What you don't know can still hurt you - underreporting in EU pesticide regulation.

The safety evaluation of pesticides in the European Union (EU) relies to a large extent on toxicity studies commissioned and funded by the industry. The herbicide glyphosate and four of its salts are currently under evaluation for renewed market approval in the EU. The safety documentation submitted by the applicant companies does not include any animal study regarding developmental neurotoxicity (DNT) that is compliant with test guidelines. For a fifth salt, not included in the present application for re-approval, such a DNT study was sponsored by one of the applicant companies in 2001. That study shows an effect of that form of glyphosate on a neurobehavioural function, motor activity, in rat offspring at a dose previously not known to cause adverse effects. Counter to regulatory requirements, these effects were apparently not communicated to authorities in EU countries where that form of glyphosate was authorised at that time. That DNT study may also be relevant for the ongoing assessment of glyphosate but was not included in the present or previous applications for re-approval.In this commentary, we highlight that it is the responsibility of the industry to evaluate and ensure the safety of their products, taking all available scientific knowledge into account. We argue that the legal obligation for industry to submit all potentially relevant data to EU authorities is clear and far-reaching, but that these obligations were not fulfilled in this case. We claim that authorities cannot reliably pursue a high level of protection of human health, if potentially relevant evidence is withheld from them. We suggest that a retrospective cross-check of lists of studies performed by test laboratories against studies submitted to regulatory authorities should be performed, in order to investigate the completeness of data submitted to authorities. We further suggest that future toxicity studies should be commissioned by authorities rather than by companies, to improve the authorities' oversight over existing data and to prevent that economic conflicts of interest affect the reporting of study results and conclusions.