April 14, 2009
Q&A: EU Testing of Genetically Modified Foods
Q: What are genetically modified foods?
A: Through modern biotechnology, numerous plants, animals and microorganisms have been modified genetically to produce characteristics desirable for agriculture (e.g., increased productivity, resistance to inclement conditions, and biological synthesis of a toxin or pesticidal chemical).
Q: What foods have been genetically modified and by which companies?
A: High-profile examples of genetically modified (GM) foods include Monsanto’s “Roundup-Ready” herbicide-tolerant soybeans and “Bt” corn, cotton and potatoes, which have been engineered to produce the Bacillus thuringinesis insecticidal toxin, as well as the cold- and rot-resistant “Flavr Savr” tomato from the California-based biotechnology company, Calgene. In addition, regulatory authorities around the globe list [PDF] numerous other varieties of alfalfa, canola, coffee, potato, rice and other crops that have been genetically modified for eventual commercialisation.
Q: Why test GM foods?
A: Whereas plants that have been consumed for generations are generally assumed to be safe based on their history of use, foods derived from GM crops may or may not be equivalent to their unmodified counterparts either nutritionally or in terms of safety to humans and/or the environment. It is therefore necessary to characterise the parent crop and donor organism(s) from which the new genetic material is derived, and evaluate the safety of newly introduced gene products (proteins and metabolites), and possibly the whole GM crop/food itself. Ultimately, a side-by-side comparison is made to determine whether or not GM and non-GM crops are substantially equivalent.
Q: How are GM foods regulated in the EU?
A: Numerous EU regulations have been established to control the creation, study, release and transport of genetically modified organisms. Among the most significant from a toxicological testing standpoint are Directive 90/220/EC on the deliberate release into the environment of GMOs, and Regulation (EC) 1829/2003 [PDF] concerning the placing on the market of GMOs intended for food or feed and of food or feed products containing, consisting of, or produced from GMOs feed. This legislation is administered by the European Commission Directorate General for Health and Consumer Protection, with scientific input from the European Food Safety Authority Panel on Genetically Modified Organisms (EFSA GMO Panel).
Q: What animal tests are carried out on GM foods?
A: The framework Directive requires the submission of a technical dossier of information including a full health and environmental risk assessment, which in turn should include information on potential "pathogenicity: infectivity, toxigenicity, virulence, allergenicity, carrier (vector) of pathogen, possible vectors, host range including non-target organism," as well as "toxic or allergenic effects of the non-viable GMOs and/or their metabolic products." Interpretive guidance developed by the EFSA GMO Panel states that requirements of toxicological testing in the safety assessment of food/feed derived from GMOs "must be considered on a case-by-case basis and will be determined by the outcome of the assessment of the differences identified between the GM product and its conventional counterpart, including available information on intended changes."
In the case of newly expressed proteins, toxicological assessments generally emphasize short-term effects (e.g., 1-month repeated-dose general toxicity studies in rodents, allergenicity), whereas the core set of tests for new constituents other than proteins normally consists of the following:
- three month repeated-dose general toxicity in rodents
- 12-24 month repeated-dose general toxicity and cancer studies in rats and mice
- Genetic toxicity studies of at least three varieties
- Reproductive toxicity in at least two generations of rodents
- Pre-natal developmental toxicity in rodents and rabbits
- Metabolism Absorption, distribution, metablolism and elimination studies in rodents
In some cases, testing of whole GM foods may be required, generally using a modified 3 month feeding study in rodents. In addition, where doubt exists regarding wholesomeness and nutritional value of a GM food, separate feeding studies in fast-growing farmed animal species (e.g., broiler chickens or lambs) may be carried out.
Q: Are animal tests accurate predictors of GM food safety?
A: Not necessarily. Animal “models” have proven to be a questionable standard for judging a GM food’s allergy-inducing potential in humans. For example, assessment of the Brazil nut albumin protein in transgenic soybean for potential allergenicity in both rats and mice did not elicit an allergic or immune response, leading to the erroneous conclusion that there was no allergenic protein transfer to the soybean . On the basis of this and similar examples, scientific authorities such as the Royal Society of Canada have concluded [PDF] that “there is no animal model currently able to predict accurately human allergic responses and therefore donor protein allergenicity.” König and colleagues have likewise noted in regard to animal tests for food allergy that “[v]arious approaches have been proposed, although it must be recognised that none of these has yet been evaluated fully or validated” .
According to EFSA, there are also major obstacles to the application of conventional toxicity test guidelines to animal feeding studies of whole foods. Firstly, whereas food additives and other chemicals are small, well-defined molecules that can easily be administered at high concentrations, whole foods are complex mixtures of thousands of chemicals that are bulky, filling, and will only be consumed until an animal’s appetite has been satisfied. This limits animals’ exposure to potentially harmful gene products, and by extension, may diminish the test’s ability to detect more subtle adverse effects of the GM food, according to EFSA. On the other hand, increasing the proportion of GM food in the animals’ diet can affect not only palatability, making the animals more or less likely to eat enough, but can also lead to nutritional imbalances (dilution or overload), “which would severely mask or overshadow possible adverse effects” . For example, rats in one study were fed the human equivalent of 13 kg of GM tomatoes daily for 91 days – slightly below the level at which the tomatoes’ high potassium content would have been expected to cause kidney toxicity . An additional limitation inextricably linked to animal tests of whole foods is the lack of adequate standardisation across laboratories conducting such tests. For example, König et al. (8) have noted that “the design and preparation of the test and control diets, and the amount and duration over which they are administered would benefit from standardisation, to allow for better comparison of results from feeding trials in different laboratories.”
Q: What is the value of animal nutrition studies?
A: Determination of a GM food’s nutritional value is accomplished primarily by means of a compositional analysis, which measures the levels of protein, carbohydrate, fat, vitamins and minerals, as well as any toxins or so-called anti-nutrients. Comparisons can then be made against established reference values for conventional foods. According to EFSA [PDF]: “Once compositional, phenotypic and agronomic equivalence [between a GM and non-GM crop] has been established, nutritional equivalence may also be assumed, and feeding trials with target animals add little to the nutritional assessment of the feed.” König and colleagues  have likewise stated that “animal nutrition studies are not generally viewed as an essential, sensitive, and specific element of the safety assessment of food, feed, or processed fractions derived from genetically modified crops….”
Q: What are the alternatives?
A: A number of non-animal techniques can be used to evaluate the potential allergenicity of GM foods. For example, the following elements could be combined in an integrated, non-animal testing strategy (adapted from Metcalfe et al.:
- Computerised structure-activity relationship analysis to identify similarities between a novel protein and known allergens.
- Analysis of the stability and digestibility of a novel protein under simulated gastric conditions in vitro, based on the knowledge that proteins resistant to digestion are more likely to be allergenic.
- Targeted blood using serum from human allergy patients.
Q: Are there other developments that could impact GM food testing on animals?
A: Despite the well-established limitations of testing whole foods on animals, some EU member states (e.g., Austria) and other organisations have called for extensive animal-based toxicological testing to demonstrate the safety and nutritional equivalence of GM foods. For example, Greenpeace has asserted that: “Allergenicity tests as well as toxicity and antinutrition tests have to undergo a detailled, comprehensive and mandatory testing regime. (Testing regimes for toxicity should at least follow pesticide regulations…).” In practical terms, the preceding statement is calling for extensive, multi-species testing using upwards of 10,000 animals to evaluate a GM food’s potential to cause reproductive toxicity and birth defects, neurological impairment, cancer, and a host of other toxic effects. Yet available evidence indicates that “no dietary proteins are known to date to be directly associated with teratogenic, mutagenic, or carcinogenic effects in animal models” . Moreover, analyses by Kroes et al.  and Cheeseman et al.  have demonstrated that specialised reproduction studies are not normally more sensitive in their detection of adverse effects than 90-day general toxicity studies. The same is true with respect to the sensitivity of 90-day versus 1+year studies – the former having been found [PDF] to detect, with very few exceptions, all relevant toxicological findings identified in the latter   . In addition, EFSA has reported [PDF] that its toxicologists “have become aware that 90-day studies may demonstrate shortcomings, such as masking effects due to aging of the animals, and that this type of effect may only increase with longer term studies.”
Q: What is HSI Europe doing to help animals used in GM food testing?
A: HSI Europe has been at the forefront of lobbying efforts to ensure that all available, validated non-animal methods and testing strategies achieve expeditious regulatory acceptance in the EU. Additionally, HSI Europe and affiliate organisations The Humane Society of the United States and Humane Society Legislative Fund have assumed a leading role in supporting implementation of the vision of “twenty-first century toxicology” articulated by the U.S. National Research Council, which would see animal tests that are decades old, costly, slow and of dubious relevance to people replaced by ultra-modern, efficient and human-relevant non-animal methods. HSI Europe is calling for a “big biology” project to meet this challenge, akin to the Human Genome Project of the 1990s, and are forging an international, multi-stakeholder consortium make this landmark vision a reality as quickly as possible.
4Noteborn HPJM & Kuiper HA. Safety assessment strategies for genetically modified plant products. Case study: Bacillus thuringinesis-toxin tomato. In: Proceedings of the 3rd International Symposium on the Biosafety Results of Field Tests of Genetically Modified Plants and Microorganism (Jones DD, ed.). Oakland, CA: University of California (1994).
5Metcalfe DD, Astwood JD, Townsend R, et al. Assessment of the Allergenic Potential of Foods Derived from Genetically Engineered Crop Plants. Critical Reviews in Food Science and Nutrition 36(Special), S165-86 (1996)
6Kroes R, Renwick AG, Cheeseman M ,et al. Structure-based thresholds of toxicological concern (TTC): Guidance for application to substances present at low levels in the diet. Food and Chemical Toxicology 42, 65-83 (2004).
8Betton G, Cockburn A, Harper E, et al. A Critical Review of the Optimum Duration of Chronic Rodent Testing for the Determination of Non-Tumourigenic Toxic Potential. A Report by the BTS Working Party on Duration of Toxicity Testing. Human and Experimental Toxicology 12, 221-32 (1994).
9Box RJ & Spielmann H. Use of the dog as non-rodent test species in the safety testing schedule associated with the registration of crop and plant protection products (pesticides): present status. Archives of Toxicology 79, 615-26 (2005).