Enantioselective pharmacokinetics of tramadol and its three main metabolites; impact of CYP2D6, CYP2B6, and CYP3A4 genotype

Abstract Tramadol is a complex drug, being metabolized by polymorphic enzymes and administered as a racemate with the (+)‐ and (−)‐enantiomers of the parent compound and metabolites showing different pharmacological effects. The study aimed to simultaneously determine the enantiomer concentrations of tramadol, O‐desmethyltramadol, N‐desmethyltramadol, and N,O‐didesmethyltramadol following a single dose, and elucidate if enantioselective pharmacokinetics is associated with the time following drug intake and if interindividual differences may be genetically explained. Nineteen healthy volunteers were orally administered either 50 or 100 mg tramadol, whereupon blood samples were drawn at 17 occasions. Enantiomer concentrations in whole blood were measured by LC‐MS/MS and the CYP2D6,CYP2B6 and CYP3A4 genotype were determined, using the xTAG CYP2D6 Kit, pyrosequencing and real‐time PCR, respectively. A positive correlation between the (+)/(−)‐enantiomer ratio and time following drug administration was shown for all four enantiomer pairs. The largest increase in enantiomer ratio was observed for N‐desmethyltramadol in CYP2D6 extensive and intermediate metabolizers, rising from about two to almost seven during 24 hours following drug intake. CYP2D6 poor metabolizers showed metabolic profiles markedly different from the ones of intermediate and extensive metabolizers, with large area under the concentration curves (AUCs) of the N‐desmethyltramadol enantiomers and low corresponding values of the O‐desmethyltramadol and N,O‐didesmethyltramadol enantiomers, especially of the (+)‐enantiomers. Homozygosity of CYP2B6 *5 and *6 indicated a reduced enzyme function, although further studies are required to confirm it. In conclusion, the increase in enantiomer ratios over time might possibly be used to distinguish a recent tramadol intake from a past one. It also implies that, even though (+)‐O‐desmethyltramadol is regarded the enantiomer most potent in causing adverse effects, one should not investigate the (+)/(−)‐enantiomer ratio of O‐desmethyltramadol in relation to side effects without consideration for the time that has passed since drug intake.

the (+)/(−)-enantiomer ratio of O-desmethyltramadol in relation to side effects without consideration for the time that has passed since drug intake.

K E Y W O R D S
CYP2B6, CYP2D6, CYP3A4, enantioselective pharmacokinetics, tramadol

| INTRODUCTION
Tramadol is a relatively weak opioid with a dual mechanism of action, acting both as a μ-opioid receptor agonist and a neurotransmitter reuptake inhibitor. Tramadol is administered as a racemate, with the (+)-and (−)-enantiomers of the parent compound and their metabolites showing different pharmacological effects that synergistically accomplish pain relief. 1 The (+)-enantiomer of tramadol is most potent in inhibiting serotonin reuptake, while (−)-tramadol preferentially inhibits noradrenaline reuptake. 2 The (+)-enantiomer of the metabolite O-desmethyltramadol (ODT) is, however, exerting most of the opioid effects, having the highest affinity and highest potency to the μ-opioid receptors. 3 The general interpretation of previously performed and published studies on humans is that (+)-ODT is the enantiomer most potent in relieving pain as well as in causing adverse effects. 4 The tramadol enantiomers have, however, been proposed to cause adverse effects related to abuse and overdose of the drug. 5 The CYP2D6 enzyme is accountable for the formation of ODT ( Figure 1). Individuals may be classified as poor (PMs), intermediate (IMs), extensive (normal) (EMs) or ultrarapid metabolizers (UMs) according to the metabolic activity of the CYP2D6 enzyme, determined by either phenotyping or predicted from genotyping. Several studies have consistently showed that PMs, in comparison to EMs, have an increased exposure to (+)-and (−)-tramadol combined with a reduced formation of especially (+)-ODT but also of (−)-ODT.
UMs, on the contrary, are expected to form higher amounts of (+)-ODT than EMs and to be more prone to opioid-related adverse effects. 4 Comparisons between EMs and UMs are, however, sparse in scientific literature. Apart from ODT, there are two additional primary tramadol metabolites, N-desmethyltramadol (NDT), and N,Odidesmethyltramadol (NODT) (Figure 1). NDT is pharmacologically inactive, while NODT may exert some opioid effects. 3 Formation of NDT is mediated by the CYP2B6 and CYP3A4 enzyme, whereas the metabolism route of NODT is less assured. However, all three enzymes; CYP2D6, CYP2B6, and CYP3A4 are candidates (Figure 1). 6 The CYP2B6 gene, in similarity with CYP2D6, is highly polymorphic although less studied. The significance of CYP2B6 polymorphisms in tramadol pharmacokinetics and pharmacodynamics has not yet been investigated. Nevertheless, both increased and decreased expression and activity of the enzyme as a result of genetic variation have been observed for other substrates. 7 For CYP3A4, there is substantial interindividual variation in expression and function. Many polymorphisms have been identified in the CYP3A4 gene, although most of them have not been associated with the phenotypical variability. 8 However, studies indicate that CYP3A4*22 results in reduced enzyme activity, as demonstrated in vivo with the CYP3A4 probe drugs midazolam and erythromycin, 9 and is affecting the pharmacokinetics and pharmacodynamics of several drugs. [10][11][12] To better understand the pharmacological and adverse effects of tramadol, its metabolism and the influence of genetic polymorphisms, the use of enantioselective analytical methods and genotyping are of importance in studies of tramadol. Only two of the tramadol sterioisomers are administered as the racemic drug; the enantiomeric pair of 1R,2R (+) and 1S,2S (−). 13 Tramadol and its three main metabolites thus give rise to four pairs of enantiomers in the blood following drug intake. To our knowledge, all eight compounds have not been simultaneously determined in a study population earlier. The overall aim of the present study was therefore to obtain such metabolic profiles in healthy volunteers receiving a single, therapeutic dose of tramadol. The enantioselective method needed for conducting such a study has previously been developed and validated. 14 Specifically, the following research questions and hypotheses were posted: 1 Can interindividual differences in enantioselective metabolic profiles be explained by CYP2D6, CYP2B6, and/or CYP3A4 genotype?
The hypothesis is that interindividual differences are better explained by the combined genotype of all three enzymes involved in the metabolism of the drug, rather than by CYP2D6 itself.
2 Is there a correlation between (+)/(−)-enantiomer ratios of tramadol or its metabolites and time following drug administration?
To our knowledge, this has not been investigated in a study population previously, although there are indicators of such a relation in literature. Rudaz et al. showed that enantiomer ratios of all four compounds in urine changed over time in one individual administered 100 mg tramadol. 15

| Study population
The present study was based on analysis of blood samples collected in conjunction to a previous survey. 16 16 and chiral analysis of tramadol and its metabolites, respectively.
The participants were not allowed to undergo any drug treatment, with the exception of contraceptive medication, which was used by seven subjects. However, one participant (subject 04), reported intake of an antihistamine about 12.5 hours before the tramadol administration. In conjunction to the last blood sampling on the first day (10 hours), the subjects were requested to fill in a form regarding their experience of drug-related symptoms (DRS). The participants were asked to grade their experiences of nausea, dizziness, headache, vomiting, dry mouth, sweating, fatigue, and any other DRS on a scale between zero to five, where zero was no symptoms at all and five was the worst imaginable symptoms. The study was approved by the Regional Ethical Review Board in Linköping (No: 2011/337-31), and written informed consent was gathered from the participants.

| Pharmacokinetic assessments
The enantioselective analysis of tramadol, ODT, NDT, and NODT in whole blood was performed at the National Board of Forensic Medicine in Linköping, Sweden, using a validated LC-MS/MS-method that has been described in detail previously. 14  (two standard deviations). Long-term stability tests at a low (1 ng/g) and a high (200 ng/g) enantiomer concentration were performed in −80°C up to 24 months. The difference in mean concentration between the stability samples and the fresh control samples at 24 months was in the range of −6% to −11% at the low concentration level and 0.5%-5% at the high concentration level. The predetermined limit of acceptance was ±25%.

| Genetic assessments
Genomic DNA was extracted using the Arrow Blood DNA 500 μL using the FDA approved xTAG CYP2D6 Kit v3 (Luminex, Austin, TX). Allele *3, *4, *5 and *6, as well as gene multiplication have been investigated also previously, using pyrosequencing technology. 16 There were no discrepancies between the two different methodologies used. There is no true consensus on how to perform the classification of certain genotypes into phenotypes. 4,17,18 In the present study, the guidelines of the Dutch Pharmacogenetics Working Group (DPWG) were used, which are also internationally recognized (www.pharmgkb.org), have been published, and which are used for dose recommendations. 19  The polymerase chain reaction (PCR) was initialized at 95°C for 5 minutes, followed by 40 cycles of denaturation at 95°C for 30 seconds, annealing at 50°C (516 G > T) or 58°C (785 A > G and 1459 C > T) for 30 seconds and extension at 72°C for 30 seconds. Final elongation was performed at 72°C for 10 minutes before the reaction was finalized at 4°C. Pyrosequencing was performed as previously described, 16 using the primer sequences and dispensation orders given in Falk et al. 20 For CYP3A4*22, Taqman analysis was used as described earlier. 21 Alleles without any of the investigated polymorphisms were designated as functional wild-type alleles (*1).

| Pharmacokinetic and statistical calculations
Pharmacokinetic and genetic assessments were successfully per-

| Metabolic profiles
The mean concentration vs time curves for the enantiomers of tramadol and its three main metabolites are shown in Figure 2. The pharmacokinetic parameters AUC, C max , t max and t 1/2 are given in Table 1, where wide ranges implicated large interindividual differences. Those differences are further depicted in Figure 3 Table 2).
Two other individuals, subject 02 and 14, showed metabolic profiles consisting of the smallest AUCs of the NDT enantiomers and, with the exception of the CYP2D6 PMs, the smallest AUCs of the NODT enantiomers (Figure 3). Both individuals were CYP2D6 EMs with the CYP2B6 genotype *6/*6 and *5/*5, respectively ( Table 2).
The differences in metabolic profiles of those two subjects compared to the other CYP2D6 EMs are further illustrated in Figure 4. Regarding both NDT and NODT, the mean AUC of the variant homozygotes was reduced by approximately 50% compared to the wild-type homozygotes and the heterozygotes. Subject 06, being a CYP2D6 IM with the CYP2B6 *6/*6 genotype did, however, not show the same marked difference in metabolic profile. ODT. The higher the concentration, the higher risk of side effects and toxicity. 4,6 However, review authors have expressed a need of more studies before a relationship between CYP2D6 metabolizer status and tramadol adverse effects might be established. 28,29 In the present study, the individual experiencing most DRS was the one The significance of CYP2B6 polymorphisms in tramadol pharmacokinetics has not been carefully investigated. In a study that elucidated potential genetic factors and covariates affecting the clearance of (+)-and (−)-tramadol in a group of neuropathic pain patients, CYP2B6 *9 was not found to significantly contribute. 31 In the present study, the metabolic profiles with the smallest AUCs of the NDT enantiomers belonged to two individuals being CYP2D6 EMs with the CYP2B6 genotype *6/*6 and *5/*5, respectively. They also showed small AUCs of the NODT enantiomers, only the PMs showed lower values. However, a gene-dose relationship, which is commonly shown for CYP enzymes, could not be demonstrated (Figure 4). Nevertheless, the difference, large or small, between wild-type homozygotes and heterozygotes may be substrate dependent. Also, regarding this particular pharmacokinetic parameter, not only CYP2B6 but also CYP3A4, is expected to influence the outcome.

| Correlation between enantiomer ratios and time following drug administration
Herein, the CYP2D6 IM with the CYP2B6 *6/*6 genotype did not show the same pronounced difference in metabolic profile, which further underlines the need of additional studies on this subject.

| Correlation between enantiomer ratios and time following drug administration
A positive correlation between the (+)/(−)-enantiomer ratio and time following drug administration was found regarding all four enantiomer pairs. If further investigated in a larger population, also taking different dosages, regular dosing and drug interactions into consideration, enantiomer ratios could potentially be used to estimate the time of tramadol intake, or with less nicety, distinguish between a recent or past administration. From a clinical and forensic perspective, that could, for instance, be helpful regarding patient adherence to medical treatment and in drugs and driving cases, respectively.
However, from the present investigation, it is obvious that there are large interindividual differences, which must also be carefully considered in further studies. Nevertheless, an advantage with the proposed method is the possibility of combining the information of four different enantiomer ratios, given that the CYP2D6 genotype is known.
Both the tramadol and ODT enantiomers have been subjects of discussion regarding adverse effects. [4][5][6] It has been clarified that the tramadol enantiomers have different pharmacological effects, although it has not been examined if there is a difference also in their ability to produce side effects. For ODT, it is the (+)-

| CONCLUSIONS
The most significant finding of the study was that (+)/(−)-enantiomer ratios of tramadol and its three main metabolites were positively correlated with the time following drug intake. If further investigated, the ratios might be used to distinguish a recent drug intake from a past one. Concerning the proposed association between concentrations of (+)-ODT and the risk of adverse effects, it is important knowing that the (+)/(−)-enantiomer ratios of ODT are affected not only by the CYP2D6 genotype, but also by the time that has passed between drug administration and blood sampling. It was confirmed that CYP2D6 PMs show a metabolic profile significantly different from the ones of IMs and EMs. Considerably larger AUCs of the NDT enantiomers were found, combined with much smaller, or noncalculable, AUCs of the ODT and NODT enantiomers, especially of the (+)-enantiomers. Homozygosity of the CYP2B6 alleles *5 and *6, not previously investigated regarding tramadol metabolism, indicated a reduced enzyme function. However, this pilot finding needs to be confirmed in a larger study population.

DISCLOSURE
The authors have nothing to disclose.