Congress wanted to protect children from dangerous pesticides when it passed a new food-safety law two years ago. But the legislation has produced an unexpected side effect.
Some of the world's biggest pesticide makers have resumed testing their products on humans, a practice that all but disappeared in the early 1980s. Hoping to convince American regulators that their bug-killers are safe, the manufacturers are paying scores of human volunteers in Britain and the U.S. to swallow what the companies describe as small doses of toxic pesticides.
"You are invited to take part in a study involving a pesticide called azinphosmethyl proposed for use in controlling nonbeneficial worms and insects," states an information sheet distributed to volunteers for a study that Bayer AG, the German chemicals and pharmaceuticals company, is conducting at Inveresk Clinical Research Ltd., a laboratory in Scotland.
The development disturbs officials at the U.S. Environmental Protection Agency, which doesn't require or encourage such studies. "This is what I would call a perverse and unanticipated consequence of a piece of legislation," says Lynn Goldman, the assistant administrator for the EPA's Office of Prevention, Pesticides and Toxic Substances. Dr. Goldman has asked for an independent scientific panel to determine whether the human studies meet ethical and safety standards.
The EPA says it will await the findings of the panel before using any human test data in administering the new law, which requires the agency to reassess hundreds of pesticides for how much residue they can safely leave in the food chain. But the EPA has already reviewed data from one company's tests.
Using human subjects to test the toxicity of pesticides isn't illegal in the U.S. or Europe. But doubts about the safety and ethics of such studies led most companies to discontinue them in the 1980s. Among regulators, there is still no consensus that the tests are risk-free for participants, and there is concern that the volunteers may play down or ignore potential hazards for the sake of the money they are paid. Regulators and others also worry that some information given to volunteers substitutes the word "drug" at times for "pesticide."
"I think that is misleading," says Richard Nicholson, editor of the Bulletin of Medical Ethics in London. "That is the sort of language that any competent ethics committee approving these documents, which they should have done, would remove."
The manufacturers say the tests pose few risks to volunteers, meet international ethical guidelines and represent their only recourse to a law that could ban or severely restrict many widely used pesticides in what amounts to a $37 billion world-wide market. The companies say the standards the Food Quality Protection Act sets are based on worst-case scenarios extrapolated from animal studies. So "the easiest way to argue" that a pesticide is safe for people, says Gurgen Frohling, a spokesman for Bayer's crop-protection group, "is to do the testing" on humans.
Besides Bayer, which says it also plans to test the pesticide methamidiphos on volunteers in Scotland, at least four other companies are conducting or planning similar studies. Switzerland's Novartis AG says it recently began testing diazinon -- a popular insecticide used to kill leaf-eating insects and cockroaches -- with about 60 volunteers at an undisclosed private U.S. laboratory. Amvac Chemical Corp., a unit of American Vanguard Corp. in Newport Beach, Calif., conducted trials with dichlorvos -- used in flea collars and agriculture -- on 21 volunteers in late 1996 at a research company owned by the University of Manchester in England.
Dow Chemical Co., based in Midland, Mich., says its Dow AgroSciences unit is awaiting approval from a U.S. laboratory for a human test with a worm and termite killer, chlorpyrifos. An EPA panel recently recommended that, to protect children, trace levels of chlorpyrifos currently permitted on food be radically reduced. And Cheminova Agro AS, based in Lemvig, Denmark, says it plans to test its biggest product, malathion, on volunteers at the Inveresk lab in Scotland "within the next two months." Malathion is used to kill head lice and insects on fruits and vegetables.
In all, at least seven human tests with pesticides are under way or planned, according to Chris F. Wilkinson, an industry consultant with Jellinek, Schwartz & Connolly Inc., in Arlington, Va.
Each of the pesticides being tested on humans is part of a group of chemicals known as organophosphates, the most widely used class of pesticide and the first to be reassessed by the EPA since the new food-safety law was passed. They work by interfering with cholinesterase, an enzyme found in the blood that is essential for the proper working of the nervous system. In both insects and humans, inhibiting the enzyme can lead to uncontrolled muscle twitching, convulsions and, in extreme cases, death.
In Small Doses
The pesticides being tested on humans range from mildly to highly toxic; some can be lethal in large doses. But the companies say the amounts given -- usually in a blend of corn oil in gelatin capsules, or mixed in orange juice -- are too small to cause harm. They say the tests are meant to show how much can be ingested without noticeable effects. Even if blood cholinesterase levels are affected, they say, the levels soon return to normal.
"They're not giving people just any old amount and saying, 'Let's see what happens,' " says John McCarthy, vice president for scientific and regulatory affairs at the American Crop Protection Association, a trade group in Washington.
EPA reviews of Amvac's dichlorvos tests in England show that some subjects reported minor symptoms, including mild headache, nausea and tiredness. But, according to the reviews, the testers concluded that the pesticide wasn't responsible.
The Money Question
Still, the EPA's Dr. Goldman and others worry that many test subjects, especially college students, may give little thought to their own safety because of the money. "You take an 18-year-old kid and offer him 800 bucks or 600 bucks, you can get him to do almost anything," says Arthur Caplan, director of the Center for Bioethics at the University of Pennsylvania. "Money can corrode informed consent."
Novartis says it pays $1,500 to volunteers, who range from 18 to 49 and must spend seven nights in a clinic as part of a 17-day test. Volunteers for Bayer's azinphos-methyl tests, which include a four-day stay in a clinic, are being offered $780, according to Robert Lonie, a 22-year-old Scottish college student who signed up in an effort to pay off a bank debt. "I thought it would be an easy way to make money," he says. He was rejected because he had taken hay-fever medication, but he says he planned to drop out anyway because he became worried about possible ill effects.
The Manchester volunteers, who included a number of college students, were paid either $560 or $492, depending on the test, which lasted up to three weeks and included several overnight stays. The compensation was "based on inconvenience," says Stephen Toon, managing director of Medeval Ltd., the university research lab that conducted Amvac's study.
Dr. Toon says money can be a motivator, noting that he volunteered for studies when he was in college, and "the vast majority of studies I did, I did for the money." But he says there was nothing "ethically wrong" in the Amvac studies.
While drugs are routinely tested on people, that isn't the case with pesticides because most have no medicinal use. Nevertheless, testing pesticides on humans isn't new. Numerous studies were carried out in the 1960s and 1970s to determine if the chemicals posed dangers to industrial and agricultural workers, and consumers. But controversy ensued over whether some of the test subjects -- who included prisoners and company employees -- were informed fully of potential risks or were in a position to be true volunteers.
In 1976, Switzerland's Ciba-Geigy AG conducted a test in which five boys under 18, including a 10-year-old, stood in a cotton field without protection and were sprayed by a crop duster. The company subsequently said it wouldn't use minors in any future test. Company officials say they believe no test subjects were hurt by the pesticide, which later was linked to cancer in animals and was removed from the market.
In more recent years, several companies, including Dow Chemical, Ciba-Geigy (which merged with Sandoz AG two years ago to form Novartis) and Rhone-Poulenc SA of France, occasionally tested pesticides on humans. Even the EPA sponsored a trial in the early 1990s.
In what became known within Ciba-Geigy as "the Directors' Study," the company recruited six top managers at its Greensboro, N.C., offices in 1987 to swallow atrazine, a herbicide used on corn and other crops, to determine its rate of elimination in urine. "I think the reason was that we wanted high-level people who could say no, and not have pressure put on people to do it," says Emilio J. Bontempo, who recently retired as chief executive of Novartis Crop Protection Inc. in Greensboro.
But most recent pesticide studies conducted by manufacturers and submitted to the EPA use animals, including rats, dogs and rabbits, with human studies limited to reviews of pesticide poisoning incidents and suicides. Animals are studied to determine how much pesticide they can withstand without any observable harmful effect. To account for differences among species and among people, the EPA has usually ruled that humans shouldn't be exposed to pesticide residues in food that are more than 1/100th the maximum safe level for animals.
Tougher Safety Margin
The 1996 Food Quality Protection Act tightens that safety net by a factor of 10 by setting a new human-exposure level that is 1/1000th the maximum safe level for animals -- a move aimed especially at protecting children and babies. The change had been recommended by a 1993 report by the National Academy of Sciences, which said standards were inadequate to protect children, who might be more sensitive to pesticides than adults and who consume proportionately more fruits and vegetables by body weight.
Under the new law, the EPA may ease or eliminate the new safety margin "only if, on the basis of reliable data, such margin will be safe for infants and children."
The bill sailed through Congress in 1996 with strong support from the pesticide industry, because it exempted bug-killers from the so-called Delaney Clause. That was a tough 1958 regulation that said processed foods couldn't contain even a single molecule of a pesticide known to cause cancer in animals. The pesticide makers "screwed up," says Dr. Wilkinson, the industry consultant. "The eyes of a lot of people were so focused on getting rid of Delaney that they never looked at the rest of the law."
Indeed, the law puts many popular pesticides at risk. In determining how much pesticide residue safely can be allowed on, say, an apple, the EPA must now also factor in what pesticides a person might be exposed to on any given day from other sources, such as a cat's flea collar or drinking water. Moreover, the agency must set one maximum exposure level for all chemically related pesticides. Preliminary EPA assessments already have indicated that some organophosphates are in trouble.
Fearful that the new law could lead to the banning of many pesticides, the industry began challenging some of the data and theoretical assumptions the agency plans to use to make its final determinations -- most of them from animal studies. "The time has come when the industry has to come up with real data -- real toxicology data and real exposure data -- that relate to humans," Dr. Wilkinson says.
In pursuing that data, many of the pesticide companies have chosen to conduct human studies in Britain, where they have done such work before and where they say the research labs are experienced and reliable. Britain also imposes fewer restrictions on clinical trials involving healthy volunteers than the U.S. does
At Manchester University's Medeval lab, which performed Amvac's dichlorvos trials, the school's human-study ethics committee didn't review the tests; the lab has its own committee, with members it recruits and compensates. Diana Kloss, a university law professor who heads the lab's committee, says that there was "considerable debate" before the dichlorvos study was approved, but that in the end, "the technical evidence was such that we thought it was OK."
What helped to convince her, she says, was that the committee was told by Medeval that dichlorvos "was used as a medicine," and she recalls hearing that it was used to treat Alzheimer's disease. "I certainly remember that it was used for clinical purposes already," she says.
Volunteers also were sometimes told the pesticide was a drug. Excerpts of test information sheets show that dichlorvos was referred to as a "drug" at least three times, although most references termed it a pesticide. And when the Washington-based Environmental Working Group, an organization that advocates tighter control of pesticides, publicized the dichlorvos tests in July, Amvac issued a statement saying, "Although the EWG identifies dichlorvos as a pesticide, they fail to mention that dichlorvos is also a U.S. FDA approved pharmaceutical agent used" to treat parasitic ailments.
Here the issue gets knotty. The FDA hasn't approved the use of dichlorvos for humans but only for animals. There is, however, a related compound, a drug called metrifonate, that has been used to treat parasitic ailments in humans and as a potential treatment for Alzheimer's disease. Bayer, which makes metrifonate, says that although it is chemically related to dichlorvos, the two aren't identical.
Bayer has been seeking approval from the FDA and European and other health authorities to market metrifonate as an Alzheimer's treatment. But last week, the company announced that, following consultation with the FDA, it was halting all world-wide clinical studies after observing "muscle weakness" in about 20 patients taking the drug, including some who required respiratory support. Bayer says it will analyze the problem.
As for statements made to volunteers and the ethics committee, Amvac says it "would not knowingly misrepresent ... any of its chemicals in studies" and didn't authorize any misstatements in the volunteer information sheets, which it said were prepared and distributed by Medeval. Medeval's Dr. Toon says volunteers and the committee were "absolutely fully informed" about dichlorvos.
Dr. Caplan, the bioethicist, says that suggesting to volunteers or ethics-committee members that dichlorvos was a drug raises serious ethical questions. If it is true that dichlorvos isn't identical to the drug metrifonate, "that's devastating, because that just blows the ethics of the study out of the water," he says. "It's a basic obligation to convey accurate information."
One of Bayer's pesticides also is referred to as a drug in the study under way in Scotland. A letter from Inveresk to Mr. Lonie, the student who volunteered for the study, first refers to azinphos-methyl as a pesticide but then says, "The drug will be administered orally." Mr. Lonie says Inveresk lab workers also referred to azinphos-methyl as a "medication."
Officials from Inveresk decline to comment. Dr. Frohling, the Bayer spokesman, says azinphos-methyl has no pharmaceutical uses and that referring to it as a drug was "just an error, but it's not misleading from my point of view."
Comments on this posting?
Click here to post a public comment on the Trash Talk Bulletin Board.
Click here to send a private comment to the Junkman.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.
Material presented on this home page constitutes opinion of Steven J. Milloy.