Propofol is widely used internationally for the provision of general anesthesia and sedation. It has many desirable qualities including rapid onset, short context-sensitive half-time, and a lower incidence of postoperative nausea and vomiting in comparison with inhaled anesthesia. The currently marketed propofol formulation is an oil in water macroemulsion of soybean oil, glycerol, and egg lecithin.1 This lipid-based formulation of propofol presents some limitations for the drug including pain on injection,2,3 risk of hyperlipidemia,4 allergic reactions,5 and support of microbial growth with the potential for postoperative infection.6,7 These limitations have fueled research efforts to reformulate propofol using lipid-free strategies.
A new, water-soluble preparation of propofol has been developed containing the drug, sulfobutylether ß-cyclodextrin (Captisol®, CyDex Pharmaceuticals Inc., Lenexa, Kansas) and water. Sulfobutylether ß-cyclodextrin acts by solubilizing and stabilizing the propofol molecule by providing a lipophilic core in which the propofol can form noncovalent complexes. This modified cyclodextrin-based formulation of propofol may have some advantages over the current formulation by mitigating some of the problems associated with the lipid-based formulation. A study in a porcine model suggests that the pharmacokinetics and pharmacodynamics of the propofol in cyclodextrin formulation are substantially similar to the propofol in lipid formulation.8
The primary objective of this first human study was to compare the effects of propofol (10 mg · mL−1) in the lipid formulation (Diprivan®, AstraZeneca, Wilmington, DE) with those of the new cyclodextrin formulation, particularly with regard to pain on injection. We hypothesized that the propofol in cyclodextrin would be associated with less pain on injection with similar sedative, hemodynamic, and respiratory effects compared with propofol in lipid.
The University of Utah IRB approved the study, and all subjects gave written, informed consent. Volunteers were healthy, nonsmoking males or nonpregnant, nonbreastfeeding females between the ages of 18 and 50 years. All subjects were within 50% of ideal body weight. Volunteers who had medical problems, abnormal laboratory findings with clinical significance, or any known or suspected hypersensitivity to any compound present in the study medications were excluded.
Study Design and Procedures
The study was a single-center, double-blind, 2-period, randomized, dose-escalating study using a completely balanced cross-over design. In the first period, each subject was randomly assigned to receive either propofol in cyclodextrin (Captisol-enabled propofol) or propofol in lipid (Diprivan) administered IV over 30 seconds at doses of 0.125, 0.25, 0.5, or 1.0 mg · kg−1 using a syringe pump (Harvard Apparatus Model 33, Instech Labs, Plymouth Meeting, PA). The treatment sequence was determined by block randomization. After a 3-hour washout period, the second drug was given at the same dose.
The initial 4 subjects were studied at the lowest dose of 0.125 mg · kg−1. If this dose did not produce a brief loss of consciousness in 4 of 4 volunteers with one of the formulations, then the next largest dose was administered to the next group of 4 subjects. Brief loss of consciousness was defined as a value of <50 on the bispectral index (BIS, Aspect Medical, Norwood, MA) of the electroencephalogram (EEG). The dose escalation continued up to a maximum of 1.0 mg · kg−1 or until 4 of 4 subjects exhibited a brief loss of consciousness from the bolus dose of one of the formulations.
A 20-gauge venous catheter was placed on the dorsal surface of the hand under local anesthesia (0.2 mL of 0.5% lidocaine) for the purpose of hydration and drug administration. A continuous IV infusion of 0.9% sodium chloride was started at 50 mL · hr−1. The rate was increased to 600 mL · hr−1 1 minute before study drug administration, then readjusted to 50 mL · hr−1 1.5 minutes after the study drug infusion was complete.
Pain on injection was compared between propofol in cyclodextrin and propofol in lipid using subject assessments of pain rated as none, mild, moderate, or severe at 0.25, 0.5, 1, 1.5, 2, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 60, 90, and 120 minutes after the start of each injection. Mild pain was defined as discomfort that is easily tolerated and would not interfere with usual activities. Moderate pain was defined as discomfort that would interrupt usual activities. Severe pain was defined as considerable discomfort that would interfere with usual activities. The responses were assigned numerical values (none = 0, mild = 1, moderate = 2, severe = 3).
In addition to the subject's assessment of pain, a single blinded observer also assessed the subjects for nonelicited indications of pain in facial expression, verbal expression, and movement at each time point. The interpretations of each category were assigned numerical values (not interpreted as pain = 0, interpreted as pain = 1). Finally, at the 60-minute time point the subject was asked, “When the pain was at its worst, was it mild, moderate, or severe?” These reflective responses were assigned values of none, mild, moderate, or severe as above. The primary outcome variable was a comparison of the pain scores derived from these assessments for both formulations.
A secondary efficacy objective, to evaluate the sedative effect of each formulation in terms of onset and duration of loss of consciousness, was to be measured by the BIS of the EEG, recorded continuously during study drug administration.
The safety profile of the study drugs was evaluated by monitoring changes in vital signs, pulse oximetry, and adverse events spontaneously reported by the subject or observed by the investigator, including pain on injection.
Demographic and background variables were summarized using descriptive statistics. Data manipulation, graphics, and analysis were accomplished in S-PLUS version 6.0 Release 1 (Insight Corporation, Seattle, WA) running under Linux OS 7.1 (Red Hat, Raleigh, NC). Analysis of contingency tables and ordinal variables was performed in StatXact 4.0.1 (Cytel Software Corporation, Cambridge, MA). Exact inference was used for the nonparametric statistics. Statistical significance was declared with 2-sided tests at P < 0.05. All 5 response variables were analyzed.
Comparison of propofol in cyclodextrin and propofol in lipid for the greatest pain response reported by each subject, the greatest pain response noted by the observer, and the worst pain recalled by the subject at 60 minutes were accomplished using the Marginal Homogeneity Test, which is an extension of the McNemar's Test. The Permutation Test and the Hodges–Lehman estimation compared the last 2 variables, the sum of the subject pain scores, and sum of the total pain for all subjects.
Both formulations, propofol in cyclodextrin and propofol in lipid, were generally well tolerated. The only adverse event reported with either formulation was pain on injection, which was the primary outcome being evaluated in the study. Table 1 and Table 2 show the greatest pain response reported by each subject and greatest pain response noted by the observer. The subjects are cross-classified by their response to both formulations. In these contingency tables, the subjects highlighted in the descending main diagonal (boldfaced values) had equal pain scores with propofol in cyclodextrin and propofol in lipid. The cells above the main diagonal denote subjects who had a higher pain score with propofol in lipid. Cells below the main diagonal denote subjects who had higher pain scores with propofol in cyclodextrin.
For the sum total of individual subject pain and sum of the total pain for all subjects, a Permutation Test statistic showed a significantly higher cumulative pain score for propofol in cyclodextrin (Permutation Test, P = 0.0065 and P = 0.0083). Maximum pain scores recalled by the subjects at 60 minutes are represented as a bubble plot in Figure 1. All 5 variables showed that propofol in cyclodextrin produced significantly greater pain scores.
The secondary efficacy objective, to compare the onset and duration of the sedative effect of each bolus administered using the BIS of the EEG, was summarized descriptively. By inspection of the raw data both graphically and numerically, there appeared to be no differences in the BIS between the 2 formulations. Therefore, no inferential statistical tests were performed for this variable.
All reported adverse events consisted of pain on injection and were considered treatment related. Most were of either mild or moderate intensity. All subjects tolerated the study well; there were no serious adverse events, and no subjects discontinued or had their dose reduced as the result of an adverse event. There were no clinically significant differences in vital signs (diastolic/systolic blood pressure, heart rate, respiration rate) or oxygen saturation during treatment administration between the 2 formulations.
Both formulations elicited pain on injection for all subjects when administered by infusion over 30 seconds into a vein on the back of the hand. Contrary to our hypothesis, the pain scores for the propofol in cyclodextrin formulation were higher than the scores for the lipid-based propofol formulation when evaluated over all doses.
Pain on injection, while admittedly not a serious complication, is a major concern to anesthesiologists and patients. In fact, a survey of anesthesiologists suggest that practitioners view propofol injection pain as one of the most important unsolved problems in the specialty.9 Recent drug development efforts in this therapeutic area, such as the current one, have been at least in part focused on reducing pain on injection. An important theme of all of these alternative propofol formulation efforts is that reformulation sometimes changes the clinical behavior in some way (i.e., pain on injection, latency to peak effort, among others).10
Pain on propofol injection is thought to correlate with the amount of propofol in the aqueous phase.11,12 In vitro and in vivo experiments appear to support this hypothesis. For example, it is clear that a novel, microemulsion formulation of propofol (i.e., Aquafol®, Daewon Pharmaceutical Company, Seoul, South Korea) that has a 7-fold higher concentration of free propofol in comparison with Diprivan produces a higher incidence of pain on injection.13 Similarly, a novel, lipid-based formulation of propofol with a lower lipid content (Ampofol®, Amphastar Pharmaceuticals, Cucamonga, CA) in comparison with Diprivan is also associated with a higher incidence of injection pain.14 Whether the pain associated with propofol in cyclodextrin injection is a function of free-propofol concentration as has been proposed for these other formulations is unknown. The free-propofol concentration in the cyclodextrin formulation is estimated to be approximately 50 to 80 mcg · mL−1 (data on file with CyDex Pharmaceuticals, Inc.). This is several-fold more than the free-propofol concentration in the lipid formulation.11
The mechanism by which propofol formulations cause pain on injection is still poorly understood, although work has focused attention on the transient receptor potential channel A1 (TRPA1). Some investigators had originally proposed that the kallikrein-kinin system was likely involved in the production of pain associated with propofol injection,15 although this view has been challenged.14 The TRPA1 receptor, a receptor associated with pain sensations,16 is activated by alkyl phenols such as propofol as demonstrated by in vitro whole-cell voltage clamp recording techniques in transfected human embryonic kidney cells expressing TRPA1 receptors.17 Interestingly, patients with high bitter taste sensitivity, a trait thought to be mediated at least in part through TRPA1 receptors,17 appear to be more susceptible to propofol injection pain.18 In any case, it is clear that the pain on injection elicited by propofol is independent of γ-aminobutyric acid A receptor activation.19 Improved understanding of the propofol injection pain mechanism may eventually lead to interventions that can control the pain associated with the nonlipid formulations, thereby making the lipid-containing formulations less attractive.
An obvious major limitation of the current study is that the protocol was designed to elicit pain. A small-gauge IV catheter in a vein on the dorsum of the hand with a relatively slow infusion rate was designed to create a strong primary outcome signal for the study. It is possible that the available interventions to mitigate propofol injection pain would reduce the magnitude of the pain considerably.
Although the propofol in cyclodextrin formulation described in this study likely addresses other shortcomings of the lipid formulation (i.e., lipid load, promotion of microbial growth), it does not appear on its own to reduce or eliminate the pain on injection associated with propofol in lipid emulsion.
Name: Crystal B. Wallentine, MD.
Conflict of Interest: This author has no conflict of interest to declare.
Name: Noriko Shimode, MD.
Conflict of Interest: This author has no conflict of interest to declare.
Name: Talmage D. Egan, MD.
Conflict of Interest: This author has no conflict of interest to declare.
Name: Nathan L. Pace, MD, MStat.
Conflict of Interest: Dr. Nathan L. Pace received consulting fees from CyDex Pharmaceuticals, Inc., Lenexa, Kansas.
This manuscript was handled by: Tony Gin, FANZCA, FRCA, MD.
Talmage D. Egan:
In 1998 I was fortunate to receive a Clinical Research Award from FAER. Entitled “A Quantitative Pharmacodynamic Analysis of Hypnotic and Opioid Interactions in Volunteers,” the grant employed study methods and analysis techniques I had learned during my clinical pharmacology fellowship training at Stanford under the mentorship of Dr. Steve Shafer. The FAER award was absolutely critical to my career development. Our FAER-supported work led to a host of publications and additional funding from both peer-reviewed and commercial sources. Along with many very capable collaborators, my lab continues to pursue various permutations of the work initially made possible by the generosity of donors to FAER. I am deeply grateful to FAER and its supporters for the impact they have had on my career and the careers of my mentees. The initial “boost” of the FAER Clinical Research Award made career “lift-off” possible and helped establish a successful trajectory.
1. Baker MT, Naguib M. Propofol: the challenges of formulation. Anesthesiology 2005;103:860–76
2. McLeskey CH, Walawander CA, Nahrwold ML, Roizen MF, Stanley TH, Thisted RA, White PF, Apfelbaum JL, Grasela TH, Hug CC Jr. Adverse events in a multicenter phase IV study of propofol: evaluation by anesthesiologists and postanesthesia care unit nurses. Anesth Analg 1993;77:S3–9
3. Tan CH, Onsiong MK. Pain on injection of propofol. Anaesthesia 1998;53:468–76
4. Eddleston JM, Shelly MP. The effect on serum lipid concentrations of a prolonged infusion of propofol–hypertriglyceridaemia associated with propofol administration. Intensive Care Med 1991;17:424–6
5. Laxenaire MC, Mata-Bermejo E, Moneret-Vautrin DA, Gueant JL. Life-threatening anaphylactoid reactions to propofol (Diprivan) [see comments]. Anesthesiology 1992;77:275–80
6. Bennett SN, McNeil MM, Bland LA, Arduino MJ, Villarino ME, Perrotta DM, Burwen DR, Welbel SF, Pegues DA, Stroud L, Zeitz PS, Jarvis WR. Postoperative infections traced to contamination of an intravenous anesthetic, propofol [see comments]. N Engl J Med 1995;333:147–54
7. Sosis MB, Braverman B. Growth of Staphylococcus aureus in four intravenous anesthetics. Anesth Analg 1993;77:766–8
8. Egan TD, Kern SE, Johnson KB, Pace NL. The pharmacokinetics and pharmacodynamics of propofol in a modified cyclodextrin formulation (Captisol) versus propofol in a lipid formulation (Diprivan): an electroencephalographic and hemodynamic study in a porcine model. Anesth Analg 2003;97:72–9
9. Macario A, Weinger M, Carney S, Kim A. Which clinical anesthesia outcomes are important to avoid? The perspective of patients. Anesth Analg 1999;89:652–8
10. Egan TD. Exploring the frontiers of propofol formulation strategy: is there life beyond the Milky Way? Br J Anaesth 2010;104:533–5
11. Doenicke AW, Roizen MF, Rau J, Kellermann W, Babl J. Reducing pain during propofol injection: the role of the solvent. Anesth Analg 1996;82:472–4
12. Doenicke AW, Roizen MF, Rau J, O'Connor M, Kugler J, Klotz U, Babl J. Pharmacokinetics and pharmacodynamics of propofol in a new solvent. Anesth Analg 1997;85:1399–403
13. Sim JY, Lee SH, Park DY, Jung JA, Ki KH, Lee DH, Noh GJ. Pain on injection with microemulsion propofol. Br J Clin Pharmacol 2009;67:316–25
14. Song D, Hamza M, White PF, Klein K, Recart A, Khodaparast O. The pharmacodynamic effects of a lower-lipid emulsion of propofol: a comparison with the standard propofol emulsion. Anesth Analg 2004;98:687–91
15. Tramer MR, Moore RA, McQuay HJ. Propofol and bradycardia: causation, frequency and severity. Br J Anaesth 1997;78:642–51
16. McNamara CR, Mandel-Brehm J, Bautista DM, Siemens J, Deranian KL, Zhao M, Hayward NJ, Chong JA, Julius D, Moran MM, Fanger CM. TRPA1 mediates formalin-induced pain. Proc Natl Acad Sci USA 2007;104:13525–30
17. Lee SP, Buber MT, Yang Q, Cerne R, Cortes RY, Sprous DG, Bryant RW. Thymol and related alkyl phenols activate the hTRPA1 channel. Br J Pharmacol 2008;153:1739–49
18. Erden V, Basaranoglu G, Korkut Y, Delatioglu H, Yangin Z, Kiroglu S. Relation between bitter taste sensitivity and incidence or intensity of propofol injection pain. Eur J Anaesthesiol 2007;24:516–20
© 2011 International Anesthesia Research Society
19. Fischer MJ, Leffler A, Niedermirtl F, Kistner K, Eberhardt M, Reeh PW, Nau C. The general anesthetic propofol excites nociceptors by activating TRPV1 and TRPA1 rather than GABAA receptors. J Biol Chem 2010;285:34781–92