Based on USEPA (1986) Table 2 for mixtures risk assessment, please select the best classification for data quality for PFAS health effects information

Answer Explanations

• Full health effects data are available but extensive extrapolation for route or duration of exposure or for species differences. These extrapolations are not directly supported by the information available

  Expert 6

  I am torn between dot points 2 and 3. While extrapolations between routes, duration of exposure and species are reasonably supported by pharmacokinetic considerations for some PFAS, there are also some significant data gaps for the PFAS less well studied.

• Full health effects data are available but extensive extrapolation for route or duration of exposure or for species differences. These extrapolations are not directly supported by the information available

  Expert 2

  Two major issues here:
  Pharmacokinetics of PFAS varies between species.
  Pharmacokinetics of most subgoups of PFAS are not fully known (lack of data).

• Certain important health effects data are lacking and extensive extrapolations are require for route or duration of exposure or for species differences

  Expert 10

  The answer here depends on the critical effect that is the basis of the risk assessment. For many types of toxic effects there is a lack of data for most PFAS, including knowledge of a common mode of action.
  However, for a few effects (e.g. liver hypertrophy) there are enough data to derive relative potency factors (RPFs). In the case of the EFSA Opinion for tolerable weekly intake of PFAS in food, it was assumed that four PFAS (PFOA, PFNA, PFOA and PFHxS) had the same elimination kinetics and a common toxicity (immunotoxicity). But, there was an extensive extrapolation regarding mode of action, with the justification that it was precautionary. The lack of common mode of action is often used as "strictly-scientific" justification for not grouping many PFAS for risk assessment (see e.g. Goodrum et al., 2021), but in reality most jurisdictions are willing, and able, to be pragmatic and precautionary.

  Goodrum, P.E.; Anderson, J.K.; Luz, A.L.; Ansell, G.K. (2021) Application of a Framework for Grouping and Mixtures Toxicity Assessment of PFAS: A Closer Examination of Dose-Additivity Approaches. Tox. Sci., 179, 2, 262–278.

• Expert 11

  I am not an expert in the literature on the toxicity or chemicals interactions of PFAS. I suspect the level of information varies greatly across the specific PFAS substances in the mixture. As a result, any of the five classifications could be applicable for certain PFAS substances.

• Certain important health effects data are lacking and extensive extrapolations are require for route or duration of exposure or for species differences

  Expert 5

  The classification most likely to be appropriate given the present state of dose response and health effects information seems to be E - certain health effects data are lacking and extensive extrapolations are required. This characterization may be only marginally better then F, lack of health effects data on the mixture or components precludes a quantitative risk assessment. Critical data gaps include quantitatively reliable information on clearance values (often discussed as "half lives"), and at least some reliable infomration on mode of action and its human relevance.

• Certain important health effects data are lacking and extensive extrapolations are require for route or duration of exposure or for species differences

  Expert 4

  The limited data does not "PRECLUDE" a quantitative risk assessment. It affects the accuracy and precision of risk estimates, which should both be discussed in the uncertainty analysis portion of the QRA.

• Full health effects data are available but extensive extrapolation for route or duration of exposure or for species differences. These extrapolations are supported by pharmacokinetic considerations, empirical observations, or other relevant information

  Expert 1

  It's difficult to pick one of these options. There is a lot of data on PFAS that contribute the most to drinking water contamination (i.e., PFOS, PFOA, PFNA, PFHxS), but very little on others which generally have smaller contributions to the total organofluorine load. In terms of extrapolation, it all depends on the studies that are being used to determine health-based guidance values. There are numerous epidemiological studies at environmentally-relevant doses and routes of exposure, and very few animal studies at environmentally-relevant doses.

• A lack of health effects information on the mixture and its components in the mixture precludes a quantitative risk assessment

  Expert 8

  It is not clear what the context for this question is. The problem formulation statement provided is not sufficiently detailed to provide the actual PFAS to be considered. If one is only considering the 6 PFAS included in the UCMR3 (PFOA, PFOS, PFNA, PFHxS, PFHpA and PFBS) then the second answer applies – “Full health effects data are available but extensive extrapolation for route or duration of exposure or for species differences. These extrapolations are supported by pharmacokinetic considerations, empirical observations, or other relevant information.” Toxicity, half-life, and pk modeling methods exist for these 6 individual PFAS, but this does not mean that it is appropriate to group these chemicals into a mixtures assessment. In fact, the toxicity data and MOA data suggest that combining them is actually inappropriate.
  If one is considering other PFAS, I am guessing the answer would be “A lack of health effects information on the mixture and its components in the mixture precludes a quantitative risk assessment”. However, since I don’t know what those PFAS are this is only a guess.

• Certain important health effects data are lacking and extensive extrapolations are require for route or duration of exposure or for species differences

  Expert 3

  Note again that the USEPA decision tree was developed in 1986. It is good advice as far as it goes. Improved extrapolation methods, read across, QSAR, use of adverse outcome pathways (AOPs) have all provided means to make more full use of incomplete data.

• A lack of health effects information on the mixture and its components in the mixture precludes a quantitative risk assessment

  Expert 9

  While there are abundant data on numbers of certain PFAS and their levels in environmental media and in living organisms, there are very little data with respect to health effects of mixtures. The data that are available are either unanalyzed as mixtures (i.e., some epidemiological studies) and concern mostly just binary mixtures of PFAS.

• Certain important health effects data are lacking and extensive extrapolations are require for route or duration of exposure or for species differences

  Expert 7

  For this question: Assuming that human health effects from PFAS are not only caused by drinking water intake and since neither the exposure pathways for all PFAS (whatever definition) nor the effects are known, I would state that at present the last bullet applies (..precludes a quantitative risk assesssment).
  BUT: I think that protective measures can be taken based on few model compounds that are well characterized and then extrapolations done. Such approaches have been done for PCB (with indicator PCB or dioxin-like PCB) or PCDD/PCDF (not all congeners have data) in pragmatic ways (with definitions).
  In order to act, no full assessments are necessary but sufficient sound basis.