Impure Perfluorocarbon Liquids: A Preventable Tragedy
In this issue of RETINA, Pastor and colleagues12 report an epidemic of catastrophic blindness caused by the use of a chemically contaminated PFO. This totally preventable complication was caused by poor regulatory oversight and the lack of expertise of a small company. Fortunately, the Spanish retinal community collaboratively was able to identify the origin of the problem before more eyes were lost. For retinal surgeons, this episode heightens the importance of knowing exactly what they are putting into the eye. In product labeling of liquid vitreous tamponades, there is little detailed information concerning the level of impurities, isomers, or small low molecular weight components. The manufacturer is implicitly assumed to have filed this information with the appropriate regulatory agency, and that the medical-grade quality of the product has been maintained.
Perfluorocarbon (perfluoroalkanes) compounds are biologically inert and nontoxic because the chemical bond between carbon and fluorine (C–F) is among the strongest in organic chemistry. When we first began studying the possible applications of perfluorocarbon liquids (PFCL) for vitreoretinal surgery, only industrial-grade perfluorocarbon liquids were available for study. In our early studies, these industrial-grade materials caused significant vitreous inflammation when placed into the vitreous. We believed that purified PFCL would be well-tolerated, and working with Dr. William Miller, a renowned fluorine chemist at Cornell University, to develop perfluoro-n-octane for use during vitrectomy. Considerable effort was made by a small company, Fluorosystems (led by R.N.S., MA, PhD, FRSC) to develop the purest medical-grade perfluoro-n-octane for eye surgery (99.99%). Specifically, a detailed chemical analysis was done to identify, reactive fluoride bonds, reducing agents, and other chemical contaminants within the product. A detailed Master File describing the synthetic pathways, sources of reagents, validated analytical procedures, including IR and UV spectroscopy, gas chromatography/mass spectrometric analysis, and nuclear magnetic analysis of the final PFO product was created and submitted to the United States Food and Drug Administration. On-site inspection of the facilities and documentation was also required, both pre- and postapproval. An equivalent approval process also applied in the UK until 1998 when it was changed to the European system.
This incident in Spain illustrates the lower standards for approval applied to medical devices by regulatory authorities in other countries. It appears that the company that sold PFO obtained batches of industrial-grade product and did not provide chemical analysis of each lot to European regulatory authorities before packaging and marketing. A simple chemical analysis using gas chromatography/mass spectroscopy would have been able to detect the presence of toxic impurities in the PFO. And for example, UV absorption at 220 nm to 180 nm could be easily performed which would detect the presence of impurities or contaminants (since perfluorocarbons are transparent in that region). These analytical techniques would be the most cost-effective method of checking for the purity of perfluorocarbon liquids. While Pastor and colleagues advocate in vitro testing for safety of these liquids in tissue culture, a chemical analysis would also detect toxic contaminants more quantitatively.