Jun 4, 2007

News

Send Add comment

Discussion about MON863 maize

“Statistically significant differences occur in practically all feeding studies”

In March 2007 the insect-resistant Bt maize variety MON863, which has been authorised in the EU as food and feed since January 2006, became the focus of public debate again. The French scientist Gilles-Eric Séralini had published a new evaluation of the feeding study on rats submitted by Monsanto during the authorisation procedure and had once again cast doubt on the safety of MON863. GMO Safety spoke to Marianna Schauzu of the Federal Institute for Risk Assessment (BfR) about it.

The scientists working with Séralini evaluated the observed statistically significant differences between the animals fed with MON863 and the control animals as toxic effects and came to the conclusion that MON863 could not be classed as safe. The European Commission then asked the European Food Safety Authority (EFSA) for an opinion and EFSA consulted the EU member states.

In its statement, the Federal Institute for Risk Assessment (BfR), which is involved in the safety assessment of genetically modified food and feed, comes to the conclusion that the new statistical analysis does not call into question the results of the earlier assessments of Mon863 maize. The BfR continues to view the food and feed produced from this maize as being as safe as conventional maize products.

GMO Safety: In the feeding study, which was submitted with the application documentation for the authorisation of MON863, statistically significant deviations were found in the blood and urine of the trial animals fed on GM maize compared with those fed on conventional maize. Unlike Séralini, you classify the differences as not toxicologically relevant in your statement – claiming that they lie within the range of “natural fluctuation”. How is it possible to have such different assessments of the same data?

Marianna Schauzu: A 90-day feeding study is regarded as being sensitive enough to detect any harmful effects that might be caused by newly introduced substances or relevant changes to natural substances contained in the food or feed.

If a statistically significant difference is found between the trial and control animals, this does not necessarily mean that we are dealing with an important effect, as the German translation of the English word ‘significant’ might imply. Statistically significant differences within individual parameters occur in almost all feeding studies and must in all cases be tested for their toxicological relevance. In the case in question, this check showed no indications of a possible risk for consumers.

GMO Safety: When, in your view, should deviations be regarded as toxicologically relevant?

Marianna Schauzu: In the statistical evaluation of the feeding study carried out by Séralini et al. and that carried out by Monsanto, significant differences were found in individual test parameters. These were usually small. The values were within the range of the historical control data, which had been obtained during earlier studies with rats of the same strain. In addition, feeding with MON863 maize led to contradictory effects in male and female rats for certain laboratory parameters, which does not point to an effect caused by the test maize. In particular, however, the changes did not follow a consistent pattern. The differences were not dose-dependent and/or they were not supported by differences in other parameters that might indicate an effect in the same organ. Neither were any effects observed in the organ and tissue examinations under the microscope that would indicate any toxicological relevance of the statistically significant differences in the laboratory parameters.

GMO Safety: In your statement, you say that the applicant, Monsanto, included in its assessment other data from additional control groups fed on commercial maize varieties, as well as historical control data from earlier studies with rats of the same strain, and that a comparison with this data is necessary to check the toxicological relevance. What data is needed for an evaluation?

Marianna Schauzu: There should always be a 90-day feeding study on rodents conducted in accordance with the internationally accepted recommendations, e.g. the relevant OECD guideline. The study design and the steps to follow for the tests are described in detail there. If there are statistically significant differences between the trial and control groups, they should be tested for toxicological relevance. As well as the aspects already mentioned – the dose dependence of an effect and the necessary clarification of possible differences with other organ-specific parameters, it is important to see whether the values lie within the natural range of fluctuation. For this, the values are normally compared with the historical control data for rats of the same strain. For studies with products from genetically modified plants, data for additional test groups fed with the relevant conventional products can also provide further relevant information.

In addition, the 90-day feeding study on rats is just one element in the comprehensive investigation programme that is required to answer the question of whether genetically modified plant products are as safe as comparable conventional products. In this investigation programme, for instance, the modifications are analysed at DNA and protein level. New proteins are tested for their toxic and allergenic potential. During field trials over several years, the genetically modified plants are compared with the parent line and with conventional varieties, at different sites in order to assess whether there are any changes in their agronomic properties. Relevant plant substances, like proteins, fats, carbohydrates, minerals, vitamins and minerals and other metabolites that are characteristic of the plant in question, such as allergens and toxins, are subjected to a comparative analysis in order to establish whether the genetic modification has caused unintended changes. If this is the case, the health relevance of the differences must be evaluated and feeding studies may have to be carried out. In the case of MON863 maize, there were no indications of unintended changes, but a 90-day feeding study was carried out anyway.

GMO Safety: Standards for feeding studies are laid down in e.g. OECD guidelines. You have many years’ experience of assessing such studies. Is your experience and that of other authorities incorporated in the standards? Are these requirements adequate?

Marianna Schauzu: The Federal Institute for Risk Assessment (BfR) is one of the German authorities involved in the safety assessment of genetically modified food and feed within the authorisation procedure. The results of our assessment are incorporated in the opinion statement, which is conveyed by the Federal Office of Consumer Protection (BVL), the coordinating authority in Germany, to EFSA, which is responsible for the safety assessment. The BfR is also involved in drawing up and updating guidelines for the safety assessment of genetically modified organisms. The current guidelines are considered by the BfR to be adequate and appropriate for the safety assessment of genetically modified food and feed.