EPA’s toxicological review of PFNA gets mixed reviews from government, industry

Background: PFAS and PFNA

• Perfluorononanoic acid (PFNA) (CAS RN 375-95-1) is part of a wider class of chemicals called per- and polyfluoroalkyl substances (PFAS). PFAS are utilized in many industrial and commercial applications due to the chemicals’ ability to withstand many environmental stressors, in addition to repelling both water and oil.
• PFNA is a perfluoroalkyl carboxylic acid (PFCA), meaning that it has a chain of fully fluorinated carbon atoms (where each hydrogen atom has been replaced with a fluorine atom) and a carboxylic acid functional group. PFNA is considered a “long chain” PFAS, containing nine carbon atoms. It is traditionally used in the production of polyvinylidene fluoride (PVDF), which is utilized as insulation for wires and circuit boards and as a coating for components that are exposed to reactive chemicals. Although PFNA is not intended to be included in PVDF, it may be present as a residual byproduct in the fluoropolymer.

Background: IRIS and the EPA

• The EPA uses the Integrated Risk Information System (IRIS) to characterize and identify the health hazards of chemicals. IRIS review is part of a larger risk management system for chemicals, representing the first two steps of the process. IRIS review is followed by risk characterization and management after the EPA publishes a corresponding IRIS assessment.
• In late 2019, the EPA published a systematic review protocol for IRIS assessment of PFAS. The agency published updates to the protocol in 2020, for four PFAS: perfluorobutanoic acid (PFBA), perfluorohexanoic acid (PFHxA), PFNA, and perfluorohexane sulfonic acid (PFHxS). All four of the assessments use the scope and methods from the systematic review documents as well as additional information such as PECO criteria (Populations, Exposures, Comparators, and Outcomes). The EPA updated its process slightly in January 2021, providing additional insight on scoping, problem formulation and key science issues with PFAS, as well as planned overall objectives for the risk assessments.

The draft toxicological review for PFNA

• On March 6, 2024, the EPA announced a draft version of the IRIS toxicological review for PFNA and its salts, sodium perfluorononanoic acid (CAS RN 21049-39-8) and ammonium perfluorononanoic acid (CAS RN 4149-60-4). The draft review (Note: link leads to an automatic PDF download) dove into the research surrounding PFNA exposure and various adverse health outcomes.
• The EPA stated there was insufficient evidence to determine whether exposure to PFNA may affect the development of some cancers. However, the agency found evidence of developmental effects linked to oral exposure, as well as evidence suggesting oral exposure may cause immune, thyroid, neurodevelopmental, or cardiometabolic effects.
• The EPA set a lifetime and subchronic reference dose (RfD) of 7 x 10^-9 milligrams per kilogram of bodyweight per day (mg/kg-d).
• The comment period on the draft toxicological review ran from publication until May 6, 2024, and received 11 comments. Of these, three come from organizations: American Chemistry Council (ACC), New Jersey Department of Environmental Protection (NJDEP), and Solvay Specialty Polymers USA, LLC (as Syensqo). The rest of the comments appear to stem from an assignment for a Harvard undergraduate class. This analysis will focus on the comments from NJDEP, ACC, and Solvay.

Insights from the comments

• ACC’s comments heavily criticize the draft toxicological review, highlighting the use of epidemiological studies, and overall recommending that the agency toss the studies in lieu of animal data to “provide coherence with the reports from studies in humans.” The organization asserts that the agency should abandon its reliance on these studies, instead utilizing experimental animal data to provide a better dose-response as well as a “human-relevant biologically plausible endpoint.” Due to perceived issues with available human data, ACC argues that the animal data should be combined for any direct quantitative use in risk assessment or for establishing public health goals.
• ACC targets the agency’s findings regarding lower body weight, noting that most studies failed to report a significant association between exposure to PFNA and lower birth weight. ACC notes that the EPA’s RfD for developmental effects relies on data where only seven of 27 studies reported lower birth weights. The organization asserted that the agency’s analysis was not supported by the strength of the underlying evidence base and was in itself not aligned with EPA procedures in the agency’s IRIS handbook.
• ACC also accuses the EPA of “significant manipulation” of study data, describing a wide variety of efforts to “convert and rescale” data to better support the agency’s conclusion.
• NJDEP’s comments were broadly supportive of the EPA’s decision, with some clarifications on areas where the EPA could improve for a final toxicological review. NJDEP suggested several studies that were not mentioned in the draft IRIS document that focus on mixtures that contain primarily of PFCA which could improve the agency’s understanding of animal studies. Although the studies themselves focused on Surflon S-111, a mixture of various PFCAs, the majority of the mixture is composed of PFNA and NJDEP concluded that the toxicity of the mixture was primarily due to PFNA based on sex-specific differences in sensitivity to toxicity to the mixture.
• NJDEP approved of the conclusions that PFNA exposure causes developmental effects. Unlike ACC, the department commented the agency’s use of the studies, calling the evaluation “exceptionally thorough,” and that developmental effects observed in rats and mice bolster a conclusion that PFNA causes developmental effects in humans.
• The department disagreed with the EPA’s characterization of a study on hepatic effects in rodents, noting that the focus of the study was drug development, indicating a limited period of time of exposure to PFNA that indicated a reversibility of effects. However, NJDEP asserted that this was not a valid reason to discount the adversity of potential effects if chronic exposures are considered.
• Syensqo’s comments focus mainly on the EPA’s focus on PVDF as a potential exposure source for PFNA, arguing that aqueous film-forming foams (AFFFs) are the primary source of PFNA contamination and releases in the environment. Syensqo asserts that the “millions of gallons of AFFF” released into the environment far outweighs the mass of PFNA utilized to produce PVDF, citing high levels of detections of PFNA near sites where AFFF was used, but PVDF manufacturing was not present. The comment also notes that for non-military specification AFFF, PFNA was often the primary fluorosurfactants utilized, highlighting its use in National Foam’s AFFF products. This assertion is bolstered by further data on the known use of National Foam AFFF around the country with concordant detections of PFNA in groundwater.

Further position analysis of the comments

• Although the EPA may make smaller changes to the final version of the toxicological review for PFNA, it is unlikely to make some of the sweeping changes requested by industry, including the dropping of epidemiological studies in favor of a focus on animal studies. However, NJDEP’s suggestion of additional animal studies may also provide the EPA with additional information to bolster its claims. Although the agency does not typically provide a detailed response to comments on IRIS toxicological reviews, it does normally collate public comments and characterize them, providing occasional notes on the EPA’s reaction to the comments. In a collection of public comments on the draft toxicological review of PFHxA, the EPA frequently noted its plans to add additional studies into the final version of the toxicological review, explaining that it “intends to incorporate newer literature as appropriate based on peer review feedback.” This indicates that although the agency is open to adding additional literature, it will run those studies through its peer reviewers before it finalizes their inclusion in the toxicological review. The agency also highlighted its willingness to reevaluate studies, characterizing a study that Crowell & Mooring LLP had submitted as “uninformative due to reporting deficiencies” but that it would reevaluate it based on newly submitted data. The agency also indicated some level of resistance to changing how it classified its levels of evidence, characterizing another comment’s text on hematopoietic effects as not accurately reflecting the EPA’s actual statements.
• The support of NJDEP on most aspects of the agency’s review, with minor edits, may signal that the agency will incorporate more of NJDEP’s work into its final review. There is also a possibility, that once the agency considers some of the studies that NJDEP raised (as well as any others that may show up), it will be able to strengthen or weaken its conclusions after consideration of the evidence. Based on the EPA’s muted response to comments on the PFHxA toxicological review, the agency was willing to implement more NJDEP studies, pending further peer review, and cites the state agency’s comments heavily throughout the piece.

To contact the author of this analysis, please email Walker Livingston ( wlivingston@agencyiq.com).
To contact the editor of this analysis, please email Alexander Gaffney ( agaffney@agencyiq.com)

Key Documents and Dates

• IRIS page for PFNA
• IRIS Toxicological Review of Perfluorononanoic Acid (PFNA) and Related Salts (Public Comment and External Review Draft, Mar 2024) (March 7, 2024) (note: clicking link will automatically start download)
• IRIS Toxicological Review of Perfluorononanoic Acid (PFNA) and Related Salts (Public Comment and External Review Draft, Mar 2024) Supplemental Information (March 7, 2024) (note: clicking link will automatically start download)