Risk assessment and risk management of transgenic pharmaplants

Pharmaplants are of special concern when possible environmental and biodiversity aspects are considered. As acknowledged in the Norway –Canada Workshop 2007 pharmaplants may pose unique risks to the environment eg because of the novelty of the pharma compound itself or the novelty of the exposure to the environment. Multi-ple interactions with herbivores, their predators, pollinators and any other organism being part of the food web and the interacting biocenosis network are to be taken into consideration. Distribution and spread via seeds, pollen or with the help of vegetative proliferation has to be prevented. Current risk assessment using concepts like substantial equivalence or familiarity as starting points are not suited for an environ-mental assessment of pharmaplants. There is the need of developing physical containment concepts which really live up to the expectations of a fully reliable containment. Out of a nature protection perspective this is somehow a prerequisite for the use of pharmaplants as production facilities for drugs or other potent substances.

I agree with Beatrix Tappeser’s comments on pharmaplants. However, in the rush to put such products to regulators there may be the temptation to argue that without an appropriate conventional comparator upon which to apply “the history of safe use”, we should apply comparators that are even less suited to the purpose of hazard identification. This is, in my opinion, what happened in at least some regulatory approvals of high lysine corn LY038. This product was acknowledged by all to be modified to be not substantially equivalent to any kind of corn or vegetable in existence. In the case of that corn, the various hazards were dismissed by what I consider to be some regulators’ far-fetched comparisons to red meat, button mushrooms and chicken eggs.

The problem with this is two-fold. First, even if one were to accept that hazards from the modification of a plant can be identified and dismissed based on comparisons to similar compounds in other kinds of foods, there remains the problem that the plant-derived hazard may be prepared differently and consumed in different amounts. Second, this approach equates the overall safety and value of a food with each of its parts. Some hazards in milk and meat are balanced by the overall nutritional benefits of these foods in the diets of many people. That does not make each component in these products safe at comparable concentrations in other plants nor does it justify moving a potential hazard into a food product that has historically been free of this potential hazard.

The use of historical literature to define acceptable ranges of compounds in a species must be abandoned in favour of descriptions that emphasise the overall physiological balance in the test and comparator at the same time and location. It is dubious to argue, for example, that because some corn plants grown somewhere in the 1920s may have had sodium levels 5 times that of the proper comparator grown in 2008 then significantly higher sodium levels in the GM plant is also justified. This is because we don’t know anything about how the 1920s plant may have compensated for this sodium or even if it resembled a proper corn plant.

It could be argued that pharma is a more blatantly hazardous modification and therefore this temptation, that has seduced regulators in the past, will not be so powerful in the future. However, this argument is speculative as developers cleverly undermine regulators’ worries about just how dangerous a pharma-protein is (e.g., because of claims that it will be digested etc.). And as developers make drugs based on multiple components raised in different plants and which they will argue only have a pharma property when combined through industrial process highly unlikely to exist in nature. None of these arguments are of scientific origin as much as they will be designed to reassure.

Thus I would argue that more attention needs to be paid to the scientific description of comparators with every effort placed on making this definition more stringent and meaningful than it has been in the past. Doing this for pharma can have retrospective benefits for first generation GM products.