                 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                            WASHINGTON, D.C. 20460
                                                                               
                                                     OFFICE OF CHEMICAL SAFETY 
                                                       AND POLLUTION PREVENTION



MEMORANDUM

DATE:	March 10, 2022

SUBJECT:	Difenoconazole.  Response to Comments on the Proposed Interim Decision.

PC Code:  128847
DP Barcode:  D464207 
Decision No.:  580572
Registration No.:  NA
Petition No.:  NA
Regulatory Action:  Registration Review
Risk Assessment Type:  NA
Case No.:  7014
TXR No.:  NA
CAS No.:  119446-68-3
MRID No.:  NA
40 CFR:  §180.475

FROM:	Bonnie Cropp-Kohlligian, Environmental Scientist 
	Minerva Mercado-Feliciano, Toxicologist
	Thurston Morton, Chemist
	Carolyn Mottley, Biologist
            Risk Assessment Branch IV
            Health Effects Division (7509P)

THROUGH:	Shalu Shelat, Branch Chief
            Risk Assessment Branch IV
            Health Effects Division (7509P)

TO:		Lauren Weissenborn, Chemical Review Manager (CRM)
            Khue Nguyen, Team Leader
            Cathryn Britton, Branch Chief
            Risk Management and Implementation Branch V
      Pesticide Re-evaluation Division (7508P)


The Pesticide Re-Evaluation Division (PRD) of the Office of Pesticide Programs (OPP) has requested that the Health Effects Division (HED) respond to comments on the difenoconazole Proposed Interim Decision (PID, EPA-HQ-OPP-2015-0401) related to human health concerns that were received from the Center for Food Safety. The HED thanks the Center for Food Safety for the comments and provides responses to those comments in this memo. 
Comment regarding "Liver Toxicity Endpoint and Chronic Reference Dose"

Comment (verbatim):  "The major target organ of difenoconazole is the liver, as demonstrated in both chronic and subchronic mouse and rat feeding trials conducted by registrants. Adverse hepatic effects included hepatocellular hypertrophy, hepatic vacuolation, cell necrosis, increased liver weight, fatty changes, bile stasis, increased serum levels of enzymes indicative of liver injury  -  alanine aminotransferase (ALA), serum alkaline phosphatase (SAP), and sorbitol dehydrogenase (SDH)  -  and an increased albumin/globulin ratio (EPA 9/18/20, pp. 5-6, 19; EPA 7/27/94).

In 1994, EPA established a chronic reference dose of 0.01 mg/kg/day based on hepatoxicity, in particular hepatocellular hypertrophy, in males in a chronic rat study (NOAEL = 0.96 mg/kg/day) (EPA 2/24/94; 7/27/94, p. 18). By 2015, EPA had retained the same chronic reference dose, based on the same rat study, but changed the endpoint from hepatocellular hypertrophy to cumulative decreases in body weight gains in both sexes (EPA 2/24/15). In EPA's latest human health assessment, the proposed chronic reference dose has been increased five-fold to 0.05 mg/kg/day, and based on a mouse rather than rat study, in which liver lesions were observed in male mice and hepatocellular hypertrophy in both sexes of mice, with an NOAEL of 4.7 mg/kg/day (EPA 9/18/20).  

EPA's dismissal of hepatocellular hypertrophy in male rats and the associated NOAEL and reference dose (0.96 and 0.01 mg/kg/day, respectively) in favor of the five-fold higher mouse study endpoint is incorrect and should be reversed. First, hepatocellular hypertrophy in rats is properly regarded as an adverse effect when accompanied by other adverse liver effects, which is the case here: the rats in this chronic study also exhibited increased liver weight and an increased albumin/globulin ratio (EPA 7/27/94, p. 18). Second, subchronic (13-week) rat and mouse studies also demonstrated low-dose adverse hepatic effects. The mouse study's effects included histopathologic alterations in the liver at an LOEL of 30 mg/kg/day and NOEL of 3 mg/kg/day. Rats in the subchronic study exhibited increased liver weights at the LOEL of 15.5 mg/kg/day (F), with an NOEL of 1.43 mg/kg/day (F) (EPA 7/27/94, pp. 12-13).

By illegitimately dismissing the chronic rat NOAEL and reference dose (0.96 and 0.01 mg/kg/day), and by ignoring the subchronic rat and mouse NOEL's, all three of which are below EPA's new chronic mouse study endpoint, EPA enables 5-fold greater chronic exposure to this hepatotoxic fungicide.

We would also note that the European Food Safety Authority continues to base its NOAEL of 1 mg/kg/day on hepatotoxicity (increased incidence and severity of hepatocellular hypertrophy), as EPA once did (EFSA 2009, p. 109)."

HED Response:  As correctly stated by the Center for Food Safety, the 2020 difenoconazole draft human health risk assessment (DRA; B. Cropp-Kohlligian et al., 9/18/2020, D457325) used the NOAEL from the carcinogenicity mouse study (4.7 mg/kg/day) to derive a chronic population adjusted dose (cPAD). That NOAEL is based on increased incidence of liver lesions (individual cell necrosis and bile stasis in males, hepatocyte hypertrophy in both sexes), and increased serum levels of serum sorbitol dehydrogenase (SDH) in males at a LOAEL of 46 mg/kg/day. The previous risk assessment (T. Morton et al., 2/04/2020, D455964) indicated that a NOAEL from a combined chronic/carcinogenicity (0.96 mg/kg/day) was used to derive the cPAD, based on cumulative decreases in body weight gains at a LOAEL of 24.4 mg/kg/day. Several risk assessments for difenoconazole completed between 1998 and 2020 used the same endpoint. However, the difenoconazole toxicology database underwent extensive review for registration review in late 2020, and most studies were updated to reflect current toxicology evaluation practices. Changes in study LOAELs at that time prompted changes to the points of departure (PODs) selected for risk assessment.

During the 2020 review, it was established that the previous review (K. Whitby, 8/26/1992, TXR 0009689) for the rat combined chronic toxicity/carcinogenicity study determined the lowest observed effect level (LOEL) for the study, not the lowest adverse effect level (LOAEL). However, current HED policy is to use only adverse effects, which are indicated by a LOAEL (not by a LOEL) as basis for risk assessment. The previous review (i.e., 1992) based the LOEL of 24 mg/kg/day on reductions in cumulative body weight gains. The current review (i.e., 2020) bases the LOAEL of 124 mg/kg/day on decreased mean absolute body weight (22-23%) in females, as well as decreased platelet counts (18-24%) and increased albumin/globulin ratio (14-48%) in males. Also, at 124 mg/kg/day, mean relative liver weights were increased in females at 52 and 104 weeks, and in males at 52 weeks. However, mean absolute liver weights were not significantly different from the control at any dose. At the terminal (104 weeks) sacrifice, an increase in the incidence of liver hypertrophy was observed at 24 (65% males, 34% females) and 124 mg/kg/day (89% males, 84% females) compared to controls (18% males, 13% females). In the absence of more severe liver findings (see mouse study discussion below) or associated clinical chemistry findings, the changes in the liver are considered adaptive. 

The mouse carcinogenicity study was also reviewed and updated in 2020. In contrast to the adaptive effects in the rat, the mouse showed individual cell necrosis and bile stasis in males, hepatocyte hypertrophy in both sexes, and increased serum levels of SDH in males at a LOAEL of 46 mg/kg/day. Cell necrosis and bile stasis are adverse effects in the liver. In the case of mice, the hepatocyte hypertrophy is also considered adverse because it occurs in the presence of increased serum SDH, cell necrosis and bile stasis. The mouse study had a lower LOAEL than the rat study and was, therefore, selected by HED to derive the chronic PAD.

HED acknowledges older reviews of the rat combination chronic toxicity/carcinogenicity study mentioned by the Center for Food Safety in their comments: Carcinogenicity Peer Review of Difenoconazole (J. Rowland and E. Rinde, 7/27/1994, TXR 0011150), and RfD/Peer Review Report of Difenoconazole (G. Ghali, 2/24/1994, TXR 0051309). These older reviews set a LOEL of 24.1 mg/kg/day based on increased incidence of hepatocytic hypertrophy in both sexes. The reviews also indicated that the albumin/globulin ratio in blood was increased in males only, and only at a dose of 124 mg/kg/day. These findings were confirmed during Registration Review and provided further evidence for the current LOAEL of 124 mg/kg/day on decreased mean absolute body weight (22-23%) in females, as well as decreased platelet counts (18-24%) and increased albumin/globulin ratio (14-48%) in males.



Comment regarding "Carcinogenicity"

Comment (verbatim):  "EPA originally classified difenoconazole as a Group C Possible Human Carcinogen in 1994, based on clear inducement of hepatocellular adenomas and carcinomas in a mouse study (EPA 7/27/94), then subsequently re-classified it under the descriptor Suggestive Evidence of Carcinogenicity (EPA 9/18/20, p. 6)."

HED Response:  The Center for Food Safety correctly indicates that "EPA originally classified difenoconazole as a Group C Possible Human Carcinogen in 1994, based on clear inducement of hepatocellular adenomas and carcinomas in a mouse study (EPA 7/27/94), then subsequently re-classified it under the descriptor Suggestive Evidence of Carcinogenicity (EPA 9/18/20, p. 6)." In 1994, the HED Carcinogenicity Peer Review Committee (CPRC) concluded that difenoconazole should be classified as Group C (possible human carcinogen) and recommended that for the purpose of risk characterization, the Margin of Exposure (MOE) approach should be used for quantification of human risk (J. Rowland and E. Rinde, 7/27/1994, TXR 0011150). Later in 2007, and in accordance with EPA's 2005 Guidelines for Carcinogenic Risk Assessment, difenoconazole was re-classified as "Suggestive Evidence of Carcinogenic Potential" based on liver tumors in male and female mice (P. Shah, 3/01/2007, TXR 0054532, D318039). The classification is consistent with current Agency guidelines.

Comment regarding "Toxicity of Metabolite Unknown"

Comment (verbatim):  "EPA has identified a major metabolite of difenoconazole that is present in humans, livestock and fish  -  CGA-205375  -  yet has practically no toxicological information on it (EPA 9/18/20). EPA should demand toxicity studies on this and other metabolites rather than rely on guesstimates based on unreliable, in silica structure-activity modeling."

HED Response:  HED has reviewed data that indicate that CGA-205375 is a residue of concern in livestock commodities and drinking water. HED uses structure-activity-relationship (SAR) analyses, as appropriate, to support decisions on residues of concern in situations where empirical data are lacking, and/or, to trigger the need for additional toxicology studies. In the case of difenoconazole, the SAR analysis of the metabolites using DEREK v.12 did not indicate any concerns for toxic effects that were not observed in the available difenoconazole toxicity database (W. Irwin, 9/19/2011, D391350).

Comment regarding "Dermal Absorption"

Comment (verbatim):  "An in vivo dermal absorption study in rats found dermal absorption of 48% of the applied dose after 24 hours in rats exposed to 0.5 ug/cm[2] of difenoconazole (EPA 9/18/20, p. 18). Rather than use this value as the dermal absorption factor, EPA reduced it to 6% based on two in vitro dermal absorption tests, one with human and one with rat skin. EPA multiplied the ratio of the in vitro absorption results (human/rat = 0.12) by the in vivo rat result of 48% to arrive at a human dermal absorption factor of 6%. There are multiple flaws that invalidate EPA's 6% estimate, utilizing the so-called "triple-pack" approach.

First, the relevant EPA regulations for the dermal penetration assay (40 CFS, Part 158.500, Guideline No. 870.7600) prescribe an in vivo rat study, and provide no support for EPA's manipulation of this figure by applying in vitro results as EPA implies (EPA 9/18/20, pp 49-50, citing Guideline No. 870.7600 for in vitro as well as in vivo tests). Guideline No. 870.7600 (see EPA 1998) details a test protocol involving live rats to determine dermal absorption, not in vitro tests, much less the use of in vitro tests to reduce the dermal absorption factor derived from the in vivo rat study.

Second, even if one accepts the triple pack method as acceptable in principle, its use was entirely inappropriate with difenoconazole because the in vitro human/rat dermal absorption ratio was derived from tests employing far higher doses (10, 100, 1000 ug/cm[2]) than the in vivo study (0.5, 1.3, 2.5 ug/cm2), rendering them incompatible, especially given the large differences in absorption that were observed as a function of dermal dose (EPA 9/18/20, pp. 49-50). Indeed, the 20-40-fold difference in dosage between the in vivo and in vitro tests violated EPA's precondition that the protocols and doses be the same in all three studies in order to utilize the triple pack approach (see EPA 10/29/13, dermal absorption of glufosinate; see also EPA 6/2/10, pp. 2-3, dermal absorption of thiabendazole, likewise citing "identical protocols in both the in vivo and in vitro studies" as a precondition for use of the triple pack approach). The difenoconazole tests also failed a second EPA criterion for application of the triple pack approach  -  that the dermal absorption factors from the in vitro and in vivo animal tests be roughly equal (i.e. their ratio approximately equal to 1) (see EPA 6/2/10; see also EPA 10/29/13, where EPA rejects the triple pack method for estimating dermal absorption of glufosinate because the rat in vitro and rat in vivo results diverged substantially, among other reasons). In the case of difenoconazole, the in vivo and in vitro rat absorption factors from the tests carried out at the same doses differed by 3- to 4-fold (10 and 100 ug/cm[2]), violating this criterion as well. 

Finally, the test substance used in these assays was not specified, and if it is the technical active ingredient, this would likely lead to an underestimate of dermal absorption relative to use of real-world formulations with absorption-enhancing surfactants. Even use of a particular difenoconazole formulation in this test would not be predictive of absorption with other formulations.

EPA should demand full dermal absorption data for various difenoconazole formulations. Until then, it should conduct residential and occupational exposure assessments that incorporate dermal absorption based on a dermal absorption factor of 48%, based on a test that most closely follows the EPA (1998) protocol, rather than the 6% from illegitimate use of the triple-pack approach."

HED Response: As indicated by the Code of Federal Regulations (CFR) Title 40 - Part §158.500, a dermal penetration study (Guideline No. 870.7600) is conditionally required. However, the general provisions in Part §158 indicates that these are minimum requirements [§158.1(b)(1)]; that EPA has the flexibility to establish or modify data needs for individual pesticide chemicals [§158.30(a)]; and that EPA may require submission of additional data beyond that specified in this part [§158.30(b)]. The dermal penetration study guideline (870.7600) of 1998 must be followed for in vivo studies in rodents, however it does not preclude use of other kinds of studies to derive a dermal absorption factor (DAF). EPA will also use skin absorption in vitro studies that follow the Organization for Economic Cooperation and Development (OECD) guideline number 428. In some cases, EPA will also use studies from the public scientific literature, or studies conducted under special protocols, if they are judged by EPA to follow sound scientific methodology and provide data that are applicable to risk assessment. Use of the triple pack approach is supported by EPA, when appropriate, to refine dermal penetration results by accounting for differences between in vitro and in vivo absorption within a test species (typically rat) as well as for species differences observed between the animal and human skin. Previously, EPA, along with other North American countries, did not accept in vitro studies alone for deriving DAFs out of an abundance of caution due to methodological concerns at the time (NAFTA 2009); however, there have been significant improvements in the in vitro methods over the past decade. Furthermore, EPA recently completed retrospective analyses of dermal triple pack data, which demonstrated that the in vitro studies alone provide similar or more protective estimates of dermal absorption, with only limited exceptions (Allen et al., 2021). As a result, EPA has shifted its use of in vitro studies and will now consider in vitro studies alone conducted in accordance with OECD 428 guidelines for DAF derivation.

During registration review, HED used a DAF of 6% that was applied in previous human health risk assessments for difenoconazole. The highest in vivo rat absorption from the available in vivo rat study using the SCORE[(R)] 250 EC end-use product (48%), an emulsifiable concentrate (EC) formulation, was multiplied by the highest ratio of human vs. rat absorption (0.12) from the available in vitro studies using technical difenoconazole. We appreciate the Center for Food Safety bringing to our attention that this DAF was calculated using a combination of the in vivo data using a formulation with in vitro data using the technical product. The resulting DAF was considered a conservative estimate of absorption by human skin for difenoconazole at the time it was derived; however, the triple pack should be applied when similar protocols are utilized across the in vitro and in vivo studies, including the same test substance and similar dosing. 

HED has reevaluated the DAF for difenoconazole and, in addition to the data previously considered, in vitro studies (MRID 46950332) with radiolabeled difenoconazole formulated as the SCORE[(R)] 250 EC end-use product were also considered. The SCORE[(R)] 250 EC end-use product is not registered in the United States; however, it is essentially the same as a registered end-use product, Inspire[(R)] (EPA Registration #100-1262) in terms of ingredients, formulation type, and the proportion of ingredients. The studies with the SCORE[(R)] 250 EC end-use product also tested dilutions that are applicable to anticipated exposures of difenoconazole based on the registered and proposed uses. Therefore, EPA considers these studies appropriate for evaluating dermal penetration potential of difenoconazole formulations registered in the United States. Furthermore, the data for this formulation indicate higher dermal penetration potential as compared to the data for the technical, and therefore a DAF based on the formulation data is considered a more health protective estimate of dermal penetration for difenoconazole.

Based on all the available data and current practices, HED has concluded that a DAF of 8% should be applied for difenoconazole moving forward. This DAF is derived from the amount absorbed in the in vitro SCORE[(R)] 250 EC end-use product study (MRID 46950332) using human skin at 24 hours (with 6 hours of exposure) to lowest concentration tested (0.5 ug/cm2 of difenoconazole). The amount measured in the skin membrane was not included in the DAF calculation because the data indicates that the skin does not act as a reservoir for continued exposure to difenoconazole after termination of exposure (i.e., difenoconazole in the skin is not bioavailable for further skin absorption after washing).

The updated DAF of 8% does not materially impact the human health risk assessment. A dermal POD has not been selected for difenoconazole. The database does not show systemic effects after exposure via the dermal route at doses that would be relevant to risk assessment. A route-specific study in rabbits showed liver effects of minimal to moderate severity at the limit dose (1000 mg/kg/day), which are considered an inflection point in the dose-response curve. Applying the 8% DAF to oral short- and intermediate-term studies in the database yields LOAELs above the limit dose, which are not relevant to risk assessment. There is no concern for increased in utero or postnatal offspring susceptibility.

HED thanks the Center for Food Safety for their comments regarding dermal absorption, which has provided EPA with an opportunity to improve the DAF that will be applied for difenoconazole human health risk assessments in the future.

Comment regarding "Need for Cumulative Exposure and Risk Assessment of Triazole Fungicides"

Comment (summarized):  Triazole fungicides clearly meet EPA's criteria for designation as a common mechanism group (CMG), for which a cumulative risk assessment must be carried out, as mandated by the Food Quality Protection Act. They have a similar chemical structure, the liver is their primary target organ, they exert similar toxic effects on the liver, and do so via common mechanisms of toxicity (share mode of action and adverse outcome pathways for several endpoints). The European Food Safety Authority conducted a cumulative assessment of triazoles over a decade ago, forming cumulative assessment groups for developmental effects observed following acute exposure (cranio-facial malformations), and for hepatotoxicity as the chronic endpoint (EFSA 2009). All or most of 10 other triazole fungicides reviewed by EFSA induced hepatocellular hypertrophy, hepatic cell degeneration or death, fatty changes, inflammation, and hepatocellular tumors (Nielsen et al., 2012). They exert these effects by activating nuclear receptors that induce the production of cytochrome P450 detoxification enzymes in the liver, causing an increase in cellular organelles that is responsible for hepatic cell enlargement (hypertrophy). There are at least two endpoints, shared by most triazoles, that should be the focus of a cumulative assessment: fatty changes and carcinogenicity.

HED Response:  EPA has not assumed that difenoconazole has a common mechanism of toxicity with other substances at this time. The Agency will utilize the Pesticide Cumulative Risk Assessment: Framework for Screening Analysis framework to determine if the available toxicological data for difenoconazole suggest that a candidate common mechanism group (CMG) may be established with other pesticides. If a CMG is established, a screening-level toxicology and exposure analysis may be conducted to provide an initial screen for multiple pesticide exposure. 

Comment regarding "Cumulative Risk Assessment of 1,2,4-Triazole and its Conjugates"

Comment (summarized):  Triazole fungicides share an eponymous structural feature, 1,2,4-triazole (1,2,4-T), which together with its conjugates (triazole alanine (TA) and triazole acetic acid (TAA)) are common metabolites of these fungicides. Due to concerns over the toxicity of these metabolites, in the year 2000 EPA delayed granting any new triazole registrations pending more toxicology and exposure data for the metabolites (EPA 2/7/06). EPA issued a data call-in for the following studies: developmental neurotoxicity, acute neurotoxicity, and carcinogenicity for 1,2,4-T; developmental toxicity study (rabbits) for TA and TAA; chronic toxicity with neurological evaluations in rat for TA; combined 90-day feeding/neurotoxicity in rat for TAA (EPA 2/7/06). The registrant group US Triazole Task Force (USTTF) did not respond to the 2002 call-in and requested waivers in 2003 that EPA denied. The studies were still outstanding in 2005, when USTTF submitted renewed waiver requests.

To our knowledge, registrants to this day have not submitted the studies EPA demanded 15 years ago as a condition for any further registrations of triazoles. We find no record of any of the above-mentioned studies been submitted. Neither did EPA cease registration of new uses and new triazoles until it had received this study, as it had demanded in 2006. EPA applied arbitrary safety factors in an attempt to compensate for the missing developmental data but has no way of knowing whether they are adequate. In any case, these safety factors are intended only as a temporary stopgap until the relevant studies are submitted, permitting a data-based assessment.

HED Response:  In 2006, HED issued aggregate human health risk assessments for 1,2,4-T and the conjugated metabolites of 1,2,4-T, TA and TAA (M. Doherty, et al., 2/7/2006, D322215). The assessment was based on sufficient data to support a risk assessment for these metabolites. HED conducted two assessments: one for 1,2,4-T and one for combined exposure to TA and TAA. Both assessments are highly conservative, screening-level evaluations in terms of the hazards associated with these compounds (e.g., use of the maximum combination of uncertainty factors) and potential dietary and non-dietary exposures (i.e., high-end estimates of both dietary and non-dietary exposures). Additionally, the assessments retain a 10X database uncertainty factor to account for the data gaps associated with the toxicological database and were designed to be extremely conservative so that the assumptions associated with the risk estimates will remain valid for the majority of the anticipated registrations. While the Agency has not received any new data following the 2006 assessments, a number of requests for new uses of triazole fungicides have been submitted to the Agency. These new use requests have been evaluated with the same conservatisms (i.e., including the 10X database uncertainty factor) in place as in the 2006 risk assessments and as a result, new dietary and aggregate exposure and risk estimates have been calculated for both 1,2,4-T and combined residues of TA and TAA, for each new use proposal as necessary. HED does not believe that exposure or risk has been underestimated through these risk assessment approaches. HED will continue to employ our protective screening approach for all actions involving the triazoles and continue to evaluate the need for additional data. 




References:

Allen et al. (2021) "Retrospective analysis of dermal absorption triple pack data", ALTEX - Alternatives to animal experimentation. https://doi.org/10.14573/altex.2101121.

Cropp-Kohlligian, B. et al., 9/18/2020, D457325.  Difenoconazole. Draft Human Health Risk Assessment for Registration Review.

Doherty, M. et al., 2/07/2006, D322215.  1,2,4-Triazole, Triazole Alanine, Triazole Acetic Acid: Human Health Aggregate Risk Assessment in Support of Reregistration and Registration Actions for Triazole-derivative Fungicide Compounds.

Ghali, G., 2/24/1994, TXR 0051309.  RfD/Peer Review Report of Difenoconazole.

Irwin, W., 9/19/2011, D391350.  Difenoconazole:  Report of the Residues of Concern Knowledgebase Sub-committee (ROCKS). 

Morton, T. et al., 2/04/2020, D455964.  Difenoconazole: Section 18 Request for Use on Spinach in Texas.

Rowland, J. and E. Rinde., 7/27/1994, TXR 0011150.  Carcinogenicity Peer Review of Difenoconazole.

Shah, P., 3/01/2007, TXR 0054532.  D318039.  DIFENOCONAZOLE (PC Code 128847). Request for Restatement of 1994 EPA Cancer Classification and Risk Assessment Approach Using Current Terminology.

Whitby K., 8/26/1992, TXR 0009689.  Petition for Tolerances of Difenoconazole in Imported Wheat, Barley, and Rye Grain Raw Agricultural Commodities. 

