By Jeffrey Smith
Institute For Responsible Technology
Saturday, Jan 1, 2011
Remember the pictures of the fish tomatoes? For years they were an unofficial emblem of the anti-GMO movement. They depicted how anti-freeze genes from an Arctic fish were forced into tomato DNA, allowing the plants to survive frost. Scientists really did create those Frankentomatoes, but they were never put on the market. (Breyers low-fat ice cream, however, does contain anti-freeze proteins from Arctic fish genes, but that’s another story.)
The tomato that did make it to market was called the Flavr Savr, engineered for longer shelf life. Fortunately, it was removed from the shelves soon after it was introduced.
Although there are no longer any genetically modified (GM) tomatoes being sold today, the FDA’s shady approval process of the Flavr Savr provides a lesson in food safety—or rather, the lack of it—as far as gene-spliced foods are concerned. We know what really went on during the FDA’s voluntary review process of the Flavr Savr in 1993, because a lawsuit forced the release of 44,000 agency memos.
(Those same memos, by the way, also showed that FDA scientists had repeatedly warned their superiors about the serious health risks of genetically modified organisms [GMOs]. They were ignored by the political appointees in charge, who allow GMOs onto the market without any required safety studies.)
Calgene, the tomatoes’ creator-in-chief (now a part of Monsanto), voluntarily conducted three 28-day rat feeding studies. Before I share the gory details, I must commend the Calgene scientists who were committed to transparency and full disclosure with the FDA. Unlike all other subsequent voluntary submissions from biotech firms to the agency, Calgene provided detailed feeding study data and full reports. Dr. Belinda Martineau, one of Calgene’s tomato makers, writes in First Fruit about their commitment to an open process while they attempted to introduce the world’s first GM food crop.
Calgene tested two separate Flavr Savr tomato lines. Both had the same gene inserted into the same type of tomato. The process of insertion and the subsequent cloning of the cells into GM plants can cause lots of unique and unpredicted consequences. The two lines, therefore, were not considered identical.
The rats that ate one of these Flavr Savr varieties probably wished they were in a different test group. Out of 20 female rats, 7 developed stomach lesions—bleeding stomachs. The rats eating the other Flavr Savr, or the natural tomatoes, or no tomatoes at all, had no lesions.
If we humans had such effects in our stomachs, according to Dr. Arpad Pusztai, a top GMO safety and animal feeding expert, it “could lead to life-endangering hemorrhage, particularly in the elderly who use aspirin to prevent thrombosis.”
The lab that performed the study for Calgene acknowledged that the results “did suggest a possible treatment related” problem. FDA scientists repeatedly asked Calgene to provide additional data in order to resolve what they regarded as outstanding safety questions. The director of FDA’s Office of Special Research Skills wrote that the tomatoes did not demonstrate a “reasonable certainty of no harm,” which is the normal standard of safety. The Additives Evaluation Branch agreed that “unresolved questions still remain,” and the staff pathologist stated, “In the absence of adequate explanations by Calgene, the issues raised by the Pathology Branch … remain and leave doubts as to the validity of any scientific conclusion(s) which may be drawn from the studies’ findings.”
Oh yeah, some rats died
The team that had obtained the formerly secret FDA documents sent the full Flavr Savr studies to Dr. Pusztai for review and comment. While reading them, he happened across an endnote that apparently the FDA scientists either did not see or chose to ignore. The text nonchalantly indicated that 7 of the 40 rats fed the Flavr Savr tomato died within two weeks. The dead rats had eaten the same tomato line as those that developed lesions. In the other groups, fed the other Flavr Savr line, a natural tomato control, or a water control, only one rat had died.
But the endnote summarily dismissed the cause of death as husbandry error, and no additional data or explanation was provided. The dead rats were simply replaced with new ones.
When I discussed this finding with Dr. Pusztai over the phone, he was beside himself. He told me emphatically that in proper studies, you never just dismiss the cause of death with an unsupported footnote. He said that the details of the post mortem analysis must be included in order to rule out possible causes or to raise questions for additional research. Furthermore, you simply never replace test animals once the research begins.
Questionable follow-up study
Calgene repeated the rat study. This time, one male rat from the non-GM group of 20, and two females from the GM-fed group of 15, showed stomach lesions. Calgene claimed success. They said that the necrosis (dead tissue) and erosions (inflammation and bleeding) were “incidental” and not tomato-related. The FDA staff pathologist, however, was not convinced. He responded that “the criteria for qualifying a lesion as incidental were not provided.” Further, he said that the disparity between the studies “has not been adequately addressed or explained.”
In reality, the new study was not actually a “repeat.” They used tomatoes from a different batch and used a freeze-dried concentrate rather then the frozen concentrate used in the previous trial. Dr. Martineau explained to me that by freeze-drying, it allowed them to put more of the concentrated tomato into each rat. But Dr. Pusztai said that altering the preparation of the food can lead to different results. He also pointed out that humans were more likely to consume frozen concentrate compared with freeze-dried.
In spite of the outstanding issues, the political appointees at the FDA concluded that the lesions were not related to the GM tomatoes. To be on the safe side, however, Calgene on its own chose not to commercialize the tomato line that was associated with the high rate of stomach lesions and deaths. The other line went onto supermarket shelves in 1994.
Faulty science rules the day
This was the very first GM food crop to be consumed in the US. It was arguably the most radical change in our food in all of human history. It was the product of an infant science that was prone to side-effects. Yet it was placed on the market without required labels, warnings, or post-marketing surveillance. One hopes that the FDA would have been exhaustive in their approval process, holding back approvals until all doubts were extinguished. But the agency was officially mandated with promoting biotechnology and bent over backwards to push GMOs onto the market. As a result, their evaluation was woefully inadequate.
Having discovered problems in the stomach, for example, Dr. Pusztai said they should have looked further down the digestive system at the intestines as well, but they didn’t. They should have increased the number of animals in the experiment to strengthen the findings, but they didn’t. And they should have used young (e.g. month-old) and pregnant animals as is done with pharmaceutical studies, but they didn’t.
They did, however, use rats with vast differences in starting weights. This invalidates any conclusions that there were no significant differences in weight gain, feed intake, or organ weights between GM- and non-GM-fed groups. The starting weights in the Flavr Savr experiment ranged from 130 to 258 grams for males, and 114 to 175 grams for females. Contrast that with the hundreds of rat feeding trials conducted by Dr. Pusztai, where the starting weights were within a range of 1 or 2 grams.
Dr. Pusztai also pointed out that the experimental tomatoes were grown at different locations and harvested at different times, which further increases the variability of results.
The FDA’s defense that the bleeding stomachs did not come from the Flavr Savr diet was also an exercise in faulty science. They blamed the lesions on mucolytic agents in the tomato (i.e. components that dissolves thick mucus); but according to Dr. Pusztai, tomatoes are not known to contain mucolytic agents. The FDA also claimed that it might be the food restriction in the rats’ diet—but the rats ate as much as they wanted. Or maybe it was the animal restraint—but the rats were not restrained.
The explanation that stuck to the wall was that the process of force-feeding the tomatoes through tubes was the reason for the stomach lesions. But as Dr. Pusztai and FDA scientists both observed, there was no adequate explanation as to why the rats fed GM tomatoes in the earlier study had the higher rate of lesions.
Dr. Pusztai said the “study was poorly designed and executed and, most importantly, led to flawed conclusions.” He warned, “the claim that these GM tomatoes were as safe as conventional ones is at best premature and, at worst, faulty.”
Fortunately, the Flavr Savr tomatoes lacked flavor. They also got mushy (unless they were handled in such a way that the company spent more money getting them to market than it could sell them for). They were taken off the market by the time Monsanto bought Calgene in 1997.
After the Flavr Savr’s superficial review and controversial approval, no subsequent GMO producer has ever presented such detailed safety test data to the FDA.
I write about the Flavr Savr in Genetic Roulette: The Documented Health Risks of Genetically Engineered Foods. Two biotech advocates, Drs. Chassy and Tribe, created a GMO disinformation site that allegedly discredits all 65 health risks highlighted in the work. I have already shown that their attack on the first risk, Dr. Pusztai’s potatoes, was based on pure PR spin and scientific sleight of hand. Below I respond to their accusations regarding the Flavr Savr.
1. (Chassy and Tribe) FDA records clearly show that experts stated that the process of introducing stomach tubes can damage the rats’ stomachs and/or end up placing test material in the lungs. . . The reader is not told that regulators approved the tomato because their concerns had been fully satisfied that the GM tomato was not toxic.
As indicated in Part 1, the actual scientists at the FDA wrote memo after memo declaring that the higher rates of lesions in the GM-fed group could not be explained away, and that they were not fully satisfied by the explanations. The discrepancy between what the political appointees at the agency stated publicly, and the concerns expressed in private memos by the scientific staff, has been clearly documented.
In fact, one memo reveals that during their Flavr Savr review, the FDA was making blatant and possibly illegal exceptions. One person wrote, “It has been made clear to us that this present submission [Flavr Savr] is not a food additive petition and the safety standard is not the food additive safety standard. It is less than that but I am not sure how much less.” According to attorney Steven Druker, who is an expert in US food safety law, the FDA’s own regulations clearly state that a lower standard should not have been applied in this instance.
As for the stomach lesions, without repeating the study with the same tomatoes, in the same concentration, with larger sample sizes, we can’t be confident that the GM line was the cause. But likewise, we can’t be confident that they were not. It’s another example of too few data.
2. No real differences were seen between groups of animals in the study. Contrary to Smith’s claims, expert pathologists stated that mild gastric erosions were seen at similar levels in both GM and non-GM fed rats.
This is quite a bizarre statement, given that seven female rats had stomach lesions in the first study, compared to none in the other feeding groups. Even the experimenters said that the results suggest a treatment-related effect. I guess if you completely ignore the main rat study in question, which apparently Chassy and Tribe would like us all to do, then you will not see significant differences. But putting blinders on to ignore inconvenient evidence does not prove safety or demonstrate good science.
3a. Rats might have been injured . . . by accidental administration of test material into the lung instead of the stomach.
3b. Gastric lesions can be caused by acidosis brought on by fasting.
Neither of these arguments address why 7 of 20 females fed GM tomatoes had lesions while the controls, reared under the same conditions, did not. Furthermore, since the rats did not fast but ate as much as they wanted, why would they throw in this irrelevant point (if not to obscure the truth)?
4. Smith is actually asking the reader to believe that the FDA would approve a lethal product.
Believe it! The FDA approves lethal products all the time. According to a report by the United States General Accounting Office, more than half of the drugs approved by the FDA between 1976 and 1985 had severe or fatal side-effects that had not been detected during the agency’s review and testing. In other words, after drug companies spent an estimated 12 years and $231 million dollars to research, test, and secure new drug approval through a very hands-on FDA approach, most of the drugs had to be taken off the market or required major label changes due to missed safety issues.
With GMOs, the situation is far more dangerous. The FDA doesn’t require a single study, the complex biology of GM crops may produce far more side-effects than drugs, GM foods are fed to the entire population, and they are not labeled or monitored, so symptoms are difficult or impossible to track.
5. There is no evidence of animal deaths. . . . Smith may have confused the words necrosis and dead cells with animal deaths. Careful reading reveals that the regulatory record does not mention any animal deaths which surely would have been of concern had they occurred. . . . This claim (in Pusztai and others 2003) appears to be blatantly untrue.
They would hope it was untrue. But just because they didn’t have access to the 44,000 documents made public from the lawsuit does not mean that the deaths did not occur. I can assure you they did, and that Dr. Pusztai, widely recognized as the world’s leading expert in his field and author of more than 300 studies, would not mistake dead cells for animal deaths.
In fact, on page 18 of the IRDC Report, it refers to “necroscopy data” on each animal. Necroscopy is an examination of a dead body, not dead cells.
The reason why the FDA scientists did not raise this issue is that they apparently either did not read the endnote, or simply accepted the unsupported conclusion on face value, which said that the necroscopy suggested that the deaths were due to a husbandry error and not test-article related. Even the Calgene scientists didn’t raise eyebrows at the finding. It wasn’t until a highly experienced animal feeding study expert like Dr. Pusztai reviewed the original papers that this oversight became apparent.
6. Interestingly, eating too many tomatoes can kill rats.
It is odd that Chassy and Tribe first claim that no rats died and then try to argue that if rats did die, tomato overdose could be the culprit. Since all the rats were fed under similar conditions, their killer-tomato argument fails to explain why 7 of 40 GM-fed animals died, compared to only 1 in the other groups.
7. These products are assessed carefully for safety before they are marketed, and—more importantly—there is no scientific reason to believe they pose and (sic) new or different risks.
To claim that there are no new potential health hazards from GMOs is absurd. Fran Sharples, the Director of the Board on Life Sciences at the US National Academy of Sciences (NAS), told me, “The academies have issued numerous reports on assessing the risks of transgenic plants. If the academy believed there were no such potential risks, why would we have delved into these matters in these reports?” One of those NAS reports even acknowledged that the current system of regulating GMOs might not detect “unintended changes in the composition of the food.”
The Royal Society of Canada stated that it is “scientifically unjustifiable” to presume that GM foods are safe and that the “default presumption” is that unintended, potentially hazardous side-effects are present. A WHO spokesperson said that current regulations are not adequate to determine the health effects; the Indian Council of Medical Research called for a complete overhaul of existing regulations; and the American Academy of Environmental Medicine called for a moratorium of GM foods altogether.
Since Chassy and Tribe are fond of using the FDA policy as support for their position, I am happy to quote Linda Kahl, an FDA compliance officer, who directly contradicts their ridiculous assertion. In a memo that summarized the position of FDA scientists about GMOs, she stated, “the processes of genetic engineering and traditional breeding are different, and according to the technical experts in the agency, they lead to different risks.”
What’s Chassy and Tribe’s real motive?
Many of the arguments presented by Chassy and Tribe are easily and completely countered by the evidence. If one were feeling especially generous, one might guess that they simply weren’t aware of the strong concerns voiced in quotes by FDA scientists, the incidence of stomach lesions in the first study, or the fact that the rats didn’t fast. But these points were contained within the very passage of Genetic Roulette that they were supposedly critiquing. If they actually read the book, which we must assume they did, then they absolutely knew that their counter-arguments were directly contradicted by FDA memos and study reports, and thus were utterly false.
Why then did they construct their website in the first place? It appears that they are not really motivated to make cogent scientific counter-arguments, but instead are hoping that the readers blindly accept their baseless condemnation of Genetic Roulette and never actually read the book.
This tactic is similar to other techniques used by the biotech industry that I describe in Genetic Roulette. GMO advocates, for example, often write up lengthy studies or reports that hardly anyone ever reads in detail. Instead, people generally look at the abstract and/or conclusion and accept the authors’ declaration that the findings demonstrate GMO safety. But when an expert actually takes the time to go through the details, he or she discovers that the conclusions are entirely unsupported and unjustified. In some cases, they are in direct opposition to the data.
It took the biotech industry three years to create their so-called academic review of Genetic Roulette. The mere fact that after all that time they could not put together even the most basic scientific arguments is a tribute to the authenticity of the book. If they could have used science to counter it, they would have. But they didn’t. They used spin.
It will continue to be my delight to go through each of their pages to expose their “scientific” sleight-of-hand. But I am much more motivated to spend my time taking steps that will end the genetic engineering of the food supply, rather than trying to convince the handful of people who accidentally wander onto Chassy and Tribe’s disinformation site. So have patience.