For reprint orders, please contact:
[email protected]
R ESEARCH R EPORT
ABCB1 polymorphisms influence the response to antiepileptic drugs in Japanese epilepsy patients Takayuki Seo1, Takateru Ishitsu2, Nao Ueda3, Naoyuki Nakada4, Keigo Yurube1, Kentaro Ueda1 & Kazuko Nakagawa1† †Author
for correspondence University, Division of Pharmacology and Therapeutics, Graduate School of Medical and Pharmaceutical Sciences, Oe-honmachi 5–1, Kumamoto 862–0973, Japan Tel.: +81 96 371 4545; Fax: +81 96 371 4545; E-mail: kazukon@ gpo.kumamoto-u.ac.jp 2Kumamoto Saishunso National Hospital, Kumamoto, Japan 3Kumamoto Chuo Hospital, Department of Hospital Pharmacy, Kumamoto, Japan 4Astellas Pharma Inc., Drug Metabolism Research Laboratories, Drug Discovery Research, Tokyo, Japan 1Kumamoto
Keywords: ABCB1, carbamazepine, drug resistance, epilepsy, haplotype, Japanese, p-glycoprotein, polymorphism
Objectives: The efflux transporter P-glycoprotein encoded by the ATP-binding cassette (ABC)B1 gene may play a role in drug-resistant epilepsy by limiting gastrointestinal absorption and brain access of antiepileptic drugs (AEDs). Our objective was to investigate the effect of ABCB1 polymorphisms on AED responsiveness and on the pharmacokinetics of carbamazepine (CBZ) in epileptic patients with the indication for CBZ therapy. Methods: The ABCB1 T-129C, C1236T, G2677T/A and C3435T polymorphisms were genotyped in 210 Japanese epileptics who had been prescribed AEDs, including CBZ, for longer than 2 years. Haplotype and diplotype frequencies were estimated by expectation–maximization algorithm. Drug resistance was determined by the presence of seizures. Association of the polymorphisms with the risk of drug resistance was estimated by logistic regression analysis and the odds ratios (ORs) were adjusted for the clinical factors affecting the outcome of AED therapy. CBZ concentrations to the dose (C/D) ratios were compared among the ABCB1 polymorphisms. Results: Drug-resistant patients were more likely to have the T allele (OR [95% confidence interval (CI)], 2.02 [1.14–3.58]) and the TT genotype at C3435T (OR [95% CI], 3.64 [1.16–11.39]), and the TT genotype at G2677T/A (OR vs the GG genotype [95% CI], 3.43 [1.01–11.72]). The frequency of the T-T-T haplotype at C1236T, G2677T/A and C3435T was significantly higher (OR [95% CI], 1.84 [1.03–3.30]), and the CC-GG-CC diplotype was lower (OR [95% CI], 0.09 [0.01–0.85]) in the drug-resistant patients than in the drugresponsive patients. None of the ABCB1 polymorphisms were observed to influence the C/D ratios of CBZ. Conclusion: We demonstrated that ABCB1 polymorphisms may influence the AED responsiveness without significant changes in the plasma concentrations of CBZ. Our findings were the inverse of previous results in European epileptics, thus the influence of ABCB1 polymorphisms on the AED responsiveness and/or the P-glycoprotein activity may vary among races.
Epilepsy is one of the most common chronic neurological disorders. Although the prognosis for the majority of patients is good, up to 30% of patients continue to have seizures despite carefully optimized antiepileptic drug (AED) treatment [1,2]. Drug-resistant epilepsy affects individual health and the quality of life, and it also imposes a heavy burden on society [1,2]. Two principal theories, alterations in drug penetration to the brain and alterations in drug targets, have been suggested to play a role in therapeutic failure in this population [2,3]. P-glycoprotein (P-gp) is a product of the ATP-binding cassette subfamily B member 1 transporter (ABCB1, also known as multidrug resistant 1 [MDR1]) gene [4–7], and its role in the etiology of drug-resistant epilepsy has been partially elucidated [2,8,9]. An overexpression of P-gp was observed in surgically excised human epileptic brain tissues and in the region of experimentally induced seizure foci [10–13]. P-gp has a wide substrate specificity for lipophilic molecules,
10.2217/14622416.7.4.551 © 2006 Future Medicine Ltd ISSN 1462-2416
including the majority of AEDs, and it excretes them out of cells as a transmembrane efflux pump [2–4,8,9]. Since it is expressed predominantly on the apical surface of intestinal epithelial cells and brain capillary endothelial cells, P-gp limits both the oral bioavailability and the brain access of substrates, thereby regulating their pharmacological effects [4,5]. P-gp expression has been demonstrated, at least in part, to be influenced by ABCB1 polymorphisms [4,7,14]. Previous studies have demonstrated a significant association between C3435T polymorphism and the functionality of P-gp in volunteers and/or patients [4,7,14]. The expression and activity of P-gp increased in the duodenum of Caucasian volunteers with a CC genotype, and their peak steady-state plasma concentrations of digoxin were lower than those with a TT genotype [14]. The T-129C, C1236T and G2677T/A polymorphisms were also associated with variations in the activities of P-gp and/or in the expression of P-gp or ABCB1 Pharmacogenomics (2006) 7(4), 551–561
551
RESEARCH REPORT – Seo, Ishitsu, Ueda et al.
mRNA [4,7]. If this genetic influence is extended to cerebrovascular P-gp, then the brain penetration of substrates, such as AEDs, might be dependent on the individual genotype [8,9]. Recently, Siddiqui and colleagues [15] reported that patients with drug-resistant epilepsy were more likely to have the CC genotype of C3435T than the TT genotype. Furthermore, as for temporal lobe epilepsy, the ABCB1 haplotype of C-GC at C1236T, G2677T/A and C3435T was found to be significantly associated with the poor response to AED treatment [16]. However, subsequent C3435T genotype–phenotype studies have failed to replicate the initial findings [17,18], and the haplotype–phenotype studies remain controversial [16–20]. These conflicting results could be due to the heterogeneity of epilepsy, incorporating numerous epilepsy syndromes with different etiologies, and the multifactorial nature of drug responsiveness in humans [1–3,21,22], as well as the diversity of methodology of association studies [23] and/or the races of the subjects [4,6,7]. Based on these facts, we investigated the effects of ABCB1 polymorphisms (T-129C, C1236T, G2677T/A and C3435T) on the responsiveness to the AED therapy among patients who had a history of carbamazepine (CBZ) therapy, considering the clinical factors possibly affecting the outcome and the pharmacokinetics of CBZ. Methods Subjects
The study included 210 Japanese epileptics (119 males; mean age, 17.4 years; range, 1.8–51.3), who had been prescribed AEDs for longer than 2 years and had a history of CBZ therapy at Kumamoto Saishunso National Hospital (Kumamoto, Japan) after January 1996. A total of 42 patients had complications of severe or profound mental retardation with significant behavior impairment, F72.1 or F73.1 of the International Classification of Diseases (ICD)-10 criteria [24]. All patients and/or their parents/guardians gave their written informed consent to participate in the study. The protocol was approved by the ethics committees of Kumamoto Saishunso National Hospital and the Graduate School of Medical and Pharmaceutical Sciences, Kumamoto University (Kumamoto). For all patients, the appropriate AED was chosen according to the treatment guideline of Societas Neurologica Japonica and literature reviews [1,25,26], based on the diagnoses made by medical history, physical and neurological examinations, electroencephalography, computed 552
tomography, magnetic resonance imaging, and extra examinations in cases where required. The drug dosage was adjusted to achieve a therapeutic plasma level as clinical circumstances dictated, with particular attention being paid to efficacy and tolerability. The patients were treated with a single drug whenever possible. The treatment was changed to another drug if the seizures remained uncontrolled, if drug-precipitated seizures were suspected or if the patient had any intolerable adverse drug reaction(s). A combination of AEDs was used in patients whose epilepsy remained uncontrolled, despite treatments with a single AED. At each follow-up visit, clinical information and the response to AED therapy were recorded. Compliance was monitored at the clinic by therapeutic drug monitoring (TDM), since poor compliance is a common cause of treatment failure in patients with epilepsy. Any patients who persistently did not comply with the treatment regimen were excluded from the study. Definitions
The types of seizures and epileptic syndromes were classified according to the guidelines of the International League against Epilepsy [25,26]. The seizures were classified as generalized seizures or as partial seizures. The epilepsy was classified as idiopathic, symptomatic, or cryptogenic, according to the putative cause and depending on factors such as the age of the patient, the type of seizure, the presence or absence of a family history of epilepsy, and the presence or absence of an underlying neurologic lesion. According to the definition by Kwan and colleagues [26], the patients were considered to be free of seizures (drug responsive) if they had not experienced seizures of any type for a minimum of 1 year while receiving the same dose of antiepileptic drug. Patients who had seizures were, by definition, considered to be drug resistant. The extent of control of seizures was assessed at the time of the patient’s last clinic visit. The CBZ concentrations were assessed by the routine TDM data. The patients examined for the concentration of CBZ fulfilled all of the following conditions: taking immediate-release CBZ for at least 2 weeks, having normal renal and hepatic functions, and were able to provide detailed medical data. The CBZ concentrations were measured using a fluorescence polarization immunoassay (FPIA; TDx, Abbott Japan Co., Ltd, Osaka, Japan). The data for TDM with suspected temporary noncompliance were excluded from the analysis. Pharmacogenomics (2006) 7(4)
ABCB1 polymorphisms and response to AEDs in Japanese epilepsy patients – RESEARCH REPORT
Genotyping & haplotype assignment
Genomic DNA was prepared from whole blood using the DNA Extractor WB kit (Wako Pure Chemical Industries, Ltd, Osaka, Japan) and/or from buccal cells using a protocol modified from Richards and colleagues [27]. The ABCB1 T-129C, C1236T, G2677T, G2677A and C3435T genotypes were essentially identified by polymerase chain reaction (PCR)-restriction fragment length polymorphism (RFLP) analyses as described previously [28–30]. The details regarding the primers, restriction enzymes, and fragment size are provided in Table 1. Pairwise linkage disequilibrium and haplotypes for ABCB1 polymorphisms at C1236T, G2677T/A and C3435T were evaluated using the SNPAlyze 3.1 program (DYNACOM Co. Ltd, Chiba, Japan), using a method previously described [31]. Haplotype frequencies were estimated by the expectation–maximization algorithm. Statistical analysis
The allele and genotype frequencies of the ABCB1 polymorphisms were assessed for any deviation from the Hardy-Weinberg equilibrium using Fisher’s exact test. Fisher’s exact test (for comparisons of the categorical data), Student’s ttest or analysis of variance (ANOVA) (for comparisons of continuous data), and the MannWhitney test or Kruskal-Wallis test (for comparisons of nonparametric continuous data) were used to compare the distributions of demographic characteristics between drug-resistant epilepsy and drug-responsive epilepsy or among each allele, genotype, haplotype and diplotype. The strength of the associations between responsiveness to the AED treatment and ABCB1 polymorphisms were measured as odds ratios (ORs). The ORs were obtained with logistic regression analysis. The ORs were adjusted for gender, age, age at the onset of epilepsy, complication, history of AED therapy and seizure type and etiology of epilepsy. The steady-state CBZ concentration to the dose (C/D) ratios was compared among each allele, genotype, haplotype and diplotype using Student’s t-test or ANOVA. A p-value of less than 0.05 was considered to be statistically significant. Multiple comparisons were corrected using Bonferroni’s method. Statistical analyses were performed using the SPSS software package (version 12.0; SPSS Inc., IL, USA). Results Characteristics of the patients
A total of 126 (60.0%) patients were classified as drug-resistant epilepsy and 84 (40.0%) patients as www.futuremedicine.com
drug-responsive epilepsy. The demographic characteristics of the groups are shown in Table 2. Age at the onset of epilepsy was significantly lower in the drug-resistant group than the drug-responsive group (3.3 ± 3.9 vs 6.6 ± 4.8 years; OR per year increase of age at the onset [95% confidence interval (CI)], 0.84 [0.78–0.90]; p < 0.001). The frequency of the F72.1 or F73.1 complication was higher in the drug-resistant group than in the drug-responsive group (27.0 vs 9.5%; OR [95% CI], 3.51 [1.53–8.04]; p = 0.003). The prevalence of drug-resistant epilepsy was higher in patients with generalized seizures than in those with simple partial seizures (OR [95% CI], 2.94 [1.55–5.58]; p = 0.001). Complex partial seizures were not proven to be a risk factor of drug resistance, and a confounder was observed between the seizure type and the basic complications (p < 0.001). The patients with symptomatic or cryptogenic epilepsy were more prone to the drug-resistant epilepsy than those with idiopathic epilepsy (OR [95% CI], 12.9 [5.04–33.1] and 7.48 [2.92–19.2], respectively; each p < 0.001). The number of AEDs was significantly higher in drug-resistant epilepsy, while the doses and concentrations of CBZ were also higher in drugresistant patients who were prescribed CBZ at the date of the last follow-up visit. Relationship between the alleles & genotypes of ABCB1 polymorphisms & AED responsiveness Table 3 shows the allele frequencies of the ABCB1 polymorphisms. The allele frequencies of the -129C, 1236T, 2677T, 2677A and 3435T variants in total patients were 6.7, 66.4, 46.9, 12.9 and 44.8%, respectively. The distribution of the -129C, 1236T and 2677A alleles was similar between drug-resistant and drug-responsive groups. However, the drug-resistant patients were significantly more likely to have the T allele than the C allele at C3435T when compared with the drug-responsive patients. The OR adjusted for gender, age, age at the onset of epilepsy, complications, history of AEDs, and classification of epilepsies and epileptic syndromes was 2.02 (adjusted OR [95% CI], 1.14–3.58; p = 0.017). In addition, the frequencies of the T allele at G2677T/A was higher in the drugresistant than in the drug-responsive group, although there was no statistically significant influence (adjusted OR [95% CI], 1.76 [0.97–3.20]; p = 0.06). The genotype frequencies of each polymorphism are shown in Table 4. The observed genotype frequency distributions were consistent
553
RESEARCH REPORT – Seo, Ishitsu, Ueda et al.
Table 1. Primer sequences used for the amplification of PCR fragments, which contained distinct ABCB1 polymorphisms and restriction endonucleases. SNP
Primer sequence
Restriction enzyme
T-129C
Forward:
5'-AGT CAT CTG TGG TGA GGC TG-3'
MspA1I
Reverse:
5'-AAC GGC CAC CAA GAC GTG A-3'
Forward:
5'-TTG AAT GAA GAG TTT CTG ATG TTT TC-3'
Reverse:
5'-CCT GAC TCA CCA CAC CAA TG-3'
Forward:
5'-TGC AGG CTA TAG GTT CCA GG-3'
Reverse:
5'-TTT AGT TTG ACT CAC CTT CCC G-3'
Forward:
5'-TGC AGG CTA TAG GTT CCA GG-3'
Reverse:
5'-TTT AGT TTG ACT CAC CTT CCC G-3'
Forward:
5'-TGA TGG CAA AGA AAT AAA GCG A-3'
Reverse:
5'-TGA CTC GAT GAA GGC ATG TAT GT-3'
C1236T G2677T G2677A C3435T
Fragment length (bp)* 74, 141 32, 74, 109
HaeIII
35, 147, 269 182, 269
BanI
26, 198 224
RsaI
106, 118 24, 82, 118
MboI
49, 144 193
*Upper
line, fragments of wild-type alleles; lower line, fragments of variant alleles. ABC: Adenosine triphosphate-binding cassette; bp: Base pair; PCR: Polymerase chain reaction; SNP: Single nucleotide polymorphism.
with the Hardy–Weinberg equilibrium (each p > 0.25). The distribution of the T-129C and C1236T genotypes did not differ significantly. The drug-resistant patients were more likely to have the TT genotypes than the GG genotypes at G2677T/A in comparison with the drugresponsive patients (adjusted OR vs GG genotype [95% CI], 3.43 [1.01–11.72]; p = 0.049). The TT and CT genotypes at C3435T were overrepresented among patients with drug-resistant epilepsy as compared with drug-responsive epilepsy (adjusted OR [95% CI], 3.64 [1.16–11.39] and 2.63 [1.02–6.81]; p = 0.027 and p = 0.046, respectively]. However, when we took multiplicity into consideration, there were no significant differences between ABCB1 polymorphisms and AED responsiveness, since all p-values were higher than 0.0125. Relationship between the haplotypes of ABCB1 polymorphisms & AED responsiveness
Significant linkage disequilibrium was detected among C1236T, G2677T/A and C3435T (each |D’| > 0.6; p < 0.001). Of the 12 possible haplotypes only five, T-T-T, T-G-C, C-G-C, C-A-C and T-T-C, were encountered with frequencies of above 5%. The frequencies of haplotypes and diplotypes are shown in Tables 5 and 6. The frequencies of T-T-T haplotypes were significantly higher for the drug-resistant patients than for the drug-responsive patients (adjusted OR [95% CI], 1.84 [1.03–3.30]; p = 0.041) (Table 5). The CC-GG-CC and TC-GGCC diplotypes were significantly infrequent in the drug-resistant patients when compared with 554
drug-responsive patients (adjusted OR [95% CI], 0.09 [0.01–0.85] and 0.22 [0.05–0.96]; p = 0.036 and p = 0.043, respectively) (Table 6). The significance of these results was not shown after Bonferroni’s correction. Relationship between the C/D ratios of CBZ
The investigation of the association between C/D ratios of steady-state CBZ and ABCB1 polymorphisms included 174 patients. Among them, 97 (55.7%) patients were male and the mean age was 15.4 ± 8.9 years (range 0.9–49.3 years). Mean body weight was 39.5 ± 17.9 kg (range 7.5–75.0 kg), the blood sampling time after dosing was 3.9 ± 2.1 hours (range 0.3–9.3 hours), the mean CBZ daily dose was 10.5 ± 5.3 mg/kg/day (range 3.2–39.5 mg/kg/day), and the mean CBZ concentration was 6.2 ± 2.4 µg/ml (range 2.2–14.0 µg/ml). Valproic acid, phenytoin and/or phenobarbital were coadministered in 38, 12 and/or 15 patients, respectively. The C/D ratio of GG genotype vs TT genotype at G2677T/A was 0.68 ± 0.31 vs 0.72 ± 0.31, and that of CC, CT and TT genotypes at C3435T were 0.72 ± 0.30, 0.65 ± 0.31 and 0.64 ± 0.27, respectively. Furthermore, the CC-GG-CC and TC-GG-CC diplotypes tended to have a higher C/D ratio in comparison with other diplotypes (0.78 ± 0.32, 0.82 ± 0.30 and 0.66 ± 0.30, respectively). We did not detect any significant association between CBZ C/D ratio and ABCB1 polymorphisms. There were no confounding factors that might influence the CBZ pharmacokinetics among each group. Pharmacogenomics (2006) 7(4)
ABCB1 polymorphisms and response to AEDs in Japanese epilepsy patients – RESEARCH REPORT
Table 2. Demographic characteristics of the patients. Demographic characteristics
Drug-resistant epilepsy (n = 126)
Drug-responsive epilepsy (n = 84)
OR* (95% CI)
p-value
Male
72 (57.1)
47 (56.0)
Referent
Female
54 (42.9)
37 (44.0)
0.95 [0.55–1.66]
0.87
18.0 ± 9.6
16.5 ± 9.5
1.02 [0.99–1.05]
0.26
Gender
Age (years) Body weight (kg)
44.0 ± 22.0
43.9 ± 19.6
1.00 [0.99–1.01]
0.96
Age at the onset or epilepsy (years)
3.3 ± 3.9
6.6 ± 4.8
0.84 [0.78–0.90]
