Influence of cytochrome P450 3A5 (CYP3A5) genetic polymorphism on dose-adjusted plasma levels of carbamazepine in epileptic patients in South Indian population.

AIM: The aim of the study was to compare the dose-adjusted plasma levels of carbamazepine (CBZ) among expressers and nonexpressers of cytochrome P450 3A5 (CYP3A5)* 3 genotypes. SUBJECTS AND METHODS: The study was carried out in 100 epileptic patients who were on CBZ monotherapy. Steady-state plasma CBZ levels were measured using reverse-phase high-performance liquid chromatography method, and genotyping of CYP3A5 was done using real-time polymerase chain reaction method.
RESULTS: Patients inheriting CYP3A5*3/*3 variant (nonexpressers) had an increased plasma concentration of CBZ (4.86 μg/ml) when compared to patients inheriting either CYP3A5*1/*1 or CYP3A5*1/*3 (expressers) (4.3 μg/ml, P = 0.004). Nonexpressers had significantly increased plasma concentrations of CBZ when adjusted for dose and weight when compared to expressers (P < 0.002 and P < 0.001, respectively). The frequency of adverse reactions in expressers and nonexpressers was 12% and 9%, respectively.
CONCLUSION: There is a significant influence of CYP3A5*3 genetic polymorphism (6986A>G) on dose-adjusted plasma levels of CBZ in epileptic patients in the South Indian population. Copyright:

Introduction

Antiepileptic drugs are the mainstay of treatment in patients with epilepsy, and long-term therapy is usually needed for most of them. Treatment is always aimed at preventing seizures preferably with a single antiepileptic drug which has more efficacy and lesser side effects, thereby improving their quality of life. Carbamazepine (CBZ) is widely prescribed antiepileptic drug for the treatment of focal seizures and generalized tonic–clonic seizures. It gets metabolized by the cytochrome P450 3A (CYP3A) enzymes mainly by CYP3A4 and CYP3A5.[1] Among the patients receiving the same dose of CBZ, it exhibits marked interindividual variation in the plasma drug concentrations, which may lead to therapeutic failure or toxicity. Genetic polymorphisms in the drug-metabolizing enzymes might play a major role in explaining such interindividual variations, thereby aiding in better selection of appropriate dosage of CBZ. Among various CYP3A5 genetic polymorphisms identified, CYP3A5*3 (6986A>G rs776746) has been found to have more interethnic variations in the allele frequencies. The individual inheriting CYP3A5*3 variant allele may poorly metabolize CBZ which may lead to elevated plasma CBZ drug levels. The genetic polymorphisms of CYP3A5 have been well studied in South Indian healthy volunteers, and the frequency of CYP3A5* 3 allele in the South Indian population was found to be 56%.[2] Moreover, it is well established that the genetic makeup of Ancestral South Indians is distinct from other major ethnic populations.[3] Hence, the results obtained from the studies which were done in other ethnic populations cannot be extrapolated to the South Indian population and need to be studied in the South Indian population. Therefore, we planned this study to explore the effect of CYP3A5* 3 genetic polymorphism on dose-adjusted plasma levels of CBZ in South Indian epileptic patients.

Subjects and Methods

The protocol was approved by the institute ethics committee. This cross-sectional study was carried out in South Indian epileptic patients after obtaining written informed consent. The study was conducted in patients with epilepsy who were on CBZ monotherapy and aged 18–65 years. Patients should have received CBZ monotherapy for a minimum of 1-month duration to achieve steady-state concentration after stabilization of autoinduction. Pregnant, lactating females, patients receiving other drugs which are substrates or inducer or inhibitor of CYP3A5, alcoholics, and patients clinically diagnosed as having hepatic dysfunction and renal dysfunction and noncompliance to CBZ therapy were excluded.

Five milliliters of venous blood sample was collected under aseptic precautions. Samples were centrifuged and the cellular component was stored at −80° for DNA extraction and genotyping, and the remaining plasma was used for estimating trough plasma CBZ levels using reverse-phase high-performance liquid chromatography method (HPLC).[4] Three such blood samples were collected, one on each occasion. The average of three readings was taken to calculate the plasma levels of CBZ. Dose corrections were carried out according to the body weight. After DNA extraction, CYP3A5* 3 genotyping (A>G, rs776746; assay ID: C 26201809_30) was carried out by real-time thermocycler (ABI Prism 7300, Foster City, USA). After genotyping, the patients were subdivided into expressers and nonexpressers.

Expressers – The expresser group comprised of patients carrying CYP3A5* 1/*1 and CYP3A5* 1/*3 genotypes (AA and AG genotypes)

Nonexpressers – The nonexpresser group comprised of patients carrying CYP3A5* 3/*3 genotypes (GG genotypes).

Statistical analysis

The frequency distribution of genotypes of CYP3A5 was checked for Hardy–Weinberg equilibrium. The normality of data was checked using Kolmogorov–Smirnov test. Clinical characteristics of epileptic patients between expressers and nonexpressers were presented as median (interquartile range) and in proportions. Dose-adjusted levels were obtained by dividing total plasma CBZ levels by dose in mg/kg. Dose- and weight-normalized plasma concentrations of CBZ (μg/ml)/(mg/kg) between expressers and nonexpressers were presented as median (interquartile range). After log transformation, the continuous variables such as plasma concentration of CBZ, dose-normalized CBZ levels, and dose- and weight-normalized CBZ levels were in normal distribution, and unpaired t-test was used to compare these variables between expressers and nonexpressers. Statistical analysis was performed using SPSS version 19.0 (Armonk, New York, IBM Corp., USA). P < 0.05 is considered statistically significant.

Results

The study was conducted in 100 epileptic patients who were on CBZ monotherapy. The observed frequencies of CYP3A5 genotypes in the study population was shown in Table 1 and they were not significantly different from the expected frequency, and they were found to be consistent with the Hardy–Weinberg equilibrium.

Table 1

Observed genotype and allelic frequency of cytochrome P450 3A5*3 (n=100) in South Indian epileptic patients

CYP3A5*3 6986 A>G	Frequency (%)
Genotype
AA	26
AG	43
GG	31
Allele
A	47.5
G	52.5

CYP3A5=Cytochrome P4503A5

The clinical characteristics such as age, gender, weight, median treatment duration, and median daily dose of CBZ were not significantly different between the two groups [Table 2].

Table 2

Comparison of patient characteristics between expressers and nonexpressers of cytochrome P450 3A5 of South Indian epileptic patients

Patient characteristics	Expressers of CYP3A5 (n=69)	Nonexpressers of CYP3A5 (n=31)	Total (n=100)
Age (years)	25 (14)	26 (15)	25.5 (14.5)
Sex, n (%)
Male$	31 (44.9)	17 (54.8)	48 (48)
Female$	38 (55.1)	14 (45.2)	52 (52)
Weight (kg)	55 (17.5)	56 (23)	55 (18)
Dose of carbamazepine (mg/day)	600 (450)	600 (400)	600 (400)
Duration of treatment (months)	48 (90)	60 (84)	72.8 (91.5)

Data are expressed in median (IQR). $Data are expressed as the number of patients (%). IQR=Interquartile range, CYP3A5=Cytochrome P450 3A5

Continuous variables such as plasma concentration of CBZ (μg/ml), dose-normalized CBZ levels (μg/ml)/mg, and dose- and weight-normalized CBZ levels (μg/ml)/(mg/kg) were found to be significantly different between expressers and nonexpressers [Tables 3 and 4].

Table 3

Comparison of carbamazepine levels between expressers and nonexpressers of cytochrome P450 3A5 (n=100) in South Indian epileptic patients

Concentration of CBZ	Expressers of CYP3A5 (n=69)	Nonexpressers of CYP3A5 (n=31)	P
Plasma concentration of CBZ (µg/ml)	4.3 (1.7)	4.86 (3.6)	0.014*
Dose-normalized CBZ levels (µg/ml)/mg	0.007 (0.004)	0.01 (0.005)	0.002*
Dose- and weight-normalized CBZ levels (µg/ml)/(mg/kg)	0.37 (0.26)	0.59 (0.28)	0.001*

Data are expressed as median (IQR). *Mann-Whitney U-test. CBZ=Carbamazepine, IQR=Interquartile range, CYP3A5=Cytochrome P450 3A5

Table 4

Comparison of log-transformed carbamazepine levels between expressers and nonexpressers of cytochrome P450 3A5 (n=100) in South Indian epileptic patients

Log-transformed variables	Expressers of CYP3A5 (n=69)	Nonexpressers of CYP3A5 (n=31)	P*
Log plasma concentration of CBZ	0.62 (0.15)	0.72 (0.14)	0.004
Log dose-normalized CBZ levels	−2.1 (0.21)	−2.02 (0.2)	0.002
Log dose- and weight-normalized CBZ levels	−0.41 (0.23)	−0.24 (0.2)	0.001

Data are expressed as mean (SD). *Unpaired t-test. CBZ=Carbamazepine, IQR=Interquartile range, SD=Standard deviation, CYP3A5=Cytochrome P450 3A5

There were 22 adverse drug reactions in total. The proportion of patients with adverse reactions in expressers and nonexpressers was 11% and 9%, respectively, and was not significant between both the groups [Table 5].

Table 5

Adverse drug reactions between expressers and nonexpressers of cytochrome P450 3A5 (n=100)

Adverse drug reactions*	Expressers of CYP3A5 (n=69)	Nonexpressers of CYP3A5 (n=31)
Sedation	4 (5.8)	2 (6.5)
Headache	2 (2.9)	5 (16.1)
Vision impairment	2 (2.8)	1 (3.2)
Nausea	2 (2.8)	0
Tremors	1 (1.4)	3 (9.7)

*Data are expressed as number of patients (%). CYP3A5=Cytochrome P450 3A5

Discussion

In our study, we found that patients inheriting CYP3A5* 3/*3 variant had an increased plasma concentration of CBZ when compared to patients inheriting CYP3A5* 1/*1 and CYP3A5* 3/*3(P = 0.004). Nonexpressers also had significantly increased plasma concentrations of CBZ when adjusted for dose and weight (P < 0.002 and P < 0.001, respectively). The power of our study was found to be 90%. Dose- and weight-normalized CBZ level in both the groups was not affected by age, gender, and duration of treatment with CBZ.

The median difference in dose-adjusted plasma concentration of CBZ between expressers and nonexpressers was 0.003 (μg/ml)/mg, and nonexpressers were found to have a 42% increased plasma concentration of CBZ when compared to expressers. This difference causes fluctuation in CBZ concentration and may predispose the epileptic patients to either toxicity or lack of efficacy due to its narrow therapeutic index, which needs to be confirmed by further studies with other genetic variants.

The explanation for the increased plasma concentration of CBZ in nonexpressers is that the CYP3A5* 3 allele has a guanine (G) nucleotide instead of adenosine (A), which creates a cryptic splice site in intron 3, leading to an alteration of mRNA splicing. This aberrant splicing causes premature termination of a codon, thereby forming a truncated nonfunctional protein. Patients who inherit CYP3A5* 3/*3 have the nonfunctional CYP3A5 enzyme.[5] Therefore, they poorly metabolize CBZ. The results of our study suggest that the presence of CYP3A5* 3 genetic polymorphism affected the metabolism of CBZ in patients with epilepsy.

Influence of genetic polymorphism on CBZ metabolism is controversial as the studies done in other populations were showing conflicting results. The results of our study are consistent with studies done in Serbian, Chinese, and Korean population. In these studies, they have found that CYP3A5* 3 polymorphism would affect the metabolism of CBZ in epileptic patients.

In the Korean population, Park et al. reported that the required dose of CBZ in the expressers was higher than that of nonexpressers though the difference was not statistically significant. However, in our study, it was found that the dose of CBZ in both expressers and nonexpressers was same.[6] Our study showed that at a given dose of CBZ, nonexpressers had an increased plasma concentration of CBZ when compared to expressers.

In the Serbian population, Milovanovic et al. reported that a lower dose of CBZ was required for patients carrying CYP3A5* 3/*3 mutant allele. Furthermore, they observed that these individuals have a higher plasma concentration of CBZ.[7] Yet, their findings were not statistically significant because of less sample size.

A study conducted by Puranik et al. in African-Americans and Caucasians reported that CYP3A5 nonexpresser genotype had a longer half-life in African-Americans when compared to expressers, but the difference was not significant in Caucasians.[8]

However, Panomvana et al. studied the effect of CYP3A5 genotypes on CBZ levels and clearance in the Thai population. This study was conducted in epileptic patients on CBZ monotherapy as well as in patients on CBZ with other antiepileptic drugs, such as phenytoin, phenobarbitone, and valproate. They found no significant difference in dose-adjusted CBZ levels between expressers and nonexpressers in epileptic patients on CBZ monotherapy. This could be due to the inadequate sample size in CBZ monotherapy group as only 36 patients were on CBZ monotherapy.[9]

Similarly, in a study conducted by Wang et al., it was found that in Chinese epileptic patients on CBZ monotherapy, there was no significant association between different CYP3A5 genotypes. However, in the same study, they reported that CYP3A5 genotypes would affect the concentration of CBZ in patients who were taking CBZ along with either phenytoin or phenobarbitone. The significant result could be due to the influence of CYP3A5 inducing drugs such as phenytoin and phenobarbitone in CBZ bitherapy.[10] Moreover, the above study was conducted in 66 patients who were on CBZ monotherapy, and the authors mentioned that the study was not adequately powered to assess the effect of CYP3A5 genetic polymorphism on dose-adjusted CBZ concentration.

The major strength of our present study is that the potential confounders which could affect the plasma concentration of CBZ such as CYP3A5 substrates, CYP3A5 inducers and inhibitors, alcohol intake, and renal and liver disorders were not included. Our study assessed the influence of CYP3A5* 3 genetic polymorphism on CBZ levels in South Indian epileptic patients who were on CBZ monotherapy with adequate power and sample size. However, our study has a few limitations. CYP3A5 enzyme is involved in the major metabolic pathway of CBZ, and it converts CBZ into CBZ10.11 epoxide. The concentration of CBZ metabolite would be a better indicator to assess the influence of CYP3A5 genotypes on CBZ metabolism; however, it was not measured in our present study. There are other variants of CYP3A5, and drug-metabolizing enzymes such as EPHX1 can also alter the metabolism of CBZ but were not analyzed. These variants need to be studied to get complete knowledge about the genetic factors affecting the concentration of CBZ.

In our study population, only 20% of the patients had adverse drug reactions with CBZ. The plasma levels of CBZ for most of the patients in our study (76.2%) were within the therapeutic range. We recommend that the future study needs to be conducted with adequate power on association between genetic polymorphisms in drug-metabolizing enzymes and adverse drug reactions of CBZ.

Conclusion

Our study revealed a significant influence of CYP3A5* 3 genetic polymorphism (6986A>G) on dose-adjusted plasma levels of CBZ in epileptic patients in the South Indian population. The individuals with the mutant genotype have an increased dose-adjusted plasma concentration of CBZ. However, further studies warranted to examine the role of other genetic variants in pharmacokinetic pathway of CBZ, including EPHX1 and ABCB1 on dose-adjusted plasma CBZ level.

Financial support and sponsorship

This study was financially supported by JIPMER intramural grant.

Conflicts of interest

There are no conflicts of interest.