EPA Science Advisory Committee on Chemicals Identifies Issues with Formaldehyde TSCA Risk Evaluation

Ever notice that sometimes the headline of an article doesn't quite match the content? A notable example of this type of misdirection, sometimes referred to in journalism as "burying the lede," comes from reading the U.S. EPA Scientific Advisory Committee on Chemicals (SACC) review of the Agency's draft Toxic Substances Control Act (TSCA) Risk Evaluation for Formaldehyde, issued in August 2024.

According to the Lautenberg Act, bipartisan amendments to TSCA adopted in 2016, EPA's TSCA risk evaluations must adhere to important scientific standards, including that they be based on "best available science" and the weight-of-scientific evidence, while integrating available information to support future regulatory actions. A fulsome review of the SACC report suggests numerous areas in which EPA's process and draft scientific conclusions have failed to achieve these high TSCA standards. The SACC review unequivocally identified many instances where the non-cancer and cancer hazard values presented in the U.S. EPA Office of Research and Development (ORD) Draft (and now Final) Integrated Risk Information System (IRIS) Hazard Assessment do not rely on the best available science. Confusingly, the Executive Summary, which fails to incorporate the fundamental scientific criticisms in the report as well as past peer reviews and may not reflect the views of the individual panel members, suggests that "the draft documents are comprehensive and rely on the best available science." This sleight of hand leaves public readers of the full peer review pondering the disconnect between what would be a key takeaway and the actual findings.

EPA urged to consider best available science used by other competent authorities

Not only does the SACC Formaldehyde Review both directly and indirectly state that the hazard information (extracted from the ORD IRIS assessment) does not meet the TSCA standard of using the best available science, but it also provides guidance on how this part of the draft risk evaluation could be revised in a manner that meets the standard. For example, the SACC recommended recognizing existing exposure limits that have been enacted by other competent authorities where the best available science was used, as stated:

The Committee recommended that EPA review approaches taken by other regulatory authorities to ensure exposure limits are consistent with the best available science. [pg. 18; Bold Added for emphasis]

And,

Review the approaches taken by other governmental authorities, which set exposure limits consistent with the best available science based on the weight of the scientific evidence (e.g., Acute Exposure Guideline Levels (AEGL), Canada (2021), EU). [pg. 24; Bold in Original Text, Formal Recommendation, Underlining Added for Emphasis]

And,

Compare conclusions reached in EPA assessments (hazard, exposure, risk) with decisions reached by other global regulatory authorities and scientific organizations...For significant differences, provide a thorough and coherent technical explanation of the differences…Acknowledgement of these differences could be a big step toward credibility. [pg. 79; Bold in Original Text, Formal Recommendation]

The SACC also provided several examples of recently enacted exposure limits which it viewed as more appropriate for achieving regulations based on "best available science," as stated:

In July 2023, the EU finalized its formaldehyde risk evaluation and, in careful consideration of the data, implemented a permissible worker exposure level of 300 parts per billion (ppb). (EU Council Directive 98/24/EU, Annex 1.) [pg. 55]

And,

The occupational exposure levels used in Ontario, Canada, are Short-term Exposure Limit (STEL) of 1 ppm and a ceiling limit of 1.5 ppm (2022) and Australia are: Maximum 8-hour Time-Weighted Average (TWA): 1 ppm or 1.2 mg/m³ and the Maximum Short-Term Exposure Limit (STEL): 2 ppm or 2.5 mg/m³. The more conservative EU OEL for formaldehyde was determined as 8-hour TWA: 0.3 ppm or 0.369 mg/m³ and 15-min STEL: 0.6 ppm or 0.738 mg/m³. (European Commission, 2016). [pg. 58]

And,

The Committee saw value in recognizing the work of other authoritative organizations and discussion as to why EPA's assessment delivered different answers than achieved by other regulatory authorities, particularly the EU. This discussion would have demonstrated the Agency had been aware of these differences and possess several explanations…again improving confidence. [pg. 93]

Key studies in IRIS do not match peer review recommendations for scientific rigor

Both the EPA Human Studies Review Board's (HSRB) October 2023 review and the SACC formaldehyde review have opined on the scientific rigor of observational epidemiology studies and controlled chamber studies. ^[1] For the purposes of calculating hazard values, controlled chamber studies represent the best available science. The SACC specifically referenced the HSRB peer review and agreed that the controlled chamber studies cited by other competent authorities are preferred as the basis for the hazard value due to the improved scientific rigor of these studies:

Follow the HSRB recommendation to rely on Mueller et al. (2013) and Lang et al. (2008) to derive a POD consistent with the best available science using a weight of the evidence approach.

Without some justification for why individuals in the Krzyzanowski et al. (1990) study would be so much more sensitive to the effects of formaldehyde than individuals in the set of controlled human exposure studies (where no pulmonary impacts were found), it is difficult to accept the POD of 0.021 mg/m3 (17 ppb) as being scientifically sound. [pg. 34; Bold Added for Emphasis]

Concerns were raised by some Committee members regarding studies selected by ORD IRIS for chronic non-cancer hazards. These studies are mainly observational and unreliable for identifying a point of departure. The studies identified by ORD IRIS for the weight of evidence for chronic human health non-cancer hazard do not adequately address the chosen endpoint due to several limitations, including but not limited to the ability to determine causality specific to formaldehyde, confounders that were not addressed and including use of self-completed questionnaires instead of measured health effects which decreases the reliability of results. In addition, the use of an uncertainty factor is not necessary when study population for all studies were mainly sensitive, asthmatic children; therefore, sensitive populations have been addressed. [pg. 34; Bold Added for Emphasis]

Because the draft and final IRIS assessments did not address either the HSRB or the SACC peer review recommendations, it sets unscientific and precautionary hazard values based on studies that were not conducted with sufficient scientific rigor, nor considered best available science. As such, the reliance on the final IRIS assessment is not consistent with TSCA's scientific standards.

The weight of scientific evidence does not support the elevated cancer risks suggested in EPA's draft risk evaluation

When it comes to the cancer hazard endpoints of mutagenicity and leukemia, the SACC Formaldehyde Review stated:

Some Discussants did not entirely dismiss the EPA's findings on mutagenesis and myeloid leukemia, whereas a number of discussants fully disagreed with the Agency's findings. However, EPA should consider all available science (the "best science" approach) and carefully justify their conclusion. [Bold Added for Emphasis]. [pg. 103]

Regarding the cancer hazard value for nasal tumors, the SACC formaldehyde review stated:

Review the literature supporting both threshold and non-threshold models of formaldehyde carcinogenicity, and on the basis of 'Best Available Science' and if appropriate, calculate a level of risk independent of the IRIS assessment. [pg. 104 Bold in Original Text, Formal Recommendation, Underlining Added for Emphasis]

The SACC also resolutely stated that the IRIS cancer hazard value for nasal tumors is not supported and should not be used in the draft TSCA risk evaluation to calculate risks:

[Mode of Action]: Cancer IUR used from the unedited, Draft 2022 IRIS Assessment on Formaldehyde is not supported by a proper holistic interpretation of the collected data and should not be used by OPPT for risk assessment. [pg. 84, Bold Added for emphasis]

Finally, the SACC provided specific reasons why such a definitive statement was made, including:

In addition, given the MOA evidence for a dose-response model of the key events, and the science supporting that air concentrations below that threshold do not significantly impact the formaldehyde concentrations in the cells of the body as supported by the extensive elegant adduct work by the Swenberg lab at UNC Chapel Hill, NC (Edrissi et al.,2013; Rager et al., 2014; Yu et al., 2015; Lai et al.,2016 ; Edrissi et al., 2017; Liu et al., 2018; Leng et al., 2019), the body of scientific evidence does not support the elevated risks suggested by EPA in the Formaldehyde Draft Human Health Risk Assessment. [pg. 89, Bold Added for Emphasis]

Despite the conclusion provided in the executive summary, it is clear the SACC determined that the hazard endpoint and hazard values that were used in the draft and final IRIS assessment do not reflect the best available science, nor do they represent the weight of the scientific evidence consistent with the rigorous TSCA scientific standards.

Action is needed for alignment with TSCA scientific standards

There is a clear path forward for EPA to complete the TSCA risk evaluation independent of the IRIS assessment. That path includes careful evaluation of scientific choices recently taken by the European Union and World Health Organization and revising the TSCA risk evaluation under the guidance of the SACC's detailed recommendations for consistency with the best available science standard.

It is imperative that EPA sticks to valid data and it does not rewrite history regarding this highly critical peer review and what is needed to satisfy scientific standards demanded by Congress and credible future regulation.

_{[1] See The EPA Human Studies Review Board Scientific Peer Review Highlights that Major Revisions Are Needed to EPA's Draft IRIS Formaldehyde Assessment - American Chemistry Council; AND EPA Human Studies Review Board Peer Review Shows that Not All Peer Reviews Are Equal: A Tale of Two Peer Reviews On Formaldehyde - American Chemistry Council}

Related Content