Portal de Periódicos da CAPES

Incorporation of Flurbiprofen in a 4‐Drug Cytochrome P 450 Phenotyping Cocktail

2014; Wiley; Volume: 115; Issue: 5 Linguagem: Inglês

10.1111/bcpt.12231

1742-7843

Marija Bosilkovska, Mallorie Clément, Pierre Dayer, Jules Desmeules, Youssef Daali,

Analytical Methods in Pharmaceuticals

Cytochromes P450 (CYP) constitute the major drug-metabolizing enzyme system in human beings. The activity of these enzymes is subjected to a great interindividual variability which can cause interindividual differences in plasma drug concentrations and result in therapeutic failure or side effects. To avoid these problems, it is of great importance to evaluate the in vivo CYP activity (phenotyping). In the beginning of the 1990s, a 'cocktail' approach was developed, aiming to simultaneously assess the activity of multiple CYPs by the administration of a cocktail of CYP-specific probe drugs 1. Several cocktails using different probe drugs have been developed in the past years 2-7. Few of these cocktails use well-validated probes such as caffeine, omeprazole, dextromethorphan and midazolam for the phenotyping of CYP1A2, CYP2C19, CYP2D6 and CYP3A, respectively 2, 4, 5. These drugs have no mutual interactions as previously demonstrated 5. Two of these cocktails also include losartan 4 or warfarin 2 as probes for CYP2C9 activity. These drugs, as well as tolbutamide 8 and phenytoin 9, have been proposed as potentially useful CYP2C9 probes. However, all of the proposed CYP2C9 substrates are associated with limitations which make them less than ideal as probes for this enzyme 10. Recently, flurbiprofen has been validated as a reliable probe for CYP2C9 phenotyping both in urine 10 and single point plasma samples 11. Flurbiprofen has also been incorporated in the Pittsburgh cocktail and was shown not to interact with the other probes. 7. In this study, we aimed to validate the incorporation of flurbiprofen as a CYP2C9 probe to a cocktail composed of caffeine, omeprazole, dextromethorphan and midazolam and verify the effect these drugs have on flurbiprofen metabolic ratio. After giving their informed consent, twelve healthy male volunteers were included in an open-labelled, two-period study approved by the Ethics Committee of Geneva University Hospitals. At the first session, after an overnight fast, the volunteers were given flurbiprofen alone (Froben® 50 mg, Abbott AG, Baar, Switzerland), and on the second session, following a 1-week wash-out, they were given the five-drug cocktail composed as follows: flurbiprofen (Froben® 50 mg), caffeine (Nescafé® packet, about 150 mg, Nestlé, Lausanne, Switzerland), dextromethorphan (Bexin® 25 mg, Spirig HealthCare, Egerkingen, Switzerland), omeprazole (Antramups® 40 mg, Astra Zeneca AG, Zug, Switzerland) and midazolam (Dormicum® 7.5 mg, Roche Pharma, Reinach, Switzerland). A single venous blood sample collected into EDTA tubes was obtained 2 hr after drug administration. Urine was collected for 8 hr after drug ingestion. Plasma samples were handled by liquid–liquid extraction and CYP probe drugs, and their specific metabolites were simultaneously analysed using a rapid and selective HPLC method in a single run as previously described 12. Dextromethorphan and dextrorphan were analysed in urine samples as previously described 13. CYP activities were assessed by the plasma metabolic ratios (MR) of [paraxanthine]/[caffeine] for CYP1A2, [4-hydroxyflurbiprofen]/[flurbiprofen] for CYP2C9, [omeprazole]/[5-hydroxyomeprazole] for CYP2C19, [1-hydroxymidazolam]/[midazolam] for CYP3A, and plasma and urinary ratios (UMR) of [dextromethorphan]/[dextrorphan] for CYP2D6. Genotyping for the major CYP polymorphisms was performed by real-time PCR. Genomic DNA was extracted from 200 μl whole blood using the QIAamp DNA blood mini kit (QIAGEN, Hombrechtikon, Switzerland). CYP2D6*3, *4, *6 were determined using a commercially available TaqMan® genotyping assay (Applied Biosystems®, Warrington, UK). Real-time PCR in the presence of fluorophore was performed to detect the presence of CYP2D6*5 allele (gene deletion) and gene multiplication *2xN. CYP2C9*2, CYP2C9*3, CYP2C19*2 and CYP2C19*genotyping was performed using the Fluorescence resonance energy transfer method. Fluorescence was detected by a LightCycler® instrument (Roche Diagnostics, Rotkreuz, Switzerland). Results are presented as the mean values ± S.D. for the MRs. Geometric mean ratio for flurbiprofen MR was calculated as the ratio of the geometric mean obtained after administration of flurbiprofen alone to that obtained after administration of flurbiprofen as part of the five-drug cocktail. Absence of pharmacokinetic interaction could be claimed if the 90% confidence interval (CI) for the geometric mean ratio fell completely inside the 0.8–1.25 interval. Based on this acceptance interval, an expected 15% intra-individual coefficient of variation (CV) for flurbiprofen MR and a μT/μR value of 1.0 (i.e. absence of interaction), 10 volunteers needed to be included to attain a power of 0.8 and α-value of 0.0514. Two additional volunteers were recruited in case of volunteer withdrawal. The 2-hr plasma MRs for CYP1A2, CYP2C19, CYP2D6 (without poor metabolizers) and CYP3A were 0.25 ± 0.08; 1.24 ± 1.19; 0.49 ± 0.26 and 1.06 ± 0.75, respectively. Good correlation was observed between 8-hr urinary and 2-hr plasma dextrorphan/dextromethorphan MRs (ρPearson = 0.69, p = 0.005). Two volunteers were homozygous CYP2C19*2 carriers; their MRs were higher than 5 which is consistent with poor metabolizer phenotype 15. CYP2D6 genotypes also correlated well with the phenotypes: one volunteer with genotype *4/*4 was identified as poor metabolizer (UMR > 0.3) and another one having a multiplication of CYP2D6*2xN as ultra-rapid metabolizer (UMR < 0.003) 13. The 2-hr flurbiprofen MR was 0.042 ± 0.014 and 0.040 ± 0.013 at the first and second session, respectively. The geometric mean ratio was 1.05, and the 90% CI was within the range 0.92–1.20, indicating that there was no significant difference in the MR when flurbiprofen was administered alone or as part of the cocktail. This can be also observed with the individual volunteer results (fig. 1). None of the volunteers in our study set were identified as poor CYP2C9 metabolizers as confirmed by the genotypes. The two volunteers presenting a CYP2C9*1*3 genotype had the lowest phenotyping indices which is in accordance with previous results 16. Despite the lack of poor metabolizers in this study, it has been previously shown that the use of 2-hr MR as phenotyping index unequivocally allows the differentiation between normal, inhibited and induced CYP2C9 activity 11. Moreover, in the latter study, it was also demonstrated that a very good correlation exists between the 2-hr plasma MR and flurbiprofen clearance making this metric a reliable phenotyping index. In conclusion, in this study, we have demonstrated that flurbiprofen as a CYP2C9 probe can be incorporated to a previously validated cocktail composed of caffeine, omeprazole, dextromethorphan and midazolam.