Article Text

Download PDFPDF

Original articles
Effects of cytochrome P450 (CYP)2C19 polymorphisms on pharmacokinetics of phenobarbital in neonates and infants with seizures
  1. Soon Min Lee1,
  2. Jae Yong Chung2,3,
  3. Young Mock Lee1,
  4. Min Soo Park1,3,
  5. Ran Namgung1,
  6. Kook In Park1,
  7. Chul Lee1
  1. 1Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea
  2. 2Department of Pharmacology, Yonsei University College of Medicine, Seoul, Korea
  3. 3Department of Clinical Pharmacology, Severance Hospital, Yonsei University Health System, Seoul, Korea
  1. Correspondence to M S Park, Department of Pediatrics and Clinical Pharmacology, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul, 120-752, Korea; minspark{at}


Background Phenobarbital (PB), commonly used as the preferred treatment for neonatal seizure, is a drug that requires careful dose adjustments based on therapeutic drug monitoring. It has been reported that PB metabolism was affected by cytochrome P450 (CYP)2C19 polymorphisms in adults requiring dose adjustment.

Aim This study aimed to evaluate the effects of CYP2C19 genetic polymorphisms on PB pharmacokinetics (PK) in neonates and infants with seizures.

Methods CYP2C19 (wild type: CYP2C19*1/*1, heterozygous extensive metabolisers: CYP2C19*1/*2, *1/*3 and poor metabolisers: CYP2C19*2/*2, *2/*3) genetic polymorphisms in 52 neonates and infants with seizures were analysed. PK parameters were compared based on genotypes. The NONMEM program was used for population PK modelling.

Results No significant difference in PB clearance (CL), volume of distribution (Vd) and concentrations were shown among the CYP2C19 genotype groups. The results of PK modelling were as follows: Vd=3590 ×(body weight (BWT)/4)0.766 ×(AGE/2)0.283 and CL=32.6×(BWT/4)1.21.

Conclusions PB PK parameters of neonates and infants with seizures were not significantly different among the groups with different CYP2C19 genotypes. The addition of CYP2C19 genotyping to PK models did not improve the dosing strategies in neonates and infants.

View Full Text

Statistics from


Genetic polymorphisms of cytochrome P450 (CYP) enzymes mainly influence drug metabolism, disposition and elimination and lead to alteration of pharmacokinetic (PK) parameters resulting in changes in drug response as well as adverse drug reactions.1 2 Differences in PK parameters caused by CYP polymorphisms could be overcome by drug dosage adjustment.3 Phenobarbital (PB) is eliminated by CYP450 2C9-dependent and/or 2C19-dependent oxidation (25%) and N-glucoside formation (25%) and renal excretion (25%).4 5 In fact, CYP2C19 genetic polymorphisms have been shown to influence the clearance (CL) of PB in adult patients requiring dose adjustment.6 7 However, because of the developmental status of hepatic metabolising enzymes, the enzyme expression patterns and activities and consequently the PK responses of neonates and infants are quite different from those of adults.8,,15 Further investigation into the PK profiles of neonates and infants with regard to genetic polymorphisms of relevant hepatic enzymes is warranted.

What is already known on this topic

  • Phenobarbital metabolism in adults is affected by genetic polymorphisms.

  • Differences in genetic polymorphisms in adults result in different dosage requirements.

What this study adds

  • The cytochrome P450 (CYP)2C19 genotype did not affect phenobarbital pharmacokinetics in infants.

  • The inclusion of CYP2C19 genotype as a covariate in population pharmacokinetic modelling would not help creating optimal dosing strategies for phenobarbital.

In the present study, we aimed to evaluate the effects of the CYP2C19 genetic polymorphisms on PB PK profiles in neonates and infants with seizures for an optimal dosing strategy using the population PK method.

Materials and methods

Study population

The protocol was approved by the institutional review board of the Yonsei University College of Medicine. In this study, 52 neonates and infants who had been diagnosed as having seizures and hospitalised from December 2007 to December 2009 with PB monotherapy were enrolled. Due to incomplete sampling before reaching steady state and unstable metabolism of the early neonatal period (first week of birth), eight neonates were excluded. Among the 44 infants (27 boys and 17 girls) there was an age range of 8 days to 6 months and a body weight range of 2.1 to 8.0 kg. Due to the wide range of age and body weight, a subgroup analysis based on age to control for the age and body weight (group 1 from 8 days to 3 months and group 2 from 4 months to 6 months) was performed. Infants with a severe structural anomaly in the central nervous system, severe systemic illnesses, hepatic or renal failure, congenital haemolytic anaemia or genetic disorders were excluded. All the infants included in the study received either breast milk or formula for their diet.

All infants were given one loading dose of 20 mg/kg PB intravenously, followed by a maintenance dose of 2.5 mg/kg separated by a 12 h interval daily. If seizure was uncontrollable, multiple loading doses of 10 mg/kg PB intravenously were added. A total of 115 trough serum concentrations (1–4 samples per patient) were obtained: first measurements were usually taken at 48 h (range 35–59 h) and the next at 11.5 days (range 8–19 days) from the initiation of treatment. The drug concentrations were measured using a fluorescence polarisation immunoassay method (Abbott TDx analyser, Abbott Laboratories, Abbott Park, IL, USA). Therapeutic drug monitoring (TDM) analysis was performed using the Abbott-based PK software program. All infants were classified into homozygous extensive metabolisers (EMs) for the wild type, heterozygous EMs and poor metabolisers (PMs) by genotyping.

CYP2C19 genotyping

Genomic DNA was isolated from peripheral lymphocytes using an extraction kit. The genotyping of CYP2C19 polymorphism (991A>G (I331V), 681G>A (P227P; splicing defect) and 636G>A (W212X)) was screened by single base primer extension assay using an ABI PRISM SNaPShot Multiplex kit and TaqMan fluorogenic 5' nuclease assay (ABI, Foster City, California, USA). After amplification, the PCR products were added to a SNaPshot Multiplex Ready reaction mixture. The final reaction samples were analysed by electrophoresis in the ABI Prism 3730xl DNA analyser and the ABI PRISM 7900 HT Sequence Detection System (ABI). Duplicate samples and negative controls were included to ensure accuracy of genotyping.

Population PK model development

Population PK modelling was carried out using the non-linear mixed effects software in NONMEM V.6.2 (ICON, Ellicott City, Maryland, USA).16 A one-compartment intravenous administration model with first-order elimination was used. The body weight (BWT), age, gender, PB daily doses, genotypes of CYP2C19 and laboratory findings (aspartate aminotransferase, alanine aminotransferase, protein, albumin, blood urea nitrogen and creatinine) were screened as covariates to test the significant influence on CL or volume of distribution (Vd).

Statistical analyses

One-way analysis of variance (ANOVA) was performed with SPSS software V.17.0 (SPSS, Chicago, Illinois, USA). p Values less than 0.05 were considered statistically significant.


The demographic characteristics of the patients by CYP2C19 genotypes are shown in table 1. No significant differences were seen in the sex ratio, age, body weight and daily doses of PB among CYP2C19 genotype groups. The serum concentrations of all the infants at the second measurement were within the steady state therapeutic concentration ranges (15–40 μg/ml).

Table 1

Patient characteristics by cytochrome P450 (CYP)2C19 genotype

In both age groups, there were no statistically significant differences in the steady state PB concentrations, CL and Vd based on CYP2C19 genotype (tables 2 and 3). However, when the two age groups were compared, we found higher PB concentrations and lower CL in group 1. A wide interindividual variation in the plasma concentrations among the genotype groups was noted though it is not statistically significant (figure 1). Renal function, as determined by plasma creatinine and liver function tests, were normal for all infants. Statistical comparisons between the different groups based on their genotype showed no differences in biochemical parameters.

Figure 1

(A) Scatter plot of the individual concentration of phenobarbital (PB) grouped by cytochrome P450 (CYP)2C19 genotypes. (B) Scatter plot of the individual clearance of PB grouped by CYP2C19 genotypes. 1=wild extensive metaboliser (EM) type (*1/*1), 2=hetero EM (*1/*2, *1/*3), 3=poor metabolisers (PM) (*2/*2, *2/*3). The horizontal lines represents the median and 95% CI.

Table 2

Pharamacokinetic analysis by cytochrome P450 (CYP)2C19 genotype in infants in group 1

Table 3

Pharmacokinetic analysis by cytochrome P450 (CYP)2C19 genotype in infants in group 2

For PB population modelling, each covariate such as PB daily dose, CYP2C19 genotypes and laboratory findings did not show a significant relationship with CL or Vd. Influence of BWT on CL and Vd and age on Vd were significant after forward selection and backward elimination. The final model was as follows: typical value of Vd (ml)=3590×(BWT/4)0.766×(AGE/2)0.283 (where coefficient of variation of Vd=31.1%); typical value of CL (ml/h)=32.6×(BWT/4)1.21 (where coefficient of variation of CL=27.0%).


The effects of CYP2C19 polymorphisms on the PK of PB in neonates and infants have not previously been evaluated or reported. In the present study, we characterised the CYP2C19 genetic polymorphisms of infants with seizures and compared the PK parameters based on the genetic profiles. Unlike in adults,6 7 the effects of genetic polymorphisms of CYP2C19 on the PB PK were not significant in infants. Population PK modelling showed that CYP2C19 as a covariate did not improve the PK models but confirmed that body weight was the most important covariate for CL and age and body weight for Vd.

PB requires therapeutic drug monitoring for efficacy and toxicity. In general, the half-life of PB varies with age from 114.2±40.3 h during the early neonatal period to 41.2±13.9 h after 1 month of age.5 17,,20 CL is also variable between 0.0053 and 0.0141 litres/kg/h during the early period and increases with age.5 19 20 Our data showed a mean CL of 0.0084 litres/kg/h and Vd of 0.9 litres/kg, which are compatible with published results.

In a report on the effects of CYP2C19 polymorphism on PB PK in Japanese adult patients, the total CL of PB significantly decreased by 18.8% in CYP2C19 PMs relative to that in EMs.7 Another report showed significant effects of CYP2C9 polymorphism on PB CL, where the total CL of PB decreased by 48% in patients who were CYP2C9 PMs in comparison with those with wild type (p<0.001).6 In our study, CL and Vd of infants who were CYP2C19 PM were not significantly different from either wild type or heterogeneous EM groups. This lack of difference may be due to the CYP450 isoform-specific developmental status in neonates.11 CYP 2C19 expression at the mRNA level reaches the adult level soon after birth. However, the protein level of infants increases slowly and takes 5 years to reach the adult level.8 The enzyme activity in infants is approximately 30% of that in adults, which is achieved by 1 year of age.12,,14 The total CYP contents in the liver are relatively constant until 1 year of age, remaining 30% to 60% of the adult level.21 Because of the relatively low metabolic enzyme activities of the wild types during early infancy, the differences in enzyme activities between wild types and PMs would not be as prominent as seen in adults.

Using population PK modelling, some discrepancies exist between the models in adults. Mamiya et al studied 74 adults (age range 17–76 years) and showed CYP2C19 PM (*2/*2, *2/*3) to have significant effects on CL in their model.7 However, Goto et al showed that body weight and CYP2C9PM (*1/*3) had influence on the CL, but not on CYP2C19 PM, where the population studied included paediatric and adult patients with a mean age 13.72 years (range 0.8–43.8 years).6 In our study, the model for CL showed that CL was mainly influenced and predicted by a function of body weight, while the model for Vd was a function of body weight and age. CYP2C19 genotype status added as a covariate did not improve the models.

There are some limitations to our study: the small number of patients, stratification of the study participants into different age groups that further decreased the number of patients in each group, and limited genotyping (2C19*1,*2,*3 only). In our study, we also genotyped and found three infants with CYP2C9*1/*3, all with CYP2C19 wild type, that did not show PK differences compared with infants with wild type of CYP2C9 and excluded them from the analysis of the effects of CYP2C19 genotype (data not shown).

In conclusion, this is the first study to analyse the effects of CYP2C19 polymorphisms on the PK of PB in infants with seizures. Unlike in adults, CYP2C19 PMs in this age group did not have a significant influence on the PK profiles of PB. Addition of the CYP2C19 genotype would not improve personalised optimal drug dosing for PB treatment for neonates or infants with seizures.


This study was supported by a faculty research grant from Yonsei University College of Medicine for 2009(6-2009-0094).


View Abstract


  • Funding This work was funded by a faculty research grant from Yonsei University College of Medicine.

  • Competing interests None.

  • Ethics approval Yonsei University IRB no. 4-2008-0029.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.