EFSA Boss: EU Food Law Overhaul ‘A Big Step Towards Transparency’

The European Commission has learned its lesson from the glyphosate controversy and is coming forward with “bold proposals” on transparency, says Bernhard Url. The raw data of all industry-funded studies related to pesticides will be made publicly available under Commission plans, he told EURACTIV in an interview Wednesday.

Source: EURACTIV

Bernhard Url is the executive director of the European Food Safety Authority (EFSA), based in Parma. He spoke to EURACTIV’s energy and environment editor Frédéric Simon.

INTERVIEW HIGHLIGHTS:

• Raw data of all industry studies will be pro-actively published by EFSA in a public register
• Even aborted industry studies will be available for public scrutiny
• It will be challenging for industry to substantiate confidentiality claims in the future
• Two landmark studies on the cumulative effects of pesticides will be published in December
• Public debate needed on what role agrochemicals should play in agriculture

The European Commission recently proposed changes to its general food law, with the aim of improving the transparency of scientific studies submitted to EFSA on pesticide approvals. Do you believe this proposal will put the issue to rest? Or are there still areas where you believe transparency could be improved?

The European Commission proposal is a big step towards transparency. They really listened to the European Citizen’s Initiative (ECI) to ban glyphosate. And they also listened to the review exercise (REFIT) on the general food law, which concluded that, although food safety has significantly improved over the last fifteen years, more could still be done in terms of risk communication, clarity and transparency for consumers.

All this encouraged the Commission to come up with a quite bold proposal, I would say, to improve transparency and accessibility to data that EFSA uses for risk assessments. Th  ese proposals also cover the governance of EFSA, so I think overall it is a big step forward, and we are very happy with it.

The European Citizen Initiative (ECI), which was the basis for the Commission proposal, demanded that “all studies used to back up regulatory approval of pesticides must be published” – meaning no more use of unpublished industry studies or only partially published studies, such as executive summaries, as was the case with glyphosate. Do you agree with this principle?

If you look at the bigger picture, this level of transparency is currently not a requirement in many areas like cosmetics, chemicals, food and feed additives. Now, with this proposal, it will become the rule for all applications related to food safety that EFSA has to deal with. And I think this is the right step forward.

However, under the proposal, personal data and confidential business information will continue to be protected. So I think this shows that there is no black or white answer in terms of the balance between transparency on the one hand and the protection of personal data and intellectual property rights on the other hand.

The European Commission made it absolutely clear that, under its proposal, EFSA will publish everything that is not protected. And I find this very useful. Of course, it’s also a challenge for us in terms of IT infrastructure. It won’t be like putting 10,000 pages of pdf files on the internet – we’re talking about structured data that should be searchable. This means industry will also have to submit data in a structured way, which will involve a lot of talks with them to make it happen.

Another aspect, which I find is missing in the current discussion, is that academic research will have access to this data. And maybe academic research will find patterns in the data that regulators didn’t see.

So there is an additional positive effect here. Maybe in the meta-analysis, there will be revelations. And that’s good. If we learn something, this is the principle of science. We call it ‘organised scepticism’.

The German government admitted that its BfR agency, which received the application for the renewal of glyphosate from Monsanto, wasn’t always clear in its report which parts came from its own assessment and which parts were merely quoting from Monsanto-funded studies. Will EFSA now be able to scrutinise also what happens at the national level?

To us, it is completely clear that the BfR properly did its job of checking and scrutinising industry information submitted to them. And only when it was accurate did they extract the information to put in their report. If you look carefully into the BfR report, you will see clearly that they have done this work because there are bookmarks that BfR inserted into the text.

But still, the German Ministry said this could maybe made even clearer in the future, for example by putting the text from the applicant dossier between quotation marks or in italics. So I think the ministry is right, the information can be made clearer.

However, there is no doubt that the information extracted from the applicant dossier was examined and found to be correct. And in this case, we are not talking about information from studies commissioned by Monsanto or other members of the Glyphosate Task Force. It was publicly available scientific literature that had to be summarised by the applicant, which is a legal provision. And what BfR did was to check whether the summary was accurate. And when it was, they took it in their report.

This ‘accuracy checking’ has to be done by the Rapporteur Member State, and is a step before risk assessment. It is just preparing the basis for later risk assessment. So from my point of view, this was completely clear, and completely ok. Still, it could have been labelled more clearly which parts were the BfR’s own analysis and which parts came from the information provided by the applicants.

The concept of plagiarism however makes no sense in a regulated procedure that starts with an applicant’s dossier. Obviously the Rapporteur Member State uses the literature summaries provided by the applicant, after checking their accuracy. And this is important to stress. Ironically, the section of the report that some claimed is evidence of ‘copy and paste’ from industry, is in fact the very same section that led to a key recommendation by EFSA: that the genotoxic potential of glyphosate formulations ‘should be further considered and addressed’.

Is ‘accuracy checking’ by national authorities something EFSA will be able to scrutinise more thoroughly in the future after the glyphosate controversy?

The level of scrutiny was always very high. It’s a three-step procedure. The first step is done by the applicant – they have to summarise the published scientific literature of the last ten years. This is a legal provision. And we’re not talking here about industry studies, which is a different story. Applicants have to do a search of publicly available scientific literature looking back at 10 years, and summarise all the articles collected. There is an EFSA guidance document providing instructions on how this should be done.

Then the rapporteur member state – the BfR in this case – looks at whether the summary has been done in an accurate way. We’re still before risk assessment here. Then the BfR does a risk assessment and forwards it to EFSA. And then EFSA organises a peer-review process, with all member states – an assessment and improvement of the work done by the Rapporteur Member State.

So altogether this is a very thorough process. There is also a public consultation in between so the public also has a chance to scrutinise and add information.

What is new now with the Commission proposal is that the raw data of industry studies will be pro-actively published by EFSA.

…unless protected by business confidentiality.

The raw data of the studies will not be protected. That is already clear for me.

The proposal also foresees the possibility for the Commission to ask EFSA to request additional studies financed by the EU budget. Can you cite examples of circumstances where this new right of initiative would be most useful?

I think there are two different aspects. The first is that the Commission can task EFSA to commission additional studies “in exceptional circumstances” – that is the wording of the proposal – to verify the data that was submitted by applicants.

And I think it is very important to be clear on that: it is not about repeating studies, or doing additional studies. In exceptional cases, when there is uncertainty or ambiguity, the risk managers can task EFSA to verify certain aspects of the information provided by the applicants.

So they wouldn’t be new studies?

It could end up in a new study, yes. It could even be a bit broader, to clarify additional aspects of science. But the right of initiative is in the hands of the Commission here because this is European taxpayers’ money so it should be used wisely. And also because the responsibility still lies on the applicant to provide data showing their products are safe. It’s not supposed to come from the public budget. This is a fundamental principle introduced years ago by the EU lawmakers.

When we spoke last year, you suggested that industry could “pay into a pool” that regulators could draw from to commission independent research. Is there now a consensus building up in favour of that?

For the moment, the Commission has decided not to go down this route and focused on making sure that the data provided by the applicants is as robust and reliable as possible. So not repeating things but making sure we get things right in the first place. And I think this is the right approach.

This is why the Commission decided to put in place a register of studies. So that none of the industry studies can be put in a drawer if they end up being unfavourable. And also strengthen the audit system for laboratories that do the studies – the contract research organisations.

So even the industry studies that were aborted mid-way through would be caught by this register?

Yes, because they have to register before the study is done. And the information has to come from two sources – the industry and the labs.

To your question about industry paying into a pool – I think in theory it sounds good. But practically, the research and innovation process for a new molecule can take up to ten years and is deeply embedded into a company’s business strategy, meaning it would be very difficult to outsource this to a public organisation.

So even though it sounds politically attractive, I think this idea of a pool would be very very difficult to put in place.

Regarding new studies, one of the most complex issues is how to deal with the so-called “cocktail effect” when substances are considered in combination. Is this an area where regulators could mandate additional research?

The cocktail effects still need a lot of research, it’s not so much about individual regulatory studies.  It’s also a big challenge for methodology development in risk assessment, which EFSA is tackling. We work closely with the Dutch Institute for Public Health and the Environment (RIVM) to develop such a methodology to examine the cumulative effects of pesticides.

And there will be two landmark publications from EFSA in December 2018, on the cumulative effects of pesticides. We will look at groups of 20-50 pesticides and for the first time, there will be a cumulative assessment of their combined effects on the thyroid gland and the nervous system. This is pretty complex scientifically and we have been working on that for many years. Critics say we are not fast enough. But it’s always a question of input, money and people. Still, we are moving forward and we are here at the very edge of regulatory research.

So I don’t think the cocktail effect is a field for the additional studies the Commission is referring to in its proposal. I think the additional studies really come in when there are controversial issues with scientific divergence, and the risk managers need clarification to make a decision.

So the typical example here would be glyphosate, where the opinions of the WHO’s International Agency for Research on Cancer (IARC) and EFSA diverged. Right?

Actually, that wouldn’t be the best typical example. The IARC looked at different evidence than EFSA. Due to legal provisions, EFSA – like ECHA other regulatory agencies in the world – had the possibility to look at more evidence. We had access to all the studies in the field whereas IARC has a policy to look only at published data.

Now, in the new system proposed by the Commission, the public will also have access to all data, including unpublished industry studies. So, regarding IARC, we did not start from the same basis.

The two will come closer together in the future, however. Because under the Commission proposal, risk assessments will have to be based mostly on published studies. So these kinds of discrepancies should in theory disappear in the future.

Yes, it’s true. IARC and others will then be in a position to access all the studies that EFSA has used. In the past, they could only use the summaries of industry studies that were published by bodies like the US Environmental Protection Agency (EPA), or the WHO joint meeting on pesticide residues (JMPR). IARC only had access to the published summaries of these studies. But in the future, they will be able to use the raw data.

So you had access to information they didn’t have?

Yes. And we had to use that because it’s a legal provision. Which is good because we want to include all the available information in our ‘weight of evidence’ approach.

Coming back to the Commission proposal, companies can still request parts of the submitted studies to be kept confidential if they believe this could hurt their commercial interests. EFSA is supposed to assess whether these claims are justified or not. So under which conditions do you believe withholding such information is justified? Where will you draw the line?

The Commission proposal says there will be a positive list describing which generic parts of an application can be considered confidential. This will cover for example information on market share or economic relations with other companies, or impurities in the productions processes.  And this positive list will apply across all legislation – on GMOs, pesticides, feed additives, etc.

For the rest, outside this list of protected items, it will be challenging for industry to substantiate confidentiality claims. They will have to substantiate this with evidence and EFSA will have to assess this evidence before taking a decision.

And EFSA will now also be able to lift confidentiality for the protected categories you mentioned?

EFSA, the European Commission and the member states always have access to all the information. There is no confidentiality claim in what we can use for the risk assessment. But whether the protected parts will be published depends on whether there is a public health risk or an emergency. So even these protected items can be published if it’s needed to protect public health.

And this is where EFSA will have to assess whether it’s justified or not?

Yes.

The Commission proposal also foresees a pretty big rise in EFSA’s budget of €62.5 million every year. How do you plan to use that money?

If you look into the Commission proposal, the €62.5 million are the sum of many individual measures that were assessed in terms of costs. There are some 20 different lines, including checking confidentiality claims, setting up this register of studies, commissioning studies from time to time, working better together with member states organisation, etc. Altogether, this adds up to €62.5 million annually.

So how to spend that money does not depend on the will of EFSA, it’s already well calculated. And it is a lot about transparency and risk communication, which is another pillar of the Commission proposal. And that’s not so much about money, it’s more about the alignment of the different actors in the risk management process, including member states and the stakeholders in the agri-food system.

Talking about public involvement, with more studies publicly available, there is also a likelihood that scientific evidence will be verified and cross-checked by a variety of people –  researchers, academics and civil society organisations. But not everyone has the qualification to do a proper scientific assessment. So how can you make sure that these contributions are actually useful? How do you prevent an increase in transparency from adding to the chaos and confusion around science?

There are different target audiences for these transparency measures. The first is, as I said earlier, the scientific community who will be able to use the raw data and scrutinise our science. And this is something we are really looking forward to. Because we want to subject our science to so-called ‘reproducibility’. And for that, we have to give the data, the methods, so the public can redo the science, recalculate and come to their own conclusions.

That could lead to more discussions on a scientific level. But I think that will be good for the progress of science.

This is like science crowdsourcing in a way, which seems like something new.

True. There is a precedent in the medicines area, with the clinical trials database run by the Food and Drugs Administration (FDA) in the US and the European Medicines Agency (EMA), which did exactly the same. The aim was to make data available for research to help promote innovation but also to see risks and patterns that regulators didn’t see.

And this crowdsourcing of science is something I like very much. Will it lead to chaos? I don’t think so. But it could lead to more progress in science, which we would be very happy about.