Phenytoin and phenobarbital stable isotope studies in neonates

Phenytoin and Phenobarbital Stable Isotope Studies in Neonates Saleem I. Malik, MD*, Michael J. Painter, MD*, Raman Venkataramanan, PhD†, and John D. Alvin, PhD† A pharmacokinetic study of phenytoin and phenobarbital with nonradioactive isotopes was performed in nine neonates in an intensive care unit setting. A single-pulse dose of either labeled phenobarbital (1,315 N, 2-13C) or labeled phenytoin (2-13C, 1, 3-15N) was administered to neonates who manifested gestation between 25 and 40 weeks and were receiving maintenance medication. Blood samples were collected at fixed intervals, and with a computerized gas chromatography mass spectrometry system, plasma concentrations of the labeled and unlabeled drug in relation to time administered were obtained. According to the calculations obtained from labeled analogue, several kinetic characteristics related to drug absorption, clearance, and elimination were determined. The use of a nonradioactive labeled isotope overcomes the limitations of conventional pharmacokinetic methodology and can be specifically useful in neonates and infants in whom volumes of distribution are rapidly changing and steady state is not achieved. © 2003 by Elsevier Inc. All rights reserved. Malik SI, Painter MJ, Venkataramanan R, Alvin JD. Phenytoin and phenobarbital stable isotope studies in neonates. Pediatr Neurol 2003;29:376-380.

Introduction A stable isotope-labeled compound has one or more of the atoms of the molecule replaced by its nonradioactive isotope, which has a different number of neutrons in its nucleus. This replacement imparts a different molecular weight to the labeled compound compared with the naturally occurring counterpart, and this difference is detectable by mass spectrometry. By interfacing a gas chromatograph with the mass spectrometer, it is possible to monitor

From the *Department of Pediatric Neurology, Children’s Hospital of Pittsburgh, and the †School of Pharmacy, University of Pittsburgh, Pittsburgh, Pennsylvania.

376

PEDIATRIC NEUROLOGY

Vol. 29 No. 5

the mass of the labeled molecule or one of its fragments, as well as the mass of natural unlabeled drug. This isotopic substitution can be detected according to the molecular weight of an ionized mass fragment of the drug. For labeled phenytoin, the mass fragment (m/z 226) is three mass units higher then the unlabeled mass fragment of the ion (m/z 223). This enrichment does not change the chemical or pharmacologic behavior of the labeled molecule but does permit the simultaneous detection and quantitation of the natural drug and its weight variant analogue [1,2]. Introduction of isotopically labeled medication and subsequent use of mass spectrometry to measure the proportion of heavy to light isotopes of the same element in plasma samples permits highly sensitive and specific extended pharmacokinetic determinations [3-5]. The intelligent use of antiepileptic agents in neonates requires an understanding of pharmacokinetic characteristics [6]. Although conventional pharmacokinetic techniques have provided useful information about the metabolism of antiepileptic medications, this methodology is limited [7-10]. In many circumstances, it assumes that steady state is achieved when in fact it is not in neonates. Administering a single dose or discontinuing therapy and monitoring plasma concentrations is not therapeutically acceptable [11-13]. The labeled pulse dose equilibrates rapidly with the existing tissue and fluid pools of unlabeled drug and is eliminated in a way that is reflective of the disposition processes acting on the total body pool of unlabeled drug. Stable isotope methodology allows measurements without interfering with existing body pools of drug or requiring extra doses. No interruption of the prescribed dosage regimen is required, which is important. As a result, the pharmacokinetics of a single dose during chronic therapy can be described mathematically, and the characteristics can be compared with a previous pulse dose in the same patient to determine whether dispositional conditions have

Communications should be addressed to: Dr. Malik; Department of Pediatric Neurology; Children’s Hospital of Pittsburgh; 3705 Fifth Avenue; Pittsburgh, PA 15213. Received March 19, 2002; accepted May 18, 2003.

© 2003 by Elsevier Inc. All rights reserved. doi:10.1016/S0887-8994(03)00304-7 ● 0887-8994/03/$—see front matter

Isotopes, Cambridge, MA. Purity and dosage forms of the drugs were determined and formulated at the School of Pharmacy, University of Pittsburgh. Stock solutions of the drugs were prepared in methanol at 1.0 mg/mL. Standard solutions of labeled phenytoin were prepared by serial dilution in human plasma to concentrations ranging from 1.0 to 10.0 ␮g/mL; of unlabeled phenytoin, from 2.0 to 30 ␮g/mL. Standard solutions of labeled phenobarbital were prepared by serial dilution in human plasma to concentrations ranging from 5.0 to 30 mg/mL; of unlabeled phenobarbital, from 10 to 50 ␮g/mL.

Analysis

Figure 1. Chemical structure of nonlabeled phenytoin. Atoms marked by jagged lines indicate positions of stable isotope substitution.

changed. Included among these characteristics are the individual constants for absorption, metabolism, half-life, distribution, and excretion. These descriptors can also be used for comparison between individual patients. The National Institutes of Health has identified this methodology as particularly well adapted to children and pregnant women [2]. Materials and Methods

Gas chromatography-mass spectrometry was performed on an HP5890 gas chromatograph interfaced with a HP5971 mass spectrometer controlled by a Hewlett Packard Chemstation software package. The column was a DB5, 30 m ⫻ 0.25 mm inside diameter, 0.5 micron film thickness, from J&W (Folsom, CA). The carrier gas was 99.95% high-purity helium; the head pressure was 2 kg/cm2; the split ratio was 25:1. The gas chromatograph column temperature was 260°C ramped to 300°C at 10°C per minute. The final temperature was maintained for 10 minutes. The ions monitored were 204 and 207 m/z for unlabeled and labeled phenobarbital, respectively, and 223 and 226 m/z for phenytoin, respectively (Figs 1, 2). Peak height ratios to the internal standard methylphenobarbital were used to determine the concentration of the species. Quantitative removal of the drugs from plasma, obtained from 0.5 mL blood for each period, was by two successive extractions with 5 mL methylene chloride. Internal standard solution (10 ␮L of a 200 ␮g/mL solution) was added to all samples and standards before extraction. No internal standard was added to blank specimens. Extracts were reduced to approximately 100 ␮L under a gentle stream of dry nitrogen. Ten microliters of concentrated extract was injected onto the gas chromatographic column for each analysis.

Chemicals Pharmacokinetic Studies Phenobarbital, phenytoin, and methylphenobarbital were obtained from Sigma Chemical Corporation (St. Louis, MO). Isotopically labeled phenobarbital (2-15N, 13C-phenobarbital) and isotopically labeled phenytoin (2-13C-1, 3 15N-phenytoin] were synthesized by the condensation method of Alvin and Bush. Isotopic precursors were obtained from KOR

A total of nine neonates with a history of neonatal seizures caused by clinically diagnosed perinatal asphyxia and with gestation between 25 and 40 weeks were identified in the neonatal intensive care unit. All neonates were treated with phenobarbital initially and then phenytoin if

Figure 2. Fragmatogram of labeled and nonlabeled phenobarbital (PB) with internal standardization (IS), showing separate peaks secondary to different molecular weights of labeled and nonlabeled phenobarbital.

Malik et al: Stable Isotope in Neonates 377

Table 1.

Metabolism of isotopically labeled phenytoin in neonates (noncompartmental analysis)

Patient

1

Weight (kg) Age (wk) Serum albumin (g/dL) r2 Route T1/2 (hr) K (hr⫺1) Dose (mg) AUC0-00 (␮g/mL/hr) AUMC0-00 MRT (hr) CL (mL/hr) CL (mL/min/kg) Bioavailability (%) Vss (mL) Vss (mL/kg)

2

4.84 40 1.8 0.993 PO

107.9 2,428 22.5

3.15 40 2.30 0.9811 IV

14.8 0.0468

12.17 0.0563 10.60 75 1,314 17.4

141.3 0.487 100 2,459 508

PO 28.4 0.0244 147 6,263 42.6

PEDIATRIC NEUROLOGY

Vol. 29 No. 5

PO

28.4 0.0244 7.50 142 6,149 43

52.8 0.279 100 2,270 721

seizures were not controlled. This study was reviewed and approved by the internal review board of Children’s Hospital of Pittsburgh. Three neonates who were receiving maintenance intravenous phenytoin for 2 to 4 days were administered a pulse dose of isotopically labeled phenytoin. One half of the calculated 24-hour maintenance dose was administered intravenously as standard phenytoin, and one half was administered orally as the 13C, 15N-labeled analogue. Blood samples were then collected at 0.5, 1, 2, 4, 6, 10, 24, 48, 72, 96, 120, and 144 hours, and phenytoin (labeled and unlabeled) concentrations were measured in serum by gas chromatography-mass spectrometry. One neonate (Patient 4) received only oral phenytoin. Linear pharmacokinetics across the observed plasma concentrations and absence of any isotope effect on the metabolism of phenytoin was assumed before data analysis. From the total concentration of phenytoin in each plasma sample, the concentration of the stable isotope-labeled phenytoin in the corresponding sample was subtracted to obtain the plasma concentration of unlabeled phenytoin. The plasma concentration of unlabeled phenytoin (because of intravenous dosing) was analyzed by a noncompartmental model by using a basic program developed at our institution. The plasma concentration of stable isotope-labeled phenytoin (because of oral dosing) was analyzed by a noncompartmental model using a Basic program developed in our institution and using a onecompartmental oral model with Win-Nonlin (Pharsight Corporation, Mountain View, CA). A minimum of three data points in the terminal linear disposition phase was used to calculate disposition rate constant (K) and apparent disposition half-life (0.693/K). The principle of reverse superposition was used to assess the area under the plasma concentration vs time, 0-␣, due to the dose studied. Area under the first moment of the curve and mean residence time, volume of distribution, clearance, and bioavailability were calculated according to standard procedures. Five neonates receiving maintenance phenobarbital for 3 to 11 days received a single dose of isotopically labeled phenobarbital orally. Blood samples were then collected at 0.5, 1, 2, 4, 6, 10, 24, 48, 72, 96, 120, and 144 hours, and the concentrations of the drug labeled and unlabeled with stable isotope were determined for each sample. Volume of distribution, clearance, and half-lives were then calculated from serum concentration vs time relationship in a similar fashion as described for phenytoin.

4

0.76 25 1.30 0.976 IV

Abbreviations: AUC ⫽ Area under the plasma concentration versus time curve AUMC ⫽ Area under the first moment of plasma concentration versus time curve Bioavailability ⫽ AUCpo/AUCiv CLIV ⫽ Doseiv/AUCiv IV ⫽ Intravenously

378

3

K MRT PO r2 Vss

30.65 0.0226 86.7 3,855 44.5

3.88 40 2.50 IV

PO

30.65 0.0266 1.75 173.6 7,367 42.4

13.2 0.0526 9.30 67.5 1,282 19

IV

—

10.1 0.22 50 428 563 ⫽ ⫽ ⫽ ⫽ ⫽

Terminal disposition rate constant Mean residence time (AUMC0/AUC0) Orally Correlation coefficient for terminal log linear segment CL ⫻ MRT ⫽ (dose/AUMC)/(AUC)2

Results Table 1 presents the various pharmacokinetic characteristics calculated after noncompartmental analysis of intravenous and orally administered data. The correlation coefficient associated with the calculation of K ranged from 0.976 to 0.997. The apparent disposition half-life could be calculated from all four patients after an oral dose of phenytoin and after intravenous administration. Patient 4 did not receive intravenous phenytoin. The mean (⫾ S.D.) apparent disposition half-life of stable isotopelabeled phenytoin after oral administration was 21 (⫾ 9.8) hours. The mean (⫾ S.D.) systemic clearance of phenytoin was 0.33 (⫾ 0.14) mL/min/kg. The mean (⫾ S.D.) volume of distribution at steady state was 597 (⫾ 111) mL/kg. The bioavailability ranged from 50% to 100%. Table 2 presents the various pharmacokinetic characteristics after compartmental analysis of the oral data. The apparent disposition half-life was a little lower (18.4 ⫾ 7 hours) but was comparable to what was obtained by noncompartmental analysis of the same data. The V/bioavailability calculated was consistent with what is obtained from the intravenous data. For the purpose of illustration, one representative plot from a patient who received labeled phenytoin is depicted in Figure 3. Similar methods were used to calculate the t1/2 in patients receiving maintenance phenobarbital (Table 3). For the purpose of illustration, one representative plot from a patient who received 7 mg of labeled phenobarbital is depicted in Figure 4.

Table 2. Metabolism of isotopically labeled phenytoin in neonates (one-compartmental analysis) Patient

1

2

Weight (kg) 4.84 3.15 Age (wk) 40 40 Serum albumin (gm%) 1.8 2.30 Dose (mg) 10.6 7.5 Cmax (mg/mL) 4 3.4 Tpeak (hr) 6 6 0.477 1.31 Ka (hr⫺1) 0.0438 0.0263 K (hr⫺1) V/F (mL) 2,122 2,042

3

4

0.76 3.88 25 40 1.30 2.50 1.75 9.3 2.15 3.7 4 4 1.67 0.848 0.0325 0.069 783 2,258

Abbreviations: Cmax ⫽ Maximum plasma concentration Ka ⫽ Absorption rate constant Kd ⫽ Disposition rate constant Tpeak ⫽ Time to reach maximum plasma concentration V/F ⫽ Apparent volume of distribution/bioavailability

Discussion Limited studies have been performed on the pharmacokinetics of phenytoin in neonates. Complete and proper characterization of the systemic clearance and volume of distribution is also lacking. In the present study, the systemic clearance, volume of distribution, and half-life of phenytoin could be calculated for three patients after intravenous administration of phenytoin. With the use of stable isotope methodology, it was possible to simultaneously evaluate the intravenous and orally administered kinetics of phenytoin. Lowest clearance (per kilogram body weight), longest half-life, and lowest bioavailability were obtained in the most premature patient. Gestation did not appear to alter the volumes of distribution at steady state normalized to body weight. Large variation in the half-life of phenytoin has been reported during the first week of age, and the half-life

Figure 3. Semilogarithmic plot of plasma concentration versus time of unlabeled (•) and labeled (‚) phenytoin in a single subject.

Table 3. Metabolism of isotopically labeled phenobarbital in neonates Patient 1 2 3 4 5

Gestation (wk)

t1/2 (hr)

35 36 35 31 34

42 34 56 68 71

decreases with age [8,10]. The half-life has also been reported to be concentration dependent [9]. The volume of distribution of phenytoin is reported to be 0.6-0.7 L/kg in adults in patients with normal albumin levels. The volume of distribution in neonates reported in a previous study was larger (0.89 L/kg) than the volume of distribution measured (0.597 L/kg) in the present study, which may be related to the different methods [dose/ theoretical concentration at time 0 in the publication and Dose ⫻ AUMC/(AUC)2, where AUMC ⫽ area under the first moment of the plasma concentration vs time curve and AUC ⫽ area under the plasma concentration vs time curve, in this study] used to calculate this parameter. The absorption rate constant for phenytoin was at least 10 times higher than the corresponding disposition rate constant. The time to reach plasma concentrations (4-6 hours) is within the range (3-12 hours) reported in the literature [14]. The bioavailability of phenytoin was 100% in two patients and 50% in the patient with the youngest gestation. Bioavailability of high-quality phenytoin product has been reported to be 100% [15-17]. Our data on bioavailability do not explain the difficulty in achieving therapeutic plasma phenytoin levels in neonates with standard oral phenytoin products. The preparation used the same diluent as the intravenous form, and there may well be differences in absorption, depending on the oral preparation.

Figure 4. Semilogarithmic plot of plasma concentration versus time of unlabeled (•) and labeled (•) phenobarbital in a single subject.

Malik et al: Stable Isotope in Neonates 379

Five neonates received an oral labeled phenobarbital dose, and the pharmacokinetic studies demonstrated good absorption and elimination. The data were obtained without altering the medically indicated therapeutic regimen or assuming the presence of steady state. Absorption and half-life of the drug in the five neonates are similar to the previously reported values [18,19]. Conventional pharmacokinetic approaches to obtain these data would have demanded achievement of steady-state, single-dose administration or cessation of therapy. These data were obtained without altering the medically indicated therapeutic regimen or assuming the presence of steady state. The use of stable, nonradioactive, isotopically labeled drug analogues in pharmacokinetic studies has been described and is safe. A single tracer study can be used to determine different pharmacokinetic values while the patient is receiving maintenance therapy. This pulse-dose study can also be used to perform studies before long-term therapy is initiated, and then one or more single pulse-dose studies can be performed during long-term therapy to determine the presence or absence of time-dependent or nonlinear pharmacokinetics. Similarly, serial studies over time of the pharmacokinetics of a drug in a cohort of children to determine the timing and magnitude of agedependent changes in distribution and biotransformation can be performed. Another attractive feature is the ability to use stable isotope methodology to study variances in the bioavailability of different preparations (suspensions, elixir chewable tablets, etc.). Traditionally, studies of relative bioavailability have been performed by administering the two preparations at different times to the same patient. With stable isotope methodology, relative bioavailability can be studied precisely by administering both forms simultaneously (labeled and unlabeled, intravenously and orally) and then quantitating the relative contributions to plasma concentrations with a gas chromatograph-mass spectrometer computer system. This method has the advantage of eliminating the need to subject the patient to a procedure more than once and reducing the number of patients required to obtain reliable data. Additionally, well-designed stable isotope methodology can provide useful information about time-dependent or dose-dependent pharmacokinetic changes, as well as different drug interactions, in the pediatric age group [20,21]. This methodology provides a valuable tool to investigate old and new antiepileptic medications in the pediatric age group, an arena in which studies often lag behind and adult data are often substituted and adopted empirically.

380

PEDIATRIC NEUROLOGY

Vol. 29 No. 5

References [1] Browne TR, Langenhove AV, Costello CE, et al. Applications of stable isotope methods to studying the clinical pharmacology of antiepileptic drugs in newborns, infants, children, and adolescents: Therapeutic drug monitoring. New York: Raven Press, 1984, 6:3-9. [2] National Institutes of Health. Stable isotopes provide clinically useful tools for a variety of metabolic studies. JAMA 1982;247:1926. [3] Klein P, Kein R. Stable isotopes: Origins and safety. J Clin Pharmacol 1986;26:378-92. [4] Horning MG, Lertratanangkoon K, Nowlin J, et al. Anticonvulsant drug monitoring by GC-MS-COM techniques. J Chromatogr Sci 1974;12:630-4. [5] Brown TR, Szabo GK, Ajami A, et al. Performance of human mass balance studies with stable isotope: Labeled drug and continuous flow isotope ratio mass spectrometry: A progress report. J Clin Pharmacol 1998;38:309-14. [6] Painter MJ, Gaus LM. Neonatal seizures, diagnosis and treatment. J Child Neurol 1991;6:101-8. [7] Chiba K, Ishizaki T, Miura H, et al. Michaelis-Menten pharmacokinetics of diphenylhydantoin and application in the pediatric age patient. J Pediatr 1996;3:479-84. [8] Bourgeois BFD, Dodson WE. Phenytoin elimination in newborns. Neurology 1983;33:173-8. [9] Dodson WE. Phenytoin kinetics in children. Clin Pharmacol Ther 1980;27:704-7. [10] Loughnan PM, Greenwalk A, Purton WW, et al. Pharmacokinetic observations of phenytoin disposition in the newborn and young infant. Arch Dis Child 1977;52:302-9. [11] Browne TR, Wilder BJ. Pharmacokinetics of antiepileptic drugs. Neurology 1998;51(Suppl. 4):S2-7. [12] Browne TR, Greenblatt DJ, Szabo GK, et al. New pharmacokinetic methods: I. Estimation of mean serum concentration from trough serum concentration. J Clin Pharmacol 1990;30:482-8. [13] Brown TR, Greenblatt DJ, Evans JE, et al. Pharmacokinetics: Dose-dependent changes. J Clin Pharmacol 1986;26:463-8. [14] Evans WE, Schentag JJ, Jusko WJ. Applied pharmacokinetics. Vancouver, WA: Applied Therapeutics Inc., 1992:25.1-25.44. [15] Nelson MH, Birnbaum AK, Nyhus PJ, et al. A capillary GC-MS method for analysis of phenytoin and [13C3]-phenytoin from plasma obtained from pulse dose pharmacokinetic studies. J Pharmaceutic Biomed Anal 1998;17:1311-23. [16] Browne TR, Davoudi H, Donn KH, et al. Bioavailability of ACC-9653 (phenytoin prodrug). Epilepsia 1989;30(Suppl. 2):S27-32. [17] Kapetanovic IM, Kupferberg HJ. Stable isotope methodology and gas chromatography mass spectrometry in a pharmacokinetic study of phenobarbital. Biomed Mass Spectrom 1980;7:47-52. [18] Painter MJ, Alvin JD, David R. Use of a nonradioactive isotope to study phenobarbital metabolism in the neonate. Ann Neurol 1983;14:3. [19] Browne TR, Szabo GK. New pharmacokinetic methods for the study of antiepileptic medications of the 1990s. Epilepsia 1991;32(Suppl. 5):S66-73. [20] Painter MJ, David R, Alvin J. The use of a stable isotope to study phenytoin metabolism in the newborn. Program and Abstracts, Child Neurology Society, 379. [21] Dodson WE, Bourgeois BFD. Phenobarbital and phenytoin in neonatal seizures [letter]. Ann Neurol 1982;32:1405.