Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Individualization of drug dosage is central to optimization of therapeutic effects and limitation of adverse drug reactions. Factors accounting for interindividual variability in drug effects include age, disease, and genetic makeup. Yet despite observations that drug dosage varies greatly in different parts of the world, the concept that ethnicity is also an important determinant of drug metabolism and responsiveness has only recently been recognized (Wood and Zhou, 1991; Kallow, 1991). Thus, although pharmacokinetic or pharmacodynamic data obtained in one population may not always apply to another, information comparing drug effects among different populations is sparse.

Ethnic differences in drug metabolism have four potential explanations: 1) the incidence of poor metabolizers (PMs) of polymorphically distributed drug metabolism varies in different ethnic populations [for example, the incidence of PMs of debrisoquin cytochrome P-450 (CYP) 2D6 is 5 to 7% in Caucasians but less than 1% in Chinese (Horai et al., 1989; Alvan et al., 1990; Bertilsson et al., 1992)], 2) the frequency distribution of the debrisoquin metabolic ratio (DMR), a measure of CYP2D6 activity in vivo, is shifted in Chinese toward higher values than in Caucasians, implying that among Chinese extensive metabolizers (EMs), the activity of CYP2D6 is lower (Bertilsson et al., 1992), 3) in Caucasians but not in Chinese, the metabolism of CYP2C19 substrates, mephenytoin and diazepam, cosegregates, indicative of possible interethnic variability in substrate specificity (Zhang et al., 1990), and 4) we have recently shown that the extent of diazepam inhibition by omeprazole, a CYP2C19 substrate and inhibitor, is ethnically dependent, being greater in Caucasians than in Chinese (Caraco et al., 1995).

Codeine is an ancient drug that exerts its effect through the production of the active metabolite, morphine (MOR; Snafilippo, 1948). The O-demethylation of codeine to produce MOR is a minor pathway accounting for about 5% of codeine's biotransformation, whereas its glucuronidation to codeine-6-glucuronide (C6G) and its N-demethylation to norcodeine (NC) account for approximately 80 to 85% and 10%, respectively (Yue et al., 1991a; Fig. 1). MOR can be further metabolized by glucuronidation to produce morphine-3-glucuronide (M3G) and morphine-6-glucuronide (M6G), and N-demethylation to normorphine (NM) and NC in turn can be glucuronidated to norcodeine-glucuronide (NCG). The production of MOR from codeine is mediated through the activity of the polymorphically distributed enzyme CYP2D6 and therefore occurs only in EMs of debrisoquin (Yue et al., 1989a; Mortimer et al., 1990). Thus, in Caucasians, codeine's respiratory, psychomotor, and miotic effects are significantly greater in EMs than PMs. Furthermore, the coadministration of quinidine, a potent CYP2D6 inhibitor, markedly inhibits MOR's production from codeine in EMs and completely abolishes the interphenotypic differences in codeine pharmacodynamic effects (Caraco et al., 1996a).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Chemical structure of codeine and the major metabolic routes involved in its biotransformation.

In keeping with the previously noted reduced CYP2D6 activity, codeine O-demethylation to MOR has been shown to be lower in Chinese EMs than in their counterpart Caucasian EMs (Yue et al., 1989b, 1991b). In addition, we have shown previously that Chinese are less sensitive to the respiratory depressant effect of MOR (Zhou et al., 1993). Thus the present study was undertaken to compare codeine's pharmacokinetics and effects, with and without quinidine coadministration, in Caucasians and Chinese EMs of debrisoquin.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. Eighteen healthy nonsmoking male subjects, ten Caucasians and eight Chinese EMs of debrisoquin, participated in the study. None were using any medications, including alcohol on a regular basis, and they were told to avoid any drug intake for 1 week before study initiation and throughout the entire study period. The Caucasian subjects participated in the study while taking part in another study comparing codeine's effect between EMs and PMs of debrisoquin (Caraco et al., 1996a). The Chinese subjects were students at Vanderbilt University (Nashville, TN) who had left mainland China 3.1 ± 0.5 (mean ± S.E.M.) years before study enrollment. The subject's weight in both ethnic groups was within 20% of their respective ideal body weight, but mean weight was significantly greater in the Caucasians than in the Chinese (78.7 ± 2.1 kg versus 66.1 ± 3.7 kg, respectively, p < .01) with no significant differences in mean body mass index (25.3 ± 0.8 kg/m² versus 22.9 ± 0.9 kg/m², respectively, p > .3) and age (32.1 ± 1.1 years versus 31.3 ± 2.4 years, respectively, p > .3; Table 1). All subjects were judged to be in good health based on their medical history, a physical examination, and routine laboratory tests. The study protocol was approved by the Vanderbilt University Hospital Committee for the Protection of Human Studies and all subjects gave written informed consent.

Study Design. The study was designed as a double blind crossover study consisting of two study days separated by a week. On the morning of each study day and after an overnight fast the subjects received in random order placebo plus 120 mg codeine phosphate (codeine-alone study day) or 100 mg quinidine plus 120 mg codeine phosphate (quinidine plus codeine study day). These treatments were given as identical looking capsules that were marked as "first" or "second" so that the first capsule was given in each study day 0.5 h before the administration of the second capsule (i.e., placebo before codeine and quinidine before codeine). Food was not permitted for the first 6 h and standardized meals were provided 6 and 10 h after codeine administration. The subjects were instructed to remain in the study room throughout the entire day and physical activity was not allowed during study sessions. Blood (5 ml) was collected into heparinized tubes just before the administration of the study treatment, at 10-min intervals during the first hour and 1.5, 2, 3, 4, 5, 6, 7, 8, 10, 12, and 24 h after the second capsule intake. Blood samples were drawn through an indwelling i.v. catheter except for the 24-h sample, which was taken by a separate venipuncture. Plasma was immediately separated and kept frozen at 20°C until analysis. Urine was collected over 48 h at three intervals: 0 to 12 h, 12 to 24 h, and 24 to 48 h. The volume of each urine collection was recorded and an aliquot was frozen at 20°C until analyzed.

Analytical Assays. The concentration of codeine and codeine's metabolites in blood and urine were measured by an ion-paired HPLC method as described previously with minor modifications (Caraco et al., 1996a). This method involves solid phase extraction (Sep-pak C₁₈, Waters Associates, Framington, MA); UV detection (Spectroflow 773 UV detector; Kratos Analytical Instruments, Ramsey, NJ) of M3G, NCG, C6G, NC, and codeine; and coulometric electrochemical detection (Coulochem 5100A with 5021 conditioning cell and 5011 analytical cell; ESA, Inc., Bedford, MA) of M6G, NM, and MOR. The chromatographic apparatus consisted of a 6000 A pump, two 730 data modules, a WISP 710 A autoinjector and µBondapack C₁₈, 10 µm, 300 × 3.9 mm column (Waters Associates). The limit of detection was 2 ng/ml for M6G, NM, and MOR and 10 ng/ml for M3G, C6G, NC, and codeine. The interday and intraday coefficients of variation were less than 10 and 13%, respectively. The concentration of quinidine in the plasma was determined by fluorescence polarization immunoassay (TD_x/TD_x FL_x, Abbott Laboratories, Chicago, IL). The assay that also measures dihydroquinidine (less than 10% of quinidine concentration) has a lower detection limit of 0.2 µg/ml.

Pharmacodynamic Evaluation. Codeine's effect on respiration, psychomotor function, and pupil diameter were determined on each study day at 0.5, 1, 1.5, 2, 3, 4, 5, and 6 h after the intake of the second study capsule. Three pharmacodynamic measurements were performed before the administration of the first study capsule and the average was taken as the baseline value for that particular study day.

Measurement of Effect on Respiration. The effect on respiration was evaluated by measuring resting minute ventilation, end-tidal CO₂ content, and the ventilatory response to rising concentration of carbon dioxide by using the rebreathing method of Read (1967). In brief, the subjects were requested to breath through a mouthpiece while wearing a nose clip until minute ventilation and end-tidal CO₂ were stable. Then they were connected through a 3-way valve to a balloon containing a volume of 1.5 times their vital capacity of a gas mixture of 93% oxygen and 7% carbon dioxide. With rebreathing, the increasing concentration of CO₂ stimulated progressive hyperventilation, which was terminated after about 2 min once the end-tidal CO₂ had reached the value of 60 mm Hg. Tidal volume, ventilatory rate, and end-tidal CO₂ were measured continuously at 10-s intervals by a computerized exercise module (Cybermedics, Boulder, Co). To familiarize the subjects with the procedure, a single rebreathing session was performed 1 week before the first study day as a part of the screening process.

Measuring Effect on Pupil Diameter. A 35-mm camera (Nikon FG) equipped with a micro-Nikonlens and a ring-flash (Nikon 5B-21B), and attached to a chin-head rest was used to photograph the subject's left eye from a fixed distance. Light conditions were kept constant and monitored by a light-meter (Minolta autometer 2). The pupil and the iris diameters in their largest axis were measured from a 5- × 7-inch color print by using a ruler and a caliper. To avoid introducing an error that might have occurred due to a slight movement of the eye relative to the camera lens, the pupil-iris diameter ratio was used for pharmacodynamic evaluation. The coefficient of variation for repeated calculations of the ratio of the pupil diameter to the iris diameter from duplicates of the same print was 2.5%.

Measuring Effect on Psychomotor Function. The effect on psychomotor function was evaluated periodically during each study day by a revised version of the digit symbol substitution test (DSST; Wechsler, 1981). The subjects were given 90 s to fill in as many empty boxes with the appropriate letters according to a figure-letter code that was provided on the top of each form. The number of correctly filled boxes was considered as the score and used for comparison. To minimize the effect of learning on the DSST results, the subjects did the test twice one week before the first study day and different versions were used at different times throughout each study day.

Data Analysis. Plasma concentrations of codeine and codeine's metabolites were plotted semilogarithmically against time and the respective elimination rate constants () were derived by least-squares regression analysis of the terminal phase. The areas under the plasma concentration-time curves (AUCs) were calculated by the log-trapezoidal rule and extrapolated to infinity (AUC₀). T_1/2 was calculated based on the equation T_1/2 = 0.693/. Codeine was considered to be 100% bioavailable and the dose of 120 mg codeine phosphate corresponded to 103 mg of codeine base (Bechtel and Sinterhauf, 1978). Apparent oral codeine clearance (CL_o) was calculated as Dose/AUC₀. The partial metabolic clearances of codeine were calculated by using the following equations:

0

06

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The description of codeine's pharmacokinetics and pharmacodynamics in Caucasians has been published previously (Caraco et al., 1996a). Out of the ten Caucasian subjects enrolled, only nine were available for analysis because one subject decided not to participate in the second study day after experiencing marked dizziness and nausea during the first study day.

Pharmacokinetics: Codeine Alone Study Day. The concentrations of codeine and metabolites are shown in Figs. 2 and 3. Codeine AUC₀ was significantly lower in the Caucasians compared with the Chinese (3160 ± 286 nM · h versus 5159 ± 783 nM · h, respectively, p < .03; Table 2). On the other hand, the concentration of the N-demethylated metabolites NM and NCG (AUC₀) were significantly higher in the Caucasians than in the Chinese (606 ± 84 nM · h versus 327 ± 60 nM · h, p < .02 and 2164 ± 243 nM · h versus 950 ± 443 nM · h, p < .03, respectively). No significant interethnic differences were noted in the concentrations (AUC₀) of the O-demethylated metabolites (MOR, M3G, M6G), C6G, or NC.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Mean (±S.E.M.) codeine, C6G, NCG, and NC concentrations over time after the administration of placebo plus codeine. A, Caucasians, closed symbols; B, Chinese, open symbols.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Mean (±S.E.M.) MOR, M3G, M6G, and NM concentrations over time after the administration of placebo plus codeine. A, Caucasians, closed symbols; B, Chinese, open symbols.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Mean (±S.E.M.) codeine (C) CLo and partial metabolic clearances by glucuronidation, O-demethylation, and N-demethylation after the administration of placebo (P) plus C () and quinidine (Q) plus C ().

Pharmacokinetics: Codeine Plus Quinidine Study Day. Quinidine could be identified in the plasma samples obtained from all subjects and no significant difference was noted in its AUC₀ between the Caucasian and Chinese subjects (4.18 ± 0.73 µg · h/ml versus 2.84 ± 0.43 µg · h/ml, respectively, p > .1; Fig. 5). The administration of quinidine before codeine significantly diminished the production of codeine's O-demethylated metabolites in both ethnic groups and thus MOR and MOR metabolites could no longer be detected in the plasma samples obtained from any of the subjects. Codeine's metabolic clearance by O-demethylation was significantly reduced by the administration of quinidine in both the Caucasian (from 162.7 ± 36.6 ml/min to 17.0 ± 5.0 ml/min, p < .004) and the Chinese (from 52.7 ± 12.7 ml/min to 5.92 ± 3.34 ml/min, p < .01) subjects. The absolute reduction in codeine clearance by O-demethylation was significantly greater in the Caucasian than in the Chinese subjects (115.8 ± 25.9 ml/min versus 46.8 ± 10.6 ml/min, respectively, p < .03); however, no significant interethnic difference was noted when the change was expressed as percentage of baseline (88.8 ± 9.0 versus 90.7 ± 3.7%, respectively, p > .2). Nevertheless, even after quinidine administration, codeine metabolic clearance by O-demethylation was still greater in the Caucasians (p < .05; Fig. 4). The reduction in codeine's metabolic clearance by the O-demethylation pathway was associated with a significant decrease in the urinary recovery of MOR and MOR metabolites in both Caucasian (from 8.68 ± 1.87% to 1.10 ± 0.31%, p < .003) and Chinese subjects (from 5.00 ± 1.50% to 0.53 ± 0.36%, p < .01; Table 3).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Mean (±S.E.M.) quinidine plasma concentrations over time in Caucasian () and Chinese () subjects.

0

Pharmacodynamics: Codeine-Alone Study Day. The administration of codeine was associated in both ethnic groups with ventilatory, psychomotor, and miotic effects. Peak effect was usually noted within the initial 2 h after codeine administration and the effect was still evident 6 h post intake (Fig. 6). The magnitude of codeine's respiratory depressant effect was significantly greater in the Caucasian than in the Chinese, resulting in lower AUE₀₆ values for VE₅₅ (p < .05) and the slope of the CO₂ response curve (p < .04) and higher AUE₀₆ values for end tidal CO₂ (p < .02). No significant interethnic differences in the effect of codeine were noted for resting minute ventilation, psychomotor performance as evaluated by DSST, or pupillary constriction (p > .3; Fig. 6). The extent of codeine effect at fixed MOR concentrations (i.e., 1.25, 2.5, 5, 10, 12.5, 15, 20, and 25 nM) was significantly greater in the Caucasian than in the Chinese subjects for the slope of the CO₂ response curve (ANOVA; p < .001), VE₅₅ (ANOVA; p < .006), end tidal CO₂ (ANOVA; p < .006), psychomotor performance as measured by the DSST (ANOVA; p < .001), and pupillary ratio (ANOVA; p < .001; Fig. 7). In addition, when the subjects in both groups were analyzed together, CYP2D6 activity as evaluated by the DMR was significantly correlated with codeine effect on the slope of the CO₂ response curve (r = 0.57, p < .02), VE₅₅ (r = 0.52, p < .03), psychomotor performance evaluated by the DSST (r = 0.54, p < .02), and pupillary diameter (r = 0.50, p < .04).

View larger version (45K): [in this window] [in a new window]   Fig. 6.   Mean (±S.E.M.) percentage of baseline effect over time after the administration of placebo (P) plus codeine (C) in Caucasians () and Chinese (). A, resting minute ventilation; B, resting end-tidal carbon dioxide concentration (PETco2); C, VE55; D, slope of the CO2 response curve; E, DSST; F, pupil ratio. AUE06 after the administration of C alone in Caucasians () and Chinese () and after the administration of C plus quinidine (Q) () is shown at right.

View larger version (35K): [in this window] [in a new window]   Fig. 7.   Mean (±S.E.M.) percent baseline effect after the administration of placebo plus codeine at different MOR concentrations in Caucasian () and Chinese () subjects. A, resting minute ventilation; B, resting end-tidal carbon dioxide concentration (PETco2); C, VE55; D, slope of the CO2 response curve; E, DSST; F, pupil ratio.

Pharmacodynamics: Quinidine Plus Codeine Study Day. Quinidine administration was associated in the Caucasians with a marked decrease in codeine's effect on the slope of the CO₂ response curve (p < .001), VE₅₅ (p < .005), resting minute ventilation (p < .01), end tidal CO₂ (p < .01), psychomotor function as evaluated by the DSST (p < .001), and pupil diameter (p < .005; Fig. 6). In contrast, quinidine administration did not significantly alter codeine's effects in the Chinese subjects (p > .3; Fig. 6). Thus, the reduction in the effect of codeine caused by quinidine was significantly greater in the Caucasians than in the Chinese subjects (Fig. 8; slope of the CO₂ response curve: 121.9 ± 15.7% versus 23.5 ± 30.6% · h, p < .01), VE₅₅ (100.7 ± 26.6% versus 9.3 ± 15.7% · h, p < .01), end tidal CO₂ (22.3 ± 6.4% · h versus 0.1 ± 3.5% · h, p < .01), and psychomotor function as evaluated by the DSST (55.9 ± 7.9% · h versus 18.5 ± 8.1% · h, p < .005). In addition, when the Caucasian and the Chinese subjects were analyzed together, the decrease in codeine's effects caused by quinidine coadministration was significantly correlated with the effect of codeine on the codeine-alone study day [slope of the CO₂ response curve: r = 0.64, p < .006; VE₅₅: r = 0.76, p < .001 (Fig. 9), end tidal CO₂: r = 0.65, p < .004; psychomotor function as evaluated by the DSST: r = 0.91, p < .004; and pupillary ratio: r = 0.54, p < .02].

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Mean (±S.E.M.) decrease in the respiratory, psychomotor, and pupillary effects of codeine caused by the addition of quinidine in Caucasian () and Chinese () subjects. RV, resting minute ventilation; PETco2, resting end-tidal carbon dioxide concentration; PRatio, pupil ratio.

View larger version (21K): [in this window] [in a new window]   Fig. 9.   Correlation between the effect of codeine on VE55 and the decrease in codeine effect on VE55 caused by the addition of quinidine in Caucasian () and Chinese () subjects.

Adverse Effects. Sleepiness and dizziness were reported by all subjects on the codeine-alone study day. It usually peaked 2 h after codeine administration and by 6 h postintake was barely noticeable. Five of the Caucasians and seven of the Chinese subjects experienced variable degree of nausea after the administration of codeine plus placebo. In both Caucasian and Chinese subjects, adverse effects were less frequent and of lower intensity after the combined intake of codeine plus quinidine.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Opioids are extremely important drugs in the treatment of chronic pain syndrome (Eggen and Andrew, 1994). Several such compounds including codeine, oxycodone, and hydrocodone may be considered as pro-drugs as their effects are mediated, at least in part, through the production of an active metabolite (Chen et al., 1991; Otton et al., 1993a,b). It has become apparent in recent years that CYP2D6, a member of the cytochrome P450 superfamily, is the major enzyme involved in the bioactivation of these drugs (Yue et al., 1989a; Otton et al., 1993a,b). Thus, in subjects who are genetically deficient in CYP2D6 (PMs) or concurrently receiving CYP2D6 inhibitors such as quinidine, negligible amounts of the metabolites are produced, resulting in loss of pharmacological effect (Sindrop et al., 1991; Desmeules et al., 1991; Otton et al., 1993a,b; Caraco et al., 1996a). However, in a recent study, although quinidine inhibited oxycodone metabolism and eliminated oxymorphone production, no change in oxycodone's pharmacodynamic effects were seen (Heiskanen et al., 1998).

Debrisoquin hydroxylase activity is lower in Chinese EM subjects than in Caucasians counterparts (Bertilsson et al., 1992). This interethnic difference in CYP2D6 activity affects the metabolism of other CYP2D6 substrates such as codeine whose metabolism by O-demethylation to MOR is reduced in Chinese compared with Caucasian EMs (Yue et al., 1989b, 1991b). This reduction in CYP2D6 activity in Chinese has been attributed to the higher frequency of C₁₈₈/T and G₄₂₆₈/C CYP2D6 mutations in Chinese compared with Caucasian EMs (Tseng et al., 1996). More than 50% of Chinese EMs exhibit these polymorphisms and have lower formation rates of MOR from codeine (Tseng et al., 1996). Although the present study did not include CYP2D6 genetic analysis, it is likely that the increased frequency of CYP2D6 mutations associated with decreased CYP2D6 activity in Chinese subjects may explain the reduced MOR formation.

The respiratory depressant effect of codeine was lower in the Chinese compared with Caucasian subjects. Our data suggest that this reduced effect was due to both reduced production of and altered sensitivity to the effect of MOR. The lower clearance of codeine by O-demethylation to MOR and MOR metabolites in Chinese and the correlation between CYP2D6 activity (i.e., DMR) and codeine's effects demonstrate the role of altered MOR production. In addition, a trend toward lower plasma concentration of M6G was noted in Chinese compared with Caucasian subjects. Thus it is possible that the reduced respiratory depressant effect of codeine in Chinese reflects decreased production of M6G, a potent µ agonist. Furthermore, the central role played by M6G in mediating codeine pharmacodynamics is implied by the significant correlation found between M6G AUC₀ and codeine's respiratory effects as evaluated by VE₅₅ (r = 0.52, p < .04) and end tidal CO₂ (r = 0.60, p < .01). Altered sensitivity to MOR is indicated by the findings that MOR administered directly (Zhou et al., 1993) or MOR formed from codeine (present study) produced significantly lower effects in Chinese compared with Caucasian subjects at identical MOR plasma concentrations (Fig. 7).

Although interethnic differences in drug disposition and effect are now well recognized, interethnic differences in susceptibility to drug interactions are less well described. We have shown previously that the inhibition of diazepam and mephenytoin metabolism by omeprazole, a CYP2C19 substrate and inhibitor, is ethnically determined with greater inhibition occurring in Caucasian than in Chinese subjects (Caraco et al., 1995, 1996b). Interestingly, the potent inhibitor of CYP2D6, quinidine, decreased codeine O-demethylation in both ethnic groups but the extent of absolute reduction was significantly greater in Caucasians. The reason for such racial variability in CYP isoforms inhibition is not fully defined, but it may possibly reflect lower enzyme activity in the uninhibited state compared with Caucasians. The possibility that higher quinidine concentrations in the Caucasians might explain the difference was excluded by the absence of significant interethnic differences in quinidine plasma concentrations or AUC. An additional explanation may be ethnic differences in the affinity of the CYP isoforms for their inhibitor, perhaps reflecting the increased incidence of CYP2D6 mutations in Chinese.

Diminished production of MOR from codeine in subjects with genetically deficient or environmentally inhibited CYP2D6 has been shown in Caucasians to be associated with marked reduction in codeine's analgesic, respiratory, psychomotor, and miotic effects (Sindrop et al., 1991; Desmeules et al., 1991; Caraco et al., 1996a). However, in Chinese subjects, codeine pharmacodynamics was unaffected by quinidine intake. This may simply indicate an incomplete inhibition of codeine's O-demethylation in Chinese subjects, but the absence of detectable amounts of MOR and MOR metabolites in the plasma samples obtained after quinidine argues against such a hypothesis. Alternatively, the reduction in the concentration of the active metabolite would be expected to be associated with the greatest reduction in effect in those subjects with the most prominent initial effect. Such a hypothesis would fit the data showing a greater decrease in codeine effect after quinidine in Caucasians, in whom codeine baseline effect was more prominent initially and with the correlation noted between the decrease in codeine's effect after quinidine coadministration and the effect measured before quinidine administration.

In conclusion, ethnicity plays an important role in determining both the disposition and effects of codeine. Chinese subjects not only produced less MOR and MOR metabolites after codeine, but also exhibited reduced sensitivity to respiratory, psychomotor, and pupillary effects of the MOR that was formed. As the analgesic effect of opioids is correlated to their other effects such as respiratory depression and pupillary constriction (Fraser et al., 1954; Seed et al., 1958), we would predict that the analgesic effect of codeine and other opioids dependent on CYP2D6 bioactivation will be attenuated in Chinese compared with Caucasian EM subjects. Inhibition of CYP2D6 by quinidine and the resultant decrease in MOR production is ethnically dependent with a greater absolute inhibition in Caucasian than in Chinese subjects.