The European Food Safety Authority (EFSA) asked the Panel on Plant Protection Products and their Residues (PPR) to deliver a scientific Opinion on the identification of pesticides to be included in cumulative assessment groups (CAGs) based on their toxicological profile, the aim being to develop cumulative risk assessment (CRA) methodology.
This Opinion was preceded by two previous Opinions (EFSA, 2008, 2009). In the first one, the PPR Panel evaluated existing methodologies on cumulative risk assessment (CRA), and recommended that a tiered approach should be adopted both for hazard and exposure assessments. Criteria for grouping active substances into CAGs were proposed, based on the chemical structure, mechanism of pesticidal action, mode/mechanism of mammalian toxicity and common toxic effects. In the second Opinion an exercise was carried out to test the proposed approach by a worked example of a group of triazole pesticides, a well-defined group in terms of structure, pesticidal mode of action and toxicological effects (EFSA, 2009). Thus, the previous opinions dealt with CRA, encompassing both hazard and exposure assessment. In the Terms of Reference of the current opinion, EFSA has requested for a scientific Opinion on the identification of pesticides to be included in CAGs on the basis of their toxicological profile and deals therefore solely with hazard assessment. The Panel is aware that the conduct of CRA is a process that involves several steps and multiple considerations, many of which go beyond the scope and Terms of Reference of this opinion. CRA has to include the outcome of the current Opinion as well as other critical elements, such as the availability of occurrence data and the scientific and technical capacity of exposure assessment methodologies. Recommendations on the conduct of CRA were outside the scope of the present opinion.
The present Opinion presents a general methodology and criteria specifically developed for establishment of CAGs for pesticides. The methodology has been applied to establish CAGs for pesticides having effects on the thyroid and nervous system, and has been developed on the basis of datasets of oral toxicity studies evaluated in draft assessment reports (DARs). The methodology was developed in order to take cumulative effects into account in the decision on applications concerning maximum residue levels (MRLs) of pesticides in or on food and feed of plant and animal origin. The CAGs derived from this methodology could in principle be used to support CRA resulting from non-dietary exposures (i.e. operator, worker, bystander and resident exposure).
The allocation of pesticide active substances into CAGs requires a standardised and thorough review of the DARs for effects on individual organs and organ systems of all approved pesticides relevant for dietary exposure. Therefore, two preparatory projects for collecting toxicological data from pesticides were initiated. In the first project, all pesticides authorised prior to 31st of May 2009 were evaluated. The contractors proposed a grouping approach starting from identifying toxicological target organs and organ systems and then subsequently refining the grouping by identifying a specific phenomenological effect. If data allowed, the grouping was further refined by identifying a common mode or mechanism of action. The data collection and approach proposed by the contractor was scrutinised and partly consolidated by the PPR Working Group. It was decided that the data collection needed to be re-evaluated and, hence, a second project was launched specifically consolidating identified pesticides having effects on the nervous system, the liver and the reproductive and developmental system. In addition, pesticides approved from 31st of May 2009 until 1st of January 2012 were included in the scope of the second project.
The PPR Panel acknowledges that EU residue monitoring programmes indicate that there is some consumer exposure to residues of non-approved pesticides which should also be included in CAGs.
Following the work undertaken by the PPR Working Group for the current Opinion on reviewing pesticides for inclusion in various CAGs, it became apparent that there are often few or no data available on mode of action, but that many compounds affect the same target organ and/or cell population. On this basis, the proposed methodology follows a phenomenological approach based on organ or system toxicity, consisting in including in a CAG for a specific effect all pesticides causing this effect, even if the underlying mode of action (MoA) is unknown. Interactions (synergisms or antagonisms) are not expected to occur at the low exposure levels of residues that are observed in monitoring programs. Thus, the PPR Panel considers that mainly dose additive effects of substances are normally relevant to CAGs that may be used in the context of MRL setting (EFSA, 2008; Boobis et a., 2008).
As there may be limited opportunity for refinement of CAGs on the basis of available information on mode/mechanism of action, the proposed grouping methodology makes a sufficient precautionary approach, which is agreed upon by the European Commission and EFSA: when insufficient or no information is available, it is assumed that chemicals with the same effects may have a similar mode of action, even though they exhibit a wide range of chemical structural features. This view is based on empirical evidence that chemically unrelated substances may have a common effect in target organs/organ systems, which can be well approximated by dose addition (Kortenkamp et al., 2009). This has to be considered within the context of pesticide evaluations by EFSA and hence the approach recommended in the present Opinion differs from the approach tentatively used by the PPR Panel in its previous work.
The stepwise methodology for grouping has been elaborated to address acute and chronic dietary efects.
The methodology comprises four main steps as follows:
- Identification of the specific effects by:
i) exclusion of local effects
ii) exclusion of non-adverse effects
iii) exclusion of effects not relevant to humans
iv) evaluation of the unambiguous nature of the effect
v) identification of non-specific effects
- Characterisation of the specific effects
- Data collection
- Grouping of pesticides into CAGs
The PPR Panel recommends that the implementation of the methodology based on specific effects should be supported by expert judgement in order to identify the effects relevant for grouping according to the criteria laid down in the opinion. In particular, expert judgement is required to identify and characterise substances that can trigger different outcomes of the same toxicity pathway (e.g. different effects on motor division of the nervous system) or that may cause toxic effects at multiple sites by a single mode of action (e.g. acetylcholinesterase inhibition).
The CAG methodology in the current Opinion has been applied to the nervous system and the thyroid system.
For the identification and characterisation of the potential neurotoxicity of pesticide active substances, the functional divisions of the nervous system (motor, sensory and autonomic) along with the cognitive domain, neurochemistry and neuropathology parameters were considered as potential targets. Indicators of specific neurotoxic effects were identified and applied to characterise the CAGs for the nervous system.
A total of 68 active substances, were identified as having specific effects on the nervous system. Additional four substances were excluded from grouping because the methodological criteria were not met and/or the exposure to these substances by the oral route was highly unlikely following their authorised use.
The CAGs of substances identified as neurotoxic are presented in two separate tables for acute and chronic effects, respectively. Data were tabulated according to the level of organisation of the nervous system, the indicator of the specific neurotoxic effect, the active substance, its mode of action and the lowest NOAELs and/or LOAELs for each indicator. Non-specific or secondary effects, as well as effects that occur after administration of high doses, resulting in severe systemic toxicity, were not included in these CAGs according to the criteria for identification of specific effects listed above.
The following groups were proposed (number of pesticides in each group):
- Acute exposure (47)
- Motor division (45)
- Sensory division (20)
- Autonomic division (29)
- Motor division (53)
- Sensory division (21)
- Autonomic division (24)
- Neuropathological changes (19)
The Panel recognises that the neurochemical parameters, i.e. brain or erythrocyte AChE inhibition, represent a level of grouping for neurotoxic substances based on mechanism of action rather than on phenomenological effect. However, AChE inhibitors play a prominent role in the risk assessment that would result in an increased sensitivity for some substances. For this reason, and to keep consistency in the grouping approach, the neurochemical parameters should be used for further refinement when this mechanism of action is recognised. In addition, neuropathological changes were considered relevant only for chronic CAGs since some pesticide active substances induced morphological changes as the only adverse effect or they were found to be the most sensitive ones.
Despite the effects of pesticides on the cognitive domain e.g. learning and memory, which are relevant for assessment of neurotoxicity, the information available in the DARs failed to identify these effects. This is very likely because these effects correspond to a higher tier of assessment that was not performed on a routine basis during the toxicological assessment of pesticides.
Owing to the absence of systematic testing of pesticides for Developmental Neurotoxicity (DNT) in the European Union, and in consideration that new data requirements for active substances used in plant protection products have just been recently introduced (Regulation (EU) No 283/2013 of 1 March 2013), results from such tests, even when in certain instances available (e.g. for dimethoate, fenamiphos, fipronil, malathion and molinate), have not been considered for CAGs in the present opinion. Since the thyroid functions as a gland that produces systemically acting hormones (calcitonin, thyroxin (T4) and triiodothyronine (T3)), the most conservative level of grouping (CAG 1) was defined by effects occurring on the organ (thyroid) or organ system (hypothalamic-pituitary-thyroid axis), e.g. through changes in thyroid hormone levels (in total 101 of 287 screened substances were identified as affecting the thyroid or thyroid hormone systems). Identification of specific effects concerning two different thyroidal cell populations/hormone systems formed the basis for further refinement, yielding two sub-groups at the second level (CAG2A and CAG2B).
Substances affecting C-cells of the calcitonin system were allocated to CAG2A (22 substances). Owing to interrelationship of the specific effects between C-cell hyperplasia and neoplasms, and absence of information on underlying mechanisms, further sub-grouping of thyroid CAG2A was not possible.
Substances affecting the thyroid follicular cells and the T3/T4 system, i.e. displaying changes in circulating T3/T4 or TSH levels, follicular cell hypertrophy/hyperplasia or follicular cell neoplasia, were allocated to CAG2B (in total 96 substances). The specific effects that were used to define the CAG2B sub-group were apparently interrelated or connected to one another by a chain of events. While the precise mechanism of action is currently unknown for many substances within CAG2B, several different mechanisms of action are expected to contribute to a final deleterious common effect (i.e. decrease in T3/T4 action). For these reasons and based on the information available in DARs, further refinement of grouping is currently not possible. In exceptional cases, where there is convincing evidence for substance-dependent direct stimulation of the thyroid or hyperthyroidism, exclusion of substances from this sub-group might be considered.
The application of grouping methodology has yielded CAGs with sometimes large numbers of pesticides.The Panel notes that although some CAGs contain a large number of pesticides, little indication of cumulative risk may be inferred from the size of CAGs per se. The Panel further notes that, even within large CAGs, the majority of pesticides might not contribute significantly to a given combination effect, either because exposure is very low, and/or because potency in relation to the effect considered is weak. Instead, cumulative effects are likely to be driven mainly by a few active substances within the group.
Comprehensive preliminary work has been done on effects on the liver, adrenals, eye and developmental and reproductive system and provides a starting point for developing CAGs also for these systems in the future.
The PPR Panel identified a number of uncertainties and limitations in grouping of pesticides according to a common or shared toxic effect. In particular, a grouping based on toxic effects rather than on mode of action will lead to more uncertainties in predicting possible combination effects. However, the Panel acknowledges that when limiting CAGs to known common mode of action, thereby excluding pesticides for which information on mode of action is not available to enable their inclusion in relevant CAGs, the degree of uncertainty in CRA would also increase. Thus, a higher level of protection can be afforded by considering a wider range of pesticides and until information on precise modes of action becomes available, the cost of this is to use an effect-based approach that introduces some uncertainties around combination effects. Additional uncertainties considered by the Panel included the levels of details of the toxicological assessments in the DARs, changes occurring over the years in regard to data requirements and study protocols of the toxicological assessments, and inconsistency and variability in terminology of the DARs.
The PPR Panel also makes recommendations for the implementation of CAG grouping methodology in CRA to support MRL setting. The PPR Panel also notes that further refinement of grouping maybe achieved when data on the precise toxicological mode of action are available. However, information that justifies any deviation from dose-addition might also be necessary to consider for such a refinement. In addition, non-approved pesticides detected in food commodities should be included in CAGs, and a sound and consistent procedure for data retrieval should be developed for both the methodology and the inclusion of new substances into the relevant groups.