Dose Setting for Dermal Absorption Studies on Dried Foliar Residues

Abstract

Currently, the standard approach to estimate systemic exposure of workers after contact with dried pesticide residues on crops during re-entry activities relies on using the highest identified dermal absorption value for aqueous spray dilutions. However, recent dermal absorption studies with dried residues and their respective in-use dilutions have shown that this is likely to significantly overestimate their dermal penetration potential and, thus, predicted systemic exposure. The choice of appropriate dose levels for these dermal absorption studies has not been defined. Moreover, actual skin loading during re-entry tasks may differ significantly from that achieved by applying a fixed volume of an aqueous dilution, which is the standard practice in generic dermal absorption studies. To address this, we propose an approach to dose setting for dried residue studies within the current European risk assessment framework. Skin loading for studies can be calculated from the existing exposure algorithms and by taking appropriate body surface areas into account. Thus, skin loading in studies will vary depending on the exact nature and duration of the task and the region of the body exposed, reflecting actual exposure scenarios.

crop worker exposure, dermal absorption, dose setting, dried residues, exposure, pesticides, re-entry tasks, risk assessment

What’s important about this paper

The current estimates of systemic exposure of workers after contact with dried pesticide residues on crops during re-entry activities rely on using the highest identified dermal absorption value for aqueous spray dilutions. However, this is likely to significantly overestimate their dermal penetration potential and, thus, predicted systemic exposure. This paper describes a proposed approach to refine skin loading estimates that is based on existing exposure algorithms but accounts for body surface areas. The proposed approach fits within the current European risk assessment framework.

Introduction

Non-dietary exposure risk assessments for agrochemicals in Europe consider the dermal route as the primary exposure pathway for operators, bystanders, residents, and re-entry workers. Estimates of the external skin dose are then converted into systemic (internal) exposures by using a dermal absorption factor, which is then compared to hazard-based reference doses, e.g. Acceptable Operator Exposure Level (AOEL). The human dermal absorption factor/estimate is derived from in vitro studies according to Organization for Economic Cooperation and Development (OECD) test guideline 428 (OECD, 2004) conducted with either a pesticide formulation, which may be relevant for operators handling the concentrated product, or an aqueous in-use dilution for all other scenarios relevant for non-dietary exposure (EFSA, 2014). However, neither condition reflects the realistic scenario of an individual re-entering a field/orchard after pesticide treatment once the substance has dried and formed a residue of the diluted product on foliage. Despite this, risk assessments for dried residues currently still rely on using the highest tested dermal absorption value, expressed as a percentage of the applied dose, for aqueous dilutions, which is generally the most dilute solution (EFSA et al., 2017). However, this likely overestimates dry residue absorption, as dermal absorption studies reveal a significantly lower absorption for dry residues as compared to their respective in-use dilutions (Romonchuk and Bunge, 2006; Belsey et al., 2011; Clarke et al., 2015, 2018; Aggarwal et al., 2019).

Over the last few years, two methods to measure dermal absorption from dried residues have been developed (Belsey et al., 2011; Clarke et al., 2015, 2018; Aggarwal et al., 2019). The only difference to standard studies is the procedure for applying the test item to the skin; the studies are otherwise equivalent to dermal absorption studies according to OECD 428. Aggarwal et al. proposed to spray ¹⁴C-labelled substance onto a polytetrafluorethylene-coated septum, which is allowed to air-dry. To transfer the dose to skin mounted in a flow-through Franz diffusion chamber, the septum is placed gently onto the pre-wetted skin and gently rotated thrice. The Clarke et al. approach was similar, using a steel disc on which to prepare the dried deposit, achieving consistent levels of transfer to the skin membrane by attaching the disc to a weighted vial and moving it randomly on the membrane. Both methods achieve a high reproducibility of the transferred amounts. After this, a standard dermal absorption study is performed according to OECD TG 428 (OECD, 2004). Although these methodologies are not yet standardized or agreed in a regulatory context, they potentially allow a better dermal absorption estimate for risk assessment for those exposure groups where foliar residues are a key factor (re-entry for workers, bystanders, and residents).

One question which has remained largely unanswered is the choice of appropriate dose levels for dried residue dermal absorption studies to use in risk assessment. Aggarwal et al. (2019) aimed to achieve an equivalent dose per unit area (skin loading) to the representative in-use dilution to give a side-by-side comparison. While this is pragmatic and addressed the scientific question, those doses do not correspond to the exposure scenario. In the current manuscript, we describe and discuss (i) basic challenges in the current systemic exposure estimation when using dermal absorption studies; (ii) the currently used approach according to EFSA to estimate the risk during re-entry activities in recently treated crops; (iii) propose an alternative approach to dose setting for dried residue studies, with worked examples, which allows exposure-based testing as a refinement of the Aggarwal et al. 2019 approach and (iv) limitations of the suggested method.

Considerations for skin absorption of dry residues

The following summarizes some basic challenges in the current systemic exposure estimation by using standard dermal absorption studies.

Physical attributes of residue, duration, frequency, and site of exposure

A consideration not addressed in standard skin absorption assays is that the transfer of chemical residue onto and subsequently through the skin is dependent on the physical attributes of the residue formed (Belsey et al., 2011), where only a part of the residues is transferred onto the skin. The contact of the worker’s skin with the foliage can be brief or extended—depending on the re-entry activity—and whether repeated contact to the treated crops occurs. Therefore, it should be considered that there is a constant transfer but also rub-off of residues towards a steady-state until, e.g. hand washing for breaks or completion of work task. This exposure scenario is different from the study design of dermal absorption studies in which a single bolus dose is applied and its cumulative absorption over time is measured. One should also consider that the concentration at the site of action in the body but not the total dose absorbed is the relevant parameter for any biological effect—detoxification and excretion processes are ignored in the standard approach (Rozman and Klaassen, 1996; Hayes, 2014). These factors can lead to the estimation of exposure of pesticides by orders of magnitude higher than those resulting from actual re-entry to the crop (Ross et al., 2000).

Units of absorption µg/cm² versus % applied dose

Dermal penetration in terms of the amount in µg/cm² is (usually) proportional to skin loading and higher skin loading leads to a higher amount entering the systemic circulation. Currently, risk assessments of dried residues rely on using the highest tested dermal absorption value and only consider a relative dermal absorption in terms of % of the applied dose (EFSA et al., 2017). Therefore, when using the percentage approach, an inverse relationship between the concentration applied and the % absorbed is noted when comparing concentrates and spray dilutes. This can be largely misleading in terms of actual exposure, i.e. in µg/cm², when comparing relatively small but uncovered areas of the hand with high skin loadings in terms of µg/cm² but lower % of absorption in relation to large body areas protected by clothing, where the actual skin loading may be orders of magnitude lower. The higher relative but much lower absolute absorption for the rest of the body plays only an insignificant role in total systemic exposure per person. This aspect was taken into account when considering an appropriate and pragmatic approach on how to define the dried residue concentrations needed for dermal absorption testing. A worked example on the significant differences in actual exposure is in Supplementary Table 1, available at Annals of Occupational Hygiene online.

Dose setting

The currently used approach according to EFSA (and the US EPA) to estimate exposure during re-entry activities in recently treated crops follows a linear equation (US EPA, 1998; EFSA, 2014):

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{D}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{e}\mathrm{x}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{e}(\mu \mathrm{g})=}& \mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\\ & \times \mathrm{T}\mathrm{C}(\mathrm{c}{\mathrm{m}}^2/\mathrm{h})\times T(\mathrm{h})\end{array}}$$

(1)

The value assigned to dislodgeable foliar residues (DFR) is a default, which covers all crops and scenarios, scaled to reflect the application rate of the active substance. The determination of measured crop and substance-specific DFR values is recommended, because it leads to a more accurate dose calculation than assuming default values. Time (T), reflects the duration of exposure for a particular scenario (EFSA, 2014). The remaining parameter is the transfer coefficient (TC). This task-based value, which has a significant potential impact on dose reflects the amount of contact with treated foliage and the level of protection given to the worker by clothing (EFSA, 2014; EFSA et al., 2017).

The 2014 EFSA guidance on non-dietary exposure (EFSA, 2014) includes a relatively limited number of re-entry worker exposure scenarios with associated TCs drawn from EUROPOEM II (2002) and, to a lesser extent, US EPA interpretation of Agricultural Re-entry Task Force (ARTF) study data (US EPA, 2017). The scenarios include different crop types and tasks related to re-entry exposure to the dry residues and are summarized, together with their associated TC values, in Table 1.

Table 1.

Worker exposure scenarios included in the EFSA guidance (EFSA et al., 2017).

Crop	Task	Main body parts in contact	TC (cm2/h),
			Total potential exposure	Covered body (workwear) and gloves (PPE)	Assuming arms, body and legs covered (workwear; bare hands)
Vegetables*	Reach/pick	Hand and body	5800	580	2500
Tree fruits*	Search/reach/pick	Hand and body	22500	2250	4500
Ornamentals*	Cut/sort/bundle/ carry	Hand and body	14000	1400	5000
Strawberries*	Reach/ pick	Hand and forearm	5800	750	3000
Grapes**	Harvesting and other activities	Hand and body	30 000	−	10100
General**	Inspection/ irrigation	Hand and body	12 500	−	1400
			7500
Bystander (acute) and residents (longer-term)***	Toddler				2250
Golf course, turf, lawns*	Maintenance	Hand and body	5800	580	2500

Source of data: * EUROPOEM II; ** US EPA; *** EFSA derived values from US EPA data.

Equation (1) can be used as the basis for a standard dose algorithm in order to use exposure-based dermal doses in dermal absorption assays, within the exposure assessment framework of EFSA (2014).

According to the equation, the higher the exposure, the higher the dermal dose to be tested. However, a high dermal dose as a consequence of default assumptions would potentially lead to lower dermal absorption values when expressed as a % of the applied dose (see section ‘Units of absorption µg/cm² versus % applied dose’). Therefore, it is important to determine the parameters that are used to define the dermal dose are as realistic as possible. To address this, we aimed to make refinement of re-entry exposure to pesticides by adjusting the dermal absorption, while still respecting regulatory precautionary principles in the absence of data.

Further, in order to use the dermal dose from equation (1) in dermal absorption assays, it needs to be normalized to a unit dose/area, since the skin area available for dermal absorption studies is obviously different to the body surface of workers. Thus, the following equation can be used generically to the exposure-based doses for a dermal absorption study.

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{D}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=}& \mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\times T(\mathrm{h})\\ & \times \mathrm{T}\mathrm{C}(\mathrm{c}{\mathrm{m}}^2/\mathrm{h})/\mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{e}\mathrm{x}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{e}\mathrm{d}(\mathrm{c}{\mathrm{m}}^2)\end{array}}$$

(2)

One key observation from Table 2 is that the main body parts in contact with the crop are considered to be hands and body. The available re-entry exposure studies support this plausible a priori assumption also by data: hands are often substantially more exposed than the rest of the body (Damalas and Koutroubas, 2016). Therefore, we propose to calculate the concentration on the hands and the rest of the body separately to estimate the skin loading as accurate as possible. In the simplest instance, the total body surface area for an adult of 16 370 cm² would seem the relevant value to derive the dose level for dried residues studies (EFSA et al., 2017).

Table 2.

Summary of dose setting for single layer, no gloves.

Crop	Adjustment factor (dose = DFR (µg/cm2) × adjustment factor)			% of exposure to hands
	Hands only	Rest of body	Total
Vegetables (hand)	21	0.15	1.2	88
Vegetables (mechanical)	5.4	0.039	0.31	88
Tree fruits	24	1.0	2.2	56
Ornamentals	39	0.51	2.4	80
Strawberries	12	Not considered	12.3	100 (+forearms)
Grapes	59	2.1	4.9	60
General (inspection)	2.0	0.077	0.17	57
Golf course, turf, lawns	21	0.15	1.2	88
Child resident/bystander	1.4	0.053	0.18	57

Hence, according to Table 2, equation (2) and the consideration of the total body surface and EFSA (2014) default values, the exposure-based dose for a dermal absorption study for harvesting vegetables would be:

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{D}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=}& \mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\times 2500(\mathrm{c}{\mathrm{m}}^2/\mathrm{h})\\ & \times T(\mathrm{h})/16,370\mathrm{c}{\mathrm{m}}^2=3.67\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2\end{array}}$$

However, this would wrongly assume uniform distribution on the body and does not account for varying levels of protection afforded by clothing, e.g. impervious clothing like a Tyvek will mainly protect the body, while wearing nitrile-coated working gloves only reduces the exposure for hands. Therefore, each TC value needs to be considered independently, by accounting for the actual distribution on the body.

Exposure calculations and implications

Actual skin loading during re-entry tasks may differ significantly from that achieved by applying a fixed volume of an aqueous dilution. The suggestion has been made that skin loading can be calculated from the existing exposure algorithms, taking appropriate body surface areas into account (EFSA, 2014). Exposure will thus vary depending on the exact nature and duration of the task and the region of the body exposed. It will also be the result of incremental exposure during the day, rather than a one-off event. The choice of the appropriate TCs and exposed body parts could have a significant impact on skin loading and, as evidenced by a large body of data for liquids (Qiao et al., 1993; VanRooij et al., 1993; Wester et al., 1994; Lehman et al., 2011), this could drive the absorption estimates. As many products have extensive use patterns as part of good agricultural practice (GAPs) tables, there is the added dimension of considering multiple exposure scenarios. It is obvious that this is potentially a very complex area of exposure science which could have repercussions on the registrability of plant protection products and may require more sophisticated model approaches.

The following section describes the calculation of doses for different exposure scenarios, using vegetable harvesting scenarios as a worked example, and discusses the consequences of using different body exposure sites. All of the calculations are based on equation (2) in which the actual areas exposed are taken into account in relation to the TC values.

Equation (2) can be further simplified to

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{D}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{d}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=}& \mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\\ & \times \mathrm{a}\mathrm{d}\mathrm{j}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}\end{array}}$$

(3)

An adjustment factor is a unitless number derived by multiplying the appropriate TC (cm²/h) by the corresponding task duration (h) and dividing by the relevant body surface area (cm²), thus accounting for all parameters in the exposure equation with the exception of DFR.

Worked example—vegetables

For vegetable reach and pick scenarios, the TC of 2500 cm²/h for a single layer of workwear but no gloves are derived from hand exposure amounting to ~2200 cm²/h and actual body exposure of about 300 cm²/h. This (single layer, no gloves) is likely the most commonly considered scenario. For the hands, taking the relevant EFSA surface area of 820 cm² (EFSA, 2014), the dose per unit area could be calculated as:

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{D}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=}& \mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\times T(\mathrm{h})\\ & \times 2200(\mathrm{c}{\mathrm{m}}^2/\mathrm{h})/820(\mathrm{c}{\mathrm{m}}^2)\end{array}}$$

For hand-harvested vegetables (T = 8 h), this could be simplified to

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=\mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\times 21.46}\end{array}}$$

For mechanically harvested vegetables (T = 2 h):

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=\mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\times 5.37}\end{array}}$$

For the rest of the body, taking the EFSA surface area of 15 550 (16 370 – 820) cm² (EFSA, 2014), dose per unit area could be calculated to be: DFR (µg/cm²) × 0.039 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). It is noted that there is a difference of 139-fold between the dose predicted for the hands and for the rest of the body, which perceivably could have a significant impact on dermal absorption of dried residues relative to applied dose. Applying the dose calculated for the hands would underestimate absorption and exposure to the body. Conversely, using the dose calculated for the body as a worst case would significantly overestimate absorption for the hands, where the bulk of exposure is expected. It would be possible to carry out a study for both predicted doses and calculate exposures separately for hands and body, where additional refinement is needed.

Returning to the simplest approach of taking the total TC value and using total body surface area of 16 370 cm², the dose is calculated to be: DFR (µg/cm²) × 0.305 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). In this case, the predicted dose is 7.9-fold higher than the dose predicted by considering the body only and 17.6-fold lower than the dose predicted for the hands only. Although measuring absorption at this intermediate dose level may overestimate absorbed dose for the hands, which contribute 88% of the total dermal exposure, the approach may be considered a more reasonable compromise as a first tier than taking the highest dose and arguing that the hands are the most exposed body part.

The majority of the remaining EFSA re-entry scenarios (hand-harvesting grapes, tree fruit, ornamentals, strawberries and other low fruit) are dealt with in the same way as described for vegetables (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculations), since there is sufficient information on the TC values given in the EFSA guidance, i.e. the distribution of exposure across the body is transparent.

Crop inspection

The TC values for crop inspection come from two Agricultural Re-entry Exposure Task Force (ARTF) studies on dried peas and sweetcorn (ARTF website) and were derived by the UK Chemicals Regulation Division (CRD) (CRD HSE website). During the review of the studies, the CRD noted that 41% and 73% of actual dermal exposure was to the hands for sweetcorn and dried peas, respectively. The combined TC value for a single layer of workwear is the arithmetic mean of the parametric 75th percentile values from the respective studies and the arithmetic mean of actual dermal exposure to the hands is 57%. Taking the total TC value of 1400 cm²/h, the TC for the hands only would be ~798 cm²/h with the rest of the body accounting for 602 cm²/h. For the hands, taking the EFSA surface area of 820 cm² (EFSA, 2014), dose per unit area could be calculated as DFR (µg/cm²) × 1.95 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). For the rest of the body, taking the EFSA surface area of 15 550 (16 370 – 820) cm² (EFSA, 2014), dose per unit area could be calculated as DFR (µg/cm²) × 0.077 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). In this instance, there is a difference of 25-fold between the dose predicted for the hands and the rest of the body. Taking the total TC value and using total body surface area, the dose is calculated to be: DFR (µg/cm²) × 0.171 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). This compromise could be used as a first-tier approach for crop inspection and other general, short-duration activities in all crops.

Re-entry scenarios for bystanders and residents

The EFSA guidance includes scenarios for bystanders (acute, single event) and residents (longer-term) who may enter treated crops (EFSA, 2014). As there are no empirical exposure data for re-entry by members of the public, the TC values used are derived from those used in the assessment of exposure for re-entry workers carrying out crop inspection tasks (relatively low contact intensity). Furthermore, the same TC values are used for acute and longer-term assessments. Adjustments are made to reflect light clothing (t-shirt and shorts) and exposure for children, the latter by considering the surface area of a child to be 30% of that for an adult. The risk assessment for re-entry by children has become a significant challenge, often exceeding predicted exposures for the operator applying the product. This observation is likely more a result of a highly conservative risk assessment paradigm than an indication of genuine safety concerns and ways to refine this assessment to give a more realistic output should be sought. Finding the most appropriate dermal absorption values for the assessments is as important as it is for the professional worker. As the child resident and bystander always present worse case scenarios than the adult resident and bystander, the following proposal for dose setting focuses on the child, although the adjustment factors could also apply to adults.

As mentioned previously, the TC values for crop inspection which, in turn, provide TCs for re-entry by children come from studies on dried peas and sweetcorn, where the arithmetic mean of actual dermal exposure to the hands was 57%. The total TC value for child residents/bystanders is 2250 cm²/h. Using the above approximation of exposure to the hands of 57%, this would result in TC values for hands and rest of the body of 1283 cm²/h and 967 cm²/h, respectively. For the hands, taking the EFSA surface area of 230.4 cm² for the relevant 10 kg toddler (EFSA, 2014), the dose per unit area could be calculated as DFR (µg/cm²) × 1.39 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). For the rest of the body, taking the EFSA surface area of 4569.6 (4800 – 230.4) cm² (EFSA, 2014), dose per unit area could be calculated as DFR (µg/cm²) × 0.053 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation). In this instance, there is a difference of 26-fold between the dose predicted for the hands and the rest of the body. Taking the total TC value and using total body surface area, the dose is calculated to be: DFR (µg/cm²) × 0.117 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for the calculation).

Refined TC values for workers in vineyards

The TCs given by the EFSA guidance for workers involved in maintenance and harvesting activities in vineyards are based on a US ARTF study (ARTF website), which may not be reflective of modern European agronomic practice. Consequently, the European Crop Protection Association (ECPA), a cross-industry collaborative group, commissioned a series of studies to generate TCs to address a data GAP. These studies were carried out between 2004 and 2017 and include hand harvesting, pruning, training and shoot lifting in the Czech Republic, Germany, France and Italy. The studies and data were reviewed by the UK CRD on behalf of European regulatory agencies, with EFSA as an observer (HSE, 2020). The proposed TC values for workers wearing a single layer of clothing and no gloves were body—340 cm²/h, hands—3200 cm²/h, total 3500 cm²/h. Although these TC values have yet to be endorsed by EFSA, they are considered here in anticipation of their potential inclusion in the revised guidance. For the hands, considering TC: 3200 cm²/h, work rate: 8 h, and area exposed hands: 820 cm², a dose adjustment factor was calculated to be (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for calculation):

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{o}\mathrm{s}\mathrm{e}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)=\mathrm{D}\mathrm{F}\mathrm{R}(\mu \mathrm{g}/\mathrm{c}{\mathrm{m}}^2)\times 31.2}\end{array}}$$

For the rest of the body, taking the EFSA surface area of 15 550 (16 370 – 820) cm² (EFSA, 2014), dose per unit area could be calculated as DFR (µg/cm²) × 0.17 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for calculation, for the calculation). In this instance, there is a difference of 178-fold between the dose predicted for the hands and the rest of the body. Taking the total TC value and using total body surface area, the dose is calculated to be: DFR (µg/cm²) × 1.71 (see Supplementary Table 2, available at Annals of Occupational Hygiene online, for calculation, for the calculation).

Limitations and uncertainties of the refinement approach

As described in the Introduction, there are some issues with the dermal absorption approach used in current risk assessment. Two appear to be most relevant for the presented approach:

• (1) The dose used in the current dermal absorption studies does not correspond to an exact exposure scenario but to the test guideline requirement of using 1–5 mg/cm² of skin for a solid and up to 10 µl/cm² for liquids. A mixture of label use rate and gross exposure scenario selects the test item, the product concentrate or/and its in-use dilutions. Hence, the dermal absorption estimate can be considered to be more a product property than a value that achieves an exact systemic exposure estimation, which is however a pragmatic and conservative approximation.

• (2) In the current risk assessment approach, the sum of an estimated or measured cumulative external dose is simply converted into an internal dose by the dermal penetration factor, which is used as a percentage of the dose applied and is thus a relative value. This can lead to unreasonable relative dermal absorption estimates: when one of two molecules in the test dose in a dermal absorption study penetrate, the dermal absorption estimate is 50%—independent of the actual exposure scenario for the test item.

While we address point 1 in the current manuscript, in re-entry situations, point 2 might be a prominent issue—as the doses that would be tested can become very low. Point 2 can probably only be remedied by changing the current risk assessment framework.

Although the proposed method primarily addresses re-entry worker scenarios, other proposals have been made for dose setting where bystander or resident exposure re-entry assessments require refinement. Additional considerations may be necessary when considering re-entry exposure for products with complex GAPs, not covered in the generic approach presented in this manuscript. This is especially important when choosing appropriate field dilutions, where there are multiple different scenarios with widely differing dose levels. Careful selection of the critical uses would be required rather than testing for every scenario. It should be recognized that some of the adjustment factors for total TC values are low (e.g. 0.17 for crop inspection). For active substances with very low application rates, this could lead to very low calculated dose levels that may be difficult to achieve with standard dermal absorption testing methodologies. In this case, an alternative approach may be necessary.

In order to use the adjustment factors, a measure of DFR is necessary. In the simplest instance, this would be the default value of 3 µg/cm²/kg a.s./ha, corrected for application rate and taking into account repeated applications and decline in the dose between applications where necessary. The approach could also be applied to specifically generate DFR data. The default DFR value of ~3 µg/cm²/kg applied/ha is acknowledged as being very conservative but appropriately precautionary in the absence of measured data. Therefore, before conducting a dermal absorption study for dry residues, we propose to conduct a DFR study to measure the actual residues on the leaves at the most critical re-entry time with regard to exposure, i.e. on the day of the application after the spray has dried. Even when considering a default DFR value of 3 µg/cm² for the dermal absorption study, the final risk assessment would not result in lower exposure values when default DFR values would be revised with measured DFR values afterwards. According to the EFSA guidance on dermal absorption (EFSA et al., 2017) a pro-rata correction assuming a linear response is necessary, when the tested concentration is higher than the actual concentration. This considered to be a conservative approach.

With regard to the TC, it may be assumed in this approach that exposure is evenly distributed over the exposed area, be this hand or body, leading to a lower dermal dose to be tested in the dermal absorption study (i.e. higher relative dermal absorption estimate). In reality, the nature of re-entry tasks means that some regions of the exposed area, e.g. the palms of the hands, will be exposed more than others, leading to higher local doses. Such spot exposures are more related to liquid challenges (splashes). However, as sufficient data are lacking, the assumption of a homogenous exposure distribution and therefore a selection of a lower dermal dose is justified to fulfil precautionary principles.

A further aspect one may take into consideration which can add another level of safety is that dermal penetration varies also significantly per body part due to different barrier properties. As skin areas of most mechanical interaction, the hands and, in particular, the palms, have to be considered in comparison to the other body parts for re-entry exposure since they exhibit the thickest skin and stratum corneum (Maibach and Patrick, 2001; Monteiro-Riviere, 2004). The in vitro studies through human skin discussed in this proposal usually use skin from breast, abdomen, back or even the upper leg and therefore may overestimate the dermal absorption through the hands.

Summary

Two in vitro methods have been developed and described by Clarke et al. (2015, 2018) to determine dermal absorption from dried residues; however, for the results of these assays to be relevant for safety assessment, appropriate dose levels of dried residue need to be employed. To address this, we have developed a strategy to refine the dose per unit area (skin loading), based on existing exposure algorithms and taking appropriate body surface areas into account.

In the first tier, the dose level is the simplest approach based on the surface area of the whole body and assumes uniform distribution. This approach is likely to lead to a significant overestimate of absorption for the hands, which contributes at least 80% to the total exposure for six of the nine worker re-entry scenarios. The most appropriate dose might, therefore, seem to be that calculated only taking the area of the hands into account. However, this ignores the fact that total exposure is not received instantaneously, but is perceived to accumulate over the course of a working day of up to 8 h (probably up to an equilibrium (Ross et al., 2000)), at the end of which, the hands may be washed. Therefore, we suggest at least a two-dose approach: one to cover hands and one to cover the rest of the body. This is preferable to too much simplification which, in the end, has the tendency to be more conservative. The dose levels for dried residue dermal absorption experiments can be derived for different plant protection products use scenarios by using EFSA guidance assumptions (EFSA, 2014; EFSA et al., 2017). A summary of the adjustment factors for dose setting assuming single layer clothing and no gloves, is shown in Table 2.

In conclusion, we propose an exposure-based method for calculating dose levels for dermal absorption studies on dried foliar residues, which offers some flexibility to deal with multiple exposure scenarios.

Acknowledgements

Neil Morgan, Edgars Felkers and Christian Küster are members of the ECPA Occupational and Bystander Exposure Expert Group. Christiane Wiemann (Chair), Felix M. Kluxen and Edgars Felkers, are members of the ECPA team on dermal absorption. Nicola Hewitt is a scientific consultant. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The research was conducted as the scientific expert contribution of the authors to the above-mentioned industry association ECPA teams. The editorial work on the manuscript was partly sponsored by ECPA. The authors have no conflicts of interest.