| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
GPI 15715 is the first water-soluble propofol prodrug that has been studied in humans. Present propofol lipid formulations have well known undesirable properties, for example, pain on injection and increased triglyceride concentrations. We investigated whether GPI 15715 is suitable to achieve and maintain moderate sedation for 2 h. Six male and six female volunteers received a target-controlled infusion of GPI 15715, with an initial propofol target concentration of 1.8 µg/mL and the possibility to adjust the propofol target once after 1 h. Propofol concentrations, the bispectral index, and modified Observers Assessment of Alertness/Sedation Scale (MOAA/S) scores were monitored. The median MOAA/S score was 4 during the first hour and was 3 during the second hour of infusion. The propofol target had to be changed to 2.4 µg/mL in seven volunteers and to 3.0 µg/mL in two volunteers. A propofol concentration of 1.9 µg/mL had the highest probability to result in an MOAA/S score of 3, which corresponds with moderate sedation. We observed no serious side effects. We conclude that GPI 15715 produces excellent sedation.
Because of its unique pharmacological properties, propofol is increasingly used to provide procedural sedation, for example, during endoscopic examinations (1), even in critically ill (2) and pediatric (3) patients. Because propofol has a very low solubility in water, it is formulated as a lipid emulsion. The marketed propofol lipid formulations have some undesirable properties, including pain on injection (4), a risk of infection because of microbial contamination (5), and increasing triglyceride concentrations (6). GPI 15715 (phosphono-2,6-diisopropylphenol) is a water-soluble prodrug of propofol that is hydrolyzed by alkaline phosphatase to liberate propofol, formaldehyde, and phosphate. In a previous study, GPI 15715 was investigated for the induction of anesthesia (7). The primary goal of this study was to investigate the suitability of GPI 15715 for moderate or procedural sedation. The use of a target-controlled infusion (TCI) to provide intraoperative sedation may have advantages when compared with manual dosing regimens (8). Therefore, we administered GPI 15715 as a TCI over 2 h to investigate whether a TCI approach can be successfully used to achieve and maintain a preselected level of sedation. As a secondary goal, measured propofol concentrations were correlated with a modified Observers Assessment of Alertness and Sedation Scale (MOAA/S) (9). The electroencephalogram (EEG) was monitored by using the bispectral index (BIS).
The study protocol was approved by the local medical ethics review board. Six female (age: mean, 28 yr; range, 2432 yr; body weight: mean, 57 kg; range, 5161 kg) and 6 male (age: mean, 32 yr; range, 2640 yr; body weight: mean, 78 kg; range, 6893 kg) volunteers gave written, informed consent to participate in this investigation. The subjects had to be healthy as assessed by medical history and physical and laboratory examinations. Exclusion criteria were a body weight of >20% outside the normal range (according to the Metropolitan Life Insurance Company Table); documented drug allergies; alcohol, drug, or medication abuse; and recent use of a medication with an effect on the central nervous system. All volunteers had to agree to use an effective method of birth control between the time of screening and the end of the study, and a laboratory and urine test for pregnancy had to be negative. Before study drug administration, subjects were required to stop consuming alcohol or caffeine for 24 h and to fast for at least 8 h. Within 2 h before the start of drug infusion, a physical examination and laboratory tests were repeated. Baseline values for arterial blood pressure, heart rate, body temperature, and oxygen saturation (Sao2) were taken, and the electrocardiogram was recorded. Heart rate, arterial blood pressure, electrocardiogram, and Sao2 were continuously monitored until 6 h after the start of the infusion. When Sao2 values decreased to less than 93%, the volunteers received oxygen 34 L/min via a nasal catheter. A venous catheter was inserted into the dominant arm for drug infusion and was constantly flushed with normal saline. An arterial catheter was inserted in a radial artery of the nondominant arm for collection of arterial blood samples. To quantify depth of sedation, the MOAA/S was used (9). The subjects response was scored from 5 (awake) to 0 (unconscious) as follows: 5, subject responds readily to name spoken in a normal tone; 4, lethargic response to name spoken in a normal tone; 3, responds only after name is called loudly, repeatedly, or both; 2, responds only after mild prodding or shaking; 1, responds only after painful trapezius squeeze; 0, does not respond to painful trapezius squeeze. During the GPI 15715 infusion, the MOAA/S score was assessed at 7, 12, 22, 32, 42, 52, 62, 67, 72, 82, 92, 102, and 112 min. After cessation of the infusion, the MOAA/S score was assessed every 2 min until 4 consecutive scores of 5 had been recorded. An MOAA/S score of no less than 2 and not more than 3 was targeted as the desired effect of moderate sedation. All volunteers received a TCI of GPI 15715 with an initial propofol plasma target concentration of 1.8 µg/mL for the first hour. The MOAA/S score measured at 52 min was used to determine whether an adjustment of the target concentration was necessary for the second hour. If the MOAA/S score was <2 or >3, the target was adjusted as follows: for an MOAA/S score of 5, increase to 3 µg/mL; for an MOAA/S score of 4, increase to 2.4 µg/mL; for an MOAA/S score of 1, decrease to 1.4 µg/mL; and for an MOAA/S score of 0, decrease to 1.4 µg/mL. This new propofol target concentration was kept constant during the second hour of infusion, and after 2 h, the infusion was stopped. TCI was accomplished with an infusion pump (Braun Perfusor fm; Braun, Melsungen, Germany) and self-programmed control software (IvFeed 5.0; Department of Anesthesiology, Erlangen, Germany) running on a notebook computer. The infusion rate was calculated by using the pharmacokinetic model determined in a previous study (7). Table 1 shows the pharmacokinetic variables of this model, which consisted of a two-compartment model for the parent drug GPI 15715 and a three-compartment model for the metabolite propofol.
Arterial blood samples of 3 mL each to determine the concentrations of GPI 15715 and propofol were collected at baseline and at 5, 10, 20, 30, 40, 50, 60, 65, 70, 80, 90, 100, 110, 120, 125, 130, 140, 150, 170, 180, 200, 220, and 240 min. The samples were collected in sodium heparin glass vials that were prefilled with 30 mg of sodium orthovanadate to block alkaline phosphatase activity. Samples were stored on ice for up to 1 h, and after centrifugation at 1200 rpm for 15 min, the plasma was transferred into transport vials and stored at 70°C. Propofol analysis was performed with reverse-phase high-performance liquid chromatography with fluorescence detection, modified from the method of Plummer (10). For evaluation of the accuracy of the TCI system, we used the measured (cm) and predicted (cp) arterial propofol concentrations to calculate the prediction error [PE = (cm cp)/cp], the median PE (MDPE), and the median absolute PE (MDAPE). As a measure for a time-related trend of the PE, we determined in each volunteer the divergence as the slope of the linear regression of |PE| versus time (11). As a measure for the intraindividual variability of the PE, we calculated for each volunteer the wobble [W = median (|PEi MDPE|, i = 1,...n)], where n is the number of concentration measurements in the volunteer (11).
Starting 30 min before drug administration, an EEG was continuously recorded until 6 h after the start of the GPI 15715 infusion. An Aspect A-1000® monitor (software Version 3.31) and pregelled silver/silver chloride electrodes (Zipprep®; both Aspect Medical Systems International, Leiden, The Netherlands) were used to record the BIS. Electrodes were placed at the frontal positions Fp1, Fp2, and Fpz (reference) and at the nasion (ground; international 10-20 system of electrode placement). Electrode impedance was kept <2000
The MOAA/S scores were dichotomized for logistic regression analysis. Logit analysis was performed by using the measured arterial propofol concentrations and the MOAA/S score values. Two logit functions were fitted for an MOAA/S score Data are reported as mean ± sd together with the range or the 95% confidence interval (CI), unless stated otherwise. To assess differences between the first and second hour of sedation, we first calculated for each volunteer the mean or median value of the relevant variable during the first and second hour. These individual values were then tested with the Students t-test for paired samples or Wilcoxons matched pairs test, respectively. The relative changes of the hemodynamic variables during the first and second hour of sedation were tested for significant differences compared with baseline values by using the one-sample Students t-test with a reference value of 0. Statistical and graphical analysis were performed with GraphPad Prism Version 4.0a for Macintosh (GraphPad Software, San Diego, CA).
All volunteers successfully completed the study in accordance with the study protocol. Figure 1 shows the time course of the measured propofol concentrations and shows the BIS and the MOAA/S scores in the individual volunteers. Table 2 shows the mean arterial propofol concentrations and the median MOAA/S scores for all volunteers. Because the median MOAA/S score at 52 min was 4 (range, 25), the propofol target concentration had to be increased to 2.4 µg/mL in 7 volunteers and to 3.0 µg/mL in 2 volunteers according to the study protocol. In these volunteers, an increased level of sedation was reached within an average of 3 min after the target was changed. In 3 volunteers, the target concentration of 1.8 µg/mL was maintained, and it was not necessary to reduce the target to 1.4 µg/mL in any volunteer. Whereas the median MOAA/S score was 4 during the first hour, the median MOAA/S score was 3 during the second hour after the single target adjustment at 1 h. BIS values changed in parallel with increasing or decreasing MOAA/S scores, with a mean BIS value of 72 ± 12 during the first hour and 61 ± 11 during the second hour of infusion (P < 0.005). The coefficient of variation of BIS was 9.8% ± 6.8% in the first hour and 11.1% ± 4.5% in the second hour. The average time for recovery to an MOAA/S score of 5 after termination of the GPI 15715 infusion was 18 min (95% CI, 1620 min). Figure 2 and Table 3 summarize MOAA/S scores and measured propofol concentrations, which showed a significant but only moderate correlation (Spearman r = 0.54; P < 0.0001), because there was a considerable overlap in the concentrations.
Figure 3 shows the probability curves of the logistic regression for an MOAA/S score of
The measured propofol concentrations were smaller than the propofol target concentrations. The MDPE and MDAPE are given in Table 4. The PE was significantly smaller during the second hour. The divergence was 18.0% ± 7.1% per hour. The W was 18.0% ± 5.8% during the first hour and 6.8% ± 3.7% during the second hour (P < 0.001).
Values of systolic blood pressure, heart rate, and Sao2 during sedation with GPI 15715 are shown in Figure 4. Systolic blood pressure decreased significantly by 11.4% (95% CI, 9.0% to 13.9%; P < 0.001) during the first hour and by 16.3% (95% CI, 14.5% to 18.1%; P < 0.001) during the second hour compared with baseline values. Heart rate increased significantly by 5.6% (95% CI, 1.0%10.2%; P < 0.05) during the first hour and by 5.2% (95% CI, 0.2%10.0%; P < 0.05) during the second hour of infusion. Sao2 decreased to values less than 93% in 4 volunteers (1 with a target concentration of 1.8 µg/mL, 2 with a target concentration of 2.4 µg/mL, and 1 with a target concentration of 3.0 µg/mL). The minimum Sao2 in these individuals was 90%92%. The desaturation could be removed with oxygen via the nasal cannula and did not last longer than 1 min. There were no clinically significant laboratory findings or serious adverse events, and there was no increase in formate concentrations compared with baseline values during GPI 15715 infusion. Eleven of 12 volunteers experienced paresthesia or a burning sensation at the lower and upper body or in the perineal region. This sensation started within 1 min after the start of the TCI infusion and lasted for approximately 2 min. It was rated as moderate in seven volunteers and as mild in four volunteers.
The IV anesthetic propofol is commercially available in slightly different lipid emulsions and different propofol concentrations. All marketed propofol lipid emulsions share common side effects, especially pain on injection (4), which has an incidence of approximately 28%37% (13), when lipid solutions of long-chain and medium-chain triglycerides are administered. Also, the use of propofol lipid emulsions increases the risk of bacterial contamination and infection (5). Even after intraoperative infusion of propofol lipid formulations, triglyceride concentrations may increase (6), and it is not known whether the lipid solvents influence the incidence and severity of propofol infusion syndrome (14). Therefore, there is considerable interest in the development of lipid-free propofol formulations (15), structural propofol analogs with anesthetic effects (16), or propofol prodrugs (17). GPI 15715 is the first water-soluble propofol prodrug that has been studied for the induction and maintenance of anesthesia in humans (7). In this study, we investigated the clinical pharmacodynamics of GPI 15715 by using a TCI to provide and maintain moderate sedation for two hours. The OAA/S scale (9) has been validated to quantify the sedative drug effect of propofol (18). We did not use the original, but a modified, sedation scale, which was, however, used in other studies on propofol sedation (19,20). For propofol target concentrations to achieve preselected OAA/S scores, Casati et al. (19) described well separated CIs, whereas for measured propofol concentrations in this study, we found a considerable overlap in and, in consequence, only a moderate correlation between propofol concentration and sedation scores. A relatively large interindividual variability of propofol pharmacodynamics was, however, also reported in previous studies (21). The propofol concentrations for an MOAA/S score of 4 and 3 found in this study are comparable to those of Casati et al. (19), but we found smaller concentrations for MOAA/S scores of 2, 1, and 0, i.e., deeper levels of sedation. However, our results are in good agreement with the measured venous plasma concentrations described by Doufas et al. (22), who found a propofol concentration of 1.9 ± 0.7 µg/mL for an OAA/S score of 1 or 2 and a concentration of 1.7 ± 0.7 µg/mL for an OAA/S score of 3.
In addition, we monitored the BIS, which has been used to measure and quantify the sedative drug effect of propofol (23,24). We found mean BIS values of 72 ± 12 for a median MOAA/S score of 4 during the first hour and of 61 ± 11 for a median MOAA/S score of 3 during the second hour of infusion. In general, BIS values measured in our investigation were at least 20% less than the BIS values described by Liu at al. (23) for comparable OAA/S scores. However, Doufas et al. (22) described fairly comparable results. They found BIS values of approximately 61 for an OAA/S score Although the TCI systematically overpredicted the achieved propofol concentrations, especially during the first hour of infusion, the overall MDPE of 17.7% is in the typical range for TCI systems of propofol (25). Vuyk et al. (26) found that different pharmacokinetic models for Diprivan® underpredicted the achieved propofol concentrations, with an MDPE between 19% (for the most accurate model) and 112% (for the least accurate model). The general overprediction and the smaller PE in the second hour, as well as the divergence of 18% per hour, indicate that there is probably a model misspecification in the central volume and in the variables k12 and k13, which describe the initial distribution into peripheral compartments. The smaller W in the second hour means that the propofol concentrations were more stable, and because the clearance is the essential pharmacokinetic variable for longer infusion, this variable therefore seems to be more accurate in the present pharmacokinetic model. In spite of the inaccuracy of the TCI system in our study, the targeted pharmacodynamic effect could be achieved within an average of three minutes after a single change of the TCI target concentration and was maintained for one hour. The rapid change of the pharmacodynamic effect was, however, partly caused by an overshoot of the propofol concentration after the target had been changed. After a 2-hour infusion of GPI 15715, the mean recovery time to an MOAA/S score of 5 was 18 minutes, which is approximately 10 minutes longer compared with the propofol lipid formulation (27). We found moderate effects on systolic blood pressure and heart rate; these were similar to the changes described by Leslie et al. (20) for sedation with propofol emulsion during colonoscopy. As with fosphenytoin, another marketed prodrug with a phosphono-O-methyl group (28), GPI 15715 caused a mild to moderate feeling of paresthesia or a burning sensation at the upper and lower body or perineal region, but in general this was well tolerated. In summary, GPI 15715 appears to be suitable to provide moderate or procedural sedation, and the presented TCI system allowed easy titration of the sedative effect. The substance also merits further investigation for sedation in intensive care unit patients.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|