Utility of Free Level Monitoring of Antiepileptic Drugs

Criteria that were developed for monitoring free (unbound) rather than total (free plus bound) concentrations of antiepileptic drugs include extensive and variable binding to plasma proteins. Phenytoin and valproic acid belong to this category. It is shown that free drug concentration is independent of total drug concentration, whereas total drug concentration depends on free concentration and free fraction. Because antiepileptic drugs are predominantly bound to albumin, free fraction will increase in the presence of hypoalbuminemia (hepatic and renal disease, burns, and pregnancy). Free fraction also increases because of saturable binding (valproic acid) and competitive binding (valproic acid displacing phenytoin). There is suggestive evidence that side effects may be more closely related to the free, rather than to the total, plasma concentration of phenytoin. The clinical evidence that side effects or therapeutic effects are better correlated to the free, rather than the total, concentration of valproic acid or carbamazepine is not yet convincing. Knowledge of the free concentration improves our understanding of therapeutic and toxic effects of low total plasma concentrations. Further clinical trials are necessary for definitive assessment of the clinical relevance for free drug monitoring of valproic acid, carbamazepine, and phenytoin in the management of epileptic patients.     

Hepatotoxicity of Sodium Valproate and Other Anticonvulsants in Rat Hepatocyte Cultures

Summary: Valproic acid has recently been shown to be associated with liver disease in man. Rat hepatocyte cultures were used to test the hepatotoxicity of valproic acid and several other anticonvulsants. Dose-related hepatotoxicity was demonstrated for valproic acid at concentrations ranging from 10 to 320 /ig/ml, therapeutic concentrations being 50–100 μg/ml. Hepatotoxicity could not be demonstrated for phenobarbital, phenytoin, primidone, and clonazepam. It is concluded that valproic acid is a dose-related hepatotoxin.     

Asterixis associated with sodium valproate

Intoxication with most anticonvulsants can produce asterixis. Asterixis rarely occurs with therapeutic serum anticonvulsant levels. We report two patients with asterixis who were taking valproic acid and had serum levels within the therapeutic range. Neither patient had clinical or laboratory evidence of hepatotoxicity. Only one other patient has been reported with valproate-associated asterixis in the absence of toxic serum drug levels or hepatotoxicity. Asterixis seems to be due to a central effect of the drug unrelated to hepatotoxicity or sedation.     

Anticonvulsive drugs and blood levels of lactate, pyruvate, and glucose in children with seizures

Lactate, pyruvate, and glucose were determined in groups of 10 patients, each receiving single drug therapy for their seizure disorder with either phenytoin, valproic acid, carbamazepine, or phenobarbital, and in patients not taking medication. All drug levels were in their therapeutic ranges. The characteristics of the groups were similar, except for the phenytoin group being significantly older. No significant differences were found in glucose or pyruvate concentrations. However, patients on valproic acid had reduced lactate (p less than 0.05) and L/P ratios (p less than 0.05). The reduction in lactate in patients on valproate therapy may mask inborn errors of metabolism that normally result in an increase in blood lactate levels.     

Effect of phenytoin on protein binding of valproic acid.

In vitro experiments using the equilibrium dialysis technique were performed to determine the binding of valproic acid to plasma components in the absence and presence of therapeutic concentrations of phenytoin. The free fraction of valproic acid was found to be dependent on the total valproic acid concentration. Phenytoin did not influence valproic acid protein binding.     

Thyroglobulin and Thyroid Hormones in Patients on Long-Term Treatment with Phenytoin, Carbamazepine, and Valproic Acid

Serum concentrations of total triiodothyronine (T3) and thyroxine (T4) as well as serum thyroid-stimulating hormone (TSH) and thyroglobulin (Tg) were measured in 24 patients with epilepsy taking anticonvulsants (either phenytoin, carbamazepine, or valproic acid as single treatment) and in a control group of 28 patients with scoliosis but without thyroid disease. The T4 as well as the TSH concentrations were depressed in patients on phenytoin or carbamazepine treatment. The T3 concentration was increased in the patients on carbamazepine or valproic acid treatment, whereas the Tg levels were unaffected by all three drugs. Thus, a slight depression of the TSH concentration within the normal range does not influence the Tg release. The lack of change in the Tg concentration also speaks against a direct effect of the antiepileptic drugs on the thyroid gland.     

Phenytoin interacts with calcium channels in brain membrances

Phenytoin at concentrations of between 30 and 300 μM inhibited binding of the calcium antagonist {³H} nitrendipine to voltage-dependent calcium channels in brain membranes. Other anticonvulsants (phenobarbital, carbamazepine, valproic acid, and clonazepam) failed to inhibit binding or did so only at concentration much higher than occur clinically. Calcium channel blockade may be important in the clinical actions of phenytoin, including certain of the adverse effects of the drug.     

Valproic acid in epilepsy: clinical and pharmacological effects.

The antiepileptic drug valproic acid was studied in an open clinical trial as adjunct medication for 23 patients with uncontrolled seizures of a generalized or partial type. Two-thirds of the patients experienced reduction in seizure frequency ranging from 25 to 100%. Extensive testing revealed no evidence of serious systemic toxicity due to the drug. Minor side effects (e.g., nausea, vomiting, or sedation) were usually transient. Sodium valproate syrup and valproic acid in capsules gave equivalent mean low (23.3 microgram/ml) and maximum (42.5 microgram/ml) serum concentrations. The drug had a relatively short half-life of 8.7 hours, necessitating administration in divided daily doses. During initiation of valproate therapy there was evidence of a decline in total serum phenytoin concentration (16.5 to 10.2 microgram/ml; p less than 0.001) while the percentage of free phenytoin increased (10.9 to 20%). The quantity of unbound phenytoin was relatively stable throughout. This observation was interpreted as a drug interaction: valproic acid competed with phenytoin for access to plasma protein binding sites.     

Simultaneous toxicities in a child on multiple anticonvulsants

It is rare to develop simultaneous toxicities while on anticonvulsants. This article presents a 3(1/2)-year-old child on valproic acid, lamotrigine, and phenytoin who developed simultaneous hepatotoxicity and bone marrow toxicity during a parainfluenza virus type 3 infection. These toxicities resolved after the cessation of anticonvulsants, and her seizures were managed acutely with scheduled lorazepam. This article discusses the possibility that simultaneous use of valproic acid, lamotrigine, and phenytoin could give this combination of toxicities and that concurrent viral infection may increase this risk.     

Anticonvulsant drug actions on neurons in cell culture

Summary Two actions of clinically used antiepileptic drugs have been studied using mouse neurons in primary dissociated cell culture. The antiepileptic drugs phenytoin, carbamazepine and valproic acid were demonstrated to limit sustained high frequency repetitive firing of action potentials at free serum concentratons that are achieved in patients being treated for epilepsy. Furthermore, an active metabolite of carbamazepine also limited sustained high frequency repetitive firing while inactive metabolites of phenytoin and carbamazepine did not limit sustained high frequency repetitive firing. Phenobarbital and clinically used benzodiazepines limited sustained high frequency repetitive firing of action potentials, but only at concentrations achieved during the treatment of generalized tonic-clonic status epilepticus. Ethosuximide did not limit sustained high frequency repetitive firing even at concentrations four times those achieved in the serum of patients treated for generalized absence seizures. Phenobarbital and clinically used benzodiazepines enhanced postsynaptic GABA responses at concentrations achieved free in the serum during treatment of generalized tonic-clonic or generalized absence seizures. However, phenytoin, carbamazepine, valproic acid and ethosuximide did not modify postsynaptic GABA responses at therapeutic free serum concentrations. These results suggest that the ability of antiepileptic drugs to block generalized tonicclonic seizures and generalized tonic-clonic status epilepticus may be related to their ability to block high frequency repetitive firing of neurons. The mechanism underlying blockade of myoclonic seizures may be related to the ability of antiepileptic drugs to enhance GABAergic synaptic transmission. The mechanism underlying management of generalized absence seizures remains unclear.     

The clinical value of free phenytoin levels

The relationship between total and free phenytoin levels and drug toxicity was studied in 80 patients. Twenty-four were taking phenytoin alone. Drug toxicity was assessed by a "blind" rater using an eight-point standardized scoring system. The mean free phenytoin fraction was 0.076 in patients taking phenytoin alone or phenytoin and carbamazepine and 0.11 in patients taking valproic acid (p less than 0.001). The free fraction did not change with the total level over the range tested (6.7 to 39.9 micrograms/ml total phenytoin). There was a strong correlation between free and total levels (r = 0.84). Both free (r = 0.59) and total (r = 0.49) phenytoin levels were positively correlated with the toxicity score. Only total phenytoin levels showed a weak positive correlation with decreasing seizure frequency. Our results suggest that routine free phenytoin level monitoring is not necessary in most clinical situations.     

Interactions of valproic acid with phenytoin

The interaction of phenytoin and valproic acid was studied in four adults. We studied serial changes in total phenytoin concentrations, protein binding, urinary hydroxyphenylphenylhydantoin (HPPH) excretion, and half-life. In all four patients valproic acid caused an increase in the free fraction of phenytoin. Total phenytoin plasma concentrations decreased transiently in three patients and remained low throughout the study period in one patient. HPPH excretion increased transiently and then decreased, corresponding to changes in total phenytoin plasma concentrations. Biologic half-life transiently decreased in three patients (not statistically significant) and subsequently increased significantly in all four patients. The data suggest that valproic acid displaced phenytoin from protein-binding sites in all four patients and subsequently inhibited phenytoin metabolism in three patients.     

Brain binding of anticonvulsants: carbamazepine and valproic acid

We investigated the binding of radioactive carbamazepine and valproic acid to brain homogenates, lipid-free extracts, brain lipid, and phospholipid fractions. Carbamazepine was bound to each of these components in a manner that is qualitatively indistinguishable from the binding of phenytoin, in spite of major physical-chemical differences in these molecules but consistent with the similarities in pharmacologic action. These data support the concept that protein and phospholipid binding may be required for the activity of membrane-stabilizing anticonvulsants. Neither valproic acid nor its metabolites exhibited binding to brain or to any of the individual components tested. Therefore, binding cannot explain the reported long duration of action of this drug.     

Variability of phenytoin protein binding in epileptic patients

The variability of protein binding of phenytoin in epileptic patients was evaluated using a rapid ultrafiltration system. Thirty-one pairs of total and free (unbound) phenytoin plasma concentrations were reviewed. Filtered plasma for free phenytoin determination was obtained using an ultrafiltration membrane system. All samples were analyzed by an enzyme-multiplied immunoassay technique. A good correlation was obtained for total phenytoin concentration v free phenytoin concentration. Comparison of obtained and predicted free phenytoin values differed significantly. Of nine patients demonstrating toxic effects, six had total phenytoin concentrations within the accepted therapeutic range with free concentrations above the therapeutic range. The results of this study indicate a wide variation of phenytoin protein binding in chronic epileptic patients, implying that total phenytoin concentration must be interpreted cautiously, and use of free phenytoin concentration appears to be a more appropriate guide to therapy. The availability of a rapid method of ultrafiltration makes free phenytoin plasma levels readily available to the clinician.     

Effects of common anti-epileptic drug monotherapy on serum levels of homocysteine, vitamin B12, folic acid and vitamin B6.

There is emerging evidence to support the unfavorable effects of some anti-epileptic drugs on the plasma homocysteine concentrations. Elevated homocysteine levels induced by anti-epileptic drug administration can theoretically increase not only the risk of vascular occlusive diseases, but also the risk of resistance to anti-epileptics and development of refractory epilepsy. To investigate the effect of common anti-epileptic drugs on the homocysteine metabolism, a total of 75 epileptic patients receiving phenytoin (n=16), carbamazepine (n=19), or valproic acid (n=22) and no anti-epileptic drug (n=18) were enrolled. Eleven age- and sex-matched healthy subjects served as the control group. Blood concentrations of homocysteine, folic acid, Vitamin B12 and pyridoxal 5'-phosphate (active circulating form of Vitamin B6) were measured. Compared to the control group, epileptic patients on anti-epileptic drug had higher blood levels of homocysteine. No difference in homocysteine concentrations was observed among epileptic patients in terms of the anti-epileptic drug used. Patients receiving phenytoin had significantly lower folic acid levels and those receiving carbamazepine had marginally lower pyridoxal 5'-phosphate levels in comparison with those using other anti-epileptic drugs. A negative correlation between homocysteine and folic acid concentrations was detected in epileptic patients on anti-epileptic drug. The duration of anti-epileptic drug use was correlated to the decrease of folic acid levels, but not with changes observed in homocysteine, Vitamin B12 and pyridoxal 5'-phosphate levels. No relationship between seizure frequency and homocysteine levels was observed in epileptic patients. Our results confirm that common anti-epileptic drugs has disadvantageous effects on homocysteine status. Because there was no significant change in homocysteine concentrations in epileptic patients who were not receiving an anti-epileptic drug, and no positive correlation between seizure frequency and homocysteine levels, we suggest that increase of homocysteine levels may be due to anti-epileptic drug use, rather than being epileptic in origin. Additionally, the underlying mechanism for homocysteine increase seems to be a decrease of cofactor molecules in patients using carbamazepine and phenytoin (pyridoxal 5'-phosphate and folic acid, respectively). However, changes observed are not related to the alteration in the levels of cofactors and remain unclear in the patients using valproic acid.     

Differential effects of anticonvulsants on developing neurons in vitro

The effects of four anticonvulsants, phenobarbital, diphenylhydantoin (DPH), carbamazepine and valproic acid, on dissociated cultures of embryonic chick dorsal root ganglia (DRG) neurons were examined. DPH and valproic acid inhibited process formation in a dose-dependent manner and reduced neuronal production of substrate attached neurite promoting activity (SANPA). Less inhibition of process formation occurred with phenobarbital and carbamazepine and these two anticonvulsants had no effect on SANPA production. Substrate attachment of neurons, their survival and receptor interactions of DRG neurons with nerve growth factor were not influenced by any of the four anticonvulsants. These data suggest a rank ordering of direct toxic effects of anticonvulsants on developing neurons. The mechanisms of this neurotoxicity may include decreased neuronal production of autostimulating neurite promoting factors.     

Successful treatment with gabapentin in the presence of hypersensitivity syndrome to phenytoin and carbamazepine: a report of three cases.

We report three consecutive patients with hypersensitivity syndrome (HSS) due to phenytoin and carbamazepine and successful treatment with gabapentin. HSS is a rare but potentially fatal reaction to multiple drugs including several anticonvulsants. Cross-reactivity among drugs may occur. Immediate withdrawal of the offending drug is the most important step in treatment. Benzodiazepines acutely and, after resolution of the hepatitis, valproic acid have been successfully used for seizure control in patients with HSS. Our cases indicate that gabapentin is also a safe anticonvulsant in HSS.     

Valproic acid and plasma levels of phenytoin.

Eight patients were treated concurrently with a constant dose of phenytoin and valproic acid for 1 year. During initial therapy with valproic acid, total plasma phenytoin levels decreased. The interaction was transient and was not observed at the end of 1 year. Total plasma phenytoin levels returned to pre-valproic-acid levels in seven patients.     

Elimination of phenytoin in toxic overdose

Phenytoin is widely used for the management of seizures. Fortunately overdosage with this drug is rare. We performed a prospective study to investigate the elimination kinetics of phenytoin in toxic overdose. All patients were only on phenytoin and not on other anticonvulsants. Phenytoin toxicity was defined by clinical features and correlated with drug levels. Daily phenytoin levels were obtained until they were less than or equal to 15 mcg/ml. Nine patients with age ranging from 20 to 66-years-old were recruited. Sex ratio was male:female, 5:4. Initial phenytoin levels ranged from 34 to 57.5 mcg/ml. Serum phenytoin levels of three patients remained relatively constant for 2-5 days before declining in a steady fashion. Phenytoin levels of the remaining patients declined in an almost linear manner. Regression analysis of all patients showed that the slope terms were highly significant (with low P-values) and corresponding R(2) values were close to 1. Different patients have different rates of metabolism; in seven of nine patients, levels declined between 4.6 and 5.9 mcg/ml per day. Knowledge of the rate of elimination assists the clinician in deciding on the best time to reinstitute phenytoin therapy.     

A case of pharmacokinetic interference in comedication of clozapine and valproic acid

In 4-6% of treatment histories, clozapine induces generalized seizures by reducing the seizure threshold. Despite the knowledge of high risks combined therapy (such as bone marrow suppression, pathological EEG changes), some authors even suggest the prophylactic combination with anticonvulsants in high dose treatment of clozapine. We report a case of a 33-year-old female patient, a heavy smoker, suffering from mixed schizoaffective disorder from 1989 onwards. At her 8th admission in 1998, she was rehospitalized after experiencing her first generalized seizure under clozapine treatment, although no seizure phenomenon or other relevant side-effects under several previous clozapine therapies had been observed. Therefore, she received a valproic acid co-medication during her clozapine therapy. Based on therapeutic drug monitoring of clozapine (weekly) under compliance-controlled conditions, the serum levels of clozapine significantly decreased, probably induced by valproic acid. According to the literature, this case report might support the clinical relevance of therapeutic drug monitoring when clozapine therapy is combined with valproic acid as co-medication.