A leaked draft EPA rule on using human volunteers for tests of pesticides and toxic chemicals has drawn a sharp response from Congress.

The proposal follows some of the recommendations of a February 2004 National Research Council report on the subject. But the draft omits several safeguards and allows certain experiments that NRC said should be prohibited.

For example, the draft proposal allows exposure studies on children if conducted during children's normal activities and not used to identify a toxic effect. NRC, in contrast, concluded that studies on pregnant women, as well as children, newborns, and prisoners, would generally be unethical. NRC also recommended that EPA establish a review board to evaluate all human dosing studies, but the EPA proposal does not require such a board. The agency says a review board "would unnecessarily confine EPA's discretion."

EPA's controversial proposal has led to conflicting amendments being passed in the Senate. On June 29, the Senate addressed pesticide testing in two amendments to the appropriations bill for interior and related agencies. One measure, by Sen. Barbara Boxer (D-Calif.), prohibits EPA from spending any funds to consider third-party pesticide studies--those conducted by industry or outside organizations--in which humans are intentionally dosed. The second measure, by Sen. Conrad Burns (R-Mont.), instructs EPA to conduct a thorough review of human pesticide dosing studies and to issue a final rule on the conduct of such studies within 180 days.

Last month, the House approved a measure similar to the Boxer amendment.

EPA says the draft proposal has not been reviewed by Administrator Stephen L. Johnson and that the agency had not planned to release the proposal until August. "EPA is taking the historic step of regulating, for the first time, scientific and ethical standards for accepting third-party human studies, and it's unfortunate that members of Congress have chosen to politicize this important issue," the agency says in a statement.

Human clinical trials with pesticides have been controversial for nearly a decade. To protect children, the 1986 Food Quality Protection Act established an additional 10-fold margin of safety to be applied to pesticide exposures when extrapolating from animal to human data. Soon after that, EPA began receiving the results of human pesticide tests from companies. If tests showed that humans were not more sensitive to a pesticide than experimental lab animals, the agency could eliminate the 10-fold safety factor for extrapolating from animals to humans, making it easier for companies to get their products approved.

In 1998, EPA stopped accepting such studies, but in 2001, it began considering them on a case-by-case basis and asked for guidance from NRC.

CropLife America, which represents major pesticide manufacturers, "believes that sound science and public health protections have affirmed the safety and ethics of human data studies."