      	ENVIRONMENTAL PROTECTION AGENCY
                         WASHINGTON, D.C. 20460      
                                       
                                              OFFICE OF PREVENTION, PESTICIDES 
                                                           AND TOXIC SUBSTANCES
MEMORANDUM
DATE:	February 22, 2010
SUBJECT:	Tetrachlorovinphos, PC Code 083701, DP Barcode 346880: Summary of Pet Collar Risk Estimates
PC Code:  083701
DP Barcode:   374168
Decision No.: 386529
Registration No.: 2596 (49, 50, 62, 63, 83, 84, 139) & 61483 (53, 54)
Petition No.: NA
Regulatory Action: NA
Risk Assessment Type:  N/A
Case No.:  N/A
TXR No.:  NA
CAS No.:  22248-79-9
MRID No.:  NA
40 CFR: NA
	              									Ver.Apr.08
FROM:		Jeffrey L. Dawson, Chemist/Risk Assessor
			Registration Action Branch VII
            Health Effects Division (7509P)
THROUGH:		Michael Metzger, Chief
      Registration Action Branch VII
            Health Effects Division (7509P)
TO:			James Parker
			Pesticide Reevaluation Division (7508P)

The purpose of this document is to summarize the risks associated with the use of pet collars containing tetrachlorovinphos (TCVP) as the active ingredient and to describe the methods used in comparison to the current practices for calculating risks associated with these types of products.  TCVP is also available in other forms used to treat pets such as dusts, sprays, and dip solutions.  TCVP collars contain 14.55 percent active ingredient with an active stated useful effective duration of 90 to 180 days depending upon which specific collar is considered.

An interim reregistration eligibility decision (IRED) was completed for TCVP in September 2002 and is available at http://www.epa.gov/pesticides/reregistration/REDs/tcvp_red.pdf .  A series of analyses were completed to support the aspects of this document focused on the potential risks which could occur after contact with a pet wearing a TCVP containing collar.  The critical documents can be identified by the following information:  

 Tetrachlorovinphos:  Addendum for Residential Exposures Using Updated Standard Operating Procedures For Residential Exposure. PC Code 083701; Submission S559765; DP Barcode 274891 (Author: S. Hanley, 6/8/01) 
   
 Addendum To The Tetrachlorovinphos Occupational and Residential Exposure Risk Revision For The Reregistration Evaluation Decision. PC Code 083701, DP Barcode 280716 (Author: S. Hanley, 2/4/02)
   
 Tetrachlorovinphos (Chemical ID No. 083701/List A Reregistration Case No. 0321).  Addendum To The HED Human Health Risk Assessment For The Interim Reregistration Eligiblity Decision (IRED). No MRID #.  DP Barcode 281763 (Author: C. Swartz, 4/5/02)

In D274891 exposures and risks from collars were calculated using several different approaches including:

 Standard Operating Procedures For Residential Exposure Assessment (1999 Version);
   
 Standard Operating Procedures For Residential Exposure Assessment (2001 Version);

 Standard Operating Procedures For Residential Exposure Assessment (2001 Version) coupled with maximum transfer rate from a TCVP non-collar study; and 

 Standard Operating Procedures For Residential Exposure Assessment (2001 Version) coupled with mean transfer rate from a TCVP non-collar study.

The risk estimates which were included in D274891 are excerpted and presented as Table 1 below.  The column titled "Only 2001 ResSOPs" would be the recommended choice for considering risks from TCVP products because the other columns are based on outdated methods (i.e., "October 1999 RED Chapter results") or they are based on fur transferable residue measurements from non-collar products which were normalized using an inappropriate basis (i.e., only sampled area of dog instead of total surface area which is current premise for estimating these values was used for normalization purposes).  The MOE (Margin of exposure) for the collar using the recommended method was 95 and the level of concern is 100 so the resulting risks were of concern but just below the target level.  The endpoint which served as the basis for these calculations was plasma/RBC cholinesterase inhibition from a subchronic neurotoxicity study in rats with a point of departure of 4.23 mg/kg/day.  A dermal absorption factor of 9.57 percent was also used.  It is important, however, to consider these results in the context of how current changes in methods might impact them as well as other applicable changes.  A discussion of current methods and other changes compared to the D274891 assessment is provided below.  

      Table 1:  Revised Toddler Aggregate MOEs compared with October 1999 Estimates (Table 5 excerpted from D274891)
                                         
                                   Application
                                     Scenario
                                         
                        October 1999 RED chapter Results*
                                         
                                Only 2001 ResSOPs
                                         
                        2001 ResSOPs with Maximum Study TR
                                         
                        2001 ResSOPs with Average Study TR
                                         
                                Dip (1800 mg use)
                                         
                                    444 to717
                                         
                                        18
                                         
                                       190
                                         
                                       440
                                         
                                       Dust
                                         
                                     45 to 61
                                         
                                        9
                                         
                                        18
                                         
                                        24
                                         
                                     Aerosol
                                         
                                     42 to 99
                                         
                                        16
                                         
                                       5.1
                                         
                                        11
                                         
                                      Collar
                                         
                                        NA
                                         
                                        95
                                         
                                        22
                                         
                                        30
   
The methods used to estimate exposure in D274891 are identical to those used currently to estimate exposure from collars with two notable changes. [Note:  Current methods for calculating exposures from pet products are described in D350531 included as Appendix A of this document.  Please refer to that document for further information.]   Exposure occurs from pet collars in a 2 step process. First, a mass of TCVP is emitted from the collar which is then distributed across the animal where some is absorbed and other residues become available for transfer with those who contact the animal.  In D274891 the amount of TCVP emissions from a collar was assumed to be at an even rate until the mass of TCVP is exhausted over the recommended use period of the collar (i.e., 6 months) and 20 percent of this daily emission was considered to be transferable which can be thought of as the amount of the TCVP emitted daily that could rub off on the skin by contact with a treated pet.  However, the assumption about the daily emission rates from collars has become circumspect as available data indicate this type of even daily emission rate over the life of a collar might not generally applicable as described in the recent FIFRA Scientific Advisory Panel meeting on further updates to the SOPs For Residential Exposure Assessment (see http://www.epa.gov/scipoly/sap/meetings/2009/100609meeting.html for further information).  Currently, it is assumed that the entire TCVP content of a collar would be emitted in a single day rather than assuming a daily even TCVP emission rate over the life of a collar which would dramatically increase risks compared to those calculated in D274891.  [Note:  Transferability of 20 percent is retained in the current method.]  However, it should be noted that this premise clearly leads to conservative estimates of exposure because it ignores the intent of collars to be efficacious over extended periods and also available data for other ingredients which indicate emissions persist for extended periods even though the patterns of such emissions are difficult to generically predict.   It should be noted that the current method used by the Agency is purposely intended to be conservative for screening purposes and that product specific transferable residue dissipation data can be developed to update this input with product specific information as needed.

The other notable change in approach from that used in D274891 compared to current practice, conversely, would dramatically reduce risks compared to those calculated in D274891.  This change involved modifying the inputs used to estimate exposures from mouthing behaviors in children.  Transferable residue sampling methods from treated animals use an intensive petting simulation activity which involves multiple contacts over several minutes.  In D274891 the mouthing behaviors for children were calculated using a frequency value of 20 events per hour for 2 hours.  In the current method only a single mouthing event is considered because the manner in which transferable residues are collected involves extensive contact with the treated animal thereby already accounting for the multiple contact mouthing behavior that the assessment method used in D274891 was intended to simulate (i.e., the transferable residue sampling method is based on multiple vigorous petting events to obtain a single sample, thus the single result reflects the amount of residues which would be anticipated on the skin from the many contacts considered for mouthing behaviors).

In addition to the changes in exposure assessment methodology the point of departure was subsequently increased, based on the same endpoint, from 4.3 mg/kg/day to 6.7 mg/kg/day which would further reduce the calculated risks (see D281763 for further information).  

Along with the impacts changes in methods and the point of departure have on the risks calculated in D274891 additional factors need to be considered relative to the conservative nature of the transferable residues used as the basis for the assessment.  A transferability factor of 20 percent was used to derive the recommended MOE value of 95. This means that of the TCVP emitted from a collar on a daily basis that 20 percent of it would be the direct source of exposure through contact with the treated pet.  This value would also be used in the current methodology in lieu of data and was used in the assessment completed in D274891.  However, research has been done on the transferability of TCVP with products other than collars (i.e., dusts, aerosols, trigger sprayers) which indicate much lower levels of transferability ranging from approximately 0.2 to 1.1 percent depending upon the type of TCVP product (D280716). These are much lower values than the 20 percent used in the assessment of collar risks.  

Additionally, as part of the analyses used to develop updates to the Agency's Standard Operating Procedures For Exposure Assessment described above the transferable residue data available for all pet products, including TCVP, was analyzed in an attempt to develop a comprehensive understanding of the transferability of pet products.  The data for solid formulations was far outnumbered by data for liquid products so the analysis focused on liquids but it should be noted that observed transferability rates for solids tended to be lower than for liquids.  The data for liquid products across several active ingredients were combined (N = 91) and the percent transferability values were calculated (Table 2 and Figure 1).  [Note:  In the document percent transferability is referred to as "Dermal Exposure Fraction of Application Rate or FAR.]  The results indicate percent transferability for all available products even at the highest percentiles is much less lower than the 20 percent value used as the basis for the TCVP assessment.  Based on this information it is highly likely that the 20 percent transferability value used to develop the TCVP pet collar assessment is conservative in nature because multiple chemicals and types of products were considered with relatively consistent results.

   Table 2 (Excerpt From Agency SAP Paper For Updated Residential SOPs  -  
        Table C-1:  Dermal Exposure Fraction of Application Rate (FAR))
                                   Statistic
                 Dermal Exposure Fraction of Application Rate
50[th] Percentile
                                    0.0045
75[th] Percentile
                                    0.0069
95[th] Percentile
                                     0.012
99[th] Percentile
                                     0.022
99.9[th] Percentile
                                     0.030
Arithmetic Mean	
                                    0.0054
SD
                                    0.0043
Range
                               0.0006  -  0.0312
         
         
In summary, the pet collar risks calculated in D274891 indicate that risks were barely of concern given the information used to develop that assessment (i.e., MOE=95 with target of 100).  When coupled with the change in the point of departure risks from collar exposure are no longer of concern (i.e., MOE~150 with target of 100).  However, the methods used to evaluate risks from pet products have evolved since that time including how collar emissions are considered and how mouthing behaviors are addressed given the contact vigor associated with transferable residue sampling methodologies.  Clearly, the current assumption that all TCVP is emitted from a collar on a single day is likely to be conservative and should be considered a screening method.  This should also be considered in context with what is known about the transferability of TCVP from pet fur for other types of products and in general which indicates that the percent of transferability is likely to be much lower than the 20 percent used to develop D274891.  When all of these factors are considered in total, it is unlikely that there is a risk concern over the use of TCVP collars but there are significant uncertainties associated with these conclusions.  A TCVP pet collar specific transferable residue study would likely refine many of these uncertainties and would be recommended if a need existed further understand them. 


    
   Figure 1: Lognormal Probability Plots for Fraction of Application Rate Transferred from Separate Petting Data Sets


















Appendix A:  D350531, Current Guidance for Residential Exposure Risk Assessment for Pet Insecticide Treatments


                 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                            WASHINGTON, D.C.  20460

                                                                  OFFICE OF
                                                            PREVENTION, PESTICIDES AND
                                                             TOXIC SUBSTANCES


MEMORANDUM
DATE:  1/14/2009

        SUBJECT:	Current Guidance for Residential Exposure Risk Assessment for Pet Insecticide Treatments
        
PC Code:  NA
DP Barcode:  350531
MRID No.:  NA
Registration No.: NA
Petition No.:  NA
Regulatory Action: General Correspondence
Assessment Type:  NA
Reregistration Case No.:  NA
TXR No.:  NA
CAS No.: NA


FROM:		Wade Britton, Industrial Hygienist
			Risk Assessment Branch V
			Health Effects Division (7509P)
			Office of Pesticide Programs

THROUGH:		William Donovan, Acting Branch Chief
			Risk Assessment Branch V
			Health Effects Division (7509P)
			Office of Pesticide Programs

TO:		Venus Eagle, Product Manager (01)
			Insecticide-Rodenticide Branch
			Registration Division (7505P)
	   		Office of Pesticide Programs


The purpose of this document is to respond to a request for guidance on the risk assessment of residential exposure from pet insecticidal spot-on formulations.  The attached memorandum, "Draft Interim Guidance for the Assessment of Human Health Exposure and Risk from Pet Insecticide Treatments" represents current, established algorithms and assumptions for the assessment of pet products.  The memo does not address the SOP revisions which are currently under development for residential pet exposure assessment.  Furthermore, this document will remain in draft format until all revised residential SOPs are finalized.  When finalized, the pet risk assessment SOP will supersede this draft guidance.   

                     Science Advisory Council for Exposure

Draft Interim Guidance for the Assessment of Human Health Exposure and Risk 
from Pet Insecticide Treatments

The purpose of this draft memorandum is to provide interim guidance for risk assessors when estimating human health risk from exposure to pet insecticides.  The residential SOP concerning exposure to treated pets is currently under revision.  Until the completion of this process, the following assumptions and algorithms regarding exposure to treated pets should be used for risk assessment purposes.  Deterministic and probabilistic methods of assessing risk from pet exposure are described and either has been deemed appropriate for use during the period of transition.  

Unless chemical-specific data is available, please refrain from the use/reference of any previously performed pet exposure assessment or any pet study data which does not coincide with the following directive.  
------------------------------------------------------------------------------------------------------------

Deterministic Risk Assessment Method

Handler Inhalation and Dermal Dose from Treating Pets

Data and Assumptions for Residential Handler (Applicator) Exposure Scenarios

The series of assumptions and exposure factors which serve as the basis for estimating the dermal and inhalation exposures from all product formulations are derived from the "HED Standard Operating Procedures (SOPs) for Residential Exposure Assessments (December 19, 1997)." 

 Daily dose is based upon the amount of active ingredient handled on the day of treatment (i.e., a single pet treatment).  The Agency always considers the maximum application rates allowed by product labeling.  The maximum rate for a single pet treatment is determined using a representative (average) animal, or a medium-sized (30 lb) dog.  Estimated risks are based on an even loading of residues across the entire surface of a 30 lb dog, where Surface Area (cm[2]) = ((12.3*((BW (lb)*454)^0.65)) from HED's 1993 Wildlife Exposure Factors Handbook.  Treatments intended for cats are defined separately, using product labels to determine a representative weight and the above algorithm to estimate surface area.
   
 Ten percent (0.1) of the active ingredient applied to the pet is assumed to be the amount the homeowner is exposed to during dipping, dusting, and shampooing. This assumption is based on the professional judgment of the OPP/HED staff and assumed to be an upper-percentile value.

 One percent (0.01) of the active ingredient applied to the pet is assumed to be available for dermal and inhalation exposure from handling flea collars. This assumption is based on the professional judgment of the OPP/HED staff and assumed to be an upper-percentile value.
   
 If the product is a spray formulation, it is assumed that the handler is applying one-half can of spray.  Unit exposure values for the application of a typical pesticide product from an aerosol can are similar to that of a pet spray and will be used for this formulation.  These can be obtained from the "Pesticide Handlers Exposure Database (PHED) Surrogate Exposure Guideline, V1.1.  Health Effects Division, Office of Pesticide Programs (August 1998)."
   
 HED assumes handler contact during application with a spot-on product to be negligible.  These products are designed to be self-contained, as it is applied directly from the tube to the pet with the tip of the applicator used to part the pet's hair.  
   
 Adults are assumed to weigh 70 kg (use 60 kg for females when the selected endpoint is from a reproductive or developmental study). A body weight of 70 kg represents the mean body weight for all adults (i.e., male and female, ages 18 years and older) and is the value recommended in U.S. EPA (1996).  A body weight of 60 kg represents the average body weight for females between ages 13 and 54 years (U.S. EPA, 1996).
   
The algorithms used to assess exposure/risk from residential handler dermal and inhalation pet exposure is presented below.

Spray Formulations:

Where:

      D (mg/kg/day) = (UE * AR * DA)/ BW 

D	   =	          daily dermal/inhalation dose (mg/kg/day)
UE	   =	          unit exposure (mg/lb ai)
AR	   =	          application rate or amount applied to animal (lb ai/ treatment)
DA		=	 dermal and/or inhalation absorption factor (%)
BW 		=	 adult body weight (kg)


All Other Formulations:

Where:

	D (mg/kg/day) =	 (AR * DA * F)/ BW 

D		=	daily dermal/inhalation dose (mg/kg/day)
AR		= 	application rate or amount applied to animal (mg ai/ treatment)
DA		=	dermal and/or inhalation absorption factor (%)
F		=	fraction of active ingredient applied which is assumed to be available for dermal 				or inhalation exposure
BW 		=	adult body weight (kg)


Dermal (adult/toddler) and Hand-to-Mouth (toddler only) Postapplication Exposure to Treated Pets

Data and Assumptions for Residential Postapplication Exposure Scenarios

The series of assumptions and exposure factors which serve as the basis for estimating the dermal and incidental oral (hand-to-mouth) exposures from all formulations are derived from the "HED Standard Operating Procedures (SOPs) for Residential Exposure Assessments (December 19, 1997)" and the 1999 Draft Policy 13, "Postapplication Exposure Assessment for Children from Treated Pets."

General assumptions and factors used in the risk calculations include:   
   
 Daily dose is based upon the amount of active ingredient handled on the day of treatment (i.e., a single pet treatment).  The Agency always considers the maximum application rates allowed by product labeling.  The maximum rate for a single pet treatment is determined using a representative (average) animal, or a medium-sized (30 lb) dog.  Estimated risks are based on an even loading of residues across the entire surface of a 30 lb dog, where Surface Area (cm[2]) = ((12.3*((BW (lb)*454)^0.65)) from HED's 1993 Wildlife Exposure Factors Handbook.  Treatments intended for cats are defined separately, using product labels to determine a representative weight and the above algorithm to estimate surface area.
   
 On the day of application, it may be assumed that 20 percent (0.20) of the maximum application rate is available on the pet's body and transferred to the adult/toddler as a dislodgeable residue. This value is based on the professional judgment and experience of the OPP/HED staff from the review of company-submitted data and is believed to be an upper-percentile assumption (US EPA, 1999 SAP).  

 If chemical-/use-specific data have been submitted to the Agency (i.e., a petting/stroking study), this data can be used as a surrogate for the fraction of transferable residue.  Results from the day of application (Day 0) should be used in place of the default transferable residue fraction assumption.  
   
 Postapplication activities must be assessed on the same day that the pesticide is applied because it is assumed that individuals could handle/touch their pets immediately after application.  For subsequent days after application, it may be assumed that residues do not dissipate because it is frequently desirable to maintain a specific level of pesticide on the pet (e.g., flea collars).

 It is assumed that one pet is contacted per day.

 Adults are assumed to weigh 70 kg (use 60 kg for females when the selected endpoint is from a reproductive or developmental study). A body weight of 70 kg represents the mean body weight for all adults (i.e., male and female, ages 18 years and older) and is the value recommended in U.S. EPA (1996).  A body weight of 60 kg represents the average body weight for females between ages 13 and 54 years (U.S. EPA, 1996).

 3 year old toddlers are expected to weigh 15 kilograms (representing an average weight from years one to six).

 The dermal absorption factor is [insert] % as determined by HED.  If a dermal toxicity endpoint is not based on a dermal toxicity study, then a default dermal absorption value must be used.  If no chemical-specific dermal absorption data are available, then a default of 100% should be used in the dermal dose calculation.

 HED default for the surface area of an adult hug is 5625 cm[2], a toddler hug is 1875 cm2 (US EPA, 1999 SAP).
   
 Saliva extraction efficiency is 50 percent (i.e., every time the hand goes in the mouth approximately half of the residues on the hand are removed).

 The approach used to address the hand-to-mouth exposure pathway has been modified since 1999 Draft Policy 13, "Postapplication Exposure Assessment for Children from Treated Pets."  In the draft policy, contact with dogs is based on 40 events per day (20 mouthing events/day for 2 hours). For each event, the palmar surface of the hands (i.e., 20cm[2]/event) is placed in the mouth of the child contributing to non-dietary ingestion exposure.  In the revised approach, the frequency term has been modified to an equilibrium approach analogous to the dermal exposure component (i.e., the frequency = 1 event/day).  The approach was revised since the data from which the transferable residue concentrations were determined rely on a continuous contact (grooming) technique that would lead to concentrations on the hands which are anticipated to be significantly higher than would result from petting/hugging.  

The algorithms used for residential postapplication dermal and incidental oral (hand-to-mouth) pet exposure scenarios are presented below.

Adult and toddler exposure from dermal activity (hug) to treated companion animal:

The following demonstrates the method used to calculate dermal exposures that are attributable to a child touching a treated companion pet.

Where:

	 D (mg/kg/day)  	= [((AR * FAR) / (SA pet)) * (SA hug) * (DA)] 
					                 BW (kg)


      D		=	daily dose from dermal pet contact (mg/kg/day)
      AR		= 	application rate or amount applied to animal (mg ai/ treatment)
      FAR		=	fraction of the application rate available as transferable residue (0.20), or fraction determined from a chemical-specific petting study
      SA pet		=	surface area of a treated dog (5986 cm[2]/ animal)
      SA hug		=	surface area of a child hug (5625 cm[2] (adult), 1875 cm[2] (toddlers))	
      DA		=	dermal absorption factor (%)
      BW		=	body weight (70 kg (adult), 15 kg (toddler))

      
Toddler exposure from hand-to-mouth activity to treated companion animal:

The following demonstrates the method used to calculate hand-to-mouth exposures that are attributable to a child touching a treated companion pet and then placing their hands in their mouth.

Where:

	 D (mg/kg/day)     = [((AR * FAR) / SApet)) * (SAL) * SAhands * Freq)]
                                           BW (kg)

      D		=	daily nondietary ingestion dose from treated pets (mg/kg/day)
      AR		= 	application rate or amount applied to animal (mg ai/ treatment)
      FAR		=	fraction of the application rate available as transferable residue (0.20), or fraction determined from a chemical-specific petting study
      SA pet		=	surface area of a treated dog (5986 cm[2]/ animal)
      SA hands		=	surface area of a child's hands (20 cm[2])	
      SAL		=	saliva extraction factor (50%)
      Freq		=	frequency of hand-to-mouth events (1 event/day)
      BW		=	toddler body weight (15 kg)


Probablistic Risk Assessment Method

In addition to the deterministic risk assessment method described above, a probabilistic risk assessment method is also appropriate for the assessment of human health risk from exposure to pet insecticides.  Until revision of the pet SOP is complete, the assumptions and algorithms as described for the deterministic risk assessment method (handler and postapplication exposure to treated pets) should be used to perform probabilistic risk assessment.  In the probabilistic model, the risk assessor should establish, characterize and validate the distributions used for the deterministic data inputs.  Current, publicly available models include CARES (Cumulative and Aggregate Risk Evaluation System), Calendex(TM), and SHEDS (Stochastic Human Exposure and Dose Simulation Model for Pesticides).  The Agency will consider probabilistic assessments submitted by registrants under chemical review for use in risk characterization.


