Pharmacodynamics and Pharmacokinetics of Propofol in a Medium-Chain Triglyceride Emulsion

The study design was randomized and double-blind with a two-period crossover. Two drug administration protocols were used with 12 different subjects in each. Subjects in protocol 1 received a single bolus of 2.5 mg/kg, and those in protocol 2 received the same induction dose followed by a 30-min infusion at 0.2 mg · kg⁻¹· min⁻¹. The study was approved by the institutional review committee, and informed consent was obtained from all subjects. Each subject received both drugs in a crossover fashion, and the order (sequence) of the drug administration was randomized. At least 1 week was allowed for washout of drug and metabolites between experiments. Because of a slight difference in the color of the drugs and the different concentrations, one unblinded investigator infused the drugs into the subject out of sight of the other blinded investigators.

The subjects were admitted to the general clinical research unit the evening prior to drug administration and fasted after midnight. On the morning of the drug administration, an intravenous catheter was placed in a large forearm vein and a second one was placed in the antecubital fossa of the contralateral arm for blood sampling. Subjects were monitored with electrocardiogram, noninvasive blood pressure, and pulse oximetry, and oxygen was provided via  an anesthesia machine. Manual positive pressure ventilation using a circle system was provided as needed to maintain a minimum respiratory rate of 4 breaths/min. In protocol 2, the electroencephalogram was monitored with the bispectral index (BIS; BIS Model A-2000, software version 2.1 upgraded to 2.21; Aspect Medical Systems, Newton, MA).

Subjects in both protocols received an induction dose of 2.5 mg/kg at a rate of 4 mg/s (manually injected). This induction dose was followed in protocol 2 by a 30-min infusion at a rate of 0.2 mg · kg⁻¹· min⁻¹via  syringe infusion pump. During induction, any complaints of pain were noted by one of the blinded investigators and scored as none, mild, moderate, or severe. The subject was asked to count out loud, and the time after the start of injection when the subject stopped counting was recorded as the induction time. After the bolus (protocol 1) or the termination of the infusion (protocol 2), the subject was asked, without any physical stimulation, to open his or her eyes at 1-min intervals by the same blinded investigator, and the time to following this command was recorded as the emergence time. If the subject spontaneously opened his or her eyes, then that time was recorded as the emergence time. Thirty minutes after recovery, the subject was asked to rate any pain with the drug injection, using a 0 (no pain) to 10 (worst possible pain) visual analog scale.

In protocol 1, blood samples were taken at time 0 (prior to the induction dose) and at 1, 2, 3, 4, 6, 8, 10, 20, 30, 45, 60, and 120 min after the start of the injection. In protocol 2, blood samples were taken at time 0 and at 1, 2, 5, 10, 20, 30, 32, 34, 36, 40, 50, 60, and 120 min after the start of the injection. In protocol 2, samples were also taken at time 0 and 30 min and at 1, 2, and 4 h for analysis of plasma triglycerides, cholesterol, high-density lipoprotein, lactate, β-hydroxybutyrate, acetoacetic acid, and octanoic acid (octanoate).

Blood samples were placed on ice, centrifuged and frozen immediately after they were obtained, and analyzed either in the clinical laboratories of the University of Rochester or sent to reference laboratories (propofol to Pharmakinetics, Baltimore, MD, and lactate, β-hydroxybutyrate, acetoacetate, and octanoate to the Metabolic Screening Laboratory, Dr. James D. Shoemaker, M.D., Ph.D., Associate Professor of Biochemistry and Molecular Biology, Director, Metabolic Screening Laboratory, St. Louis University, St. Louis, MO).

The sample size of 12 subjects in each protocol (total of 24 subjects) was considered sufficient to determine the pharmacokinetic profile of IDD-D Propofol and bioavailability relative to Diprivan without exposing an undue number of subjects to the new preparation. The study was designed to compare the induction and emergence times as the primary pharmacologic effects, but the measurement of the BIS was also used as a secondary outcome in protocol 2. Comparison of the pharmacokinetics was made by calculating the area under the curve, clearance, and volume of distribution, and in addition a two-compartment (biexponential) model (calculations provided by Covance, Inc., Madison, WI, using WinNonlin Professional, Pharsight Corp., Mountain View, CA) was fitted to the data and described by the half-time and volume of the fast (T1/2α[minutes] and V¹[liters per kilogram]) and slow (T1/2β[minutes] and V²[liters per kilogram]) components. The plasma sample at 1 min was not used in this calculation since, for many subjects, the bolus infusion was still in progress at that time.

Data are reported as mean ± SD except as noted. Statistical analysis was performed using STATA statistical package (Stata Corp., College Station, TX). Comparison between drugs was done via  two-way (drug and subject) analysis of variance. Gender analysis was performed by analysis of variance with these factors: gender, subject nested within gender, drug, and drug-by-gender interaction. P  values without correction for multiple comparisons are reported. P  values > 0.05 are reported as not significant.

All subjects tolerated both drugs well and completed both arms of the study. In protocol 1, the mean age of the 12 subjects (8 men and 4 women) was 35 ± 9.9 yr, and the mean weight was 76 ± 15 kg. In protocol 2, the mean age of the 12 subjects (6 men and 6 women) was 26 ± 8.7 yr, and the mean weight was 75 ± 12 kg. There were no adverse effects, and generally the drugs were clinically indistinguishable to the blinded investigators.

Overall, IDD-D Propofol showed higher pain scores, but no subject's pain was rated as severe. Since the subjects’ rating of pain may be lessened in protocol 2 because of the length of time between induction and the rating, only the observer's rating of pain is combined between protocols 1 and 2. For the observer's rating of pain on induction (single bolus) in both protocols, with Diprivan, 21 subjects were rated as having no pain and 3 with mild pain, but with IDD-D Propofol, 11 were rated with no pain, while 10 had mild pain and 3 had moderate pain (P = 0.007, Pearson chi-square test). The pain visual analog scale scores for protocol 1 (single bolus) with IDD-D Propofol was 25.5 ± 14.1 and with Diprivan was 6.1 ± 8.5 (P = 0.0006), but for protocol 2 (single bolus plus 30-min infusion) the visual analog scale scores were lower and not significantly different (9.4 ± 16.5 and 3.1 ± 8.1 for IDD-D Propofol and Diprivan, respectively).

The averages of the plasma concentrations for protocol 1 are shown in figure 1, and the pharmacokinetic parameters are given in table 1. Missing samples (because of inability to withdraw the blood) in one subject prevented calculation of pharmacokinetic parameters in this subject. There were no apparent or statistical differences in the pharmacokinetics between the two drugs.

In one subject in protocol 2 who received IDD-D Propofol, there was an infusion pump error that caused the subject to receive an infusion dose of 0.131 mg · kg⁻¹· min⁻¹. The data from that subject are excluded from the pharmacokinetic analysis where noted by the indication of 11 subjects. Figure 2shows the plot of the individual plasma concentrations for Diprivan and IDD-D Propofol for protocol 2. Again, there were no statistical differences between the concentrations at any time point. The results for the pharmacokinetic analysis are given in table 1. There was no significant difference by drug for any of the pharmacokinetic parameters.

Both drugs caused the expected decrease in blood pressure with induction and infusion, but there was no significant difference. Mean baseline blood pressure was 92 ± 12 mmHg for both IDD-D Propofol and Diprivan; blood pressure decreased to 58 ± 9 mmHg in the IDD-D Propofol group (n = 11) and to 60 ± 9 in the Diprivan group (n = 12) at the end of the infusion. The heart rate was unchanged from baseline for both drugs.

Because the same induction dose was used in both protocols, the analysis of induction time was combined (N = 24) and was 53.3 ± 12.1 s for IDD-D Propofol and 46.9 ± 7.8 s for Diprivan (P = 0.002). The emergence times were also slightly longer for IDD-D Propofol. In protocol 1, it was 575 ± 228 s for IDD-D Propofol and 546 ± 216 s for Diprivan (nonsignificant); in protocol 2, it was 1,197 ± 445 s (n = 11) for IDD-D Propofol and 1,005 ± 422 s (n = 12) for Diprivan (P = 0.07). Interestingly, if the subject in the IDD-D Propofol group in protocol 2 who received the low dose in error is included, the emergence time increases to 1,254 ± 468 s, and the difference in times barely reaches statistical significance (P = 0.04). This is because this subject had a relatively longer emergence time despite the lower dose. Since it would be expected that the correct larger dose would have resulted in a longer emergence time, it would appear to be conservative to include this data.

The data from the BIS monitor in protocol 2 was summarized by ensemble averaging the BIS readings after interpolating at 5-s intervals. BIS data were not obtained in one subject in the IDD-D Propofol group because of technical problems with recording the BIS data. The subject who received the erroneous infusion rate was not included; therefore, n = 10 for the IDD-D Propofol group, and n = 12 for the Diprivan group. The values and the time course of the BIS readings were similar for both drugs (fig. 3). The time to minimum BIS and the minimum BIS value were 117 ± 38 s and 16 ± 7 for IDD-D Propofol and 113 ± 27 s and 17 ± 6 for Diprivan, respectively (nonsignificant for both). At the time of the infusion termination, the average BIS reading (averaged over the last minute of data prior to the end of the infusion) was not significantly different: 34.2 ± 3.9 for IDD-D Propofol and 36.2 ± 3.9 for Diprivan.

The plasma concentrations of triglycerides confirmed the expected differences (fig. 4). Octanoate is a significant metabolite of medium-chain triglycerides; it was elevated in the IDD-D Propofol group but not in the Diprivan group. In one of the male subjects, the reference laboratory reported elevated concentrations of octanoate when he received Diprivan but not when he received IDD-D Propofol. In this subject, the triglyceride value (performed on a different sample in the University of Rochester clinical laboratories) was apparently correct (elevated during Diprivan and unchanged for IDD-D Propofol). While the statistical results are not affected regardless of how this subject's data are handled, for the purposes of this analysis, we deleted the data from this subject assuming that it was erroneous. The data from the subject receiving the incorrect infusion rate of IDD-D Propofol is not included. Plasma concentrations of cholesterol, high-density lipoprotein, lactate, acetoacetate, or β-hydroxybutyrate did not demonstrate any large or consistent changes with either formulation of propofol.

It has been reported that there are gender differences affecting the awakening time from propofol. 10The plasma concentrations for each subject shown in figure 2indicates that the female subjects tended to have a higher initial peak but a lower final concentration. Since there was no difference between the two drugs, the data were combined and plotted by gender in figure 5. While the initial peak was higher in females, the difference was not significantly different; however, the plasma concentrations at 20 and 30 min were significantly lower in female subjects, and this difference persisted after the infusion was terminated. Table 1also gives the comparison between female and male subjects for calculated pharmacokinetic parameters; only V²did not show a statistical difference.

Table 2compares the awakening time, plasma concentration at the end of the infusion, and the BIS level at the same time between male and female subjects. There was a highly significant difference in the wake-up times between the male and female subjects. This difference was greater than the difference between the two drugs. There was also a significant difference in the plasma concentration but not in the BIS levels at the time of the termination of the infusion.

Propofol has become the induction agent of choice when speed of awakening without any “hangover” is important. 1It is also increasingly administered as a continuous infusion to maintain anesthesia or provide sedation for patients requiring intensive care. Since the drug is extremely insoluble in water, it is necessary to formulate it in a lipid-based emulsion. The currently marketed products contain propofol as a 1% emulsion in a soybean-based lipid. This formulation contains the long-chain triglycerides from the soybean base. While this formulation has been clinically very useful, it has some drawbacks, including pain on injection, elevation of plasma triglyceride concentrations during prolonged infusion, and the risk of bacterial contamination. 1,2 

This new RTP formulation of propofol attempts to address some of these issues by changing the emulsion base. In this formulation, propofol is a 2% oil-in-water emulsion stabilized with phosolipids, where the oil phase comprises a mixture of medium-chain triglycerides (C⁸and C¹⁰fatty acids, primarily caprylic and capric acids). The lack of long-chain triglycerides in the formulation precludes their elevation in the plasma following a prolonged infusion. However, this new formulation may result in changes in pharmacodynamic or pharmacokinetic properties of propofol.

The plasma concentrations for both drugs were similar in both protocols, indicating no pharmacokinetic difference between the two formulations. Although peak drug concentrations following the initial bolus injection were higher for IDD-D Propofol, these differences were not significant because of the expected individual variability. None of the calculated pharmacokinetic parameters showed a significant difference, and the values are similar to those reported by Shafer et al. , 11who also analyzed venous samples.

The induction times were similar. Although the 6.4-s longer time with IDD-D Propofol was statistically longer, it is not clinically significant, and the time to the peak BIS effect was the same for both drugs. Both the induction times and the time to minimum BIS are similar to those found by Flaishon et al.  12following a 2-mg/kg induction dose given over 20 s. An explanation for this difference may also be found in the work of Kazama et al.  13In a study of induction time with a wide range of induction infusion rates using 1% undiluted and 0.05% diluted propofol, they found that the more concentrated propofol had a longer induction time. In our protocol, the induction dose of the drug was calculated on a milligram-per-kilogram basis (2.5 mg/kg), but the speed of injection was not (4 mg/s). This resulted in induction dose rates (protocols 1 and 2 combined) of 196 ± 33 and 267 ± 45 mg · kg⁻¹· h⁻¹adjusted for lean body mass (calculated using the formulas given by Kazama et al.  13; at approximately these rates, Kazama et al.  13found an induction time of 51 ± 5 s for undiluted propofol and 34.1 ± 3.2 s for the diluted propofol. If these results apply to the more concentrated IDD-D Propofol, then the longer induction time may be largely explained by this concentration difference. In addition, Van Hemelrijck et al.  14gave Diprivan at a slightly slower rate than we did and found the induction time to be 75 ± 25 s.

While the difference in emergence times was only marginally statistically significant in protocol 2 and not significant in protocol 1, there was a tendency for IDD-D Propofol to prolong emergence slightly. However, a complete interpretation of this result must be tempered by the fact that we obtained only venous plasma samples and that there may be a significant difference between arterial and venous propofol during (arterial greater than venous) and after (venous greater than arterial) continuous infusions. 15Determining the pharmacodynamics of a drug with only venous samples may lead to erroneous conclusions. 16However, the longer awaking time for IDD-D Propofol may indicate that it is slightly more potent.

While in our protocol we did not find any major differences between IDD-D Propofol and Diprivan, we did observe large differences between our male and female subjects. The emergence time was faster for women, and the plasma concentrations at the termination of the infusion were statistically significantly lower. This leads us to conclude that this gender difference is primarily a result of pharmacokinetic differences, which are shown by the differences in pharmacokinetic parameters (table 1). However, since the BIS levels at the termination of the infusion were the same for these different plasma concentrations, there may also be a gender-related difference in pharmacodynamic effect. Stepwise multiple regression analysis of the awakening time showed that gender (P = 0.002) and BIS at the time of termination of the infusion (P = 0.02) and drug (P = 0.07) were the most significant factors, while the octanoate and drug concentrations at the end of the infusion and lean body mass were not significant. While some previous reports have not shown gender to be a significant cofactor (in addition to lean body mass) in a pharmacokinetic analysis, 17others have. 11It seems that the gender difference may be both a pharmacokinetic and a pharmacodynamic effect. 10,18In this study, the gender-related difference was larger than the differences found between the propofol formulations.

Medium-chain triglyceride administration to humans has been extensively studied and is without significant toxicity when given either orally and parenterally. 9However, there has been some indication that octanoate concentrations of 3–8 μmol/ml in animals 9can cause somnolence and coma, and infusion of a large dose (plasma concentrations > 10 μmol/ml) of octanoate in Rhesus monkeys also caused somnolence and coma. 19However, a 6-h infusion in human volunteers of a 10% solution of a lipid emulsion containing 75% medium-chain triglycerides was without any neurologic effects, and the plasma concentrations reached 2–3 μmol/ml. 20The highest concentration seen in our study, 0.189 μmol/ml, is well below these concentrations. While the octanoate concentration has returned almost to normal by 90 min after the termination of the infusion, clinical trials with larger doses (longer infusions) will be needed to determine the pharmacokinetics of this metabolite; it will be important to monitor for any possible neurologic effects. Although large doses of medium-chain triglycerides can result in the formation of ketone bodies, there was no elevation in β-hydroxybutyrate or acetoacetate seen with IDD-D Propofol.

Pain on injection was not one of our primary outcome variables, but there seemed to be slightly more pain on injection with IDD-D Propofol. This difference in reported pain might have been caused by an increased amount of free propofol in the 2% formulation, since pain has been related to the amount of free propofol in the injectate. 3In contrast to the common reporting of severe pain in the operating room, the general pain concentrations were low in this study, most likely because of the use of a large antecubital vein. We did not use any common clinical technique to reduce pain (e.g. , mixing with lidocaine). The fact that the injection was made relatively slowly may have resulted in less pain than might occur with a more rapid injection into a more peripheral vein. 4Larger clinical trials, using techniques more commonly used in clinical practice (e. g. , use of a small hand vein) will be needed to determine if there is more injection pain with this formulation and if it is caused by the higher propofol concentration.

In this phase 1 trial involving 24 normal, healthy subjects, we found that the new formulation of propofol was well tolerated and generally indistinguishable from one of the current propofol formulations (Diprivan 1%). We did find strong gender differences in emergence time and pharmacokinetics that were similar for both drugs. Since the new formulation may have significant advantages, further clinical trials are warranted.

The authors thank Ann Holberton, R.N., and Linda Palmer, R.N. (Clinical Study Coordinators, Department of Anesthesiology, University of Rochester, Rochester, NY), for help in the organization and performance of this research.