Analysis of currently available data for characterising the risk of engineered nanomaterials to the environment and human health--lessons learned from four case studies.

Production volumes and the use of engineered nanomaterials in many innovative products are continuously increasing, however little is known about their potential risk for the environment and human health. We have reviewed publicly available hazard and exposure data for both, the environment and human health and attempted to carry out a basic risk assessment appraisal for four types of nanomaterials: fullerenes, carbon nanotubes, metals, and metal oxides (ENRHES project 2009(1)). This paper presents a summary of the results of the basic environmental and human health risk assessments of these case studies, highlighting the cross cutting issues and conclusions about fate and behaviour, exposure, hazard and methodological considerations. The risk assessment methodology being the basis for our case studies was that of a regulatory risk assessment under REACH (ECHA, 2008(2)), with modifications to adapt to the limited available data. If possible, environmental no-effect concentrations and human no-effect levels were established from relevant studies by applying assessment factors in line with the REACH guidance and compared to available exposure data to discuss possible risks. When the data did not allow a quantitative assessment, the risk was assessed qualitatively, e.g. for the environment by evaluating the information in the literature to describe the potential to enter the environment and to reach the potential ecological targets. Results indicate that the main risk for the environment is expected from metals and metal oxides, especially for algae and Daphnia, due to exposure to both, particles and ions. The main risks for human health may arise from chronic occupational inhalation exposure, especially during the activities of high particle release and uncontrolled exposure. The information on consumer and environmental exposure of humans is too scarce to attempt a quantitative risk characterisation. It is recognised that the currently available database for both, hazard and exposure is limited and there are high uncertainties in any conclusion on a possible risk. The results should therefore not be used for any regulatory decision making. Likewise, it is recognised that the REACH guidance was developed without considering the specific behaviour and the mode of action of nanomaterials and further work in the generation of data but also in the development of methodologies is required.

Thresholds of toxicological concern for endocrine active substances in the aquatic environment.

The threshold of toxicological concern (TTC) concept proposes that an exposure threshold value can be derived for chemicals, below which no significant risk to human health or the environment is expected. This concept goes further than setting acceptable exposure levels for individual chemicals, because it attempts to set a de minimis value for chemicals, including those of unknown toxicity, by taking the chemical's structure or mode of action (MOA) into consideration. This study examines the use of the TTC concern concept for endocrine active substances (EAS) with an estrogenic MOA. A case study formed the basis for a workshop of regulatory, industry and academic scientists held to discuss the use of the TTC in aquatic environmental risk assessment. The feasibility and acceptability, general advantages and disadvantages, and the specific issues that need to be considered when applying the TTC concept for EAS in risk assessment were addressed. Issues surrounding the statistical approaches used to derive TTCs were also discussed. This study presents discussion points and consensus findings of the workshop.

Use of terrestrial model ecosystem data in environmental risk assessment for industrial chemicals, biocides and plant protection products in the EU.

Risk assessment approaches within the regulatory framework of the European Union (EU) based on single species tests were compared to those using data from terrestrial model ecosystems (TMEs). In a case study with the fungicide carbendazim, single species data led to ratios of the predicted environmental concentration (PEC) and predicted no effect concentration (PNEC) of above 1000, depending on available data and related assessment factors, indicating concern for the terrestrial environment. Considering the high degree of realism of the TME studies with multiple endpoints measured, but also residual uncertainty related to higher variability of endpoints, an assessment factor of 5 was applied on TME data. The most sensitive reliable endpoint was earthworm biomass. With the TME studies yielding slightly higher effect thresholds compared to laboratory data, and due to the lower assessment factor, the PEC/PNEC ratio was lowered to 5. This means that there would be concern for high application rates of carbendazim.