The protocol and subsequent amendments were reviewed and approved by the Independent Ethics Committee (Plymouth) (Non-NHS Phase 1). All subjects gave written consent to participation. Clinical Trial Exemption Certificate, CTX, was not required for this study. The clinical trials registration number is NCT01062867.
Setting and study subjects
The study was conducted in a specialist clinical trials unit with standard equipment for induction and maintenance of anaesthesia and subsequent supervised recovery. We studied healthy male subjects who were non-smokers and consumed less than 20 units of alcohol ( < 160 g) per week. All subjects underwent standard medical screening including laboratory investigations within 3 weeks of scheduled dosing.
Org 25435 was supplied as a freeze-dried cake stored at 2-8°C and reconstituted with water to 20 mg/ml.
Short Infusion study
This was a dose escalation study. Dosing began at 0.25 mg/kg infused over one minute and increased to 0.5, 1.0, 2.0, 3.0, 4.0, 5.0 and 6.0 mg/kg in subsequent subjects. Beginning with the lowest dose, two subjects were studied at each dose. Once general anaesthesia occurred in any volunteer, the number of subjects studied at that dose and upwards increased to five. The study proceeded to one dose level above the first level at which three or more of the five subjects experienced general anaesthesia. Anaesthesia was defined as loss of speech contact, loss of eyelash reflex and dropping of a water-filled syringe. Where there was ambiguity or doubt, the decision as to whether or not anaesthesia occurred was taken by the senior anaesthetist present. Video recordings were available for review if necessary. Recovery endpoints included time of eye-opening on command and time of limb movement on command.
Target Controlled Infusion (TCI)
A TCI study was performed after completion of the Short Infusion study and appropriate evaluations of safety. An interim evaluation of the pharmacokinetic data of the subjects treated in the Short Infusion study was performed to provide the pharmacokinetic parameter estimates which were used to control the infusion rate of the pump during the TCI. The TCI was given by a Harvard 22 pump controlled by STANPUMP software running on a personal computer. The 30 minute TCI aimed to achieve stable arterial plasma concentrations providing for a stable period of anaesthesia. Stable anaesthesia was defined as anaesthesia lasting for a consecutive period of 10 minutes or more.
TCI infusion scheme
The 30 minute TCI was divided into two sections: 0-15 minutes and 15-30 minutes.
First TCI subject
The initial target concentration was based on the mean Org 25435 concentration occurring at the time at which eyes were opened on command in the short infusion study and evaluated using simulation to ensure that the total dose received (including any upwards titration), did not exceed 1.5 times the maximum dose given in the Short Infusion study.
If anaesthesia was achieved and maintained during the 0-15 minute section, the target concentration was to be maintained for the 15-30 min section; otherwise, after the first 0-15 minute section, the target concentration was to be increased by 50% for the remaining 15-30 min section.
Subsequent TCI subjects
The target concentration for the 0-15 minute infusion interval was to be set to the target concentration of the previous volunteer during the 15-30 minute infusion interval. However, if anaesthesia was always observed from the first volunteer onwards during the 0-15 minute infusion interval, the TCI target concentration during the 0-15 minute infusion interval for the next volunteer was to be set to 2/3 times the TCI target concentration of the previous volunteer during the 15-30 minute infusion interval.
Once stable anaesthesia was identified using the above rules in two volunteers at the same target concentration over the complete period of 30 minutes, the remaining volunteers to a maximum of seven were to be studied at target concentrations not more than 1.5 times that at which previous volunteers were studied. Dose escalation proceeded only after evaluation of safety and efficacy including review of all clinical and laboratory data (haematology and biochemistry) the response of subjects to dosing and comments from attending investigators.
The performance of the TCI model was assessed retrospectively by calculation of the bias, accuracy, divergence and wobble [7
Hartmann's Solution was infused intravenously (3 ml kg-1 hr-1) from at least 3 hours prior to the anticipated time of drug administration to ensure adequate hydration whilst fasting. An arterial cannula was inserted and a second venous cannula placed in an ante-cubital vein for infusion of Org 25435. Arterial blood samples (5 mL) were collected at 0, 1, 1.5, 2, 3, 4, 5, 8, 12, 16, 20, 45, 60, 120, 180, and 240 minutes after the start of infusion in the short infusion study and at 0, 3, 7, 11, 15, 18, 22, 26, 30, 30.5, 31, 32, 34, 37, 41, 45, 49, 59, 74, 89, 149, 209, and 269 minutes in the TCI study. The sample at 15 minutes was taken just prior to the potential alteration in target concentration. The sample at 30 minutes was taken just prior to stopping the infusion. A sample was also collected at the time eyes were opened to command.
Venous blood samples (5 mL) were collected at 60, 120, 180, 240, 360, and 720 minutes in the short infusion study and at 389, 509, and 749 minutes in the TCI study. Blood samples were processed to serum and stored at 20°C until analysis.
Urine was collected for 24 hours from subjects participating in the 4 mg/kg dose group of the Short Infusion study and in the TCI study.
After a stabilization period, during which baseline measurements were obtained, the anaesthetic infusion was started without notice. The times of eye closure, dropping water filled syringe, loss of eyelash reflex, eye opening on command and limb movement on command were recorded.
For safety reasons adverse events were recorded throughout the study.
Frontal electroencephalogram (differential recordings between FP1, FP2 locations international 10/20 system) was recorded from all volunteers using custom built equipment. Absolute power in two frequency bands, delta power (0.5-3.0 Hz) and theta power (3.25-8.0 Hz) were continuously calculated in subsequent epochs of 4 s.
Quantification of Org 25435
Org 25435 and its internal standard, Org 24446, were isolated from serum using solid-phase extraction. Briefly, 10 ng Org 24446 was added to each 0.5 mL serum sample and the sample was vortex mixed for 3 seconds. The solid phase extraction cartridge (C18 (end-capped), 50 mg, 1 mL (ICT)) was conditioned with 1 mL methanol, followed by 1 mL water and then the serum sample was applied to the cartridge and allowed to run through under vacuum. The cartridge was then washed twice with 1 mL water. Org 25435 and Org 24446 were eluted from the cartridge with 1 mL acetonitrile.
The extracted samples were quantified by liquid chromatography (LiChroCART® 55-2 Purospher® STAR RP-18 endcapped (3 μm)), coupled to a mass spectrometer (API 3000, Applied Biosystems) using turbo ion spray in multi reaction monitoring (MRM) mode, operating at 350°C and 5000 V. The mobile phase was run isocratically at a flow rate of 250 μL/min and consisted of 80% acetonitrile, 20% 0.01 mol/L ammonium acetate solution, pH = 4.2. The column oven temperature was set to 40°C. Operating conditions for the API 3000 were as follows: nebuliser gas 13, collision gas 5, turbo heater gas 8 L/min. The protonated M + H + peak of Org 25435 (m/z 369.9) was used as the precursor ion and in the MRM-mode m/z 174.2 was measured as the product ion. The protonated M+H + peak of Org 24446 (m/z 355.9) was used as the precursor ion and in the MRM-mode m/z 174.2 was measured as the product ion.
The lower limits of quantification of the assays were 5 ng/mL (serum) and 1 ng/mL (urine). The plasma assay was linear across the range 5 to 2000 ng/mL (r = 0.992). The plasma intra-assay coefficient of variation (CV) was less than 13% across the calibrated range, while the inter-assay CV was less than 15%. The urine assay was linear across the range 1 to 500 ng/mL (r = 0.997). The urine intra-assay coefficient of variation (CV) was less than 6% across the calibrated range, while the inter-assay CV was less than 12%.
Pharmacokinetic modelling and simulation
Moment analysis was performed using standard techniques. Mixed-effects population models [8
] were fitted to the Org25435 plasma concentration versus time data using the first order estimation method of the program NONMEM [9
]. Arterial plasma concentrations were used preferentially (over venous concentrations) where available. When not available e.g. time points post 240 minutes (post 269 minutes in the TCI studies) after administration of the study drug, venous concentrations were used for analysis. Three-compartment kinetics were assumed for the interim model whereas three- and four-compartment models were compared for the final model. Models were parameterized in clearances and volumes. Parameters describing inter- individual variability (etas) were tested for their significance one at a time and were retained in the final population model if statistically justified. The relationship evaluated was as follows: Pi
) where Pi
is the parameter value in the i
th subject, PTV is the typical value of the parameter in the population, and etai
is a random variable with a mean of 0 and a variance of omega 2
. Various residual errors models (proportional, additive and combined) were explored.
Two population models were produced:
1) An interim model describing Org25435 kinetics following short infusion. This was subsequently used for computer-controlled drug administration in the TCI study. The structural parameters (volume and clearances) for this model were assumed to vary according to volunteer body weight.
2) A final population model derived from Org25435 plasma concentration versus time data from both the short infusion and TCI studies. For the final population analysis which combined data from the short infusion study and the TCI study, the potential relationships between body weight and individual parameter estimates were explored.
The statistical significance of each proposed covariate-parameter relationship and the requirement for ETA parameters was assessed using the likelihood ratio test (where appropriate i.e. for nested models) and by consideration of the Akaike Information Criterion (non-nested models) and the precision of the final parameter estimates (all models). For nested models, the justification for each additional effect was for it to improve the goodness-of-fit statistic (-2 log likelihood) by more than 6.6 (evaluated against the chi-square distribution, this is equivalent to significance at the 0.01 level). The improvement (or lack of) in model goodness-of-fit was also assessed visually by the examination of diagnostic plots. After completing the model build, the necessity for each added component was re-assessed by removing it from the model and evaluating the resulting impact on the model fit. If the removed component caused an increase in the goodness-of-fit statistic equivalent to significance at the 0.005 level (i.e. the removal of the component significantly worsened the model), it was allowed to remain.
A sequential approach was taken to pharmacokinetic-pharmacodynamic modelling. That is, the final population pharmacokinetic model was used to derive individualised (post-hoc Bayesian) pharmacokinetic parameter estimates (e.g. rate constants and central compartment volume) for each volunteer in the short infusion groups. These were used as inputs, along with the drug dose and effect measure data, for the estimation of pharmacodynamic parameters. The absolute power in the theta band (3.25-8.0 Hz) was used as an effect measure and was described using a sigmoid Emax
model of Org 25435 concentration (CE) in an effect site compartment, according to the following equation:
where E0 is the baseline value, Emax is the maximum response, EC 50 is the effect site concentration corresponding to half of the maximum effect, and γ (Hill coefficient) describes the steepness of the drug concentration-effect relationship. The incorporation of a hypothetical effect compartment accounted for the hysteresis (temporal delay) between serum Org 25435 concentration and the onset of drug effect. This delay was characterised by the estimation of Ke0, the blood-brain equilibration rate constant. NONMEM's first order conditional estimation method was used. Random residual error was described using an additive model. A proportional variance model was used to describe the inter-individual variability in Emax, EC50 and Ke0 and E0.