The article by Dr. Kharasch entitled "Sevoflurane: The Challenges of Safe Formulation," published in the APSF Newsletter, Fall 2007, contains a number of inaccuracies and errors of omission. These lead to a number of scientifically unfounded concerns regarding the packaging of sevoflurane, the formulation (water content), and the compatibility of generic sevoflurane in anesthetic vaporizers.
As it now stands, there are 3 different types of containers and water content being used for sevoflurane:
The first company to receive regulatory approval for the marketing of sevoflurane as an anesthetic agent, Maruishi Pharmaceutical (Japan), was granted marketing approval to enter the Japanese market in 1990. The Maruishi product was packaged in USP Type III Amber glass. Subsequently, Maruishi licensed Abbott Laboratories, and in 1995 Abbott received FDA approval for the product in the United States and began to market sevoflurane in USP Type III Amber glass bottles.
There were no problems with the product, packaged in USP Type III Amber glass, in Japan or the US, until Abbott had a recall of three lots of sevoflurane (FDA D-054-7) in November 1996 due to contamination of the bulk API with rust (ferric oxide). There was no mention that the packaging, USP Type III Amber glass, caused the degradation nor was the degradation attributed to the water content of their product. In their correspondence sent to the FDA, they assured that the cause was known and that actions had been taken to correct the problem.1
There was a second recall of 22 lots of sevoflurane (FDA D-011-4) in September 1997 for a similar problem.1 In correspondence with the FDA, it was again stated that the recall was due to contamination of the bulk API with rust (ferric oxide). As before, there was no mention that the packaging, USP Type III Amber glass, caused the degradation nor was the degradation attributed to the water content of their product.
Over the years it has been known that ferric oxide, and certain other metal oxides, would autocatalytically decompose certain halogenated ethers. Good manufacturing practices (cGMP) and great care are taken to avoid these contaminants.
The fact that USP Type III Amber glass was not mentioned is not surprising. Enflurane, isoflurane, and desflurane have all been packaged in this type glass for over 15 years without stability problems. Similarly, unadulterated sevoflurane has been packaged in this type glass by Maruishi and Abbott and is being packaged in this type glass by MINRAD without any degradation occurring. MINRAD has stability data at 40°C for a year and at 25°C for over 2 years, all with very low water content, and there is no indication of product degradation.
The surface of the USP Type III Amber glass consists of silicon hydroxide (SiOH), commonly called Silanol, which is formed by the reaction of silicon dioxide with moisture in the air as the glass is annealed in the Lehr to relieve stresses. Silanol is a weak Bronsted Acid, not a Lewis Acid, and does not cause degradation of any of the halogenated ether anesthetic agents, no matter what the water content.
In 1985, we, then at Anaquest, carefully investigated the possibility of packaging enflurane and isoflurane in polymer containers as a way to reduce the shipping costs of the products. We found that none of the then available polymers were as suitable as USP Type III Amber glass for the packaging. More recently MINRAD, for the same reason, reviewed the newer polymers. A few polymers, polyethylene terephthalate (PET) and polyethylene naphthalate (PEN), appeared to have the required structural strength; however, both were rejected because of the potential for extractables. Both PET and PEN are polyesters and in the process of blow forming bottles with these polymers, there is a thermal degradation reaction that produces acetaldehyde and possibly ethylene glycol and acrolein. Since Minrad had data proving that USP Type III Amber glass had no extractables from the glass, we decided that the appropriate action was to continue to package our product in USE Type III Amber glass in order to preserve the quality of our product.
As a result of the Baxter recall of Penlon Sigma Delta vaporizers showing degradation of sevoflurane, there is now concern about the need for some sort of inhibitor to avoid the degradation of sevoflurane in vaporizers. Water may slow the rate of degradation when sevoflurane is exposed to an incompatible metal such as rust. But, as Clarine M. Callan, MD, stated in the Spring 1997 APSF Newsletter
After initiation by the valve material, a necessary part of the reaction involves the preferential consumption of moisture in sevoflurane. This reaction continues at a very slow rate until all the water is consumed, at which point reaction with the solution container rapidly occurs resulting in the pungent odor (SiF4).
The "initiation" described is the rust catalysis of sevoflurane producing FW and other degradation products. In the APSF Newsletter, Summer 1997, a Letter to the Editor by Beverly C. Collins, CRNA, discusses the analysis of the many rust initiated degradation products. With HF being produced in the rust contaminated sevoflurane, the product is out of specification, and the pungent odor of HF will be detected independent of the container.
In his article, Dr. Kharasch creates the impression that all metal oxides are "potential Lewis Acids," an assertion that is not supported by scientific evidence. Abbott has circulated data of laboratory tests of sevoflurane vaporizers manufactured by CIE-Ohmeda, Dräger, and Penlon, using Minrad, Baxter, and Abbott-marketed sevoflurane under accelerated conditions, to many parties. One series of tests inventory the metal oxides, and their areas, present in the vaporizers. The results of these tests would lead one to the conclusion that degradation should be worse in Dräger vaporizers, a conclusion not supported by the second test series that quantified the actual degradation of the sevoflurane products in the vaporizers.
In the latter tests no degradation was detected in either GE-Ohmeda or Dräger vaporizers, independent of the water content, a fact not mentioned by Dr. Kharasch. Equally important was the fact that for all manufacturers, sevoflurane degraded in the Penlon vaporizers. Although the rate of degradation was apparently reduced by water, water did not prevent degradation from occurring. Dr. Kharasch is aware of this outcome because in a letter, also circulated by Abbott, from Dr. Kharasch to Mario Saltarelli, MD, PhD, Divisional vice president, Abbott Laboratories, Dr. Kharasch states the following:
Further in the same letter, Dr. Kharasch ignores the fate of the methyl fluoride group that was cleaved at the ether oxygen in order to produce HFIP (hexafluoroisopropyl alcohol).
Subsequent investigation by Penlon concluded that the problem was caused by defects in the barrier coating of the aluminum alloy vaporizer body that exposed the product to the aluminum alloy surface. Penlon has taken steps to correct the defect and Quality Assurance testing is in progress to insure that their vaporizer is fully compatible with sevoflurane.
Since US 5,990,176, "Fluoroether Compositions and Methods for Inhibiting Their Degradation in the Presence of a Lewis Acid" specifically mentions the aluminum oxide in the USP Type III Amber Glass as a source of the supposed Lewis Acid, in 2000, Dr. Terrell did studies of the stability of sevoflurane using Type 3A molecular sieves (these molecular sieves are composed of aluminum oxide and silicon dioxide). The results of Dr. Terrell's experiments clearly showed that there was no degradation of sevoflurane independent of water content varying from 40 ppm to several hundred ppm. Recently the retained samples from Dr. Terrell's work were retested after more than 5 years of contact with unactivated alumina and silicon dioxide; there was still no indication of degradation of sevoflurane.