This technical report presents a systematic framework for the identification of emerging chemical risks occurring in the food and feed chain with a likely direct or indirect impact on human, animal and/or plant health within the EFSA’s remit. Such risks may arise from intentionally and non-intentionally produced industrial chemicals, as well as certain natural contaminants that may be transferred to the food/feed chain through the environment. The proposed framework uses a variety of data sources as an input, relating to, for example, industrial chemicals, environmental occurrence of chemical contaminants, as well as software models that can be used to predict the environmental behaviour and potential toxicity of chemical substances from their structural features and physico-chemical properties. The procedure consists of a multi-step selection process that starts with a list of chemicals to which a sequence of selection criteria is applied to identify the substances of potential concern. The selection criteria take into account a number of parameters such as volumes of production or import, persistence in the environment, potential for bioaccumulation, dispersive uses, toxicity, and any available outcomes of risk assessments. The procedure has two main entry points either for industrial chemicals registered under REACH Regulation or for substances consistently detected in the environment with a subset of more specific entry points depending on specific objectives and relevant data availability. The procedure proposed in this report needs to be tested and further developed using specific examples of chemical substances, preferably through a pilot project. The results of the pilot project should inform on additional activities that might be needed for further refinement of the proposed approach. This iterative process will provide a basis for a proactive approach to understanding emerging risks from materials, products, and applications of the continually evolving scientific and technological developments.
This technical report derives from a self-task mandate of EFSA and describes a framework applicable, in principle, for the identification of emerging chemical risks, as defined by Regulation (EC) No 178/2002, occurring in the food and feed chain with a likely direct or indirect impact on humans, animals, plants or any other organisms under EFSA’s competence, mainly associated with intentionally and non-intentionally manufactured industrial chemicals as well as with natural contaminants transferred to the food/feed chain through the environment. However, in view of the need for undergoing a validation step before undertaking a wider application of the proposed methodology, the present proposal is focused on available data sources/bases and criteria more relevant for the identification of emerging risks for humans.
The proposed procedure uses in a structured manner: (i) a variety of data sources that have recently become available on industrial chemicals produced in the EU, or on environmental occurrence of chemical contaminants; and (ii) software models for the prediction of the environmental behaviour, biological and toxic activity of specific chemicals from molecular structures and physico-chemical properties.
In general, the framework consists of a multi-step selection procedure that starts from a list of chemical substances (referred to as “entry point”) to which a sequence of selection (inclusion/exclusion) criteria is applied to identify the chemicals of potential concern in the present context. The selection criteria take into account volumes of production or import, persistence in the environment, bioaccumulation, dispersive uses, toxicity, and any available risk assessment.
The procedure is discussed in terms of: (i) main entry points in the selection procedure (e.g. list of chemicals to be screened, such as industrial chemicals registered under the REACH Regulation or chemical contaminants consistently found in the environment) with a subset of more specific entry points depending on particular objectives characterizing the application of the procedure and relevant data availability; (ii) several selection (inclusion/exclusion) criteria, including production volume, dispersive use, persistence, bioaccumulation, toxicity, evidence from existing Regulations or previous risk assessments; and (iii) selection process for the chemicals: multi-step procedure with a varying number of steps in which the outcome of each step becomes the entry point for the next step, and the last leads to the identification of emerging risks.
Some selection criteria are better applied through a deterministic approach (i.e. yes/no or an exclusion threshold), whereas for others a probabilistic approach (i.e. inclusion/exclusion range) can be applied. The proposed procedure may also be applied in an iterative manner in order to evaluate the impact of different threshold values adopted for specific selection criteria on the outcome of the procedure or to test different combinations of specific selection criteria.
The proposed procedure includes two main entry points - industrial chemicals registered under the REACH Regulation, and the non-intentionally produced or natural chemicals detected in different environmental compartments (e.g. water, soil and biota).
The first main entry point is the REACH Registered Substances Information of industrial chemicals produced or imported in the EU. The first two main procedural steps aim at selecting those chemicals which are produced in high volumes (first step) and used with dispersive modalities (second step). The third step of the procedure consists of parallel selection of: (i) high volume industrial chemicals characterised by highly dispersive use modalities, high persistence and tendency to bioaccumulate; and (ii) high volume industrial chemicals characterised by highly dispersive use modalities and high toxicity. The fourth step consists of a probabilistic combination of these two criteria and results in the selection of industrial chemicals characterised by highly dispersive use modalities, high persistence and tendency to bioaccumulation or high toxicity. The fifth procedural step is intended to exclude from the selected chemicals those that are already regulated as food contaminants, as undesirable substances in feed, or authorised or prohibited for specific uses in the food chain. The sixth procedural step, aiming at the exclusion of chemicals already assessed by EFSA and other scientific bodies, identifies chemicals classified as emerging issues (i.e. unregulated toxic chemicals likely to occur in the food chain). The seventh, and last, procedural step, consists of the selection of chemicals classified as constituting emerging risks according to the EFSA operational definition (i.e. toxic chemicals likely to occur in the food chain that have not been regulated in food/feed and have neither been evaluated by the European Commission or EFSA, nor authorised for use in food/feed).
The second main entry point aims at identifying chemicals, not included in the REACH register, using several different databases, including the Norman Network. This entry point includes, for example, chemicals of natural origin (e.g. mycotoxins, phytotoxins), or substances detected in specific environmental compartments (e.g. water, soil, sediments, biota or wildlife) which may be contaminants of the food/feed chain. After the exclusion of industrial chemicals included in the REACH Registered Substances Information, the previously-described procedure from step three to step seven applies.
In the two examples described, the exclusion criteria are applied in a deterministic manner, whereas some inclusion criteria (i.e. persistence/bioaccumulation and toxicity) are applied probabilistically and others (i.e. production volume and dispersive use) deterministically. Moreover, it is important to stress that the sequence of the selection criteria is intended to be flexible. Namely, to save time and resources, criteria that can be easily and quickly applied to large numbers of chemicals in the relevant context should be always used as soon as possible in the selection procedure.
The procedure proposed in this report needs to be tested and further developed, preferably through a pilot project. Testing this procedure starting from a specific list of chemicals will show whether the procedure works effectively, and it may also provide additional inputs for improvement and refinement. It is expected that a software program will be needed to manage efficiently the large amount of data in the different selection steps to test the procedure. Further specific suggestions are provided on how to streamline the pilot phase to refine the proposed approach, and to test it with specific examples. The results of the proposed pilot project should then be used to decide on additional activities for the further refinement of the proposed approach.