Flowchart ("Roadmap") for risk assessment: the necessary steps to conduct risk assessment according to Annex III of the Protocol

Dear All,


My name is Piet van der Meer. I have been working in the field of biosafety since 1986, the first 13 years for the Dutch Government and the last 9 years providing support to Governments and various international organizations, among which the Public Research and Regulation Initiative (PRRI).

As with the on line discussion on capacity building, I found the previous discussions on risk assessment for GM trees, GM fish, Pharmaplants and GM micro-organisms a fascinating exercise.

Picking up from the debate under those previous discussions, I post the following general thoughts for your consideration and feedback.

As the background document produced by Hans Bergmans reminds us, the general principles and methodology of risk assessment as outlined in Annex III of the CPB are scientifically sound and based on the experience of over 2 decades of risk assessment.

Those general principles and methodology have shown to be adequate for GM crop plants with altered agronomic traits, and they are equally valid and appropriate for risk assessment for relatively new cases, such as GM trees, pharmaplants, GM fish and GM micro-organisms. As with GM crop plants with altered agronomic traits, practical guidance how to apply this methodology in these specific areas will certainly be useful.

However, the fact that additional guidance may be useful should not be interpreted as that for those cases adequate risk assessments cannot be done. This is a misperception.  The methodology of Annex III can be applied, also where there are areas of uncertainty. As the Protocol explains in Annex III: “Where there is uncertainty regarding the level of risk, it may be addressed by requesting further information on the specific issues of concern or by implementing appropriate risk management strategies and/or monitoring the living modified organism in the receiving environment.

What the previous on line discussions on RA also showed is that in conducting risk assessment we do not look at the GMOs “in a vacuum”. As we can see from Annex III of the Protocol, a key characteristic of risk assessment is that it is comparative.  Risk assessment follows a number of steps and comparisons are made in the subsequent steps of the risk assessment.

Reading some of the postings of the previous on line discussions, I realized that we need to discuss further that the nature of these comparisons is different for each step, and by consequence, the “comparator” we choose can be different for each step. This is why for example the EFSA guidelines refer to the “appropriate comparator”.

The first step of the risk assessment as outlined in Annex III, is to identify whether there are any novel genotypic and phenotypic characteristics associated with the living modified organism that may have adverse effects.

The first question here is therefore which novel characteristics result from the genetic modification. Here we typically make a distinction between intended changes (e.g. the genes inserted) and possible unintended changes.

To identify whether there are any unintended changes, we compare some key characteristics with the characteristics non-modified host organism. For plants, often a near isogenic line is used for this first comparison. Once a “change” from the non-modified host is observed, we ask ourselves whether that change is within the normal variation of that species.  Differences from a near isogenic line do not mean there is a concern, since each species is normally quite variable.

The next step in the risk assessment is to identify any of the identified changes may have an adverse effect.

As was explained in one of the postings in the previous discussions, this step is tied to what we call the “problem formulation”, which in turn is related to protection goals. I hope we can discuss this further too.

(NB: It is important to remain aware that a “change” in itself is not ‘risk’. It only means “change”. We should remember that in conventional plant breeding, there are typically many unintended changes observed, many of which are not desired by the plant breeder and therefore thrown away, and some are actually of interest to plant breeders and kept for further development.)

In the next step of the risk assessment we evaluate the likelihood of the potential adverse effects being realized.

This is followed by an assessment of the consequences should these adverse effects be realized. In this step we make again a comparison, this time to see whether a certain effect on biodiversity would be significant. Significance here is to be seen in the context of the protection goals and natural fluctuations, which brings us back to the “problem formulation”.
Next follows An estimation of the overall risk posed by the living modified organism based on the evaluation of the likelihood and consequences of the identified adverse effects being realized.
And finally we move to the last step of the risk assessment in Annex III, i.e. we ask ourselves whether or not the identified risks – if any - are acceptable or manageable.

In evaluating whether identified risks are acceptable, we have again a comparison step. This time we make a comparison between an identified risk associated with the GMO with the risks posed by the non-modified recipients and their use.

As this sequence shows, in the different steps of the risk assessment we make different comparisons, taking into account different parameters.

As I often explain in RA training workshops, the methodology of risk assessment is in a way similar to a recipe for baking a cake. (I say “in a way” to avoid the suggestion that risk assessment is a piece of (baking a) cake). A recipe tells us to follow a number of steps and which ingredients to use at which steps. The same is the case for risk assessment, annex III tells us in paragraph 8 which steps to follow and in paragraph 9 which ingredients to use.

I therefore very much support the suggestion Hans Bergmans made in the back ground document that in assisting the development of the ‘roadmap’ for risk assessment MOP4 has asked for, we identify which points to consider mentioned in paragraph 9 of Annex III can be of relevance in which steps mentioned in paragraph 8 of Annex III.

This is more to follow up on the comments by Dr. Piet van der Meer, my historical colleague since the negotiation of the Protocol.

The Raodmap is like a blueprint without a building at the beginning when starting RA. But now we have  a lot of field release of LMOs-FFP.  Withe feedback of the experiences, there could be extension and elaboration on tow ways. 

One for elaborating on different categories of LMOs for further tune-up to facilitate efficiency and efficacy of RA.

Another could be future outlet of the experiences reflected to the simplification or more details tuning of the direction.

Kind regards,