Concentration of metabolites of valproic acid in plasma of epileptic patients

Astroglia in Autism Spectrum Disorder

Autism spectrum disorder (ASD) is an umbrella term encompassing several neurodevelopmental disorders such as Asperger syndrome or autism. It is characterised by the occurrence of distinct deficits in social behaviour and communication and repetitive patterns of behaviour. The symptoms may be of different intensity and may vary in types. Risk factors for ASD include disturbed brain homeostasis, genetic predispositions, or inflammation during the prenatal period caused by viruses or bacteria. The number of diagnosed cases is growing, but the main cause and mechanism leading to ASD is still uncertain. Recent findings from animal models and human cases highlight the contribution of glia to the ASD pathophysiology. It is known that glia cells are not only "gluing" neurons together but are key players participating in different processes crucial for proper brain functioning, including neurogenesis, synaptogenesis, inflammation, myelination, proper glutamate processing and many others. Despite the prerequisites for the involvement of glia in the processes related to the onset of autism, there are far too little data regarding the engagement of these cells in the development of ASD.

Role of environmental factors and epigenetics in autism spectrum disorders

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder thought to be caused by predisposing high-risk genes that may be altered during the early development by environmental factors. The impact of maternal challenges during pregnancy on the prevalence of ASD has been widely studied in clinical and animal studies. Here, we review some clinical and pre-clinical evidence that links environmental factors (i.e., infection, air pollution, pesticides, valproic acid and folic acid) and the risk of ASD. Additionally, certain prenatal environmental challenges such as the valproate and folate prenatal exposures allow us to study mechanisms possibly linked to the etiology of ASD, for instance the epigenetic processes. These mechanistic pathways are also presented and discussed in this chapter.

Increased BDNF expression in fetal brain in the valproic acid model of autism

Human fetal exposure to valproic acid (VPA), a widely-used anti-epileptic and mood-stabilizing drug, leads to an increased incidence of behavioral and intellectual impairments including autism; VPA administration to pregnant rats and mice at gestational days 12.5 (E12.5) or E13.5 leads to autistic-like symptoms in the offspring and is widely used as an animal model for autism. We report here that this VPA administration protocol transiently increased both BDNF mRNA and BDNF protein levels 5-6-fold in the fetal mouse brain. VPA exposure in utero induced smaller increases in the expression of mRNA encoding the other neurotrophins, NT3 (2.5-fold) and NT4 (2-fold). Expression of the neurotrophin receptors, trkA, trkB and trkC were minimally affected, while levels of the low-affinity neurotrophin receptor, p75(NTR), doubled. Of the nine 5'-untranslated exons of the mouse BDNF gene, only expression of exons I, IV and VI was stimulated by VPA in utero. In light of the well-established role of BDNF in regulating neurogenesis and the laminar fate of postmitotic neurons in the developing cortex, an aberrant increase in BDNF expression in the fetal brain may contribute to VPA-induced cognitive disorders by altering brain development.

Can we develop improved derivatives of valproic acid?

Valproic acid is one of the major antiepileptic drugs. In animal models, valproate showed less anticonvulsant potency than the other three established antiepileptic drugs: phenobarbital, phenytoin and carbamazepine. In addition, two major side-effects, teratogenicity and hepatotoxicity, have been associated with valproate therapy. Due to the above and the shortage of new antiepileptic drugs there is a substantial need to develop improved derivatives of valproate. This paper analyses three kinds of valproate derivatives: valpromide, the primary amide of valproate, and its analogues; monoester prodrugs of valproate and an active metabolite of valproate, 2-n-propyl-2-pentenoate. The comparative evaluation was carried out by pharmacokinetic and pharmacodynamic analyses in animals. From the data accumulated so far, we can conclude that 2-n-propyl-2-pentenoate and/or a valpromide isomer, which does not undergo amide-acid biotransformation and preferably is not an epoxide hydrolase inhibitor, may prove to be improved derivatives of the parent compound valproic acid.

Pathogenic role of glutamate in hyperthermia-induced seizures

Hyperthermia induces seizures in both humans and rodents, but the underlying mechanism remains unknown. The present study showed that hyperthermia, causing rapid increase in body temperature, increases the concentration of glutamate (Glu) released into a cortical perfusate before onset of seizures in rats and that this increase in Glu concentration correlated with a decrease in seizure threshold temperature. These results indicate that increased cortical extracellular Glu induced by hyperthermia contributes to onset of seizures. The same mechanism may be involved in clinical seizures induced by fever in patients with febrile convulsions or epilepsy.

Dose-dependent inhibition in plasma protein binding of valproic acid during continued treatment in guinea-pigs

Plasma protein binding of valproic acid over a wide range of steady-state plasma concentration (11.3 +/- 2.6-1303.0 +/- 122.9 micrograms mL-1: s.e.m., n = 5) in guinea-pigs has been studied. Valproic acid was given by intravenous constant infusion. At steady-state the plasma protein binding of valproic acid was analysed. Nonlinear binding was observed. Unbound fraction (fu) of valproic acid increased from 25 to 95% with the increase of steady-state plasma concentration (Css). The plasma protein-bound drug concentration (Cb) of valproic acid increased initially with Css but decreased after the Css exceeded 345.0 micrograms mL-1, where the Cb was 152.5 +/- 26.8 micrograms mL-1. At a Css of 1303.3 +/- 122.9 micrograms mL-1 the Cb was significantly (P less than 0.05) decreased to 72.8 +/- 20.2 micrograms mL-1. Binding characteristics of valproic acid in-vitro were studied using drug-free guinea-pig plasma with added valproic acid (10-1000 micrograms mL-1). The binding behaviour was also nonlinear in-vitro. The fu increased from 14 to 79% with the increase of valproate concentrations. No decrease in Cb was observed throughout the range. The study demonstrated that binding characteristics of valproic acid in-vivo and in-vitro are not parallel. The results suggest that valproic acid may produce or induce plasma protein binding competitors; metabolites of valproic acid may be implicated.

Total cholesterol, high-density lipoprotein cholesterol, and triglycerides in children receiving antiepileptic drugs

The influence of antiepileptic drug (AED) therapy on total cholesterol (TC), high-density lipoprotein (HDL) cholesterol, and triglycerides was studied in 208 epileptic children compared with 175 normal children. A significant increase in TC plasma levels was observed with carbamazepine (CBZ), phenobarbital (PB), and phenytoin (PHT). The patients receiving valproate (VPA) showed levels very similar to those of the control population. The results may be explainable by the different biotransformation pathway of these drugs. HDL cholesterol and triglycerides were not altered by any of the AEDs. We recommend monitoring TC level in patients receiving CBZ, PB, and PHT and prescription of diet treatment, at least during the time of AED treatment.

Influence of co-medication on the metabolism of valproate

Valproate is extensively metabolized in the liver and at least six main pathways which produce about 50 metabolites have been identified in man. The enzyme-inducing antiepileptic drugs phenobarbital, primidone, phenytoin and carbamazepine increase total valproate clearance by 30-85%, whereas cimetidine and the new anticonvulsant compound striripentol display a small inhibitory effect (10-20%). Both carbamazepine and phenytoin induce a two-fold increase in the formation of delta 4-valproate and stimulate omega-oxidation and omega-1-oxidation. Acetylsalicylic acid causes a fall of 60-70% in the content in the urine of the metabolites of the beta-oxidative pathway, i.e. delta 2-valproate, 3-OH-valproate and 3-oxo-valproate, and an increase of glucuronidation (approximately 30%) and delta-dehydrogenation (approximately 20%). Stiripentol inhibits the formation clearance of delta 4-valproate by 30%. In the light of the possible therapeutic and toxic effects of some valproate metabolites, drug interactions with valproate at metabolic level may have important clinical implications.

Ion trap detector--capillary gas chromatography of valproic acid and its mono-unsaturated metabolites in serum using methyl ester derivatives

A quantitative method was developed for valproic acid and five of its mono-unsaturated metabolites using capillary gas chromatography-mass spectrometry with selected ion monitoring. The method was applied to serum and all metabolites were measured in a single run. Methyl esters were synthesized as the derivatives suitable for gas chromatography. Calibration curves were found to be linear and the sensitivities in the order of 0.1 micrograms ml-1. Patients' data are presented. By this method it is possible to separate the stereoisomers of 2-n-propyl-2-pentenoic acid and of 2-n-propyl-3-pentenoic acid.

Pharmacokinetics, anticonvulsant efficacy, and adverse effects of trans-2-en-valproate after acute and chronic administration in amygdala-kindled rats

The trans isomer of 2-en-valproate (trans-2-en-VPA), an unsaturated metabolite of the clinically established antiepileptic valproate (VPA), was examined in the kindling model of epilepsy. As in the case of VPA, trans-2-en-VPA exerted potent anticonvulsant effects against partial seizures and secondarily generalized clonic seizures in amygdala-kindled rats after i.p. administration of acute doses. The anticonvulsant potency of trans-2-en-VPA appeared to be higher than that of VPA, especially in the case of secondarily generalized seizures. However, as previously reported for effects of valproate in the kindling model, trans-2-en-VPA exerted anticonvulsant effects against kindled seizures only at doses which were associated with motor impairment. Pharmacokinetic experiments with trans-2-en-VPA indicated non-linear kinetics with dose-dependent elimination rate and enterohepatic recirculation. According to the initial rapid decline in plasma concentrations of trans-2-en-VPA, the duration of anticonvulsant action in kindled rats was short-lasting so that an experimental protocol with 3 daily administrations was chosen for chronic experiments with this drug. During chronic treatment of kindled rats with 3 times daily injection of 100 mg/kg trans-2-en-VPA for 2 weeks, there was a marked reduction of anticonvulsant activity during the second week of treatment. This loss of anticonvulsant activity was not due to metabolic tolerance, i.e. reduction of drug levels by increased drug metabolism. Furthermore, additional experiments with altered experimental protocol indicated that the loss of anticonvulsant activity was not due to contingent tolerance, i.e. involvement of learning processes due to too frequent drug testing. However, the size of chronic treatment dose was important for the rate and degree of tolerance development, since an increase of dosage to 150 mg/kg 3 times daily resulted in significant anticonvulsant effects throughout the period of treatment with almost no indication of tolerance. The date indicate that trans-2-en-VPA is as effective as valproate in the kindling model. In view of previously reported experimental evidence that trans-2-en-VPA might have a lower hepatotoxic and teratogenic potential as valproate, the present study substantiates that trans-2-en-VPA might be an interesting alternative to valproate in antiepileptic therapy.

Inhibition of brain glutamate decarboxylase activity is related to febrile seizures in rat pups

Because previous work showed that in the newborn brain, but not in the adult brain, glutamate decarboxylase (GAD) is notably susceptible to heat, we have studied the possible involvement of GAD inhibition in febrile convulsions and the related changes in gamma-aminobutyric acid (GABA) content. Rats of different ages were subjected to hyperthermia, and GAD activity was determined in brain homogenates by measuring the release of 14CO2 from labeled glutamate and by measuring the formation of GABA. The latter method gave considerably lower values than the former in the youngest rats, and was considered more reliable. With this method, we found a 37-48% inhibition of GAD activity in rat pups 2-5 days old, which showed febrile seizures at progressively higher body temperatures, whereas in 10- and 15-day-old animals, which did not show convulsions, GAD activity was not affected by hyperthermia. Whole-brain GABA levels, however, did not change at any age. In contrast to GAD, choline acetyltransferase and lactic dehydrogenase activities were not altered by hyperthermia at any of the ages studied. These results suggest that a decreased efficiency of the inhibitory neurotransmission mediated by GABA, consequent to the inhibition of GAD activity, may be a factor related to febrile convulsions.

Metabolic profiling of valproic acid in patients using negative-ion chemical ionization gas chromatography-mass spectrometry

A negative-ion chemical ionization gas chromatographic-mass spectrometric method for the determination of valproic acid (VPA) and fourteen of its metabolites in a single chromatographic run is reported. The assay features the use of four internal standards and is applicable to the analysis of small serum and urine volumes. A combination of pentafluorobenzyl and trimethylsilyl derivatization resulted in the [M - 181]- ion as the base peak for all the metabolites measured. When these ions were monitored sensitivities in the low picogram levels were achieved. The VPA metabolite profile was determined in pediatric patients on VPA monotherapy and on combined VPA therapy with either carbamazepine or clobazam. The recently characterized diene metabolite, (E,E)-2,3'-diene-VPA, was found to be a major serum metabolite of VPA. In the patient groups taking VPA in combination with carbamazepine, the induction of omega and omega-1 pathways of VPA metabolism was apparent, while the levels of the beta-oxidation products were significantly decreased.

The enzymatic basis for the metabolism and inhibitory effects of valproic acid: dehydrogenation of valproyl-CoA by 2-methyl-branched-chain acyl-CoA dehydrogenase

Five distinct acyl-CoA dehydrogenases are currently known. These are short, medium, long and 2-methyl-branched-chain acyl-CoA dehydrogenases, and isovaleryl-CoA dehydrogenase. We tested these five acyl-CoA dehydrogenases for their ability to dehydrogenate valproyl-CoA using pure enzyme preparations isolated from rat liver mitochondria. The activities of the pure human short-chain, medium-chain and isovaleryl enzymes purified from post-mortem livers, and a long-chain acyl-CoA dehydrogenase preparation partially purified from placental mitochondria, were also tested. Valproyl-CoA was dehydrogenated at a significant rate (0.167 mumol/min per mg protein) only by rat 2-methyl-branched-chain acyl-CoA dehydrogenase. Human 2-methyl-branched-chain acyl-CoA dehydrogenase has not been purified; therefore, it could not be tested. Since four other human acyl-CoA dehydrogenases did not dehydrogenate isobutyryl-CoA, 2-methylbutyryl-CoA (obligatory intermediates from valine and isoleucine, respectively) nor valproyl-CoA, it is reasonable to assume that valproyl-CoA is dehydrogenated by 2-methyl-branch-chain acyl-CoA dehydrogenase in man as well. We identified 2-propyl-2-pentenoyl-CoA as the reaction product from valproyl-CoA by mass spectral analysis of the acyl moiety. Valproyl-CoA, at 0.3 mM, moderately inhibited human acyl-CoA dehydrogenases with the exception of the long-chain enzyme. 5 mM free valproic acid inhibited the activities of various acyl-CoA dehydrogenases only very weakly.

Beta-oxidation of valproic acid. I. Effects of fasting and glucose in humans

The effect of fasting and glucose-infusion on valproate (VPA) disposition was investigated to determine the involvement of endogenous fatty acid (FA) beta-oxidation in the metabolism of VPA. Fifteen healthy volunteers received multiple oral doses of VPA to achieve steady state under both conditions. Depressed plasma FA concentrations in the glucose-infused state (median 51%, p less than 0.0001) were associated with lower unbound plasma VPA fractions (median 17%, p less than 0.0001). Unbound plasma VPA concentrations were notably lower in the glucose-infused state due to significantly higher (median 41%, p less than 0.0001) metabolic clearance, beta-oxidative metabolite formation clearance, representing the largest urinary dose fraction recovered, was significantly higher (median 60%, p less than 0.004) in the glucose-infused state. This finding is consistent with competition between endogenous FA and VPA for the enzymes of beta-oxidation modulated by conditions which affect FA mobilization to the site of catabolism.

Inhibitory and excitatory amino acids in cerebrospinal fluid of chronic epileptic patients

We studied the levels of excitatory and inhibitory amino acids in the cerebrospinal fluid (CSF) of 28 epileptic patients (24 with partial type seizures, 4 with primary generalized seizures) and 12 controls. The levels of aspartate were 63% (p less than 0.01), glutamine 129% (p less than 0.001), and homocarnosine 127% (p less than 0.005) that of controls. The concentrations of glutamate, asparagine, total GABA, free GABA, taurine, and glycine did not differ between epileptic patients and controls. Patients with partial epilepsy had a pattern of amino acids in CSF similar to that in patients with primary generalized seizures. In the present study we did not observe increased excitation or decreased inhibition in the seizure-active brains of epileptics, as far as the CSF levels of amino acids reflect their levels in the brain.

Clinical pharmacokinetics of valproic acid--1988

Sodium valproate (valproic acid) has been widely used in the last decade and is now considered a relatively safe and effective anticonvulsant agent. Recently, several investigators have proposed its use in the treatment of anxiety, alcoholism and mood disorders, although these indications require further clinical studies. Valproic acid is available in different oral formulations such as solutions, tablets, enteric-coated capsules and slow-release preparations. For most of these formulations bio-availability approaches 100%, while the absorption half-life varies from less than 30 minutes to 3 or 4 hours depending on the type of preparation used. Once absorbed, valproic acid is largely bound to plasma proteins and has a relatively small volume of distribution (0.1 to 0.4 L/kg). Its concentration in CSF is approximately one-tenth that in plasma and is directly correlated with the concentration found in tears. At therapeutic doses, valproic acid half-life varies from 10 to 20 hours in adults, while it is significantly shorter (6 to 9 hours) in children. Valproic acid undergoes extensive liver metabolism. Numerous metabolites have been positively identified and there is reasonable evidence that several of them contribute to its pharmacological and toxic actions. In fact, several valproic acid metabolites have anti-convulsant properties, while many of the side effects it may cause (e.g. those related to hyperammonaemia or liver damage) are most often observed in patients previously treated with phenobarbitone. This could indicate that induction of liver enzymes is responsible for the formation of toxic valproic acid metabolites.

Capillary gas chromatographic assay for valproic acid and its 2-desaturated metabolite in brain and plasma

A capillary gas chromatographic method has been developed for quantitating the antiepileptic drug valproic acid and its pharmacologically active metabolite, E-delta 2-valproic acid, in brain and plasma. The method was designed to cover large concentration ranges of both valproic acid and E-delta 2-valproic acid for pharmacokinetic and pharmacodynamic studies in laboratory animals and human subjects. Careful optimization of the extraction and chromatographic procedures was needed to resolve the analytes from a variety of endogenous constituents and other known metabolites of valproic acid. A sensitivity limit of 0.10 micrograms/g for 300 mg of brain tissue or 0.20 micrograms/ml for 150 microliters of plasma was achieved using flame ionization detection. The within-batch coefficients of variation for both analytes were less than 8%. Reproducible calibration data were observed over a period of three to eleven months.

Valproate metabolites and hepatotoxicity in an epileptic population

Idiosyncratic hepatotoxicity, although rare, is of major concern when one is treating patients with valproate (VPA). Several clinical criteria are associated with an increased risk of developing this complication, but more specific predictors are needed. It has been postulated that 4-en-VPA or one of its further metabolites may be responsible for the hepatic toxicity and that under certain conditions the metabolism of VPA is shifted to this product. We postulated that measurement of serum concentrations of 4-en-VPA or another metabolite might be a simple technique that would be predictive of risk for developing idiosyncratic hepatotoxicity. Because this complication is rare, we chose to analyze our data by a multiple linear regression model, exploring associations between VPA or three of its metabolites and clinical risk factors for hepatotoxicity. 4-en-VPA correlated with older age and absence of encephalopathy. 4-en-VPA was only seen in patients receiving polytherapy; all patients were also receiving CBZ. 2-en-VPA correlated with poor nutritional status. We conclude that routine measurement of serum 4-en-VPA is unlikely to be a useful predictor of risk for developing fatal hepatotoxicity. Serum concentrations of 4-en-VPA may not reflect presence or effects in the liver as it may be metabolized to further intermediates or be bound to tissue. Thus, serum levels of 4-en-VPA do not reflect its important role in the pathogenesis of hepatotoxicity. This metabolite was detected only in patients receiving polytherapy, a potent risk factor for developing this rare complication.

Valproate-associated hepatotoxicity and its biochemical mechanisms

Intake of the anticonvulsant drug valproic acid, or its sodium salt, has been associated with occasional instances of severe and sometimes fatal hepatotoxicity. Probably at least 80 cases have occurred worldwide. The syndrome affects perhaps 1 in 10,000 persons taking the drug, and usually develops in the early weeks or months of therapy. Most instances have involved children, usually those receiving more than 1 anticonvulsant. Multiple cases have occurred in 2 families. The typical presentation is of worsening epilepsy, increasing depression of consciousness, and progressive clinical and biochemical evidence of liver failure. The liver has sometimes shown hepatocyte necrosis, and on other occasions widespread microvesicular steatosis, while cholestatic changes have also occurred. The appearances are interpreted as consistent with a drug toxicity reaction. During the hepatotoxicity increased amounts of unsaturated metabolites of valproate, notably 4-en-valproate, have been found in blood and urine. In 4 cases there has been evidence of impaired beta-oxidation of valproate with, in 1 case, accumulation of isomers of valproate glucuronide caused by intramolecular rearrangement of the conjugate. There are molecular structural similarities between 4-en-valproate and 2 known hepatotoxins (4-en-pentanoate and methylenecyclopropylacetic acid, the latter being responsible for hypoglycin poisoning). There are also clinical and histopathological similarities between valproate hepatotoxicity and both hypoglycin poisoning and certain spontaneous disorders of isoleucine metabolism (one pathway of valproate metabolism is analogous to oxidative degradation of isoleucine). Unsaturated metabolites of valproate, in particular 4-en-valproate, may contribute to the hepatotoxicity of the drug. However, since the hepatotoxicity appears to involve an element of idiosyncrasy, the primary defect in some cases may be an inherited or acquired deficiency in the drug's beta-oxidation. This defect may divert valproate metabolism towards omega-oxidation, with increased formation of the toxin 4-en-valproate, but may also allow increased formation of a toxic metabolite derived from isoleucine, since beta-oxidation of isoleucine derivatives will also be impaired.

Lack of relationship between sodium valproate-induced adverse effects and the plasma concentration of its metabolite 2-propylpenten-4-oic acid

The concentrations of valproic acid (VPA) and of its metabolites 3-oxo-VPA and 4-en-VPA were measured in the plasma of 12 selected epileptic patients 1, 2, 3, and 4 h after administration of a loading dose of VPA. Four of the patients, all on polytherapy, had had short-term adverse effects during chronic VPA treatment, and in them there has been abnormal NH3-values after a test doese of VPA. Eight patients (4 on monotherapy and 4 on polytherapy) had been free from adverse effects. No significant difference in the VPA, 3-oxo-VPA and 4-en-VPA concentrations was found between the three groups of patients. Accumulation of 4-en-VPA is not involved in the short-term adverse effects and hyperammonaemia induced by VPA.

Accumulation and washout kinetics of valproic acid and its active metabolites

There is growing evidence that the metabolites of valproic acid (VPA) may be pharmacologically active and could contribute to both the therapeutic and toxic effects of the drug. The accumulation and washout kinetics of VPA and its oxidative metabolites were, therefore, examined in five healthy volunteers. Valproic acid (250-mg capsules) was administered bid for 15 days. Blood samples were obtained periodically during the 15 days of drug administration and for seven days following termination of treatment. Urine was also collected over the final dosing interval. Steady-state serum concentrations of VPA were achieved within three to four days of treatment. The accumulation of all metabolites in serum lagged behind that of the parent compound, with the mono-desaturated metabolites accumulating more slowly than the hydroxylated species. Furthermore, the apparent washout half-life of each metabolite was longer than the elimination half-life of VPA. In general, the unsaturated metabolites were eliminated more slowly than the hydroxylated metabolites. The serum and urinary metabolite profiles of VPA observed in the healthy volunteers were comparable with those reported for epileptic patients. The differences in the disposition kinetics of VPA and of its potentially active metabolites may explain the previously observed dissociation between the pharmacokinetics and pharmacodynamics of the drug in epileptic patients.

Capillary gas chromatography-mass spectrometry of valproic acid metabolites in serum and urine using tert.-butyldimethylsilyl derivatives

A quantitative method has been developed for valproic acid and twelve of its metabolites using capillary gas chromatography--mass spectrometry with selected-ion monitoring. The method is applicable to serum or urine and all metabolites are measured in a single run. Ions selected for quantitative purposes were the characteristic (M-57)+ ions of the tert.-butyldimethylsilyl (tBDMS) derivatives. The 4-hydroxyvalproic acid was measured as the gamma-lactone. Calibration curves were found to be linear and the sensitivities in the order of 0.1 microgram/ml. Patient data are presented. A comparison of tBDMS and trimethylsilyl (TMS) derivatives showed that tBDMS gave superior sensitivity for the unsaturated metabolites and a shorter analysis time. Mixed tBDMS-TMS derivatives were also investigated.

GABA levels in cerebrospinal fluid of patients with epilepsy

The lumbar cerebrospinal fluid (CSF) gamma-aminobutyric acid (GABA) levels were measured in 27 patients with epilepsy, another three epileptic patients with status epilepticus and three epileptic patients with chronic cerebellar ataxia. The mean lumbar CSF GABA levels of the 27 patients with epilepsy were not significantly different from those of normal controls. Six of these 27 patients who had daily partial complex and partial motor seizures showed significantly low CSF GABA levels as did the six other patients, three each with status epilepticus and chronic cerebellar ataxia. These findings suggest that some epileptic patients have impaired brain GABAergic neurons.

An update on sodium valproate

Sodium valproate has been in clinical use for the treatment of epilepsy in Great Britain since 1973 and in the United States since 1978. It is chemically quite different from the existing antiepileptic drugs. Although most authorities concentrate on its modification of GABAergic inhibitory transmission in the central nervous system, its mechanism of action remains obscure. It has been shown to be an effective antiepileptic drug in a wide variety of seizure types, but clinically, its major use to date has been in generalized seizures. It is particularly effective in photosensitive epilepsy and myoclonus. Most adverse reactions to sodium valproate are mild and reversible, but with increasing experience, the drug's rare, idiosyncratic, adverse effects are becoming apparent, particularly hepatotoxicity and teratogenicity. The role of therapeutic drug monitoring in the management of patients taking sodium valproate is controversial.

A single therapeutic dose of valproate affects liver carbohydrate, fat, adenylate, amino acid, coenzyme A, and carnitine metabolism in infant mice: possible clinical significance

We have previously reported that chronic valproate administration reduced ketonemia in suckling mice and fasting epileptic children. The present study demonstrates that even a single dose of valproate in the therapeutic range for man caused a prolonged reduction of plasma beta-hydroxybutyrate levels in normal infant mice; the plasma glucose concentration was also significantly lowered. In the livers of these animals, there were extraordinary decreases in levels of free coenzyme A, acetyl CoA and free carnitine. Concomitantly concentrations of acid-soluble fatty acid (short-chain, non-acetyl) coenzyme A esters and of acid-insoluble (long-chain) fatty acid carnitine esters increased. There was evidence for inhibition of the metabolic flux through the Krebs citric acid cycle at those enzyme reactions which require coenzyme A. While valproate doubled liver alanine levels, concentrations of liver aspartate, glutamate and glutamine were reduced. All of the valproate-induced metabolite changes can be explained by the decrease of coenzyme A due to the accumulation of acid-soluble (non-acetyl) coenzyme A esters (presumably valproyl CoA and further metabolites). Decreased coenzyme A would limit the activities of one or more enzymes in the pathway of fatty acid oxidation and the Krebs citric acid cycle. Secondary decreases in acetyl CoA would limit both ketogenesis and gluconeogenesis. Decreased levels of selected hepatic amino acids could reflect their use as alternative fuels. The effect of clinical doses of valproate in infant mice may relate to the valproate-associated syndrome of hepatic failure and Reye-like encephalopathy in some infants and children and suggest a simple screen for those who may be at particular risk.

Pharmacokinetics of valproic acid in volunteers after a single dose study

The pharmacokinetics of valproic acid (VPA) was investigated in six healthy volunteers. This was done by monitoring total and free (unbound) valproic acid levels in the serum, and the amount of one of its metabolites, VPA glucuronide, in the urine as a function of time, after a single dose administration of the parent drug. VPA half-life calculated from the urine data of the metabolite was shorter than the half-life calculated from the blood data. About 15 to 20 per cent of the administered oral dose of VPA was excreted in the urine as VPA glucuronide. The average free fraction of VPA obtained in this study, by using the EMIT technique, ranged from 1.5 to 11.5 per cent with a mean value of 4.9 per cent.

Aspects of the metabolism of valproic acid

The metabolic routes of valproic acid (VPA) were studied by i.p. administration of the mono-unsaturated and hydroxylated metabolites to rats. Conjugation with glucuronic acid was a major metabolic route for VPA and its metabolites. Conjugation with glycine was a minor route for VPA, but was of more importance with the unsaturated metabolites. The hydroxylated metabolites, which were further oxidized to oxo-derivatives and subsequently to the dicarboxylic acids, were not metabolically dehydrated to form unsaturated metabolites. Multiple metabolic pathways, including dehydrogenation, isomerization, hydration, hydroxylation, reduction and epoxidation were inferred from the metabolites obtained after dosage of the unsaturated metabolites. Six dien-VPA metabolites were detected in VPA-treated rats, four of which are present in patients. It was concluded that 3-en-VPA and 4-en-VPA pathways, originating through dehydrogenation, are distinct from the omega- and omega-1-hydroxylation pathways. Enzyme induction from co-administration of phenobarbital caused enhancement of the minor omega-1-oxidation pathway, yet the largest effect on clearance came from increases in glucuronidation. Mitochondrial processes were unaffected, resulting in decreased contribution to the total clearance.

Reassessment of the concept of a therapeutic range of anticonvulsant plasma levels

The significance of plasma-level measurements of anti-epileptic drugs is reviewed, especially as it applies to the treatment of childhood epilepsy. The problems associated with the determination and interpretation of 'therapeutic' levels are discussed, including the justification of applying to children the results of studies of adults. It is concluded that plasma-level determinations are of great value in specific circumstances, but that more study is needed to take into account the multiplicity of variables involved.

Treatment of epileptic patients with valproic acid does not modify plasma and urine short-chain-fatty acids

The purpose of this study was to evaluate the possible modifications of the plasma and urine short-chain-fatty acid (SCFA) patterns indiced by treatment with valproic acid (VPA). Increased amounts of SCFAs in patients under VPA treatment may explain the presence of VPA-induced hyperammonemia, toxic encephalopathies and rarer Rey-like syndromes recently observed. For this reason we assayed SCFA levels in the plasma and looked for propionic acid in the urine of 10 epileptic patients to whom it was decided to add VPA to their previously unsatisfactory anti-epileptic treatment. This was carried out prior to and during therapy with VPA. 5 of these patients developed toxic encephalopathy with hyperammonemia induced by VPA. Our data show that plasma and urine SCFAs are not modified by VPA treatment. This is so even in patients who have toxic encephalopathy with hyperammonemia indiced by this drug.

Abnormal metabolism of valproic acid in fatal hepatic failure

A 7-year-old boy developed a severe unilateral grand mal seizure at the age of 5 years (phenobarbitone therapy); 1.5 years later valproate (2-propylpentanoic acid, VPA) was added to the therapy. After a seizure-free period of 3 months the patient died from hepatic failure resembling Reye syndrome. Several plasma and urine samples from the final stage before and during peritoneal dialysis were analyzed by GC/MS. The predominant feature was the abnormally increased formation of both 3 mono- and 4 double unsaturated metabolites of VPA amounting in plasma to 58%-71% of the sum of VPA plus all analyzed metabolites (controls maximal 15%) and in urine to 34%-61% (controls maximal 10%). The beta-oxidation pathway of VPA was shown to be suppressed (lack of 3-keto-VPA), whereas metabolites from the omega-oxidation pathway could still be measured (urinary 5-OH-VPA plus 2-propylglutaric acid ca. 1.6%, controls more than 10%). 4-en-VPA (2-propyl-4-pentenoic acid) (5%-21% in plasma) and 4,4'-dien-VPA (2(2-propenyl)-4-pentenoic acid) (4%-7%) have been found as abnormal unsaturated metabolites not detectable in controls. Additional typical findings were the high excretion of adipic acid, suberic acid, and 4-octen-1,8-dicarboxylic acid demonstrating the enhanced capacity of omega-oxidation in fatty acid oxidation.

Influence of valproic acid on hepatic carbohydrate and lipid metabolism

Valproic acid (dipropylacetic acid), an antiepileptic agent known to be hepatotoxic in some patients, caused inhibition of lactate gluconeogenesis, fatty acid oxidation, and fatty acid synthesis by isolated hepatocytes. The latter process was the most sensitive to valproic acid, 50% inhibition occurring at ca. 125 microM with cells from meal-fed female rats. The medium-chain acyl-CoA ester fraction was increased whereas coenzyme A (CoA), acetyl-CoA, and the long chain acyl-CoA fractions were decreased by valproic acid. The increase in the medium chain acyl-CoA fraction was found by high-pressure liquid chromatography to be due to the accumulation of valproyl-CoA plus an apparent CoAester metabolite of valproyl-CoA. Salicylate inhibited valproyl-CoA formation and partially protected against valproic acid inhibition of hepatic metabolic processes. Octanoate had a similar protective effect, suggesting that activation of valproic acid in the mitosol is required for its inhibitory effects. It is proposed that either valproyl-CoA itself or the sequestration of CoA causes inhibition of metabolic processes. Valproyl-CoA formation also appears to explain valproic acid inhibition of gluconeogenesis by isolated kidney tubules. No evidence was found for the accumulation of valproyl-CoA in brain tissue, suggesting that the effects of valproic acid in the central nervous system are independent of the formation of this metabolite.

Decreased serum carnitine in valproate induced Reye syndrome

A 3-year-old girl developed acute liver disease after treatment with valproate for 6 months. She developed the typical features of Reye syndrome. Serum free carnitine was decreased as well as 3-keto-valproic acid, the main metabolite of beta-oxidation of valproate. The serum valproate concentration was at the upper limit of the therapeutic range. The possible importance of carnitine in the pathogenesis of valproate induced liver disease is discussed.

Valproate-induced hepatic steatogenesis in rats

The administration of high-dose valproic acid (VPA) (750 mg per kg) consistently produced significant microvesicular steatosis in mature Sprague-Dawley rats after 48 hr. Similar changes occurred in animals pretreated with phenobarbital which received a lower dose of VPA (350 mg per kg), but no steatosis was seen in animals treated with the low-dose VPA alone. The steatogenic effect of VPA is most likely mediated by a toxic metabolite. It can also be speculated that phenobarbital, by enhancing the inducing effects of the hepatic mixed-function oxidase system, may lead to increased conversion of VPA to a toxic metabolite. Young and weanling rats appeared to be resistant to the steatogenic effects of VPA. Reproduction of microvesicular steatosis in this experimental; model may permit exploration of factors that enhance or inhibit VPA-induced hepatic injury.

Valproate-induced hepatic injury: analyses of 23 fatal cases

Analyses of 23 fatal instances of hepatic injury in patients taking valproic acid reveals that all but three were less than 20 years old, and all but four had been taking the drugs for more than 1 month. Convulsions, facial edema, lassitude, and vomiting were prominent clinical features. Hypoglycemia was recorded in six patients. Rash and eosinophilia were not seen. Values for transaminases were modestly elevated in most patients. Most levels of SGOT were below 500 IU, and SGPT levels were below 200 IU. Livers showed microvesicular steatosis in most patients, usually accompanies by necrosis. Four patients had cirrhosis. Overt valproic acid-induced hepatic injury appears to be rare and hence, by definition, idiosyncratic. That it may be an idiosyncratic exaggeration of a much more frequent phenomenon is suggested by the higher incidence of seemingly trivial injury. The idiosyncrasy appears to be metabolic rather than immunologic, and the available information leads to the plausible hypothesis that a metabolite is responsible for the microvesicular steatosis seen in most fatal cases. The steatosis resembles that of Reye's syndrome and Jamaican vomiting sickness, and there is reason to believe that the metabolite responsible for the steatosis resembles the agent responsible for Jamaican vomiting sickness. A different metabolite is presumably responsible for the necrosis seen in many of the cases.

Anticonvulsant and antiaggressive properties of di-n-propyl acetate after repeated treatment

Swiss albino mice were treated twice daily with 200 mg/kg (i.p.) di-n-propylacetate (DPA), for 14 days and sacrificed 15 hr after the last injection, with or without an additional injection of 400 mg/kg at 45 min before death. Without the additional injection, the brain distribution of GABA followed closely that of control mice; with the additional injection, the pattern of regional increase of brain GABA was very similar to that following an acute injection of DPA. The anticonvulsant and antiaggressive properties of DPA were similar in rodents treated acutely or additionally after repeated DPA treatment. Neither the anticonvulsant nor the antiaggressive properties could be observed 15 hr after the last of the repeated injections of DPA. A residual effect of DPA against seizures induced by pentetrazole was shown, i.e. a potentiation of the anticonvulsant properties of the second of two successive injections of DPA, the latter being given after the decrease of the GABA levels in whole brain to control values.

Plasma levels of valproic acid and its metabolites during continued treatment in dogs

Dogs were treated with the anti-epileptic drug valproic acid (VPA) for 2 weeks in order to determine whether therapeutic plasma levels could be maintained. VPA (as the sodium salt) was administered in a dosage of 170-180 mg/kg/day divided into three equal doses in the form of enteric-coated tablets. Due to the rapid elimination of this drug in dogs, therapeutic plasma levels were reached only intermittently. However, the concentrations maintained may be sufficient for anti-epileptic therapy in dogs because the lower plasma protein binding of VPA in this species gives rise to higher central concentrations compared with man. Furthermore, some VPA metabolites previously shown to exert anticonvulsant activity (2-en-VPA and 3-keto-VPA) were found in relatively high concentrations in dog plasma. We would recommend the dosages used in this study as a basis for assessment of VPA in the treatment of canine epilepsy.