National Research Council issues two reports reviewing EPA's chemical testing and science.

In the past week, the National Research Council (NRC) has released two reports on aspects of EPA’s chemical testing and science programs. Today, the NRC published a report reviewing EPA’s Integrated Risk Information System (IRIS) process for assessing chemical hazards and concluding that the agency has made substantial improvements in response to the NRC’s general recommendations made in 2011. The report published on May 2 found several key flaws in EPA’s draft “state-of-the-science” evaluation of nonmonotonic dose-response relationships for endocrine disruptors. Both reports were authored by committees of the NRC’s Board on Environmental Science and Toxicology.

The IRIS report was commissioned by Congress in response to a 2011 NRC report on EPA’s IRIS assessment of formaldehyde that found deficiencies in the agency’s general assessment methods. According to the new review, EPA is on track to make the IRIS process more effective and efficient. The agency has followed several of NRC’s 2011 recommendations; for example, EPA has implemented a new document structure that streamlines assessments, added a standard preamble explaining the process and its underlying principles, and drafted a handbook describing the process in greater detail. Noting that EPA is still in the process of making the changes recommended by the NRC, the study makes additional suggestions to strengthen to the IRIS process. Looking forward and considering the constantly evolving nature of the science involved, the report identifies critical steps to improve assessments in the future, calling for continuous updates to assessment methods, systematically identifying and addressing inefficiencies, and regular evaluation of chemical-assessment teams’ expertise and training.

The NRC took a dimmer view of EPA’s draft evaluation of nonmonotonic dose-response relationships, which intended to answer basic questions about the phenomenon wherein lower doses may be associated with larger responses, which may not be detected by traditional toxicological models. EPA’s evaluation tried to address, for example, the implications of nonmonotonic dose-response relationships for toxicity testing, weight of evidence conclusions, and risk assessment determinations. However, the authoring committee found that the process adopted by EPA in conducting the evaluation was “poorly described and inconsistent” in its approach to analyzing and summarizing data. The NRC report recommends that EPA develop and apply an “analytic plan” to predefine and document literature search strategies and establish criteria for selecting and analyzing studies, and analyze toxicity-testing strategies under more specific constraints. In addition, the NRC cited the lack of analysis supporting EPA’s conclusion that nonmonotonic dose-response relationships can have both qualitative and quantitative effects which would be considered appropriately using current risk-assessment practices. The report urges EPA to address how nonmonotonic dose-response relationships for estrogen, androgen, and thyroid pathways would be treated under current risk-assessment practices. Although the report does not address more fundamental questions – such as whether endocrine-disrupting chemicals are harmful, or the adequacy of EPA’s testing techniques – the NRC’s critique may raise issues in other areas, including the U.S.’ trade in chemicals with the European Union, which has adopted a more aggressive approach to regulating potential endocrine disruptors.