Letter re: "Cyclosiloxanes produce fatal liver and lung damage in mice".

It is EHPs stated editorial policy to serve as a forum for discussion of issues of environmental health, encouraging the expression of scientific opinion and fostering healthy scientific debate. Your editorial policy states that "all scientific articles are subject to rigorous peer review. The primary criteria are environmental significance and scientific quality." Based on these criteria , it appears that the journal has failed to hold the paper of Lieberman et al. (1) to these standards. This paper is a blatant and deliberate misdirection of the reader, providing misinterpretation of a poorly designed study that is not up to the standards of modern toxicology or EHP. Lieberman et al. (1) indicate that they distilled breast implant gel at 180°C at reduced pressure for 24 hr and imply that this material represents what would leak from implants. It is well known that silicone polymers can thermally depolymerize to form cyclic siloxanes under the authors' distillation conditions (2), but this does not represent "real life" conditions. Lieberman et al. (1) administered this distillate to mice by intraperitoneal (ip) injection at doses up to 35,000 mg/kg-surely an unacceptably high dose that would cause direct irritation. It is therefore no surprise that the identified ip median lethal dose (LD 0) was 28,000 mg/kg and that lung and iver lesions were noted. Perhaps the animal care committee should have requested a revision of the testing protocol before the study was initiated. Lieberman et al. (1) reported that one of the individual components of the distillate, identified as CS-4 based on an ip LD50 of 6,000-7,000 mg/kg, is equivalent in toxicity to oral exposures to carbon tetrachloride; they characterized the distillate and individual distillate materials as highly toxic. For regulatory purposes, any material with an LD50 greater than 2,000 mg/kg (3) or 5,000 mg/kg (4-9) is considered to be the highest dose necessary to test. Materials with LD50 values greater than these dose levels are considered to be virtually nontoxic. If this misinterpretation of toxicity data were to remain quietly in the annals of EHP, it would be merely a problem of editorial carelessness. However, this paper has been picked up by several of the news services (e.g., Reuters, BBC), with online and print media declaring "Silicones Kill Mice!" and no longer noting that the dose and dose route are responsible for the lethality, not the inherent toxicity of the material. By publishing this paper, EHP …