An August 25, 2005 dispatch from ChemWeek.com describes ACSH's petition to stop the EPA from using high-dose rodent tests alone to dub things "carcinogens" (and a longer version of the piece appears in the August 24/30, 2005 issue of the affiliated print magazine Chemical Week):
The American Council on Science and Health and the Washington Legal Foundation have petitioned EPA to stop using rodent studies as a basis for determining whether a chemical is a "likely" human carcinogen. The petition says that EPA's use of rodent studies to classify carcinogens violates the Data Quality Act (DQA), which requires information disseminated by federal agencies be objective, transparent, and reproducible.
EPA recently adopted guidelines for agency staff to use when determining whether a substance is a "known" human carcinogen or a "likely" human carcinogen. Of central concern to WLF and ACSH is that EPA officials are instructed to classify a substance as a likely human carcinogen if rodent health studies indicate a substance causes cancer. EPA's risk assessment guidelines "have led to numerous substances being deemed likely human carcinogen, despite the absence of evidence that the substances have caused any cancer in humans," WLF says.
The groups are also concerned about EPA's use of "default options," where regulators assume that if a large dose of a chemical causes cancer in rodents, then a smaller dose is also likely to cause cancer. WLF and ACSH say that the Data Quality Act "does not permit a federal agency to label a substance a 'likely' human carcinogen in the absence of any sound evidence to support such a label." EPA says the use of "default options" is justifiable because the agency's mission is to protect public health, and the agency must err on the side of caution. WLF says that EPA may be subverting its own mission to protect public health by labeling a substance -- which may have significant public health benefits -- as a likely human carcinogen.