RISK ASSESSMENT OF COMBINED EXPOSURES TO MULTIPLE CHEMICALS:

A WHO/IPCS FRAMEWORK

1

2

3

4

5

EXAMPLE CASE-STUDY B: CARBAMATES

Contents 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

1. CONSIDERING THE NEED FOR A FRAMEWORK ANALYSIS .................................... 4 2. PURPOSE AND FOCUS OF THE ASSESSMENT ............................................................. 6 3. THE FRAMEWORK ANALYSIS ........................................................................................ 7

3.1 Tier 0 ................................................................................................................................. 7 3.1.1 Exposure assessment .................................................................................................. 7 3.1.2 Hazard assessment ...................................................................................................... 8 3.1.3 Tier 0 summary........................................................................................................... 9

3.2 Tier 1 ................................................................................................................................. 9 3.2.1 Exposure assessment .................................................................................................. 9 3.2.2 Hazard assessment .................................................................................................... 11 3.2.3 Tier 1 summary......................................................................................................... 13

3.3 Tier 2 ............................................................................................................................... 13 3.3.1 Exposure assessment ................................................................................................ 13 3.3.2 Hazard assessment .................................................................................................... 13 3.3.3 Tier 2 summary......................................................................................................... 13

3.4 Tier 3 ............................................................................................................................... 14 3.4.1 Exposure assessment ................................................................................................ 14 3.4.2 Hazard assessment .................................................................................................... 14 3.4.3 Tier 3 summary......................................................................................................... 15

4. CONCLUSIONS.................................................................................................................. 15 5. REFERENCES..................................................................................................................... 16 APPENDIX TO CASE-STUDY B: PERCENTAGE OF TOTAL CROP AREA TREATED WITH N-METHYL CARBAMATE INSECTICIDES FOR SELECTED CROPS IN 1992, 1997 AND 2002 ....................................................................................................................... 18

2

Note to readers 1

2 3 4 5 6 7 8 9

10 11

This case-study was prepared by Anna Lowit, United States Environmental Protection Agency, and Elizabeth Shipp, Bayer Cropscience on behalf of the European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC), and finalized by the Planning Group Chair, Bette Meek. It was prepared for the purpose of illustrating and testing the draft World Health Organization/International Programme on Chemical Safety (WHO/IPCS) framework on the risk assessment of combined exposures to multiple chemicals. It is presented here to facilitate the process of public comment.

3

This case-study addresses the N-methyl carbamate (NMC) insecticides as an example of a group of substances with a well understood mode of action and a relatively complete toxicity database. Additionally, they are used on a number of food crops, and the possibility may exist for dietary exposure to more than one NMC insecticide during one day. They are considered as priorities for a combined assessment, given the potential for co-exposure and their potency as acetylcholinesterase inhibitors.

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23

24

25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

Dietary exposure, however, is not the only route by which exposure to the NMC insecticides could occur. Exposure could also occur by ingestion of drinking-water contaminated with NMCs, by the occupational route during pesticide application or crop re-entry tasks, or by residential exposure following “home and garden” application of NMC insecticides. Owing to the high potency of the NMC insecticides and the possibility of exposure via multiple pathways, the United States Environmental Protection Agency (USEPA) has conducted an extensive risk assessment of combined exposure to NMC insecticides via multiple routes and pathways (USEPA, 2007e). This case-study briefly examines the toxicity of the NMC insecticides, their uses and potential exposures via the dietary route, and the assessment of the risk of dietary exposure to more than one of these substances as an example of potential tiers of consideration. The focus of the assessment is the risk of adverse health effects posed by potential combined exposure to multiple NMC insecticides via the dietary route of exposure (excluding drinking-water), given their use on multiple crops and detection in multiple agricultural commodities. 1. CONSIDERING THE NEED FOR A FRAMEWORK ANALYSIS The draft WHO/IPCS framework presents a series of questions designed to determine whether or not it is appropriate to group substances in an assessment group for a risk assessment. Those questions will be addressed here for the NMC insecticides. • What is the nature of exposure? Are the key components known? Are there data available

on the hazard of the mixture itself? There are several potential routes of exposure to NMCs. Dietary exposure is possible when carbamate insecticides are present on food or in drinking-water (used in food preparation); this exposure may be to only one carbamate insecticide or to multiple carbamates. Occupational exposure may occur during the application of an insecticidal formulation on crops or in a residential or other setting. Finally, residential exposure may occur through uses on lawns and gardens or on pets. The key components of this group are known, although their toxicity has largely been assessed on the basis of single compounds. • Is exposure unlikely or very low, taking into account the context? Exposure to at least one of the carbamates in the course of daily life is quite possible. The application of carbamate insecticides to all crops, including for both food and non-food uses, in the United States in 1992, 1997 and 2002 (CropLife America [CLA] database) is summarized in Table 1.

4

1 2 3 4

Table 1. Total quantity of carbamate insecticides applied to food and non-food crops

in the United States in 1992, 1997 and 2002.

Amount of active ingredient applied per year (lbs)aCompound 1992 1997 2002

Aldicarb 4 022 459 4 277 526 3 419 186 Carbaryl 4 452 694 4 857 299 2 985 342 Carbofuran 5 101 362 3 397 983 1 015 337 Formetanate hydrochloride 290 445 134 505 64 721 Methomyl 2 754 783 1 997 326 917 639 Oxamyl 945 828 938 782 747 693 Thiodicarb 1 705 515 821 242 224 663

a One pound (lb) = 0.45 kg. 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

Quantifiable residues of some of these NMCs have been measured in food monitoring programmes in the United States (USDA, 2007). Additionally, these insecticides are relatively stable in water, and thus it can be expected that some of the applied amounts will be present in drinking-water after either agricultural or residential application. As three of the NMCs are registered for residential uses, there is also the potential for residential or non-occupational exposure to NMC residues. However, given that these two potential pathways of exposure are less significant than exposure through dietary routes, they are not included in this illustration of the framework. • Is there a likelihood of co-exposure within a relevant timeframe? The CLA database lists a number of instances in which greater than the total area for a specific crop in a specific state was treated with carbamate insecticides (e.g. 144% of the total sweet corn crop in Alabama in 1992). It can be assumed that at least part of the area in each case was treated with more than one carbamate. A wide variety of fruit, vegetable and grain crops are treated each year with NMC insecticides. Given the high use of NMC insecticides, the large number of crops treated and the frequency with which NMC insecticides are detected in monitoring programmes, the possibility of exposure to multiple carbamate insecticides from multiple foods during one given day needs to be investigated. • What is the rationale for considering compounds in an assessment group? The NMC insecticides all act, in both target insects and mammals, by a rapidly reversible inhibition of the acetylcholinesterase enzyme. This inhibition leads to an accumulation of the neurotransmitter acetylcholine, rather than the usual degradation of acetylcholine shortly after its release into a nerve synapse. Accumulation of acetylcholine then leads to continuous stimulation of cholinergic receptors throughout both the central and peripheral nervous systems, leading to acute cholinergic toxicity. The NMC insecticides are not the only insecticides that inhibit acetylcholinesterase; rather, they share this same mode of action with the organophosphate insecticides, such as chlorpyrifos. However, the NMC insecticides cause rapid onset of effects as well as rapid recovery following inhibition, with maximal inhibition usually occurring between 15 and 45 min after exposure and recovery following within a matter of hours at most. The organophosphate insecticides inhibit acetylcholinesterase in a largely non-reversible manner

5

and thus do not share the same time course of recovery from inhibition. Owing to this difference, organophosphate and NMC insecticides have not been considered jointly in one common assessment group. Combined exposure to the organophosphate insecticides has been considered separately (USEPA, 2002, 2006). The NMC insecticides were established as an assessment group (in the USEPA terminology, a “common mechanism group”) by the USEPA in 2001 (USEPA, 2001), and the combined risk assessment subsequently prepared for the NMC insecticides (USEPA, 2007e) constitutes the basis for this document.

1 2 3 4 5 6 7 8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

24 25 26 27 28 29 30 31 32

2. PURPOSE AND FOCUS OF THE ASSESSMENT The NMC insecticides considered in this risk assessment have food uses or maximum residue limits (tolerances) in food crops in the United States, and thus there may be exposure to multiple NMCs via the dietary route. They have also been assessed by the USEPA as an assessment group (USEPA, 2007e), which constitutes the principal basis for the current evaluation. Figure 1 shows the WHO/IPCS framework for consideration of grouped compounds. It is important to note that only as many tiers as are necessary to set the compounds aside as non-priorities for further consideration or as a basis to inform subsequent risk management should normally be worked through; if crude estimates indicate that margins between estimated exposure and hazard are large (and protective), there is no need to go on to higher tiers and develop more refined data, such as Monte Carlo exposure estimates.

Yes, no further action required

No, continue

Input from exposure or hazard

assessments(iterative process)

Is the margin of exposure adequate

?

Increasing refinement of exposure models

Increasing refinement of hazard models (MOA)

Sample Tiered Exposure and Hazard Considerations

Mixture or Component Based

Tier 0

Simple semi-quantitative estimates of

exposure

Tier 1

Generic exposure scenarios using

conservative point estimates

Tier 2

Tier 3

Probabilistic Exposure Estimates

Tiered Exposure Assessments

Tier 0

Dose addition for all components

Tier 2

More refined potency (RPF) and grouping

based on MOA

Tier 3PBPK or BBDR; probabilistic

estimates of r isk

Tier 1

Refined potency based on individual POD, refinement of

POD

Tiered Hazard Assessments

Refined exposure assessment, increased use of actual measured

data

Figure 1: The framework for consideration of an assessment group, as proposed by WHO/IPCS (see text for details).

The current case-study, however, is based on a pre-existing risk assessment for an assessment group, rather than being an original work. Although not all agricultural or residential pesticides require detailed exposure assessments prior to registration or re-registration, such data were generated, in view of the toxicity of the NMCs and the potential for exposure to residues of multiple NMCs at one time. For this reason, the type and amount of data available

6

for this case-study for exposure, particularly, do not align with the tiers outlined in the WHO/IPCS framework. This evaluation is not meant to stand as an independent, original risk assessment of a specific group of substances, but rather as an example in application of the WHO/IPCS framework to an existing assessment in order to determine whether the framework is logical and workable. The uncertainty factors and benchmark doses (BMDs) are those presented in the USEPA assessment.

1 2 3 4 5 6 7

8 9

10 11 12 13 14 15 16 17 18 19 20 21 22

3. THE FRAMEWORK ANALYSIS A framework risk assessment on grouped compounds would be based on information required to deprioritize the compounds from further consideration or to inform risk management. In the current case of NMC insecticides, all of the tiers in the WHO/IPCS framework are examined regardless of whether or not they would have actually been required; this is done in order to provide an example of information that might be used. The compounds to be considered in this evaluation, listed in Table 2, were selected based on their membership in the NMC class of compounds and on the existence of maximum residue limits (tolerances) in food items and/or registered food uses in the United States.

Table 2. United States food use registration status of NMC insecticides evaluated in this framework analysis.

Compound Registration Tolerancea

Aldicarb Yes Carbaryl Yes Formetanate hydrochloride Yes Methomyl Yes Oxamyl Yes Pirimicarb Pending Thiodicarb Yes

a Existence of a tolerance is given only where there is no registered food use. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

3.1 Tier 0 3.1.1 Exposure assessment The Tier 0 exposure assessment, according to the WHO/IPCS framework, could rely on a simple, semiquantitative estimate of exposure. Such a screening-level evaluation could take the form of totalling usage of NMC insecticides in each state and for each crop for specific years. Selected data are presented in Tables 8–10 of the appendix to this case-study; the crops presented in those tables were chosen on the basis of frequent consumption, including by children, and on the basis of apparent treatment of more than 100% of the total crop area in at least one state and on at least one crop in each year examined (CPRI, 2006). Although not all of the NMC insecticides included in those tables are included in the USEPA’s combined exposure risk assessment on which the current document is based, there is sufficient overlap between the compounds included in this assessment and the compounds used to prepare the appendix to case-study B. As can be seen in Tables 8–10 in the appendix to this case-study, as well as in Table 1 above, the usage of NMC insecticides decreased in the 10-year period from 1992 to 2002. The

7

number of crops and states in which more than 100% of the crop was apparently treated with NMC insecticides decreased as well, with only one such exceedance (cucumbers in Illinois) evident in 2002. Nonetheless, the data available suggest that exposure to residues from a number of NMC insecticides on several dietary components could occur. For example, using the data from 2002 NMC crop applications, a diet in which one day’s meals contain apples from Washington state, cucumbers and grapes from New York state, green beans from North Carolina and sweet corn from South Carolina could expose the consumer to a number of different NMC residues (food items and states selected from Table 10). In order to ensure that such exposure does not pose undue risk, further exposure assessment is needed under higher tiers of the WHO/IPCS framework.

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

3.1.2 Hazard assessment At Tier 0 of the hazard assessment (see Figure 1), it could be assumed that only simple dose addition is required for all components of the assessment group under examination and that all members of the group share an equal potency. The data available on the NMCs are examined below to test that assumption. The NMC insecticides are known to share the same mode of action—namely, a rapidly reversible inhibition of the acetylcholinesterase enzyme. A recent paper from the USEPA (Padilla et al., 2007) demonstrated that when NMC insecticides are administered separately to rats, they share largely similar time courses for reversal of inhibition. Table 3 summarizes inhibition and reversibility data for the seven compounds investigated in that paper.

Table 3. Acetylcholinesterase inhibition characteristics of selected NMCs in rat brain and red blood cells after oral administration (Padilla et al., 2007).

Maximum % inhibition Time to 100% recovery (h) Compound Dose

(mg/kg bw)

Time to peak

effect (h) Brain Red blood

cells Brain Red blood

cells Carbaryl 30 0.5 40 10 2 24 Carbofuran 0.5 0.5 40 55 6 6–24 Formetanate 10 0.5 85 85 4–24 4–24 Methiocarb 25 0.5 60 75 4–24 4–24 Methomyl 3 0.5 53 67 4 4 Oxamyl 1 0.5 50 80 4 4 Propoxur 20 0.25 68 80 4 4

27 28 29 30 31 32 33 34 35 36 37 38 39 40

In general terms, it is clear that the NMCs tested here show similar characteristics in their time to peak inhibition of acetylcholinesterase and in the time to recovery of enzyme activity. However, comparison of the doses and their effects shows that the potency of the NMC insecticides is not equivalent and that significant brain inhibition (>50%) occurs at relatively low doses for several NMCs. For example, an oxamyl dose of 1 mg/kg bw produces 50% inhibition of brain cholinesterase at 30 min after administration, whereas a carbaryl dose of 30 mg/kg bw produces only 40% inhibition of brain cholinesterase at the same 30-min time point. Clearly, the potency of the different NMCs assessed in Table 3 (and presumably those examined in the current document, which were not studied by Padilla et al., 2007) varies with regard to maximum inhibition of acetylcholinesterase. While there are differences in potency of the various NMCs, should the margin between estimated total exposure to all and the potency of the most toxic NMC be large, additional assessment would not be required.

8

9

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

3.1.3 Tier 0 summary A recent comparative study (Padilla et al., 2007) showed that while the NMCs share a number of characteristics, including rapid onset and recovery of acetylcholinesterase inhibition in both brain and red blood cells, there are data on differences in potency that enable an additional tier of the hazard assessment. Simple analysis of NMC use on selected crops shows that there is the potential for exposure to multiple NMC insecticide residues on multiple dietary components, and thus further exposure assessment beyond Tier 0 is required. As there is the potential, after this first screening tier of the framework analysis, for combined exposure to levels of NMC insecticides that could exceed reference values, further assessment in Tier 1 and possibly above is required both to refine the exposure estimates and to compare them with more refined reference values/hazard values. 3.2 Tier 1 3.2.1 Exposure assessment At Tier 1, the exposure assessment could be based on a generic exposure scenario using conservative point estimates of exposure. With regard to NMC insecticides, this could include an evaluation of estimated consumption compared with the acute reference dose (ARfD) for each compound. The minimum and maximum consumption, where available, of each ARfD are listed in Table 4, as well as the derivation of the reference dose used. The data presented there show that while there is only low estimated exposure to some NMCs (pirimicarb, exposure to 7–10% of the ARfD is expected for all groups examined), some NMCs that are used with greater frequency and on more food crops are estimated to carry a greater likelihood of exposure (carbaryl, with exposure estimated to be 43–68% of the ARfD). These data were extracted from either the United States Reregistration Eligibility Documents (USEPA, 1994, 1998, 2007a,b,c,d) or the United States Tolerance Petition notices for each compound (specific documents and their online availability are listed in the references). The data in Table 4 show that it is possible for a varied diet to contain significant proportions of the ARfDs of compounds within the same assessment group; for example, children 1–6 years of age could consume 19% of the ARfD for aldicarb by eating citrus, sweet potato or potato, 81% of the ARfD for oxamyl by eating citrus, sweet potato, cucumber or apple, and 7% of the ARfD for pirimicarb through consumption of asparagus and leafy petiole crops such as celery. The ARfDs for aldicarb and oxamyl are both based on measured inhibition of acetylcholinesterase, whereas that for pirimicarb is based on clinical signs of neurotoxicity most likely related to acetylcholinesterase inhibition.

10

Table 4. Acute reference values for NMC insecticides with food use registrations or maximum residue limits/tolerances, and estimated consumption of those reference values.a

Compound Reference value

(mg/kg bw per day)

Source study End-point Safety factors

Age group % reference value

Aldicarb 0.001 Human volunteer AChE inhibition 100 Children, 1–6 years old 19 General population 43 Carbaryl 0.01 Rat developmental

neurotoxicology FOB changes 100

Children, 1–2 years old 68 Adults, 20–49 years old 16 Formetanate

hydrochloride 0.000 65 Comparative AChE study AChE inhibition 100

Infants 56 Infants <1 year old 27 Methomyl 0.02 Rabbit teratology Maternal/fetal

toxicity 300

Children, 1–6 years old 72 Oxamyl 0.001 Rat acute neurotoxicology AChE inhibition 100 Children, 1–6 years old 81

General population 10 Children, 1–2 years old 10

Pirimicarb 0.01 Rat neurotoxicology Clinical signs 1000

Children, 1–6 years old 7 Children, 1–6 years old 31 Thiodicarb 0.01 Rat teratology Body weight gain 1000 Infants 60

AChE, acetylcholinesterase; FOB, functional observational battery. a Where available, the minimum and maximum per cent consumption for a given compound are given to indicate the overall range.

3.2.2 Hazard assessment In Tier 1, the hazard assessment calls for refining the potency of the individual members of the assessment group, based on their individual points of departure, as well as refining those points of departure. Based on the limited data shown in Table 3 above, the acetylcholinesterase inhibitory potential of each of the NMCs varies, mostly in the maximum extent of inhibition and somewhat in the time to recovery of acetylcholinesterase activity. Although not all of the compounds tested are considered in the risk analysis (owing to lack of food uses or maximum residue limits/tolerances in food items) and not all compounds in the risk analysis were tested, it is safe to assume that this variation in potency extends to those compounds that were not tested by Padilla et al. (2007) (i.e. aldicarb, pirimicarb and thiodicarb). In order to derive relative potency factors, the inhibition of brain cholinesterase by each of the NMCs considered in the assessment (USEPA, 2007e) was determined. Benchmark dose estimation included a meta-analysis of data submitted by the registrants in combination with data generated by the USEPA for each of the substances listed, as well as published or other experimental data. The benchmark dose and half-life to recovery estimates (not provided here) were derived using a sophisticated dose–time–response exponential model. The limit of sensitivity for detecting cholinesterase inhibition in either brain or red blood cells is considered to be 10% based on an analysis by the USEPA (2002) using data from over 100 studies and more than 30 organophosphates. Thus, the central estimate of the BMD10 was selected for use in deriving relative potency factors. The BMD10, BMDL10 (lower 95% confidence limit on the BMD10) and relative potency factor for each of the members of the NMC group are presented in Table 5.

Table 5. Benchmark doses and relative potency factors for the NMC insecticides examined in this combined risk assessment.

Brain Compound

BMD10 (mg/kg bw) BMDL10 (mg/kg bw) RPF

Aldicarba 0.05 (female) 0.06 (male)

0.03 (female) 0.03 (male)

4

Carbaryl 1.60 (registrant female) 1.21 (registrant male) 5.46 (NHEERL male) 1.58 (combined male) 2.63 (Moser)

1.35 (registrant female) 0.99 (registrant male) 4.15 (NHEERL male) 1.11 (combined male) 2.03 (Moser)

0.15

Formetanate hydrochloride

0.11 0.06 2.18

Methiocarb 1.31 0.56 0.18 Methomyl 0.36 0.2677 0.67 Oxamyl 0.24 0.18 1.00 Pirimicarb 11.96 6.98 0.02 Thiodicarb 0.27 0.23 0.89

NHEERL, National Health and Environmental Effects Research Laboratory.

11

a The sulfone and sulfoxide metabolites of aldicarb are not listed in the current risk assessment, although they are represented in the higher-tier exposure assessments.

The modelling summarized here was conducted in order to cover dietary, residential and occupational uses of the NMCs. As residential and occupational exposures assessed by the USEPA’s (2007e) combined exposures risk assessment, in addition to exposures through food and water, may be by dermal and inhalation routes as well as by the oral route, the index chemical for relative potency factor derivation was selected on the basis of a robust database for acetylcholinesterase inhibition after administration by oral, dermal and inhalation routes. As the most robust database by all three routes belonged to oxamyl, that substance was selected as the index chemical. The data available for oxamyl included six acute oral studies, including one study in humans and a comparative cholinesterase study in adult and juvenile (postnatal day 11) rats, and covered a dose range in the rat studies from 0.005 to 15.3 mg/kg bw, thus providing a broad dose–response range. Although the current case-study does not cover non-dietary exposures, oxamyl will be retained as the index chemical for this case-study. The relative potency factors for each compound for children and adults, incorporating relevant uncertainty factors, are shown in Table 6.

Table 6. Relative potency factors and adjusted relative potency factors for children and adults for each of the NMC insecticides of interest.

Compound Oral RPF Adjusted RPF, children Adjusted RPF, adults Aldicarb 4 16 8 Carbaryl 0.15 2.7 1.5 Formetanate hydrochloride

2.18 44 22

Methiocarb 0.18 18 1.8 Methomyl 0.67 10 3.3 Oxamyl 1.00 10 3 Pirimicarb 0.02 2 0.2 Thiodicarb 0.89 89 8.9

Finally, the assumption called for in the preliminary (Tier 0) hazard assessment was that exposure to more than one NMC insecticide would produce an additive response in inhibition of acetylcholinesterase activity. As the potency of the various NMC insecticides tested was apparently different, this assumption was tested by conducting a mixture study in which the proportion of each compound within the mixture was based on its calculated potency. A dose–response curve was generated using overall doses of the mixture that were predicted to yield between 5% and 60% inhibition of brain and red blood cell acetylcholinesterase activity. The results obtained indicated the predicted additivity, and the inhibition at each dose was very similar to the predicted inhibition of acetylcholinesterase activity (data not shown). This work confirms simple additivity of effects after simultaneous exposure to more than one NMC, despite their differences in potency. The adjusted relative potency factors for each of the NMCs in this analysis (Table 6) were then used in a refinement of the initial exposure assessment.

12

Commonly, animal-to-human risk assessment for comparison of exposure with standard reference values uses default inter- and intraspecies safety factors of 10 or, alternatively, a margin of exposure between animals and humans of 100. As the interspecies factor is already included in the relative potency factor for each NMC (based on human data), only the intraspecies factor remains. Thus, in the exposure assessment, a margin of exposure of 10 (representing the intraspecies safety factor) or better is considered adequately protective of human health. 3.2.3 Tier 1 summary Simple additivity of the effects of NMC insecticides was confirmed through experimental investigation, and refined potency factors for each of the compounds of interest were developed. However, as there is the possibility for significant exposure to residues of multiple NMCs in one day, even if consumption of each individual compound is below the reference dose, the Tier 1 assessment does not provide a sufficient margin of exposure to end the evaluation at that tier. Further assessment of exposure is needed, although refinement of the hazard assessment may or may not be needed. 3.3 Tier 2 3.3.1 Exposure assessment At Tier 2, refinement of potential exposure could include deterministic estimates based on actual measured data. For the NMC insecticides or for any group of food-use pesticides, source data could include those available from the United States Department of Agriculture (USDA) Pesticide Data Program on detection of pesticide residues and metabolites in dietary components. In an original evaluation of combined exposures, this assessment would be conducted before moving on to probabilistic estimates of exposure (Tier 3). However, as an existing assessment is being used as a data source, the probabilistic exposure assessments generated in that assessment are used in this case-study. This should not be taken as an indication that such highly refined estimates of exposure are called for in all cases; this is simply a use of existing data in order to test the WHO/IPCS framework. 3.3.2 Hazard assessment Until the Tier 3 exposure assessment is tested against the refined potency estimates derived in Tier 1 of the hazard assessment, no further refinement of the potency or grouping based on mode of action will be conducted. Clearly, if the Tier 1 refinement of hazard assessment is not sufficient, additional refinements will be conducted. 3.3.3 Tier 2 summary Ordinarily, a complete tier of both exposure and hazard assessment would not be omitted. However, as the data on which this assessment is based began with very generic exposure estimates and moved directly to probabilistic assessment of dietary intake, an exposure assessment for this tier is not available. Likewise, hazard assessment would not be further refined until it is clear whether or not such refinement is needed.

13

3.4 Tier 3 3.4.1 Exposure assessment Food consumption data used in the combined risk assessment for NMCs were obtained from the USDA food consumption survey in which dietary recall data were collected from over 20 000 people ranging from infants less than 1 year old to adults over 50 years old. Actual residue information for each of the compounds was obtained from the USDA Pesticide Data Program (USDA, 2007), which analyses samples of domestic and imported food at distribution centres. Unlike the data used in Tier 0 or 1, the Pesticide Data Program data represent residue levels after the food item is processed for consumption, shortly before purchase by the consumer, and thus present a more realistic picture of actual exposure. Monte Carlo simulations for exposure to the NMC compounds were conducted for each age group using dietary consumption data and pesticide residue data from the above-mentioned USDA programmes. Residue concentrations for each compound were expressed as equivalents to the index chemical, rather than as residues of each specific compound. The margin of exposure for combined exposure to the NMCs was then determined at the 95th, 99th and 99.9th percentiles for each age group (Table 7). Table 7. Probabilistic analysis of dietary exposure to NMC insecticides and the

resulting margins of exposure.

95th percentile 99th percentile 99.9th percentile Age group mg/kg bw per

day

MOE mg/kg bw per day

MOE mg/kg bw per day

MOE

General population 0.0004 404 0.0023 79 0.0115 15 Infants <1 year 0.0005 342 0.0024 74 0.0106 16 Children 1–2 years 0.0013 141 0.0051 35 0.0229 7.9 Children 3–5 years 0.0010 185 0.0044 40 0.0209 8.6 Children 6–12 years 0.0006 323 0.0028 63 0.0145 12 Youth 13–19 years 0.0003 576 0.0017 106 0.0098 18 Adults 20–49 years 0.0001 1278 0.0008 236 0.0042 42 Adults 50+ years 0.0002 1035 0.0009 193 0.0044 40 Females 13–49 years 0.0004 505 0.0019 97 0.0101 17

Although predicted exposure of children between 1 and 5 years of age exceeds the margin of exposure of 10 at the 99.9th percentile level of protection, the margin of exposure is equal to 10 at the 99.8th percentile for these age groups. The margins of exposure included in Table 7 are considered by the USEPA to be adequately protective of human health. This was supported, in part, by extensive sensitivity analyses conducted on assumptions, data and models used in the risk assessment. 3.4.2 Hazard assessment As the combined exposure risk assessment passes with Tier 2 exposure assessment and Tier 1 hazard assessment data, there is no need to further refine the hazard assessment in either Tier 2 or 3.

14

3.4.3 Tier 3 summary Because the compounds examined here share a common mode of action but have differing potencies, an index chemical was selected, and potency factors for each of the other substances were derived relative to that index chemical using a highly refined meta-analysis benchmark dose approach. Probabilistic estimates of exposure were determined, and margins of exposure were determined for various age groups. This highly refined analysis demonstrates that there is no concern for combined dietary exposure to NMC insecticides. 4. CONCLUSIONS Refinement of the hazard assessment by deriving relative potency factors and interspecies assessment factors and of the exposure assessment by using realistic values for pesticide residues in food consumed results in the conclusion that combined dietary exposure to the NMC insecticides does not raise concerns for any age group. Based on this, further refinement of either hazard or exposure as provided for in the WHO/IPCS framework is unnecessary. However, such refinements could be conducted in other cases where combined dietary exposure to pesticide residues exceeds the acceptable margin of exposure. If the margins of exposure had not been adequate with the NMCs considered here, further refinements of the hazard assessment might include: • conducting comparative cholinesterase inhibition assays in juvenile and adult

animals; or • using pharmacologically based pharmacokinetic (PBPK) modelling, given the

rapid nature of onset of inhibition and rapid recovery, to better evaluate the dynamic nature of exposure to and effect of NMC insecticides.

The exposure assessment presented above addresses the highest tier. However, even the scenarios listed above could be refined if the margins of exposure were not sufficient. The probabilistic exposure models used in the exposure assessment totalled consumption of all foods during a 24-h period, rather than separating them into meals or eating events. With compounds that have a rapidly reversible biological effect, such as NMC-based acetylcholinesterase inhibition, it is possible that the inhibition from NMC consumption at one meal would be reversed before the subsequent meal and potential NMC consumption. Thus, evaluation of the contribution of each meal to the total daily NMC consumption might have been used to potentially reduce apparent exposure to NMCs; consumption at breakfast of sufficient NMC equivalents to the index chemical to cause 20% acetylcholinesterase inhibition followed by consumption at dinner of NMC equivalents to cause 30% inhibition differ in predicted hazard from consumption in one 24-h period of sufficient NMC equivalents to inhibit enzyme activity by 50%. To the extent possible with available models, the USEPA has evaluated the within-day exposures to NMCs through food. This analysis is beyond the scope of this case-study. In conclusion, use of the WHO/IPCS framework for combined exposure risk assessment of food-use pesticides is generally productive. The hazard assessment tiers are fairly easily adapted to the available data or to the data that can be generated specifically for the purpose of the combined exposure risk assessment. The exposure

15

assessment tiers are less neatly divided, as registration of an agrichemical pesticide generally requires fairly refined data, which may go as far as Tier 3. However, the overall approach is still applicable.

5. REFERENCES CPRI (2006). National Pesticide Use Database: 2002. Washington, DC, CropLife Foundation, Crop Protection Research Institute (http://www.croplifefoundation.org/cpri_npud2002.htm). Padilla S, Marshall RS, Hunter DL, Lowit A (2007) Time course of cholinesterase inhibition in adult rats treated acutely with carbaryl, carbofuran, formetanate, methomyl, methiocarb, oxamyl, or propoxur. Toxicology and Applied Pharmacology, 219: 202–209. USDA (2007) Pesticide Data Program: Annual summary, calendar year 2006. December 2007. Washington, DC, United States Department of Agriculture (http://www.ams.usda.gov/AMSv1.0/ams.fetchTemplateData.do?template=TemplateG&topNav=&leftNav=ScienceandLaboratories&page=PDPDownloadData/Reports&description=Download+PDP+Data/Reports&acct=pestcddataprg). USEPA (1994) Reregistration eligibility decision document: Methiocarb. Washington, DC, United States Environmental Protection Agency, 30 March (http://www.epa.gov/oppsrrd1/REDs/old_reds/methiocarb.pdf). USEPA (1998) Reregistration eligibility decision for methomyl. Washington, DC, United States Environmental Protection Agency, December (http://www.epa.gov/oppsrrd1/REDs/0028red.pdf). USEPA (2001) Memorandum from M. Mulkey to L. Rossi: Implementation of the determinations of a common mechanism of toxicity for N-methyl carbamate pesticides and for certain chloroacetanilide pesticides, dated 12 July 2001. Washington, DC, United States Environmental Protection Agency (http://www.epa.gov/oppfead1/cb/csb_page/updates/carbamate.pdf). USEPA (2002) Organophosphate pesticides: Revised cumulative risk assessment. Washington, DC, United States Environmental Protection Agency, June (http://www.epa.gov/pesticides/cumulative/rra-op/). USEPA (2006) Organophosphate pesticides (OP) cumulative assessment—2006 update. Washington, DC, United States Environmental Protection Agency, 31 July (http://www.epa.gov/pesticides/cumulative/2006-op/index.htm). USEPA (2007a) Reregistration eligibility decision for aldicarb. September 2007. Available online at http://www.epa.gov/pesticides/reregistration/REDs/aldicarb_red.pdf. USEPA (2007b) Reregistration eligibility decision for carbaryl. Washington, DC, United States Environmental Protection Agency, 24 September (http://www.epa.gov/pesticides/reregistration/carbaryl/).

16

USEPA (2007c) Reregistration eligibility decision for formetanate hydrochloride. Washington, DC, United States Environmental Protection Agency, 24 September (http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC5064786). Reports from other years are available following links from: http://www.epa.gov/pesticides/reregistration/formetanatehcl. USEPA (2007d) Reregistration eligibility decision for oxamyl. Washington, DC, United States Environmental Protection Agency, 24 September (http://www.epa.gov/pesticides/reregistration/oxamyl/). USEPA (2007e) Revised N-methyl carbamate cumulative risk assessment. Washington, DC, United States Environmental Protection Agency, Office of Pesticide Programs, 24 September (http://www.epa.gov/oppsrrd1/REDs/nmc_revised_cra.pdf).

17

Appendix to case-study B: Percentage of total crop area treated with N-methyl carbamate insecticides for selected crops in 1992, 1997 and 2002

Table 8. Percentage of total crop area treated with NMC insecticides for selected crops in 1992 (CPRI, 2006).

% of crop area treated State

Apples Cucumbers Grapes Green beans

Potatoes Sweet corn

Alabama 50 57 144 Arizona 71 Arkansas 29 60 79 California 49 72 27 142 19 41 Colorado 46 21 25 97 Connecticut 92 7 60 Delaware 95 136 100 78 79 Florida 99 40 70 173 Georgia 36 100 87 39 215 Idaho 75 62 2 97 Illinois 159 24 38 151 24 Indiana 75 24 38 152 3 Iowa 11 151 9 Kansas 46 Kentucky 26 34 18 Louisiana 15 114 Maine 56 11 27 Maryland 191 76 110 78 100 Massachusetts 28 46 7 25 58 Michigan 108 27 76 4 42 66 Minnesota 12 11 39 18 Mississippi 104 Missouri 7 60 76 21 70 New Hampshire 110 27 New Jersey 88 35 53 9 25 83 New York 118 46 53 8 25 118 North Carolina 71 38 87 112 97 110 North Dakota 81 Ohio 18 89 42 100 13 44 Oklahoma 40 35 54 Oregon 86 21 5 62 22 8 Pennsylvania 194 46 68 75 92 118 Rhode Island 37 25 58 South Carolina 94 34 87 70 83 115

18

% of crop area treated State Apples Cucumbers Grapes Green

beans Potatoes Sweet corn

South Dakota 23 Tennessee 26 34 100 18 110 Texas 46 164 35 54 Vermont 102 22 Virginia 114 37 87 75 95 93 Washington 93 21 50 62 8 11 West Virginia 99 93 Wisconsin 45 27 11 4 7

Table 9. Percentage of total crop area treated with NMC insecticides for

selected crops in 1997 (CPRI, 2006).

% of crop area treated State Apples Cucumbers Grapes Green

beans Potatoes Sweet

corn Alabama 18 85 Arizona 11 4 97 Arkansas 36 75 95 California 32 46 18 32 5 88 Colorado 50 80 4 10 38 Connecticut 28 15 62 Delaware 42 50 27 52 Florida 139 64 96 135 Georgia 19 47 80 18 106 Idaho 25 56 33 38 Illinois 95 130 15 20 38 Indiana 43 10 100 15 25 5 Iowa 59 10 51 Kansas 18 Kentucky 75 67 10 Louisiana 43 Maine 86 3 45 Maryland 111 24 60 51 77 Massachusetts 51 3 15 65 Michigan 89 13 85 3 27 38 Minnesota 21 24 6 Mississippi 43 Missouri 36 40 10 Montana 95 Nebraska 7 New Hampshire 110 59 New Jersey 67 58 33 99 1 76

19

% of crop area treated State Apples Cucumbers Grapes Green

beans Potatoes Sweet

corn New Mexico 12 New York 83 10 64 17 62 North Carolina 51 73 115 19 39 66 North Dakota 74 Ohio 42 116 91 3 25 38 Oklahoma 50 50 63 Oregon 119 21 4 56 62 38 Pennsylvania 111 32 69 60 11 62 Rhode Island 50 47 South Carolina 77 20 80 50 24 106 Tennessee 41 50 5 10 175 Texas 92 80 40 62 63 Utah 30 Vermont 110 25 Virginia 51 12 90 52 37 75 Washington 93 21 4 56 92 38 West Virginia 96 45 Wisconsin 36 13 4 24 3

Table 10. Percentage of total crop area treated with NMC insecticides for

selected crops in 2002 (CPRI, 2006).

% of crop area treated State Apples Cucumbers Grapes Green

beans Potatoes Sweet

corn Arkansas 39 75 20 California 30 2 2 5 2 2 Colorado 75 4 8 Connecticut 38 32 1 Delaware 15 20 20 Florida 2 1 Georgia 80 1 1 1 Idaho 38 Illinois 118 5 10 Indiana 15 100 10 10 5 Kansas 50 27 4 Kentucky 25 15 5 Maine 60 1 10 Maryland 5 Massachusetts 60 15 32 14 Michigan 39 17 58 1 2 6 Mississippi 12

20

% of crop area treated State Apples Cucumbers Grapes Green

beans Potatoes Sweet

corn Missouri 67 Nebraska 5 Nevada 5 New Hampshire 32 14 New Jersey 17 12 15 11 New York 63 2 56 1 1 1 North Carolina 8 13 80 47 5 2 Ohio 35 40 40 6 20 10 Oklahoma 3 Oregon 43 20 8 15 4 Pennsylvania 10 67 6 11 Rhode Island 20 South Carolina 1 20 11 Tennessee 50 25 10 27 Texas 1 Utah 50 Vermont 32 14 Virginia 70 90 11 Washington 67 4 47 West Virginia 15 Wisconsin 1

21