Twelve healthy male volunteers, median age 31 years (range 26–42 years), median weight 77 kg (range 63–95 kg), median BMI 23.9 kg/m² (range 19.9–28.4 kg/m²) participated in the study. One subject had to be replaced because of additional drug intake during the study.

The study was approved by the ethics committee of the local medical board (Landesärztekammer Baden Württemberg) according to the Declaration of Helsinki (1996 Sommerset). All volunteers gave their written informed consent prior to study inclusion.

The volunteers were healthy according to history, physical examination and laboratory tests, had no history of drug abuse and did not take any regular medication. The volunteers had previously been genotyped for CYP2D6 by allele-specific PCR 2122 and all were predicted to be extensive metabolisers (EM). In addition, the subjects were phenotyped with respect to sparteine oxidation polymorphism with a single dose of 100 mg sparteine sulphate and were classified as EMs according to the metabolic ratio (MR) of sparteine and its 2- and 5-dehydrometabolite.23 In order to limit the effect of the variability of CYP2D6 on metabolic capacity, volunteers with a MR<1 were generally included in the study, with one subject having a MR of 2.8 (intermediate metaboliser).

The study was designed as a randomised, placebo-controlled, double-blind, cross-over trial. Randomisation was performed using the computer program Sampsize 2.0 (Blackwell Science Ltd., Machin, Campbell, Fayers, Pinol). Each volunteer received on the study days 1 and 8 in random order either 100 mg codeine phosphate + 50 mg diclofenac sodium or 100 mg codeine phosphate + placebo. Codeine phosphate was provided as tablets (Codeinum Phosphoricum forte Compretten^®, Glaxo Welcome GmbH/Cascan GmbH, Hamburg, Germany), diclofenac sodium and placebo were provided as capsules of identical appearance (manufactured by Contract Pharma GmbH & Co KG, Murr, Germany).

On study days 1 and 8 the following identical procedures were carried out: Blood samples were taken before drug administration as well as 0.25, 0.5, 0.45, 1, 1.5, 2, 2.5, 3, 4, 6, 8, 10, 12, 24, 35 and 36 hours after drug administration. Blood samples were centrifuged after 30 min. and serum was stored at -20°C until analysed. Urine was collected before drug administration and from 0–12, 12–24 and 24–36 hours after drug administration. The volume was measured and an aliquot of each collection was stored at -20°C until analysed.

The analgesic effects of codeine with and without simultaneous administration of diclofenac or placebo were assessed using the cold pressor test as previously described.17 Briefly, the cold pressor test apparatus consisted of temperature-controlled water baths of 35 ± 0.5°C and 1.0 ± 0.5°C (ice-water bath). The nondominant forearm was placed into the warm-water bath for exactly 2 minutes. Fifteen seconds before transferring the forearm into the cold-water bath, a blood pressure cuff was inflated to 20 mmHg below the diastolic blood pressure and the eyes were covered with eye-patches. Subjects placed their forearm in a fixed position with the fingers wide apart into the cold-water bath (for a maximum time period of 2 min). They were instructed to clearly indicate the time of the first pain sensation as well as the time of intolerable pain, at which the forearm was removed from the cold-water bath. The time from the immersion of the forearm into the cold-water bath to the first pain sensation was measured, and is defined as the pain threshold. The pain tolerance is defined as the time from immersion of the forearm into the cold-water bath until the time of intolerable pain. A training session was carried out before the study to screen out volunteers with a pain tolerance of less than 15 sec or more than 120 sec and to familiarise the volunteers with the study procedure. The pain experiments were conducted on study days before drug administration (10–15 min before drug administration) and 1, 1.5, 2, 2.5, 3, 3.5, 4 and 6 hours after drug administration. Subsequently, the changes from baseline were determined and for each study day the area under the pain threshold change versus time curve and the area under the pain tolerance change versus time curve were calculated.

Adverse events were evaluated by the following method: The volunteers were required to list each symptom (fatigue, headache, dizziness, blurred vision, nausea, itching, exanthema and any other symptom) on a visual analogue scale (VAS) rated from 0 (not present) to 10 (most severe) at zero h (baseline, before drug intake), 2, 4 and 6 hours after drug intake. The values were corrected for the baseline measurement at t = 0 h.

Codeine and its metabolites were determined using HPLC-electrospray mass spectrometry analogous to a previously published method24 with minor modifications. Samples (1.0 ml of plasma, or 100 μl of urine diluted with 900 μl of water) were spiked with the deuterated internal standards (10 μl of standard mix with 50 pmol/μl codeine-G-glucuronide-d₃, 5 pmol/μl codeine, 10 pmol/μl morphine-3-glucuronide-d₃, 5 pmol/μl morphine-6-glucuronide-d₃ and 1 pmol/μl morphine-d₃) and extracted automatically using end-capped C₂ solid-phase extraction columns. For the determination of the conjugates of norcodeine and normorphine, urine was hydrolysed25 prior to extraction. The mobile phases used for HPLC were: (A) 1 % acetic acid in water and (B) 1 % acetic acid in acetonitrile. HPLC separation was achieved on a LiChrospher 100 RP-18 (Merck, Darmstadt, Germany) end-capped analytical column (125 × 3 mm I.D., 5 μm particle size) at a flow rate of 0.5 ml/min using a linear gradient from 8 % B to 40 % B in 8 min. The mass spectrometer (HP 1100 MSD, Hewlett-Packard, Waldbronn, Germany) was operated in the selected ion monitoring mode using the respective MH⁺ ions, m/z 476 for codeine-6-glucuronide, m/z 300 for codeine, m/z 303 for codeine-d₃, m/z 286 for norcodeine, m/z 462 for morphine-3-glucuronide and morphine-6-glucuronide and m/z 465 for the deuterated morphine glucuronides, m/z 286 for morphine, m/z 289 for morphine-d₃ and m/z 272 for normorphine. The limits of quantification achieved with this method were 0.5 pmol/ml for the morphine-glucuronides and morphine, 2 pmol/ml for codeine, normorphine and norcodeine and 5 pmol/ml for codeine-6-glucuronide with coefficients of variation less than 12 %.

Following the addition of 100 μl of a 10 μg/ml solution of 4'-methoxydiclofenac in water for internal standardisation and 2 ml 10 mM phosphate buffer pH 6.5 to 1 ml of serum, samples were extracted by an automated solid phase extraction procedure using an ASPEC XL (Gilson, Villiers Ie Bel, France). Samples were applied on Bakerbond SPE™ C₁₈ extraction columns (3 ml, 500 mg) which were equilibrated with 3 ml methanol and 3 ml of phosphate buffer. The columns were washed with 2 ml of phosphate buffer and 2 ml of phosphate buffer: methanol 6:4 (v/v) and eluted with phosphate buffer: methanol 2:8 (v/v). The solvents were evaporated and the residue was resuspended in 150 μl of the mobile phase. A 100 μl aliquot was injected on a HPLC-system consisting of an autosampler SIL 9A, a solvent delivery system LC 9A and an UV-detector (Shimadzu, Duisburg, Germany). Separation was performed on a Waters Spherisorb ODS1 5 μm 125 × 4.6 mm column equipped with a guard column filled with the same material (Bischoff, Leonberg, Germany) using 8 mM tetrabutylammonium bromide in 10 mM phosphate buffer pH 6.5:acetonitrile:tetrahydrofuran 65:30:5 (v/v/v) at a flow rate of 1 ml/min. Diclofenac and the internal standard 4'-methoxydiclofenac were detected at 282 nm at a retention time of 11.0 and 12.4 min, respectively. Recovery was between 85 and 100.5%. Calibration curves were linear over a concentration range from 10 to 2000 ng/ml (r² ranging from 0.9993 to 1). Inter-assay variability (n = 9) was 14.2, 8.3, and 7.8% for 30, 300, and 2000 ng/ml, respectively. Accuracy was within 10% at these concentrations.

Standard noncompartmental analysis for calculations using serum concentration-time data was performed using TOPFIT 2.0 (Gustav Fischer Verlag 1993).

The following pharmacokinetic parameters were determined from serum concentration-time data and urine concentration data for codeine (cod), codeine-6-glucuronide (C-6-G), norcodeine (NC), morphine (morph), morphine-3-glucuronide (M-3-G), morphine-6-glucuronide (M-6-G) and normorphine (NM):

C_max- peak serum concentration [pmol ml^-1], obtained from the visual inspection of the serum concentration-time curves

t_max- time to reach peak serum concentrations [h], obtained from the visual inspection of the serum concentration-time curves

t_1/2- terminal phase half-life [h] calculated according to ln(2)/λ_z, the terminal elimination rate constant λ_z was determined from the slope of the regression line of In(concentration) vs time

AUC_0-∞- area under the serum concentration curve [pmol ml^-1*h], calculated by the trapezoidal rule, the segment to infinity was calculated from the last concentration measured by dividing the last concentration measured by the elimination rate constant (λ_z)

ratio AUC' s- AUC_0-∞ codeine/AUC_0-∞ morphine

Ae- cumulative amount of codeine, C-6-G, norcodeine, norcodeine-glucuronide, morphine, M-3-G, M-6-G, normorphine and normorphine-glucuronide [% of dose] excreted in urine

CLo- apparent oral clearance of codeine calculated using dose/AUC_0-∞ [ml min^-1]

CL_cod→c-6-G- metabolic clearance of codeine to C-6-G calculated as Ae C-6-G/AUC_0-∞ codeine [ml min^-1]

CL_cod→NC- metabolic clearance of codeine to norcodeine calculated as (Ae norcodeine + Ae norcodeine-glucuronide)/AUC_0-∞ codeine [ml min^-1]

CL_cod→morph- metabolic clearance of codeine to morphine calculated as (Ae morphine + Ae M-3-G + Ae M-6-G + Ae normorphine + Ae normorphine-glucuronide)/AUC_0-∞ codeine [ml min^-1]

CL_{morph→M-3-G}- metabolic clearance of morphine to M-3-G calculated as Ae M-3-G/AUC_0-∞ morphine [ml min^-1]

CL_{morph→M-6-G}- metabolic clearance of morphine to M-6-G calculated as Ae M-6-G/AUC_0-∞ morphine [ml min^-1]

CL_morph→NM metabolic clearance of morphine to normorphine calculated as (Ae normorphine + Ae normorphine-glucuronide)/AUC_0-∞morphine [ml min^-1]

CL_{ren cod}- renal clearance of codeine calculated as Ae codeine/AUC_0-∞ codeine [ml min^-1]

CL_{ren morph}- renal clearance of morphine calculated as Ae morphine/AUC_0-∞ morphine [ml min^-1]

CL_{ren C-6-G}- renal clearance of C-6-G calculated as Ae C-6-G/AUC_0-∞ C-6-G [ml min^-1]

CL_{ren NC}- renal clearance of norcodeine calculated as Ae norcodeine/AUC_0-∞ norcodeine [ml min^-1]

CL_{ren M-3-G}- renal clearance of M-3-G calculated as Ae M-3-G/AUC_0-∞ M-3-G [ml min^-1]

CL_{ren M-6-G}- renal clearance of M-6-G calculated as Ae M-6-G/AUC_0-∞ M-6-G [ml min^-1]

CL_{ren NM}- renal clearance of normorphine calculated as Ae normorphine/AUC_0-∞ normorphine [ml min^-1]

In this cross over study, 12 subjects were required to ensure a 80% power of detecting a potentially clinically relevant difference in pharmacokinetic parameters of 30% with a variability of 25% at the 5% significance level for α. Pharmacokinetic data are presented as mean ± 95% CI. Pharmacokinetic data of codeine and its metabolites after administration of codeine + placebo vs. codeine + diclofenac are compared using the Wilcoxon matched pairs test. In addition, AUC and C_max of codeine and C-6-G were investigated in terms of bioequivalence: Point estimates (geometric means) and 90% CI were given for the ratios test/reference (test: pharmacokinetic parameters under consideration after codeine + diclofenac, reference: pharmacokinetic parameters under consideration after codeine + placebo). Test was considered bioequivalent to reference if 90% CI of log transformed AUC ratios were within 0.80–1.25 and C_max ratios were within 0.75–1.34.

Adverse events are presented in two different manners: 1.) The number of subjects reporting any side effect after administration of codeine + placebo vs. codeine + diclofenac were compared using the chi square test. 2.) At different time points (2, 4 and 6 h after drug administration) VAS-rated mean sum scores of all reported side effects after administration of codeine + placebo vs. codeine + diclofenac were calculated; the mean sum scores are compared using the paired Wilcoxon matched pairs test at the different time points. The area under the pain threshold versus time curve and the area under the pain tolerance versus time curve (corrected to baseline at t = 0 h) are presented as mean ± SD and are compared using Wilcoxon matched pairs test. A p value of less than 0.05 was considered significant.

Similar serum concentrations of codeine and its metabolite C-6-G were observed after treatment with codeine + placebo vs. codeine + diclofenac (Figure 2). Peak serum concentrations (C_max) of codeine (764; 602–924 pmol ml^-1 vs. 821; 663–980 pmol ml^-1) and C-6-G (5263; 4738–5788 pmol ml^-1 vs. 5621; 5167–6074 pmol ml^-1), did not differ significantly between the two different treatments, nor did the time to attain peak serum concentrations (t_max) and the terminal half-life t_1/2 (Table 1). The concentrations of codeine and its metabolites in the one intermediate metaboliser (MR = 2.8) were within the range of the concentrations of the other volunteers with MR<1, however as expected the metabolites formed via O-demethylation (morphine, morphine glucuronides and normorphine) were the lowest or second lowest observed.

The AUCs of codeine did not differ between the two treatments (Table 2). In addition, the two treatment, based on AUCs and C_max of codeine and C-6-G, could be considered bioequivalent: Point estimates of the test/reference ratios and 90% CI were 1.007 (0.998–1.017) for AUC codeine, 1.004 (1.001–1.008) for AUC of C-6-G, 1.015 (0.990–1.042) for C_max of codeine and 1.008 (1.001–1.016) for C_max of C-6-G. However, a small but significant increase in the AUC of C-6-G (+ 5.5%) was observed in subjects after codeine + diclofenac (Table 2). In terms of morphine formation, the ratios of AUC codeine/AUC morphine did not reveal a significant difference (53.4; 10.5–96.4 95% CI vs. 45.2; 16.4–74.0 95% CI) between the two treatments.

Peak serum concentrations (C_max) of morphine (22.4 pmol ml^-1; 12.3–32.5 95% CI vs. 24.3 pmol ml^-1; 13.7–35.0 95% CI) and M-6-G (63.4 pmol ml^-1; 38.7–88.1 95% CI vs. 68.8 pmol ml^-1; 41.2–96.5 95% CI) did not differ significantly between the two different treatments, as well as the time to attain peak serum concentrations (t_max) and the terminal half-life t_1/2 (Table 1). AUC of morphine did not differ between the two treatments (Table 2). A small but significant increase of the AUC of M-6-G (+10.8%) was observed after codeine + diclofenac (Table 2). Serum concentration-time curves of morphine, M-3-G and M-6-G are displayed in Figure 3.

No significant differences were observed in either the apparent oral clearance of codeine (CLo) (1805 ml min^-1; 1498–2112 95% CI vs. 1700 ml min^-1; 1470–1929 95% CI), metabolic clearance of codeine to C-6-G, norcodeine and morphine, of morphine to M-3-G, M-6-G and normorphine or renal clearance of codeine and its metabolites between the two treatments (Tables 3 and 4). The total amount excreted of codeine and its metabolites excreted was 81.3% of dose (76.3–86.3 95% CI) vs. 84.9% of dose (81.8–88.0 95% CI) and did not differ between the two treatment groups. The urinary recoveries of codeine and its metabolites are listed in Table 5.

Diclofenac was detected in the serum of all subjects after codeine + diclofenac with mean peak serum concentration of 4.4 nmol/ml (3.77–5.06 95% CI) after 1.73 h (1.21–2.26 95% CI) (see Table 6).

The area under the pain threshold-time as well as the area under the pain tolerance-time curves did not differ significantly after administration of codeine + placebo vs. codeine + diclofenac (8.68 ± 12.7 vs. 4.87 ± 9.82 s*h and 23.3 ± 57.5 vs. 13.9 ± 35.6 s*h, respectively). The time courses of pain threshold and pain tolerance are displayed in Figure 4.

Both codeine + placebo and codeine + diclofenac were well tolerated by the volunteers. Only minor side effects occurred and no rating higher than 5 was reported on a VAS scale. The number of subjects reporting any adverse event from VAS rated from 0 (not present) to 10 (most severe) at 2, 4 and 6 hours after drug intake is displayed in table 7 and did not differ significantly between the two treatments. Also, no significant differences could be observed concerning the mean sum scores of all reported side effects at the given time points (Table 8).

CI: confidence interval

MR: metabolic ratio

NSAIDs: non steroidal anti-inflammatory drugs

C-6-G: codeine-6-glucuronide

M-6-G: morphine-6-glucuronide

UGT: UDP-glucuronosyltransferase