From the archives - LQM evaluation of the updated Toxicological Review of Benzo[a]pyrene (USEPA 2017)

This popular post was from 13/04/2017

The US Environmental Protection agency (USEPA) has a long history of providing peer-reviewed expert opinions on the health risks posed by environmental pollutants in documents known as “Toxicological Reviews”.  Summaries of the reviews are also published on the USEPA’s Integrated Risk Information System (IRIS) .  But this is just one source of such reviews amongst the many US executive agencies, for example the Agency for Toxic Substances and Disease Registry (ATSDR) publish equally authoritative Toxicological Profiles.

In considering the toxicological effects of contaminants, the USEPA derive Reference Doses (RfD) and Reference Concentrations (RfC), which represent oral and inhalation exposures respectively, that are protective of any threshold effects of the contaminant.  Where appropriate, they also derive oral Slope Factors (SF) and Inhalation Unit Risk (IUR) values, which indicate lifetime cancer risks via oral and inhalation routes respectively.  If a contaminant exhibits both threshold (non-cancer) and non-threshold (cancer) effects, and suitable toxicological studies are available, then both an RfD (or RfC) and a SF (or IUR) will be derived and risk assessors are expected to consider which is the most significant or applicable on a case-by-case basis.

In a similar manner, the UK toxicological guidance contained in SR2 dictates the use of Tolerable Daily Intakes (TDIs) in relation to threshold behaviour and Index Doses (IDs) for non-threshold behaviour.  USEPA Slope factors and IURs are normally derived by linear extrapolation of tumour incidence observed in experiments involving animals exposed to the test substance.  However, it is important to recognise that in 2012 UK Department of Health Committee on Carcinogenicity (CoC) reported that it ”considers that it is not valid to extrapolate from the observed dose range in animal carcinogenicity studies through many orders of magnitude to give an estimate of excess lifetime risk at environmental levels of exposure, e.g. 1 case of cancer in a population of 1 million (1 in 106 ), because of the uncertainties involved. This methodology generates a false sense of precision which cannot be justified.”  So using SL and IUR values derived by linear extrapolation of animal carcinogenicity studies as the basis for deriving IDs for use in human health risk assessments in the UK is problematic, at best.

The USEPA published their first assessment of benzo[a]pyrene in 1987.  However, in light of new information, policy and methodology, the USEPA has reconsidered its original assessment.  In January 2017, a new, revised Toxicological Review of Benzo[a]pyrene was published.

The USEPA clearly state that in relation to the RfD and RfC: “These organ- or system-specific reference values may be useful for subsequent cumulative risk assessments that consider the combined effect of multiple agents acting at a common site.” – i.e. when considering non-cancer effects only on specific common sites these reference values could be applied.

The USEPA (2017) also recommended a SF of 1 per mg/kg BW/day, which was based on linear extrapolation (of the BMDL10) from observed alimentary tract tumours in mice, and a IUR of 6 x 10-4 per µg/m3, which was based on linear extrapolation (of the BMCL10) from observed upper respiratory and upper digestive tract tumours in male hamsters.  As mentioned above (CoC, 2012), linear extrapolation should not be used as the basis for deriving an Index Dose.  Further, the USEPA (2017) make it clear that the Age-Dependant Adjustment Factors (ADAFs) recommended by the USEPA (2005) should be applied, to account for increased risk of cancer for individuals exposed during early life to mutagenic mode of action carcinogens (i.e. BaP).  Applying the USEPA (2005) ADAFs would result in daily dose equivalents of 6.04 x10-3 µg/kg BW/day and 2.87 x10-3 µg/kg BW/day for oral and inhalation exposure, respectively, based upon lifetime exposure (0-70 years).

It should also be noted that the USEPA (2017) clearly state that the oral SF and inhalation IUR for BaP is derived with the intention that it will be paired with EPA’s relative potency factors for the assessment of the carcinogenicity of PAH mixtures. The RfD, RfC, SF and IUR values recommended in USEPA (2017) are the same as those proposed in the earlier draft version published for external review in September 2014.

The LQM/CIEH ‘Suitable 4 Use Levels’, or S4ULs, were published in 2015 as generic assessment criteria for use in UK human health risk assessments of contaminants in soil based on a “minimal or tolerable” level of risk and capable of demonstrating that developments are ‘safe’ and ‘suitable for use’, as required by the NPPF and the relevant planning guidance in Wales, Scotland and Northern Ireland.   The S4ULs for BaP alone (i.e. not as a surrogate marker for PAH mixtures for which a separate S4UL is presented) are based on the critical effect of both oral and inhalation exposure to BaP being the development of various types of cancer.  The S4ULs for the various land uses are based on an IDoral of 3.1 x10-2 µg/kg BW/day, and an IDinhal of 0.3 x10-3 µg/kg BW/day.  The draft version of the revised USEPA Toxicological Review was available during the preparation of the LQM/CIEH S4ULs, which recommended the same vales as those in the final version (USEPA, 2017).  The values recommended in this draft were considered in deriving the S4UL, but the SF and IUR values were not considered the most appropriate basis for deriving UK assessment criteria, even though this would have resulted in more cautious S4UL.

Now that the final version of the revised USEPA review has been published, there has been some discussion about whether the values recommended could, or should, be used as the basis for revised assessment criteria for use in the UK.  In deciding whether we need to revise the LQM/CIEH S4UL for BaP we have considered the following:

  • Assessments based on recommended RfD and RfC would only be protective of the developmental threshold effects of BaP exposure and would not protect against its more significant cancer-causing effects (it should be noted that any assessment of threshold effects must also take account of background exposure by applying the relevant MDIs). Consequently, as the critical effects of BaP relate to cancer formation, the USEPA RfD and RfC are not a suitable basis for deriving assessment criteria in the UK.
  • The use of IDs calculated from the SF and IUR for BaP involves linear extrapolation methods that are not valid under UK toxicological policy. It should also be noted that their use would result in even lower assessment criteria for BaP.  For example, a revised LQM/CIEH S4UL for BaP alone (residential with home-grown produce landuse, 1% SOM) would be about 20% of the current value if health criteria values using IDs derived from the USEPA (2017) SF and IUR were used, and excludes the impact on the assessment criteria for the remaining PAHs related to BaP by their relative potency factors.

Consequently, in the absence of re-evaluation by an authoritative UK body, to adopt more stringent toxicological health criteria values, the IDs underpinning the current S4UL remain the most appropriate for use in UK human health risk assessments intended to demonstrate “minimal or tolerable” levels of risk.

Finally, the USEPA Toxicological Review only considered BaP, and the USEPA (2017) favour the use of potency factors to assess PAH mixtures, as opposed to the Surrogate Marker Approach recommended by Public Health England.  Consequently, the toxicological recommendations of the USEPA (2017) for BaP have no relevance to the Index Doses used as the basis of the LQM/CIEH S4ULs for “Coal Tar (BaP as surrogate marker)”.