Rethinking GMO risk assessment - repensando a avaliação de riscos de OGMs

Dear readers

Suddenly a stream of fresh air is flowing in the GMO risk assessment field: the old paradigm of event-specific evaluation is being challenged. Moreover, respected researchers claim risk assessment must be directed to the product, not to the technology. And, finally, the same authors clarify that regulation must be risk-based.

The Brazilian GMO regulatory framework, established more than 10 years ago, requires a case-by-case evaluation for all GMOs and regulates the technology, not the product. The strategy worked well for a while, but had a very high cost and demanded a lot of time from CTNBio specialists to evaluate the large dossiers for commercial release. Every genetic transformation, even if with the same genes and the same host, was considered potentially different and was named an “event”. The idea behind this approach was that the random gene insertion made every transformation very different from one anothe, even if the constructs were essentially the same. However, after 20 years of genetic engineering, it is clear that events harboring the same or similar genes behave in a very similar way, irrespective on how many copies they have and where they are inserted in the genome. It is clear now that the genomic intrincacies are irrelevant for risk assessment: the phenotypic changes are to be valued. Under this new perspective, the event-based approach has no support.

CTNBio has already recognized the limits of the event-based approach and issued a normative by which a recurrent construction used in a new event could be assessed in a more streamlined, fast systematics. But a list of “safe” events was never produced, even if this is a provision of the Cartagena Protocol, also never fulfilled. It is time to open the windows and reduce the burden on safe products, using the principles of familiarity and history of safe use.

On the other hand, a change on the regulatory viewpoint from technology to products depends on a revision of the Biotechnology law. However, this is imperative and should be a priority as many new technologies are coming to the market and will make the scenario rather confusing. The Obama administration started a discussion on this subject one year ago and USA may have a new regulatory framework in a near future. The changes in our regulatory framework are certainly less extensive, but a meaningful discussion should be started as soon as possible. For the next Brazilian GMO regulatory framework some premises have to be kept very clear in the regulator´s and in the legislator´s minds:

a)       A GMO is an organism with a new phenotype, derived from any new technique encompassed under the large umbrella of the “new biotechnology” – it is cumbersome to split regulations among technologies, as the final product is what has to be regulated, not the way it was produced.

b)      The risk assessment procedure is product-directed and is now consolidated: in proceeds step by step, case by case, but takes into account the previous history of safe use and the familiarity of the genes, construct and host. This is to say there should be no difficulty in producing a reasonable list of safe organisms and a template to quickly review new products and put them on the list of safe GMOs or, otherwise, send the product to a more thorough assessment.

c)       Any change in the regulatory framework should be risk-driven. Restrictions should be proportionate to risks.

Here we add two new papers to the first one we recommended some days ago to risk assessors (Conko et al, 2016). All of them stress essentially the same points:

a)       regulation must be risk-based

b)      products should be regulated, not technologies

A risk-based approach to the regulation of genetically engineered organisms   pp493 - 503
Gregory Conko, Drew L Kershen, Henry Miller and Wayne A Parrott
doi:10.1038/nbt.3568

Ending event-based regulation of GMO crops   pp474 - 477 Steven H Strauss and Joanna K Sax doi:10.1038/nbt.3541

Regulate genome-edited products, not genome editing itself   pp477 - 479 Dana Carroll, Alison L Van Eenennaam, Jeremy F Taylor, Jon Seger and Daniel F Voytas doi:10.1038/nbt.3566

We wish all of you a good reading. Please contact me if you don’t have access to any of these papers (andrade@ufpe.br).

Paulo Andrade

Dept.Genetics/ Universidade Federal de Pernambuco