Population pharmacokinetics of phenobarbital by mixed effect modelling using routine clinical pharmacokinetic data in Japanese neonates and infants: an update.

WHAT IS KNOWN AND OBJECTIVE: Optimal use of phenobarbital in the neonatal population requires information regarding the drug's pharmacokinetics and the influence of various factors, such as different routes of administration, on the drug's disposition. However, because of sampling restrictions, it is often difficult to perform traditional pharmacokinetic studies in neonates and infants. This study was conducted to establish the role of patient characteristics in estimating doses of phenobarbital for neonates and infants using routine therapeutic drug monitoring data. METHODS: The population pharmacokinetics of phenobarbital was evaluated using 109 serum concentration measurements obtained from routine phenobarbital monitoring of 70 neonates and infants. The data were analysed using the non-linear mixed effects model. A one-compartment pharmacokinetic model with first-order elimination was used. Covariates screened were current total bodyweight (TBW), gestational age, postnatal age (PNA), post-conceptional age, gender and neonates-infants clearance factor (serum concentration of phenobarbital; Conc). RESULTS AND DISCUSSION: The final pharmacokinetic parameters were CL/F (mL/h) = (5.95.TBW (kg) +1.41.PNA (weeks)) Conc (serum phenobarbital concentration >50 µg/mL)(-0.221),Vd/F(L) =1.01.TBW (kg), and F = 0.483 for oral administration and F = 1 was assumed for suppository. Conc(-0.221) is 1 for phenobarbital concentration <50 µg/mL. The important variables for predicting phenobarbital clearance in this study were TBW, PNA and Conc. Phenobarbital clearance increases proportionately with increasing TBW, and an older newborn was expected to have a higher rate of clearance than a younger newborn of equal bodyweight. Moreover, the clearance of phenobarbital decreased nonlinearly with increasing serum concentration of phenobarbital >50 µg/mL (Conc(-0.221) ). WHAT IS NEW AND CONCLUSION: We developed a new model for neonate and infant dosing of phenobarbital with good predictive performance. Clinical application of our model should permit more accurate selection of initial and maintenance doses to achieve target phenobarbital concentrations in Japanese neonates and infants, thereby enabling the clinician to achieve the desired therapeutic effect. A similar approach can be used to validate our model for use in other neonate and infant populations.

Effect of obesity on serum amiodarone concentration in Japanese patients: population pharmacokinetic investigation by multiple trough screen analysis.

OBJECTIVE: To evaluate the influence of obesity on pharmacokinetics of amiodarone (AMD) using Non-Linear Mixed Effects Modelling (NONMEM) in Japanese patients treated with oral therapy. METHOD: Serum concentrations of AMD were determined by high performance liquid chromatography. One hundred and fifty-one trough concentrations from 23 patients receiving repetitive oral AMD were collected. Body mass index (BMI) and body fat percentage were measured. RESULTS: Estimates generated using NONMEM indicated that the clearance of AMD was influenced by BMI, age and daily dosage of AMD. The final pharmacokinetic model was CL (L/h) = 0*16 * TBW * 0.53(AGE >or= 65 ) * 0*78(BMI >or= 25) * DD(0.51), V(d) (L) = 10*2 * TBW, where CL is total body clearance, TBW is total body weight (kg), DD (mg/kg/day) is daily dosage of AMD, AGE (years) >or=65 = 1 for patient was 65 years old or over and 0 otherwise, BMI (kg/m(2)) >or=25 = 1 for patient was 25 kg/m(2) or over and 0 otherwise and V(d) is apparent volume of distribution. The clearance of AMD decreased significantly by 22.3% with a BMI higher than 25 kg/m(2). The clearance of AMD also decreased significantly by 46.9% when patient age was more than 65 years. CONCLUSION: Population pharmacokinetic analysis confirms that obesity affects the pharmacokinetics of AMD.

Optimisation of antiepileptic drug therapy. The importance of serum drug concentration monitoring.

The ability to measure the serum concentrations of antiepileptic drugs, and the widespread use of this procedure, has markedly improved the treatment given to patients with epilepsy during the past 3 decades. The monitoring of antiepileptic drug concentrations in serum is necessary for the optimal drug therapy of seizures, because the therapeutic and toxic effects of these drugs are better related to serum concentration than to administered dosage. Monitoring appeared to have a major impact on improving the effectiveness and safety of antiepileptic drug therapy. The age-related variability of pharmacokinetic parameters may also require the individualisation of therapy, with subsequent re-evaluation as the person grows older. Monitoring serum concentrations of antiepileptic drugs may help to optimise the dose. A drug concentration, however, can only be regarded as a guide around which to alter the dosage according to the patient's clinical condition. Serum drug concentration monitoring is particularly useful to ensure compliance and in helping to manage combinations of antiepileptic drugs that invariably interact. The addition or deletion of other antiepileptic drugs may change dosage requirements. Therefore, routine monitoring of antiepileptic drug serum concentrations would be extremely useful, especially in the paediatric population, and in patients who require associated antiepileptic medication.

Population pharmacokinetics of digoxin in Japanese patients: a 2-compartment pharmacokinetic model.

OBJECTIVE: To clarify the observed variability of digoxin disposition by performing a population pharmacokinetic analysis in a Japanese population. DESIGN: Retrospective analysis of clinical pharmacokinetic data. PATIENTS AND PARTICIPANTS: Data were obtained from 106 patients with heart failure and atrial fibrillation (43 males and 63 females). METHODS: Digoxin concentrations in serum were measured by fluorescence polarisation immunoassay. Population pharmacokinetic analysis was performed using a 2-compartment open pharmacokinetic model with the computer program NONMEM. RESULTS: 246 serum concentrations were obtained. Final pharmacokinetic parameters were: CL (L/h) = (0.036 x TBW + 0.112 x CL(CR)) x 0.77SPI x 0.784CCB, V1 = 1.83 L/kg, V2 = 22.6 L/kg and Q = 0.629 L/h/kg, where CL is total body clearance, V1 and V2 are the apparent volumes of distribution in the central and peripheral compartments, Q is intercompartmental clearance, TBW is total bodyweight (in kg), CL(CR) is creatinine clearance (in ml/min), SPI = 1 for concomitant administration of spironolactone (and zero otherwise) and CCB = 1 for concomitant administration of calcium antagonists (and zero otherwise). Concomitant administration of digoxin and spironolactone resulted in a 23% decrease in digoxin clearance. Concomitant administration of digoxin and calcium antagonists (diltiazem, nicardipine, nifedipine or verapamil) resulted in a 21.6% decrease in digoxin clearance. CONCLUSIONS: The estimated population parameter values may assist clinicians in the individualisation of digoxin dosage regimens.

Detection of carbamazepine drug interaction by multiple peak approach screening using routine clinical pharmacokinetic data.

Multiple peak approach screening was used to detect the effects of other antiepileptic drugs on the population estimates of relative clearance of carbamazepine. Routine clinical pharmacokinetic data (N = 1,010) collected from 466 patients receiving carbamazepine were analyzed according to a simple steady-state pharmacokinetic model with the use of NONMEM, a computer program designed for population pharmacokinetic analysis that allows pooling of data. NONMEM estimates indicated that carbamazepine clearance decreased nonlinearly with increasing total body weight (TBW) in the maturation process (over an age range of 5 months to 15 years) and increased nonlinearly with increasing daily dose (mg/kg) of carbamazepine. Concomitant administration of other antiepileptic drugs resulted in an increased in clearance of carbamazepine as follows: valproic acid alone (VPA), 7%; phenobarbital alone (PB), 16%; more than two antiepileptic drugs (POLY), 27%. Final regression model of relative clearance (Cl) for all data was: Cl (mL/hr/kg) = 64.9. TBW (kg)-0.336.DOSE (mg/kg/day)0.465 1.07VPA 1.16PB 1.27POLY.

Pharmacokinetic and pharmacodynamic analysis of subcutaneous recombinant human granulocyte colony stimulating factor (lenograstim) administration.

A total of 72 adult healthy volunteers were administered 1 microgram/kg of rhG-CSF. There was no correlation between Cmax and an increase in peripheral neutrophil count, and there was a negative correlation between AUC and this increase. The mechanism of this is probably based on the correlation between the elimination rate constant (ke) and neutrophil increase. The ke probably has a close relationship with uptake by neutrophil and its progenitor via the G-CSF receptor. An individual with higher ke should therefore show a greater increase in neutrophil count. Therefore, AUC is proportional to the rhG-CSF remainder, that is, the proportion that is not consumed in the course of increasing the neutrophil count. In such a situation, the bioavailability calculated from the AUC is unlikely to indicate the absorbed amount. The authors also analyzed the pharmacokinetics using a two-compartment model with zero-order absorption and first-order elimination. This model was sufficient to obtain a good curve fit, and this demonstrates that the absorption process is not a first-order but a zero-order process. Therefore, there might be an upper limit to the rhG-CSF transfer rate from subcutaneous tissue to blood.

Population-based investigation of valproic acid relative clearance using nonlinear mixed effects modeling: influence of drug-drug interaction and patient characteristics.

Nonlinear mixed effects modeling (NONMEM) was used to estimate the effects of drug-drug interaction on valproic acid relative clearance values using 792 serum levels gathered from 400 pediatric and adult patients with epilepsy (age range, 0.3-54.8 years) during their clinical routine care. Patients received valproic acid as monopharmacy or in combination with either the antiepileptic drugs, phenobarbital, or carbamazepine. The final model describing valproic acid relative clearance was CL (mL/hr/kg) = 15.6.TBW (kg)-0.252.DOSE (mg/kg/day)0.183.0.898GEN.COPB.COCBZ, where COPB equals 1.10 if the patient is treated with phenobarbital, a value of unity otherwise, and COCBZ equals 0.769.DOSE (mg/kg/day)0.179 if the patient is treated with carbamazepine, a value of unity otherwise. Valproic acid relative clearance was highest in the very young and decreased in a weight-related fashion in children, with minimal changes observed in adults. This pattern was consistent whether valproic acid was administered alone or coadministered with phenobarbital or carbamazepine. When valproic acid was coadministered with phenobarbital or carbamazepine, valproic acid relative clearance increased as compared with that in monopharmacy. Its magnitude in the presence of carbamazepine increased in a valproic acid daily dose-related fashion. Concomitant administration of phenobarbital and valproic acid resulted in a 10% increase on valproic acid relative clearance. The clearance in female patients was approximately 10% less than that in male patients.

Population-based investigation of relative clearance of digoxin in Japanese patients by multiple trough screen analysis: an update.

The steady-state concentrations of digoxin at trough levels were studied to reestablish the role of patient characteristics in estimating doses for digoxin using routine therapeutic drug-monitoring data. The data (n = 548) showing steady-state serum concentrations of digoxin after repetitive oral administration in 385 hospitalized patients were analyzed using NONMEM, a computer program designed to analyze pharmacokinetics in study populations by allowing pooling of data. Analysis of the pharmacokinetics of digoxin was accomplished with a simple steady-state pharmacokinetic model. The effect of a variety of developmental and demographic factors on the clearance of digoxin was investigated. Estimates generated by NONMEM indicated that clearance of digoxin was influenced by the demographic variables of age, total body weight, serum creatinine, estimated creatinine clearance, gender, the coadministration of spironolactone, the presence or absence of congestive heart failure, and the administration of a half-tablet. The interindividual variability in the clearance of digoxin was modeled with proportional error with an estimated coefficient of variation of approximately 22%; the residual variability was approximately 25.0%. An a priori method, based on the value for clearance of digoxin obtained by NONMEM analysis, was proposed as a useful adjunct for the prediction of the steady-state concentration of digoxin at trough level as a function of the maintenance dose of digoxin.

Circadian rhythm of fever induced by interferon-alpha in mice.

The influence of dosing time on the fever induced by interferon-alpha (IFN-alpha) was investigated in ICR male mice under light-dark cycle. There was a significant circadian rhythm in rectal temperature, as an index of fever, at 1 hr after IFN-alpha (10 MIU/kg, i.v.) injection. The rhythmic pattern resembled overall the rhythm occurring in the nondrugged state. IFN-alpha increased rectal temperature during the light phase, but not during the dark phase. The fever induced by IFN-alpha was blocked by indomethacin (10 mg/kg, i.p.) pretreatment. There was no significant difference of plasma IFN-alpha concentrations at 0.167, 0.5 and 1 hr after IFN-alpha injection. Therefore the dosing time dependent difference of fever induced by IFN-alpha may be caused via that of PGE2 level elevated by IFN-alpha. The choice of the most appropriate time of day for drug administration may help to achieve rational chronotherapeutics of IFN-alpha.

Influence of dosing time on pharmacological action of G-CSF in mice.

The role of the susceptibility of living organisms and the pharmacokinetics of G-CSF on the rhythm of granulocyte colony-stimulating factor (G-CSF) activity was investigated. ICR male mice were housed in a standardized light-dark cycle (lights on at 0700, off at 1900) with food and water ad libitum. The leukocyte counts at 2 and 24 hr after G-CSF (250 microg/kg, i.v.) injection were significantly higher in mice injected with the drug at 0700 than at 1900 (p<0.01, respectively). The higher leukocyte-increasing effect corresponded to drug dosing at the time in which the granulocyte colony formation stimulated by G-CSF and DNA synthesis increased and the lower effect corresponded to drug dosing at the time in which they decreased. The rhythmicity corresponded to that in plasma G-CSF concentration. The present study suggests that the rhythm of G-CSF activity is caused by that of the sensitivity of bone marrow cells to the drug and the pharmacokinetics of the drug.

Relationship between diurnal rhythm of cell cycle and interferon receptor expression in implanted-tumor cells.

Whether the diurnal rhythm of cell cycle is associated with that of interferon-alpha/beta receptor (IFNAR) expression was investigated in implanted-tumor cells. The expression of IFNAR mRNA significantly increased when the proportion of tumor cells in DNA synthesis (S) phase increased in vitro. A diurnal rhythm was observed for cell cycle distribution in implanted-tumor cells. The specific binding of interferon-alpha to receptor and IFNAR mRNA increased when the proportion of tumor cells in S phase increased in vivo. The time-dependent expression of IFNAR was supported by that of transcription factor level induced by interferon-beta. The present result suggests that the rhythm of IFNAR expression is closely related to that of cell cycle distribution in implanted-tumor cells.

Gender- or age-related binding characteristics of valproic acid to serum proteins in adult patients with epilepsy.

The aim of the present study was to determine the gender- or age-related binding characteristics of valproic acid (VPA) to serum proteins in the adult population. Serum samples examined in the study were obtained from 70 adult patients (36 males, 34 females) with epilepsy on VPA monotherapy. Their age ranged from 16 to 68 years (mean age with (SD), 37.7 (15.7) years; <45 years, n=44; >/=45 years, n=26). The in vivo population binding parameters of VPA to serum proteins and theoretical minimal unbound serum VPA fraction (Fu) were determined using an equation derived from the Scatchard equation in: (1), all; (2), male and female subgroups; and (3), younger (<45 years) and older (>/=45 years) subgroups. There was a significant difference in serum concentration of unbound VPA between male and female patients. The mean association constant (K) was 0.010 microM(-1) in all, male, and female patients. The mean total concentration of binding sites (n(Pt)) was 1453 microM for all patients, and 1561 and 1394 microM for male and female patients, respectively. The Fu was 0.064 for all patients, and 0.060 and 0.067 for male and female patients, respectively. There were no significant differences in the binding characteristics of VPA to serum proteins between the male and female groups. On the other hand, there were significant differences in the serum albumin concentration and molar concentration ratio of free fatty acids to albumin in serum between the younger and older patients. The mean value of K was 0.016 microM(-1) for the younger patients and 0.007 microM (-1) for the older patients. The mean n(Pt) was 1157 microM for the younger patients and 1703 microM for the older patients. The Fu was 0.051 for the younger patients and 0.077 for the older patients. Thus, significant differences were observed in the binding characteristics of VPA to serum proteins between the younger and older groups. Our results show that age, but not gender, has significant influences on the binding characteristics of VPA to serum proteins in our patient population.

Estimation of rhG-CSF absorption kinetics after subcutaneous administration using a modified Wagner-Nelson method with a nonlinear elimination model.

The clearance of recombinant human granulocyte-colony stimulating factor (rhG-CSF) is known to decrease with dose increase, and to be saturable. The average clearance after intravenous administration will be lower than that after subcutaneous administration. Therefore, the apparent absolute bioavailability with subcutaneous administration calculated from the AUC ratio is expected to be an underestimate. The absorption pharmacokinetics after subcutaneous administration was examined using the results of the bioequivalency study between two rhG-CSF formulations with a dose of 2 microg/kg. The analysis was performed using a modified Wagner-Nelson method with the nonlinear elimination model. The apparent absolute bioavailability for subcutaneous administration was 56.9 and 67.5% for each formulation, and the ratio between them was approximately 120%. The true absolute bioavailability was, however, estimated to be 89.8 and 96.9%, respectively, and the ratio was approximately 108%. The absorption pattern was applied to other doses, and the predicted clearance values for subcutaneous and intravenous administrations were then similar to the values for several doses reported in the literature. The underestimation of bioavailability was around 30%, and the amplification of difference was 2.5 times, from 8 to 20%, because of the nonlinear pharmacokinetics. The neutrophil increases for each formulation were identical, despite the different bioavailabilities. The reason for this is probably that the amount eliminated through the saturable process, which might indicate the amount consumed by the G-CSF receptor, was identical for each formulation.

New and simple method for estimating metildigoxin dosing regimens by multiple trough screen analysis.

The pharmacokinetics of digitalis glycosides were studied using routine therapeutic drug monitoring data to evaluate the role of patient characteristics for estimating metildigoxin dosing regimens. The 232 serum glycoside concentration data at steady-state after repetitive oral administration in 144 hospitalized patients was analyzed using NONMEM, a computer program designed for population pharmacokinetic analysis that allows pooling of data. Pharmacokinetic analysis of digitalis glycosides was described using a simple steady-state pharmacokinetic model. The effect of a variety of developmental and demographic factors on glycoside clearance was investigated. NONMEM estimates indicated that this digitalis glycoside clearance was influenced by the demographic variables of age, total body weight, serum creatinine, gender, daily dose and the coadministration of spironolactone. An elderly patient was expected to have a lower rate of clearance than a young patient of equal body weight and serum creatinine. The interindividual variability in glycoside clearance was modelled with proportional error with an estimated coefficient of variation of 19.7% and the residual variability was 21.8% The dosing method based on glycoside clearance value obtained by NONMEM analysis allowed the prediction of the minimum steady-state glycoside concentration as a function of metildigoxin maintenance dose with acceptable error for therapeutic drug monitoring.

A feasibility study of the multiple-peak approach for pharmacokinetic screening: population-based investigation of valproic acid relative clearance using routine clinical pharmacokinetic data.

The multiple-peak approach was used to evaluate the effect of age, body mass, dose and gender on the population estimates of valproic acid relative clearance. Routine clinical pharmacokinetic data (n = 474) was collected from 250 patients receiving valproic acid and no other drug. The data was analysed according to a simple steady-state pharmacokinetic model using NONMEM, a computer program designed for population pharmacokinetic analysis allowing pooling of data. The final regression model for relative clearance (CL, mL kg-1 h-1) was: CL = 18.9 x body weight(-0.276) x daily dose(0.142) x gender where gender = 1 for males, 0.877 for females. NONMEM estimates suggested that the rate of valproic acid clearance decreased nonlinearly with increasing body weight in the maturation process, and increased nonlinearly with increasing valproic acid dose (mg kg-1 day-1). The clearance in females was about 11% less than in males. These estimates were similar to those detected from previous studies.

Phenobarbitone population pharmacokinetics from routine clinical data: role of patient characteristics for estimating dosing regimens.

Routine clinical pharmacokinetic data collected from patients receiving phenobarbitone have been analysed to evaluate the role of patient characteristics for estimating dosing regimens. The data were analysed using NONMEM, a computer program designed for population pharmacokinetic analysis that allows pooling of data. The pharmacokinetic model of phenobarbitone was described using a one-compartment steady-state model. The effect of a variety of developmental and demographic factors on clearance was investigated. NONMEM estimates indicated a nonlinear function of total body weight as the optimum adjustment of phenobarbitone clearance. Concomitant administration of phenobarbitone and other antiepileptic drugs showed a decrease of phenobarbitone clearance in young children. The dosing method based on clearance values obtained by NONMEM analysis allowed the prediction of the steady-state concentration as a function of maintenance dose with acceptable error for therapeutic drug monitoring.

One-point feedback control method for phenytoin dosage adjustment.

Routine clinical pharmacokinetic data collected from patients receiving phenytoin have been analysed to propose a new and simple equation to aid the dosage adjustment of this drug. The data were analysed using NONMEM, a computer program designed for population pharmacokinetic analysis that allows pooling of data. The rate equation for the elimination of phenytoin can be written as Do = kCssn, which fits the steady-state serum concentration (Css) and daily dose data (Do). The parameter n is the kinetic order and the parameter k is an arbitrary rate constant. From the above equation, D2 = D1C1 -nC2n can be derived, which forms the basis of predicting the dosage, D2, to obtain a desired Css, C2, using one initial Css, C1, obtained with an initial dose, D1, and using a population value of n. The value of n for phenytoin was estimated to be 0.312 in this study. The predictive performance of this equation was compared with the Richens and Dunlop nomogram and Bayesian feedback method using two or more steady-state concentration/dose pairs from each of 78 outpatients. This equation allowed the prediction of a dose needed to produce a desired steady-state concentration with errors comparable with the Bayesian feedback method for therapeutic drug monitoring.

Digoxin population pharmacokinetics from routine clinical data: role of patient characteristics for estimating dosing regimens.

Routine clinical pharmacokinetic data collected from patients receiving digoxin have been analysed to evaluate the role of patient characteristics for estimating dosing regimens. The data were analysed using NONMEM, a computer program designed for population pharmacokinetic analysis that allows pooling of data. The pharmacokinetic model of digoxin was described using a one-compartment steady-state model. The effect of a variety of developmental and demographic factors on clearance was investigated. NONMEM estimates indicate that digoxin clearance was influenced by the demographic variables of age, total body weight, serum creatinine and sex. The interindividual variability in digoxin clearance was modelled with additive error with an estimated standard deviation of 46.15 L day-1 and the intraindividual variability, or residual error was 0.209 ng mL-1. The dosing method based on clearance values obtained by NONMEM analysis allowed the prediction of the steady-state concentration as a function of maintenance dose with acceptable error for therapeutic drug monitoring.

Detection of carbamazepine-induced changes in valproic acid relative clearance in man by simple pharmacokinetic screening.

Selecting the optimum dose of valproic acid is difficult because the pharmacokinetics are complicated by inter-patient variability and by effects arising as a result of co-administration with other antiepileptic drugs. The multiple peak approach has been used to evaluate the effect of age, total body weight, dose, gender and comedication (carbamazepine-induced change) on population estimates of valproic acid relative clearance. Routine clinical pharmacokinetic data (n = 479) were collected from 207 epilepsy patients on combination therapy. The data were analysed by a simple steady-state pharmacokinetic model with the use of NONMEM, a computer program designed for population pharmacokinetic analysis that enables pooling of data. NONMEM estimates suggested that the rate of valproic acid clearance in patients receiving concomitant administration of valproic acid and carbamazepine decreased non-linearly with increasing total body weight in the maturation process, and increased non-linearly with increasing valproic acid dose. The clearance in females was 5.7% less than in males. NONMEM estimates also suggested that the rate of valproic acid clearance increased non-linearly with increasing carbamazepine dose. Concomitant administration of valproic acid and carbamazepine with other antiepileptic drugs resulted in an increase in valproic acid clearance of 10%. The final regression model of valproic acid relative clearance was CL = 6.06TBW-0.168 x DOSE0.414 x CBZDOSE0.093 x 0.943GEN x 1.10CO, where CL is the clearance (mL kg-1 h-1), TBW is the total body weight (kg), DOSE is the dose of valproic acid, CBZDOSE is the dose of carbamazepine, GEN = 0 for males and 1 for females and CO = 0 for concomitant administration of valproic acid and carbamazepine and 1 for concomitant administration of valproic acid and carbamazepine with other antiepileptic drugs. This technique can be used to estimate the pharmacokinetic parameters of a population from sparse data collected during routine clinical care and to determine the extent to which patient characteristics influence drug pharmacokinetics.

Comparison of two binding equations for prediction of the concentration of unbound valproic acid in the serum of adult epileptic polytherapy patients.

Because binding of valproic acid to plasma proteins affects the efficacy of the drug in the treatment of epilepsy (only the unbound fraction of the drug is effective) we have compared two methods which use different binding parameters to predict the in-vivo concentration of unbound valproic acid in serum. The study was performed on 46 serum samples from 29 polytherapy adult patients with epilepsy. Mean prediction error, mean absolute prediction error and root mean squared error were calculated for each method; these values served as a measure of prediction bias and precision. The mean absolute prediction errors and root mean squared errors for the two methods were similar in magnitude (Method 1, mean absolute prediction error = 10.0 microM, root mean squared error = 15.0 microM; Method 2, mean absolute prediction error = 10.3 microM, root mean squared error = 13.5 microM). Method 2 had a general tendency to over-predict unbound valproic acid; both methods had a tendency to over-prediction for total concentrations above 500 microM. Method 1 had a tendency to under-prediction at total concentrations below 250 microM. Within the total concentration range of valproic acid investigated, Method 1 was superior to Method 2 for prediction of unbound serum valproic acid. Our approach using Method 1 may be useful for prediction of unbound serum valproic acid concentration in patients with total valproic acid concentrations ranging from 250 to 500 microM; Method 2 may be useful for patients with total valproic acid below 500 microM. Our results suggest that there is wide and unpredictable variability in valproic acid binding to serum proteins among study populations.