Comments of Friends of the Earth
before the FIFRA Scientific Advisory Panel (SAP)

Open Meeting on Bt Crop Re-Registrations
October 20, 2000

Introduction

My name is Bill Freese. I'm speaking on behalf of Friends of the Earth, an international environmental organization with 20,000 members and supporters in the U.S. and affiliate groups in 66 countries. My comments today will focus mainly on the potential allergenicity of StarLink corn, particularly in light of the widespread contamination of the food supply. But I will also address some broader questions concerning the failure of our regulatory system with respect to genetically engineered foods.

Friends of the Earth welcomes the recalls of contaminated corn products carried out by Kraft, Safeway and Mission Foods, as well as Aventis' withdrawal of its license for StarLink corn. These were clearly prudent and necessary steps given the very real concerns about the potential allergenicity of Cry9C, the bacterial endotoxin engineered into StarLink corn, that were raised in this Scientific Advisory Panel's report of June 29, 2000 .

However, we cannot consider the matter ended here. Contaminated corn products are still on supermarket shelves. As we speak, consumers may be developing allergies or experiencing allergic reactions to Cry9C. Farmers and mill workers are likely exposed to high levels of this Bt toxin through inhalation of pollen and corn dust. In addition, farmers who planted StarLink are suffering economically because of the need to segregate corn, test for contamination of other varieties of corn, and arrange for separate storage and shipping. So I would like to comment on lessons that might be learned from this regulatory breakdown in the hope that such reckless action can be prevented in the future.

Massive, uncontrolled experiment on the American public

With the approval of StarLink corn in May of 1998, Aventis and the EPA launched a massive and uncontrolled allergenicity experiment on the American public. I say this because contamination of conventional corn with StarLink was an inevitable outcome of planting it. The EPA's belief that limiting the StarLink license to corn for animal feed and industrial use would be sufficient to keep the corn out of the food chain reveals a surprising ignorance of real world conditions on the farm. Bureaucratic barriers cannot stop pollen flow; one cannot rely on a corporation headquartered in France to effectively communicate planting and sales restrictions to farmers in Kentucky, especially when the corporation and its agents are motivated to sell as much seed as possible; and busy grain elevator operators cannot be expected to run DNA tests to segregate shipments of animal feed from food-grade corn.

In fact, there are numerous reports that Aventis and its agents failed to inform many farmers that StarLink was meant only for animal feed. Some farmers were not told that a buffer strip was required to prevent contamination of other corn varieties with the toxin gene. In short, Aventis clearly failed to honor its stewardship agreement with the EPA. The root problem here concerns a peculiarity in EPA's regulation of genetically engineered plant pesticides as opposed to chemical pesticides. With chemical pesticides, the EPA retains enforcement authority at point of use, while with plant pesticides this enforcement authority is delegated to the seed manufacturer by way of a stewardship agreement. This StarLink stewardship agreement created a blatant conflict of interest for Aventis. The company's interest in selling as much seed as possible was clearly in conflict with its duty to fully disclose to farmers the burdensome sales and planting restrictions imposed by the StarLink license. By informing farmers of these restrictions, Aventis risked loss of sales. In addition, the stewardship arrangement effectively puts the biotech company in charge of enforcing EPA's regulations. In this way, EPA passes the buck for regulating genetically engineered Bt corn to the corporation which engineers the seed. The result, as we have seen, is massive contamination of our food supply. EPA needs to immediately overhaul this aspect of genetically engineered crop regulation so as to eliminate this conflict of interest situation. The agency should retain and exercise authority to enforce the licensing conditions for genetically engineered crops at the point of use.

Downplaying the risk

Friends of the Earth is also concerned by numerous recent statements downplaying the allergenic risk of Cry9C. Steve Taylor, a food scientist with the University of Nebraska, actually told St. Louis Post Dispatch readers:
"Nobody's going to get sick. Eat your tacos and don't be concerned."

Fortunately, Kraft recalled its taco shells just days later. Mr. Taylor has also categorically denied that Cry9C is allergenic, though in the next breath he admits it could give people allergies . These irresponsible statements are unfortunate, but what concerns us more is what our federal authorities are saying. In an EPA press release on StarLink dated October 12th, Stephen Johnson, EPA's Deputy Assistant Administrator for Pesticides, said that "the agency believes the risks, if any, are extremely low." This statement is a far cry from the evaluation offered less than a month ago by federal officials, who called the discovery of StarLink corn in Kraft taco shells a "very serious" development . It also differs greatly from Mr. Johnson's own statement of September 20th in an NPR interview concerning StarLink, in which he said: "some people can get very, very sick if it's a severe food allergy. So… we take the issue of allergenicity very, very seriously." My question for the EPA is this: Why has the allergenicity risk of Cry9C changed from "very serious" to "extremely low"? Has our scientific understanding of this protein suddenly advanced in the past month? Does the EPA have new information that lays to rest the concerns expressed in the SAP's June 29th report? If so, I would like to hear it.

If EPA officials are waffling, what do food allergy experts have to say? According to this Panel, Cry9C has the two features regarded as "the best available criteria presently known [for food allergens]: 1) heat stability and 2) resistance to digestion." The Panel goes on to say that "these properties can be used as screening tools…" Cry9C's molecular weight of 68 kDa also fits into the upper end of the weight range of known allergens (10-70 kDa). Clearly, what the "old" EPA stated one month ago was correct. Contamination of food with Cry9C was and is a very serious matter. The "new" EPA needs to explain the basis for its revised assessment of "extremely low" risk. Well, that will be difficult to do, for a number of reasons.

Shoddy corporate science

First, the studies submitted by AgrEvo (Aventis' predecessor) purporting to show that Cry9C is not allergenic are deficient in a number of ways. EPA criticized AgrEvo's rat study as follows: "The brown Norway rat model is not a validated test for food allergy at this point and the study submitted is significantly flawed…" These significant flaws include AgrEvo's failure to report crucial data, including the concentration of Cry9C in the extracts injected into the rats, the number of rats that were tested, and the time that elapsed between injections of serum and challenge antigens . These are not minor oversights, but rather serious breaches of basic scientific protocol; a high school biology student would be flunked for less. Also dubious are the so-called "testimonial letters" of field employees of the Garst Seed Company, the largest seller of StarLink corn, which were submitted as evidence of no allergic reactions. A closer look reveals that the testimony comes not from Garst workers, but rather from ten Garst field managers. These managers submitted form letters reporting that they had not observed allergic reactions to StarLink in the employees under their supervision. Given Garst's obvious conflict of interest, this is hardly "testimony" that can be given much weight.

The amino acid homology comparisons - conducted to determine whether the Cry9C protein is similar in structure to any known allergens - were criticized by the SAP as not taking account of the crucial three-dimensional nature of the allergy-causing parts of allergenic proteins. In addition, there are still many allergens and epitopes whose sequences are unknown.

Most of the registrant's studies were not conducted according to standards of good laboratory practice; test protocols and other crucial data are routinely missing. In one study on the stability of Cry9C, the supposedly non-StarLink controls were contaminated with the Cry9C protein!

In short, the EPA and the SAP based their evaluations of StarLink corn on shoddy corporate science. Given AgrEvo's obvious interest in approval, we cannot help but wonder why it left out crucial data. What did the company have to hide if this corn were in fact safe for animals and people, as it claimed?

No firm basis for risk assessment

To return to Mr. Johnson's assessment of Cry9C's "extremely low" risk. With possibly millions of people exposed to this potential allergen, a seat-of-the-pants qualitative risk assessment will not do. The stakes here demand a much more serious quantitative appraisal of risk. Quantitative risk assessment requires a threshold dose as a reference point - in the case of carcinogens, this threshold dose is the amount of substance which gives one in one million exposed people cancer. Once that's established, you can define what a low or high risk dose is - put a solid number to it. But without this reference point, high and low risk are just relative terms, really little more than subjective opinion. So, what is Mr. Johnson's reference point for stating that Cry9C poses an "extremely low" risk, or none at all? Well, no threshold dose has been established for Cry9C because, like all other Bt proteins, it has been exempted from the requirement of a tolerance - this means that there are absolutely no legal limits on how much of this stuff can be expressed in genetically engineered corn.

Now, the public has been told that Cry9C is not an allergenic threat because it is present at such low levels . But what is low? What does low mean in the context of allergenicity? Mere nanograms of peanut allergen - billionths of a gram - can elicit an allergic response, even anaphylaxis, in sensitive individuals. Dust mites can also trigger allergic symptoms at this level. With respect to developing an allergy, the SAP report states that "smaller doses are more likely to sensitize the immune system than to induce tolerance."

According to EPA's data , whole plant levels of Cry9C in StarLink are 250 micrograms per gram of fresh corn tissue. This is extremely high relative to peanut allergen. And it's also more than 50 times higher than the level of Cry protein found in any other Bt corn.

Another factor to consider is the variability that is inherent to living things. For instance, Dr. Allan Felsot , an environmental toxicologist, cites data on levels of Cry protein in Bt corn that differ greatly from those found in the EPA's Registration Action Document, in one case by a factor of 50 times. How does one account for these discrepancies? One likely explanation is that we are dealing with living things here, not inert chemicals. Living organisms interact with the environment, they behave quite differently under differing weather and soil conditions, they cross-pollinate, and they have a large range of natural genetic variability. You simply cannot do a study on a Bt corn plant and say that the results represent all other plants in the field - much less all plants of its variety. Genetically engineered plants are alive, and cannot be regulated with a system developed for inert chemicals.

This natural variability is especially worrisome when we consider children eating pesticidal plant products. We know that children are especially prone to allergies, as reflected in their much higher incidence of allergic conditions. Federal law requires the EPA to apply a tenfold margin of safety to account for the special sensitivity of infants and children to pesticides, including plant pesticides. This means that the exposure limit for children is set at one-tenth of that considered safe for adults. Nevertheless, the EPA apparently did not require any special studies of Cry9C's risk to infants and children. And the tenfold margin of safety cannot be applied because the agency has failed to establish any reference exposure limit for adults.

So the question becomes this: We have reams of data on genetically engineered Bt crops, but of what use are all these data when the EPA, FDA and USDA have basically washed their hands of their responsibility to regulate these crops?

Regulatory mayhem

To recap, here's the situation we are facing: The EPA approves a variety of genetically engineered corn that contains a potential food allergen for animal feed only. It waives all limits as to how much of this potential allergen the corn can contain. It then delegates on-site regulatory authority for this corn to the biotech seed company. This seed company, anxious to sell its GE corn, apparently neglects to tell many farmers it can't be sold for food. The corn is planted on over 300,000 acres. It gets into the food supply. Millions of consumers eat food laced with a potential allergen. Neither the EPA nor the FDA knows this; neither agency even has the means to test for this potential allergen. Friends of the Earth and other environmental and consumer groups have products tested and find the suspect corn. A food company confirms the test. The FDA calls in the biotech seed company to teach it how to do the test. The food company spends tens of millions on product recalls. Consumers complain to the FDA. The FDA passes the buck to the EPA. The EPA pressures the seed company to withdraw its license. The USDA begins buying back the suspect corn, costing the seed company millions. Then, after finally taking appropriate action, the EPA suddenly turns around and claims the risk from the corn contaminant is "extremely low" or non-existent, with no reliable evidence to back up this claim.

Clearly, our regulatory system for genetically engineered foods is broken, and desperately needs to be fixed. As a start, Friends of the Earth calls for a halt to the re-registration of Bt corn crops due to the EPA's failure to use allergy testing capability available to the agency for over a year. We also call for mandatory, stringent, long-term safety testing of every genetically engineered food; GE foods should only be permitted on the market if they fully pass all relevant tests. We demand labeling of all genetically engineered foods; not only do large majorities of consumers favor this, but labels are also the crucial first step in tracing back any health problems that may arise from eating these novel foods. Labels and identity preservation from farm to consumer could have prevented the ongoing StarLink contamination. Finally, we also demand clear rules that place the liability for any damage caused by GE foods squarely on the biotech originators of these crops. Farmers should never be made to pay for the irresponsible actions of the biotech seed companies.

In conclusion, I should mention that Friends of the Earth, the Center for Food Safety and the FDA have received at least 20 reports from individuals who experienced adverse effects after eating Kraft Taco Bell taco shells; there has also been one reported death alleged to be related to the consumption of contaminated taco shells. We call on the FDA to thoroughly investigate these reports for any possible connection to StarLink corn and its Cry9C endotoxin. All potential adverse health effects from StarLink corn must be investigated carefully and immediately. Even if none are discovered, research on possible allergens in all GE foods should be stepped up. Nothing less than the safety of our food supply and the health of all Americans is at stake.