Drug Metab Drug Interact 2014; 29(3): 129–141

Review Paulo Magalhães, Gilberto Alves*, Adrián Llerena and Amílcar Falcão

Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers Abstract: Venlafaxine (VEN) is one of the safest and most effective drugs used in the treatment of selective serotonin reuptake inhibitors-resistant depression, and thereby it is nowadays one of the most commonly prescribed antidepressants. Nevertheless, patients treated with antidepressant drugs including VEN have exhibited large inter-individual variability in drug outcomes, possibly due to the influence of genetic and nongenetic factors on the drug pharmacokinetics and/or pharmacodynamics. Among them, an increased interest has emerged over the last few years on the genetic and/or phenotypic profile for drug-metabolizing cytochrome P450 isoenzymes and drug transporters such as potential predictive pharmacokinetic-based biomarkers of the variability found in drug biodisposition and antidepressant response. The integration of some of these key therapeutic biomarkers with classic therapeutic drug monitoring constitutes a promising way to individualization of VEN’s pharmacotherapy, offering to clinicians the ability to better predict and manage pharmacological treatments to maximize the drug effectiveness. Thus, this review provides an extensive discussion of the pharmacokinetics of VEN focusing *Corresponding author: Gilberto Alves, PharmD, PhD, CNC – Centre for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal; and Faculty of Health Sciences, University of Beira Interior, CICS-UBI – Health Sciences Research Centre, University of Beira Interior, Av. Infante D. Henrique 6200-506 Covilhã, Portugal, Phone: +351 275 329002, Fax: +351 275 329099, E-mail: [email protected] Paulo Magalhães: Laboratory of Pharmacology, Faculty of Pharmacy, University of Coimbra, Pólo das Ciências da Saúde, Azinhaga de Santa Comba, Coimbra, Portugal; CNC – Centre for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal; CICS-UBI – Health Sciences Research Centre, University of Beira Interior, Covilhã, Portugal; and CICAB, Clinical Research Centre, Extremadura University Hospital and Medical School, Badajoz, Spain Adrián Llerena: CICAB, Clinical Research Centre, Extremadura University Hospital and Medical School, Badajoz, Spain Amílcar Falcão: Laboratory of Pharmacology, Faculty of Pharmacy, University of Coimbra, Pólo das Ciências da Saúde, Azinhaga de Santa Comba, Coimbra, Portugal; and CNC – Centre for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal

in particular on metabolism issues, without forgetting the clinically relevant sources of pharmacokinetics variability (mainly the genetic sources) and aiming on the identification of phenotypic and/or genetic biomarkers for therapy optimization. Keywords: biomarkers; metabolism; pharmacokinetics; therapeutic drug monitoring (TDM); venlafaxine (VEN). DOI 10.1515/dmdi-2013-0053 Received October 2, 2013; accepted January 23, 2014; previously published online March 7, 2014

Venlafaxine Venlafaxine (VEN) is a hydroxycycloalkylphenylethylamine-derivative bicyclic antidepressant belonging to the pharmacological class of serotonin-noradrenaline reuptake inhibitors (SNRIs) (Figure 1, [1–5]) and it was first introduced in 1993 by Wyeth for the treatment of depression and anxiety disorders [6–11]. Presently, VEN is often used as an alternative drug in the treatment of selective serotonin reuptake inhibitors (SSRIs)-resistant depression being, therefore, one of the antidepressant agents most commonly prescribed worldwide [9, 12–15]. In truth, although several meta-analyses and reviews suggest that SNRIs and, in particular VEN, may be even more efficacious than SSRIs in the remission of depressive symptoms [16–20] and in the attainment of higher remission rates [20–22], not all studies support this conclusion [14, 23, 24]. However, VEN has been considered as a good choice for nonremitting patients [25, 26]. Moreover, VEN has been proposed to have a favorable tolerability profile compared to other antidepressant drugs (e.g., tricyclic antidepressants and tetracyclic antidepressants) [27]. These findings may be at least partially explained by the particular pharmacodynamics of VEN. Indeed, VEN is more potent as a serotonin reuptake inhibitor rather than as a noradrenaline reuptake inhibitor, and it also weakly inhibits dopamine reuptake [8, 28]; additionally, this drug has no

Brought to you by | Carleton University OCUL Authenticated Download Date | 6/29/15 11:46 AM

130      Magalhães et al.: Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers

Figure 1 Proposed metabolic pathways of venlafaxine (VEN) in humans [1–5]. The bold arrow indicates the major metabolic pathway and the main metabolizing enzyme(s) catalyzing each metabolic step is/are ­indicated above the arrows, in a higher font size. CYP, cytochrome P450; UDPGT, uridine diphosphate glucuronosyltransferase.

significant affinity for α1-adrenergic, muscarinic cholinergic, H1 histaminergic, benzodiazepine, or opioid receptors and does not inhibit monoamine oxidase [8, 29]. VEN is clinically used as a racemic mixture of two pharmacologically active enantiomers [S-(+)-VEN and R-(-)VEN] which present similar absorption and disposition properties [30]. The S-(+)-enantiomer primarily acts as a serotonin reuptake inhibitor, whereas the R-(-)-enantiomer inhibits both serotonin and noradrenaline reuptake [27, 28]. Therefore, VEN essentially acts as a serotonin reuptake inhibitor under low daily doses (75 mg/day) and the noradrenaline reuptake inhibition becomes meaningful only at higher doses (150 mg/day or higher) [8, 31, 32]. Unfortunately, despite the large armamentarium of older and newer antidepressants currently available the desired results with antidepressant therapy have not been successfully achieved [33–35]. Actually, patients treated with antidepressant drugs including VEN have exhibited large inter-individual variability in drug outcomes

(efficacy vs. adverse effects), a fact explained by genetic and nongenetic factors [8, 15, 34–39]. Thus, bearing in mind that the development of novel antidepressant agents has failed, efforts should be concentrated in establishing more effective and safer regimens taking into account patients’ individual characteristics [35, 38, 40–42]. The integrated use of classic therapeutic drug monitoring (TDM) based on drug concentrations (as a pharmacokinetic phenotyping approach) and the genotyping of predictive biomarkers for inter-individual variability in drug disposition and drug response (as a pharmacogenetic approach) constitutes a promising way to individualize the dosage regimens of drugs for which the therapeutic response is difficult to evaluate [43]. In line with this fact, it is important to characterize in detail the metabolite profiling of drugs in order to search for useful phenotypic and/or genetic biomarkers predictive of the inter-individual variability in pharmacokinetics and clinical response. The identification of drug exposure-effect relationships

Brought to you by | Carleton University OCUL Authenticated Download Date | 6/29/15 11:46 AM

Magalhães et al.: Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers      131

determined at least in part genetically could help clinicians to better predict drug response based on patients’ pharmacogenetics testing. Therefore, in this review the pharmacokinetics of VEN is extensively addressed, focusing in particular on the drug metabolic profile. The available data on the clinical impact factors that can affect the pharmacokinetics of VEN are also herein critically discussed in order to identify potential phenotypic and/or genetic biomarkers, which may provide a rational basis for drug therapy optimization.

Pharmacokinetic properties VEN is rapidly and extensively absorbed from the gastrointestinal tract after oral administration to humans. Indeed, mass balance studies support that at least 92% of VEN is absorbed following a single oral dose. Nevertheless, VEN has an absolute oral bioavailability of only 40%–45% due to its extensive first-pass metabolism to the major metabolite O-desmethylvenlafaxine (ODV; Figure 1), also known as desvenlafaxine, which is currently available in some countries as an antidepressant drug [1, 7, 9, 44]. Despite the plasma concentrations of ODV being higher (two- to three-fold) than those of the parent drug in most individuals, the therapeutic consequences of the extensive presystemic metabolism are not expected to be meaningful because the metabolite ODV is substantially equivalent to the parent drug in terms of pharmacological activity and potency [29]. Besides ODV, other minor metabolites [N-desmethylvenlafaxine (NDV), N,O-didesmethylvenlafaxine (DDV) and N,N,O-tridesmethylvenlafaxine (TDV)] are also formed by secondary metabolic pathways [1, 2, 28] (Figure 1), which have been described as less-active derivatives devoid of clinical relevance [45, 46]. At this point, it is still important to refer that the dietary conditions (fasting vs. presence of food) have no significant effects on the rate and extent of absorption of VEN; the presence of food can only slightly delay the rate of absorption but does not affect the bioavailability of VEN and ODV [1, 47]. VEN is currently marketed as two types of oral formulations: immediate-release and extended-release formulations. Both products have differences in the release profiles of VEN which, consequently, also determines differences in the systemic exposure to the drug [29]. VEN extendedrelease formulations are usually administered once-daily and the drug has a prolonged absorption profile, resulting in a lower maximum plasma concentration (Cmax) when compared with that obtained after an immediate-release formulation; even so the total absorption of VEN appears

to be equivalent independently of the type of pharmaceutical formulation used [1, 29]. Thus, a delayed absorption profile is evidenced after the use of a VEN extended-release formulation, so that the peak plasma concentrations of VEN and ODV are reached at approximately 5.5 and 9  h postdose, respectively; following treatment with a VEN immediate-release formulation, the time to reach Cmax is of about 2 h for the drug (VEN) and 3 h for the metabolite ODV. Considering the formulation-dependent pharmacokinetic profiles of VEN, it is absolutely ­comprehensible the potential benefits of the clinical use of an extendedrelease vs. an immediate-release formulation in terms of patient convenience and compliance, and also in terms of the fluctuation index. Actually, the lower peak-to-trough fluctuation in plasma VEN concentrations improves the drug’s tolerability profile and reduces the Cmax-related side effects (e.g., nausea and dizziness) [1, 29, 48–50]. Once VEN and ODV have entered into the systemic circulation they are widely distributed throughout the body. Bearing in mind the estimated values for the steady-state apparent volume of distribution of VEN (7.5 ± 3.7 L/kg) and ODV (5.7 ± 1.8 L/kg), it is evident that both compounds are well distributed beyond the total body water [9, 51]. The limited extent of binding of VEN and ODV to human plasma proteins, 27% and 30%, respectively, can also contribute for the large apparent volumes of distribution exhibited by both compounds [9, 51]; hence, the occurrence of drug interactions in plasma protein binding involving VEN is unlikely [1]. Also under this context, it is worthy of note that VEN and ODV cross the placenta and they are also extensively distributed into breast milk [51–53]; in truth, the area under the plasma concentration-time curve (AUC) was shown to be approximately three- to five-fold higher in breast milk than those in maternal plasma [51]. Another relevant issue to be considered when discussing aspects on the distribution and/or disposition of drugs is the involvement of efflux transmembrane transporters, such as the P-glycoprotein (P-gp), which may significantly determine the pharmacokinetics and pharmacodynamics of many drugs. In fact, various authors have claimed that VEN and its main metabolite (ODV) are P-gp substrates [35, 54–57]. This information is especially important seeing that P-gp is expressed in the intestinal epithelium and brain vascular endothelium, which can influence the oral bioavailability and the distribution of VEN and ODV to the brain (biophase) [35, 54, 56]. Recently, Karlsson et al. (2011) found that the brain concentrations of VEN and some of its metabolites were two- to four-fold times higher in P-gp knockout vs. wild-type mice. These data show that the expression of P-gp plays an important role in limiting brain access of VEN and its metabolites [58]. Therefore, it is likely

Brought to you by | Carleton University OCUL Authenticated Download Date | 6/29/15 11:46 AM

132      Magalhães et al.: Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers that differences in expression and functionality of the P-gp are able to explain, at least in part, the observed variations in therapeutic outcomes of patients receiving VEN, such as side effects, drug resistance, and even discrepancies between plasma levels and clinical response [35, 54, 56]. Additionally, some in vitro studies have also shown that VEN and ODV have weak or no inhibitory effects on the P-gp activity [58–60]. Nevertheless, there is no consensus about it. With regard to the induction phenomenon, studies performed in Caco-2 cells [60] and in human brain endothelial cells (blood-brain barrier model) [54] clearly demonstrated that VEN is, on the contrary to ODV, an inducer of the expression of drug efflux transporter proteins. However, to better understand this induction phenomenon mediated by VEN, and in order to assess how these in vitro findings are translate to in vivo conditions, further nonclinical and clinical researches are required. At last, with regard to the body’s excretion routes of VEN, Howell et  al. (1993) described the urinary metabolic profile for VEN radioactively labeled after its oral administration to different species, including man. These authors showed that VEN-related compounds are primarily excreted by the kidneys (92.1%), representing the excretion by feces only approximately 2% [3]. In addition, as VEN is also distributed into breast milk, the drug excretion through this route may also occur in breastfeeding women [36]. More specifically, Howell et al. (1993) demonstrated that VEN is excreted in human urine as unchanged VEN (4.7%), unconjugated ODV (29.4%), conjugated ODV (26.4%), unconjugated DDV (9.8%), conjugated DDV (6.2%), NDV (1.0%) and TDV (1.0%). Therefore, approximately 78.5% of the VEN-related compounds excreted in urine were identified, but 13.6% of the total dose excreted in urine was not identified [3]. Accordingly, VEN can display other metabolic pathways in humans beyond those shown in Figure 1. Overall, the terminal elimination halflife of VEN is approximately 5 h and that for ODV is about 11 h [36, 40], determining the time to reach the steady state, and also the time of washout [40]. Particularly, under oral multiple-dose therapy, the steady-state plasma concentrations of VEN are reached within 3 days [29, 44]. From the pharmacokinetic perspective, an important advantage of the VEN and its active metabolite (ODV) is that they exhibit a linear relationship between dose and plasma concentration within a wide range of daily doses (75–450 mg). Therefore, if dose adjustments of VEN are needed, they will not lead to unexpected disproportionate increases in plasma concentrations of the active entities [40]. Taking into consideration the aim of the present article and the eligibility of the metabolic phase as a prominent source of the inter/intra-individual variability,

even after having performed a comprehensive overview of the absorption, distribution, metabolism and excretion of VEN, we decided to discuss in more detail the metabolic pathways of this drug.

Metabolic pathways VEN undergoes extensively hepatic biotransformation catalyzed by cytochrome P450 (CYP) isoenzymes, and in vitro and in vivo studies have indicated that CYP1A2, CYP2D6, CYP2C9, CYP2C19, and CYP3A4 are the CYP isoforms involved. Specifically, the O-demethylation of VEN to ODV is the main metabolic pathway in humans, being approximately 56% of the dose metabolized through this process, which is primarily mediated by CYP2D6 [1, 8, 27, 28, 45]. Consequently, the plasma concentrations of the active metabolite ODV are usually higher (two- to threefold) than those of VEN in man. Other minor metabolic routes implicated in the oxidative metabolism of VEN and its metabolites are additional N- and O-demethylation reactions [2, 19, 29, 45]. Furthermore, there are evidences that the metabolites ODV and DDV participate in conjugation (phase II) metabolic reactions leading to the formation of the corresponding aryl O-glucuronide metabolites [3]. A schematic overview of the known metabolic pathways for VEN in humans is provided in Figure 1. Although VEN is not yet clinically available as a single enantiomer, but instead as a racemic mixture, it is consensual that O-demethylation reactions occur with strong stereoselectivity towards the R-enantiomer [61–63], and these type of reactions are predominantly mediated by CYP2D6; N-demethylation reactions appear to be at least partially mediated via CYP3A4 [1, 8, 28, 45, 63]. As expected, a recent review refers that CYP2D6 poor metabolizers (PMs) have significantly lower plasma levels of ODV and higher levels of VEN as compared to extensive metabolizers (EMs) [38]. However, in contrast to what was observed for VEN, significant variations in the metabolism of ODV were not found between these two different human CYP2D6 metabolizer phenotypes [2, 38], reinforcing the influence of other isoenzymes in reactions of N-demethylation. Unquestionably, there is evidence from several studies that CYP2C9 and CYP2C19 also participate in metabolic pathways of O- and N-demethylation [1, 4-7, 29, 63–65]. In a study using human liver microsomes and microsomes containing specific human cytochromes expressed in cDNA-transfected cells, Fogelman et  al. (1999) demonstrated that CYP2C9, CYP2C219, and CYP2D6 are isoenzymes involved in the biotransformation of VEN to ODV, with CYP2D6 being the dominant isoform responsible for this

Brought to you by | Carleton University OCUL Authenticated Download Date | 6/29/15 11:46 AM

Magalhães et al.: Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers      133

metabolic reaction [lowest Michaelis-Menten constant (Km) and highest intrinsic clearance] [5]. CYP2C9, CYP2C19, and CYP3A4 were also associated with the metabolism of VEN to NDV, having been attributed the highest intrinsic clearance to CYP2C19 and the lowest to CYP3A4 [5]. However, despite the lower CYP3A4-mediated intrinsic clearance when compared to that determined via CYP2C19, Fogelman et  al. (1999) explained that the higher in vivo abundance of 3A isoforms will increase the relevance of the CYP3A4 isoform [5]. Wellington and Perry (2001) also refer that the N-demethylation of VEN may be carried out by CYP3A4 and CYP2C19 [29]. In spite of these evidences, other authors fundamentally identified the CYP3A4 as the main isoenzyme involved in metabolic reactions of N-demethylation without highlighting the role of CYP2C19 [4]. In contrast, McAlpine et  al. (2011) investigated the associations between the blood concentrations of VEN and its metabolite ODV, and the genetic polymorphisms of CYP2D6 and CYP2C19 isoenzymes in human subjects. These authors concluded that CYP2C19 catalyzes both O- and N-demethylation reactions in humans [4], but the presence of genetic variants of the CYP2C19 isoenzyme was not considerably associated with the concentrations found for ODV. The lack of association between CYP2C19 genetic variants and ODV concentrations may possibly be the result of the involvement of CYP2C19 in the metabolic pathways leading to the formation (from VEN) and metabolism (to DDV) of ODV. CYP2C19 genotypes showed a strong association with the concentration levels of total active moiety (VEN plus ODV); a fact explained by the involvement of CYP2C19 isoenzyme in conversion of VEN to NDV, as well as in the conversion of ODV to DDV. In another research work, Reis et al. (2002) suggest that CYP1A2 can also participate in the biotransformation of VEN, but further studies are indispensable to improve the understanding of the impact of CYP1A2 on the metabolism and clinical response of VEN [27].

Factors affecting the pharmacokinetics Nowadays, it is well known that the sources of variability in drug response are multifactorial, and apart from patients’ differences in their genetic background, a variety of other conditions such as demographic, physiological, pathophysiological, and environmental (e.g., nutrition, co-medication) factors may determine a profound impact on the pharmacokinetics and/or pharmacodynamics affecting, thereby, the clinical response [34, 35]. The knowledge of all

these factors as well as the understanding of their therapeutic relevance constitutes the first step toward the individualized medicine. Therefore, it is important to identify the key variables shared by patient subpopulations, which could be determinant to guide the pharmacotherapeutic interventions. Thus, considering the great interest of VEN as an effective therapeutic option for SSRIs-resistant depressive disorders and in an attempt to optimize the drug treatment, the co-variables that can affect the pharmacokinetics profile of VEN will be discussed.

Age and gender Despite the mentions in the summary of product characteristics of Efexor® (a trade name for VEN; John Wyeth & Brothers Ltd., Havant, UK) indicating that age and gender do not significantly affect the pharmacokinetics of VEN and ODV [44], there is scarce information on this topic and the available data are contradictory. Klamerus et al. (1996) examined the effects of age and gender on the pharmacokinetic profiles of VEN and its active metabolite (ODV) after single and multiple-dose regimens [66]. These authors concluded that no dosage adjustments are required based on the pharmacokinetic changes induced by age and gender. However, in elderly patients the steady-state half-life increased 24% for VEN and 14% for ODV. In addition, considering together VEN and ODV (pharmacologically active moieties), an increase of 16% in the extent of drug exposure was estimated in elderly at steady-state conditions (as assessed by AUC), but it was considered to be nonsignificant by authors [66]. On the contrary, a recent retrospective evaluation of 478 TDM analyses of VEN found that patients older than 60 years had about 46% higher dose-corrected serum levels of VEN and ODV than the younger ones. Moreover, women had about 30% higher dose-corrected serum levels of VEN and ODV than men [67]. Thus, these results support that TDM is a useful approach to identify factors affecting the drug pharmacokinetics, and age and gender may contribute to the inter-individual pharmacokinetic variability of VEN and these variables should be considered for optimal dosing of patients under VEN therapy.

Body weight and smoking behavior Patients with depressive disorders frequently have an elevated or a reduced body mass index and the patient’s body weight can affect the apparent volume of distribution of drugs. Nonetheless, few studies have assessed

Brought to you by | Carleton University OCUL Authenticated Download Date | 6/29/15 11:46 AM

134      Magalhães et al.: Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers the influence of body weight on the pharmacokinetics of VEN. Specifically, in a retrospective study Unterecker et al. (2011) found that the dose-corrected serum concentrations did not significantly correlate with patient’s body weight for various antidepressant agents, including VEN [68]. Accordingly, these findings suggest that there is no relevant influence of body weight on the serum levels of VEN [68]. However, these data must be interpreted with caution, attending to the small sample size for the individuals treated with VEN (n = 24) and, therefore, more robust investigations are needed. Another relevant condition that affects the concentrations of VEN and its metabolites is cigarette smoking; there are indications that smokers are more likely to have lower plasma levels of VEN and VEN-related metabolites (ODV and DDV) than nonsmokers, a fact probably associated with the higher metabolic clearance in smokers due to the smoking-mediated induction of CYP1A2 [27, 67]. However, the impact of the smoking behavior on the clinical response to VEN should be further evaluated.

Hepatic and renal impairment In patients with hepatic and renal dysfunction, the clearance of VEN is reduced, increasing the drug elimination half-life, which can lead to drug accumulation in the body and, consequently, may determine a higher predisposition to concentration-dependent toxicity [1, 29, 44, 45, 69]. Particularly, in patients with hepatic cirrhosis the VEN and ODV half-lives were prolonged by approximately 30% and 60%, respectively, and their plasma clearances decreased by 50% and 30% as compared to individuals with normal hepatic function [29]. Similarly, in patients with renal dysfunction, the half-life of VEN was prolonged by approximately 50% and the plasma clearance was reduced by approximately 24% [29]. Consequently, the dose of VEN should be reduced by approximately 50% in patients with mild to moderate hepatic impairment or with severe renal impairment (creatinine clearance rates  

Venlafaxine pharmacokinetics focused on drug metabolism and potential biomarkers.

Venlafaxine (VEN) is one of the safest and most effective drugs used in the treatment of selective serotonin reuptake inhibitors-resistant depression,...
676KB Sizes 3 Downloads 3 Views