By Lynn L. Bergeson, Charles M. Auer, Richard E. Engler, Ph.D., and Oscar Hernandez, Ph.D.
The U.S. Environmental Protection Agency’s (EPA) release of its Toxic Substances Control Act (TSCA) Section 5(a)(3)(C) determination for P-16-0510 represents a significant step in EPA’s implementation of the New Chemicals Program under new TSCA. The substance is a polymer (a copolymer of ethylene glycol and propylene glycol end-capped with acrylamide groups). It is intended to be used as a deodorizer in a variety of products, including floor cleaners, cat litter, fabric freshener sprays, and other consumer products.
Notably, EPA’s determination document specifies the conditions of use that are intended, known, and reasonably foreseen. EPA states that there are no known or reasonably foreseen conditions of use other than those intended by the submitter. This may appear to be a controversial statement. Based on EPA’s interpretation of “conditions of use,” it would not pass legal muster to speculate that “anybody could manufacture or use it for anything” and, hence, impose use restrictions to prevent purely speculative applications with no basis in fact or reality. EPA has repeatedly stated that it would base what is reasonably foreseen on information, knowledge, or experience, not on any conceivable condition of use.
EPA identifies the new chemical’s potential health hazard endpoints based on the acrylate/acrylamide category. The concerns are based on acrylamide itself and some low molecular acrylamide analogs and include mutagenicity, developmental toxicity, reproductive effects, neurotoxicity, and a “marginal potential” for oncogenicity. This too may sound alarming. The real question, however, is how toxic is the new chemical and are exposures expected to exceed a “safe” level.
In this case, EPA specifically considers the low-molecular weight (LMW) components of the polymer (i.e., the “worst case”) in its assessment and identifies two analogs of the LMW components. Both analogs have similar structural features (they are end-capped with acrylates), so both are expected to share the same mode of action, and have similar molecular weights as the LMW components of the premanufacture notification (PMN) substance. EPA states that it also considered the toxicity of acrylamide in its assessment. We note, however, that acrylamide is not a good analog because it is substantially lower molecular weight than the LMW components of the PMN substance. Based on the identified analogs, EPA set a no observable adverse effect level (NOAEL) of 250 mg/kg/day for systemic toxicity based on a combined repeat dose/reproductive/developmental toxicity screening test (OECD 422). This study tests for a chemical’s potential to cause toxicity and the primary endpoints of concern relevant to the category (developmental, reproductive, and neurotoxicity). This NOAEL would put the substance in the low-to-moderate toxicity category. Note that despite the nominally alarming set of health endpoints identified in EPA’s category assessment, the 422 study shows the analog is not especially toxic to mammals. By way of contrast, EPA’s most recent Integrated Risk Information System (IRIS) assessment of acrylamide identified a NOAEL of 0.5 mg/kg-day.
EPA also identifies ecotoxicity concerns. Using its predictive models, EPA predicts toxicity levels for both acute and chronic effects to aquatic species and sets concentrations of concern (CoC) at 425 ppb for acute exposures and 43 ppb for chronic exposures. These levels put the substance in the “moderate” category for environmental hazard.
EPA then applies exposure modeling to predict exposures to workers, the general population, and consumers. EPA found that predicted exposures are sufficiently below EPA’s concern level to not present an unreasonable risk to workers, the general population, or consumers. EPA even found that at the “worst case” of 100 percent PMN substance, exposures would still be sufficiently below EPA’s concern level. EPA also evaluated surface water concentrations and found that the estimated maximum acute and chronic concentrations did not exceed the CoCs.
EPA reviewed the PMN, reviewed likely and potential exposures to workers, the general population, consumers, and aquatic species, and did not identify any foreseeable conditions of use that would lead EPA to predict that unreasonable risk was likely.
This is a marked and welcomed departure from previous TSCA Section 5(a)(3)(C) decisions. In nearly all cases in the past, EPA only made a not likely determination if it identified a low hazard for both health and ecological effects (“low/low” cases). Absent a low/low finding, EPA seemingly believed that there could be some conditions of use that could contribute exposures that could exceed EPA’s concern levels. Based on our review, EPA did not explain how it differentiated between “any possible/foreseeable” and “reasonably foreseeable” conditions of use. Instead, if EPA could imagine a set of circumstances that could elicit an exceedance, EPA was of the view that new TSCA precluded it from making a Section 5(a)(3)(C) finding. (While some stakeholders might applaud an approach based on concepts such as the European Union’s Precautionary Principle, this is not how new TSCA, or U.S. environmental legislation more generally, is structured.)
In the case of P-16-0510, EPA more carefully applied new TSCA as written when it identified a low/moderate health concern and a moderate eco concern, and nevertheless took a reasonable approach grounded on the law to go beyond mere consideration of potential hazard and to interpret the “reasonably foreseen” conditions of use and assess unreasonable risk as new TSCA requires. We support this more measured approach and believe it better meets the statutory intent and requirements. As we have written previously, in our view, a “not likely” finding is not limited to cases in which toxicity is so low that exceedances are unimaginable. Rather, EPA must limit its consideration to those conditions of use that are reasonably likely to occur and must evaluate unreasonable risks, not merely hazard, and regulate to protect to the “extent necessary” to protect against such unreasonable risks.
We applaud EPA’s more measured approach that likely indicates a maturation of its understanding of what is needed to meet a not likely determination. We urge EPA to articulate its thinking on what is and is not “reasonably foreseeable” and what PMN submitters can do to help EPA understand not only what is intended, but what might be reasonably foreseen to occur.