The risk assessment of GM plants and derived food/feed is based on comparison with non-GM counterparts. The GMO Panel considers that a comparative approach following international risk assessment standards using molecular, compositional and other analyses[1] remains an appropriate basis for determining whether animal feeding studies are needed.

The Panel recommends that any risk assessment of GM food/feed should first consider whether initial studies using in silico[2] and in vitro[3] approaches may answer some of the safety questions and indicate whether there is a need for subsequent in vivo[4] studies, such as animal feeding trials. This report discusses in depth the strengths and weaknesses of repeated dose animal feeding trials for the risk assessment of whole GM food or feed. Such feeding trials on the whole food or feed should be considered in cases of substantial differences between the GM plant and its conventional counterpart or where there are indications for the potential occurrence of unintended effects. In addition, the Panel considered when livestock feeding studies are needed for the safety and nutritional evaluation of feed derived from GM plants.

The Panel made a number of recommendations to help improve the scientific basis for assessing the safety and nutritional aspects of whole foods, such as further development of in silico and in vitro tests as well as recommendations for appropriate design of in vivo tests.