For the purposes of conducting a screening level risk assessment, how do you recommend the potential hazards/risks from PFAS mixtures be assessed?

Answer Explanations

• Assume dose additivity

  Expert 1

  As mentioned above, screening should assume dose additivity and include all PFAS as measured through adsorbable/extractable organic fluorine analyses. This approach may be very conservative, but it is relatively cheap and would help quickly identify environmental/biological samples that may exceed health-based guidance values. These samples could subsequently be analyzed using targetted methods, which would allow identifying the contribution of known PFAS to the total organic fluorine load.

• Use of a relative potency approach Assume dose additivity

  Expert 6

  Development of a Hazard Index (HI) type risk assessment approach is conventional for "screening purposes" but there will be difficulties in choosing the right data sets for PFAS (see also answers to 1.7, 1.8 1.9 and 1.11). The development of suitable Relative Potency Factors (RPFs) may assist with this process where no suitable RfDs are available (see also answer to 1.8)

• Expert 11

  I am not sufficiently familiar with the existing monitoring or toxicity data on PFAS to offer an opinion.

• Use of a relative potency approach

  Expert 2

  I would use relative potency based on a common mode of action, although however such activity is still not known. Some authors claim thyroid hormone disruption/in vitro approaches are available, but the validity of such assays is still not vetted.

• Other (please explain)

  Expert 7

  For which purpose a screening; this would not help a risk assessor nor risk management. Guidance should be provided for risk managers to take action (and not to initiate years of research)
  we have recently published a paper on occurrence of PFOS, PFOA and PFHxS in human milk (Fiedler and Sadia, 2021)
  We did not find high concentrations of PFAS in surface waters in three continents (Baabish et al., 2021)
  Initial assessment of foodstuffs did not find very high concentrations (not as high as reported by EFSA 2018, 2020 (EFSA CONTAM Panel et al., 2020) (Sadia et al., 2020)

  Additional references:
  Baabish, A., Sobhanei, S., and Fiedler, H. (2021). Priority perfluoroalkyl substances in surface waters - A snapshot survey from 22 developing countries. Chemosphere 273, 129612.

  Fiedler, H., and Sadia, M. (2021). Regional occurrence of perfluoroalkane substances in human milk for the global monitoring plan under the Stockholm Convention on Persistent Organic Pollutants during 2016–2019. Chemosphere 277, 130287.

  Sadia, M., Yeung, L.W.Y., and Fiedler, H. (2020). Trace level analyses of selected perfluoroalkyl acids in food: Method development and data generation. Environmental Pollution 263, 113721.
  EFSA CONTAM Panel, Schrenk, D., Bignami, M., Bodin, L., Chipman, J.K., Jesús del Mazo, Grasl-Kraupp, B., Hogstrand, C., Hoogenboom, L.R., Leblanc, J.-C., et al. (2020). Public Draft: Risk to human health related to the presence of perfluoroalkyl substances in food. EFSA Journal.

• Assume dose additivity

  Expert 8

  A screening level risk assessment relies on high end and conservative estimates of exposure and hazard. If this conservative estimate of risk does not indicate a problem then the case is dropped. If it does indicate a problem, then the risk assessment is further refined. This is an iterative process. A generally accepted method is dose addition or use of the hazard index. Depending again on the problem formulation statement, and the exact group of PFAS being addressed, this may be an appropriate method. It should be noted that the HI approach will likely be driven by the PFAS with the longer half-lives.
  Another way to look at this issue would be to utilize the NHANES data, if the assessment includes any or all of the chemicals included in the NHANES survey. The blood levels in the general population represent total exposure – i.e. exposure through all routes. If no risk is determined at those serum levels than it may be prudent to drop the case. If a potential risk is identified then further work would need to be done to determine the likely source(s) of exposure.

• Use a whole-mixtures approach Other (please explain)

  Expert 9

  As summarized in the background document, some entities have applied a concentration addition approach to a relatively small number of individual PFAS found in drinking water/drinking water sources. However, also as summarized in the background document, these approaches are hampered by lack of data on PFAS that may be encountered in such environmental mixtures. Some environmental mixtures, for example at sites contaminated by aqueous film-forming foams (AFFFs), the number of identified PFAS may underestimate the number of PFAS precursors and environmental/biological transformation products of those precursors that have not yet been fully described/identified. As such, screening level "risk assessments" at this point in time and for environmental assessments will be limited to hazard identification.

• Other (please explain)

  Expert 10

  For a screening approach, I would consider using the extrable organic fluorine approach to determine e.g. those water samples with high EOF levels. This would be a fast and cheap way to screen a lot of samples. Those with high EOF could then be scritnised further. In Sweden, the EOF screening approach is being evaluated and the intital published results are promising (see: https://www.kemi.se/en/publications/pms/2021/pm-5-21-interlaboratory-comparison-of-extractable-organofluorine-eof).

• Use of a relative potency approach Assume dose additivity

  Expert 3

  Response additivity is not really applicable (US EPA 2000 and others), and whole mixture testing is largely impractical for PFAS.

  Dose additivity is a a reasonable assumption for chemical mixtures in general and for PFAS in particular. As noted in several papers, it will likely result in an overestimate of hazard / risk. This is based both on the few results for PFAS (showing no interaction in the few instances tested) and on experience with other component-based assessments. The few well-studied instances of greater than additive effects are in high-dose situations; this is a general statement referring to compounds other than PFAS. Studies of interactions generally report antagonism or less than additive effects among components. For assessment of risk from environmental exposure, an assumption of dose additivity is reasonable.

  As a conservative assumption for a screening assessment, all PFAS in the mixture could be considered equitoxic. This is not unreasonable given the results in Goodrum et al (2020) and (Bil et al 2020). This is a conservative assumption if based on most potent member of the group. It is clear that not all PFAS are equipotent, or have the same MOA. With conservative assumptions in a screen it is feasible to rank scenarios for potential human risk, including elimination from consideration of those showing negligible risk under conservative assumptions.

• Use of a relative potency approach Assume dose additivity

  Expert 4

  Screening assessments should be practical to complete. Their goal is to determine whether a more detailed assessment, screening assessments is needed, so screening assessments need to be easy, and should generate upper-bound estimates of risk.

  Doses, in practical terms, are easier to measure or estimate than responses, so dose additivity is a more practical approach (there's not enough data to know which is right).

  Similarly, there's almost never relevant whole mixture data for assessing PFAS mixtures, so the practical alternative is to make assumptions about how potent each mixture component is.

• Assume dose additivity

  Expert 5

  Inclusion of a whole mixture approach here may be somewhat of a disconnect. In the context of EPA guidance, “Whole Mixture” data or testing refers to testing and evaluation of the whole mixture, whether it be the actual mixture of concern, or a mixture deemed “sufficiently similar”, which is a technical term. Testing the whole mixture, itself, does not represent a screening approach. However, several screening approaches do apply to the mixture at hand, and (under EPA guidance) are conducted via dose additive models. Response additivity, being somewhat less conservative than dose additivity, is not recommended for screening approaches. Because RPF is, itself, a dose additive model, the most pertinent answer to this question seems to be “dose additivity”, which can be conducted using (e.g.,) a, RPF approach or a Hazard Index approach. Among these approaches, the constraint of data limitation and the circumstance of varying degrees of confidence in the data required to complete either of these approaches may well guide the decision between the RPF or HI approach.

  The preferred approach for a PFAS screening assessment might be one consistent with the screening approach established by EPA. Under this approach, a screening Hazard Index is applied, where the exposure is divided by the acceptable level (e.g., RfD value) and the Hazard Quotient values for each chemical are summed, regardless of target organ/tissue/system. This approach will be complicated by the lack of established acceptable exposure values, and perhaps by a lack of certainty in exposure values.

  If data are limited, an alternative screening approach may be considered, based loosely on the RPF approach, with further adjustments. Under such an approach, assignment of RPF values based on knowledge of potency differences for known substances may be considered, and adjustments made to account for structural alerts of other information necessary for data-limited components.

  Regardless of the approach taken, data limitations will play a large role in the interpretation of the results.