethosuximide, vigabatrin, or clobazam (1.9 +/- 1.6, n = 26). The LTG C/D ratios significantly increased with increasing plasma valproic acid concentrations, and significantly decreased with increasing phenytoin concentrations. The effect of enzyme-inducing antiepileptic drugs increased with the number of associated drugs. A clear age effect was demonstrated in the homogeneously treated subgroups, in all of which adults had higher C/D ratios than children.

Plasma levels of primidone and its metabolite phenobarbital: effect of age and associated therapy.

The effects of age and associated therapy on plasma primidone (PRM) and derived phenobarbital (PB) concentrations, and on plasma concentrations-to-PRM dose ratios (L/D ratio) were evaluated retrospectively from 408 consecutive PRM and derived PB determinations in 238 chronically treated epileptic patients (153 children and adolescents between 5 months and 15 years of age and 85 adults between 16 and 55 years of age). The correlation between PRM administered and both plasma PRM and derived PB levels was significant; the correlation between PRM and PB plasma levels was also significant, but the scatter of values for the linear regressions was such that the relationship had no predictive value. Significant differences in mean plasma PRM and PB L/D ratios were found between patients aged 0-3 years, 4-9 years, 10-15 years, and adults (16-55 years), with higher values in the older groups. The PB/PRM concentration ratios were significantly lower in children than in adolescents and adults. Concomitant treatment with carbamazepine affected PRM disposition and led to increased L/D ratios for PB and decreased L/D ratios for PRM, whereas phenytoin increased the L/D ratios for PB without any significant change in the L/D ratios for PRM. The variability in the results indicates the need for routine monitoring of PRM and derived PB plasma levels, particularly in pediatric populations, in order to tailor the dose to each patient.

Comparison of the effectiveness of several formulations of sodium valproate: tablets, enteric-coated capsules, solutions and rectal capsules.

The effects of an oral enteric-coated preparation (VAL 579 capsules) of sodium valproate (VPA) were compared with those of conventional oral and rectal preparations in 24 epileptic patients. Enteric-coated capsules are useful for avoiding gastric intolerance to VPA. Plasma VPA level peaks after the enteric- coated form were later than after other forms. The bioavailability appeared to be the same for all the forms.

Comparative pharmacokinetic study of chewable and conventional carbamazepine in children.

Fifteen children (7 boys and 8 girls) with generalized tonic-clonic seizures (GTCS) and partial seizures with elementary or complex symptomatology, treated with carbamazepine (CBZ) alone (n = 7) or in combination with either phenobarbital (PB, n = 6) or clobazam (CLB, n = 2) given for at least 3 months at stable individualized doses and regimens, entered an open, within-patient, change-over study of consecutive periods, each lasting 2 weeks. During period 1, conventional CBZ was given; during period 2, a chewable CBZ formulation was substituted for conventional CBZ and given at the same total daily dosage with the same schedule as in period 1. Blood samples for measuring plasma concentration of both total CBZ and CBZ-10,11 epoxide (CBZ-E) were taken on the last day of each period. No significant difference between the two periods was noted in the mean +/- SD of Cmax, Css mean, and area under the curve (AUC) of total CBZ and CBZ-E. The two different CBZ formulations, administered at the same total daily dosage, can be considered bioequivalent.

Lamotrigine in resistant childhood epilepsy.

Fourteen children (6 M, 8 F) suffering from refractory epilepsy received LTG as add-on therapy. LTG was administered twice daily at dosages increasing up to 2 mg/kg/day (for patients taking VPA) or to 10 mg/kg/day for patients taking AEDs that induce hepatic metabolism. The drug was withdrawn for side effects in 3 cases (rash: two cases, hirsutism: one), because of increased seizure frequency in 2 cases and because of unchanged seizure frequency in one. One patient died from acute respiratory failure, after repeated respiratory tract infections. A decrease in seizure frequency after one year of treatment with LTG was observed in 6 of the 7 patients who completed the study. The median total seizure frequency decreased from 10.7 +/- 7.3 to 3.8 +/- 4.6 seizures per day. At the end of the study, seizure frequency had decreased by more than 50% in 2 patients, by more than 75% in 2 patients, and 2 patients were seizure-free; in the remaining patient seizure frequency was unchanged. The best results were obtained with plasma LTG concentrations ranging from 0.5 to 5.4 micrograms/ml; no further improvement was observed at higher LTG concentrations.

Schizencephaly: neuroradiologic and epileptologic findings.

PURPOSE: Nine patients affected by schizencephaly were analyzed, and the epileptologic findings prospectively studied, to define the relations between the anatomic brain malformations and clinical outcome. METHODS: The schizencephaly was diagnosed by means of magnetic resonance imaging (eight cases) or computed tomography (one case). The clinical histories of all the patients were analyzed, and a psychometric evaluation was made. The electroclinical features and course of epilepsy in the six patients with epilepsy were prospectively followed up for a period ranging from 3 to 14 years. RESULTS: The patients were divided into those who were unilaterally (six) and those bilaterally (three) affected. The former were characterized by mild neurologic deficits and late-onset epilepsy; their epileptologic features were consistent in terms of age of onset, seizure semiology, the absence of secondary generalization, and resistance to antiepileptic treatment. The patients with bilateral schizencephaly associated with other brain malformations were characterized by severe neurologic deficits but were only rarely affected by epilepsy, which was always completely controlled by antiepileptic treatment. CONCLUSIONS: Our data show that the extent of anatomic malformation is strictly related to the severity of motor and mental impairment but not to the presence or severity of epilepsy. The absence of prenatal risk factors for brain damage in our series, previously described familial cases of schizencephaly, and the recent report of mutations in homeobox gene EMX2 associated with cases of schizencephaly all indicate that genetic factors may play a key role in the pathogenesis of this brain malformation.

Intrauterine growth in the offspring of epileptic mothers.

The present paper concerns the fetal growth of 315 newborns of epileptic mothers prospectively followed from the beginning of pregnancy. In comparison with Italian standards, neonatal weight, length and head circumference at birth were below the 10th percentile in respectively 15.7%, 1.1% and 19.2% of the newborns. Weight at birth was above the 90th percentile in 8 cases. Observed frequencies were significantly higher than expected frequencies for both weight and head circumference. The percentage of newborns with a small head circumference increased significantly according to the number of drugs taken by the mother during the first three months of pregnancy: 7.1% with no drug, 16.8% with one drug, 23.6% with two drugs and 50% with three drugs. A statistically significant correlation was found between gestational age-adjusted head circumference and drug-level scores during the first trimester. Head circumferences below the 10th percentile were fewer among newborns treated with CBZ than among newborns treated with either PB or VPA.