Download PDF
January 9, 2024

EPA Releases Decision Framework to Assess Eye Irritation or Corrosion in New Chemicals

Bergeson & Campbell, P.C.

The U.S. Environmental Protection Agency (EPA) released on January 8, 2024, a decision framework for identifying eye irritation or corrosion hazards for new chemicals reviewed under the Toxic Substances Control Act (TSCA). 89 Fed. Reg. 1093. EPA states that the “New Chemicals Program Decision Framework for Hazard Identification of Eye Irritation and Corrosion” provides a decision framework for use by the Office of Pollution Prevention and Toxics (OPPT) New Chemicals Division (NCD) to identify eye irritation or corrosion hazards for new chemical substances based on prioritization of reproducible, human-relevant data. EPA notes that the framework supports its mandate under TSCA to promote the development and implementation of New Approach Methodologies (NAM) to reduce, refine, or replace vertebrate animal testing and provide information of equivalent or better scientific quality and relevance for assessing risks of injury to health or the environment.

According to EPA, it based the framework on the prioritization of reproducible and human-relevant data and upon existing peer-reviewed literature, accepted Organization for Economic Cooperation and Development (OECD) Test Guidelines, and other existing accepted risk assessment approaches. EPA provides the following brief overview of the decision framework steps:

  • Available data on chemical substance evaluated for quality and applicability.

    • Data are prioritized in the following order:

      • Data from human cell or tissue test methods.

      • Data from in chemico, in vitro, or ex vivo test methods.

      • Data from in vivo test methods.

    • Where data from eye irritation or corrosion tests are not available, test methods that assess skin irritation or corrosion potential may be considered.

  • If eye and skin irritation or corrosion data are not available, other information may be considered. Where no data exist, a hazard determination may not be possible.

  • Any relevant human data will be evaluated and incorporated on a case-by-case basis.

Full details of how existing data are prioritized are provided in the framework document.


Bergeson & Campbell, P.C. (B&C®) applauds EPA’s announcement. It demonstrates EPA’s commitment to developing NAMs to reduce vertebrate testing and doing so in a manner that provides stakeholders an opportunity to review the proposed NAM and comment on its appropriateness as an alternative to in vivo testing.

EPA’s Framework discourages prospective use of in vivo acute eye irritation/corrosion studies in albino rabbits (i.e., OECD 405). This represents a departure from OECD guidance document 263, which reserves performing an OECD 405 study “[a]s a last resort.” There are, however, several aspects of the Framework for which EPA could have provided more clarity to ensure its expected benefits, namely “transparency for stakeholders” and “improved consistency across final new chemical risk assessments,” as discussed below.

EPA stated in the Framework that “Methods for evaluating analogues for suitability are outside the scope of this framework.” We view this as a significant limitation, given that analog selection is critical in ensuring that the analog appropriately addresses the data gap for the new chemical. It is important that EPA provide a scientific justification for its analog selection. It is not clear why EPA did not reference analog selection resources that it has referenced in previous TSCA documents (e.g., Points to Consider When Preparing TSCA New Chemical Notifications).

EPA stated that “Further details regarding scientific quality considerations are outside the scope of this framework.” While that may be true, the quality of the data used to inform EPA’s decisions is a key factor to consider when evaluating the appropriateness of using data. It is important to have objective, high-quality criteria to ensure that an Administration does not over- or underweight studies to support a preferred outcome. This is an issue across TSCA regulatory matters, so while it is outside the scope of the irritation/corrosion framework, it is still an important issue for EPA to address.

B&C notes that EPA stated that it “uses three categories to identify new chemicals for their eye irritation potential: corrosive, irritating and nonirritating.” Our view is that EPA could strengthen its framework also to consider concentration thresholds. Understanding the thresholds for irritation and corrosion can help inform whether certain conditions of use may or may not be unreasonable risks.

It is also not clear if EPA will take this approach to its test orders. This Framework was developed at or about the same time EPA was preparing a TSCA Section 4(a)(1) test order on 2,3,3,3-Tetrafluoro-2-(heptafluoropropoxy)propanoyl fluoride (HFPO-DAF). In the TSCA Section 4(a)(1) test order for HFPO-DAF that EPA issued on August 15, 2023, EPA ordered a “Bovine Corneal Opacity and Permeability Test (OECD 437),” and if that test shows corrosivity to the eyes, then no further ocular testing is required. If the test does not show corrosivity to the eyes, then EPA requires the test order recipients to perform a Reconstructed Human Cornea-Like Epithelium (RHCE) Test Method (OECD 492B). In the proposed framework, EPA specifies that OECD 492B can be used as the sole test (rather than a tier subsequent to the OECD 437).

Finally, EPA stated that “The complete absence of any relevant data or structural alert information for the new chemical substance may preclude a hazard determination.” This is an important consideration for submitters because in the absence of this information, EPA may conclude it has “insufficient” data to inform its assessment and could impose up-front testing prior to commercialization. As with other endpoints, submitters should be sure to address irritation/corrosion along with the other health and environmental endpoints, whether those endpoints are addressed with models, analogs, or testing on the substance itself.

B&C views EPA’s Framework as a positive outcome for replacing the use of in vivo animal testing with information sources that may provide more relevant data for protecting human health. EPA’s Framework will likely be well received by entities that use TSCA regulated substances for other applications such as cosmetic products, which have separate certification programs to ensure “cruelty-free” testing of products. Notwithstanding these benefits, we are hopeful that EPA will consider developing guidance for submitters that provides transparency with EPA’s selection and justification of analogs, its criteria for determining study quality, and consistent approaches to testing for both new and existing chemical substances.