HAZARD /DENTIFICATION -

TOXICOLOGY J^NDPOINT SELECTION PROCESS
                A GUIDANCE DOCUMENT
                     Prepared By:

                     Jess Rowland
                   Executive Secretary
         Hazard Identification Assessment Review Committee
                  Health Effects Division
                 Office of Pesticide Programs
                    August 11,1998

-------

-------
I.
II.
                         TABLE OF CONTENTS

BACKGROUND	

DIETARY RISK ASSESSMENT	
      A. ACUTE REFERENCE DOSE (ACUTE RfD)
            1. Objective	2
            2. Relevant Studies for Acute Dietary Risk Assessment	2
            3. Hazard identification Process for Acute Dietary Risk Assessment	3
            4. Acute Hazard Identification for Various Population Subgroups	3

      B. CHRONIC DIETARY RISK ASSESSMENT (CHRONIC RfD)
            1. Objective	.4
            2. Relevant Studies for Chronic Hazard Identification	4  •
            3. Hazard Identification Process for Chronic Dietary Risk Assessment	5

HL    OCCUPATIONAL OR RESIDENTIAL DERMAL EXPOSURE RISK ASSESSMENT	6

      A. SHORT-TERM DERMAL RISK ASSESSMENT
          '  1. Objective	6
            2, Relevant Studies for Short-Term Dermal Risk Assessment 	.....,, 6
            3. Hazard Identification Process for Short-Term Dermal Risk Assessment		7,

      B. INTERMEDIATE-TERM DERMAL RISK ASSESSMENT
            1, Objective			:	.9
            2. Relevant Studies for Intermediate-Term Dermal Risk Assessment	.,	.9
            3. Hazard Identification Process for Intermediate-Term Dermal Risk-Assessment .... 10

      C. LONG-TERM DERMAL RISK ASSESSMENT
            1. Objective .....;.; ....'t	,...	..........	.. 11
            2.-Relevant Studies forLong-Term Dermal Risk Assessment .........:........... 11
            3. Hazard Identification Process for Intermediate-Term Dermal Risk Assessment .... 11

IV.    OCCUPATIONAL OR RESD5ENTIAL INHALATION EXPOSURE RISK ASSESSMENT. 12

      A. Objective .-.-..-;	,';..-..:...,..../.	...,	j.....	 12
      B. Need for Risk Assessment.:.,. ......I...'................. .,v	 12
      C. Relevant Studies for Inhalation Hazard Identification  ..... •........................ .>.-12
      D. Hazard Identification Process for Inhalation Exposure Risk Assessment ...../......... 13

V.    THE INFLUENCE OF DERMAL ABSORPTION IN RISK ASSESSMENT.............. 14

VI.    THE USE.OF NOELs, LOELs, AND ENDPOINTS IN TES PROCESS	:	 15

VH.   APPLICATION OF TES IN RISK ASSESSMENT	-.'	,....: 18

-------

-------
I. BACKGROUND

The Office of Pesticide Programs (QPP), has historically, focused on comprehensive risk assessment for
chronic dietary [Reference Dose (RfD)] exposure, or for endpoints based on cancer, developmental or
reproductive hazards. During the re-registration process, it became apparent that there were pesticide   '
uses and exposures other than dietary which needed to be addressed. Therefore, the decision was made
to perform comprehensive risk assessments for acute dietary as well as occupational and residential
exposures. In order to accomplish this, based on the use pattern, several exposure scenarios have been
developed and the toxicology data base is systematically evaluated for hazard identification for the
various exposure scenarios*.  For hazard identification, the toxicology endpoint selection (TES)] entails
identification of doses and endpoints from appropriate lexicological studies that most closely matched
the route and duration of exposure for which a risk assessment is required. Hazard identification is
accomplished at the Health Effects Division's Hazard Identification Assessment Review Committee
(HIARC) meetings.                                         -

Exposure scenarios evaluated by the HIARC are divided into dietary and non-dietary, the latter of which
is then subdivided by duration and route of exposure. Acute Dietary refers to a one day or 24 hour -
dietary exposure and Chronic Dietary refers to life-time dietary exposure. Non-dietary exposures
include Occupational or Residential Exposures via the dermal and inhalation routes that are divided
into three time periods: a Short-Term exposure period of 1 to 7 days; an Intermediate-Term exposure
period with a duration of 7-days to several months; and a Long-Term exposure period covering a
substantial portion of the life time (i.e., several months to life-time).

Occupational and residential exposure involves three categories of exposed individuals:  I) pesticide
handlers, which includes those who mix, load and apply pesticides in their work; 2) reentry workers,
which includes agricultural workers who reenter treated.fields or greenhouses to harvest crops or   -
undertake any other work in a treated area; and non^agricultural workers whose work area is treated with
pesticides; and 3) individuals who are exposed at home or in the workplace, to pesticides applied by
themselves, family members, or professional applicators.                 -.-. ;  '.'.-•  '     ",'        '
               ••••.'     ,•         ••     "     • • V      •  '''.'••   . • '•  .      •.-...   '•'.'•'
This guidance document describes the procedures used in toxicology endpoints selection for acute and*
chronic dietary, as well as occupational and residential risk assessments; 'For each exposure scenario,  -
guidance is provided for  1) evaluation of toxicity studies that are relevant for use (i.e., route and -;  "  '
duration of the study being similar to the exposure duration of interest); 2) selection of appropriate,
endpoints for hazard identification (i.e., doses and. endpoints that best define the potential hazard in
association with the exposure scenario); 3) the process for hazard identification (i.e., use of a weight-of—
evidence type approach in which all available studies are considered together as opposed to the results of
a single study); 4) the influence of dermal absorption in hazard identification; 5) the criteria for the use
of NOEL, LOEL, and the appropriate endpoints in the hazard identification process; and 6) the use of
MOEs in risk assessments.          '•'•'-

-------
II. DIETARY RISK ASSESSMENT

   A. ACUTE REFERENCE DOSE (ACUTE RFD)

      1. Objective                                                       •        .

      The objective is to identify acute hazard (dose and endpoint) based on the toxic effects observed
      in a study following a single oral exposure (dose) of the pesticide to establish an Acute Reference
      Dose (Acute RfD).
                       »*          .
      2. Relevant Studies for Acute Hazard Identification

          a. Acute Neurotoxicitv Study in Rats

          This study (§81 -8) is pertinent because: 1) animals receive a single oral dose and therefore all
          lexicological effects can be attributed to that one dose and 2) the use of three dose levels
          yields a NOEL that can be used in risk assessments. This study, however, is not available in
          the existing database for most of the pesticides.

          b. Prenatal Developmental Toxicitv Studies

        . The use of the prenatal developmental toxicity studies (§83-3a,b) for acute hazard
          identification process presumes that the developmental effects could result from exposure to
          a single dose (US EPA, 1986b, 1991). The cells, tissues and organ systems are part of an
          ever-changing environment in a developing animal. An adverse effect is likely to occur when
          an exposure of sufficient magnitude occurrs during a critical period of fetal development.
          and/or of a particular organ system. The nature of the critical period to a large extent defines
          the exposure conditions such as the dose, duration and frequency, necessary, to result in an
          altered development Consequently, decisions concerning the appropriateness of endpoints
          from these studies must be based on professional judgement The prenatal developmental : •
          toxicity studies) are relevant because: 1) Jhe treatment route is oral, 2) a single dose of a  .'
          substance, administered at a critical point in the development of the jorganism, can elicit
         'developmental effects and 3) frequently it is possible to ascertain the relationship between the
          day of posing and the manifestation of the maternal or developmental effects. It is noted,  ; •
          however, that the treatment period consists of repeated dosing (9 days, for rats and 15 for   •
          rahhits) tn pfggqftnt animals.   -  <   *    .                '  "             '         .

          c. Other Studies     .

          The subchronic, chronic, or reproductive toxicity, or carcinogenicity studies (conducted via
          the oral route) mav be used ONLY if the acute hazards can be identified to have occurred.'•  -
          during the first few days of the treatment and therefore are appropriate for extrapolation.  ^
         - Human data when there is information on the exposure levels associated with an appropriate
          endpoint. The human data, when available, are given first priority, with the -animal toxicity
          studies serving to complement them.                       .                      .

-------
    3' Hazard TdMirtg^ffrn P
    The critical element is the selection of toxicology endpoints observed after a single oral
    administration of a pesticide. When a potential acute endpoint is identified, a determination
    should be made as to whether the toxicological effect is, in fact, very likely to be manifested
    as a result of a single or (at most) very few doses.

    If the acute hazard identified is from an acute neurotoxicity study, the following data are
    evaluated: 1) type and degree of neurotoxicological effects observed; 2) time of occurrence of
    these effects; 3) dose-response curves; 4)similarity and/or differences in toxicity between the
    sexes; and 5) the appropriateness of the NOEL/LOEL established.

    If the acute endpoint identified is from a developmental toxicity study, (based on either maternal
    or developmental effects), a determination is made if these, are in fact potentially manifested
   after a single or (at most) after very few doses and when these effects are of the greatest concern.
   Also, results of the developmental toxicity studies are compared for any similarity or differences
   in the observed toxicity between pecies. Acute hazards may also be identified by extrapolating
   data from subchronic or chronic toxicity studies if the hazard is identified early in the study.
   A  *   -  —~
   4- Acute Hazard TH.ntifin,nnn Knr VaH«n» popul
  The HED Dietary Risk Evaluation System (DRES) contains food consumption data for U.S
  populaiton and 22 subpopulaito and can provide acute dietary risk assessments for various
  subpopulations such as Females 13 + years, pregnant, Nursing, infants, Non-nursing infants.
  Children 1-6 years old and 7-12 years old, Males, 13-19 years, Males 10 years and older,
  Females13-19 years, and Females20years and older.

  Therefore, for acute dietary risk assessments only, the population subgroups are divided into two
  main categories; Females 13+ (i.e., child bearing age) and the General Population which
  includes infants and children and adult males (i.e, excluding/Females 13+). - The hazard
  identification (i.e, endpoints selection), therefore, most be pertinent to .these population
  subgroups in establishing the acute RflX.
 When the qndpoint selected is from an acute neurotoxicity study (e.g., cholinergic signs.,' '•••'••'
 cholinesterase inhibition, behavioral alterations and/or neuropathology), or from subchronic or
 chronic toxicity studies (e.g.M toxicity attributable early expsoure), it Is appropriate for ajl
 population subgroups (i.e, Females 13+ as well as the General Population which includes infants
 and children) s.

 On the other hand, when a developmental endpoint is selected from one of the developmental
 toxicity studies (rats/rabbits), it is appropriate only for Females 13+ because: 1) the
 developmental effects are in utero effects (i.e., occurs only during pregnancy) and thus
 applicable only to females of childbearing age; 2) developmental effects are not relevant for adult
males; and 3) the effects can not occur posmataily and therefore not applicable to infants and
children. The acute RfD established on a developmental endpoint .should be used for acute
dietary risk assessment only for females of child bearing age (i.e, Females 13 +).

-------
Therefore, when a developmental endpoint is selected, another endpoint (Le., a "non-
developmental)MUSTt be selected for the General Population (i.e, adult males, infants and
children). This can be accomplished by evaluating the maternal toxicity observed in the
developmental toxicity study that was used for Females 13+, or the acute neurotoxicity study,     _
and/other oral toxicity studies. For example, if maternal toxicity (e.g., clinical signs, abortions, body"
weight loss in early dosing period, etc.) is attributable to a single exposure, it can be used for acute
hazard identification for the General Population.  Even though the endpoint identified is maternal
toxicity and occurrs in pregnant animals, it is appropriate for use for the General Population because
effects are seen in-the most sensitive population (i.e., pregnant animals) and thus (the maternal
NOEL) would be protective of potential adverse effects on the developing fetuses as well as the
general population. This acute RID should be used for acute dietary risk assessments for the General
Population including infants and children.

When it is not possible to identify a dose and endpoint (i.e., a "non-developmental endpoint")
attributable to a single exposure from any of the available oral toxicity studies, then a determination
is made that no lexicological effects attributable to a single exposure (dose) were observed in oral
toxicity studies. Thus an acute RfD can not be. established and an acute dietary risk assessment will
not be required for the General Population.

B. CHRONIC DIETARY RISK ASSESSMENT  (CHRONIC RfD)

   1. Objective

   The objective is to identify chronic hazard (dose and endpoint) based on the toxic effects
   observed in a study following repeated oral exposure (dose) of the pesticide to establish an
   Chronic Reference Dose (Chronic RfD).                .

   2. Relevant Studies for Chronic Hazard Identification
   ,                       '•'.'.       # "
       a. Chronic toxicity Study in Dogs (§ 83-lb)
                                           '             .                     '  '   •
       b. Chronic Toriciry/Carcinogenicity Study m Rats (§83-5)

       c. Cacfinogeniciry Study in Mice (§83-2b)

       d Two-Generation Reproduction Study in Rats (§83-4)

       e. Human Data    -

-------
    3- Hazard Identification Process fo
                                                                                    .
    The quantification of lexicological effects of a pesticide consists of an assessment of
    noncarcinogenic and carcinogenic effects. Pesticides that do not produce carcinogenic effects are
    believed to have a threshold dose below which no adverse, noncarcinogenic health effects occur,
    while carcinogens are assume to act without a threshed.  For chronic dietary risk assessment, a  '  '
    chronic Reference Dose (chronic RID) is established based on the assumption that thresholds
    exist for certain toxic effects such as cellular necrosis, but may not exist for other toxic effeots
    such as carcinogenicity.
                   *

    The Chronic RfD is an estimate (with an uncertainty spanning perhaps an order of magnitude) of
    a daily exposure to the human population (including sensitive subgroups) that is likely to be
   without an appreciable risk of deleterious effects during a lifetime. The critical element is the
   selection of toxicology endpoints observed following repeated oral exposure of a pesticide.
   Duration of exposure in the laboratory animal study selected for establishing the chronic RfD
   should be comparable to the expected human exposure. In .general, a weight-of-the-evidence
.   approach should be used in which appropriate endpoints'from all available studies are considered
   together. Considerations would include the similarity of effects among species, the onset and
   development of effects (if this can be determined from' the studies), and similarity or differences
   in effect levels among species.                          •

  When an animal study is selected for deriving the RfD, the principal (critical) study must be
  evaluated for the following factors: 1) adequacy of the study design (e.g., adequate number of
  animals, good animal care, accounting for competing causes of death, sufficient number of dose
  groups and sufficient magnitude of dose levels, the statistical tests used etc.); 2) appropriateness
  of the NOEL and LOELs established; .3) significance of the effect or endpoint identified for both "
  humans and animals; 4) relationship of the study conclusions to the-overall database (i.e.^toxicity
  profile of the pesticide) and 5) the relevancy of the species (tested), dose (selected) and endpoint
  (identified) for human health risk assessments..

  When human data, are available and there is .information on the exposure levels associated with
  an appropriate endpoint, these data-are given first priority with the.animal toxicity studies serving— •--
  to complement them. When aTiuman study is selected for deriving the RE), this' study must be
 evaluated ifcr the following: 1) adequacy of the studydesigri; 2) reliable ^ exposiire.aiid/oic . '
 monitoring data; 3) sufficiently long period of exposure to account for, health effects observed; 4)
 adequate control for confounding factors; and 5) appropriateness of the statistical tests used.
              '.  ' '  ."'.'•  "'••[..   "'•"'"'.••"'..•    ' '•- ••" ..  •' '"  .  •' '  ./'.''  .".''' •'•/•  '" •  ';.'"''
 Once the critical study demonstrating the toxic effect of concern has been identified, the selection
 of the endpoint and dose (usually the NOEL but some times LOELj results from an objective
 examination of the data available on the pesticide. The critical endpoint selected should be the
effect exhibiting the lowest NOEL. The RfD is then derived by dividing the NOEL or the LOEL
by an appropriate Uncertainty Factor (s). Selection of the Uncertainty Factor to be .employed in
the calculation of the RfD is based on professional judgement while considering the entire data
base of lexicological effects for the pesticide.                           '
                           X-

-------
III. OCCUPATIONAL OR RESIDENTIAL EXPOSURE RISK ASSESSMENTS

   A. SHORT-TERM DERMAL RISK ASSESSMENT

      1. Objective

      The objective is to identify Short-Term hazard based on the toxic effects observed in studies
      where the treatment conditions in experimental animals are similar to the route (dermal) and
      duration (1-7 days) of human exposure.
                       *


      The guideline study that is most directly applicable to this route (dermal) and exposure period of
      concern (l-7days), is the 21-day dermal toxicity study. In the absence of a 21-day dermal
      toxicity study or other dermal studies, toxicity studies in which the route of administration is oral
      may be used for short-term hazard identification. If an oral study is used for dermal exposure risk
      assessment, the magnitude of dermal absorption must be ascertained and a dermal absorption
      factor must be identified for route-to-route extrapolation. The procedure for identifying the
      dermal absorption factor is discussed in detail in Section V.

      In general, a Short-Term dermal exposure risk assessment will not be necessary if no systemic
     . toxicity was seen at the Limit-dose (1000 mg/kg/day) in the 21-day dermal study. However, if.
    .  the toxicity profile of the pesticide indicates serious concerns for toxicological effects not
      evaluated in the 21-day study (e.g., neurological or developmental effects), then a weight-of-
      evidence approach must be used in which all available studies (oral and dermal) are considered
      concomitantly for endpoint selection. .

   .   Studies that are considered to be most suitable for this risk assessment are:       -   .

          a. 21-dav DermalToxicitv Study •  '   •••'-              V:      "• •    •   :      .

          This study in rats or rabbits (§82-2) is pertinent because: 1) the experimental conditions-" .• _
          (dermal' applications) simulate human dermal exposure depending on the intended pesticide
          formulation and use scenario, and 2) the treatment period (6iiours/day, 5 days/week for 3
          weeks), although longer; does encompass the exposure period of concern (i.e.f 1-7 days).
v
'
          b.  Prenatal Developmental Toxicitv Studies (oral and/or dermalV
            '        •. ' '      • •   ••."••';.      '  .  '        ' ' • '     '•'-•'. : •• s
          These studies (§83-3a,b) are considered to be appropriate when:

             (1).A 21-day dermal toxicity study is not 'available;

             (2).Nb systemic toxicity is seen in the available 21-day study but data from the
             developmental toxicity studies indicate a serious concern for developmental effects in the
             absence of maternal toxicity.  Thus, the concern for the developmental effects outweighs
             the lack of adverse effects in the dermal study since the dermal studies do not evaluate
             parameters that characterize developmental toxicity endpoints;

-------
       (3) The relationship between the day of dosing and the manifestation of systemic toxicity
       can be ascertained in these studies since clinical observations are made throughout dosing
       period (9 days in rats and 15 in rabbits) making toxicity data available and comparable
       for the 1-7 day period;

       (4) The nature and severity of the developmental effects observed across species are of
       concern or multiple observations of developmental toxicity (which constitute a
       syndrome) were observed in a single or multiple species; and/or

       (5) Using the NOEL from the developmental toxicity study would increase the protection
       against potential adverse effects on the developing fetuses as well as the general
       population.

    c. Acute Neurotoxicitv Battery foral)                               •

    This study (§81-8) is suitable when:

       (1) There is a concern for neurotoxicological effects and/or
                                                    v
       (2) The NOEL for acute neurotoxicity or other endpoints is lower than that for.
       developmental toxicity or any other Short-Tenn endpoint

    d. Other Studies  ,                  :                                             .
                                                            '      '•            - .'
    Range-finding (if available), subchronfc (oral or dermal), chronic and/or reproductive toxicity
    (oral) studies may be used if treatment-related toxic effects appear early and are most
    appropriate to extrapolate for short-term exposures; Endpoints that can be used fionUhese
    studies include effects that appear to occur early in the study (i.e.; within 1-1 days).  This
    extrapolation from long-term studies for short-term hazard identification is valid only if the  :
  . endpoints are established early in the study and are thus most appropriate to extrapolate to   .
    short-term exposures. Early indications of effects in subchronic studies might include, but
    are not limited to, cage-side observations, hematblogy (e.g., anemia), clinical chemistry    .
    (indicate .development of abnormal pathology) and histopathology^ire-rieoplastic lesions).
    data. Body weight data from dietary studies are not appropriate due to palatability problems.
    Body weight data from oral (gavage), dermal and inhalation studies art appropriate as an
    indicators of early animal stress. Therefore, extrapolationfrom subchronic studies as
    opposed to chronic studies may yield a higher confidence in selecting an endpoint for this
    exposure period. On the other hand, cases in which effects appear late (i.e., after several
    months) are not appropriate for hazard identification for short-term risk assessment.
       • .            •      x    *.       '    • •    -      ••..''.                .   -'-''.
3. Hazard Identification Process for Short-Term Jgrma/Eiposare Risk Assessment:  .
         •*
                 *        '                 .                       '                -• _
Toxicology endpoint selection should be made using toxicity data-'generated by the same route  as
the likely exposure (i.e., dermal). Therefore, in identifying the Short-Tenn enpdoints, the
primary preference should be the 21 -day dermal toxicity study.

-------
 When a 21-day dermal toxicity study is available and an appropriate endpoint is identified
 this study, then a determination is made on the: 1) type and degree of toxicity observed; 2)
 relationship between the day of dosing and the manifestation of toxic effects during the first
 week of the study; 3) dose-and time-response curves; 4) toxicity between the sexes; 5)
 relationship between dermal and systemic toxicity; 6) significance of dermal absorption; and 7)
 the appropriateness of the NOEL/LOEL established.
             "V
                  *
 When a 21-day dermal toxicity study is available and: 1) no systemic toxicity is seen; of 2) the
 effects seen are not appropriate for this exposure period (1-7 days)of concern (i.e., body weigh
 gain depression, alterations in hematological or clinical chemistry parameters measured only at
 termination, and/or histopathological lesions,.etc.) and 3) the endpoint is not appropriate for
 making regulatory decisions (e.g., piloerection, etc), then this study should not be used in hazard
 identification.

 When a 21-day dermal toxicity study is available but the toxicity profile of the pesticide indicates
 a potential for neurotoxicitv concerns via the oral route, then die dermal study should be
 examined to determine wheather neurotoxicity parameters were evaluated. If they were not
 evaluated, then this study must be considered concomitantly with the acute neurotoxicity study.
 The toxicity observed in both these studies (oral and dermal) will be evaluated to ascertain the
 influence of dermal absorption as well as the relevancy of the effects based on the routes (oral vs.
 dermal) of administrations.                                                      .

 When a 21-day dermal study is available but 1) either no systemic toxicity is seen, of 2) the
 systemic toxicity seen does not reflect the toxicity profile, of the pesticide (such as potential for -(
 developmental-effects of concern), .or 3) a developmental endpoint was used for establishing an •
.acute RfD, then this study may not be appropriate for this risk assessment Under these •   v '
 conditions, the 21-day dermal study is not appropriate for this risk assessment because; 1) of the
 concern for the fetal effects seen; 2) developmental effects ace considered to be appropriate for
this exposure period of concern; 3) fetal parameters are not evaluated in the dermal toxicity study
 (not a Guideline requirement) and these adverse effects can not be determined for the dermal
 route 
-------
   When an oral NOEL is selected for short-term dermal risk assessment, a dermal absorption factor
   should be determined for appropriate route-to-route (oral to dermal) extrapolation. The dermal
   absorption factor can be obtained either from a dermal absorption study (if available) or
   estimated from the oral and dermal toxicity studies (See Section V)..                  '

B. INTERMEDIATE-TERM DERMAL RISK ASSESSMENT,
                    *                                    .                  '    • ' •
   1. Objective   *

   The objective is to identify intermediate-term hazard based on the toxic effects observed in
   studies where the treatment conditions in experimental animals are similar to the route (dermal)
   and duration (1-week to several months) of human exposure.

   2. Relevant Studies for Intermediate-Term Dermal Hazard Identification

   The current toxicology data requirements contain a number of studies that are relevant for
   intermediate-term hazard identification. In these studies, the duration of treatment corresponds to
   the exposure period of concern (one week to several months). However, studies via the dermal
   route are limited to the 21-day (§82-2) and the 90-day studies in rats or rabbits (§82-3). Other
   studies that may be relevant for intermediate-term hazard identification conducted via the oral
   route include: 1) the subchronic neurotoxicity study in rats (§82-6); 2) the subchronic feeding
   studies in rats and dogs (§82-la,b); 3) the prenatal developmental toxicity studies in rats and
   rabbits (§83-3a ,b); 4) the 2-gerieration reproduction study in rats (§83-4); and 5) interim data
   from chronic toxicity (§83-1 a,b) and caicmogenicity (§83-2 a,b) studies.       /

       a. 21-day and 90-Dav Dermal Toxicity Studies        .    '                     .'"••
  •     "           ,      ' ., •        , "      •                    .           .     -
                                 •i  ,,"'.*•".     -'•'•.  'x •*..."  ,.     .,     , "
   ,,.   These studies are considered to be directly applicable for the purpose of hazard identification
       since the experimental conditions (dermal applications) simulate the real-life exposure
       (dennal)depradingonrnemtendedpestitidefonnuJati^
       treatment period (6 hours/day, 5 days/week for 2 l-or90-days) encompasses the exposure
       period of concern (1 week to several months). However; for most.-of the pesticides, the 21-  .;•
       day dermal toxicity study is conducted instead of the 90-day sriidy based on the criteria ' .
       specified in the 40 CFR. Part 158.          '    •   -•>•;" ^         •;•*-•. /'   ;

       b. Other9Q-Day Toxicity Studies foraD  •.         .         .  ,1

       Studies of 90-days duration are conducted in multiple species (rats and dogs) by the oral
       route. In me absence of eimer a 21-day of 90-day dermal study, the 90-day oral or
       neurotoxicity studies may.be used for intermediate-term hazard identification. However,
       since multiple species and routes are involved,* selection of-an appropriate dose and endpoint
       shoul^ include a comparison of results across these studies and the intended formulation and
       use scenario.                                           -*

-------
       c. Prenatal Developmental Toxicitv Studies foraH

       These studies are relevant ONLY when:

           (1) Dermal toxicity studies are not available to ascertain toxicity via this route;

           (2) No systemic or dermal toxicity is seen in the 21 or 90-day dermal toxicity studies but
           developmental toxicity studies indicate a concern for developmental toxicity in the
           absence of maternal toxicity (since developmental parameters are not examined hi the
           dermal studies);.             •                                 .

           (3) The nature and severity of the developmental toxicity observed across species are of
           concern or multiple observations of developmental toxicity (which constitute a
           syndrome) were observed in a single or multiple species; and/or

           (4) Using the NOEL from the developmental toxicity studies would increase the'
         ,  protection against potential adverse effects on the developing fetuses as well as the
           general population.         '                      ,       '

       d. The 2-Generarion Reproduction Study (dietary):

       .The period of dosing (70 days prior to mating) represents exposure of intermediate duration;
       the route of exposure, however, is oral. Therefore, this study should be used only when the
       parental (systemic), reproductive or fetal toxicity are the major concerns and are the most
.    .   appropriate endpoint .for intermediate-tenn hazard identification:     •';

     ,  e. Long-TermStudies (oral);      .    :   .      . •                  ,-'--•'

       When a relevant study 'Jhat is appropriate for the ratennediate-term exposure period is :
       unavailable, then the interim dak from the bag-torn studies may ;be used.  However; the   .
       interinrdata should be included as a part ofthe; overall evaluation of subchronic effects when
       the weight-of-the-evidence evaluation is conducted. -These data may provide a broader data '\
      ' base for evaluation of appropriate endpoints and hazard identification for mtennediate-terin
       exposure scenario! -'     '-..  ^'  ^"   =v  -    -';.:v.';  . :'•;.  .':"'   ..  •  :.'. '••'  '•[ • '''.'-.•'

3. Hazard Identification Process for Intermediate-Term Dgymo/ Exposure RiskAssessment;.

In identifying hazards for Intermediate-Term risk assessment, critical endpoints and considerations
are similar to those that are currently used in establishing a NOEL/LOEL in any toxicity study.
Generally, a weight-of-the-evidence approach should boused in which appropriate endpoints from
all available 90-day studies are considered together. Considerations would include the similarity of
effects among species, time course of development of effects (if this can^be determined from the
studies), and similarity or differences in effect levels among species. .The dose identified for .  *
intermediate-term risk assessment should not be higher than the dose selected for short-term risk
assessment  If the dose identified is from an oral study, then a dermal absorption factor (estimated,
not known) must be used.

                             .  A          '10   '      '               :"

-------
When studies with appropriate duration, species and route are available, selection of dose and
endpoint should include a comparison of results across studies. Points for considerations include,
but are not limited to: 1} whether the effect occurs in multiple species, 2) if the effect is route
specific,3) time of onset (if it can be determined), and 4) the nature of the dose-response in different
studies. Hazard identification from developmental or reproductive toxicity studies are relevant if the
pesticide has shown the potential to be a developmental or reproductive toxicant and if the dose
identified from these jtadies will increase the protection against potential developmental or
reproductive effects on the developing fetus as well as the general population. .

Comparison should be made to other subchronic studies or to interim data from long-term studies to
evaluate any potential discrepancy between the various data sets. Interim data should be included as
a part of the overall evaluation of subchronic effects when the weight-of-ttie-evidence evaluation is
conducted.  These data may provide a broader data base for evaluation of the appropriate endpoints.
Increase in severity of a toxic response can be evaluated by comparing the NOELs/LOELs from the
chronic and subchronic studies. If the NOELs/LOELs are similar, greater latitude can be given in
extrapolating from long-term data to intermediate-term situations. Where NOELs/LOELs differ
greatly, the effects may be cumulative in nature,: and direct extrapolation from a long-term study to
an intermediate-term time frame may be an overly conservative estimate of hazard.

3. LONG-TERM DERMAL RISK ASSESSMENT.

   1- Objective

   The objective is to identify long-term hazard based on the toxic effects observed in studies where
   the treatment conditions in, experimental animals are similar to the route (dermal) and duration
   (several months to life time) of himian exposure.. ,             ,  -.   ;    ;      '    u
                '  v •           ,      . .   • '     .* '      •           'I   J" ' , ''     . '   "•  •• '
                '  ,   ••  r  t    • , • ' i       ' '   ''•'.-    '        "    ,         '   '   '.-*•"•-•
   2. Relevant Studies for LoneVTenn Dermal Hazard laentfficatfoiii) V f :\ x-     ^  : U

   If a long-term dermal study is available, that study is considered first The current toxicology
   data inurements, however, fo                                                       '•"
   identification and risk assessment Long-term studies are usually available by -the oral route, in'
   which the duration of treatment (major portion 6f the animals life span) corresponds to the:  \  .
   exposure period (several months to.iife time): Due to this limitation^the Committee must rely on
   long-term oral studies for long-term hazard identification and risk assessment and take dermal
   absorption into account These  studies include the chronic toxicity studies in rodents and non-
   rodents (§83-1); the carcinogenicity study injnice and rats (§83-2); and the 2-generation
   reproduction study uvrats (§83-4).

   3.  Hazard Identification Process for Long-Term Dermal Exposure Risk Assessment
                                                 '                         '  ^
   For most pesticides, a Reference Dose (RfD) is used for chronic dietary, risk assessment -.
   _Since long-term.dermal toxicity studies are rarely available, the Committee has to depend on the
    long-term oral studies for identification of long-term hazards. In doing so, the Committee will
    evaluate the appropriateness of the dose and hazard identified for establishing the RfD in relation

                                           11  ••      .    .           •- •    '

-------
       to long-term hazard for occupational or residential exposure requirements. The Committee
       recommend the use of the same dose and endpoint, used for deriving the RfD, for long-term
       dermal rrisk assessment However, if the use pattern and exposure scenario indicates special
       concerns, if the toxicity of the pesticide warrants a hazard identification different from that of the
       RfD, or if the effects seen by the oral route are not expected by the dermal route (i.e., route-
       specific effects), then the Committee will identify a dose different from that used for deriving the
       RfD. If an oral hJOEL (i.e., the same NOEL used to derive the chronic RfD) are selected for
       Long-Term dermal risk assessments, the Committee will select a dermal absorption factor for use
       in risk assessments. At times the 21- day or 90-day dermal (when available) study may be used
       if the Committee determines that the endpoints observed in this study is appropriate for Long-
       Term risk assessments and/or if it believes (based on other studies) that toxicity would not be
       expected to increase over time.
IV. OCCUPATIONAL OR RESIDENTIAL INHALATION EXPOSURE RISK ASSESSMENTS.

   A.  Objective                        .

   The objective is to identify inhalation hazards based on the toxic effects observed in inhalation
   studies where the treatment conditions in experimental animals is similar to the duration of human
   exposure; 1-7 days for short-term, one week to several months for intermediate-term and several
   months to life time for long-term.

   B.  Need for Risk Assessment

   In general, the dermal route of exposure for occupational or residential uses is thi most significant
   For certain pesticides (and use patterns), such as the fijmigants, the inhalatibn route is the most
   significant Inhalation exposure is also a concern when the dermal exposureiias been successfully
   mitigated or when dermal exposure has a very^maU unpact (based oh extremely low dermal   .-
   absorption) on the total .exposure scenario. Therefore, the need for risk-assessment via this route is
   contingent on:, 1) the type of pesticide formulation, 2) use-pattern r Long-Term).

   When there is a concern for potential inhalation exposure (based on the use pattern) and there are noj

                                             12                        ~

-------
 inhalation toxicity studies (except for the acute LCSO study) available in the database, the Committee
 is left with no option but to resort to the use of an oral NOEL for inhalation risk assessments (i.e.,
 route-to-route extrapolation).  While it is generally recognized that route-to-route extrapolations
 should be avoided, in the absence of appropriate inhalation toxicity studies, route-to-route
 extrapolation overcomes the obstacle of inadequate data by allowing one route to substitute for
 another and provides a way to combine risk for multiple routes.
                      «
                    «                                 .
>D, Hazard Identification Process for Inhalation Exposure Risk Assessment;

 When a subchronic inhalation toxicity study (14,21 day or a 90-day) is available, the results should
 be carefully evaluated to ascertain if the toxicity observed is in concordance with the overall toxicity
 profile of the pesticide. If the toxicity observed via the inhalation route is similar to those seen in
 other studies via the oral route (e.g., cholinesterase inhibition, neurptoxicity, hepatotoxicity, etc.),
 then the endpoint can be used for risk assessments. Since this is the only study that is available in
 the database,  it can be used for risk assessments for any time period (i.e., Short-, Intermediate-, and
 Long-Term).
                                              »                          '

 In contrast, if the toxicity observed in that study are clinical signs (such as alopecia and piloerection),
 decreases in body weight gain, and/or gross or histopathology of the respiratory tract but the toxicity
 profile of the pesticide indicate a concern for developmental, neurotoxicity, or hepatotoxicity, then
 this study is not appropriate for use in risk assessments since the endpoints observed in that study did
 not correspond to the toxicity profile of thepesticide. Similarly, the endpoints of concern selected for
 oral and dermal exposure risk assessments should be compared with the endpoints observed in the
 inhalation study.  If me endpomts for oral and/or dermal reassessments are based on^^^
 developmental toxicity, alterations in hematolpgical parameters orneurdpathdlpgy, then the
 inhalation toxicity study is not suitable for risk assessments because these parameters (i.e,
 developmental, hematological or neurological) are not measure in this study. Under these '   •
 conditions, route-to-route extrapolations; should be used.   -   '          -;    .:   V

 la order to conduct route-to-route extrapolations, one must convert exposure levels to mg/kg/day    .
 since the oral NOELs are usually reported in mg/kg/day.  In route-to-route extrapolations, the step
 that allows the conversion from one route to the other is the dosage expressed in mg/kg/day. In  '
 order to convert the exposure (dermal and inhalation) levels to an oral equivalent dose (i.e.,
 mg/kg/day, the following steps are recommended: 1) convert the inhalation exposure (^itg/lb a.i) and -
 the dermal exposure (mg/Ib aj) to oral equivalent doses (mg/kg/day;); 2) combine the converted oral
 equivalent doses to get a combined dose for total (dermal + inhalation) exposurej'and 3) the
 combined dose (mg/kg/day) should then be compared with the oral NOEL (mg/kg/day) to calculate
 the Margins of Exposure.  (The procedure for the "Three C'sw are as follows:

 Step I:    Convert the inhalation and dermal exposures to oral equivalent doses (mg/kg) as follows:

              "* a.  unit inhalation exposure (ug/lb ai) x absorption rate (100% default) x application rate (Ib si/acre) x
               acres treated x 1  mg/1000 fig/kg -*- body weight (70 kg or 60 kg for developmental endpoints)

               b.  unit dermal exposure (mg/lb ai) x absorption late ( % recommended or 100% default) x application
               rate (Ib ai/acre) x acres treated + body weight (70 kg or 60 kg for developmental endpoints)
                                             13

-------
    Step II:    Combine the converted oral equivalent doses to obtain total exposure via these roui
              (inhalation and dermal).           -

    Step HI:   Compare the combined dose to the oral NOELs for the appropriate exposure period (i.e.,
              Short-, Intermediate-, or Long-Term).
 V. THE INFLUENCE OF DERMAL ABSORPTION IN RISK ASSESSMENT.

 Dermal absorption is a significant factor in occupational or residential exposure risk assessments since
 these exposures occur most frequently via the dermal route. Under the current toxicology data
 requirements, often data are not available to perform the route-specific risk assessment due to the lack of
 either dermal absorption data or appropriate dermal studies. One of the greatest difficulties encountered
 by the Committee is the task of estimating the proportion of the pesticide that is absorbed through the
 skin.  Dermal absorption is critical especially when the hazard identified is from an oral study in which
 the experimental conditions (oral dosing) do not simulate the real-life exposure (dermal) scenario.
 Therefore, a correction must be made forthis difference in absorption rates (i.e., oral vs. dermal). This
 can be accomplished by the use of a dermal absorption factor to adjust'(correct) oral (systemic)
 absorption to potential dermal absorption. This correction should be described as an additional
 uncertainty in the final risk assessment. In determining dermal absorption, the Committee considers the
 weight-of-evidence approach including dermal absorption studies as well as comparison of the dermal  .
 and oral studies as discussed below. '

 When ,dermal absorption data are avaUable, caution must to exercised as to what dermal absorption •
 value to use. Some factors mat effect percent dermal absorption include; application site; the type and
 amount of vehicle used; total time of application; total dose applied; and the distribution of the >   *
/administered dose (e.g., quantity in skin wash and on the protective cover, material remaining in or on
 the washed skin, material in selected organs, if;collected, and the residue; uime carcass). When, dermal
 absorption studies are available, me Committee reviews the data and selects the dermal absorption value
 (percent) reported for a 8-10 hour period, tije time period that reflects an average work day for the  r''
 pesticide handlers/mixer/loader/applicator).                           •        1    -       :

 When dermal absorption data are not available, the Committee estimates dermal absorption by:  1)
 comparing the LOELs established in the dial and .dermal studies in me same Species;2) evaluating (on a
 case by case basis) the physical.nature ofme.pesticide(i.e., granular, powder etc); and 3) examining the •
 similarity of the concerned pesticide to other chemicals or classes of chemical compounds. When
 neither the appropriate LOELs and/or data on the pesticide nor structurally related chemicals are
 available (i.e., as a default), the Committee will assume 100% dermal absorption (a likely overestimate).

 When a dermal absorption factor is estimated by comparing the appropriate LOELs, the Committee
 relies on oral and dermal studies conducted in the saine species. When such data are available, the
 Committee will compare: 1) the LOEL from a 90- feeding study in rats to that of the LOEL established
 in the 90-day dermal study in rats, 2) .the maternal or developmental LOEL (as the case may be) from a
 developmental study in rabbits to the LOEL in a 21 -day or 90-day dermal toxicity study in rabbits, 3) the
 LOEL from the acute neurotoxicity study in rats to the LOEL in the 21-day or 90-day dermal toxicity  ^|
 study in rats, or 4) the LOEL from a reproductive study to the LOEL established in'the 21-day or 90-day^P
 dermal toxicity study, and so on.   _                ,                      ~
                                             14

-------
When oral and dermal studies are not available in the same species (i.e., appropriate LOELs), the
Committee will make a comparison of the acute oral and dermal LDJO values to gain some "knowledge"
on dermal absorption. This comparison, however, is less reliable because of the ambiguous nature of the
measurement. If an attempt is made to use the LDJO values, the nature of toxic response (other than
death) must be evaluated, and more than one dose level should be available. In other words, single dose
studies or Limit-Dose studies are not used in such comparisons.

When data are available on the physical-chemical properties of the pesticide or onvstructurally related
compounds, the Committee will use these data in estimating dermal absorption. For example, for one of
the pyrethroids, the Committee estimated a dermal absorption value of 50% based on the physical and
chemical nature as well as the similarity of the dermal absorption data for this class of chemical
compounds.

The Committee, in estimating a dermal absorption factor, will routinely use the weight-of-evidence
approach which considers all of the options discussed above. However, the Committee will assume
100% dermal absorption as a default when: 1) there is a concern for high occupational or residential
dermal exposure; 2) an oral study was used for hazard identification; and 3) there is low confidence in
the studies as well as the data available for estimating a dermal absorption factor or in the estimation
itself. Although the 100% default is in most cases an overestimation of exposure, no better default has
been identified.
VL THE USE OF NOELs, LOELs AND ENDPOINTS IN TES PROCESS.

The dose identified for calculating the Margins of Exposure (MOE) for the various exposure scenarios is
usually the NOEL (sometime LOEL) established in the appropriate study. Identification of the dose is
based, not on the results of a single isolated study, but upon the entire data base using a weight-of-fhe-
evidence approach. The criteria for use of the NOEL or the LOEL are discussed below.

   A. No Observed Effect Level fNOELl

   A dose should not be selected routinely by default simply because it is the NOEL. The entire dbse
   response curve should.be reviewed to determine how the NOEL relates to the dose at which effects
   actually begin to appear (i.e., the LOEL); Similarly, the toxicology endpoint with the lowestNOEL
   should not be automatically selected because that endpoint may not be relevant to the exposure
   scenario under consideration.'In some cases, data from two studies may be considered together to
   determine the most appropriate NOEL.

   B. Lowest Observed Effect Level (LOEL)

   The LOEL is not routinely used in risk assessments because it is identified as the lowest dose in the
   study where treatment related effects actually begin to appear. However, a LOEL may be used if a
   NOEL is not established in the critical study, when severity of the effects observed at this dose is of
   negligible concern for human risk, or when there is a data gap.  Therefore, when a LOEL is
   identified for risk assessment, additional modifying factors (range of 3 to 10) may be used in
   addition to the total Uncertainty Factor of 100 (i.e., 10 for intra- and 10 for inter-species variation).
                                  .<
                                  ''•"          15

-------
C. Toxicology Endpoints used in Hazard Identification Process

Toxicology endpoint selection is based on the NOELs and LOELs established in the Data Evaluation
Records (DERs) prepared for each study. Examples of appropriate endpoints used for establishing
the NOELs/LOELs may include the following:

   1. Acute Neurotoificitv Battery

   »•   Mortality

   *   Clinical signs indicative of neurotoxicity including cholinergic signs either in the presence or
       absence of cholinesterase inhibition

   *•   Cholinesterase inhibition

   >   FOB and/or motor activity measurements evaluated at the estimated time of peak effect (i.e.,
       within 8 hours of dosing).

   »•   Histopathological findings

   *   Other non-neurotoxic but biologically significant findings (i.e., systemic and/or biochemical
       effects)
   2. Developmental Toricitv Studies - The use of developmental studies in the TBS process often
   assumes that toxic effects could result from exposure to a single dose. Consequently, decisions, ? '
   concerning the appropriateness of endpoints for acute hazard identification from these studies
   must be based on professional judgement   ^       •         '

   *  Maternal toxicity, characterized as deaths, clinical signs (e.g., cholinesterase inhibition,
       neurotoxic/cholinergic clinical signs) are appropriate if the possibility that these effects   "
       occurred after a single dose can not be discounted

   >  Developmental toxicity, characterized as deaths (embryonic/ fetal resorptions, fetal/pup
       deaths, post implantation loss), fetal/pup body weight decrements, and permanent alterations
       (malformations or visceral and skeletal variations), are appropriate if the possibility that these
       effects occurred after a single dose oan not be discounted.        .           .

   Other factors considered in the selection of developmental toxicity endpoints include:

    *•  Developmental toxicity observed at maternally'toxic or excessive doses generally lessens
       concern that effects could result from a single dose
                                                 4
    +  Developmental toxicity observed in the absence of maternal toxicity generally increases
     * concern that effects could result from a single dose
       Slope of the dose-response curve
                                            16

-------
»•  Developmental toxicity observed across species or the effects are clearly dose-related

••  Multiple observations of developmental toxicity (which constitute a syndrome) observed in a
   single or multiple species

*  Developmental^endpoints are not appropriate if equivocal or marginal developmental toxicity
   occurred only at the highest dose tested, were seen only in one species and were not
   accompanied by any other developmental effects.

3. 21-Day Dermal Toxicity Study

*  Mortality

>•  Decreases in body weight gain and/or food consumption

»•  Clinical and/or cholinergic signs indicative of neurotoxicity

*  Cholinesterase inhibition

»  Systemic toxicity

4.90-Dav Studies

*  Neurotoxicity (e.g., clinical/chdlinergic signs, ChEI, FOB, motor activity)

*>  Systemic toxicity (e.g., mortality, body weight changes, alterations in clinical pathology
   parameters, changes in organ weights)
                                        '
*•  Gross and/or histopathological lesions

5.2-Generation Reproduction Study;
                  i            '         *
>  Reproductive and fertility parameters of both sexes

*•  Systemic toxicity (e.g., mortality, clinical signs, body weight and food consumption, and  '
   organ weight changes) of parental animals of either generation  .

>  Alternations in reproductive parameters (conception rate, number of corpora lutea etc.)

»  Offspring data (e.g., number of pups bom, pup 'viability, body weight changes, gross
   abnormalities)

»  Reproductive organ toxicity (e.g., changes in the weights of the testis, prostate, seminal
   vesicle and epididymis, and ovaries

»•  Histopathological lesions of the testis, prostate, seminal vesicle, vas deferens, epididymis,
                            •^
                            /
                                        17

-------
          vagina, cervix, uterus, fallopian tubes and ovaries as well as the adrenal and pituitary glands^
          (both sexes)

       »•  Other data, such as semen analysis, accessory cell function (testosterone levels secreted by
          Leydig cells), and hormonal status [follicle-stimulating hormone (FSH) and luteiriizing
          hormone (LH)], when available.

       6. Long-Term Toxicitv Studies

       »  The selection of appropriate endpoints for long-term hazard identification are essentially
          identical to those currently used in establishing the RfD such as neurotoxicity (e.g.,
          cholinergic signs, ChEI, FOB and motor activity, histopathology), systemic toxicity (e.g.,
          changes in body weight and body weight gains, food consumption, clinical pathology
          parameters, organ weights, arid gross as well as-histopathological lesions), and evidence of
          chronic toxicity or carcinogenicity or other significant toxic effects (e.g., target organ
          toxicity).
VL APPLICATION OF THE TOXICOLOGY ENDPOINT SELECTION IN IN RISK
ASSESSMENTS

The dose and endpoints selected in the TES process for the various exposure scenarios are used to
characterize risks. Risk characterization is the discussion of the strengths, weaknesses, uncertainties
assumptions that are part of risk assessment Risk assessment generally involves the integration of
on hazard identification, dose response evaluations and exposure assessment to determine the likelihood
mat humans will experience any of the various forms of toxicity associated with a, pesticide. The results
of risk characterization and risk assessment are used by- the Agency to support regulatory actions.
                                              18

-------
 IX.   REFERENCES.


 Barnes, D.G., and Dourson, M.  1988. Reference Dose (RfD): Description and Use in Health Risk
 Assessments. Regulatory Toxicology and Pharmacology 8:47'1-486.

 Engler, R. 1994. Guidance Document and Process Statement for Less Than Lifetime Hazard  .
 Assessment. Memorandum.  R. Engler, January 10, 1994.
                       *   •         '                     .
 Makris, S., and Rowe, J. 1994. Acute Risk Assessment Policy For Developmental Effects Observed in
 the Presence or Absence of Maternal Toxicity. -Memorandum: S. Makris and J. Rowe, Health Effects
 Division, Office of Pesticides Program. November 21,1994.


 U.S.Environmental Protection Agency. 1986. Guidelines for the Health Assessment of Suspected
 Developmental Toxicants. Fed. Reg.. 51:34028-34040.                             -


 U.S. Environmental Protection Agency. 1991 Gudielines for Developmental Toxicity Risk Assessment.
Fed. Reg. 56:63798-63826.


Whalan, J and Pettigrew, 1998. DRAFT: Inhalation Risk Assessments and the Combining of Margin of
Exposures. John Whalan and  Hugh Pettigrew, Health Effects Division, Office of Pesticide Programs,
March 6,1998.
                                           19

-------

-------