Concepts and Principles

For the past 10,000 years agriculture has relied on plant and animal breeding, which except for the past 100 years, has been conducted on a trial and error basis. In the case of plants, this meant selection based on healthy appearance, vigorous growth, higher yields, and desirable appearance, taste and smell of the edible portions. Since the beginning of the 1900s, breeding has been performed on a more scientific basis. In the past 20 years, the application of molecular biological tools has allowed the production of plants and animals with traits that could not have been introduced through traditional breeding techniques.

Food is more than just a source of nutrition and energy, it has a highly symbolic significance and is one of the ways in which human cultures differentiate themselves. Not surprisingly, our beliefs about the safety and acceptability of food are based more on culture and tradition than on objective food safety testing. Although traditional plant selection and breeding might have included an evaluation of safety, it was not formally recognized and in any event, there was little documentation of processes for establishing the safety of new foods.

The diversity of new traits that can be introduced into food using biotechnology has challenged our traditional approach to food safety. Novel foods produced using biotechnology are being scrutinized more than any other food, and the public expectation with respect to assuring the safety of these products is higher than for any food produced using more conventional methods. Significantly, this more rigorous evaluation of novel foods has also served to emphasize our limited knowledge of the normal levels and variability of food constituents, including their short and long-term health effects.

Although modern biotechnology broadens the scope of genetic changes that can be introduced into organisms used for food, it does not inherently result in foods that are less safe than those produced by more conventional techniques. This principle has important ramifications for the safety assessment of genetically engineered and novel foods. It means that a new or different standard of safety is not required, and that previously established principles for assessing food safety still apply. Moreover, the inherent precision of molecular biological methods for introducing specific genetic changes should enable a more direct and focused assessment of safety.

Substantial Equivalence

To date, the safety assessment of genetically engineered and novel foods has been based on the principle that these products can be compared with traditional foods that have an established history of safe use. Furthermore, this comparison can be based on an examination of the same risk factors (i.e., hazards) that have been established for the traditional counterpart. The objective is to determine if the novel food presents any new or greater risks in comparison with its traditional counterpart, or whether it can be used interchangeably with its traditional counterpart without affecting the health or nutritional status of consumers. The goal is not to establish an absolute level of safety, but rather the relative safety of the new product such that there is a reasonable certainty that no harm will result from intended uses under the anticipated conditions of processing and consumption.

Accounting for processing and consumption patterns is important even for traditional foods. A number of plants consumed by humans are acutely toxic in the raw state, but are accepted as food because processing methods alter or eliminate their toxicity. For example, the cassava root is quite toxic, but proper processing converts it into a nutritious and widely consumed food. Soybeans and lima beans, among other crops, contain antinutrients (e.g., soybean trypsin inhibitor and lectins) and require proper processing. Potatoes and tomatoes can contain toxic levels of the glycoalkaloids solanine and alpha-tomatine, respectively. Thus, the presence of a toxicant in a plant variety does not necessarily eliminate its use. In considering the safety of the novel food, it is therefore important to examine the range of possible toxicants, critical nutrients or other relevant factors, as well as its processing, intended use and exposure.

This comparative approach has been embodied in the concept of substantial equivalence, which was developed before any new genetically engineered foods came to the market. It was first described in an Organization for Economic Cooperation and Development (OECD) publication in 1993 (OECD 1993). This document was the result of some 60 experts from 19 OECD countries who spent more than two years discussing how to assess the safety of genetically engineered foods. The concept of substantial equivalence was further endorsed by a Food and Agriculture Organization (FAO)/World Health Organization (WHO) joint expert consultation in 1996 (FAO/WHO 1996). It recognized that the establishment of substantial equivalence is not a safety assessment per se, but that establishing the characteristics and composition of the novel food as equivalent to those of a familiar, conventional food with a history of safe consumption means that the new product will be no less safe under similar consumption patterns and processing practices.

In the 1996 Joint Consultation, three possible scenarios arising out of a substantial equivalence evaluation were envisaged (FAO/WHO 1996):

1. When substantial equivalence has been established for an organism or food product, it is considered as safe as its conventional counterpart and no further safety evaluation is needed.


2. When substantial equivalence has been established apart from certain defined differences, further safety assessment should focus on these differences. A sequential approach should focus on the new gene product(s) and the(ir) structure, function, specificity and history of use. If a potential safety concern is indicated for the new gene product(s), further in vitro and/or in vivo studies may be appropriate.


3. When substantial equivalence cannot be established, this does not necessarily mean that the food product is unsafe. Not all such products will require extensive safety testing. The design of any testing programme should be established on a case-by-case basis, taking into account the reference characteristics of the food or food component. The objectives should be clear, and care should be taken in experimental design. Further studies, including animal feeding trials, may be required, especially when the new food is intended to replace a significant part of the diet.

One important benefit of the substantial equivalence concept is that it provides flexibility that can be useful in food safety assessment. It is a tool, which helps identify any difference, intended or unintended, that might be the focus of further safety evaluation. Because it is a comparative process for evaluating safety, the determination of substantial equivalence can be performed at several points along the food chain (e.g., at the level of the harvested or unprocessed food product, individual processed fractions, or the final food product or ingredient).

The most recent FAO/WHO joint expert consultation on foods from biotechnology (FAO/WHO 2000) concluded that application of the substantial equivalence concept contributes to a robust safety assessment framework. The Consultation was satisfied with the approach used to assess the safety of the genetically engineered and novel foods that have been approved for commercial use. Safety assessments based on the concept of substantial equivalence have been the most practical approach to addressing the safety of foods and food ingredients developed through modern biotechnology. In fact, there are presently no alternative strategies providing a better assurance of safety (FAO/WHO 2000).

As an example of the application of substantial equivalence, consider potatoes that have been genetically engineered to express the coat protein from potato virus Y (PVY), and thus display resistance to disease caused by this virus:

1. In non-transformed potatoes, the presence of viral coat proteins is due to natural viral infection. This is a common occurrence; therefore, these proteins have a long history of safe consumption.


2. Plant viral coat proteins (whether from PVY or any other plant virus) have never been associated with toxicity or cases of allergic reaction.


3. Consequently, transgenic potatoes expressing the PVY coat protein gene would be considered to be substantially equivalent to non-transformed potatoes infected with this same virus provided that the amounts expressed were not grossly different from the levels occurring in naturally infected tubers.

This analogy applies only to plant viral coat proteins expressed in the parts of the plant normally consumed (tubers) taking into account any adverse effects on glycoalkaloid levels and key nutrient starches, as well as consumption patterns.

Limitations of Substantial Equivalence

Applying the concept of substantial equivalence requires that sufficient analytical data be available in the literature, or be generated through analysis, to allow effective comparison between the novel food and its traditional counterpart. This suggests a basic limitation of the substantial equivalence concept: dependence on a comparator, and on the information that is available or can be generated for the comparator, means safety assurance is relative to the components assessed for the particular comparator. The choice of comparator is therefore crucial to effective application of the concept of substantial equivalence to establish the safety of a novel food. An appropriate comparator must have a well-documented history of use. If adverse effects have been associated with the particular food type, specific components considered to cause these effects should be described and well characterized to permit effective comparison.

Without exception, all of the novel foods approved to date have been the result of incorporating (or selecting for) one or two rather simple single-gene traits into plants. Except for canola and soybean varieties with modified oil composition (e.g., high lauric acid; high oleic acid) and delayed softening tomatoes for improved shelf-life, all of these traits have been targeted toward reducing agricultural inputs by conferring resistance to insects and/or viruses, or tolerance to environmentally friendly broad spectrum herbicides. These products were purposefully designed to be comparable in composition and nutritional quality with their traditional counterparts, thus making the demonstration of substantial equivalence (with defined differences) straightforward.

The next generation of products will be much more complex and will blur the boundary between foods and therapeutics. The product mix will include nutraceuticals, edible vaccines, and biopharmaceuticals produced in plants and animals. For these products, it will be much more difficult to find appropriate comparators and approaching the safety assessment from a substantial equivalence viewpoint may not be effective.

Unexpected Effects

As the following example(s) demonstrate, new products with intentionally altered nutritional profiles will challenge our ability to assess unintended consequences.

The first example relates to genetically engineered low-glutelin rice, created by introducing the glutelin-encoding gene in the antisense orientation, for commercial sake production. The decrease in glutelin levels was, however, associated with an unintended increase in levels of prolamins. The change in prolamin levels was not detected by standard nutritional analyses, such as total protein and amino acid profiles, but was only observed following sodiumdodecylsulfate (SDS)-polyacrylamaide gel electrophoresis (PAGE). While the change in prolamin levels did not affect the industrial application, it could affect nutritional quality and allergenic potential if the rice were used as a food.

A second example relates to genetically engineered “golden rice” designed to express increased levels of beta-carotene, a precursor to vitamin A. Unexpectedly, it was found that this modification was accompanied by higher levels of xanthophylls, a change that would not have been apparent from standard nutritional analyses but was detected from high-pressure liquid chromatography (HPLC) analyses for carotenoids.

As these two examples illustrate, targeting a single nutrient can lead to unintended alterations in the levels of other constituents, and specialized analytical methodologies may be required to assess changes in overall nutrient profile.

Another consequence of the introduction of significant nutritional changes in a food may be the requirement for post-market monitoring to determine whether the overall diet has been altered and to what degree, before an accurate assessment of the impact on nutritional status can be made.

Safety Considerations

The goal of the risk assessment process for genetically engineered foods is to examine the intentional and unintentional consequences of the specific modification on food components, including toxicants, in comparison with a counterpart food that has a history of safe use. Within this general framework, “case-by-case” variations must be considered.

International discussion between OECD countries, and within the United Nations FAO/WHO expert consultations, have resulted in a consensus on the specific safety issues that should be considered when evaluating a novel food (OECD 2000). They include:

1. a description of the host organism that has been modified, including information on nutrient composition, known antinutrients, toxicants and allergenic potential, and any significant changes in these that may result from normal processing;


2. a description of the donor organism, including any known associated toxicities and allergenicities, and the introduced gene(s);


3. molecular characterization of the genetic modification, including a description of the modification process and the stability of the introduced trait;


4. identification of the primary and secondary gene products, including a description of the characteristics of the inserted gene;


5. evaluation of the safety of expected novel substances in the food, including an evaluation of any toxins produced directly by the modification;


6. assessment of the novel food’s potential allergenicity; and


7. evaluation of unintended effects on food composition, including (a) assessment of changes in the concentration of nutrients or naturally occurring toxicants, (b) identification of antinutrient compounds that are significantly altered in novel foods, and (c) evaluation of the safety of compounds that show a significantly altered concentration.

In evaluating these safety issues, due consideration should be given to the processed version of the food, if the food normally undergoes manufacturing or processing, and food consumption issues, including: (a) identification of the potential human population consuming the genetically engineered foods and the amount they are expected to consume, and (b) assessment of any effects that may occur if intake of the modified food differs from intake of its conventional counterpart.

References

1. FAO/WHO (1996): Biotechnology and food safety, Report of a joint FAO/WHO consultation. FAO Food and Nutrition Paper 61, Food and Agriculture Organization of the United Nations, Rome.


2. FAO/WHO (2000): Safety aspects of genetically modified foods of plant origin, FAO/WHO consultation 29 May – 2 June 2000. World Health Organization, Geneva, Switzerland.


3. OECD (1993): Safety evaluation of foods derived by modern biotechnology, Concepts and Principles. Organization for Economic Cooperation and Development, Paris.


4. OECD (2000): Report of the task force for the safety of novel foods and feeds. C(2000)86/ADD1 Organization for Economic Cooperation and Development, Paris.