PHARMACOGENETICS AND GENOMICS Tramadol as a new probe for cytochrome P450 2D6 phenotyping: A population study Background and Objective: Polymorphic cytochrome P450 (CYP) 2D6 activity has been shown to be a determinant of the pharmacokinetics and pharmacodynamics of tramadol via hepatic phase I O-demethylation of (؉)-tramadol to (؉)-O-desmethyltramadol. Our objective was to investigate whether tramadol can be used as a probe for CYP2D6 phenotyping by determining the concordance between the 8-hour tramadol and 12-hour sparteine metabolic urinary ratios. Methods: Sparteine phenotyping test was carried out in 278 healthy, white subjects. At a minimum of 2 weeks later, each subject took 50 mg tramadol hydrochloride followed by 8-hour urine collection, and a venous blood sample was drawn from 276 subjects. Urine and plasma concentrations of (؉/؊)-tramadol and ,؉/؊)-O-desmethyltramadol were determined. CYP2D6 genotyping was performed with regard to *3, *4) *6, and *9 alleles. Results: There were 28 poor metabolizers of sparteine (10.1% [confidence interval, 6.8%-14.2%]). Very low recoveries of (؉)-M1 were found in poor metabolizers (0.53% [range, 0.1%-1.1%]) compared with extensive metabolizers (8.7% [range, 1.7%-23.2%]). A bimodal distribution of the metabolic ratio of (؊)-M1/(؉)-M1 was found. The visual antimode was 2.0. This new phenotype test had only 1 misclassified subject compared with sparteine phenotyping (sensitivity and negative predictive value, 100%; specificity, 99.6%; positive predictive value, 96.6%). Of the 28 sparteine poor metabolizers, 26 were found to be genotypically poor metabolizers with regard to the inactivating mutations *3, *4, and *6. Conclusion: Fifty milligrams of tramadol is an alternative CYP2D6 phenotype probe by use of the 8-hour urinary ratio of (؊)-M1/(؉)-M1. The poor metabolizers have a metabolic ratio of 2.0 or higher. (Clin Pharmacol Ther 2005;77:458-67.) Rasmus Steen Pedersen, MSc, Per Damkier, MD, PhD, and Kim Brosen, MD, DMSc Odense, Denmark Cytochrome P450 (CYP) 2D6 phenotyping is a valu- polymorphism in 1977,1 a number of substrates have able tool in drug metabolism research. Since the been proposed and used for CYP2D6 phenotyping discovery of the debrisoquin (INN, debrisoquine) —bufuralol,2 dextromethorphan,3 metoprolol,4 and sparteine.5 Sparteine has become a classical and gold- standard CYP2D6 probe.6,7 Sparteine is metabolized From the Institute of Public Health, Faculty of Health Sciences, by CYP2D6 to the 2 oxidative metabolites 2,3- Clinical Pharmacology, University of Southern Denmark. didehydrosparteine and 5,6-didehydrosparteine.8 The Received for publication Aug 27, 2004; accepted Jan 14, 2005. Reprint requests: Rasmus Steen Pedersen, MSc, Institute of Public population is divided into 2 general phenotypes— Health, Faculty of Health Sciences, Clinical Pharmacology, Uni- extensive metabolizers (EMs) and poor metabolizers versity of Southern Denmark, Winslowparken 19, DK-5000 (PMs)—according to the activity of CYP2D6 defined Odense C, Denmark. by a metabolic ratio of the parent substrate to the E-mail: [email protected] metabolites. CYP2D6 is encoded by the functional al- 0009-9236/$30.00 Copyright © 2005 by the American Society for Clinical Pharmacology leles in EMs, and the absence of the active enzyme in 9 and Therapeutics. PMs is caused by mutations and deletions. About 7% doi:10.1016/j.clpt.2005.01.014 of the Danish white population are PMs.10 458 CLINICAL PHARMACOLOGY & THERAPEUTICS 2005;77(6):458-67 Tramadol as a new probe for CYP2D6 phenotyping 459 Fig 1. Major metabolic pathways for tramadol in vivo. Tramadol is a racemic mixture of the 2 enantiomers (Ͻ24 hours) administration (6011 patients) is quite (ϩ)-tramadol hydrochloride and (Ϫ)-tramadol hydro- similar, quantitatively and qualitatively, to that of long- chloride: (1RS,2RS)-2-[(dimethylamino)methyl]-1-(3- term administration.16 Rare cases of seizures and ana- methoxyphenyl)-cyclohexanol hydrochloride (Fig 1). phylactic reactions have been reported, with an in- Tramadol is a synthetic 4-phenyl-piperidine analog of creased risk at tramadol doses above the recommended codeine and thus an aminocyclohexanol derivative.11 It dose. A British study evaluated the risk of seizures in is a centrally acting synthetic analgesic with a wide- 10,916 patients receiving tramadol. Eleven definite spread clinical use. It has an analgesic efficacy similar cases of idiopathic seizures were identified. With ex- to weak opioids and is devoid of some of the side posure to tramadol alone, the risk of idiopathic incident effects seen with other opioids of comparable effi- seizures was found not to be increased.17 Tramadol has cacy.12,13 Tramadol produces analgesia in short-term been prescribed in Europe for close to 3 decades. More and long-term pain states by synergistically combining than 5 million patients in the United States and more weak opioid and monoaminergically mediated mecha- than 40 million patients worldwide have received tra- nisms.14,15 Safety data for tramadol have been summa- madol.18 Therefore phenotyping with 50 mg tramadol rized by Cossmann et al.16 It is a well-tolerated anal- is considered safe and can be conducted worldwide, gesic. Information from phase II to IV clinical studies because tramadol is registered in more than 100 coun- and postmarketing surveillance studies, covering safety tries. It has been licensed for children aged over 1 year data from more than 21,000 patients, were considered. in several European countries since 1977, and several The most frequent adverse events were nausea (6.1%), studies have shown that it is safe in children. Murthy et dizziness (4.6%), drowsiness (2.4%), tiredness/fatigue al19 showed that none of the pharmacokinetic parame- (2.3%), sweating (1.9%), vomiting (1.7%), and dry ters were significantly different from adults. Payne et mouth (1.6%). The profile for single-dose or short-term al20 concluded that tramadol at 3 mg/kg in children has CLINICAL PHARMACOLOGY & THERAPEUTICS 460 Pedersen, Damkier, and Brosen JUNE 2005 no clinical respiratory depressant effect and behavior ducted as an open, 2-phase, crossover trial. The mini- and recovery are unaffected. Kaabachi et al21 also con- mum washout time between the 2 phases was 2 weeks. cluded that tramadol is safe and efficient in children. Subjects. The study included 278 healthy, white vol- According to Food and Drug Administration (FDA) unteers. The 155 men and 123 women, aged 19 to 40 guidelines, tramadol should not be used in pregnant, years, were primarily Danish students at the University delivering, and nursing women because no adequate of Southern Denmark (Odense, Denmark). None of the and well-controlled studies have been conducted in volunteers had a daily intake of alcohol or drugs (ex- these populations. The FDA also states that healthy cept oral contraceptives), and none of the volunteers elderly subjects aged 65 to 75 years have plasma con- were pregnant or breastfeeding. centrations and elimination half-lives comparable to Phenotyping. In phase A of the trial, the CYP2D6 those observed in healthy subjects aged less than 65 phenotype was determined. Oral intake of 100 mg years. sparteine sulfate pentahydrate (supplied by the Central The in vivo metabolism of tramadol is complex, with Pharmacy, Odense University Hospital, Odense, Den- 23 metabolites having been identified: 11 phase I metab- mark) was followed by urine collection for 12 hours. olites and 12 phase II conjugates.22 The major metabolic The urine volume was recorded, and two 10-mL ali- pathways are O-demethylation to O-desmethyltramadol quots were stored at Ϫ20°C until analysis. The amounts (M1) and N-demethylation to N-desmethyltramadol (M2) of excreted sparteine, 2,3-didehydrosparteine, and 5,6- (Fig 1).23 The O-demethylation of tramadol is mediated didehydrosparteine were determined by gas chromatog- by CYP2D6 both in vitro and in vivo.24-27 Paar et al26 raphy.6 The design of phase B was identical to a pilot suggested that tramadol metabolism via CYP2D6 is study that separated EMs and PMs completely.28 The stereospecific. The fact that the O-demethylation of volunteers were given 50 mg tramadol hydrochloride as (ϩ)-tramadol to (ϩ)-M1 is catalyzed by CYP2D6 has a single oral dose (Nobligan; Grünenthal, Germany; been verified.28,29 Tramadol is also directly excreted in distribution in Denmark by Meda, Allerod, Denmark). the urine.28 CYP2D6 activity has also been shown to be Urine was collected for 8 hours. The urine volume was a determinant of the pharmacodynamics of tramadol. recorded, and two 10-mL aliquots were stored at Laugesen et al29 concluded that paroxetine, a very Ϫ20°C until analysis. The concentrations of (ϩ)- potent CYP2D6 inhibitor, inhibits the metabolism of tramadol, (Ϫ)-tramadol, (ϩ)-M1, and (Ϫ)-M1 in urine tramadol to M1 and reduces the hypoalgesic effect of and plasma were analyzed in duplicate by an HPLC tramadol in human experimental pain models, particu- method.28 The HPLC method for urine samples was larly in opioid-sensitive tests. validated and published with a range of linearity from Because sparteine is no longer marketed as a drug, it 0.1 to 3.0 ␮mol/L and a maximum coefficient of vari- is becoming increasingly difficult to obtain commercial ation of 5.7% for precision and reproducibility (n ϭ formulations and reference substances. Therefore there 10). Samples with higher concentrations were diluted is an interest for new validated probe drugs for in vivo with millipore water before analysis. The two 10-mL phenotyping of CYP2D6. The metabolism of total tra- aliquots of venous blood were obtained with ethyl- madol is well described, but the direct relationship enediaminetetraacetic acid as anticoagulant from each between the CYP2D6 phenotype and the stereospecific volunteer approximately 2 hours after medication dur- tramadol O-demethylation is to some extent still un- ing phase B. The blood sample was centrifuged for 10 known.