Technical Report

TR 112 – Refined Approaches for Risk Assessment of PBT/vPvB Chemicals

TR 112 : Refined Approaches for Risk Assessment of PBT/vPvB Chemicals | October 2011
ISSN-0773-8072-112 (print)
ISSN-2079-1526-112 (online)


Building upon a previous ECETOC report to develop a framework for the risk assessment of PBT chemicals (ECETOC, 2005a), this report reviews the scientific developments that are available to date and details of the on going research that is being carried out with the specific aim of reducing the uncertainty of risk assessments of PBT/vPvB chemicals. Several case studies have been analysed and the literature on newly developed methodologies has been reviewed.

PBT assessment is in many cases a worst-case assessment of intrinsic properties and is not always linked to exposure assessment or the likelihood that the exposure or hazard may be manifest through routine use of the chemical. A review of some recent risk assessments of PBT/vPvB and PBT/vPvB candidate substances revealed that in most cases standard modelling techniques wereused to predict environmental exposure concentrations (PECs) from emission data, physico-chemical properties and fate data and compared to predicted no effects concentrations (PNECs) derived from standard toxicity testing data and application of assessment factors.

In some cases refinements were used to reduce the uncertainty in the exposure assessment. These refinements include: The use of the overall persistence calculated with models like RAIDAR[1]or The Tool[2]; the use of environmental and biomonitoring data; predictions of spatio-temporal variation in environmental concentrations using advanced fate models and use of probabilistic models; and physiologically-based pharmacokinetic (PBPK) modelling in appropriate species to support conclusions on uptake and bioaccumulation potential. For the effects assessment refinements including the use of the critical body burden approach (CBB), more detailed insights in the mode of action and the derivation of tolerable intake levels based on available toxicological and kinetic information were applied.

The review of recent methodological developments revealed that the main improvements arerelated to enhancements in environmental fate and effects testing, modelling and (bio)monitoring techniques.

The approach proposed in ECETOC Technical Report 98 (2005a) that advocated the use of all data used to include the PBT categorisation within environmental risk assessment is still considered a valid starting point. This aims to reduce the inherent uncertainties involved when attempting to characterise exposure and risks of these substances. ECETOC (2005a) also recommended that, due to the special fate properties of these substances, exposure estimations should be extended to areas that are remote from emission sources (PEC refinement) and to exposure through secondary poisoning (PECoralconsideration) to account for food chain bioaccumulation. It also suggested that potential effects of these substances are assessed in greater detail (PNEC refinement), by considering higher tier effects data, such as chronic data, critical body burden measurements, and identifying the mode of action. Conventional risk assessment strategies as applied to non-PBTs will not necessarily take into account all of these considerations and therefore will not necessarily come to a conclusion about risk that is both comprehensive and protective for PBT/vPvB substances. vPvB substances do not fulfill the T criteria and are considered of concern due to a precautionary assumption that increasing concentrations may cause adverse effects after prolonged exposure. For these substances it is important to base the hazard assessment on information on chronic toxicity ideally supplemented with kinetic data that allow the derivation of critical body burden or internal dose information. The comparison of these values with steady state concentrations derived by refined assessment of exposure and fate of these substances, will allow a sound risk assessment including a possible no risk conclusion. Risk assessment of PBT/vPvB chemicals can also benefit from the use of probabilistic methods to better account for uncertainties within both exposure and effects assessment. By conducting a comprehensive risk assessment of PBT substances critical uses can be identified that can be used to form the basis for appropriate risk refinement and risk management strategies where the PEC:PNEC ratio is greater than 1.

Through the examples described within this report, this task force has come to the conclusion that the choice of the appropriate refinement strategy is strongly dependent on the nature of the respective chemical and its exposure characteristics. Therefore the selection of methodologies needs to be made on a case-by-case basis. Chemical space mapping could be a tool to improve decision-making on the best refinement strategy to use for a particular chemical depending on equilibrium partitioning data (octanol-air, air-water, octanol-water). It can predict the primary compartment(s) of concern. Combined with knowledge on the exposure situation this can be used as a starting point to select and prioritise the most appropriate refinement strategy. This will target subsequent empirical based work and field-based observations to address the areas of greatest uncertainty.

Exposure assessments for PBT substances have profited from recent developments in single andmultimedia models and their combination with high quality monitoring data. Single media persistence models have been developed at screening and confirmatory levels, allowing individual half-lives to be derived in individual media. These refined half-lives can be used ifneeded with the relevant partitioning data, e.g. for chemicals with multiple environmental discharges or undergoing multimedia transport, in multimedia models to evaluate their relative importance or determine overall persistence. However, it would be important to develop improved test methods to generate more realistic biodegradation kinetics data in order to reduce uncertainty of the model results. At present there are no criteria for overall persistence, as thereare for individual environmental compartments (e.g. water, air, sediment and soil). Consequently, further scientific discussion is required to discuss the importance of overall persistence, its applicability domain, and its inclusion at a policy level.

Several models and improvements in experimental methods have also been developed or are still in development to better describe bioaccumulation and the potential to biomagnify in the food chain. A combination of both experimental and modelling data could be a promising approach for more differentiated assessments of the bioaccumulation and biomagnification potential in the relevant organisms and food chains for a respective chemical.

An environmental effects assessment for PBT/vPvB chemicals should specifically focus onlonger-term tests in the environmental compartments of concern and provide a refined foodchain/ secondary poisoning assessment. Relevant information on the mode of action of a chemical, both from environmental and toxicological data as well as a growing database of molecular biology (?omics) information can be used in a weight-of-evidence approach to determine the mode of action and develop tailored strategies for further testing. CBB or human bioequivalent dose approaches combining information on external exposure and dose response with toxicokinetic data provide better estimates of internal safe levels and also consider possible critical metabolites.

In conclusion the risk assessment of PBT/vPvB type chemicals needs to include higher tier evaluations and refinement. The additional studies that need to be carried out and any refinement options will need to be considered on a case-by-case basis and mayinclude:

· Measurement of half-live ranges in appropriate environmental media using improved biodegradation tests under appropriate environmentally relevant experimental conditions (e.g.temperature, duration, inoculum source and amount) and further evaluation of degradation pathways.

· Further modelling of the environmental distribution and overall multi-media fate.

· The use of physiologically-based pharmacokinetic (PBPK) modelling in appropriate species to support conclusions on uptake and bioaccumulation potential.

· Chronic effects testing on appropriate test species based on likelihood of exposure andsensitivity.

· Targeted environmental monitoring work in appropriate media (e.g. air, sewage effluent, river water, sediment, marine water and biota).

· Long-term monitoring programmes to investigate persistence and bioaccumulation potential in the field on a temporal and spatial scale.

Whilst there are a number of generic elements that can be included within such a risk assessment(e.g. high level of refinement in both exposure and effects assessment, the need for food chain assessment, monitoring of environmental concentrations along both a spatial and a temporal scale), the detail of the strategy that would be required to adequately reduce uncertainties associated with each chemical in question needs to be considered on a case by case basis. Each high tier assessment must take into account the specific properties and use scenarios for the chemical in question to address its specific concerns. The case studies in this report can be used as examples for the targeted application of refined methodologies for the risk assessment of PBT and vPvB properties. Further research is ongoing and other areas of research have been identified that will further increase the possibilities of reducing uncertainties in risk assessment of PBT and vPvB substances.

An appropriate risk assessment, incorporating the refinement options suggested within this report, would require significantly more data to be generated in a targeted manner. This would require a significant investment to be made in research, which may include an ongoing commitment (e.g.ongoing monitoring of environmental concentration on a spatial and temporal scale).


[1] RAIDAR (RiskAssessment IDentification And Ranking) (Arnot and Mackay, 2008) is a mass balance screening level steady-state model incorporating detailed emissions, chemical fate and effect considerations. It is capable of calculating both exposure assessment factors (EAFs) and Pov. RAIDAR estimates body burden of contaminants in a representative individual.

[2] TheTool (Wegmannetal, 2007) estimates the characteristic travel distance (CTD) of each substance in air. Body burden in the remote region is estimated by scaling the body burden in the source region by the fraction of the chemical that is transported to theremoteregion.