Polymorphisms in CYP2D6 have a Greater Effect on Variability of Risperidone Pharmacokinetics than Gender

Transcription

1 Basic & Clinical Pharmacology & Toxicology, 2015, 116, Doi: /bcpt Polymorphisms in CYP2D6 have a Greater Effect on Variability of Risperidone Pharmacokinetics than Gender Teresa Cabaleiro 1, Dolores Ochoa 1, Manuel Roman 1, Isabel Moreno 1, Rosario Lopez-Rodrıguez 2,3, Jesus Novalbos 1 and Francisco Abad-Santos 1,3 1 Service of Clinical Pharmacology, Hospital Universitario de la Princesa, Instituto Teofilo Hernando, Instituto de Investigacion Sanitaria Princesa (IP), Madrid, Spain, 2 Liver Unit, Gastroenterology Service, Hospital Universitario de la Princesa, Instituto de Investigacion Sanitaria Princesa (IP), Madrid, Spain and 3 Centro de Investigacion Biomedica en Red de Enfermedades Hepaticas y Digestivas (CIBERehd), Instituto de Salud Carlos III, Madrid, Spain (Received 22 April 2014; Accepted 3 June 2014) Abstract: Within-subject coefficient of variation (CV w ) plays a decisive role in the determination of sample size in bioequivalence clinical trials. Highly variable drugs may require the participation of a large number of subjects. The aim of this study was to investigate whether gender and polymorphisms in CYP2D6 affect the CV w of risperidone. Two single-dose, two-period crossover studies of risperidone (n = 70) were reanalysed to calculate CV w for AUC t and C max. Subjects were classified into four different CYP2D6 phenotype groups [poor metabolizers (PM), intermediate metabolizers (IM), extensive metabolizers (EM) and ultrarapid metabolizers (UM)]. The effect of gender was evaluated in EM and IM. CV w was lower in PM (13.3% for AUC t and 10.9% for C max ) and UM (17.4% and 8.7%) than in EM (28.7% and 34.7%) and IM (33.2% and 27.3%). Variability was slightly lower in women (27.9% for AUC t and 25.7% for C max ) than in men (33.3% and 37.2%, respectively). Genetic polymorphisms affect within-subject variability more than gender and could considerably affect sample size calculation. Therefore, subjects participating in bioequivalence trials should be genotyped. Bioequivalence studies are conducted to determine whether a new formulation of a drug has the same rate and extent of absorption as its corresponding reference drug. The maximum plasma concentration (C max ) is used as an index of the rate of absorption, and the area under the curve (AUC) is used as an index of the extent of absorption. A test product and its corresponding reference product are bioequivalent if the 90% confidence intervals (CI) of the geometric mean of the test/ reference ratios for both pharmacokinetic parameters fall within the bioequivalence limits of 80% and 125% [1]. The number of subjects to be included in the study should be based on an appropriate sample size calculation to ensure adequate power within the bioequivalence limits for AUC and C max. The size of the sample in bioequivalence trials is affected by the within-subject variability (coefficient of variation [CV w ]) of these parameters [2 4], that is, the higher the CV w, the greater the number of volunteers that should be included and, consequently, the higher the costs of the trial. CV w depends on the type of drug, as it increases with highly variable drugs [3 5], which are generally defined as drugs for which the CV w for C max, AUC or both is 30% [1,6 9]. Therefore, the bioequivalence of these formulations is limited, because the high variability means that large numbers of subjects are required to ensure adequate statistical power to meet FDA or EMA bioequivalence limits [6,7,10]. Author for correspondence: Francisco Abad-Santos, Clinical Pharmacology Service, Hospital Universitario de la Princesa, Diego de Leon 62, Madrid, Spain (fax , francisco. abad@salud.madrid.org). TC and DO contributed equally to the manuscript. Pharmacokinetic variability can be associated with gender and genetic polymorphisms, both of which affect clearance [11]. Drug clearance must remain stable throughout study periods so that the bioavailable dose fraction for both the reference product and the test product can be accurately compared. Consequently, subjects whose clearance varies significantly could show differences in plasma concentrations between products that are not completely related to differences in bioavailability but to variations in their own metabolism [12]. About 25% of the most frequently prescribed drugs are metabolized by CYP2D6. The CYP2D6 gene is highly polymorphic, and its alleles are associated with an absence of catalytic activity [13]. Approximately 5 10% of Caucasians carry only inactive CYP2D6 copies and therefore have genetically determined deficient metabolic capacity (poor metabolizers, PM), with the most common inactive alleles being CYP2D6*3, *4, *5 and *6 [14 16]. Moreover, approximately 5 10% of the Spanish population has an ultrarapid metabolism (ultrarapid metabolizers, UM), because they have more than two active alleles [14 18]. One of the many drugs metabolized by CYP2D6 is risperidone, an atypical antipsychotic agent prescribed to treat major psychiatric diseases such as schizophrenia and bipolar disorder. Risperidone is rapidly and completely absorbed, with peak plasma concentrations occurring 1 hr after oral administration, and <1% is excreted unchanged in faeces [19]. Risperidone undergoes extensive hepatic metabolism predominantly via CYP2D6 to an active metabolite, 9-hydroxyrisperidone [20], whose pharmacological activity is

2 PHARMACOGENETICS AND VARIABILITY OF RISPERIDONE PHARMACOKINETICS 125 approximately equal to that of risperidone [19]. Steady-state plasma concentrations of risperidone exhibit large interpatient differences at therapeutic dosages [21]. Biotransformation of polymorphisms in CYP2D6 leads to an effect on the pharmacokinetic parameters of risperidone, with an increase in the AUC for PM compared with that of extensive metabolizers (subjects with two active alleles, EM) [22,23]. In addition, these pharmacokinetic studies have documented that CYP2D6 genotype has a major effect on the pharmacokinetics of risperidone, but not on the total active moiety [22,23]. Drugs with active CYP2D6 metabolism, such as risperidone, could act as probes to assess the potential effect of genotypes on CV w. Therefore, the main objective of this study was to investigate whether gender and polymorphisms in CYP2D6 affect the CV w of risperidone. Genotyping. CYP2D6 was genotyped in all 70 subjects. Genomic DNA was extracted from 4 ml of peripheral blood using an automatic DNA extractor (MagNa Pure LC; Roche; Indianapolis, Indiana, USA). The DNA concentration was quantified spectrophotometrically using a NanoDrop â ND-1000 Spectrophotometer (Wilmington, DE, USA). Samples were analysed using PharmaChip [24], a commercially available tool whose suitability in pharmacogenetic genotyping has been previously described [25,26]. According to standard procedures, the CYP2D6*1 allele was defined as the one lacking the mutations analysed in the CYP2D6 gene. Predicted phenotypes were determined following the activity score (AS) system published by Gaedigk et al. [27], who assigned points as follows: 0 points for the inactivating alleles *3, *4, *5, *6 and *7; 0.5 points for alleles *9, *10 and *41; 1 point for the wildtype alleles *1, *2 and *35; and 2 points for the duplication of *1 or *2. The AS value was reclassified in four phenotypic groups, as detailed in table 1. An AS higher than two indicated UM, a value of Materials and Methods Study design and subjects. Two risperidone bioequivalence studies were carried out at Hospital Universitario de la Princesa (Madrid, Spain) in Each trial was a randomized, two-period, crossover study [22]. They were open for investigators but blind for the analyst responsible for plasma risperidone and 9-hydroxyrisperidone concentration measurement. Risperidone was administered as singledose 1-mg tablets under fasting conditions (10 hr fasted), each risperidone formulation separated by a 14-day washout period. Risperdal (Janssen-Cilag, Madrid, Spain) was the reference formulation in both studies. As both test formulations proved to be bioequivalent with the reference one, data were analysed together. The study sample comprised 70 healthy volunteers (35 males and 35 females) enrolled from both clinical trials. Trials were conducted under the principles of good clinical practice and following approval by the Ethics Committee of Hospital Universitario de la Princesa (Madrid, Spain) and authorization by the local regulatory agency (Spanish Agency for Medicines and Health Products). All the subjects gave their written informed consent prior to enrolment. Pharmacokinetic analysis. Blood samples were collected in 10-mL sterile EDTA-K3 tubes before dosing and at different times after dosing. Plasma concentrations of risperidone and 9-hydroxyrisperidone were measured using high-performance liquid chromatography with tandem mass spectroscopy, as previously described [22]. The limit of quantification was 0.1 ng/ml for risperidone and 9-hydroxyrisperidone, and the values of the intra-assay and interassay CV w were <5% at all concentrations ( ng/ml) of calibration curves for both compounds [22]. Pharmacokinetic parameters for risperidone were calculated using non-compartmental methods. In this study, we only analysed the two main parameters for evaluating bioequivalence: the peak plasma concentration (C max ), which was obtained directly from the raw data, and the AUC for the time of administration to the last measured concentration (AUC t ), which was calculated using trapezoidal integration [1]. Only risperidone was considered because European guidelines state that evaluation of bioequivalence should be based upon measured concentrations of the parent compound [1]. After logarithmic transformation, AUC t and C max were analysed using ANOVA, which takes into account the terms sequence, subject within sequence, period and formulation. The value of mean square error (MSE) was obtained from the ANOVA. All calculations were carried out using WinNonlin 2.0 (Scientific Consulting, Inc., Gaithersburg, Maryland, USA). Table 1. Distribution of polymorphisms in CYP2D6 according to gender in the 70 study subjects. Genotype AS Men Women UM *1 9 n/*1 >2 1 *1 9 n/*2 >2 3 *1 9 n/*41 >2 1 *2 9 n/*2 >2 1 by gender 5 1 (%) 6 (8.57) EM *1/* *1/* *1 9 n/*4 2 1 *1/* *2/* *2 9 n/*4 2 1 *2/* *35/* *1/* *1/* *1/* *2/* *2/* by gender (%) 34 (48.57) IM *1/* *1/* *2/*3 1 1 *2/*5 1 1 *2/*6 1 1 *4/* *6/* *41/* *4/* *4/* by gender (%) 24 (34.29) PM *4/* *4/*5 0 2 *4/*6 0 1 by gender 4 2 (%) 6 (8.57) AS, activity score; PM, poor metabolizers; IM, intermediate metabolizers;

3 126 TERESA CABALEIRO ET AL indicated EM, a value of indicated intermediate metabolizers (IM), and a value of 0 indicated PM. Evaluation of C w and determination of sample size. The CV w was calculated using the following formula [1]: CV w ¼ p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðe MSE Þ 1 Where MSE is the MSE obtained from ANOVA. Although this formula does not exactly determine CV w, as might be estimated from a study with replicate administration of the same formulation, it provides the appropriate variance term for computing confidence intervals for the difference between formulation means [1,11]. The sample sizes for the bioequivalence study were estimated using the formula based on a multiplicative model with a power of 80%, a significance level of 0.05 and a bioequivalence range of [28], as follows: N 2 CV 2 w 2 ta 2 0:2 ;N 2 þ t b;n 2 where N is sample size, N 2 the degree of freedom, t the t-test value, and a and b statistical errors. To avoid the influence of CYP2D6 polymorphisms, we calculated the CV w of men and women in IM and EM. Results The study sample comprised 70 Caucasian healthy volunteers, each of whom received a single 1-mg dose of risperidone and was included in the pharmacokinetic analysis. The male group comprised 35 volunteers (mean age, years; range, years); the female group also comprised 35 volunteers (mean age, years; range, years). All the volunteers were genotyped for CYP2D6 and divided into four phenotypic groups (PM, IM, EM and UM), as shown in table 1; most of the subjects were EM and IM (82.9%). Table 2 shows the MSE and CV w for each genotype group. PM subjects showed the lowest variability. CV w for risperidone AUC t and C max was two to three times higher in IM and EM than in PM and UM. Table 2. Analysis of MSE and within-subject variability for each phenotypic group in risperidone bioequivalence studies. PM (n = 6) IM (n = 24) EM (n = 34) UM (n = 6) (n = 70) MSE CV w (%) % MSE CV w (%) % Ln, log-transformed; MSE, mean square error; CV w, within-subject coefficient of variation; PM, poor metabolizers; IM, intermediate metabolizers; 1 Per cent increment in CV w with respect to the PM group. Sample size estimation for bioequivalence studies is shown in table 3. Considering the C max, the number of subjects is reduced by 82% if only PM were included [from 34 to 6 subjects (test/reference ratio of 1.00)]. Similarly, a reduction of 85% was observed when UM were included [from 34 to 5 subjects (test/reference ratio of 1.00)]. In contrast, when only EM or IM were included, the trend was in the opposite direction (table 3). Within-subject variability was slightly lower in women (CV w 27.1% for AUC t and 25.7% for C max ) than in men (31.6% and 35.1%, respectively; table 4). If we calculate variability in EM and IM to avoid the effect of polymorphisms in Table 3. Estimation of sample size of risperidone bioequivalence studies according to phenotype group and ratio test/reference (T/R). T/R PM IM EM UM Sample size estimated with the formula reported in Methods, for test/ reference ratios of 1.00 and 1.05 (or 0.95). Ln, log-transformed; PM, poor metabolizers; IM, intermediate metabolizers; Table 4. Analysis of mean square error, within-subject variability and sample size for men and women in risperidone bioequivalence studies, considering all subjects and extensive metabolizers (EM) intermediate metabolizers (IM) subjects. All subjects Women (n = 35) Men (n = 35) (n = 70) MSE CV w (%) % N MSE CV w (%) % N EM IM subjects Women (n = 32) Men (n = 26) (n = 58) MSE CV w (%) % N MSE CV w (%) % N Ln, log-transformed; MSE, mean square error; CV w, within-subject coefficient of variation. 1 Per cent increment in CV w with respect to women. 2 N = sample size for a test/reference ratio of 1.00.

4 PHARMACOGENETICS AND VARIABILITY OF RISPERIDONE PHARMACOKINETICS 127 CYP2D6 (table 4), variability is much lower in women (27.9% for AUC t and 25.7% for C max ) than in men (33.3% for AUC t and 37.2% for C max ). In the case of risperidone, the number of subjects can be reduced by around 30% when only women were included. Discussion Within-subject variability can considerably affect sample size in bioequivalence clinical trials: the higher the CV w, the higher the number of volunteers that should be included. Genotypes of CYP3A5 and CYP2D6 were previously associated with CV w, thus showing that genotyping could help to improve the design of bioequivalence studies and therapeutic drug monitoring [28,29]. C max usually has a higher CV w than AUC [30]. Among 212 bioequivalence studies submitted to the FDA between 2003 and 2005, 33 showed a high CV w in AUC and/or C max ; in 28, only C max, but not AUC, showed a CV w higher than 30% [8]. These results agree with ours, as, when all subjects were taken into consideration, only CV w for C max exceeds the 30% limit and was higher than that of AUC. However, this is not the case in subjects with lower variability, such as females, UM and PM, in whom CV w was higher for AUC than C max. Recruiting individuals with decreased CYP2D6 activity (PM) due to lower CV w would have made it possible to reduce the sample size [28]. In addition, the CV w of the AUC and C max of tacrolimus could be associated with the CYP3A5 genotype: the CV w of these parameters and, subsequently, sample size were greater in the CYP3A5*3/*3 group than in the CYP3A5*1*1 +*1/*3 group [29]. We also observed that subjects with absent CYP2D6 activity (PM) have a lower pharmacokinetic CV w than the other phenotypic groups (table 2). PM would have reduced variability in drug clearance between periods, owing to the lack of elimination via the CYP2D6 metabolic pathway. In addition, PM could more accurately reveal differences between the bioavailabilities of drug products, given the reduced variability in their own metabolism between periods, because PM have a lower risk of interaction with other drugs, food or endogenous substances such as hormones. Gonzalez-Vacarezza et al. [28] evaluated CV w according to the presence of CYP2D6 in subjects participating in two bioequivalence clinical trials with mirtazapine. As this drug is also metabolized by CYP3A4, it is not the best example for evaluating the effect of polymorphisms in CYP2D6. Risperidone also seems to be metabolized by CYP3A4, to a lesser extent [20]. In this sense, CV w for C max in PM was higher for mirtazapine (21.4%) than for risperidone (10.9%). Therefore, our model seems to better explain sample size reduction according to CYP2D6 genotype. Surprisingly, CV w was lower in women than in men, to the extent that the number of subjects could be reduced by 30% if only women were included. Higher variability in women may be more frequent, although men and women generally have similar within-subject variability [11]. Moreover, other investigations found no significant sex-related differences in plasma concentrations of risperidone and 9-hydroxyrisperidone [31]. Therefore, enrolment of women in bioequivalence clinical trials does not increase variability, at least in the case of drugs metabolized by CYP2D6. The effect of polymorphisms on variability is much stronger than the effect of gender. One limitation of this study is the small sample size in the PM and UM groups. With a small number of subjects, it would be less likely to find a pharmacokinetic outlier. However, if an outlier appears, the influence in calculation of CV w and sample size would be very relevant. Bioequivalence studies must be conducted in homogeneous but representative samples of the general population to guarantee the external validity of the results. Such an approach is essential for the analysis of drug safety and efficacy and to ensure that results can be extrapolated to the general population. Nevertheless, as the main purpose of bioequivalence studies was to assess differences in bioavailability between drug products, the decrease in within-subject variability might improve the accuracy of pharmacokinetic parameters such as absorption, metabolism and elimination. Our results show that there are differences in CV w according to CYP2D6 genotype, which may explain variability in bioequivalence studies. Although gender also affects CV w, its effect is smaller than that of polymorphisms. Nevertheless, these factors can considerably affect sample size calculation. Therefore, genotyping should be considered in bioequivalence trials if a drug is known to be affected by major genetic polymorphisms. Acknowledgements This study was partially funded by Fondo de Investigacion Sanitaria (ref. FISEC ) and Fundacion Teofilo Hernando. This study would not have been possible without the cooperation of the volunteers. We are grateful to Thomas O Boyle for editorial assistance. References 1 European Medicines Agency. Guideline on the investigation of bioequivalence. London, United Kingdom, January 20, Doc. Ref.: CPMP/EWP/QWP/1401/98 Rev. 1/Corr. europa.eu/docs/en_gb/document_library/scientific_guideline/2010/ 01/WC pdf (last accessed on 8th march 2014). 2 Hauck WW, Parekh A, Lesko LJ, Chen ML, Williams RL. Limits of 80% 125% for AUC and 70% 143% for C max. What is the impact on bioequivalence studies? Int J Clin Pharmacol Ther 2001;39: Chow S-C, Wang H. On sample size calculations in bioequivalence trials. J Pharmacokinet Pharmacodyn 2001;28: Hauschke D. A note on sample size calculation in bioequivalence trials. J Pharmacokinet Pharmacodyn 2002;29: Ramirez E, Guerra P, Laosa O, Duque B, Tabares B, Lei SH et al. The importance of sample size, log-mean ratios, and intrasubject variability in the acceptance criteria of 108 bioequivalence studies. Eur J Clin Pharmacol 2008;64: Diliberti CE. Why bioequivalence of highly variable drugs is an issue. Advisory Committee for Pharmaceutical Sciences Meeting Transcript, April 14, /transcripts/4034T2.pdf/July 2013 (last accessed on 11th march 2014).

5 128 TERESA CABALEIRO ET AL. 7 Midha KK, Rawson MJ, Hubbard JW. The bioequivalence of highly variable drugs and drug products. Int J Clin Pharmacol Ther 2005;43: Davit BM, Conner DP, Fabian-Fritsch B, Haidar SH, Jiang X, Patel DT et al. Highly variable drugs: Observations from bioequivalence data submitted to the FDA for new generic drug applications. AAPS J 2008;10: Blume HH, Midha KK. Bio-International 92, conference on bioavailability, bioequivalence, and pharmacokinetic studies. J Pharm Sci 1993;82: Haidar SH, Davit B, Chen ML, Conner D, Lee L, Li QH et al. Bioequivalence approaches to highly variable drugs and drug products. Pharm Res 2008;25: Chen ML, Lee SC, Ng MJ, Schuirmann DJ, Lesko LJ, Williams RL. Pharmacokinetic analysis of bioequivalence trials: implications for sex-related issues in clinical pharmacology and biopharmaceutics. Clin Pharmacol Ther 2000;68: Vaan Peer A. Variability and impact on design of bioequivalence studies. Basic Clin Pharmacol Toxicol 2010;106: CYP2D6 Allele Nomenclature. htm/july 2013 (last accessed on 5th march 2014). 14 Bertilsson L, Dahl ML, Dalen P, Al-Shurbaji A. Molecular genetics of CYP2D6: clinical relevance with focus on psychotropic drugs. Br J Clin Pharmacol 2002;53: Menoyo A, del Rio E, Baiget M. Characterization of variant alleles of cytochrome CYP2D6 in a Spanish population. Cell Biochem Funct 2006;24: Stamer UM, Bayerer B, Wolf S, Hoeft A, St uber F. Rapid and reliable method for cytochrome P450 2D6 genotyping. Clin Chem 2002;48: Bernal ML, Sinues B, Johansson I, McLellan RA, Wennerholm A, Dahl ML et al. Ten percent of North Spanish individuals carry duplicated or triplicated CYP2D6 genes associated with ultrarapid metabolism of debrisoquine. Pharmacogenetics 1999;9: Gonzalez I, Perez B, Alvarez M, Dorado P, Llerena A. Study of debrisoquine hydroxylation polymorphism (CYP2D6) in the Cuban population compared to Spaniards. Med Clin (Barc) 2007; 128: Heykants J, Huang ML, Mannens G, Meuldermans W, Snoeck E, Van Beijsterveldt L et al. The pharmacokinetics of risperidone in humans: a summary. J Clin Psychiatry 1994;55(Suppl): Fang J, Bourin M, Baker GB. Metabolism of risperidone to 9-hydroxyrisperidone by human cytochromes P450 2D6 and 3A4. Naunyn Schmiedebergs Arch Pharmacol 1999;359: DeVane CL. Brief comparison of the pharmacokinetics and pharmacodynamics of the traditional and newer antipsychotic drugs. Am J Health Syst Pharm 1995;52:S Novalbos J, Lopez-Rodrıguez R, Roman M, Gallego-Sandın S, Ochoa D, Abad-Santos F. Effects of CYP2D6 genotype on the pharmacokinetics, pharmacodynamics, and safety of risperidone in healthy volunteers. J Clin Psychopharmacol 2010;30: Scordo MG, Spina E, Facciola G, Avenoso A, Johansson I, Dahl ML. Cytochrome P450 2D6 genotype and steady state plasma levels of risperidone and 9-hydroxyrisperidone. Psychopharmacology 1999;147: Tejedor D, Castillo S, Mozas P, Jimenez E, Lopez M, Tejedor MT et al. Reliable low-density DNA array based on allele-specific probes for detection of 118 mutations causing familial hypercholesterolemia. Clin Chem 2005;51: Almoguera B, Riveiro-Alvarez R, Gomez-Dominguez B, Lopez- Rodriguez R, Dorado P, Vaquero-Lorenzo C et al. Evaluating a newly developed pharmacogenetic array: screening in a Spanish population. Pharmacogenomics 2010;11: Cuyas E, Olano-Martın E, Khymenets O, Hernandez L, Jofre- Monseny L, Grandoso L et al. Errors and reproducibility of DNA array-based detection of allelic variants in ADME genes: PHAR- MAchip. Pharmacogenomics 2010;11: Gaedigk A, Simon SD, Pearce RE, Bradford LD, Kennedy MJ, Leeder JS. The CYP2D6 activity score: translating genotype information into a qualitative measure of phenotype. Clin Pharmacol Ther 2008;83: Gonzalez-Vacarezza N, Abad-Santos F, Carcas-Sansuan A, Dorado P, Pe~nas-Lledo E, Estevez-Carrizo F et al. Use of pharmacogenetics in bioequivalence studies to reduce sample size: an example with mirtazapine and CYP2D6. Pharmacogenomics J 2012;13: Yong Chung J, Jung Lee Y, Bok Jang S, Ahyoung Lim L, Soo Park M, Hwan Kim K. CYP3A5*3 genotype associated with intrasubject pharmacokinetic variation toward tacrolimus in bioequivalence study. Ther Drug Monit 2010;32: Verbeeck RK, Musuamba FT. The revised EMA guideline for the investigation of bioequivalence for immediate release oral formulations with systemic action. J Pharm Pharm Sci 2012;15: Aichhorn W, Weiss U, Marksteiner J, Kemmler G, Walch T, Zernig G et al. Influence of age and gender on risperidone plasma concentrations. J Psychopharmacol 2005;19: