- Email: [email protected]
Methadone
J Pharmacol Toxicol 42 (1999) 61–66
Reviews
Methadone: a review of its pharmacokinetic/pharmacodynamic properties María J. Garrido, Iñaki F. Trocóniz* Department of Pharmacy and Pharmaceutical Technology, Faculty of Pharmacy, University of Navarra, Pamplona, 31080, Spain Received 8 February 2000; accepted 8 February 2000
Abstract During the past decades the use of methadone has been increased as a result of the interest of optimizing its therapeutics in opioid addicts, one of the groups with higher risk for AIDS infection. However standard dose of methadone are far from being the appropriate for relief pain or prevent withdrawal signs in maintenance programs in many patients. To achieve an optimal dose regimen for an individual, the knowledge of the relationship between the pharmacokinetics/pharmacodynamics (pk/pd) drug properties and the demographic and physiopathological characteristics of the subject is required. Unfortunately, there is a lack of studies dealing with the population pk/pd properties of methadone. In the current study, a review of the pk/pd properties of methadone is presented with the aim of understanding the sources of variability in response. This will help in the design of prospective pk/pd studies; in particular, individual data including sex, weight, ␣1-acid glycoprotein levels in plasma, concomitant medications, time after starting treatment with methadone and previous exposure to other opioids should be requested. In addition, designs for drug administration should allow the characterization of the plasmaversus-biophase distribution and the development of tolerance processes. Because methadone is usually administered as a racemic mixture, the use of enantioselective techniques to determine both enantiomers in plasma is also highly recommended. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Methadone; Pharmacokinetics; Pharmacodynamics; Opioids
1. Introduction Methadone was introduced to the market during the sixties. It has generally been used to prevent the abstinence syndrome occurring after rapid interruption of continuous opioid administration (Dole & Nyswander, 1965). The spread of AIDS in the past two decades, as well as the effort of the health-care institutions to control the rapid increase in the number of infected people, has resulted in an increasing interest in optimizing methadone therapeutics in opioid addicts, one of the groups with higher risk for AIDS infection (Wolff et al., 1991). Methadone is also used as an analgesic drug, but to a lesser extent compared with morphine, codeine, or buprenorphine (Vigano et al., 1996). It is a common feature for patients treated with standard doses of methadone to complain about discomfort, pain, or some signs of withdrawal in maintenance programs. The literature contains many studies exploring the kinetics and the dynamics of this compound. Notwithstanding, a frequent feature in methadone therapy is continuous disagreement between authors (Ripamonti et al., 1997). Optimizing therapy
* Corresponding author. Tel: 34 948 425600 (ext 6507); Fax: 34 948 425649. E-mail address: [email protected]
means finding an appropriate dosing regimen for a particular person. Such a dosing regimen should be established on the basis of the factors determining interindividual variations in pharmacokinetics (PK) and pharmacodynamics (PD). We will now proceed to briefly review the PK and PD characteristics of methadone, which are useful to understand and anticipate sources of variability in response. 2. Pharmacokinetics Methadone is a liposoluble basic drug with a pKa of 9.2, which is usually administered orally as a racemic mixture of two enantiomers: R-methadone (R-Met) and S-methadone (S-Met). There is evidence showing that the PK of these two enantiomers significantly differ with respect to distribution and elimination. Notwithstanding, most of the available data on methadone PK properties have been obtained from studies in which the racemic mixture was employed (Nilsson et al., 1982). 2.1. Absorption Absorption after oral administration of methadone given as solution or tablets is fast and almost complete. The mean time to achieve peak plasma drug concentrations has been reported to be 2.5 h for methadone in solution (Wolff et al.,
1056-8719/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 1 0 5 6 - 8 7 1 9 ( 0 0 ) 0 0 0 4 3 - 5
62
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
1993) and 3 h for methadone in tablets (Nilsson et al., 1982). Oral bioavailability is high, ranging from 0.67 (Kristensen et al., 1996) to 0.95 (Rostami-Hodjegan et al., 1999). 2.2. Distribution Methadone exhibits a considerable tissue distribution, because it is a lipophylic drug (Säwe, 1986). To our knowledge, there are no published reports dealing with a physiological model for methadone distribution in humans. Nonetheless, the work carried out by Gabrielsson et al. (1985) in pregnant rats showed that methadone was distributed at least to brain, gut, kidney, liver, muscle, and lung with tissue to plasma partition coefficients of 4.6, 37.2, 76.6, 44.2, 14.7, and 156.3, respectively. These results are consistent with the high volume of distribution reported in humans. Despite high differences in the mean estimates reported for the apparent volume of distribution at steady-state (Vss) by several authors, in most cases, values are much higher than the actual physiological volume, indicating that tissue binding predominates over binding to plasma proteins. In the work published by Wolf et al., (1993) on opioid addicts, Vss ranged from 4.2 to 9.2 l/kg. On the other hand, in patients with chronic pain, Vss ranged from 1.71 to 5.34 l/kg (Inturrisi et al., 1987). Methadone binds to plasma proteins to a high degree— 86% (Inturrisi et al., 1987). This binding was found to be similar in other species, such as rats (Garrido et al., 1996; Gómez et al., 1995). Owing to its basic properties, methadone binds predominantly to ␣1-acid glycoprotein (AAG) (Romanch et al., 1981; Wilkins et al., 1997). AAG is an acute-phase reactant protein that exhibits variations in its plasma levels, depending on physiological or pathological conditions (Olsen, 1973). It is generally recognized that, during stress conditions, AAG levels show a significant increase. Such an increase was found to be the main factor responsible for lower free fractions (fu) in plasma for methadone in cancer patients and opiate addicts compared with healthy volunteers (Abramson, 1982; Calvo et al., 1996). These observations open up an interesting discussion about the effect of changes in plasma protein binding on both total (Cp) and unbound (Cu) plasma drug concentrations. After a rapid drug input (i.e., bolus), a decrease in fu will be indicated in the early period after drug administration by an increase in Cp, because Vss is proportional to fu. On the other hand, Cu levels will remain unchanged. If Cu is assumed to be the pharmacological active concentration, a decrease in fu will not modify the maximum observed response. The rate of tissue distribution, however, will be increased because, for a lipophylic drug, this rate depends on Cp and not on Cu (Rowland & Tozer, 1995). These theoretical considerations were pointed in a study published by Garrido et al. (1999). Methadone was administered as an intravenous bolus of 0.35 mg/kg dose to control (healthy) and abstinent rats. AAG levels in plasma were increased in rats exhibiting abstinence syndrome and, consequently, fu was decreased.
The PK of Cp were significantly modified in abstinent rats, but the Cu-versus-time profiles were very similar between the two groups. With respect to the access of methadone to the brain, the peak concentrations in abstinent rats appeared almost immediately after drug injection. In the control group, maximum concentrations were located 10 min after drug administration. Similar results have been reported for another basic drug, lidocaine, in dogs (Marathe et al., 1991). On the basis of the preceding considerations, a study with the aim of prospectively characterizing methadone distribution should pay attention to demographics such as weight and sex (factors affecting body composition) and AAG. In fact, in a recent study in opioid addicts, sex and weight were the two major covariates accounting for 33% of the interindividual variability in Vss. This parameter was found to be higher in female than in male opiate users, and it was directly related to weight. A time-dependent decrease in Vss was associated with the observed time-dependent increase in AAG (Rostami-Hodjegan et al., 1999). 2.3. Elimination Methadone undergoes hepatic metabolism and renal excretion. As a result of its basic (pKa ⫽ 9.2) and lipophylic properties, changes in the pH of the urinary tract can be an important determinant in the elimination of methadone. In fact, at a urinary pH above 6, renal clearance constitutes only 4% of total drug elimination. However, when urinary pH is below 6, the unchanged methadone excreted by the renal route can increase to 30% of the total administered dose (Änggård et al., 1975; Inturrisi et al., 1987). Rostami-Hodjegan et al. (1999) reported a 27% decrease in the estimate of the interindividual variability in methadone clearance when urine pH was incorporated in the model as a covariate. With regard to its hepatic elimination, methadone can be considered to be a restrictive cleared or a low extraction ratio drug. Mean estimates of 3.1 and 1.5 ml/min/kg for clearance have been reported in opioid addicts and in patients with chronic pain, respectively (Änggård et al., 1975; Inturrisi et al., 1987). With the use of the 0.75 estimate for the blood-toplasma ratio reported by Inturrisi et al. (1987) and the assumption of a blood flow rate of 1500 ml/min, the resulting hepatic extraction ratio is 0.16 and 0.08 for opioid addicts and for patients with chronic pain, respectively. These elimination characteristics explain the high bioavailability after oral administration of methadone and have highly relevant consequences regarding interindividual variability in clearance. Clearance will depend on both, fu and intrinsic clearance (Clint), which represents the intrinsic enzymatic activity of the liver. On the other hand, bioavailability will remain mostly unaffected by changes in fu, Clint, and liver perfusion. It has previously been stated that AAG levels and fu are modified in cancer patients and opioid addicts (Abramson, 1982; Calvo et al., 1996). Therefore, plasma protein binding should be considered a potential factor responsible, at least in part, for the interindividual variations in clearance.
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
In Section 2.2, a brief discussion of the effect of alterations in fu on Cp and Cu after rapid administration was presented. In a study of the kinetics of a drug and the effect of changes in fu in a steady-state situation, is important to realize that steady-state concentrations (Cpss) depend on the rate of administration (k0) and drug clearance (CL), as determined by the following equation: Cpss ⫽ k0/CL. For a drug with a low extraction ratio, such as methadone, a decrease in fu will decrease CL, and consequently Cpss will be increased proportionally. However, steady-state unbound methadone concentrations (Cuss) will remain constant because Cuss ⫽ fu*Cpss. For both, Cp and Cu, the time to reach steady-state is not modified, because V and Cl both decrease proportionally with fu. Methadone biotransformation is mainly mediated by CYP3A4, although CYP2C9 and CYP2C19 also appear to take part, but to a much lesser extent (Foster et al., 1999). Interindividual variations in the expression of CYP3A4 will be the main factor responsible for interindividual variability in clearance. On the other hand, although methadone has been shown to be able to alter several CYP2D6 substrates metabolism, this enzyme does not seem catalyze the biotransformation of methadone. CYP3A4 is an inducible enzyme. This characteristic contributes to making the understanding and prediction of individual clearances of methadone more difficult. RostamiHodjegan et al. (1999) reported a 3.5-fold increase in the total clearance between the first dose of methadone and steady-state doses in opiate users. This finding was explained by the capability of the drug to induce its own metabolism. On the basis of the preceding results, time from the start of treatment should be considered a covariate to be included in clearance. Another covariate that investigators should consider is the presence of concomitant medication with drugs such as carbamazepine, phenytoin, rifampicin, zidovudine, barbiturates, spironolactone, verapamil, diethylstilboestrol, and amitriptyline (Wolff et al., 1993). There are data in the literature showing that these drugs induce methadone elimination kinetics. On the other hand, in vitro studies have shown that fluoxetine inhibited, in about 50%, the formation of the main metabolite of methadone. Similar results were found with omeprazole and carbamazepine (Foster et al., 1999). Inducers of CYP3A4, such as rifampicin, also have been shown to be capable of up-regulating the production of AAG, increasing the plasma binding of basic drugs (Abramson & Lutz, 1986). This up-regulation phenomenon has been suggested to be implicated in the slight increase in AAG seen after long treatment with methadone (Rostami-Hodjegan et al., 1999). Although distribution characteristics in humans and small laboratory animals were found to be similar, this is not the case for drug elimination. Methadone elimination in rats has been reported to be nonrestrictive, with a mean clearance of 0.1 l/min/kg. Differences between species in the elimination process are probably due to differences in the expression of CYP isoforms in methadone metabolism.
63
2.4. Disposition After single doses, plasma drug concentrations of methadone-versus-time profiles have been frequently described by using a two-open compartmental model. However, Inturrisi et al. (1987) used a triexponential model to describe the kinetics in plasma for patients with chronic pain. Despite the pharmacokinetic model used to describe the time course of plasma concentrations, the reported estimates for the terminal halflife are similar: 26.8 h (Wolff et al., 1993) and 23 h (Inturrisi et al., 1987). Rostami-Hodjegan et al. (1999) reported a time dependency in the terminal half-life, which was explained by the ability of methadone to induce its own metabolism. It should be noted, however, that variations in clearance do not always have any effect on half-life, because this secondary parameter also depends on the volume of distribution. 2.5. Stereoselective pharmacokinetics The estimates of volume of distribution reported for R- and S-Met were 496.6 l and 289.1 l, respectively, in chronic pain patients (Kristensen et al., 1996). These same authors reported the clearance of R-Met and S-Met as 0.158 and 0.129 l/min, respectively. Schmidt et al. (1994) found that clearance for R-Met in beagle dogs also was slightly higher than that obtained for S-Met: 0.357 versus 0.316 l/min, respectively. Notwithstanding, in a recent study, Foster et al. (1999) found a lack of stereoselective metabolism in human microsomes. These results suggest that the differences found in plasma clearance between enantiomers could be due to differences in plasma protein binding. Eap et al. (1990) reported fu values for R-Met and S-Met of 0.13 and 0.10, respectively. Such difference, although small, can explain, at least in part, the observed differences in plasma clearance between the two enantiomers. Although the discussed results show differences in kinetics between enantiomers, on the basis of the estimates of volume of distribution and clearance, it can be assumed that changes in protein binding, body composition, metabolism, and other demographic data will affect both enantiomers in a similar way. Regarding absorption processes, Kristensen et al. (1996) reported no differences between enantiomers in lag time and bioavailability. It should be nonetheless mentioned that published data on the absorption of methadone enantiomers are scarce.
3. Pharmacodynamics 3.1. Basic pharmacodynamic properties Methadone is a synthetic opioid agonist with morphinelike properties, mostly used as a maintenance drug for opioid addicts. Notwithstanding, it has also proved to be a powerful analgesic in patients with malignant and postoperative pain. Methadone elicits its pharmacodynamic properties through binding to , ␦ and opioid receptors. Methadone-receptor binding has been characterized in in vitro
64
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
studies. The affinity constant (Ki) to receptors was reported to be 3.51 nM, whereas the Ki value for morphine is 1.41 nM. These results indicate that methadone has a lower affinity for this receptor than does morphine (Blake et al., 1997). The receptor is the principal factor responsible for most of the important opioid pharmacodynamic properties, because its activation produces analgesia, respiratory depression, physiological dependence, and tolerance. It is sometimes difficult to dissociate therapeutic effects from adverse effects (Ling et al., 1985; Matthes et al., 1996). Methadone is described as inducing rapid dependence in rats. The degree of such dependence can be quantified in terms of the incidence of specific withdrawal signs after naloxone administration (Pierce et al., 1996). However, in humans, methadone has shown a lower potential for abuse than morphine, methadone being the drug of choice in maintenance therapy programs for opioid users (Blake et al., 1997). Another important complication during the use of opioids is the development of tolerance. Tolerance can be defined as a decrease in the analgesic potency (or other opioid effect) of opioid agonists after prior treatment with the opioid (Chan et al., 1997). Although this phenomenon is well known in experimental and clinical pharmacology, its mechanism of action is still poorly understood. Several authors suggest that tolerance is due, at least in part, to a change in the expression and function of receptors (Yoburn et al., 1993). The use of selective agonist and antagonist ligands for the different opioid receptors has been proposed to be a suitable tool to investigate the molecular alterations in the opioid-receptor system after opioid exposure in both in vivo and in vitro studies. Different strategies can be used to study the development of tolerance and the capacity of different full agonists to produce such phenomenon. The use of an irreversible opioid antagonist such as -funaltrexamine (-FNA), which produces a dose-dependent inactivation of the -receptor population, provides information about the relative efficacy between full agonists (Pitts et al., 1998). Adams et al. (1990), in an in vivo rat study, found that a 5 g i.c.v. dose of -FNA was capable of reducing the maximum analgesic effect of morphine but not that of methadone. These results suggest that methadone has a higher intrinsic efficacy than morphine does. Cross-tolerance is a phenomenon that results in a diminished response to an opioid agonist after previous treatment with another opioid agonist. Paronis and Holtzman (1992) studied the development of tolerance to the analgesic effects of several opioid agonists, including methadone, after continuous infusion of morphine, meperidine, or fentanyl in rats. They found that the ED50 for methadone before and after morphine exposure did not change, suggesting that methadone is less susceptible to inducing tolerance than is morphine and that this property is inversely related to the drug efficacy. On the other hand, the in vivo description/ prediction of the development of tolerance phenomenon can also be explored by using pharmacokinetic/pharmacodynamic models (Ekblom et al., 1993; Mandema & Wada, 1995; Oullet & Pollack, 1995).
3.2. Stereoselective pharmacodynamics R-Met exhibits a 10-fold higher affinity for the and ␦ opioid receptors and as much as a 50 times higher analgesic activity than S-Met does in human and animal models of pain (Foster et al., 1999). R-Met has been shown to prevent the occurrence of opioid withdrawal symptoms, whereas S-Met appears to be inactive as an opioid. Nonetheless, S-Met is active as an NMDA receptor antagonist, blocking morphine tolerance after systemic and intrathecal administration and NMDA-induced hyperalgesia. These results suggest that S-Met could be used alone or in combination with morphine to treat the hyperalgesic component of neuropathic pain, as well as to improve the analgesic efficacy of chronic morphine administration (Davis & Inturrisi, 1999).
4. Pharmacokinetic/pharmacodynamic relationships Pharmacokinetic/pharmacodynamic (PK/PD) relationships allow the description/prediction and characterization of the time course of the in vivo drug effect. There are several studies dealing with such relationships for opioids commonly used in anesthesia, such as alfentanyl, fentanyl, sufentanyl, and remifentanyl (Cox et al., 1999; Mandema & Wada, 1995). In regard to opioids used for the treatment of pain, studies of PK/PD relationships are less frequent, being available mainly for morphine (Ekblom et al., 1993; Gårdmark et al., 1993; Oullet & Pollack, 1995). A priori, some issues should be taken into consideration when the goal is to establish the PK/PD relationship for an opioid drug. It should be noted that the plasma is not the effect site and that usually a certain time is needed before the drug distributes into the central nervous system. This phenomenon has important implications, because Cp cannot be used directly to describe the observed effect. To overcome this difficulty, the effect-versus-time data can be analyzed by using the “effect-compartment model,” or “link model.” This model has frequently been used in PK/PD modeling for opioid drugs (Ekblom et al., 1993; Gårdmark et al., 1993). Another important factor is the possibility for tolerance development, which means that the pharmacodynamic relationship found after a single drug administration will not be adequate to represent the dynamics of the drug after continuous exposure. In addition to the “link model,” different models have been proposed in recent years to describe the development of tolerance to opioids (Ekblom et al., 1993; Gårdmark et al., 1993; Oullet & Pollack, 1995). Unfortunately, the adequate characterization of both biophase distribution and eventual tolerance development requires complex experimental designs. It is recommended to use of a variety of doses, administered by different routes and varying infusion lengths. Experimental designs are even more complicated for drugs with one or more active metabolites, as is the case with morphine and remifentanyl (Cox et al., 1999). To our knowledge, only four studies address the PK/PD
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
relationship of methadone. Inturrisi et al. (1987; 1990) studied patients with chronic pain. In the first study (Inturrisi et al., 1987), they administered methadone as a single 10–30 mg i.v. bolus dose to eight patients who had been receiving opioid treatment other than methadone. In addition to plasma drug concentrations, they determined the analgesic effect by means of a visual analog scale, blood pressure, pulse rate, respiratory rate, and pupil size. However, only the analgesic response was subjected to PK/PD modeling. The experimental design allowed an estimation of ke0, the first-order rate constant governing the distribution between plasma and biophase, and C50, the steady-state methadone concentration eliciting 50% of maximum pain relief. The mean estimated values were 0.19 l/min and 0.29 g/ml for ke0 and C50, respectively. Both parameters showed high interindividual variability, probably because of different grades of cross-tolerance between patients. A previous mean estimate of C50 for methadone, obtained after administration to relatively nontolerant subjects, was 0.06 g/ml (Gourlay et al., 1984). In the second study, Inturrisi et al. (1990) also studied the PK/PD relationship for methadone in patients with chronic pain who had previous exposure to opioids other than methadone. However, in this study, methadone was given as an infusion, rather than as a bolus. The drug was continuously infused for 3 to 4.5 h, and drug plasma concentration, pain relief, and sedation were measured during the infusion and 4 to 5 h after the end of the infusion. Pain relief and sedation data were reported by each subject by using visual analog scales. The mean estimates of the pharmacodynamic parameters were very similar for the two measured effects. C50 was 0.36 and 0.34 g/ml, and ␥ (the slope of the effect against steady-state drug concentration curve) was 4.4 and 5.8 for pain relief and sedation, respectively. Both parameters showed high interindividual variability. It should be noted that the estimate of C50 for pain relief was very similar to that obtained in the previous study with the use of i.v. bolus administration (Inturrisi et al., 1987), indicating that both designs seem to be adequate to estimate the dynamic properties of methadone. However, for almost 50% of the patients, an estimate of ke0 could not be obtained for pain relief and sedation. These results, although they could be interpreted in principle as a very rapid equilibrium between plasma and biophase, indicate that the design did not allow an adequate characterization of the plasma to biophase distribution kinetics. On the other hand, the long-infusion design is appropriate for identifying tolerance. As constant plasma concentration is maintained and plasma-versusbiophase equilibrium is achieved, a decrease in the effect can only be attributed to tolerance. No tolerance development was seen for methadone in this case. Dyer et al. (1999) studied the methadone PK/PD relationship in methadone-maintenance patients. Methadone plasma concentrations were directly related to pain relief, pupil size, and withdrawal score data by using a sigmoidal Emax model. In this study, 18 patients received constant
65
doses of methadone once daily for at least 2 months. In the PK/PD analysis, two major assumptions were made: (1) concentrations of methadone in plasma and brain were in equilibrium (because all the patients were at steady state) and (2) tolerance to effects was present but stable. Estimates of C50 for withdrawal scores, pupil size, and pain relief were 0.3, 0.28, and 0.18 g/l, respectively, and were also in the range of the previously reported estimates of C50 for pain relief and sedation (Inturrisi et al., 1987, 1990). Garrido et al. (1999) designed an experiment to assess the implications of the abstinence syndrome in the PK/PD relationships for methadone in rats. To induce physical dependence, a morphine solution was delivered continuously at a rate of 0.5 l/h for 6 days with the use of minipumps implanted under the skin. The spontaneous withdrawal syndrome emerged after removal of the minipumps. At the time of maximal withdrawal syndrome, an i.v. 0.35 mg/kg bolus dose of methadone was administered. Plasma and antinociception (measured by the tail-flick test) were simultaneously determined. An effect-compartment model was used to describe the time course of the antinociceptive effect. An increase in C50, from 24.2 (control rats) to 33.6 ng/ ml (abstinent rats) was observed. A cross-tolerance phenomena was suggested as responsible for the increase in C50. 5. Concluding remarks Methadone shows some peculiarities in its pharmacokinetics and pharmacodynamics, which should be taken into consideration if a prospective PK/PD study has to be designed. Regarding pharmacokinetics, weight, sex, and, if possible, AAG levels in plasma should be part of the known individual demographics. In addition, owing to the fact that CYP3A4 is the principal enzyme in the metabolism of methadone as well as other numerous drugs, data about concomitant medications are required. Time after starting treatment with methadone also can be considered a potential covariate because the drug has been shown to be able to induce its own metabolism. Pharmacodynamic properties of methadone are determined by its affinity for the -opioid receptor, which implies tolerance development, cross-tolerance, and dependence. Information regarding previous exposure to other opioids will be useful. Moreover, designs for drug administration should be devised to achieve a proper characterization of both plasma-versus-biophase distribution and development of tolerance processes. Because the pharmacokinetic and pharmacodynamic properties of methadone enantiomers are different, the use of an enantioselective technique for the determination of the drug in plasma also is recommended. References Abramson, F. P. (1982). Methadone plasma protein binding: alterations in cancer and displacement from ␣1-acid glycoprotein. Clin Pharmacol Ther 32, 652–658.
66
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
Abramson, F. P., & Lutz, M. P. (1986). The kinetics of induction by rifampicin of ␣1-acid glycoprotein and antipyrine clearance in the dog. Drug Metab Dispos 14, 46–51. Adams, J. U., Paronis, C. A., & Holtzman, S. G. (1990). Assessment of relative intrinsic activity of mu-opioid analgesics in vivo using -funaltrexamine. J Pharmacol Exp Ther 255, 803–808. Änggård, E., Gunne, L. M., Holmstrand, J., McMahon, R. E., Sandberg, C. G., & Sullivan, H. R. (1975). Disposition of methadone in methadone maintenace. Clin Pharmacol Ther 17, 258–266. Blake, A. D., Bot, G., Freeman, J. C., & Reisine, T. (1997). Differential opioid agonist regulation of the mouse opioid receptor. J Biol Chem 272, 782–790. Calvo, R., Aguirre, C., Troconiz, I. F., López, J., & Garrido, M. J. (1996). Alpha1-acid glycoprotein and serum protein binding of methadone in heroin addicts during withdrawal. Proceedings of the Sixth World Congress on Clinical Pharmacology and Therapeutics, Buenos Aires, Argentina, p. 174. Cox, E. H., Langemeijer, M. W. E., Gubbens-Stibbe, J. M., Muir, K. T., & Danhof, M. (1999). The compartive pharmacodynamic of remifentanyl and its metabolite, GR90291, in a rat electroencephalographic model. Anesthesiology 90, 535–544. Chan, K. W., Duttory, A., & Yoburn, B. C. (1997). Magnitude of tolerance to fentanyl is independent of -opioid receptor density. Eur J Pharmacol 319, 225–228. Davis, A. M., & Inturrisi, C. E. (1999). D-Methadone blocks morphine tolerance and N-methyl-D-aspartate-induced hyperalgesia. J Pharmacol Exp Ther 289, 1048–1053. Dole, V., & Nyswander, M. (1965). A medical treatment for diacetylmorphine (heroin) addiction. JAMA 193, 646–650. Dyer, K. R., Foster, D. J. R., White, J. M., Somogyi, A. A., Menelaou, A., & Bochner, F. (1999). Steady-state pharmacokinetics and pharmacodynamics in methadone maintenance patients: comparison of those who do and do not experience withdrawal and concentration-effect relationships. Clin Pharmacol Ther 65, 685–694. Eap, C. B., Cuendet, C., & Baumann, P. (1990). Binding of d-methadone, l-methadone and dl-methadone to proteins in plasma of healthy volunteers: role of the variants of ␣1-acid glycoprotein. Clin Pharmacol Ther 47, 338–346. Ekblom, M., Hammarlund-Udenaes, M., & Paalzow, L. (1993). Tolerance development and rebound effects to morphine in relation to different administrations in rats. J Pharmacol Exp Ther 266, 244–252. Foster, D. J. R., Somogyi, A. A., & Bochner, F. (1999). Methadone N-demethylation in human liver microsomes: lack of stereselectivity and involvement of CYP3A4. Br J Clin Pharmacol 47, 403–412. Gabrielsson, J. L., Johansson, P., Bondesson, U., & Paalzow, L. K. (1985). Analysis of methadone disposition in the pregnant rat by means of physiological flow model. J Pharmacokinet Biopharm 13, 355–372. Gårdmark, M., Eblom, M., Bouw, R., & Hammarlund-Udenaes, M. (1993). Quantification of effect delay and acute tolerance development to morphine in the rat. J Pharmacol Exp Ther 267, 1061–1067. Garrido, M. J., Jiménez, R., Gómez, E., & Calvo, R. (1996). Influence of plasma protein binding on analgesic effect on methadone in rats with spontaneous withdrawal. J Pharm Pharmacol 48, 281–284. Garrido, M. J., Valle, M., Calvo, R., & Trocóniz, I. F. (1999). Altered plasma and brain disposition and pharmacodynamics of methadone in abstinent rats. J Pharmacol Exp Ther 288, 179–187. Gómez, E., Martínez-Jordá, R., Suárez, E., Garrido, M. J., & Calvo, R. (1995). Altered methadone analgesia due to changes in plasma protein binding: role of the route of administration. Gen Pharmacol 26, 1273–1276. Gourlay, G. K., Willis, R. J., Wilson, P. R. (1984). Postoperative pain control with methadone: influence of supplementary methadone doses and blood concentration-response relationships. Anesthesiology 61, 19–26. Inturrisi, C. E., Colburn, W. A., Kaiko, R. F., Houde, R. W., & Foley, K. M. (1987). Pharmacokinetics and pharmacodynamics of methadone in patients with chronic pain. Clin Pharmacol Ther 41, 392–401. Inturrisi, C. E., Portenoy, R. K., Max, M. B., Colburn, W. A., & Foley, K. M. (1990). Pharmacokinetic-pharmacodynamic relationships of metha-
done infusions in patients with cancer pain. Clin Pharmacol Ther 47, 565–577. Kristensen, K., Blemmer, T., Angelo, H. R., Christrup, L. L., Drenck, N. E., Rasmussen, S. N., & Sjøgren, P. (1996). Stereoselective pharmacokinetics of methadone in chronic pain patients. Ther Drug Monitor 18, 221–227. Ling, G. S. F., Spiegel, K., Lockhart, S. H., & Pasternak, G. W. (1985). Separation of opioid analgesia from respiratory depression: evidence for different receptor mechanisms. J Pharmacol Exp Ther 232, 149–155. Mandema, J. M., & Wada, D. R. (1995). Pharmacodynamic model for acute tolerance development to the electroencephalographic effects of alfentanyl in the rat. J Pharmacol Exp Ther 275, 1185–1194. Marathe, P. H., Shen, D. D., Artru, A. A., & Bowdle, A. (1991). Effect of serum protein binding on the entry of lidocaine into brain and cerebrospinal fluid in dogs. Anesthesiology 75, 804–812. Matthes, H. W. D., Maldonado, R., Simonin, F., Valverde, O., Slowe, S., Kitchen, I., Befort, K., Dierich, A., Le Meur, M., Dollé, P., Tzavara, E., Hanoune, J., Roques, B. P., & Kieffer, B. L. (1996). Loss of morphineinduced analgesia, reward effect and withdrawal symptoms in mice lacking the -opioid-receptor gene. Nature 383, 819–823. Nilsson, M. J., Widelöv, E., Meresaar, U., & Änggård, E. (1982). Effect of urinary pH on the disposition of methadone in man. Eur J Clin Pharmacol 22, 337–342. Olsen, G. D. (1973). Methadone binding to human plasma proteins. Clin Pharmcol Ther 14, 338–343. Oullet, D. M., & Pollack, G. M. (1995). A pharmacokinetic-pharmacodynamic model of tolerance to morphine analgesia during infusion in rats. J Pharmacokinet Biopharm 23, 531–549. Paronis, C. A., & Holtzman, S. G. (1992). Development of tolerance to the analgesic activity of mu agonists after continuous infusion of morphine, meperidine or fentanyl in rats. J Pharmacol Exp Ther 262, 1–9. Pierce, T. L., Hope, W., & Raper, C. (1996). The induction and quantitation of methadone dependence in the rat. J Pharmacol Toxicol Methods 36, 137–146. Pitts, R. C., Allen, R. M., Walker, E. A., & Dykstra, L. A. (1998). Clocinamox antagonism of the antinociceptive effects of mu opioids in squirrel monkeys. J Pharmacol Exp Ther 285, 1197–1206. Ripamonti, C., Zecca, E., & Bruera, E. (1997). An update on the clinical use of methadone for cancer pain. Pain 70, 109–115. Romanch, M. K., Piafsky, K. M., Abel, J. G., Khouw, V., & Sellers, E. M. (1981). Methadone binding to orosomucoid (␣1-acid glycoprotein): determinant of free fraction in plasma. Clin Pharmacol Ther 29, 211–217. Rostami-Hodjegan, A., Wolff, W., Hay, A. W. M., Raistrick, D., & Calvert, R. (1999). Population pharmacokinetics of methadone in opiate users: characterization on time-dependent changes. Br J Clin Pharmacol 48, 43–52. Rowland, M., & Tozer, T. N. (1995). Physiologic concepts and kinetics: distribution. In M. Rowland & T. N. Tozer (Eds.), Clinical Pharmacokinetics: Concepts and Applications, 3rd ed. (pp. 137–183). Media, PA: William & Wilkins. Säwe, J. (1986). High-dose morphine and methadone in cancer patients: clinical pharmacokinetic considerations of oral treatment. Clin Pharmacokinet 11, 87–106. Schmidt, N., Brune, K., Williams, K. M., & Geisslinger, G. (1994). Stereoselective pharmacokinetics of methadone in beagle dogs. Chirality 6, 492–495. Vigano, A., Fan, D., & Bruera, E. (1996). Individualized use of methadone and opioid rotation in the comprehensive management of cancer pain associated with poor prognostic indicators. Pain 67, 115–119. Wilkins, J. N., Ashofteh, A., Setoda, D., Wheatley, W. S., Huigen, H., & Ling, W. (1997). Ultrafiltration using the Amicon MPS-1 for assessing methadone plasma protein binding. Ther Drug Monit 19, 83–87. Wolff, K., Hay, A., & Raistrick, D. (1991). High-dose methadone and the need for drug measurements in plasma. Clin Chem 37, 1651–1664. Wolff, K., Hay, A. W. M., Raistrick, D., & Calvert, R. (1993). Steady-state pharmacokinetics of methadone in opioid addicts. Eur J Clin Pharmacol 44, 189–194. Yoburn, B. C., Barbara, B., & Duttaroy, A. (1993). Opioid receptor regulation in mice. J Pharmacol Exp Ther 265, 314–317.
Reviews
Methadone: a review of its pharmacokinetic/pharmacodynamic properties María J. Garrido, Iñaki F. Trocóniz* Department of Pharmacy and Pharmaceutical Technology, Faculty of Pharmacy, University of Navarra, Pamplona, 31080, Spain Received 8 February 2000; accepted 8 February 2000
Abstract During the past decades the use of methadone has been increased as a result of the interest of optimizing its therapeutics in opioid addicts, one of the groups with higher risk for AIDS infection. However standard dose of methadone are far from being the appropriate for relief pain or prevent withdrawal signs in maintenance programs in many patients. To achieve an optimal dose regimen for an individual, the knowledge of the relationship between the pharmacokinetics/pharmacodynamics (pk/pd) drug properties and the demographic and physiopathological characteristics of the subject is required. Unfortunately, there is a lack of studies dealing with the population pk/pd properties of methadone. In the current study, a review of the pk/pd properties of methadone is presented with the aim of understanding the sources of variability in response. This will help in the design of prospective pk/pd studies; in particular, individual data including sex, weight, ␣1-acid glycoprotein levels in plasma, concomitant medications, time after starting treatment with methadone and previous exposure to other opioids should be requested. In addition, designs for drug administration should allow the characterization of the plasmaversus-biophase distribution and the development of tolerance processes. Because methadone is usually administered as a racemic mixture, the use of enantioselective techniques to determine both enantiomers in plasma is also highly recommended. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Methadone; Pharmacokinetics; Pharmacodynamics; Opioids
1. Introduction Methadone was introduced to the market during the sixties. It has generally been used to prevent the abstinence syndrome occurring after rapid interruption of continuous opioid administration (Dole & Nyswander, 1965). The spread of AIDS in the past two decades, as well as the effort of the health-care institutions to control the rapid increase in the number of infected people, has resulted in an increasing interest in optimizing methadone therapeutics in opioid addicts, one of the groups with higher risk for AIDS infection (Wolff et al., 1991). Methadone is also used as an analgesic drug, but to a lesser extent compared with morphine, codeine, or buprenorphine (Vigano et al., 1996). It is a common feature for patients treated with standard doses of methadone to complain about discomfort, pain, or some signs of withdrawal in maintenance programs. The literature contains many studies exploring the kinetics and the dynamics of this compound. Notwithstanding, a frequent feature in methadone therapy is continuous disagreement between authors (Ripamonti et al., 1997). Optimizing therapy
* Corresponding author. Tel: 34 948 425600 (ext 6507); Fax: 34 948 425649. E-mail address: [email protected]
means finding an appropriate dosing regimen for a particular person. Such a dosing regimen should be established on the basis of the factors determining interindividual variations in pharmacokinetics (PK) and pharmacodynamics (PD). We will now proceed to briefly review the PK and PD characteristics of methadone, which are useful to understand and anticipate sources of variability in response. 2. Pharmacokinetics Methadone is a liposoluble basic drug with a pKa of 9.2, which is usually administered orally as a racemic mixture of two enantiomers: R-methadone (R-Met) and S-methadone (S-Met). There is evidence showing that the PK of these two enantiomers significantly differ with respect to distribution and elimination. Notwithstanding, most of the available data on methadone PK properties have been obtained from studies in which the racemic mixture was employed (Nilsson et al., 1982). 2.1. Absorption Absorption after oral administration of methadone given as solution or tablets is fast and almost complete. The mean time to achieve peak plasma drug concentrations has been reported to be 2.5 h for methadone in solution (Wolff et al.,
1056-8719/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 1 0 5 6 - 8 7 1 9 ( 0 0 ) 0 0 0 4 3 - 5
62
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
1993) and 3 h for methadone in tablets (Nilsson et al., 1982). Oral bioavailability is high, ranging from 0.67 (Kristensen et al., 1996) to 0.95 (Rostami-Hodjegan et al., 1999). 2.2. Distribution Methadone exhibits a considerable tissue distribution, because it is a lipophylic drug (Säwe, 1986). To our knowledge, there are no published reports dealing with a physiological model for methadone distribution in humans. Nonetheless, the work carried out by Gabrielsson et al. (1985) in pregnant rats showed that methadone was distributed at least to brain, gut, kidney, liver, muscle, and lung with tissue to plasma partition coefficients of 4.6, 37.2, 76.6, 44.2, 14.7, and 156.3, respectively. These results are consistent with the high volume of distribution reported in humans. Despite high differences in the mean estimates reported for the apparent volume of distribution at steady-state (Vss) by several authors, in most cases, values are much higher than the actual physiological volume, indicating that tissue binding predominates over binding to plasma proteins. In the work published by Wolf et al., (1993) on opioid addicts, Vss ranged from 4.2 to 9.2 l/kg. On the other hand, in patients with chronic pain, Vss ranged from 1.71 to 5.34 l/kg (Inturrisi et al., 1987). Methadone binds to plasma proteins to a high degree— 86% (Inturrisi et al., 1987). This binding was found to be similar in other species, such as rats (Garrido et al., 1996; Gómez et al., 1995). Owing to its basic properties, methadone binds predominantly to ␣1-acid glycoprotein (AAG) (Romanch et al., 1981; Wilkins et al., 1997). AAG is an acute-phase reactant protein that exhibits variations in its plasma levels, depending on physiological or pathological conditions (Olsen, 1973). It is generally recognized that, during stress conditions, AAG levels show a significant increase. Such an increase was found to be the main factor responsible for lower free fractions (fu) in plasma for methadone in cancer patients and opiate addicts compared with healthy volunteers (Abramson, 1982; Calvo et al., 1996). These observations open up an interesting discussion about the effect of changes in plasma protein binding on both total (Cp) and unbound (Cu) plasma drug concentrations. After a rapid drug input (i.e., bolus), a decrease in fu will be indicated in the early period after drug administration by an increase in Cp, because Vss is proportional to fu. On the other hand, Cu levels will remain unchanged. If Cu is assumed to be the pharmacological active concentration, a decrease in fu will not modify the maximum observed response. The rate of tissue distribution, however, will be increased because, for a lipophylic drug, this rate depends on Cp and not on Cu (Rowland & Tozer, 1995). These theoretical considerations were pointed in a study published by Garrido et al. (1999). Methadone was administered as an intravenous bolus of 0.35 mg/kg dose to control (healthy) and abstinent rats. AAG levels in plasma were increased in rats exhibiting abstinence syndrome and, consequently, fu was decreased.
The PK of Cp were significantly modified in abstinent rats, but the Cu-versus-time profiles were very similar between the two groups. With respect to the access of methadone to the brain, the peak concentrations in abstinent rats appeared almost immediately after drug injection. In the control group, maximum concentrations were located 10 min after drug administration. Similar results have been reported for another basic drug, lidocaine, in dogs (Marathe et al., 1991). On the basis of the preceding considerations, a study with the aim of prospectively characterizing methadone distribution should pay attention to demographics such as weight and sex (factors affecting body composition) and AAG. In fact, in a recent study in opioid addicts, sex and weight were the two major covariates accounting for 33% of the interindividual variability in Vss. This parameter was found to be higher in female than in male opiate users, and it was directly related to weight. A time-dependent decrease in Vss was associated with the observed time-dependent increase in AAG (Rostami-Hodjegan et al., 1999). 2.3. Elimination Methadone undergoes hepatic metabolism and renal excretion. As a result of its basic (pKa ⫽ 9.2) and lipophylic properties, changes in the pH of the urinary tract can be an important determinant in the elimination of methadone. In fact, at a urinary pH above 6, renal clearance constitutes only 4% of total drug elimination. However, when urinary pH is below 6, the unchanged methadone excreted by the renal route can increase to 30% of the total administered dose (Änggård et al., 1975; Inturrisi et al., 1987). Rostami-Hodjegan et al. (1999) reported a 27% decrease in the estimate of the interindividual variability in methadone clearance when urine pH was incorporated in the model as a covariate. With regard to its hepatic elimination, methadone can be considered to be a restrictive cleared or a low extraction ratio drug. Mean estimates of 3.1 and 1.5 ml/min/kg for clearance have been reported in opioid addicts and in patients with chronic pain, respectively (Änggård et al., 1975; Inturrisi et al., 1987). With the use of the 0.75 estimate for the blood-toplasma ratio reported by Inturrisi et al. (1987) and the assumption of a blood flow rate of 1500 ml/min, the resulting hepatic extraction ratio is 0.16 and 0.08 for opioid addicts and for patients with chronic pain, respectively. These elimination characteristics explain the high bioavailability after oral administration of methadone and have highly relevant consequences regarding interindividual variability in clearance. Clearance will depend on both, fu and intrinsic clearance (Clint), which represents the intrinsic enzymatic activity of the liver. On the other hand, bioavailability will remain mostly unaffected by changes in fu, Clint, and liver perfusion. It has previously been stated that AAG levels and fu are modified in cancer patients and opioid addicts (Abramson, 1982; Calvo et al., 1996). Therefore, plasma protein binding should be considered a potential factor responsible, at least in part, for the interindividual variations in clearance.
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
In Section 2.2, a brief discussion of the effect of alterations in fu on Cp and Cu after rapid administration was presented. In a study of the kinetics of a drug and the effect of changes in fu in a steady-state situation, is important to realize that steady-state concentrations (Cpss) depend on the rate of administration (k0) and drug clearance (CL), as determined by the following equation: Cpss ⫽ k0/CL. For a drug with a low extraction ratio, such as methadone, a decrease in fu will decrease CL, and consequently Cpss will be increased proportionally. However, steady-state unbound methadone concentrations (Cuss) will remain constant because Cuss ⫽ fu*Cpss. For both, Cp and Cu, the time to reach steady-state is not modified, because V and Cl both decrease proportionally with fu. Methadone biotransformation is mainly mediated by CYP3A4, although CYP2C9 and CYP2C19 also appear to take part, but to a much lesser extent (Foster et al., 1999). Interindividual variations in the expression of CYP3A4 will be the main factor responsible for interindividual variability in clearance. On the other hand, although methadone has been shown to be able to alter several CYP2D6 substrates metabolism, this enzyme does not seem catalyze the biotransformation of methadone. CYP3A4 is an inducible enzyme. This characteristic contributes to making the understanding and prediction of individual clearances of methadone more difficult. RostamiHodjegan et al. (1999) reported a 3.5-fold increase in the total clearance between the first dose of methadone and steady-state doses in opiate users. This finding was explained by the capability of the drug to induce its own metabolism. On the basis of the preceding results, time from the start of treatment should be considered a covariate to be included in clearance. Another covariate that investigators should consider is the presence of concomitant medication with drugs such as carbamazepine, phenytoin, rifampicin, zidovudine, barbiturates, spironolactone, verapamil, diethylstilboestrol, and amitriptyline (Wolff et al., 1993). There are data in the literature showing that these drugs induce methadone elimination kinetics. On the other hand, in vitro studies have shown that fluoxetine inhibited, in about 50%, the formation of the main metabolite of methadone. Similar results were found with omeprazole and carbamazepine (Foster et al., 1999). Inducers of CYP3A4, such as rifampicin, also have been shown to be capable of up-regulating the production of AAG, increasing the plasma binding of basic drugs (Abramson & Lutz, 1986). This up-regulation phenomenon has been suggested to be implicated in the slight increase in AAG seen after long treatment with methadone (Rostami-Hodjegan et al., 1999). Although distribution characteristics in humans and small laboratory animals were found to be similar, this is not the case for drug elimination. Methadone elimination in rats has been reported to be nonrestrictive, with a mean clearance of 0.1 l/min/kg. Differences between species in the elimination process are probably due to differences in the expression of CYP isoforms in methadone metabolism.
63
2.4. Disposition After single doses, plasma drug concentrations of methadone-versus-time profiles have been frequently described by using a two-open compartmental model. However, Inturrisi et al. (1987) used a triexponential model to describe the kinetics in plasma for patients with chronic pain. Despite the pharmacokinetic model used to describe the time course of plasma concentrations, the reported estimates for the terminal halflife are similar: 26.8 h (Wolff et al., 1993) and 23 h (Inturrisi et al., 1987). Rostami-Hodjegan et al. (1999) reported a time dependency in the terminal half-life, which was explained by the ability of methadone to induce its own metabolism. It should be noted, however, that variations in clearance do not always have any effect on half-life, because this secondary parameter also depends on the volume of distribution. 2.5. Stereoselective pharmacokinetics The estimates of volume of distribution reported for R- and S-Met were 496.6 l and 289.1 l, respectively, in chronic pain patients (Kristensen et al., 1996). These same authors reported the clearance of R-Met and S-Met as 0.158 and 0.129 l/min, respectively. Schmidt et al. (1994) found that clearance for R-Met in beagle dogs also was slightly higher than that obtained for S-Met: 0.357 versus 0.316 l/min, respectively. Notwithstanding, in a recent study, Foster et al. (1999) found a lack of stereoselective metabolism in human microsomes. These results suggest that the differences found in plasma clearance between enantiomers could be due to differences in plasma protein binding. Eap et al. (1990) reported fu values for R-Met and S-Met of 0.13 and 0.10, respectively. Such difference, although small, can explain, at least in part, the observed differences in plasma clearance between the two enantiomers. Although the discussed results show differences in kinetics between enantiomers, on the basis of the estimates of volume of distribution and clearance, it can be assumed that changes in protein binding, body composition, metabolism, and other demographic data will affect both enantiomers in a similar way. Regarding absorption processes, Kristensen et al. (1996) reported no differences between enantiomers in lag time and bioavailability. It should be nonetheless mentioned that published data on the absorption of methadone enantiomers are scarce.
3. Pharmacodynamics 3.1. Basic pharmacodynamic properties Methadone is a synthetic opioid agonist with morphinelike properties, mostly used as a maintenance drug for opioid addicts. Notwithstanding, it has also proved to be a powerful analgesic in patients with malignant and postoperative pain. Methadone elicits its pharmacodynamic properties through binding to , ␦ and opioid receptors. Methadone-receptor binding has been characterized in in vitro
64
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
studies. The affinity constant (Ki) to receptors was reported to be 3.51 nM, whereas the Ki value for morphine is 1.41 nM. These results indicate that methadone has a lower affinity for this receptor than does morphine (Blake et al., 1997). The receptor is the principal factor responsible for most of the important opioid pharmacodynamic properties, because its activation produces analgesia, respiratory depression, physiological dependence, and tolerance. It is sometimes difficult to dissociate therapeutic effects from adverse effects (Ling et al., 1985; Matthes et al., 1996). Methadone is described as inducing rapid dependence in rats. The degree of such dependence can be quantified in terms of the incidence of specific withdrawal signs after naloxone administration (Pierce et al., 1996). However, in humans, methadone has shown a lower potential for abuse than morphine, methadone being the drug of choice in maintenance therapy programs for opioid users (Blake et al., 1997). Another important complication during the use of opioids is the development of tolerance. Tolerance can be defined as a decrease in the analgesic potency (or other opioid effect) of opioid agonists after prior treatment with the opioid (Chan et al., 1997). Although this phenomenon is well known in experimental and clinical pharmacology, its mechanism of action is still poorly understood. Several authors suggest that tolerance is due, at least in part, to a change in the expression and function of receptors (Yoburn et al., 1993). The use of selective agonist and antagonist ligands for the different opioid receptors has been proposed to be a suitable tool to investigate the molecular alterations in the opioid-receptor system after opioid exposure in both in vivo and in vitro studies. Different strategies can be used to study the development of tolerance and the capacity of different full agonists to produce such phenomenon. The use of an irreversible opioid antagonist such as -funaltrexamine (-FNA), which produces a dose-dependent inactivation of the -receptor population, provides information about the relative efficacy between full agonists (Pitts et al., 1998). Adams et al. (1990), in an in vivo rat study, found that a 5 g i.c.v. dose of -FNA was capable of reducing the maximum analgesic effect of morphine but not that of methadone. These results suggest that methadone has a higher intrinsic efficacy than morphine does. Cross-tolerance is a phenomenon that results in a diminished response to an opioid agonist after previous treatment with another opioid agonist. Paronis and Holtzman (1992) studied the development of tolerance to the analgesic effects of several opioid agonists, including methadone, after continuous infusion of morphine, meperidine, or fentanyl in rats. They found that the ED50 for methadone before and after morphine exposure did not change, suggesting that methadone is less susceptible to inducing tolerance than is morphine and that this property is inversely related to the drug efficacy. On the other hand, the in vivo description/ prediction of the development of tolerance phenomenon can also be explored by using pharmacokinetic/pharmacodynamic models (Ekblom et al., 1993; Mandema & Wada, 1995; Oullet & Pollack, 1995).
3.2. Stereoselective pharmacodynamics R-Met exhibits a 10-fold higher affinity for the and ␦ opioid receptors and as much as a 50 times higher analgesic activity than S-Met does in human and animal models of pain (Foster et al., 1999). R-Met has been shown to prevent the occurrence of opioid withdrawal symptoms, whereas S-Met appears to be inactive as an opioid. Nonetheless, S-Met is active as an NMDA receptor antagonist, blocking morphine tolerance after systemic and intrathecal administration and NMDA-induced hyperalgesia. These results suggest that S-Met could be used alone or in combination with morphine to treat the hyperalgesic component of neuropathic pain, as well as to improve the analgesic efficacy of chronic morphine administration (Davis & Inturrisi, 1999).
4. Pharmacokinetic/pharmacodynamic relationships Pharmacokinetic/pharmacodynamic (PK/PD) relationships allow the description/prediction and characterization of the time course of the in vivo drug effect. There are several studies dealing with such relationships for opioids commonly used in anesthesia, such as alfentanyl, fentanyl, sufentanyl, and remifentanyl (Cox et al., 1999; Mandema & Wada, 1995). In regard to opioids used for the treatment of pain, studies of PK/PD relationships are less frequent, being available mainly for morphine (Ekblom et al., 1993; Gårdmark et al., 1993; Oullet & Pollack, 1995). A priori, some issues should be taken into consideration when the goal is to establish the PK/PD relationship for an opioid drug. It should be noted that the plasma is not the effect site and that usually a certain time is needed before the drug distributes into the central nervous system. This phenomenon has important implications, because Cp cannot be used directly to describe the observed effect. To overcome this difficulty, the effect-versus-time data can be analyzed by using the “effect-compartment model,” or “link model.” This model has frequently been used in PK/PD modeling for opioid drugs (Ekblom et al., 1993; Gårdmark et al., 1993). Another important factor is the possibility for tolerance development, which means that the pharmacodynamic relationship found after a single drug administration will not be adequate to represent the dynamics of the drug after continuous exposure. In addition to the “link model,” different models have been proposed in recent years to describe the development of tolerance to opioids (Ekblom et al., 1993; Gårdmark et al., 1993; Oullet & Pollack, 1995). Unfortunately, the adequate characterization of both biophase distribution and eventual tolerance development requires complex experimental designs. It is recommended to use of a variety of doses, administered by different routes and varying infusion lengths. Experimental designs are even more complicated for drugs with one or more active metabolites, as is the case with morphine and remifentanyl (Cox et al., 1999). To our knowledge, only four studies address the PK/PD
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
relationship of methadone. Inturrisi et al. (1987; 1990) studied patients with chronic pain. In the first study (Inturrisi et al., 1987), they administered methadone as a single 10–30 mg i.v. bolus dose to eight patients who had been receiving opioid treatment other than methadone. In addition to plasma drug concentrations, they determined the analgesic effect by means of a visual analog scale, blood pressure, pulse rate, respiratory rate, and pupil size. However, only the analgesic response was subjected to PK/PD modeling. The experimental design allowed an estimation of ke0, the first-order rate constant governing the distribution between plasma and biophase, and C50, the steady-state methadone concentration eliciting 50% of maximum pain relief. The mean estimated values were 0.19 l/min and 0.29 g/ml for ke0 and C50, respectively. Both parameters showed high interindividual variability, probably because of different grades of cross-tolerance between patients. A previous mean estimate of C50 for methadone, obtained after administration to relatively nontolerant subjects, was 0.06 g/ml (Gourlay et al., 1984). In the second study, Inturrisi et al. (1990) also studied the PK/PD relationship for methadone in patients with chronic pain who had previous exposure to opioids other than methadone. However, in this study, methadone was given as an infusion, rather than as a bolus. The drug was continuously infused for 3 to 4.5 h, and drug plasma concentration, pain relief, and sedation were measured during the infusion and 4 to 5 h after the end of the infusion. Pain relief and sedation data were reported by each subject by using visual analog scales. The mean estimates of the pharmacodynamic parameters were very similar for the two measured effects. C50 was 0.36 and 0.34 g/ml, and ␥ (the slope of the effect against steady-state drug concentration curve) was 4.4 and 5.8 for pain relief and sedation, respectively. Both parameters showed high interindividual variability. It should be noted that the estimate of C50 for pain relief was very similar to that obtained in the previous study with the use of i.v. bolus administration (Inturrisi et al., 1987), indicating that both designs seem to be adequate to estimate the dynamic properties of methadone. However, for almost 50% of the patients, an estimate of ke0 could not be obtained for pain relief and sedation. These results, although they could be interpreted in principle as a very rapid equilibrium between plasma and biophase, indicate that the design did not allow an adequate characterization of the plasma to biophase distribution kinetics. On the other hand, the long-infusion design is appropriate for identifying tolerance. As constant plasma concentration is maintained and plasma-versusbiophase equilibrium is achieved, a decrease in the effect can only be attributed to tolerance. No tolerance development was seen for methadone in this case. Dyer et al. (1999) studied the methadone PK/PD relationship in methadone-maintenance patients. Methadone plasma concentrations were directly related to pain relief, pupil size, and withdrawal score data by using a sigmoidal Emax model. In this study, 18 patients received constant
65
doses of methadone once daily for at least 2 months. In the PK/PD analysis, two major assumptions were made: (1) concentrations of methadone in plasma and brain were in equilibrium (because all the patients were at steady state) and (2) tolerance to effects was present but stable. Estimates of C50 for withdrawal scores, pupil size, and pain relief were 0.3, 0.28, and 0.18 g/l, respectively, and were also in the range of the previously reported estimates of C50 for pain relief and sedation (Inturrisi et al., 1987, 1990). Garrido et al. (1999) designed an experiment to assess the implications of the abstinence syndrome in the PK/PD relationships for methadone in rats. To induce physical dependence, a morphine solution was delivered continuously at a rate of 0.5 l/h for 6 days with the use of minipumps implanted under the skin. The spontaneous withdrawal syndrome emerged after removal of the minipumps. At the time of maximal withdrawal syndrome, an i.v. 0.35 mg/kg bolus dose of methadone was administered. Plasma and antinociception (measured by the tail-flick test) were simultaneously determined. An effect-compartment model was used to describe the time course of the antinociceptive effect. An increase in C50, from 24.2 (control rats) to 33.6 ng/ ml (abstinent rats) was observed. A cross-tolerance phenomena was suggested as responsible for the increase in C50. 5. Concluding remarks Methadone shows some peculiarities in its pharmacokinetics and pharmacodynamics, which should be taken into consideration if a prospective PK/PD study has to be designed. Regarding pharmacokinetics, weight, sex, and, if possible, AAG levels in plasma should be part of the known individual demographics. In addition, owing to the fact that CYP3A4 is the principal enzyme in the metabolism of methadone as well as other numerous drugs, data about concomitant medications are required. Time after starting treatment with methadone also can be considered a potential covariate because the drug has been shown to be able to induce its own metabolism. Pharmacodynamic properties of methadone are determined by its affinity for the -opioid receptor, which implies tolerance development, cross-tolerance, and dependence. Information regarding previous exposure to other opioids will be useful. Moreover, designs for drug administration should be devised to achieve a proper characterization of both plasma-versus-biophase distribution and development of tolerance processes. Because the pharmacokinetic and pharmacodynamic properties of methadone enantiomers are different, the use of an enantioselective technique for the determination of the drug in plasma also is recommended. References Abramson, F. P. (1982). Methadone plasma protein binding: alterations in cancer and displacement from ␣1-acid glycoprotein. Clin Pharmacol Ther 32, 652–658.
66
M.J. Garrido, I.F. Trocóniz / J Pharmacol Toxicol 42 (1999) 61–66
Abramson, F. P., & Lutz, M. P. (1986). The kinetics of induction by rifampicin of ␣1-acid glycoprotein and antipyrine clearance in the dog. Drug Metab Dispos 14, 46–51. Adams, J. U., Paronis, C. A., & Holtzman, S. G. (1990). Assessment of relative intrinsic activity of mu-opioid analgesics in vivo using -funaltrexamine. J Pharmacol Exp Ther 255, 803–808. Änggård, E., Gunne, L. M., Holmstrand, J., McMahon, R. E., Sandberg, C. G., & Sullivan, H. R. (1975). Disposition of methadone in methadone maintenace. Clin Pharmacol Ther 17, 258–266. Blake, A. D., Bot, G., Freeman, J. C., & Reisine, T. (1997). Differential opioid agonist regulation of the mouse opioid receptor. J Biol Chem 272, 782–790. Calvo, R., Aguirre, C., Troconiz, I. F., López, J., & Garrido, M. J. (1996). Alpha1-acid glycoprotein and serum protein binding of methadone in heroin addicts during withdrawal. Proceedings of the Sixth World Congress on Clinical Pharmacology and Therapeutics, Buenos Aires, Argentina, p. 174. Cox, E. H., Langemeijer, M. W. E., Gubbens-Stibbe, J. M., Muir, K. T., & Danhof, M. (1999). The compartive pharmacodynamic of remifentanyl and its metabolite, GR90291, in a rat electroencephalographic model. Anesthesiology 90, 535–544. Chan, K. W., Duttory, A., & Yoburn, B. C. (1997). Magnitude of tolerance to fentanyl is independent of -opioid receptor density. Eur J Pharmacol 319, 225–228. Davis, A. M., & Inturrisi, C. E. (1999). D-Methadone blocks morphine tolerance and N-methyl-D-aspartate-induced hyperalgesia. J Pharmacol Exp Ther 289, 1048–1053. Dole, V., & Nyswander, M. (1965). A medical treatment for diacetylmorphine (heroin) addiction. JAMA 193, 646–650. Dyer, K. R., Foster, D. J. R., White, J. M., Somogyi, A. A., Menelaou, A., & Bochner, F. (1999). Steady-state pharmacokinetics and pharmacodynamics in methadone maintenance patients: comparison of those who do and do not experience withdrawal and concentration-effect relationships. Clin Pharmacol Ther 65, 685–694. Eap, C. B., Cuendet, C., & Baumann, P. (1990). Binding of d-methadone, l-methadone and dl-methadone to proteins in plasma of healthy volunteers: role of the variants of ␣1-acid glycoprotein. Clin Pharmacol Ther 47, 338–346. Ekblom, M., Hammarlund-Udenaes, M., & Paalzow, L. (1993). Tolerance development and rebound effects to morphine in relation to different administrations in rats. J Pharmacol Exp Ther 266, 244–252. Foster, D. J. R., Somogyi, A. A., & Bochner, F. (1999). Methadone N-demethylation in human liver microsomes: lack of stereselectivity and involvement of CYP3A4. Br J Clin Pharmacol 47, 403–412. Gabrielsson, J. L., Johansson, P., Bondesson, U., & Paalzow, L. K. (1985). Analysis of methadone disposition in the pregnant rat by means of physiological flow model. J Pharmacokinet Biopharm 13, 355–372. Gårdmark, M., Eblom, M., Bouw, R., & Hammarlund-Udenaes, M. (1993). Quantification of effect delay and acute tolerance development to morphine in the rat. J Pharmacol Exp Ther 267, 1061–1067. Garrido, M. J., Jiménez, R., Gómez, E., & Calvo, R. (1996). Influence of plasma protein binding on analgesic effect on methadone in rats with spontaneous withdrawal. J Pharm Pharmacol 48, 281–284. Garrido, M. J., Valle, M., Calvo, R., & Trocóniz, I. F. (1999). Altered plasma and brain disposition and pharmacodynamics of methadone in abstinent rats. J Pharmacol Exp Ther 288, 179–187. Gómez, E., Martínez-Jordá, R., Suárez, E., Garrido, M. J., & Calvo, R. (1995). Altered methadone analgesia due to changes in plasma protein binding: role of the route of administration. Gen Pharmacol 26, 1273–1276. Gourlay, G. K., Willis, R. J., Wilson, P. R. (1984). Postoperative pain control with methadone: influence of supplementary methadone doses and blood concentration-response relationships. Anesthesiology 61, 19–26. Inturrisi, C. E., Colburn, W. A., Kaiko, R. F., Houde, R. W., & Foley, K. M. (1987). Pharmacokinetics and pharmacodynamics of methadone in patients with chronic pain. Clin Pharmacol Ther 41, 392–401. Inturrisi, C. E., Portenoy, R. K., Max, M. B., Colburn, W. A., & Foley, K. M. (1990). Pharmacokinetic-pharmacodynamic relationships of metha-
done infusions in patients with cancer pain. Clin Pharmacol Ther 47, 565–577. Kristensen, K., Blemmer, T., Angelo, H. R., Christrup, L. L., Drenck, N. E., Rasmussen, S. N., & Sjøgren, P. (1996). Stereoselective pharmacokinetics of methadone in chronic pain patients. Ther Drug Monitor 18, 221–227. Ling, G. S. F., Spiegel, K., Lockhart, S. H., & Pasternak, G. W. (1985). Separation of opioid analgesia from respiratory depression: evidence for different receptor mechanisms. J Pharmacol Exp Ther 232, 149–155. Mandema, J. M., & Wada, D. R. (1995). Pharmacodynamic model for acute tolerance development to the electroencephalographic effects of alfentanyl in the rat. J Pharmacol Exp Ther 275, 1185–1194. Marathe, P. H., Shen, D. D., Artru, A. A., & Bowdle, A. (1991). Effect of serum protein binding on the entry of lidocaine into brain and cerebrospinal fluid in dogs. Anesthesiology 75, 804–812. Matthes, H. W. D., Maldonado, R., Simonin, F., Valverde, O., Slowe, S., Kitchen, I., Befort, K., Dierich, A., Le Meur, M., Dollé, P., Tzavara, E., Hanoune, J., Roques, B. P., & Kieffer, B. L. (1996). Loss of morphineinduced analgesia, reward effect and withdrawal symptoms in mice lacking the -opioid-receptor gene. Nature 383, 819–823. Nilsson, M. J., Widelöv, E., Meresaar, U., & Änggård, E. (1982). Effect of urinary pH on the disposition of methadone in man. Eur J Clin Pharmacol 22, 337–342. Olsen, G. D. (1973). Methadone binding to human plasma proteins. Clin Pharmcol Ther 14, 338–343. Oullet, D. M., & Pollack, G. M. (1995). A pharmacokinetic-pharmacodynamic model of tolerance to morphine analgesia during infusion in rats. J Pharmacokinet Biopharm 23, 531–549. Paronis, C. A., & Holtzman, S. G. (1992). Development of tolerance to the analgesic activity of mu agonists after continuous infusion of morphine, meperidine or fentanyl in rats. J Pharmacol Exp Ther 262, 1–9. Pierce, T. L., Hope, W., & Raper, C. (1996). The induction and quantitation of methadone dependence in the rat. J Pharmacol Toxicol Methods 36, 137–146. Pitts, R. C., Allen, R. M., Walker, E. A., & Dykstra, L. A. (1998). Clocinamox antagonism of the antinociceptive effects of mu opioids in squirrel monkeys. J Pharmacol Exp Ther 285, 1197–1206. Ripamonti, C., Zecca, E., & Bruera, E. (1997). An update on the clinical use of methadone for cancer pain. Pain 70, 109–115. Romanch, M. K., Piafsky, K. M., Abel, J. G., Khouw, V., & Sellers, E. M. (1981). Methadone binding to orosomucoid (␣1-acid glycoprotein): determinant of free fraction in plasma. Clin Pharmacol Ther 29, 211–217. Rostami-Hodjegan, A., Wolff, W., Hay, A. W. M., Raistrick, D., & Calvert, R. (1999). Population pharmacokinetics of methadone in opiate users: characterization on time-dependent changes. Br J Clin Pharmacol 48, 43–52. Rowland, M., & Tozer, T. N. (1995). Physiologic concepts and kinetics: distribution. In M. Rowland & T. N. Tozer (Eds.), Clinical Pharmacokinetics: Concepts and Applications, 3rd ed. (pp. 137–183). Media, PA: William & Wilkins. Säwe, J. (1986). High-dose morphine and methadone in cancer patients: clinical pharmacokinetic considerations of oral treatment. Clin Pharmacokinet 11, 87–106. Schmidt, N., Brune, K., Williams, K. M., & Geisslinger, G. (1994). Stereoselective pharmacokinetics of methadone in beagle dogs. Chirality 6, 492–495. Vigano, A., Fan, D., & Bruera, E. (1996). Individualized use of methadone and opioid rotation in the comprehensive management of cancer pain associated with poor prognostic indicators. Pain 67, 115–119. Wilkins, J. N., Ashofteh, A., Setoda, D., Wheatley, W. S., Huigen, H., & Ling, W. (1997). Ultrafiltration using the Amicon MPS-1 for assessing methadone plasma protein binding. Ther Drug Monit 19, 83–87. Wolff, K., Hay, A., & Raistrick, D. (1991). High-dose methadone and the need for drug measurements in plasma. Clin Chem 37, 1651–1664. Wolff, K., Hay, A. W. M., Raistrick, D., & Calvert, R. (1993). Steady-state pharmacokinetics of methadone in opioid addicts. Eur J Clin Pharmacol 44, 189–194. Yoburn, B. C., Barbara, B., & Duttaroy, A. (1993). Opioid receptor regulation in mice. J Pharmacol Exp Ther 265, 314–317.