RA&RM of LMOs with Stacked Genes or Traits

I would like to propose an anlysis of what the approach of the document on stacked genes could be.
If you agree with this analysis, we could have a subsequent look at the proposed outline and see if and where it would need to be restructured.

In the report of the first session of the AHTEG, paragraph 29, a general structure of the guidance documents is agreed:
‘In the context of the steps contained in paragraph 8 of Annex III of the Protocol, the Group also agreed that the general structure of these guidance documents should be organized by providing: (i) points to consider; (ii) rationales for the points to consider; and (iii) relevant bibliographies and supporting documents.’

Taking this into account, I have the following considerations for the structure and approach of the document on stacked genes.

• The document should provide a definition of ‘stacked event’, and the scope of what we consider to be stacked events.
For instance: There is no fundamental difference between stacked events that are introduced ('united' may be a better term) into one recipient, e.g. by multiple transformation events or by crossing, compared to stacked events that arise from one transformation event with a construct with several different genes.
• Provide a general rationale for RA of stacked events:
The RA of the separate genes of the stacked event should be (or already have been) done in separate RAs. If risk assessments of the separate genes have already been done earlier, they are also valid for the situation in which these genes are stacked. The RA of the stacked event only has to take into account the interaction between the gene products of the separate events.
• The steps of the Roadmap are followed to derive rationales and points to consider for the RA of the stacked event.
Step 1 may betaken here in two sub-steps : An identification of any novel genotypic and phenotypic characteristics: this identification has been done already for the single genes in their seperate RA; specific point to consider here is whether any source of instability has to be taken into account, or whether the molecular characterization of the inserts remains unchanged.
Step 1, second sub-step: An identification whether the stacked genes taken together as a ‘novel genotype or phenotype’, may have adverse effects on biological diversity in the likely potential receiving environment, taking also into account risks to human health:
This step is not essentially different from what is done for a single event. It requires the establishment of a credible causal pathways whereby the combination of stacked genes and their combined gene products may give rise to adverse effects.
The main point to consider here is to establish the physiological consequences of the interaction of two or more new genes and gene products in the same recipient. If no interaction is to be expected, the RA finishes at this point, as any consequences of the single genes have already been taken into consideration.
Steps 2 – 5: These steps are the same as the steps for a single event: they follow from the adverse effect(s) identified in step 1, specifically the adverse effects based on the interaction of two or more gene products. The points to consider are in principle not different from the RA of adverse effects that arise from a single gene.
• In conclusion: except for the points mentioned for Step 1, I see no a priori reason why the points to consider for stacked genes should be different from the points to consider for single gene events. If we come to the conclusion the new points to consider would be necessary for the case of stacked genes, we would have to discuss whether they are or are not necessary for the Roadmap in general, and they should in that case be added to the general Roadmap, not in a document on stacked genes. (Of course, I may have overlooked some that are specific for stacked events, but they would become apparent in this exercise.)
• The available literature on stacked events should be screened whether they underpin this approach (that I think follows logically from the Roadmap), or follow a different approach. Consequently, the bibliography should contain an analysis and discussion of the available literature, and not only contain references to documents that underpin our point of view.

Taking into account what has been agreed by AHTEG I wqant to support the proposal by Hans Bergmans. I also would like to draw your attention to the chapter on stacked events of the EU report BEETLE on long-term effects of genetically modified crops which is available in the internet at the home page of DG Environment of the EU Commission.
http://ec.europa.eu/environment/biotechnology/reports_com_stud.htm  I have copied that chapter:

5.1.8 Stacked Events
Where GM events have been approved under Regulation (EC) No 1829/2003 or Directive 2001/18/EC, genotypes produced by crossing plants containing these events with non-GM plants are not required to undergo further risk assessment. However, where applications involve the crossing of plants to stack GM events, a risk assessment is required in the European Union. The stacking of approved events can arise from intentional crosses as well as from unintentional crosses. Stacked events have become more important during the last few years. The first cultivation started in 1997-1999 in the US with a stacked event of insect resistance (IR) and HT in cotton and maize. Today, the most common stacked events are a combination of (i) different IR genes or (ii) of an IR and a HT gene, obtained by the crossing of single trait paternal lines. An increasing number of stacked events are submitted for culti-vation in the EU. This raised the question if the safety of stacked events has to be assessed differently from single trait plants. The ERA should take into account the evaluation of the individual events and additional data from molecular characterisation and comparative com-positional analysis of the stacked events when determining potential interactions between genes or between gene products.
5.1.8.1 Assessment within the sections 5.1.1 - 5.1.7
To our knowledge, no detrimental negative interaction has been observed between stacked genes in GM crops so far. Although stacked events have been cultivated for about 10 years, very little research has been published and investigations addressing potential long-term effects are missing. However, first studies on the nutrient composition of double resistant maize and its impact on feeding of chickens did not show any significant differences.
In the OSE, the experts were asked for their view on stacked events and a potential differentiation with regard to intended or unintended stacks for each category mentioned (see Annex 2, 3.10, p. A2-13). For none of the processes the majority of the experts (56 - 68%; see Annex 2, A2-15 - A2-38) tended to make a difference in assessment between intended or unin-tended stacks. About 10% of the experts were of the opinion that the data basis is insufficient for the processes concerning interactions with NTO, effects on ecological functions and ef-fects on cultivation and management.
5.1.8.2 Conclusions
For the assessment of long-term effects resulting from stacked events, the initial ERA according to Annex II of Directive 2001/18/EC should already consider the occurrence of unintended stacks (Annex 3, 3.3, p.A3-10).