Safety evaluation of genetically modified foods.

The concept of substantial equivalence has been accepted as the cornerstone of the health hazard assessment of genetically modified (GM) foods (OECD 1993). Substantial equivalence is the most practical approach to address the safety of foods or food components derived from GM crops and is based on comparison of the phenotypic and compositional characteristics of the parent crop and the GM crop. Basically, three categories of GM crops can be considered (FAO/WHO 1996; EU 1997): (a) GM crops which have the same composition as the parent crop, (b) GM crops which have the same composition as the parent crop with the exception of a well-defined trait, and (c) GM crops which are different from the parent crop. For the safety assessment of the first category of GM foods only a molecular characterisation of the genetic insert is sufficient, whereas for the second category a safety assessment of the expressed protein(s) is also required. For the last category an extensive evaluation including bioavailability and wholesomeness studies are required, beside the molecular characterisation and safety assessment of the expressed protein(s) and their products. By molecular characterisation is meant the position, nature, stability and number of copies of the inserted DNA. Substantial equivalence is established by the determination of the phenotypic characteristics (e.g. resistance against diseases, agronomic properties) and the complete chemical composition of the plant including nutrients, toxicants, antinutrients, and allergens. The toxicity of the expressed protein(s) is assessed by their homology with known protein toxins, degradation in the gastro-intestinal tract, stability to food processing and acute toxicity in rodents. The possible allergenicity of the expressed proteins is evaluated by comparison of their amino acid sequence with that of known allergens and determination of their stability to digestion and food processing. If the source of the genetic insert is allergenic then the use of solid-state immunoassays, skin prick tests and even food challenge tests can be considered.