Subjects were male or female Japanese and Caucasian volunteers aged 20–45 years who were in good health according to medical history, physical examination, vital signs, electrocardiogram (ECG) and laboratory tests, and who had a body mass index (BMI) of 18.0–25.0 kg/m2 and body weight of 50–90 kg. Japanese and Caucasian cohorts were sex- and BMI matched, as closely as possible. Subjects were defined as Japanese if they were born in Japan, had two parents of Japanese descent and had not resided outside Japan for > 5 years. Key exclusion criteria included pregnancy and lactation; a history of significant gastrointestinal, liver or kidney disease, or other conditions that might affect drug pharmacokinetics; use of opioids or opioid antagonists within 30 days of enrolment; and use of any medication within 7 days of, or during, the study. Full inclusion and exclusion criteria are included in Online Resource 1.
This phase 1 study (EudraCT No 2015-003071-29) was conducted between 20 October 2015 and 17 December 2015. The trial was a randomised, open-label, two-cohort, three-period, crossover, single-dose bioavailability study, conducted in a single centre in the UK by a contract research organisation (Richmond Pharmacology Ltd). Screening and subject eligibility assessments took place between days − 21 and − 1, and enrolment and randomisation took place between day − 1 and day 1. Subjects were randomised, via an automated system and central randomisation list, to receive each of three oral doses of CTC once over the course of three dosing periods, according to six pre-determined treatment sequences (Fig. 1). The treatments were 100 mg CTC (1 × 100 mg tablet consisting of 44 mg rac-tramadol.HCl and 56 mg celecoxib; Laboratorios del Dr. Esteve, S.A.U.), 150 mg CTC (1 × 150 mg tablet consisting of 66 mg rac-tramadol.HCl and 84 mg celecoxib) and 200 mg CTC (2 × 100 mg tablets, total dose consisting of 88 mg rac-tramadol.HCl and 112 mg celecoxib). Each dose was separated by a washout period of at least 10 days. For each dosing period, subjects entered the study clinic on day − 1 and remained until pharmacokinetic and safety assessments were completed at 48 h post-dose on day 3. Dosing occurred on day 1 of each period, under fasting conditions (subjects fasted from 10 h pre-dose until at least 4 h post-dose). Subjects received standardised meals and fluid intake during the remainder of each in-patient stay. During each in-patient stay and for 48 h before check-in to each study period subjects had to refrain from the consumption of grapefruit juice and any food or drinks containing alcohol, caffeine, poppy seeds or xanthine. Tobacco use was not permitted during the study or for 90 days before the first administration of CTC. A follow-up visit was conducted 7–10 days after final CTC administration. The maximum study length was 52 days per subject.
The study drug was manufactured and packaged according to Good Manufacturing Practice and supplied to the principal investigator in labelled containers by the manufacturer (Laboratorios del Dr. Esteve, S.A.U.).
The study protocol was approved by an ethics committee (South Central-Berkshire B Research Ethics Committee, UK; IRAS:188440; REC reference: 15/SC/0549). The study sponsor was Mundipharma Research Ltd. The study was performed in accordance with Good Clinical Practice guidelines, according to the principles of the Declaration of Helsinki and the International Conference on Harmonization Good Clinical Practice guidelines. All participants provided written informed consent.
Blood samples were taken for pharmacokinetic analysis pre-dose and at 0.5, 0.75, 1, 1.25, 1.5, 1.75, 2, 2.5, 3, 3.5, 4, 5, 6, 8, 10, 12, 16, 24, 36 and 48 h after dosing in each of the three dosing periods. Methods for blood sample collection have been described in detail elsewhere . In brief, blood samples were collected in K2-EDTA-containing tubes and centrifuged for 15 min at 1500 g and 4 °C, within 30 min of collection. Following centrifugation, plasma was separated and stored at −80 °C.
Plasma concentrations of tramadol, O-desmethyltramadol (M1; tramadol’s primary metabolite) and celecoxib were measured using validated high-performance liquid chromatography with tandem mass spectrometry. Samples (0.05 mL) were extracted by solid-phase extraction and separated by reverse-phase liquid chromatography on a XBridge C18 column (3.5 μm 50 × 2.1 mm, Waters Corporation, Milford, MA, USA), using a gradient of water with 0.1% formic acid, and acetonitrile with 0.1% formic acid as a mobile phase. Detection was performed via tandem mass spectrometry on an API 4000 triple quadrupole mass spectrometer (Sciex, Ontario, Canada) using Turbo Ion spray ionisation. Propranolol and E-6087, a cyclooxygenase-2 inhibitor, were used as internal standards. Linearity and range; selectivity; specificity in the presence of concomitant medication; intra- and inter-run precision and accuracy; limits of quantification; dilution integrity; carry-over; recovery; matrix effect; and stability in an autosampler at room temperature, after freeze–thaw cycles and long-term, were evaluated during method validation. The lower and upper limits of quantification were 4.00 and 640.00, 1.00 and 160.00, and 2.50 and 1000.00 ng/mL, for tramadol, M1 and celecoxib, respectively. Samples below these limits were not quantified and were considered to be below the lower limit of quantification. Samples above the limit of quantification were reassayed after dilution, using a dilution factor equal to or lower than the maximum dilution factor tested during method validation (dilution factor of 10). The inter-run accuracy (nominal values) varied from 99.81 to 103.65% for tramadol, 96.21 to 103.69% for M1, and 105.84 to 108.68% for celecoxib, and 94.2, 92.7 and 109.5% for limits of quantification, respectively. The inter-run precision (coefficient of variation) varied from 4.77 to 8.67% for tramadol, 7.67 to 9.38% for M1 and 6.19 to 7.05% for celecoxib, and 15.5, 15.3 and 10.1% for limits of quantification, respectively. Assay specificity was assessed using six independent matrix sources from Caucasian and Japanese volunteers, verified for the absence of interference and compared with the respective limits of quantification at the retention times and mass transitions of analytes and internal standards. Absolute recovery was 97.9–99.2, 96.5–101.0 and 74.1–78.0% for tramadol, M1 and celecoxib, respectively. The stability of the samples at the bench was 25 h and samples were stable under three freeze–thaw cycles. The maximum sample storage duration was 76 days; long-term stability of frozen samples was 160 days.
The primary pharmacokinetic parameters were tramadol and celecoxib Cmax and area under the plasma concentration–time curve measured from the time of dosing to the last measurable concentration (AUCt). Other pharmacokinetic parameters included area under the plasma concentration–time curve measured from the time of dosing and extrapolated to infinity (AUC∞); time to Cmax (Tmax); and terminal phase half-life (T½Z).
Safety was evaluated via the assessment of adverse events (AEs), standard laboratory evaluations, physical examination, vital signs (including pulse oximetry) and 12-lead ECG. AEs were monitored throughout the study and severity was graded by the investigator. All other safety assessments were conducted during initial screening and at the end-of-study follow-up visit. In each dosing period, vital signs were also assessed pre-dose and at 1, 2, 4, 6, 8, 12, 16, 24, 36 and 48 h post-dose; physical examination was conducted pre-dose and at 48 h post-dose; and ECG at 2 h post-dose.
The primary objective was not associated with any formal statistical hypotheses and therefore sample size was calculated based on the number of subjects required to assess the dose proportionality of CTC in each cohort (a secondary study objective). Using a dose-adjusted bioavailability analysis of variance (ANOVA) model, dose proportionality of two CTC doses would be demonstrated if the 90% confidence intervals (CI) for the geometric least-squares means of AUCt and Cmax fell entirely between 80.00 and 125.00%. The sample size calculation was performed using nQuery (Cork, Ireland) applying the t test of equivalence in ratio of means for crossover designs, assuming true ratios of 100% between CTC doses and a standard deviation of the period differences on the log scale of 0.330. No adjustment was made for multiple testing. Twenty-six subjects were calculated to be required per cohort to provide 90% power to determine dose proportionality. With an assumption that approximately 10% of randomised subjects would not be eligible for inclusion in the pharmacokinetic population or would not provide sufficient valid pharmacokinetic parameters to be included in the statistical analysis for a comparison, 30 subjects were randomised within each cohort.
The intention-to-treat (ITT) population consisted of all randomised subjects. The pharmacokinetic population consisted of all subjects who received one dose of CTC and from whom at least one valid pharmacokinetic parameter was recorded. The safety population consisted of all subjects who received at least one dose of CTC.
Pharmacokinetic parameters were summarised descriptively by CTC dose in each cohort. AUC∞ was calculated from the ratio of the final observed plasma concentration to the terminal phase rate constant; this was added to AUCt. (AUCt was calculated using the linear up/log down trapezoidal method). Pharmacokinetic calculations were conducted using Phoenix WinNonlin, Version 6.4.
Within-cohort dose proportionality for tramadol and celecoxib was calculated by evaluating AUCt, AUC∞ and Cmax using two methods; the dose-adjusted bioavailability ANOVA model and a power model. In the ANOVA, proportionality of two CTC doses was demonstrated if 90% CI values for the ratio of geometric least-squares means for AUCt, AUC∞ and Cmax fell between 80.00 and 125.00%. Dose proportionality was calculated using 100 mg CTC as a reference dose (i.e. 150 mg CTC and 200 mg CTC were each compared to 100 mg CTC). In the power model, dose proportionality was demonstrated if the 90% CIs of the slope estimates fell entirely between the critical region (0.6781–1.3219). The key modelling assumption in the power model was that the logarithm of each pharmacokinetic variable was linearly related to the logarithm of the dose.
For comparisons of pharmacokinetic exposure across cohorts, Cmax, AUCt and AUC∞ were analysed on a log scale using an ANOVA model with fixed effects for sequence, period and cohort. Test/reference (Japanese cohort/Caucasian cohort) ratios were calculated by transforming the difference between the natural log least-squares means back to the original scale and multiplying by 100. Comparability of tramadol, M1 or celecoxib pharmacokinetic exposure between cohorts was to be assumed if the 90% CIs for a test/reference ratio were > 80 and < 125%. A post hoc weight-adjusted inter-cohort comparability analysis was also performed. Statistical programming and analyses were performed using SAS® version 9.3 or higher (SAS Institute, Cary, USA).