washington

The US Environmental Protection Agency (EPA) said last week that it was “deeply concerned” that some pesticide manufacturers appeared to be engaging in health-effects studies on human subjects “as a way to avoid more protective results from animal tests”.

The statement came as it emerged that a Californian pesticide company has helped to pay for a study at the University of California, Davis, of human volunteers who were exposed to methyl isothiocyanate, the active ingredient in metam sodium, a potent soil fumigant (see box).

The company, Amvac Chemical Corporation, belonged to a coalition of six chemical companies — including Zeneca Ag Products — which spent roughly half a million dollars on the experiment, designed to meet California regulations. The study was carried out in 1994 and involved 70 paid volunteers.

Revelations last week that the same company had funded human experiments on British volunteers of an organophosphate pesticide, dichlorvos, have already focused attention in both Washington and London on a previously little-asked question: should pesticide makers be allowed to carry out tests on human volunteers to provide data to federal regulators, and, if so, under what conditions?

In response to news of the US-funded tests in Britain, the EPA said in a statement that “no human test data have been used by EPA for any final decisions about acceptable levels of pesticide use” under new food safety laws that critics claim the companies are trying to circumvent through human testing. (California's approach appears less stringent).

According to Ken Cook, president of the Washington-based Environmental Working Group (EWG), which advocates tougher pesticide controls, the Californian study is the first known pesticide experiment to have been conducted on human volunteers in the United States in recent years. He describes the experiment as further evidence that the EPA “needs to turn on its radar system and find out the extent of human experimentation on pesticides”.

The British tests were described in a report by EWG that documents the testing on humans of dichlorvos for Amvac by the Medeval Laboratories in Manchester in 1997. The Guardian newspaper reported this week that human tests were still being carried out on the organophosphate insecticide azinphos-methyl by the Inveresk Clinical Laboratory in Scotland, for which students and others were being paid $480 (US$780).

In its report, EWG criticizes the EPA for using data from the British experiments in determining pesticide safety levels, and calls for a moratorium on the use of human data in pesticide regulation. “EPA is accepting and evaluating human experimental studies that it does not require and, in fact, actively discourages,” the report says.

The EWG report contends that EPA used the human data in deciding to relax the safety margins for aldicarb. The group also alleges that industry is deliberately using human data to circumvent the tougher safety standards required in the Food Quality Protection Act of 1996, which overhauled pesticide regulation. By testing on humans, it says, companies hope to avoid an extra, tenfold margin of safety in determining safe exposure levels.

The EWG report also criticizes the EPA for failing to ensure that pesticide companies abide by government ethics rules — the so-called Common Rule for human subjects' protection — adopted by the agency in 1991.

Amvac defends the use of humans in experiments by arguing that it meets the conditions in the Helsinki Declaration for Testing and Protection of Human Subjects, requiring informed consent of participants and approval by independent ethics boards.

Ian Chart, the company's director of regulatory affairs, says that testing pesticides on humans “is good science when carried out under the Declaration of Helsinki”, which makes no distinction between the testing of pharmaceuticals and pesticides.

The company points out that dichlorvos, the pesticide used in the Manchester experiment, has been used for about 30 years to treat schistosomiasis.