On June 15 the EU Commission (EC) issued its highly anticipated “scientific criteria” for identifying Endocrine Disrupting Chemicals (EDCs). Now that the dust has settled, and stakeholders around the globe have had a chance to offer their thoughts, the time is ripe to explore to the heart of the criteria – what the EC selected (and what it didn’t select), and the potential impact their choices will have on consumers, industry and the global regulatory arena. 

The criteria are outlined in the following separate documents:

• Communication From the Commission to the European Parliament and the Council on endocrine disruptors 

• Draft Commission legal acts setting out scientific criteria for their determination in the context of the EU legislation on plant protection products and biocidal products.

• An Impact Assessment so as to estimate and understand the potential impact of the proposed four options outlined in the EC Roadmap on endocrine disruptors.

First look – a deft political compromise

Any way you slice the proposed EC “scientific criteria”, they strike those of us who have been steeped in the issue for decades and who have followed the EU debate closely as hugely underwhelming and represent at best a mixed bag for stakeholders.

It’s not hard to interpret the EC “scientific criteria” to be a deft political compromise between the insistence by very vocal health, environmental and consumer focused NGOs that “potency” be ignored when identifying EDCs, and industry’s plea to the Commission to adhere to scientific and risk-based principles and avoid using multiple categories of EDCs (e.g., “suspected”, “endocrine active”) that would only confuse EU consumers and global markets.  

After much deliberation, the EC opted for the second of the four potential options they had outlined earlier in their Roadmap on endocrine disruptors— simply using the WHO/IPCS definition of an EDC, something the EC admits EU regulatory agencies have already been applying in the absence of any existing direction on “scientific criteria”.  One could legitimately ask “Where is the beef?” as the “criteria” lack any meaningful substance beyond recommending adoption of the approach the European Food Safety Authority (EFSA) has already been taking to dealing with endocrine disruptors.  Although one could highlight a number of potential grievances associated with the proposed EC “scientific criteria” this blog post will focus on three key ones:

(1) The potential for the EU and the US to disagree on which chemicals to label, list and regulate as EDCs leading to potential trade issues and confusion in the marketplace and among consumers. 

(2) The EC’s decision to reject the use of categories of EDC’s such as “suspected endocrine disruptors” and “endocrine active” and its rejection of consideration of potency in identifying which EDC chemicals are unlikely to pose a risk to the health of humans and wildlife.

(3) Some stakeholders are criticizing how the EC and its regulatory agencies apply the proposed “scientific criteria”— in particular, how they conduct a weight of the evidence evaluation to assess the likelihood that a chemical agent’s interaction or interference with one or more components of the endocrine system actually causes an adverse effect.

Potential for Disagreement Between the EU and US in Identifying EDC’s

Let’s first consider the potential for disagreement between the EU and the US in labeling specific chemicals as EDC’s.  It can easily happen owing to a lack of harmonized approaches between the respective government agencies, and also due to a lack of coordination and communication.  The EC’s proposed “scientific criteria” are based on the WHO/IPCS definition of an EDC which requires that a substance act through an endocrine mode of action to cause an adverse effect.  To its credit, the EC has proposed conducting a robust systematic review of all relevant scientific evidence, using a weight of an evidence-based approach to make a determination whether a substance meets the WHO/IPCS definition.  However, the EC proposal does not mandate collection of any new data for making its weight of evidence determination on individual substances.

By contrast, the US EPA’s EDSP mandates new data collection. Manufacturers of chemicals selected by EPA for screening in the EDSP must submit their chemicals to a battery of 11 validated Tier 1 assays designed to elicit whether they have the potential to interact with estrogen, androgen, or thyroid pathways in mammals or amphibians.  EPA then conducts a weight of the evidence evaluation of the results of those assays as well as considering other available data, and determines whether additional Tier 2 testing to determine adverse effects is warranted.  In making their determination EPA also evaluates dose-response relationships to consider whether additional testing is likely to yield meaningful data that could change the prevailing EPA risk assessment for that chemical. If they judge it unlikely, then no further testing is required.  

Another key difference between approaches is that the US EPA makes no determination whether the endocrine activity detected in Tier I testing is causally-related to any adverse effects observed in Tier 2 bioassays. Thus far, Tier 1 screening and EPA weight of evidence determinations have been completed for 52 chemicals, which are mostly active ingredients in pesticides. 

EPA has recently announced a pivot in the EDSP program to substitute high throughput assays for three of the current Tier 1 assays to greatly accelerate the pace of the program and reduce the numbers of laboratory animals used.

Oddly enough, the EC ‘scientific criteria” do not reference the US EDSP or any other regulatory approaches being taken around the world (e.g., Japan, Canada, etc), so one is left to wonder how it will weight the data being generated by the EDSP program.  The EC did “field test” its “scientific criteria” by conducting an Impact Assessment.  For determining whether an active substance in a Plant Protection Product (PPP) or Biocidal Product (BP) would be identified as EDC under each of the four options outlined in the Roadmap, a screening study was performed by an external contractor. The screening was based on available evidence, no additional testing and was completed under a significant time constraint. The EC cautions that results of the screening “…therefore do not constitute evaluations of individual substances to be carried out under the respective chemical legislation and in no way prejudge future decisions on active substances to be taken pursuant to these two Regulations.” It would thus be erroneous to consider that the substances identified in the screening are considered as endocrine disruptors within the meaning of the EU legislation.  Despite those caveats the results of the screening provide us with some insight as to how the EU might apply their “scientific criteria” to identify EDC’s.  To estimate the potential for disagreement between the EU and the US, I compared the results of the EC contractor screening with the US EPA’s weight of evidence assessment of the List 1 chemicals.  The results are summarized in the following table.  

The EC contractor screened 324 PPP and 98 BP actives and identified 26 and 5 of them, respectively as potential EDCs using the proposed “scientific criteria” (note that two of the five BP actives identified as potential EDCs were also considered PPP actives).   29 of the actives screened by the EC contractor were also among the EPA’s List 1 chemicals for which Tier 1 test data and EPA weight of evidence determinations were available.  For 11 of the 29 chemicals both the EC contractor and EPA each independently came to the conclusion that they were unlikely to be EDCs.  They also concurred in five instances that the chemicals had the potential to be EDCs; however, for four of those five chemicals the US EPA concluded that no Tier 2 testing was required because it would not change the prevailing risk assessment for them. In two instances (2,4-D and Malathion), the EC contractor concluded that the chemicals were potential EDCs whereas EPA found no evidence of potential interaction with the endocrine system and recommended no Tier 2 testing.  For a further eleven chemicals, the EC contractor concluded they were not likely to be EDC’s whereas US EPA concluded that there was potential for interaction with the endocrine system; however, EPA argued no Tier 2 testing was required for five of them again because the results would not change EPA’s prevailing risk assessments.

So, what are the key takeaways from this analysis?

1. Unless the EU and US take concerted action to share data and coordinate and communicate their draft EDCs assessments in an effort to harmonize them, there exists a real possibility that they will come to different determinations on specific chemicals, leading to trade disputes and marketplace and consumer confusion.
2. The differences in EDC assessments go both ways, i.e., it is not a case where one government consistently over- or under-identifies EDCs compared to the other.
3. By considering potency, the US EPA will allow use of some PPP and BP actives that the EU might otherwise ban from commerce.   Use of these PPP and BP actives is entirely justified scientifically, however, since the EPA has shown that there are no unacceptable risks to humans or wildlife.

Since hazardous properties are inherent to a substance, and do not vary by jurisdiction, common sense would seem to dictate the need for greater cooperation and coordination between the EU and the US on the identification and characterization of EDCs.  As a first step, I would urge the EC to explicitly acknowledge the US EPA’s EDSP program in its “scientific criteria” and pledge to seek international cooperation and coordination in identifying and characterizing EDCs.

EC Rejection of EDC Categories and Potency

The EC rejected the NGOs’ and the ES’ stated preference for Option #3 from their Roadmap which would have placed chemicals into several categories of potential EDCs (“known”, “suspected”, and “endocrine active”) based on some vague strength of evidence arguments.  The EC rightly concluded that such categories were not helpful to the process of identifying what an EDC is and that the use of such categories would decrease “legal certainty” for regulators and other stakeholders with no “established benefits” for protecting health or the environment.  In making their case for multiple categories, the NGO’s have cited the existence of analogous categories for carcinogens and reproductive toxins in the EU; however, endocrine disruption is a mode of action and is therefore more akin to mutagenesis.  The EU doesn’t utilize multiple categories of mutagens, and so by analogy can easily justify not using multiple categories of EDCs.

But the Commission also rejected option #4 from its Roadmap which was the one preferred by industry, farmers and several member states and would have further narrowed the list of chemicals identified as EDCs by discounting those which had only weak potency — weaker than many naturally occurring EDCs found in commonly consumed foods and beverages. In doing so, they claimed that use of potency would have run counter to scientific consensus.  They do not cite any specific source to support their claim of scientific consensus, but the implication is that it came from the recently released BfR consensus statement, which itself has come under criticism by several scientists who did not participate in the consensus meeting (see Borgert http://www.euractiv.com/section/science-policymaking/opinion/disrupted-the-endocrine-disruptor-debate/; Nance http://www.science20.com/risky_business/endocrine_disruption_a_new_strategy-172707 and Bond http://www.science20.com/endocrine_policy_perspectives/blog/distinguished_european_scientists_challenge_endocrine_pseudoscience-173159.).  

The only way the BfR could coax consensus from the scientists it had convened was to dupe them into thinking that identification of EDCs was merely the hazard identification step in risk assessment.  The consensus statement argues that potency is critical to hazard characterization, but not to hazard identification.  But it has been pointed out to me (personal communication with Dr. Christine Palermo) that this logic is flat out wrong because the WHO/IPCS definition of an EDC includes mode of action which itself is not usually included in hazard identification, but only in hazard characterization and/or later steps in a risk assessment.  Furthermore, as Borgert pointed out in his critique, the BfR consensus statement would allow chemicals to be labelled endocrine hazards in the EU, despite lacking sufficient potency to affect anyone’s endocrine system. By such standards, caffeine or cocoa would be endocrine hazards.  This hardly seems helpful to regulators nor would consumers find it credible if it was explained to them in this manner. 

How the EC Proposes to Apply its Criteria

The EC rightfully points out that the most difficult determination to make in identifying EDCs is assessing whether the interaction or interference with the components of the endocrine system actually is the cause of any adverse effects that are observed.  They again rely on precedent set by EFSA “… a reasonable evidence base for a biologically plausible causal relationship between the [endocrine mode of action] and the adverse effects seen in intact organism studies.”  The EC rightfully rejected a much higher standard of conclusive evidence of causality which would require observing direct evidence of harm already in humans and wildlife.  Although it would have been useful for the EC to reference existing systematic frameworks for assessing causality such as the nine viewpoints offered by Sir Bradford Hill, the language in the draft legal acts does specifically outline several key decision criteria which explicitly include four of Hill’s nine viewpoints: biological plausibility, consistency of findings across studies, dose-response and coherence of the evidence.  It would have been advantageous if they had specifically mentioned the five others, including: strength of the association, specificity between agent and adverse effects, temporality (exposure must precede the adverse effect in time), effects of intervention, and analogy.

Reading the various press releases issued by the NGO community following the release of the EC’s scientific criteria they are claiming that the EC has set the barrier on determining causation too high, but this is pure hyperbole.  In my opinion, the EC has struck an appropriate balance between not requiring conclusive proof, yet ensuring that there is reasonable scientific evidence to support classifying a substance as an EDC.  Of course, final judgment must await evidence of how the various EU regulatory agencies actually apply the criteria and whether they do so with any amount of consistency. In the meantime, the EC should resist pressures from the NGOs to lower its scientific standards.

Conclusion

In conclusion, the EC’s proposed “scientific criteria” are at worst underwhelming and at best represent a mixed bag for stakeholders.  To strengthen them, the EC should explicitly acknowledge the role that the US EPA’s EDSP is playing to assist with identification of EDCs and should commit to efforts at international cooperation and coordination to avoid disagreements on labeling specific chemicals as EDCs that will only lead to trade problems and confusion in the marketplace and among consumers.  The EC should also seriously revisit its decision to not incorporate potency in identifying EDCs or it risks losing many useful products and undermining consumer confidence.  Finally, the EC should resist efforts by the NGOs to weaken its scientific standards, and maintain and strengthen its criteria for assessing whether endocrine active substances actually cause adverse effects through endocrine mediated pathways.