Anesthesiologists in the United States now have a choice of sevoflurane products that has not been available since the launch of this anesthetic. Sevoflurane was first marketed for human use in 1995 by Abbott Laboratories, Inc. (Abbott Park, IL) under the tradename Ultane® (1). In early 2006, Baxter Healthcare Corp. (Dearfield, IL) launched a generic version of this anesthetic (Sevoflurane Inhalation Anesthetic, Food and Drug Administration (FDA) approved July 2002) in the United States. In approving Baxter’s product, FDA determined that Baxter’s sevoflurane is therapeutically equivalent to Ultane, the reference listed drug.
Even though therapeutically equivalent, the products are marketed in different containers. Ultane is supplied in amber plastic bottles (PEN, polyethylene naphthalate polymer), and Baxter’s sevoflurane is marketed in aluminum bottles lined with an epoxyphenolic resin. Before 2002, Ultane was marketed in amber glass bottles. The change in Ultane container and the use of an aluminum bottle for the Baxter product could reasonably raise questions as to why different containers are used and whether there are differences in the liquid anesthetic.
DIFFERENCES IN SEVOFLURANE MANUFACTURE
The patent on sevoflurane as an anesthetic was issued in 1972 and consequently has long since expired. This patent disclosed several methods for synthesis of sevoflurane, but none was patented (2). The lack of development of sevoflurane after its initial discovery was due to various reasons, including concerns about its instability in CO2 absorbents and the cost of manufacture by the methods available at the time. Eventually, the desire for a faster-acting volatile anesthetic prompted reconsideration of this compound. For potential manufacturers, the need for better synthetic methods provided grounds by which sevoflurane manufacture could be further developed and the resulting intellectual property patent protected.
Currently, the basic method by which sevoflurane is manufactured is to react 1,1,1,3,3,3-hexafluoroisopropanol (HFIP) with reagents that create the fluoromethyl ether moiety directly from the alcohol. This process can involve either a single step or multiple steps. The currently used one-step method was patented by Baxter Travenol (3) and licensed to Maruishi Pharmaceuticals Co., Ltd. (Osaka, Japan) (1). This method is used for the manufacture of sevoflurane distributed by Maruishi and Abbott. This synthetic process involves the reaction of HFIP (United States Pharmacopeia (USP) Sevoflurane Related Compound C) with formaldehyde (or paraformaldehyde) and hydrogen fluoride (HF) in the presence of oleum (H2SO4–SO3 mixture) (Fig. 1). As a single step, this process is efficient and adaptable to a continuous production (4).
A three-step method is used by Baxter (5) (Fig. 1). It involves, 1) reaction of HFIP with a methylating agent to form sevomethyl ether (SME, USP Sevoflurane Related Compound B), 2) followed by the addition of a chlorine atom on the methyl group to form chlorosevo ether (CSE) and, lastly, 3) substitution of the chlorine of CSE with fluorine using a fluorinating agent (6,7). The reagents to synthesize sevoflurane by the three-steps have not been disclosed, but some likely methods use dimethylsulfate as the methylating agent, chlorine gas (Cl2) via photochemical chlorine activation, as the chlorinating agent, and a patented step using a tertiary amine HF salt as the fluorinating agent (7). The economical synthesis of HFIP is also not trivial and a number of methods to synthesize this alcohol have been developed (8).
All synthetic reactions generate some impurities and, for a drug product, they must be removed or decreased to levels acceptable to the FDA and ultimately specified by the USP. It is usually most difficult to remove those impurities that are most similar in physical characteristics to the final product. The current draft USP monograph on sevoflurane specifies that sevoflurane contain not <99.97% and not >100.00% sevoflurane, calculated on an anhydrous basis; i.e., not >300 ppm impurities excluding water (9). It should meet the requirements for identification by infrared absorption, have a refractive index of 1.2745–1.2760 at 20°C, contain not >0.1% water (1000 ppm), contain not >2 μg/mL fluoride, contain not >1.0 mg nonvolatile residue/10 mL liquid, and have a limit of peroxide of not >0.22 μg/mL.
Given the two synthetic methods, some differences in quantity and quality of impurities are to be expected. A comparison of the two products in a study released by Baxter showed that the lots tested were within the USP specifications (5). Ultane and the Baxter sevoflurane were found to be >99.99% sevoflurane. Both contained trace amounts of HFIP (boiling point (bp) = 57–60°C) and Compound A (USP Sevoflurane Related Compound A, bp = 46°C), the latter of which is the product of dehydrofluorination of the hexafluoroisopropyl moiety of sevoflurane (10). On the other hand, HFIP was approximately 50 times higher in the Baxter product. Ultane contained SME at detectable levels, but SME (bp = 50°C) was not detected in Baxter’s sevoflurane. CSE, which is only produced in the three-step method, was not found in either product in this study. Its lower volatility (bp = 76°C), compared with the other impurities, probably facilitates its removal during distillation (11). As validated, the quantities of impurities are low and qualitative differences minor. If they remain so, the differences in impurities upon manufacture are not expected to be of clinical significance.
The significance of containers to sevoflurane storage and distribution relates to the discovery that sevoflurane can be unstable after manufacture in the glass bottles in which it was initially supplied. In 1996, there were reports of sevoflurane liquid in some bottles having a cloudy appearance and a pungent odor (12). Investigation of these containers by Abbott Laboratories found them to have a high acidity (pH <1) and to contain HF, a toxic volatile acid, in concentrations up to 863 ppm (13,14). An increased concentration of HF indicates sevoflurane degradation.
The reason for this degradation was ultimately linked to Lewis acid-dependent defluorination reactions which, when they occurred in the presence of glass, resulted in a cascade of reactions involving the fluoride ion and silicon dioxide (15). These reactions were favored under anhydrous conditions. The initial source of the Lewis acids in the 1996 incident was traced to a single valve connected to a bulk sevoflurane shipping container (13). The Lewis acid on the valve was identified as iron oxide, i.e., rust (16). Upon further investigation, Abbott Laboratories concluded that the equipment used in manufacturing and shipping the glass bottles and anesthesia equipment all contain components (i.e., Lewis acids), which when contacted with sevoflurane could lead to sevofluane degradation (17).
Lewis acids are molecules, usually metal-containing molecules, which have empty orbitals in the outer shells of their central atoms (18). They act as Lewis acids by accepting electrons from another compound, a Lewis base, which can result in Lewis base degradation. Lewis acids are ubiquitous. Some examples are the metal halides and oxides: FeCl3, Fe2O3, AlCl3, Al2O3, SbCl3, SnCl4, TiCl4, BCl3, and BF3. Unlike other fluorinated anesthetics, which are supplied in glass containers in which they are stable, sevoflurane is particularly susceptible to Lewis acid attack due to its monofluoromethyl ether group (R―O―CH2F). Lewis acids are, in fact, useful reagents in fluorocarbon synthetic reactions that involve the cleavage of α-monofluoroalkyl ether bonds (R―O―CH2F―R) (19).
Type III glass, a soda lime glass previously used for sevoflurane and currently used to package the other fluorinated anesthetics, contains silicon dioxide (SiO2), with lower amounts of calcium oxide (CaO), sodium oxide (Na2O), and aluminum oxide, Al2O3 (20). Small quantities of other metal oxides, which give the glass its amber color, are added to block ultraviolet and short wavelength visible light (21). Aluminum oxide is thought to be the Lewis acid that initiates and propagates sevoflurane degradation in glass (15). A study of the direct reaction between activated aluminum oxide, i.e., dehydrated alumina, and sevoflurane showed that the reactions are complex and multiple sevoflurane-derived products occur. The initial reaction and most subsequent ones release HF (22) (Fig. 2).
HF, if not able to escape, acidifies the liquid and reacts with SiO2 generating various silicon fluorides (23). These silicon oxide reactions have several important consequences. They can cause the glass to lose its structural integrity and result in glass thinning and glass particle formation. They expose the contents of the container to more Lewis acids, e.g., Al2O3, to propagate the reactions, and the silicon fluorides generated can also act as Lewis acids due to the strong electronegative effect of fluoride on the silicon to which it is bound. Ultimately, silicon tetrafluoride (SiF4) can form (23):
Silicon tetrafluoride, like HF, is a toxic volatile compound with a pungent odor. When SiF4 contacts moisture it hydrolyzes to form HF and silicic acid (24).
The degradation of Ultane was first detected by odor (12). Fortunately, HF is pungent below its occupational safety and health administration permissible exposure level: 3 ppm average concentration over an 8-h period (25). As little as 0.04 ppm can be detected by odor. Unfortunately, inhalation of as little as 50 ppm for 1 h can result in serious adverse effects. HF is highly corrosive to human tissue (26). Exposure to any HF, or sevoflurane thought to be degraded, should be avoided.
Although the occurrence of such degradation in Ultane glass bottles was limited, the pungency and toxicity of HF and SiF4 provided a strong impetus to find ways to prevent these reactions. One solution is the addition of small quantities of a compound to the liquid anesthetic to inhibit the Lewis acid reactions, i.e., a Lewis acid inhibitor. A number of compounds, both organic and inorganic, serve this purpose and some have been patented by Abbott Labs (22). These include H2O, butylated hydroxytoluene, methyparaben, propylparaben, propofol and thymol. The presence of low concentrations of an added organic compound to sevoflurane is potentially acceptable and would not be unlike adding thymol to liquid halothane. The most convenient Lewis acid inhibitor, however, is simply water. Water can act as a Lewis base where the oxygen of the water molecule coordinates with the Lewis acid via the unbound electrons of the oxygen atom (15). This impedes the interaction of the Lewis acid with an alternate Lewis base, in this case, sevoflurane. When water is absent or low, sevoflurane can donate electrons to Lewis acids via the unbound electrons of the ether oxygen, a process which, as noted, can ultimately lead to sevoflurane degradation. Therefore, in sufficient quantity water can inhibit Lewis acid-dependent sevoflurane degradation. This protective effect of water prompted Abbott to add water to a concentration of 330 ppm in their final sevoflurane product to produce a so-called “water-enhanced” or “wet” sevoflurane (17,27).
The patenting of Lewis acid inhibitors, including water, by Abbott Labs is an impediment to competitors marketing sevoflurane containing Lewis acid inhibitors. Most manufacturing processes for sevoflurane, however, result in the product having some water content. Water is a direct product of sevoflurane synthesis by the one-step method (4). Furthermore, water can be absorbed into anhydrous sevoflurane from the atmosphere. The Baxter product does not contain added water, but contains ≤130 ppm water naturally resulting from its manufacture (27). The marketing of sevoflurane by Baxter was legally challenged by Abbott on the basis that the water it contains functions as a Lewis acid inhibitor and protects sevoflurane. In 2005, US courts concluded that the Baxter product did not infringe the Abbott patent (16).
NEW SEVOFLURANE CONTAINERS
An alternative solution to adding Lewis acid inhibitors to sevoflurane is to package it in containers that do not cause or minimize defluorination. Baxter markets sevoflurane in aluminum bottles [(28) Fig. 3]. There are several potential concerns in using these containers. For example, the process of producing an aluminum container can result in small particles of aluminum in the container that could contaminate its contents. Aluminum oxide can be formed on the surface of aluminum by exposure to air. As noted, aluminum oxide can degrade sevoflurane under appropriate conditions (22). The aluminum bottles in which sevoflurane is packaged have an internal flexible lacquer liner made of an epoxyphenolic resin (16). This liner prevents the liquid from being contaminated with aluminum particles, minimizing sevoflurane contact with the metallic aluminum, and possibly serving as a Lewis acid inhibitor where it contacts the metallic aluminum. A drawback to plastic liners is that fluorinated anesthetics are good organic solvents and could leach polymer components. Epoxyphenolic resin liners are commonly used in metallic food containers, and studies have shown that resin monomers can be found in the packaged food substances in low concentrations (29). An additional drawback to aluminum containers is that the liquid anesthetic cannot be visually inspected for particulates or cloudiness.
In spite of the apparent effectiveness of added water (>130 ppm) in protecting sevoflurane from degradation in glass, Abbott now markets sevoflurane in amber plastic bottles (30). Abbott has stated that although plastic reduces the level of potential contaminants, water is still required to inhibit the formation of HF (17). This suggests that the plastic contains some Lewis acids. The polymer used by Abbott is a PEN (Fig. 4). A patent has been granted for this and other polymeric containers for fluorinated anesthetics (31). Unlike glass and aluminum, plastics are potentially permeable to vapor, and they are more susceptible to weakening with heat. Furthermore, as an organic polymer, there is the potential of polymer components leaching into the anesthetic. Various tests have shown PEN to be compatible with sevoflurane (32). PEN permeability increases with increases in temperature; however, no differences in physical properties of PEN containers were noted when sevoflurane was stored at 40°C for 3 mo. Low PEN permeability to gases and vapors was exemplified by a study that showed no loss of CO2 from a PEN container over a 24-mo period. Scanning electron microscopy showed no evidence of flaking or cracking due to sevoflurane contact with the polymer. Acetaldehyde is a product of high temperature PEN manufacture and as such could contaminate the product if not allowed to completely desorb from the bottles before filling. The PEN container used by Abbott has sufficient transparency for visible inspection of the liquid.
In spite of tests of the suitability of PEN with sevoflurane, one recall has been issued on several lots of Ultane supplied in PEN bottles (33). The recall notice stated that the bottles may have pinholes that could cause the product to leak or evaporate. Presumably, this recall was due to improper container manufacturer.
In summary, while the sevoflurane products are rated therapeutically equivalent and clinical differences are not expected, there are differences in Ultane and generic sevoflurane that have arisen in the course of product development by the different companies. As exemplified by its reactions with basic CO2 absorbents (10), Lewis acids in glass, and Lewis acids from other sources, sevoflurane has a unique chemistry among the fluorinated anesthetics. A potential remains for sevoflurane instability in all types of containers in which it is packaged; therefore, some vigilance regarding product integrity remains prudent.
1. Callan C, Delgado-Herrera L, Guzek D, Blahunka K. An historical perspective of the successful development of sevoflurane. Drug Inf J 1998;32:119–27.
2. Regan BM, Longstreet JC. Method of anesthesia. US Patent 3,683,092. 1972.
3. Coon CL, Simon RL. Method of synthesizing fluoromethylhexafluoroisopropyl ether. US Patent 4,250,334. 1981
4. Marynowski CW. Scale-up: three easy lessons. Chemtech 1987;560–3.
5. Comparative analysis of sevoflurane produced by Baxter Healthcare Corporation and sevoflurane produced by Abbott Laboratories. Baxter Healthcare Corp., New Providence, NJ. Presented Feb 24, 2006, University of Iowa.
6. Kudzma LV, Lessor RA, Rozov LA, Ramig K. Method of preparing monofluoromethyl ethers. US Patent 5,886,239. 1999.
7. Rozov LA, Lessor RA. Process for recovery of 1,1,1,3,3,3-hexafluoroisopropanol from the waste stream of sevoflurane synthesis. US Patent 6,987,204 B2. 2006.
8. Katsuhara Y, Nakamichi T, Kawai T, Nakazora T. Preparation of 1,1,1,3,3,3-hexafluoropropane-2-ol by hydrogenolysis of hexafluoroacetone hydrate. US Patent 4,564,716. 1986.
9. Section IV: Chemistry and Compendial Requirements – Sevoflurane. USP DI, Vol. 3: Approved Drug Products and Legal Requirements. United States Pharmacopeial Convention, Inc., 2006;576–7.
10. Huang C, Venturella VS, Cholli AL, Venutolo FM, Silbermann AT, Vernice GG. Detailed investigation of fluoromethyl 1,1,1,3,3,3,-hexafluoro-2-propyl ether (sevoflurane) and its degradation products. Part I: Synthesis of fluorinated, soda lime-induced degradation products. J Fluor Chem 1989;45:239–53.
11. Bieniarz C, Behme C, Ramakrishna K. An efficient and environmentally friendly synthesis of the inhalation anesthetic sevoflurane. J Fluor Chem 2000;106:99–102.
12. Leary JP. Contaminated sevoflurane use reported from NY State (letter to editor). Anesthesia Patient Safety Newsletter, Winter 1997.
13. Callan C. Maker follows up on sevoflurane problem (letter to editor). Anesthesia Patient Safety Newsletter, Spring 1997.
14. Bertolini JC. Hydrofluoric acid: a review of toxicity. J Emerg Med 1992;10:163–8.
15. McLesky CH. The science and art behind the enhanced stabilization of sevoflurane. Bull HK Coll Anaesthesiol 2006;15:11–3.
16. Bench Opinion. Abbott Laboratories, et al vs Baxter Pharmaceutical Products, Inc, et al. In the United States District court for the Northern district of Illinois Eastern Division. Case 1:01-cv-01867, Document 244, Filed Sept 22, 2005.
17. McLeskey CH. Anesthesiologist executive reports how Abbott made sevoflurane safer. Water stops formation of highly toxic acid (letter to editor). Anesthesia Patient Safety Newsletter, Fall 2000.
18. Brown TL, LeMay HE Jr, Bursten BE. Acid–base equilibria. In: Challice J, ed. Chemistry: the central science. Upper Saddle River: Prentice Hall 2000;593–639.
19. Engl DC. Catalytic conversion of fluoroalkyl alkyl ethers to carbonyl compounds. J Org Chem 1984;49:4007–8.
20. Flament-Garcia MJ, Cromack KR, Loffredo D, Raghavan R, Ramsay GM, Speicher ER. Container for an inhalation anesthetic. US Patent 6,162,443. 2000.
21. Special glass containers for primary packaging: Amber glass, Part 1. Pharma Information Letter, 4th ed, Newsletter. 3/2003. Schott-Rohrglas GmbH.
22. Bieniarz C, Chang SH, Cromack KR, et al. Fluoroether compositions and methods for inhibiting their degradation in the presence of a Lewis acid. US Patent 5,990,176. 1999.
23. Oxtoby DW, Nachtrieb NH, Freeman WA. Chemistry of the halogens. In: Chemistry: the science of change. 2nd ed. Philadelphia: Saunders College Publishing, 1994;905–36.
24. U.S. Department of Labor Occupational Safety & Health Administration. Available at www.osha.gov/dts/chemicalsampling/data/CH_267175.html
. Accessed Sept 14, 2006.
25. Medical management guidelines for hydrogen fluoride (HF). Agency for toxic substances and disease registry (ATSDR). Department of Health and Human Services. Available at www.atsdr.cdc.gov/
. Accessed Sept 14, 2006.
26. Sticht G Fluorine. In: Seiler HG, Sigel H, Eds. Toxicity of inorganic compounds. New York: Marcel Dekker, 1988;283–91.
27. Court rules generic version of Abbott anesthetic does not infringe; Abbott appeals. Court Proceedings, Pharmaceutical Law & Industry Report, BNA. 2005;3:1009.
28. Rudzinski RV, Lessor RA. Container for inhalation anesthetic. US Patent Application Publication No. 2002/0068767 A1. June 6, 2002.
29. Goodson A, Summerfield W, Cooper I. Survey of bisphenol A and bisphenol F in canned foods. Food Addit Contam 2002;19:796–802.
30. PEN bottle protects volatile anesthetic. Elements (BP, PLC) 2002;1:9.
31. Flament-Garcia MJ, Chang SH, Cromack KR, et al. Container for an inhalation anesthetic. US Patent 6,074,668. 2000.
32. Scientific Discussion. European Agency for the Evaluation of Medicinal Products (EMEA), Sevoflo. Document CVMP/208/03. 2003.
33. Enforcement report. Recalls and field corrections: foods and cosmetics – Class I. Recall No. D-194-6. US Food and Drug Administration. March 16, 2006.