Impact of quinidine on plasma and cerebrospinal fluid concentrations of codeine and morphine after codeine intake

Objective: The analgesic effect of codeine depends on its O-demethylation to morphine via sparteine oxygenase (CYP2D6) in the liver and presumably also via this enzyme in the CNS. We studied the ability of quinidine, which is a potent inhibitor of CYP2D6, to penetrate the blood brain barrier and its pssible impact on codeine O-demethylation in CNS. Methods: The study comprised 16 extensive and one poor metaboliser of sparteine, who underwent spinal anaesthesia for urinary tract surgery or examination. Eight patients were given an oral dose of 125 mg codeine and 9 patients (including the poor metaboliser) were given 200 mg quinidine 2 h before the same dose of codeine. Plasma and spinal fluid samples were collected 2 h after codeine intake. Results: Free concentrations of quinidine were 11-times lower in cerebrospinal fluid than in plasma, and ranged from 9–15 nmol·l^–1. Morphine concentrations were significantly lower in patients pre-treated with quinidine, both in plasma (median 1.45 nmol·l^–1, range 0.74–1.95 nmol·l^–1 vs 9.86 nmol·l^–1, range 4.59–28.4 nmol·l^–1) and in cerebrospinal fluid (0.23, 0.16–0.61 nmol·l^–1 vs 3.63, 0.6–8.09 nmol·l^–1). The morphine/codeine concentration ratio in plasma (3.07×10^–3, 1.68–3.68×10^–3 vs 19.87×10^–3, 9.87–66.22×10^–3) and in cerebrospinal fluid (0.83×10^–3, 0.58–1.45×10^–3 vs 7.19×10^–3, 2.03–17.7×10^–3) was also lower. The morphine/-codeine concentration ratios were significantly lower in cerebrospinal fluid both without and with quinidine, but the difference between the plasma and spinal fluid ratios was significantly smaller with quinidine than without (p=0.0002). Conclusion: Quinidine penetrates the blood brain barrier poorly, but quinidine pre-treatment leads to pronounced lowering of the cerebrospinal fluid concentration of morphine after codeine intake. However, the O-demethylation of codeine in CNS may not be totally blocked by quinidine.     

Effect of Age on Distribution of Zidovudine (Azidothymidine) into the Cerebrospinal Fluid of Macaca nemestrina

The brain tissue is an important target for anti-HIV drug therapy. Since the permeability of the blood–brain and blood–cerebrospinal fluid (CSF) barriers may differ between neonates and adults, we have determined the effect of age on the distribution of zidovudine (ZDV or azidothymidine) into the CSF in the macaque (M. nemestrina). Five newborn macaques were administered ZDV (iv bolus, 5 mg/kg) at various ages (2 days to 4 months). Both CSF (cisternal) and venous blood samples were obtained at approximately 60 and 90 min after drug administration. In another series of experiments, adult female macaques received ZDV as either an iv bolus (5 and 10 mg/kg) or an infusion for at least 12 hr. CSF (lumbar) and venous blood samples were obtained at approximately 60 and 90 min after iv bolus and at more than 12 hr after iv infusion. ZDV concentration in the CSF and the plasma samples was determined by high-performance liquid chromatography. The CSF/plasma concentration ratio of ZDV in the newborn and adult macaques, after iv bolus administration, was independent of time. In addition, no significant (P > 0.05) difference was observed in the pooled iv bolus ZDV CSF/plasma concentration ratio between the adult group (0.236 ± 0.058) and the newborns (0.213 ± 0.039). Moreover, the ZDV CSF/plasma concentration ratio in the adults and the newborns, after iv bolus administration, was found not to be significantly (P > 0.05) different from the ratio obtained at steady state in the adults (0.224 ± 0.094). These data indicate that the distribution of ZDV into the CSF in macaque neonates and adults is similar.     

Interactions of naloxone with morphine,amphetamine and phencyclidine on fixed interval responding for intracranial self-stimulation in rats

Rats were implanted with stimulating electrodes aimed at the medial forebrain bundle-lateral hypothalamus (MFB-LH) and were trained to lever-press for brain self-stimulation on a fixed interval: 60 s schedule of reinforcement. The effects of graded doses of naloxone (0.1–30 mg/kg), morphine (0.3–5.6 mg/kg), naloxone plus morphine,d-amphetamine (0.03–1.0 mg/kg), naloxone plusd-amphetamine, phencyclidine (0.3–5.6 mg/kg), and naloxone plus phencyclidine were tested. Naloxone produced a significant decrease in rates at 30 mg/kg. Naloxone (0.1–1.0 mg/kg) plus morphine blocked the dose-dependent decrease produced by morphine alone. In contrast, naloxone (1.0–10 mg/kg) plusd-amphetamine attenuated the graded increase in response rates produced byd-amphetamine. Naloxone (1.0–10 mg/kg) plus phencyclidine did not reliably change the increase in response rates produced by phencyclidine alone. The use of the fixed interval schedule of brain self-stimulation to study these drug interactions is novel, and further demonstrates that the highly reinforcing aspects of brain stimulation, known to be influenced by dopamine, may also be modulated by the endogenous opiate system.     

The 5-hydroxytryptamine content of mouse brain and whole mice after treatment with some drugs affecting the central nervous system

A number of drugs were examined for their ability to change the concentration of 5-hydroxytryptamine in mouse brain and in whole mice treated with 5-hydroxytryptophan. After β-phenylisopropylhydrazine or iproniazid, two inhibitors of monoamine oxidase, the brain 5-hydroxytryptamine rose to a maximum value in 8 hr., after which it declined, although a slight rise remained for as long as 6 days. Dose-effect relationships, determined 6 hr. after administration, showed β-phenyl-isopropylhydrazine to be approximately 60 times as effective as iproniazid in raising the brain 5-hydroxytryptamine. When mice were given 5-hydroxytryptophan and the amine content of the whole mice estimated, pretreatment with β-phenylisopropylhydrazine increased their 5-hydroxytryptamine content whereas pretreatment with iproniazid did not change it. The concentration of the amine in mouse brain and in whole mice was lower after reserpine, but was raised when reserpine and β-phenylisopropylhydrazine were given together. A small rise in brain 5-hydroxytryptamine was found after chlorpromazine; when chlorpromazine was given with iproniazid, however, the resulting increase was less than that found after iproniazid alone. Brain 5-hydroxytryptamine was unaltered after prolonged treatment with morphine.     

Morphine Antinociception Is Enhanced in mdr1a Gene-Deficient Mice

Purpose. Previous studies have suggested that P-glycoprotein (P-gp)modulates opioid antinociception for selected -and -agonists. Thisstudy was undertaken to assess morphine antinociception in micelacking the mdr1a gene for expression of P-gp in the CNS. Methods. Morphine (n = 4–5/group) was administered as a single s.c.dose to mdr1a(–/–) mice (3–5 mg/kg) or wild-type FVB controls(8–10 mg/kg). Tail-flick response to radiant heat, expressed as percentof maximum response (%MPR), was used to determine theantinociceptive effect of morphine. Concentrations in serum, brain tissue, andspinal cord samples obtained immediately after the tail-flick test weredetermined by HPLC with fluorescence detection. Parallel experimentswith R(+)-verapamil, a chemical inhibitor of P-gp, also were performedto further investigate the effect of P-gp on morphine-associatedantinociception. Results. Morphine-associated antinociception was increasedsignificantly in the mdr1a(–/–) mice. The ED₅₀ for morphine was > 2-foldlower in mdr1a(–/–) (3.8 ± 0.2 mg/kg) compared to FVB (8.8 ±0.2 mg/kg) mice. However, the EC;i5;i0 derived from the brain tissuewas similar between the two mouse strains (295 ng/g vs. 371 ng/g).Pretreatment with R(+)-verapamil produced changes similar to thoseobserved in gene-deficient mice. P-gp does not appear to affectmorphine distribution between spinal cord and blood, as the spinalcord:serum morphine concentration ratio was similar betweengene-deficient and wild-type mice (0.47 ± 0.03 vs. 0.56 ± 0.04, p > 0.05). Conclusions. The results of this study are consistent with thehypothesis that P-gp attenuates the antinociceptive action of morphine bylimiting the brain:blood partitioning of the opioid.     

Plasma levels of morphine and morphine glucuronides in the treatment of cancer pain: relationship to renal function and route of administration

Summary There is growing evidence that renally-impaired patients receiving morphine therapy are at greater risk of developing opiate toxicity, due to the accumulation of an active metabolite, morphine-6-glucuronide (M6G), which is usually excreted by the kidneys. This study examined the relationships between morphine dosage, renal function, and trough plasma concentrations of morphine and its glucuronide metabolites in 21 patients (aged mean: 68.5 years; 11 males) receiving either oral or subcutaneous morphine for terminal cancer pain. The median daily morphine dosages (mg · kg^–1) were: orally 1.87 (range 0.37–6.82) and subcutaneously 1.64 (range 0.22–3.60).The median plasma concentrations of morphine, morphine-3-glucuronide (M3G), and M6G (ng · ml^–1) were: 36.0, 1035.2, and 142.3, respectively. The plasma concentrations of morphine, M3G and M6G were each significantly related to the daily morphine dosage (n=21, Spearman r=0.79, 0.91, and 0.88 respectively). Accumulation of the morphine glucuronides was dependent on renal function. The plasma concentrations of M3G and M6G, when divided by the morphine concentration, were significantly related to the caluclated creatinine clearance of the patient. Patients receiving oral morphine had higher plasma concentration ratios of glucuronide/morphine than those receiving subcutaneous therapy, presumably due to first-pass glucuronidation.The results of this study confirm that accumulation of the pharmacologically active M6G is related to renal function, which probably explains the observation that morphine dosage requirements are generally reduced in patients with renal impairment.     

Screening of Multidrug-Resistance Sensitive Drugs by In Situ Brain Perfusion in P-Glycoprotein-Deficient Mice

Purpose. This study was conducted to assess the influence of P-glycoprotein (P-gp) on brain uptake of multidrug resistance sensitive drugs using an in situbrain perfusion technique in P-gp-deficient (mdr1a[–/–]) and wild-type mice. Methods. The blood-brain transport of radiolabeled vinblastine, vincristine, doxorubicin, colchicine, and morphine was evaluated in mdr1a(–/–) and wild-type CF-1 mice with the in situ brain perfusion technique. Brain uptake of drugs after intravenous pretreatment with P-gp reversal agents, (PSC 833, GF 120918, or (±)-verapamil), or vehicle also was studied in wild-type mice. In all experiments, cerebral vascular volume was determined by co-perfusion of sucrose. Results. Cerebral vascular volume was preserved during perfusion, indicating maintenance of blood-brain barrier integrity in both types of mice within the concentration range of substrates in the perfusate. The apparent brain transport of colchicine, vinblastine, doxorubicin, and morphine was increased 3.0, 2.7, 1.5, and 1.4-fold, respectively, in mdr1a(–/–) mice compared with the wild-type; the brain uptake of vincristine was not affected by P-gp. Preadministration of PSC 833 or GF 120918 in wild-type mice led to a 3-fold increase in the brain transport of colchicine and vinblastine, but no effect was observed for the other compounds. Intravenous verapamil enhanced colchicine brain transport (1.8-fold), but failed to increase the brain uptake of vinblastine and morphine. Conclusion. The in situ brain perfusion technique appears to be a sensitive and powerful tool for medium throughput screening of the brain uptake of multidrug resistance sensitive drugs. The effect of P-gp is characterized more efficiently with mdr1a(–/–) mice than by using modulators of P-gp in wild-type mice.     

Analgesic effect and plasma concentrations of codeine and morphine after two dose levels of codeine following oral surgery

Summary A double blind randomised cross over investigation was carried out in 25 male patients undergoing two oral surgical extractions, one for each lower wisdom tooth. The two extractions were performed about 6 weeks apart and were carried out under local anaesthesia. One hour after each extraction the patients randomly received 90 or 45 mg codeine. During the following 5 h the patients rated the intensity of their pain on a visual analogue scale. Blood was simultaneously sampled and assayed for codeine and its metabolite morphine.Mean pain intensity difference was just significantly higher after 90 mg codeine compared to 45 mg. The mean plasma concentrations of codeine and morphine were significantly higher after the 90 mg dose. However, for the two dose levels of codeine there was no obvious relationship between the difference in analgesic effect and the difference in the plasma concentration of codeine or morphine. The plasma concentrations of morphine were 2–3% of those of codeine and the levels were relatively low. Local formation of morphine from codeine within the human brain should therefore be investigated.Four patients were unable to demethylate codeine to a detectable plasma concentration of morphine after 90 mg codeine. In those patients the analgesic effect during the first hours was better after 90 mg codeine than after 45 mg.This suggests some analgesic effect of codeine itself.The work was presented in part at the Fourth World Congress on Pain, Adelaide, Australia, April 1–6, 1990     

Pharmacokinetics of diazepam following multiple-dose oral administration to healthy human subjects

Six healthy subjects between the ages of 21 and 31 years received diazepam tablets orally at a dose of 5 mg t.i.d. atO, 5, and 10hr on days 1–13. On day 14, the dose was 5 mg at 0 and 5 hr and 15 mg at 10 hr. Subsequently, the dose was 15 mg once daily on days 15–24. Numerous plasma samples were obtained during the multiple-dose regimen, and appropriate equations were fitted to all the multiple-dose data. Diazepam absorption was satisfactorily described by a first-order process, with disposition characterized by a linear two-compartment open model. The harmonic mean absorption half-life was 32 min, and the harmonic mean terminal exponential half-life was 57hr. The mean apparent oral total drug plasma clearance was 22.7ml/hr/kg. Steady-state plasma levels of the primary metabolite, desmethyldiazepam, were reached after 5–8 days of dosing. Steady-state diazepam plasma concentration-time profiles suggested that once daily administration of the total daily dose at bedtime might be a satisfactory dosing regimen.     

Pharmacokinetics of epidural morphine in man

Summary Cerebrospinal fluid (CSF) and plasma morphine concentrations were determined in 5 patients after epidural administration of 6 mg morphine; plasma samples were collected frequently during the initial 6 h and 6–7 CSF samples were obtained from each patient over a 24 h period. Morphine was analysed using gas chromatography and electron capture detection. Individual morphine concentration-time curves were plotted for plasma and CSF and various pharmacokinetic variables were calculated. Plasma morphine concentrations after epidural injection were similar to those found after intramuscular administration; C_max (66±8 mg/ml: mean±SEM) appeared within 12±3 min, and the terminal elimination half-life in plasma was 213±24 min. In CSF, morphine reached a peak (1575±359 ng/ml) after 135±40 min. The terminal elimination half-life for morphine in CSF was 239±10 min. The CSF bioavailability of morphine after epidural administration was calculated to be 1.9±0.5%.The study showed that epidural administration of morphine resulted in CSF concentrations many times higher than those in plasma, but still only 2% of the dose administered was available to the CSF compartment. Morphine was eliminated with similar speed from CSF and plasma.     

Pharmacokinetics of diminazene in sheep

The pharmacokinetic behavior of diminazene in plasma after administration of 2 mg/kg i.v. and 3.5 mg/kg i.m. was studied in four healthy Dala × Ryggja rams. Following i.v. injection, the data were satisfactorily described by a triexponential equation; the apparent volume of distribution at the steady-state was 0.56±0.04 L/Kg (X±sd; n=4); total body clearance averaged 1.1±0.09 ml/kg/min and elimination half-life was 9.30±1.40 hr. After intramuscular administration peak plasma levels of 6.30–7.57 g/ml were reached in 20 to 45 min and the mean absorption time averaged 5.83±1.61 hr. Systemic availability relative to the intravenous dose was 95.10±23.21% and mean residence time averaged 14.16±1.55hr. The partition of diminazene between erythrocytes and plasma averaged 0.64±0.10; plasma protein binding was high (65–85%) and concentration-dependent. Based on the experimental data obtained, an initial i.m. dose of 2.5 mg/kg followed by 2 mg/kg 24 hr later should be safe and effective in cases of babesiosis and trypanosomiasis sensitive to diminazene. A preslaughter withdrawal period of 14–26 days was estimated.     

The human pharmacokinetics of cefotaxime and its metabolites and the role of renal tubular secretion on their elimination

The pharmacokinetics of cefotaxime were investigated in human volunteers given constant intravenous infusions, intravenous bolus, and intramuscular doses of the drug. After intravenous dosing, the plasma levels of cefotaxime declined in a biphasic manner with a terminal half-life varying between 0.92 and 1.65 hr. Moreover, the pharmacokinetics were linear up to at least a 2.0 g dose for volume of distribution based on area (23.3–31.31), plasma clearance (249–288 ml/min), and renal clearance (151–177 ml/min). Renal tubular secretion of intact cefotaxime and each of its metabolites was demonstrated by its interaction with probenecid, although the ratio of drug to metabolites ultimately excreted in urine after probenecid was similar to that seen normally (54±6, 19±4, 6.5±0.7 and 5.5±0.7% for cefotaxime, DACM, M2, and M3, respectively, when calculated as a percentage of the dose). The observed half-lives of DACM, M2, and M3 were 2.3±0.4, 2.2 ±0.1 and 2.2 hr, respectively. However, when the true half-life of DACM was calculated (0.83±0.23 hr) it was not only significantly shorter than that observed but also shorter than that for intact cefotaxime. The plasma clearance of DACM (744 ±226 ml/min) was much higher than that of cefotaxime while the volume of distribution was of a similar order (56 ±241). When administered intramuscularly, there was good absorption of cefotaxime from the site of injection (92–94%) giving maximum plasma levels of the drug of between 30 and 35 mg/l at approximately 40 min after dosing. Thereafter, the plasma levels of cefotaxime declined in a monophasic manner with a half-life (1.0–1.2 hr) similar to that of the terminal half-life seen after intravenous administration. Lidocaine had no significant effect on either its absorption or elimination kinetics.     

Pharmacokinetics of bendroflumethiazide after low oral doses

The pharmacokinetics of bendroflumethiazide after oral administration of 1.25, 2.5, and 5.0 mg were studied in nine healthy male volunteers. Bendroflumethiazide was analyzed by GLC after extractive alkylation. After the lowest dose, the plasma concentration could be followed to 14 hr, and the data were adequately fitted by a one-compartment model; the half-life was 3.1 hr. After the 2.5 and 5.0 mg doses, the plasma concentration was followed for 24 hr, and the data were fitted by a two-compartment model with half-lives of 8.9 hr. The urinary sodium concentration was doubled after bendroflumethiazide intake, but the urinary potassium concentration remained almost constant. The renal clearance of bendroflumethiazide was around 30 ml · min ^–1 .     

An in Vitro Study of Diamorphine Permeation Through Premature Human Neonatal Skin

The permeation kinetics of diamorphine through human premature neonatal cadaver skin over a range of gestational ages between 24 and 36 weeks was investigated using small diffusion cells. A strong inverse correlation was noted between the apparent permeability coefficient and the gestational age of the skin (P < 0.01; n = 26). The calculated apparent permeability coefficients decreased with gestational age from 6.0 × 10 ^–2 cm · hr^–1 at 24 weeks' gestation to 5.2 × 10^–6 cm · hr^–1 at 36 weeks' gestation. The amount of diamorphine remaining bound within the skin at the end of the in vitro experiments did not change significantly with gestational age of the skin. Diamorphine was subject to degradation over the course of the in vitro experiments to produce significant amounts of 6-mono-acetylmorphine and evidence is presented to suggest that this was due to residual skin esterase activity. It is calculated that the steady-state flux rate of diamorphine through neonatal skin observed in these experiments would be sufficient to obtain a therapeutic plasma concentration of morphine assuming a 2-cm² area for application and a delivery rate of 15 µg hr ^–1 kg^–1. However, the prolonged half-life of morphine in the premature neonate would result in a delay of some hours before the attainment of this level.     

The disposition of intravenous doxapram in man

Summary The pharmacokinetics of intravenous doxapram in healthy individuals is consistent with a three-compartment open model. Doxapram was administered by bolus injection (1.5 mg · kg^–1) and by intravenous infusion (6.5 mg · kg^–1 for 2 h) to 5 subjects on separate occasions. There was no significant difference in mean terminal plasma half-lives (355 and 448 min) or in mean total body clearances (5.9 and 5.6 ml · min^–1 · kg^–1) following i. v. bolus injection or infusion respectively. In 3 subjects plasma doxapram concentrations during and after i. v. infusion agreed with those predicted from pharmacokinetic values obtained from the bolus injection study. Since mean steady-state concentrations (9.9 µg · ml^–1) would be reached only after an extended interval (mean 15.2 h), a variable-rate infusion regimen was calculated to produce and maintain a concentration of 2 µg · ml^–1 from 15–25 min onwards. A regimen in which the infusion rate is reduced step-wise is recommended to achieve early near-constant plasma doxapram concentrations.     

Disposition of sulfadimethoxine in swine: Inclusion of protein binding factors in a pharmacokinetic model

Sulfadimethoxine was administered intravenously and orally to five swine. More than 75% of the dose was excreted into urine as the acetyl metabolite with 4–6% excreted unchanged. Plasma and urine data were not consistent when a linear pharmacokinetic model was used to describe the data. Sulfadimethoxine has a high affinity for plasma protein, and the data were subsequently fitted to a nonlinear model, which included saturable protein binding. The choice of a nonlinear model was further supported by a minimum value for the Akaike information criteria. The protein binding constant obtained was 2.8× 10⁴ M^–1 and the total protein binding site concentration in plasma was 4.6×10^–4 m. Both values are comparable with in vitrodata. This result suggests that the nonlinear model involving protein binding can be successfully applied to pharmacokinetic data. The apparent biological half-life of Sulfadimethoxine (free and bound) in plasma was 14 hr; however, the half-life of elimination of free drug was 1.25 hr. Following oral administration, all of the dose was absorbed with an apparent absorption half-life of 2.9 hr.Supported in part by FDA contract 74–178. Presented at the APhA Academy of Pharmaceutical Sciences 25th National Meeting, Hollywood, Fla, November 12–16, 1978.     

Pharmacokinetics of Multiple Daily Transdermal Doses of Nicotine in Healthy Smokers

The plasma concentration–time profiles and pharmacokinetics were characterized for nicotine and its major metabolite, cotinine, after multiple daily application of a nicotine user-activated transdermal therapeutic system (UATTS) to nine healthy smokers. The volunteers abstained from smoking 24 hr prior to and during the course of the study. A 10-cm² system (designed to deliver 75 µg/cm²/hr) was applied every 24 hr for 5 days, with serial blood samples taken on Days 1 and 5 and after system removal on Day 5. Generally, the nicotine UATTS was well tolerated. Predose nicotine concentrations on Days 3 to 5 indicated that steady state was reached by Day 3. The nicotine pharmacokinetic parameters for Day 1 and Day 5 were similar: the mean (SD) AUC(0-24) values for Days 1 and 5 were 271.7 (50.7) and 311.7 (55.0) ng · hr/ml, the mean (SD) C _max values were 16.3 (2.6) and 16.8 (2.9) ng/ml, and the median (range) T _max values on Days 1 and 5 were 12 (9–24) hr and 12 (0–24) hr, respectively. There was only slight or no accumulation of nicotine after multiple dosing as indicated by the Day 5 to Day 1 AUC and C _max ratios of 1.15 (0.09) and 0.98 (0.06), respectively. Overall, the UATTS system maintained relatively constant plasma nicotine concentrations and is suitable for once-daily application.     

Concentration of ondansetron in cerebrospinal fluid following oral dosing in volunteers

This study measured the concentrations of ondansetron in plasma and cerebrospinal fluid (CSF) in six volunteers after oral dosing to steady state. Ondansetron concentrations ranged from 39.5–147 ng ml^–1 in plasma and from 2.6–15.4 ng ml^–1 in CSF. There was good correlation between plasma and CSF concentrations (r=0.89,p=0.017). CSF concentrations were less than 15% of plasma concentrations in all cases, indicating that the rate of penetration of the blood brain barrier by ondansetron is low.     

Effects of d-amphetamine,morphine, naloxone,and drug combinations on visual discrimination in rats

The effects of d-amphetamine, morphine, and naloxone on visual discrimination were investigated using a two-choice discrete-trial procedure in which rats were trained to discriminate the position of a lightflash. Morphine (0.3–5.6 mg/kg) but not amphetamine (0.1–1.0 mg/kg) caused a significant dose-dependent disruption in discriminative performance. Both amphetamine and morphine increased response latencies. Naloxone (1.0 mg/kg) prevented the disruption of any aspect of performance by up to 100 mg/kg morphine. Performance after naloxone/amphetamine co-administration was not significantly different from that observed after amphetamine alone. Naloxone alone (0.3–10 mg/kg) had no effect on discrimination, spatial bias or response latencies. These results suggest that morphine and amphetamine affect different components of discrimination performance. Offprint requests to: S.G. Holzman     

The postoperative pharmacokinetics of codeine

Summary The metabolism and systemic availability of codeine have been studied in 12 patients after cholecystectomy. They were given 20 mg codeine as an IV bolus dose on the first day after surgery and 50 mg codeine as a single oral on the fourth day after surgery. Codeine had a medium to high extraction ratio and a total plasma clearance of 10.8 (4.3) ml·min^–1·kg^–1. The clearance varied fourfold between subjects. All the patients were extensive metabolizers with regard to the debrisoquine/sparteine polymorphism, as tested using dextromethorphan as the probe drug. Nevertheless, the formation of morphine from codeine was very small and plasma morphine concentrations were below the detection limit of 3.3 nmol·1^–1 (1 ng·ml^–1). As a corollary, the morphine/codeine ratio in the the concentration-time curves was less than 3% in all the patients.The systemic availability of codeine varied extensively between subjects (range 12–84%). This might partly explain differences in the dose of codeine required as an analgesic.