Theme: 07. Per- and Polyfluorinated Alkyl Substances (PFAS)

Topic Area: 7l. PFAS Human Health and Ecological Risk Assessment and Toxicity

Formatted Title:

Comparison of Approaches for Evaluating Relative Toxicity and Risk of PFAS without Regulatory Toxicity Values

Background/Objectives:

Numerous per- and polyfluoroalkyl substances (PFAS) are detected in environmental media at hazardous waste sites. However, there are limited regulatory toxicity factors for PFAS for assessing human health risk when characterizing a site and making risk-management decisions. The need to address toxicity of mixtures for all PFAS detected at a site is reflected in USEPA’s PFAS Strategic Roadmap (2021), National PFAS Testing Strategy (2021), and draft Framework for Mixtures of PFAS (2021), as well as various state agency approaches. Various methods are currently being considered and/or implemented to derive surrogate/relative toxicity values for PFAS without regulatory criteria. Four of these methods predict toxicity and risk based on: 1) relative potency factors (RPFs) compared to PFOA, 2) PFAS grouping/subgrouping, 3) molecular weight of the specific PFAS, and 4) USEPA’s generalized read-across (GenRA) new approach methods (NAMs) tool (https://www.epa.gov/comptox-tools/genra-resource-hub). The objective of this study was to compare these four methods and discuss the strengths and limitations of each.

Approach/Activities:

Four methods developed and/or used by regulatory agencies were compared:

Hawaii Department of Health (Method 1)
Texas Commission on Environmental Quality (TCEQ) and Health Canada (Method 2)
Heads of EPA Australia and New Zealand (Method 3)
USEPA GenRA read-across NAM tool (Method 4).

USEPA’s GenRA tool was used to identify surrogate chemicals based on chemical structural similarity. Surrogates in the same primary or secondary chemical subgroup were retained. In-vivo point of departure (POD) toxicity data for the surrogates based on liver effects were obtained from USEPA toxicity documents. Weighted PODs were used and adjusted for toxicokinetic differences using pharmacokinetic modeling of blood serum concentrations. The POD-based RPFs for target PFAS were derived based on reference PFAS (PFOA or PFOS).

Three PFAS with toxicity values recently published by USEPA (PFHxA, PFBA, and PFDA) were selected for relative toxicity value predictions using the four methods described above. Results were compared with USEPA’s recently published toxicity values in order to assess how well the methods predicted toxicity values derived using traditional methods.

Results/Lessons Learned:

Surrogate/relative toxicity values for the three target PFAS were identified using the four methods and compared to USEPA’s draft or final toxicity values. USEPA’s GenRA read-across NAM tool (Method 4) was the best relative toxicity method for predicting traditionally derived regulatory toxicity values, suggesting this tool may be useful for predicting relative toxicity of PFAS without regulatory toxicity values. The key differences, strengths, and limitations in the methods (e.g., incorporation of PFAS grouping/subgrouping, carbon chain length, mechanism of toxicity, blood serum data, toxicity data considered, health endpoints) will also be discussed, as well as the challenges and uncertainties.

Menu

Neha Gowadia