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

Date:  		December 21, 2015 

SUBJECT:	Tetrachlorvinphos (TCVP):  Responses to Arguments Presented in the Natural Resources Defense Council, Inc.'s (NRDC) Aug. 5, 2015 Opening Brief in NRDC v. EPA, Case No. 15-70025 (9[th] Cir.)

PC Code:  083701, 083702
DP Barcode: 430589
Petition No.:  NA
Registration Nos.: NA
Risk Assessment Type:  NA
Regulatory Action: Response to Comments
TXR No.:  NA  
Case No.:  NA
MRID No.:  NA
CAS No.:  22248-79-9 


FROM:       	Wade Britton, MPH, Environmental Health Scientist
		  Risk Assessment Branch IV
  		  Health Effects Division (HED, 7509P)

THROUGH:   Michael Metzger, Branch Chief
             Risk Assessment Branch V/VII 
	Health Effects Division (HED, 7509P)

TO:		 James Parker, Chemical Review Manager	
		 Melissa Grable, Team Lead
		 Risk Management and Implementation Branch I
		 Pesticide Re-evaluation Division (7508P)


This document constitutes the EPA's response to arguments presented in the Natural Resources Defense Council, Inc.'s (NRDC) Aug. 5, 2015 Opening Brief in NRDC v. EPA, Case No. 15-70025 (9[th] Cir.) (Opening Brief).  EPA herein responds, on a point by point basis, to NRDC's arguments specifically pertaining to the Health Effects Division's (HED) 2014 residential exposure assessment for TCVP pet product uses, conducted in response to NRDC's 2009 petition to cancel all pet uses of TCVP.  In addition, the 2015 TCVP human health draft risk assessment for registration review takes into account, where appropriate, the arguments presented in NRDC's Opening Brief.  For any argument presented by NRDC that is not explicitly addressed in the 2015 draft risk assessment, this document explains EPA's consideration of that argument in the context of drafting the 2015 draft risk assessment.


 NRDC Argument:
NRDC argues that the U.S. Environmental Protection Agency (EPA or the agency) incorrectly eliminated the presumptive tenfold margin of safety identified in section 408(b)(2)(C) of the Federal Food, Drug, and Cosmetic Act (FFDCA), 21 U.S.C. § 346(b)(2)(C), to protect infants and children.  NRDC argues that, in conducting the 2014 TCVP residential assessment, EPA did not adequately explain why the "10X safety factor" was reduced to 1X, or why blocking ten percent of acetylcholinesterase was not harmful.  NRDC points out that, following the release of the 2014 TCVP residential assessment reducing the 10X safety factor to 1X, the revised risk assessment for another chemical - chlorpyrifos  -  completed in the context of Registration Review of that chemical retained the 10X safety factor.   

EPA Response:
Since the time of the 2014 TCVP residential assessment, EPA has revised its approach relative to consideration of the FQPA 10X factor based on its recent risk assessment for chlorpyrifos and evaluation of available research for all organophosphates (hereafter OPs).  Newer lines of research on OPs in the areas of potential adverse outcome pathways (AOPs), in vivo animal studies, and notably epidemiological studies in mothers and children, have raised uncertainty about the agency's risk assessment approach with regard to the potential for neurodevelopmental effects in fetuses and children.  Many of these studies have been the subject of review by the agency over the last several years as part of efforts to develop a risk assessment for chlorpyrifos (D424485, D. Drew et al, 12/29/2014).  A detailed review of the epidemiological studies used in this review can be found either in the 2014 chlorpyrifos risk assessment or in the 2015 literature review for other OPs (OPP/USEPA; D331251; 9/15/15).  Ultimately, the agency retained the FQPA 10X Safety Factor (SF) in the 2014 chlorpyrifos revised risk assessment based, in large part, on the findings of these epidemiological studies.  

In 2015 a literature review was conducted expanding the scope of that completed for chlorpyrifos to include consideration of the epidemiological data on any OP pesticide.  This included study designs besides prospective cohort studies, and non-U.S. based studies; eight relevant studies were evaluated.  The 2015 review concluded that, at this time, MOA(s)/AOP(s) have not been established for neurodevelopmental outcomes.  The growing body of literature does demonstrate; however, that OPs are biologically active on a number of processes that affect the developing brain.  Moreover, there is a large body of in vivo laboratory studies which show long-term behavioral effects from early life exposure, albeit at doses which cause AChE inhibition.  EPA considers the results of the toxicological studies relevant to the human population, and as qualitatively supported by the results of epidemiology studies.  The agency acknowledges the lack of established MOA/AOP pathway and uncertainties associated with the lack of ability to make strong causal linkages and unknown window(s) of susceptibility.  These uncertainties do not undermine or reduce the confidence in the findings of the epidemiology studies.  The epidemiology studies reviewed in the 2012/2014 and 2015 literature reviews represent different investigators, locations, points in time, exposure assessment procedures, and outcome measurements.  Despite all these differences in study design, with the exception of two negative studies in the 2015 literature review (Guodong et al, 2012; Oulhote et al, 2013), authors have identified associations with neurodevelopmental outcomes associated with OP exposure across four cohorts and twelve study citations.  Specifically, there is evidence of delays in mental development in infants (24-36 months), attention problems and autism spectrum disorder in early childhood, and intelligence decrements in school age children who were exposed to OPs during gestation.  Investigators reported strong measures of statistical association across several of these evaluations (odds ratios 2-4 fold increased in some instances), and observed evidence of exposures-response trends in some instances, e.g., intelligence measures.

Section 408(b)(2)(C) of the FFDCA instructs EPA, in making its "reasonable certainty of no harm" finding, that "[i]n the case of threshold effects, an additional tenfold margin of safety for the pesticide chemical residue and other sources of exposure shall be applied for infants and children to take into account potential pre- and postnatal toxicity and completeness of data with respect to exposure and toxicity to infants and children."  FFDCA § 408 (b)(2)(C) further states that "the Administrator may use a different margin of safety for the pesticide chemical residue only if, on the basis of reliable data, such margin will be safe for infants and children."  Given the totality of the evidence, there is sufficient uncertainty in the human dose-response relationship for neurodevelopmental effects which prevents the agency from reducing or removing the statutory 10X FQPA Safety Factor.  For the 2015 TCVP draft risk assessment (DRA) (D411095), a value of 10X has been retained.  Similarly, a database uncertainty factor of 10X will be retained for occupational risk assessments.  The agency will continue to evaluate the epidemiology studies and pursue approaches for quantitative or semi-quantitative comparisons between doses which elicit AChE inhibition and those which are associated with neurodevelopmental outcomes and will consider such outcomes in future risk assessment refinements if warranted.  

A full discussion of the literature review on neurodevelopmental effects can be found referenced in Section 4.4 of the TCVP DRA (D411095).  

 NRDC Argument:
NRDC argues that without a chemical-specific study, EPA should have used the 2012 Residential SOPs default of two percent to assess risks from pet collar exposures.  NRDC argues that EPA's 2012 Residential SOPs state that, "if chemical specific [transfer residue] measurements are not available, then a standard value for the fraction of active ingredient available (FAR) for transfer is used [NRDC Opening Brief, p.62]."  Further, EPA's TCVP assessment states that, "[i]t is [EPA] policy to use the best available data to assess exposure," and "[w]here chemical or formulation-specific exposure data were available, th[o]se data were used [NRDC Opening Brief, p. 53]." As such, NRDC argues that EPA should not have relied on the surrogate pet fur residue transfer data from a recently study conducted for the active ingredient, propoxur.  Further, although EPA discussed two other studies using two different chemicals, propoxur and deltamethrin, and used the more conservative mean percent residue transfer resulting from the propoxur study for estimation of TCVP post-application risk, NRDC argues that the recommended default should have instead been applied.

EPA Response:
The 2014 residential assessment addressed in detail the decision to use non-chemical-specific exposure data for assessment of risks from the TCVP pet collar formulation.  The assessment describes that it is EPA policy to use the best available data to assess exposure.  Further, where chemical- or formulation-specific exposure data are available, these data are used to conduct the assessment of residential handler and post-application exposures or integrated with other data as appropriate.  No exposure data were available specific to TCVP pet collars at the time the risk assessment was completed.  It should be noted that TCVP-specific exposure (i.e., residue transfer) data are available for the TCVP containing dust/powder and pump spray products but these data would best be considered in risk assessment only for the specific product types (i.e., formulations) because of their inherent basic differences.  In comparison to other formulations used for pet pest control (i.e., dusts, spot-ons, liquid sprays), it is well known that pet collars are unique and, thus, result in considerably lower percent residue transfer values which was the rationale used as the basis for opting to the use of the propoxur pet collar transferable residue data (i.e., product type is the key consideration).  This was clearly outlined in the 2014 residential risk assessment which stated the following:

       "When other pet product formulations are applied to a treated pet, it can be reasonably assumed that the entire amount of that applied (i.e., the application rate) is available on the hair coat of the animal for transfer to the exposed individual.  For example, the labeled amount of the TCVP pump/trigger spray product can be used directly in the assessment of exposure/risk because it is assumed this amount is available on the surface for contact.  Pet collar formulations operate by slowly releasing the ai from the collar medium, making them efficacious for pest treatment over longer periods of time.  Due to this activity, it can be difficult for the risk assessor to determine the amount which is initially available on the animal because the majority is retained within the collar."  

Given the above considerations, EPA relied on exposure data specific to the assessment of post-application risks from a pet collar product.  While TCVP-containing collar residue data were not available for assessment of post-application exposures from the use of TCVP pet collars, two pet collar formulation-specific post-application exposure studies were available at the time of the 2014 residential assessment; a propoxur pet collar study (MRID 448589901) and a deltamethrin pet collar residue transfer study (MRID 49350101).   Both studies were conducted in a manner consistent with current human ethical standards and were considered scientifically acceptable for use in post-application risk assessment.  The propoxur pet collar study resulted in mean percent residue transfer on Day 0 in excess of that of the deltamethrin study; 0.072% and 0.003% of the collar contents, respectively.  Therefore, for the 2014 residential assessment, the more conservative mean percent residue transfer resulting from the propoxur study was used as a surrogate for estimation of TCVP post-application risks.  

Since the 2014 residential assessment, an amitraz pet collar residue transfer study was submitted that was also conducted in a manner consistent with current standards.  This study results in a higher mean percent residue transfer on the day of application than the other available studies (0.14%) and, therefore, has been used as the basis for the assessment of post-application risks from the TCVP pet collars.  Further, as noted in the subsequent response, HED has also used the Davis study for estimation of post-application risks from TCVP pet collar exposures.  These data (summarized in greater detail in Section 5.3 of the 2015 Occupational and Residential Exposure Assessment for Registration Review) result in a mean residue transfer (Study 1, 12 days after collar placement) of 0.40%.  The Davis study, which is pending review by the Office of the Science Advisor (OSA) Human Studies Review Board (HSRB), further supports that pet collar formulations differ from other types of pet products and that the default pet residue transfer value of 2% would significantly overestimate residue transfer from collars.  

While the three surrogate pet collar residue transfer studies are not chemical-specific, these data are the best available surrogate for TCVP collars and offer a realistic refinement from the high-end default fraction application rate (FAR) transfer factor recommended (2%) in the updated 2012 Residential SOPs which reflects other pet product types (e.g., dusts and sprays).  The default FAR, 2%, stipulated in the 2012 Residential SOPs: Treated Pets, is the 90th percentile of Day 0 (i.e., the first 24 hours following application) measures from all transferable residue data available at that time regardless of product type.  Of the 7 studies which are the basis of the 2% default residue transfer value, 4 are conducted with spot-on formulations and 3 with either aerosol or spray formulations.  In development of the 2012 Residential SOPs and today, it is well understood that the active ingredients in these liquid formulated products are more readily available for transfer to an individual than that of pet collars which are designed for slow release over time; however, at the time of the 2012 SOP development, no residue transfer data were available for use in recommendation of a default residue value specific for pet collar formulations.  As a result, the use of the default FAR is expected to result in a significant overestimate of the risk from pet collar exposures. 

Along with the non-chemical-specific pet collar transfer residue study, a number of high end inputs define the assessment of exposure/risk to a pet collar treated pet, including the following:

:: The application rates used in the assessment of the TCVP pet collars represents the maximum amount of ai in the entire collar.  TCVP pet collar labels direct users to cut off and dispose of any excess length once the product is fit and buckled into place.  As described in the 2012 Residential Treated Pet SOP, the removal of excess length is anticipated; however, because the exact length cannot be determined, the corresponding ai loss cannot be determined and exposure is assessed assuming the full collar length.  The use of the maximum amount of ai in the pet collar being assessed and the FAR transfer factor results in the potential for greater exposure to the individual than if excess length were assumed to have been removed. 
:: The Treated Pet SOP recommends daily exposure durations of 0.77 hours for adults and 1 hour for children (1 to 4 years old) based on survey data from which participants reported the time spent with an animal while performing household activities.  As described in EPA's response to NRDC's petition claim #5 below, the SOP methods used for assessment of adult and child post-application risks imply continual, vigorous contact occurs with the treated pet during the entire duration of exposure.  
:: The highest mean percent residue transfer of all three non-chemical-specific exposure studies (i.e., the amitraz study) was selected for the assessment of post-application exposures from the TCVP pet collars.  The use of these data result in the highest estimated exposures of all formulation-specific exposure data sources and, therefore, the most health protective estimates of risk.  Further, pending HSRB review, the Davis study would be used as it results in higher mean percent residue transfer than the amitraz study.   
:: As defined in the Treated Pet SOP, exposure rates resulting from residue transfer associated with a given formulation and activity is a value, known as the transfer coefficient (TC).  The dust and liquid TCs used to assess post-application exposure to a dog treated with the TCVP collars were derived from studies representing applicator and grooming activities with dogs, respectively.  Since the applicators and groomers directly handled pesticide products and conducted activities that required vigorous contact with the treated dogs, it is expected that their resulting exposures are a reasonable approximation of upper bound estimates of adult and child contact with an animal treated with a pet collar.  In the absence of direct exposure data for this scenario (e.g., homeowner activity with a treated pet), the agency assumes that the application and grooming activities are likely to result in an estimate which is protective of individuals petting, hugging, or sleeping with a TCVP-treated cat or dog.  

 NRDC Argument:
NRDC argues that EPA failed to consider the Davis study for the estimation of post-application risks for exposures to the TCVP pet collar.  The Davis study measured how much TCVP transferred to people's hands after five minutes of petting dogs wearing TCVP collars.  NRDC argued that the Davis study would yield exposures well above the safe level determined through EPA's exposure assessment of the pet collar formulation.   

EPA Response:
EPA considered the potential effect of using the Davis study as the basis for risk assessment as discussed above in the response to NRDC Argument #2.  Formal use of the Davis study is on hold at this time pending a scheduled review of its possible use by the HSRB which is scheduled in January 2016.  Therefore, for purposes of the current TCVP draft human health risk assessment in support of registration review, OPP has used the Davis study data to estimate preliminary, comparative post-application risks.  EPA would rely on these data for regulatory decision making if HSRB determines that the study is scientifically valid and it meets appropriate human ethics requirements.    

The EPA review of the applicable human ethical standards (as outlined in 40 CFR Part 26 Subpart Q) has been conducted and it concluded that:

"there is no clear or convincing evidence that the conduct of the Davis study was fundamentally unethical; that is, the research was not intended to harm the participants and did not fail to obtain informed consent.  Similarly, the conduct of the study was not deficient relative to the ethical standards prevailing at the time the research was conducted; the studies did not place participants at increased risk of harm (based on knowledge available at the time the study was conducted) or impair their informed consent."   

The ethics review also states that: 

"OPP wishes to rely on the TCVP glove residue data generated.  The data may be crucial to a potential EPA decision to improve public health protection by imposing a more stringent regulatory restriction than could be justified without the data. If EPA proceeds under §26.1706, EPA needs to obtain the views of the Human Studies Review Board, provide an opportunity for public comment, and publish a full explanation of its decision to rely on the data, including a thorough discussion of the ethical deficiencies of the underlying research and the full rationale for finding that EPA met the standard in 40 CFR §26.1706 (c) (i.e., that the research is essential to a more stringent regulatory action to improve protection of public health)."

The science review for the Davis study has also been conducted and concluded that the method used to collect transferable TCVP residues from petting/rubbing of the dogs treated with TCVP pet collars is scientifically valid.  In addition to the transferable residue data, the Davis study also includes 1) plasma cholinesterase (ChE) from treated dogs 2) t-shirt samples collected from children exposed to TCVP treated dogs in their residences and 3) urinary biomonitoring for adults and children exposure to TCVP treated dogs in their residences.  For purpose of the TCVP risk assessment, EPA only would rely on the transferable residue data as these data are the only element from the study that result in potentially greater risks.  The other study elements are either not applicable to human exposures (i.e., dog plasma ChE measures), or have scientific limitations (i.e., a physiologically based pharmacokinetic (PBPK) model applicable to TCVP is not available to interpret the urinary monitoring data).  

The EPA intends to present the Davis study at the next meeting of the HSRB which is scheduled for January 12-13, 2016.  If the HSRB concludes that the Davis study does not constitute a human study and is scientifically valid, then EPA can rely on these data for regulatory risk decision making.  Alternatively, if HSRB concludes the Davis study constitutes a human study, then under 40 CFR §26.1706, OPP is required to provide an opportunity for public comment and publish a full explanation of its decision to rely on the otherwise unacceptable data, including a thorough review of the ethical deficiencies of the underlying research and the full rationale for finding that reliance on the data is crucial to imposing a more stringent regulatory restriction.

Until such time that these data have undergone HSRB review, post-application risk estimates for exposures to pet collar treated pets are to be considered preliminary and are presented for purpose of comparison only.   
	
 NRDC Argument:
NRDC argues that EPA incorrectly considered the TCVP pet collars to be a liquid formulated product, as opposed to a solid formulation.  NRDC states that the EPA "failed to research the TCVP flea collar label; instead it ignored the information in the label right on the box regarding the chemical formulation [NRDC Opening Brief, p.67]."  The label for the Hartz UltraGuard Flea and Tick Collar for Dogs (EPA Reg. No. 2596-84) states that "as the collar begins to work, a fine white powder will appear on the surface."  As a result, NRDC argues that the transfer coefficient (TC) recommended for solid formulations should have been used instead of the transfer coefficent for liquid formulations as is recommended by the 2012 Residential SOPs.  

EPA Response:
Per EPA's 2012 Residential SOPs: Treated Pets, pet collar products are categorized as a liquid formulation.  This position was based on research conducted at the time of SOP development that supported that pet collars function by means of diffusion, transferring from the collar to the surrounding area.  More specifically, the active ingredient, which is embedded in the collar matrix, diffuses slowly through the matrix, thus controlling the amount of the active ingredient at the collar's surface.  The active ingredient available on the surface of the pet collar then "rubs off" or transfers from the collar to the animal's hair coat via embedded lubricants which function as transfer agents at the surface of the collar.  Based on the categorization of pet collars as liquid formulations, the assessment of post-application exposures for these product types would be conducted with use of the TCs, and the fraction active ingredient on the hands from TC studies (Faihands) recommended for the assessment of liquid formulated products as recommended in the 2012 Residential SOPs.  

The information provided by NRDC states that the label for the Hartz UltraGuard Flea and Tick Collar for Dogs (EPA Reg. No. 2596-84) states that "as the collar begins to work, a fine white powder will appear on the surface."  HED has confirmed that this statement is present on the current labeling for the identified product and that an identical statement is also found on the following TCVP pet collar products (5 of 9 total pet collar products): EPA Reg. Nos. 2596-62, 2596-63, 2596-83, 2596-84, and 2596-139.  Taking label statements into account, and based upon further research which suggests that some pet collars may act by extrusion of the active ingredient from the collar matrix as a fine dust, HED has reconsidered the position that the TCVP pet collars are all liquid formulated products.  As a result of this uncertainty, in the TCVP draft human health risk assessment in support of registration review, HED has updated the assessment of post-application risks from TCVP pet collars in consideration of both the dust- and liquid-specific TCs and Faihands recommended SOP values.  

 NRDC Argument:
NRDC argues that EPA underestimates children's exposures to their pets through use of an assumption that toddlers spend one hour per day with their pets.  NRDC argues that common sense and experience tell them that many children spend far more time with their pets that one hour daily.  NRDC addresses the study used by EPA to determine the number of hours spent in daily contact with pets stating that "at least two of the nine children in the study spent more than an hour and a half a day with their pets each day, and at least one spent two hours [NRDC Opening Brief, p.48]."  NRDC asserts that EPA should use the upper portion of the distribution (90[th] to 95[th] percentile) and assume that children spend two hours per day with their pets as was assumed in the 2006 Reregistration Eligibility Decision (RED) for dichlorvos (DDVP).   

EPA Response:
In EPA's 2014 residential assessment conducted in response to NRDC's 2009 petition to cancel all pet uses of tetrachlorvinphos, EPA stated the following regarding the daily exposure time assumed for post-application exposures to TCVP treated pets:  

"The exposure time (ET) assumption used to assess residential postapplication exposure to TCVP pet products is derived from a study [Tsang and Klepeis, 1996 (1997 EPA Exposure Factors Handbook)] which sought to evaluate the times that individuals spend performing different activities around the home.  Based upon the 2012 Residential SOPs, the point estimates recommended for adult and child ET with pets are 0.77 and 1 hours, respectively." 

In the study, animal care is defined as "care of household pets including activities with pets, playing with the dog, walking the dog and caring for pets of relatives, and friends."  The data identified the time spent with an animal while performing household activities as recorded in 24 hour diaries by study volunteers.  While the activities defined do not necessarily represent the time volunteers were actively engaged in constant contact with the animal as is implicit in the post-application dermal and incidental oral algorithms, the data are the most accurate representation of time spent with pets available and, therefore, it is assumed that contact is continual throughout the timed activity.  The agency assumes the ET value (adult, 0.77 hours; children, 1 hour) reflects a reasonable high end estimate of time spent in contact with a dog treated with TCVP pet products. 

When use of the study data are coupled with high end assumptions of pet contact, the result is an exposure assessment that inherently implies vigorous, continual contact for the entire duration of contact.  While it is possible that an adult or child may be in close contact with a pet intermittently throughout the day, they would not be actively engaged in the highly vigorous contact implied by use of the TCs based on the applicator exposure data for the full exposure duration assumed. Further, it is possible that adults or children may be exposed from sleeping with a treated pet; however, they are not actively engaged in a high level of contact, or the repeated mouthing behaviors exhibited by children during waking hours, which are inherently assumed in the assessment conducted.

NRDC asserts "that common sense and experience tell them that many children spend far more time with their pets that one hour daily [NRDC Opening Brief, pp. 47-48]."  It must be emphasized that when an adult or child "spends time" with a pet, it is unlikely that the entirety of this time is spent engaged in vigorous, continual contact during this entire period as is implied by the inputs and methods used in the 2012 Residential SOPs to calculate exposures from contact with treated pets.  As described above, the study which supports the exposure time inputs for the 2012 Residential SOPs: Treated Pets is based on measures of time spent in "animal care" where animal care is defined as "care of household pets including activities with pets, playing with the dog, walking the dog and caring for pets of relatives, and friends."  It is critical to understand the connection between time spent in animal care and SOP exposure time inputs which inherently implies continual, vigorous contact during the entire exposure duration.  As defined, animal care could imply direct vigorous contact with a treated pet for a limited period of time, but could just as likely have involved activities which required little to no contact of the pet while it was being cared for.  

The assumption of vigorous, continual contact is based upon the algorithm used to estimate post-application dermal and incidental oral exposures from contact with treated pets.  Of the inputs used, the dust and liquid TCs were derived from studies representing applicator and grooming activities with dogs, respectively.  Since the applicators and groomers directly handled pesticide products and conducted activities that required continual, vigorous contact with the treated dogs, it is expected that their resulting exposures are a reasonable approximation of upper bound estimates of adult and child contact with a treated animal.  In the absence of direct exposure data for this scenario (e.g., homeowner activity with a treated pet), the agency assumes that the application and grooming activities are likely to result in an estimate which is protective of individuals petting, hugging, or sleeping with a TCVP-treated cat or dog.  

Further, EPA supports the use of mean values for default inputs, including the exposure time input.  When the default values are considered together in the exposure algorithm, the combination of these inputs result in upper percentile exposure estimates.  If EPA were to assume an upper percentile for all inputs (e.g., 90th to 95th percentile), the result would be exposure estimates that are overly conservative and unrealistic; particularly given the already health protective nature of the assessment methods and inputs.  

NRDC identified a prior EPA pet assessment conducted for the DDVP RED which used a daily exposure time assumption of 2 hours.  This assessment was conducted in 2006, prior to release of the updated 2012 Residential SOPs.  At that time, residential exposures were assessed based on the recommendations of the 1997 Standard Operating Procedures (SOPs) for Residential Exposure Assessments.  In this outdated document, an exposure time of 2 hours per day was recommended based on, "the best professional judgement and experience of the OPP staff.  The timeframe (2 hours) is approximately the same time children play outside per day.  This is assumed to be an upper-percentile value."  EPA contends that the recommendations of the 2012 Residential SOPs are based on the best data available regarding time spent with pets and that the one hour exposure time recommendation is a health protective input for determination of high-level, continual contact with treated pets.

 NRDC Argument:
NRDC argues that EPA failed to account for an established route of exposure, toddlers who touch an object or food with pesticide-contaminated hands, and then put that object or food into his/her mouth (or "indirect hand to mouth activity"), raised by NRDC in its petition.  NRDC argues that published studies show that "there is actually noticeable indirect hand to mouth activity in infants and children [NRDC Opening Brief, p.50]," and argues that "one study found that, on average, a toddler will touch an object and then put that object into his or her mouth 15 times in one hour [NRDC Opening Brief, p.50]."  NRDC acknowledges that EPA's 2012 Residential SOPs recognize that oral exposures consist generally of two pathways, children contacting treated surfaces and putting their hands in their mouth and from putting objects or other toys in their mouth that had been in contact with treated surfaces.  EPA did not conduct an assessment of indirect oral exposures (i.e., object-to-mouth) in its 2014 residential assessment and instead explained, "the quantification of hand-to-mouth exposures following direct contact with the treated surface always results in exposures of a greater magnitude and, as a result, a more health protective assessment of risk ... than from indirect object-to-mouth exposures."

EPA Response:
The 2012 Residential SOPs describe that residential post-application exposures may occur through dermal, inhalation, and/or incidental oral (non-dietary ingestion) routes.  Post-application dermal exposure can occur from surface-to-skin residue transfer for individuals contacting treated surfaces.  Post-application inhalation exposure can occur from concentrations of pesticide active ingredient remaining in the air following treatment.  Finally, post-application oral exposures are based on the ingestion of pesticide residues that can result from transfer of residues from hand-to-mouth, object-to-mouth, or via direct ingestion of residues through soil ingestion, dust ingestion, or ingestion of pesticide granules or baits.  The assessment of object-to-mouth exposures are recommended only for treated turf and treated carpets/hard floors as it is only from these surfaces that secondary exposures are expected following contact with the treated surface.  The methods used for determination of pesticide residues available for post-application dermal and oral exposures from residential pesticides are described in the 2012 Residential SOPs.  

NRDC acknowledges that EPA's 2012 Residential SOPs recognize that oral exposures consist generally of two pathways, hand-to-mouth and object-to-mouth exposures.  However, NRDC argues that 1) EPA should account for object-to-mouth exposures from objects which have come into contact with a TCVP treated pet and then mouthed by the child and 2) EPA should also account for "indirect hand-to-mouth" exposures whereby children touch an object or food with pesticide-contaminated hands, and then put that object or food into his/her mouth.  
In response to the first argument, EPA described in the 2014 response to the NRDC petition that, "with use of the methods described in the indoor environments and lawn/turf SOPs, the quantification of hand-to-mouth exposures following direct contact with the treated surface always results in exposures of a greater magnitude and, as a result, a more health protective assessment of risk (i.e., a lower MOE) than from indirect object-to-mouth exposures."  NRDC asserts that because, "it (object-to-mouth exposure) would share a common toxicological endpoint, the object-to-mouth exposure would be combined with the dermal and hand-to-mouth exposures."  To clarify, when quantified, the greater magnitude of hand-to-mouth to object-to-mouth exposures is such that the addition of object-to-mouth exposure would have little to no impact on the combined risks.  When hand-to-mouth and object-to-mouth risks are compared for turf, the resulting difference is a ~ 30 fold difference and, for indoor environments, the difference is ~ 4 to 7 fold for hard surfaces and carpet, respectively.  Further, EPA policy recommends that incidental oral exposure scenarios (i.e., hand-to-mouth and object-to-mouth) should be considered inter-related and it is likely that they occur interspersed amongst each other across time.  Combining these scenarios with the dermal exposure scenario would be overly-conservative because of the conservative nature of each individual assessment.  Therefore, the post-application exposure scenarios which should be combined for children 1 < 2 years old are the dermal and hand-to-mouth scenarios only.  This combination is considered a protective estimate of children's exposure.  Therefore, in the case of the post-application assessment of TCVP which has been determined to have no dermal hazard, it is most appropriate to present the hand-to-mouth and object-to-mouth exposures separately.  

Regarding NRDC's second argument, the assessment of hand-to-mouth activity as conducted with use of the 2012 Residential SOPs accounts for the "indirect hand-to-mouth" exposures described from children touching an object or food with pesticide-contaminated hands, and then putting that object or food into his/her mouth.  In order for the pesticide residues to have been transferred to an object or food from pesticide-contaminated hands would first require transfer of residues from a treated surface to the hands.  If EPA were to attempt to assess an "indirect" exposure, a mass/balance measure would have to be undertaken that would require removal of the residues determined to be on the hands to the object or food item handled, thus subtracting from the total on the hands.  Therefore, the quantification of combined oral risks from "indirect hand-to-mouth" and "direct hand-to-mouth" oral exposures would be, in effect, adding back those indirect oral exposures which were subtracted resulting in the same outcome had the indirect exposure never been calculated.

 NRDC Argument: 
NRDC argues that EPA's algorithm for calculating the amount of "transferable residue" is inconsistent with the findings of the Davis study which found that the residues were concentrated at the neck of the animal following application.  NRDC argues that when calculating the residues available to transfer from the surface of a treated pet to an exposed individual, EPA's dermal algorithm divides the residues available to transfer by the surface area of the treated animal.  In the Davis study, researchers found that residues were concentrated at the neck, and that very low levels of residue were found at the tail.  Therefore, regardless of pet size, the bulk of the exposure is from contact with the neck.  

EPA Response:
The algorithm for quantification of dermal exposures from the 2012 Residential SOPs: Treated Pets calculates transferable residues, or those residues anticipated to transfer from the surface of the treated animal to the exposed individual, by means of multiplying the maximum application rate of the product applied by the fraction of the application rate anticipated to be available for transfer (as determined by means of residue transfer "petting" studies) and then dividing by surface area of the treated animal.  The design of the algorithm to account for the surface area of the treated animal was intentional.  

Contact with a treated pet is not expected to be limited to one area of the pet, particularly when these exposures are anticipated to be vigorous and continual for up to one hour daily.  To assess exposures as proposed by NRDC would be equivalent to an adult or child having a high level of continuous contact with only the neck of their pet for a full hour daily, which is highly unlikely.  During this period of exposure, it is just as likely, if not more likely, that contact with a treated animal would occur elsewhere on the animal's body (e.g., along the back, belly, chest, and ribcage) as it would the neck.  

The majority of pet product formulations are not applied in a manner which would limit the concentration of residues to one area of the pet's body.  For example, an application of a shampoo, liquid spray, or dust/powder would typically be applied to the entirety of the animal's body.  In the case of pet spot-on products, while the initial application may be isolated, pet residue transfer studies show that within hours of application residues begin to spread over the surface area of the animal.  In the case of the Davis study, only the neck (over the collar and under the collar) and tail measures were collected.  In accordance with NRDC's claim, these residues were concentrated to a larger degree around the neck versus the tail.  However, it is important to note that the Davis study researchers limited their sampling (petting) to only these areas of the dogs' bodies.  Had the researchers collected samples from other parts of the dogs' bodies, such as the back or along the ribcage, they may have found residues to have spread to other areas of the body as well and it is possible that these could be in excess of those found on the tail.  Further, as evidenced by the Davis study results, measurable residues were present on the tail of the animals, thus indicating that the residues are spreading over the animal and not, as suggested by NRDC, concentrated solely around the neck area.  

In conclusion, an exposure assessment which assumes all exposures were isolated to one area of the animal's body and that all contact for the entire duration of exposure were to occur to this area of the body only is unrealistic.  For the majority of pet product formulations, it is more likely that residues are distributed over the entire pet surface area and that an adult or child's contact with the pet is not limited to one area for the entire duration of exposure.  For pet collar formulations, while the residues may potentially be concentrated around the neck area, it is unlikely that the adult or child contacting the treated pet would only contact this area.  He or she would likely also contact other areas of the body, which, based on the Davis study, would have less residue to be transferred (the Davis study residue measures from the tail were well below those measured around the neck), thus reducing overall exposures from these areas alone.  In effect, if exposures were estimated assuming a greater concentration about the neck, the exposed individual would be exposed more highly when contacting the neck area and less so when contacting other areas of the pet, thus having an effect approximately equivalent to the current design that assumes the residues are distributed across the entire surface area of the animal. 
