Dimethoate:

        Issues Related
      to the Hazard and
Dose Response Assessment
           November 2, 2004
  Office of Prevention, Pesticides & Toxic Substances
      U.S. Environmental Protection Agency
          Washington, D.C. 20460

-------
                                       TABLE OF CONTENTS


               LIST OF ABBREVIATIONS	Page 3 of 48

               LIST OF APPENDICES 	Page 4 of 48

               PREFACE	Page 6 of 48

               I	BACKGROUND      Page?
                    A	Introduction      Page 7
                    B.Interpretation of the Dimethoate Developmental Neurotoxicity and Related Studies    Page 8

               II.DIMETHOATE DNT STUDY AND RELATED CHOLINESTERASE AND CROSS-FOSTERING STU
                    A	Introduction      PageS
                    B	Summary of Study Design for Relevant Data Sources      Page 
                          1. Develop mental Neurotoxicity Study, Companion Cholinesterase Study, and Range-fine
                               a	Range-finding Study      Page 1
                               b	Main DNT Study      Page 1
                               c	Comparative Cholinesterase Study      Pajjfel
                          2	Cross-fostering Study      Pa^Fl
                          3	Reproductive Toxicity Studies      Page 1
                          4	Benchmark Dose Analyses      Page 1
                    C	Results      Page 1
                          1	Brain Cholinesterase Inhibition Data      Page 1
                          2.Pup Mortality: Main DNT Study, Comparative ChE Study, & Range-finding DNT Study
                               a	Study Results      Page 2
                               b	Benchmark Dose Analysis      Page 2
                          3	Dimethoate Cross-fostering Study      Page 2
                          4.,., Standard Reproductive Toxicity Studies in Dimethoate and Omethoate      Page 2
                          5	Discussion      Page 2
                               a	Relationship Between Maternal Toxicity and Pup Mortality
                                     	Page 36 of 48
                               b	Litters vs. Pups as the Unit of Analysis      PageS
j                               c	Pup Mortality      Page 4
i                               d.Brain Cholinesterase Inhibition: Protective for Pup Death Endpoint?     Page 4
                    D	Weight of the Evidence and Summary      Page 4

               III	REFERENCES      Page 4

-------
LIST OF ABBREVIATIONS
AChE
ChE
DCI
DER
DNT
FOB
FQPA
FRN
GD
LOAEL
NOAEL
OP
OPP
PMRA
PND
RBC
SAP
Acetylcholinesterase
Cholinesterase
Data call-in (refers to notice)
Data evaluation record
Developmental neurotoxicity
Functional observational battery
Food Quality Protection Act
Federal register notice
Gestation day
Lowest-Observed-Adverse-Effect Level
No-Observed-Adverse-Effect Level
Organophosphate pesticide
Office of Pesticide Programs
Pest Management Regulatory Agency (refers to Canada)
Post-natal day
Red blood cells
Scientific Advisory Panel

-------
LIST OF APPENDICES

Appendix 1
E-filename:  035001 ha.003.wpd
DIMETHOATE: 2nd Report of the Hazard Identification Assessment Review
Committee. Paul Chin.  March 26, 2002.

Appendix 2
E-file name:  45529703.der.wpd
DATA EVALUATION RECORD.  DIMETHOATE/035001. STUDY TYPE:
DEVELOPMENTAL NEUROTOXICITY STUDY - RAT; OPPTS 870.6300.  MRID
45529703. EPA Reviewer: K. Raffaele. January 14, 2002.

Appendix 3
E-file name:  45529701 .der.wpd
DATA EVALUATION RECORD.  DIMETHOATE. Study Type: DOSE-FINDING
DEVELOPMENTAL NEUROTOXICITY [NON-GUIDELINE] MRID 45529701.  EPA
Reviewer: K. Raffaele. January 14, 2002.

Appendix 4
E-file name:  45529702.der.wpd
DATA EVALUATION RECORD.  DIMETHOATE. Study Type: SPECIAL STUDY,
CHOLINESTERASE INHIBITION [NON-GUIDELINE].  MRID 45529702. EPA
Reviewer: K. Raffaele. January 18, 2002.

Appendix 5
E-filename:  D273221.me2.wpd
D273221: Dimethoate (035001). Review of Data on Developmental Neurotoxicity
Based on: a 6(a) 2 Report; Preliminary Data Submissions from a Range Finding Study
(CHV/068), a Developmental Neurotoxicity Study (CHV/069), and a Cholinesterase
Study (CHV/070); and a Data Audit of these 3 Studies. Kathleen Raffaele and William
F. Sette. March 22, 2001.

Appendix 6
E-file name:  46214501 .der.wpd
Cross Fostering Study (Non Guideline) - Rat (MRID 46214501).  Elissa Reaves and
Susan Makris.  June 24, 2004.

Appendix 7
E-file name:  035001_0013000_030393_TX010065_R014928.tif
EPA ID# 035001: Dimethoate - Review of Reproductive Toxicity in Rats. Paul Chin.
March 3,1993.

-------
Appendix 8
E-file name: dimethoate_appendix8_fmal.pdf
BMD Analysis of Pup Death Mortality Data

Appendix 9
E-file name: dimethoate_appendix9_final.pdf
BMD Analysis of Brain Cholinesterase Data

Appendix 10
E-file name: 46181001.der.2-gen repro.wpd
DATA EVALUATION RECORD. MRID 46181001. STUDY TYPE: 83-4;
Multigeneration Reproduction Study in Rats.

Appendix 11
E-file name: 46348201 .der.1-gen repro.wpd
DATA EVALUATION RECORD. MRID 46348201. STUDY TYPE: Non-guideline;
Range-finding One-generation Reproduction Study in Rats

Appendix 12
E-file name: omethoate_reproductivetox.wpd
Results of Reproductive Toxicity Studies with Omethoate

Appendix 13
E-file name: Meta analysis report-v1 of 2.pdf  AND  Meta analysis report-v2 of2.pdf
A Meta Analysis of Pup Death and Cholinesterase Inhibition Data for Dimethoate.
September 1,2004.

Appendix 14
E-file name: 46288001 .der.28-day oral tox in rats.wpd
DATA EVALUATION RECORD. MRID 46288001. STUDY TYPE: Repeated Dose (28-
day) Oral Toxicity Study in Rats

-------
PREFACE

      A Scientific Advisory Panel (SAP) meeting to discuss issues related to the hazard
assessment for dimethoate was originally scheduled for July 29-30, 2004.  Shortly
before the meeting, EPA became aware of additional information that could impact the
issues scheduled for discussion. This information included three additional toxicity
studies with dimethoate (two reproductive toxicity studies and a 28-day dietary toxicity
study) and additional data analyses conducted by the registrant [Cheminova's Position
Concerning the Appropriate lexicological Endpoints for the Regulation of Dimethoate
(MRID 46245901) and A Meta Analysis of Pup Death and Cholinesterase Inhibition
Data for Dimethoate,  September 1, 2004 (MRID 46386001, Appendix 13)]. Following
receipt of this information, EPA conducted additional  data analyses, including
benchmark dose (BMD) modeling, to further evaluate the dose-response relationships
for Cholinesterase (ChE) inhibition and pup mortality following exposure to dimethoate.
These analyses resulted in different interpretations from the July 6, 2004 EPA
document, in particular the new BMD analyses supported the selection of brain ChE
inhibition as an appropriate endpoint for all exposure scenarios in the dimethoate risk
assessment  Given this new analysis, the availability of a 28-day  dermal toxicity study
on dimethoate, which included brain ChE measurements, obviates the need for route-
to-route extrapolation.  Therefore, the dermal absorption discussion, originally in the
July 6, 2004 EPA document, is not included.  The results of the analyses and the
additional data have been incorporated into the revised background documents and
materials made available for the current SAP meeting.

-------
I.
BACKGROUND
      A.    Introduction

            The Food Quality Protection Act of 1996 requires EPA to reassess all
      previously approved pesticide tolerances by August 2006. As part of the
      reassessment process, EPA's Office of Pesticide Programs (OPP) is developing
      risk assessments for each of the individual organophosphate (OP) pesticides,
      including dimethoate, in addition to a cumulative risk assessment for the OP
      common mechanism group (EPA, 2002).  The primary mode of toxic action for
      OPs is inhibition of acetylcholinesterase through phosphorylation of the enzyme
      active site. This inhibition leads to accumulation of acetylcholine and results in
      cholinergic toxicity due to continuous stimulation of cholinergic receptors
      throughout the central and peripheral nervous systems. Some OPs  require
      activation in vivo to the oxon  metabolite prior to the ChE inhibition. In the case of
      dimethoate, the oxon metabolite is called omethoate. Although not registered for
      use in the US or Canada, omethoate is registered as a pesticide in some other
      countries. Dimethoate is used to control a wide variety of insect pests on a range
      of foods, feeds, ornamentals, and non-crop areas.  Potential sources of exposure
      to dimethoate are food, drinking water, and occupational exposures  (dermal and
      inhalation).
 Figure 1: Chemical structures of Dimethoate and Omethoate
 Dimethoate
                                Omethoate
            The present document was developed jointly between the U.S. EPA and
      Canada's Pest Management Regulatory Agency (PMRA). The purpose of this
      document is to discuss specific scientific issues related to the dimethoate hazard
      and dose-response assessment. These issues relate to data evaluating
      developmental neurotoxicity (DNT) which have become available to EPA and
      PMRA since the 1999 release of EPA's Preliminary Human Health Risk
      Assessment for Dimethoate. Several key studies characterizing DNT for
      dimethoate have become available. These include: a development neurotoxicity
      (DNT; OPPTS 870.6300) study, a comparative ChE study on dimethoate, and a
      cross-fostering study.

-------
B.    Interpretation of the Dimethoate Developmental Neurotoxicity and
      Related Studies

      In 1999, EPA issued a data call in (DCI) notice [Federal Register Notice
(FRN) October 5, 1999 OPP-34192 FRL 6097-9] requiring DNT studies to be
submitted for all registered OP pesticides. In addition to the information collected
in the standard guideline study (which includes behavioral and neuropathological
assessments of offspring following exposure of dams during gestation and early
lactation), the DCI included several additional requirements: (1) that exposure be
continued throughout lactation  (through  post-natal day 21);  (2) that adequacy of
exposure to pups be considered; and (3) that studies specifically evaluate the
relative, or comparative, sensitivity of pups and adults to ChE inhibition.

      Following discussions with the EPA regarding the protocol,  the DNT study
for dimethoate was initiated in October, 2000. On January 3,  2001, Cheminova
(registrant for dimethoate), notified the EPA of unanticipated adverse effects in
the DNT study for dimethoate (L000608).  Specifically, an apparent increase in
the number of deaths in young  pups  when the dams were exposed to dimethoate
during gestation and early lactation was observed.  Following a complete review
of data from the DNT, range-finding,  and comparative ChE studies, including an
on-site audit of the performing laboratory, the no-observed-adverse-effect-level
(NOAEL) for offspring in the main DNT study was determined by EPA and PMRA
to be 0.1 mg/kg/day, with a lowest-observed-adverse-effect-level (LOAEL) of 0.5
mg/kg/day, based on increased pup  death and increases in motor activity. In the
companion ChE study, the NOAEL for ChE inhibition was 0.5 mg/kg following
acute administration and 0.1 mg/kg/day  following repeated administration; similar
inhibition levels were seen in adults and young animals and there were no
differences in the NOAELs among age groups.  Because these effects occurred
at doses lower than effects seen in previous studies, the endpoints from the DNT
study were used in the revised  dimethoate hazard assessment (Appendix 1).

      Section II of this document discusses the results of the DNT, comparative
ChE inhibition, and  cross-fostering studies. EPA and PMRA are seeking
comment from the SAP on key aspects of these studies, namely the
interpretation of available pup mortality data, the possible impact of maternal
toxicity and pre-natal and post-natal  dam exposures on pup mortality, and the
interpretation of a benchmark dose analysis for brain ChE inhibition and pup
mortality data.

-------
II.    DIMETHOATE DNT STUDY AND RELATED CHOLINESTERASE AND
      CROSS-FOSTERING STUDIES

      A.    Introduction

            Several studies that characterize DNT and/or potential maternal neglect
      are available for dimethoate. Along with the main DNT study, a range-finding
      study, and a companion, comparative ChE inhibition study were submitted.
      Several reproductive toxicity studies are also available for dimethoate (2 full
      studies and one range-finding study),  and the registrant has recently performed a
      cross-fostering study to provide further information regarding the relative
      contribution of pre- and post-natal maternal exposure  to post-natal pup mortality.
      The effects observed in these studies are key to the hazard characterization of
      dimethoate. The design of the DNT, the range-finding, and comparative ChE
      inhibition studies, in addition to the cross-fostering study are described in Section
      B below.  Section C contains discussion of results related to pup mortality along
      with the potential cause. Full data evaluation reviews (DERs) for each study are
      provided in Appendices 2-7, 14, 15, and 18.

      B.    Summary of Study Design for Relevant Data Sources

            1.     Developmental Neurotoxicity Study, Companion
                  Cholinesterase Study,  and Range-finding study

                  The DNT study guideline (OPPTS 870.6300) is designed to
            evaluate potential neurotoxic effects in offspring following developmental
            exposure to pesticides or other toxic substances. Briefly, dams are
            exposed to a test substance starting 6 days after mating and continuing
            until 11 or 21 days afterbirth of the young  (post-natal day [PND] 11 or 21).
            Birth and development of the offspring are monitored (including litter size,
            survival, body weight, clinical signs, etc.).  Offspring are also evaluated
            behaviorally, using specific tests of neurological function (including
            detailed observations, motor activity, auditory startle habituation, and
            learning and memory testing), at several time points during and after
            weaning. Neuropathological evaluations (including brain weights and
            measurements as well as histopathology) are conducted at two time
            points (PND 11 or 21, and as adults).

                  a.     Range-finding Study

-------
      Prior to the start of the main DNT study, a range-finding
study was conducted to determine appropriate dose levels for use
in the main study (MRID 45529701, Appendix 3). The range-
finding study used the same exposure regimen used in the main
study. Dams (8-10/dose) were dosed from gestation day (GD) 6
through PND 10 at 0, 0.2, 3.0, and 6.0 mg/kg/day, via gavage.
Pups were potentially exposed in utero starting on GD 6 and via
milk through lactation day 10.  Pups were dosed directly, using the
same doses (on a mg/kg basis) administered to their dams, from
PND 11 until weaning on PND 21.  Reproductive parameters were
collected as in the main study. ChE inhibition was evaluated at GD
20 in dams and fetuses, and at PND 21  in pups only.  Post-dose
sampling times for purposes of ChE inhibition measurements were
the same as those used in the comparative ChE study (see below).
                    10

-------
b.    Main DNT Study

      In the main DNT study (MRID 45529703, Appendix 2), dams
(23-24/dose) were dosed from gestation day (GD) 6 through post-
natal day (PND) 10 at 0, 0.1, 0.5, or 3.0 mg/kg/day, via gavage.
Pups were potentially exposed in utero starting on GO 6 and via
milk through lactation day 10.  Pups were dosed directly, using the
same doses (on a mg/kg basis)  administered to their dams, from
PND 11 until weaning on PND 21 (Fig. 2, DNT). All parameters
detailed in the guideline were evaluated in the main DNT study,
including  reproductive data, pup development and survival, and
behavioral and neuropathological evaluations.
      Figure 2
                    11

-------
c.    Comparative Cholinesterase Study

      ChE activity levels (in whole brain, plasma, and red blood
cells [RBC]) were measured in a companion study (MRID
45529702) to the main DNT study, using the same dosing and
exposure regimen as the main study (DER Appendix 4) with 8-10
dams/dose.  ChE activity was evaluated at several time points: in
fetuses and dams on GD 20, and in pups on PNDs 4, 21, and 60.
In addition, to evaluate ChE inhibition following a single exposure,
pups from a satellite group of unexposed dams received  a single
oral dose of dimethoate on PND 11, at the same doses used in the
DNT study (Fig. 3, ChE). To enable comparison of pup versus
adult sensitivity following a similar dosing regimen, ChE inhibition
was also assessed in a group of adult males and females following
a single dose or 10 repeated doses of dimethoate, at the same
doses received by pups  in the DNT study.

      The study report stated that times for assessment were
chosen based on previously available data for time to peak effect
for adult rat behavioral observations.  Sampling times were:
      GD 20 (fetuses and dams):
      PND 4 (pups):
      PND 11 (pups):
      PND 21 (pups):
      PND 60 (adult offspring):
      Day 1 (adults):
      Day 11 (adults):
3 h post-dosing
4 h after dosing to dams
2 h post-dosing
2 h post-dosing
39 days post-dosing
2 h post-dosing
2 h post-dosing
                         Figure 3
                    12

-------
2.    Cross-fostering Study

      To obtain further information regarding the relative contribution of
pre- and post-natal maternal exposure to post-natal pup mortality, a cross-
fostering study (MRID 46214501, DER Appendix 6) was conducted by the
registrant. Cross-fostering allows maternal animals to rear unrelated
offspring, allowing separate assessment of the effects of pre-natal or post-
natal exposure to pups. In this study, dimethoate was administered to
dams at doses of 0, 3,  or 6 mg/kg, from GD 6 through PND 11, in a
regimen similar to that  used in the main DNT study.  On PND 1, pups from
dams treated with dimethoate at 3 or 6 mg/kg were cross-fostered to
control dams, and control pups were cross-fostered to dimethoate-treated
dams; pups from two additional groups {control or 6 mg/kg dams) were not
cross-fostered. Thus, half of the litters containing 12 or more pups from
dams receiving 0 and 6 mg/kg/day were cross-fostered to different dose
groups whereas all of the litters containing  12 or more pups from dams
receiving 3 mg/kg/day were cross-fostered. Detailed behavioral
observations of pups and dams were conducted at several time points
during gestation and lactation. At PND 11,  surviving pups and dams were
sacrificed and a number of clinical chemistry and hematological
parameters were measured. ChE inhibition was not evaluated in the
cross-fostering study.

      3.     Reproductive Toxicity Studies

            In a reproductive toxicity  study, parental (FO generation)
animals are continuously exposed for 6-8 weeks before mating, and
during mating, gestation, and lactation, typically in feed or drinking water.
First generation offspring (F1 generation) are exposed in utero followed by
exposure before mating, during mating, gestation, and lactation typically in
feed or drinking water.  Second generation  offspring (F2 pups) are
exposed in utero and during lactation.  Reproductive parameters, including
litter size and pup survival, are evaluated.  Some ChE inhibition data are
also available from the reproductive toxicity studies for dimethoate.

4.    Benchmark Dose Analyses

      Benchmark dose modeling offers an alternative approach to the
use of NOAELs and LOAELs as points of departure for purposes of
developing risk assessments (USEPA, 2000).  In benchmark dose (BMD)
modeling, mathematical and statistical techniques are used to estimate
the dose at which  a defined level of response (benchmark response, or
BMR) is expected to occur. Typically, the lower limit on the BMD (BMDL)
is also calculated and reported.  As discussed in EPA's Draft Benchmark
Dose  Technical Guidance (2000), a lower confidence limit is placed on the
BMD to obtain a dose (BMDL) that assures with high confidence (e.g.,
                          13

-------
95%) that the BMR is not exceeded. In cases where data are sufficiently
robust to support an analysis, BMD modeling is preferred over the use of
NOAELs and LOAELs since NOAELs and LOAELs do not necessarily
reflect the relationship between dose and response for a given chemical
but instead reflect the dose levels selected for testing. In a recent
submission, the registrant for dimethoate has performed several BMD
analyses of ChE inhibition and pup mortality in available studies for
dimethoate (see Appendix 17).  EPA and PMRA have also conducted
several BMD analyses for these data; the results of these analyses are
discussed below.
                         14

-------
C.    Results

      Findings from the main DNT study, comparative ChE and range-finding
studies, and a cross-fostering study, as well as relevant reproductive toxicity
studies, are briefly described below. Results of the BMD analyses of brain ChE
inhibition and pup mortality are also provided.

      1.     Brain Cholinesterase Inhibition Data

            Following  exposure to dimethoate, inhibition of brain ChE most
      often occurs at doses similar to or below those causing ChE inhibition in
      the blood compartments (see Appendix 1, DIMETHOATE: 2nd Report of
      the Hazard Identification Assessment Review Committee). Results from
      the DNT companion ChE study are consistent with this pattern: for all
      ages, brain ChE  inhibition was seen at or below doses causing inhibition
      in other compartments (see Appendix 4, DER for companion  ChE study).
      This document will focus on  analysis and interpretation of brain ChE
      measurements from the comparative ChE and the DNT range-finding
      studies. ChE inhibition data for plasma and red blood cells (RBC) are
      provided in Appendix 4.

                  As  seen in Table 1, statistically significant decreases in brain
      ChE activity  of 12-18% have been observed following acute (single dose)
      exposures to 3.0 mg/kg dimethoate in pups and adults. Following acute
      exposures at the dose of 0.5 mg/kg, consistent, but small, decreases in
      brain ChE activity of 2-4% were observed in pups and adults.  Although
      the decreases at 0.5 mg/kg following single exposures were statistically
      significant for two treatment groups (adult and PND11 males), they were
      not of sufficient magnitude to be considered lexicologically significant.

            Following  multiple exposures to 3.0 mg/kg in the comparative ChE
      and range-finding studies, statistically significant decreases in brain ChE
      activity of 22-75% were observed in GD 20 dams, PND 21 offspring, and
      day 11 adults (Tables 1 and 2). Statistically significant decreases in brain
      ChE of 6-13% were consistently observed at 0.5 mg/kg/day across all
      groups of pups and adults.  Results observed at the lowest dose of 0.1
      mg/kg/day following repeated exposures were less consistent, ranging
      from increases of 4% to decreases of 12%, when compared to control
      values, with  no clear pattern of effect across age or sex (Table 1).
                                15

-------
Table 1.  Brain Cholinesterase Activity in Adults, Fetuses, and Offspring of Rats Treated with Dimethoate in
Comparative ChE Inhibition Study.
Dose (mg/kg/day)
                                                                               BMPio    JBHIDLio
                  Brain ChoBnesterase Activity (p/kg)*
                                                             Bench marie Doses8
                                                             (nrtg/kg/day)
Adult Males
13,794 i 247
13,544*802
(2)a
13,294*t 241
(4)
12,131** t 1096(12)
2.6
2.0
Adult Females
14,150  555
13,625  445
(4)
13.850 t 687
(2)
12,106** 827 (14)
2.2
1.8
PND11 Males
6475  244
6363  236
(2)
6144* 360
(5)
5375*" 290 (17)
1.8
1.5
PND 11 Females
6256 i 195
6350  338
6125 298(2)
5144** 532 (18)
1.5
1.3
GD 20 Dams
(n=8/group)b
12,838 i 1373
13,044 530
(-2)
11,563*i 300
(10)
5094**  1081 (60)
0.3
0.3
GD 20 Fetuses
(n=8/group)l>
1781  175
1569*i 173
(12)
1600*  136
(10)
1188** 164(33)
0.9
0.7
PND 4 Males (n=14-
19/group)c
3137  322
2817*  434
(10)
2889* 215
(8)
2744** 335 (13)
           2.3@
PND 4 Females
(n=12-16/group>e
2823  310
2941  253
H)
2650  287 (6)
2638  269 (7)
45@
2.3@
PNO 21 Males
(n=8/group)d
10,375 207
9944* * 331
(4)
9044**  340
(13)
5675** i 551 (45)
0.4
0.3
PND 21 Females
(n=8/group)d
10,275 376
9906 313
(4)
9019** t 248
(12)
5956**  965 (42)
0.4
0.3
Adult Males
(n=8/group)*
14,100  529
13.988 662
(1)
12,700* 548
(10)
7469**  2484 (47)
0.5
0.2
Adult Females
(n=8/group)*
14,869  1400
13,913 446
(7)
12,881 ** 845
(13)
6188** 1078(58)
0.4
0.3
PND 60 Males'
13,000* 450
13,100 411
(-D
12,988 422
(0)
13,044 756(0)
NE
NE
PND 60 Females'
13,275 277
12,950  317
(2)
12,738*i 243
(4)
12,744*  586(4)
NE
NE
a Results in parenthesis () are percent inhibition relative to control
b Animals exposed from gestation day 6 to 20
 Animals exposed from gestation day 6 to post-natal day 4
* Animals exposed from gestation day 6 to post-natal day 21
* Animals exposed for 11 days
' Animals exposed from gestation day 6 to post-natal day 21
* = ps 0.05, **p a 0.01
                                                             ME=not evaluated
                                                               =poor model fit or
                                                             values outside dose
                                                             range
                                                             9See Appendix 8 for
                                                             details of analysis
                                                             doses in mg/kg/day
                                                 16

-------
      Data from the range-finding study present similar findings to those
from the comparative ChE study (Table 2).  Although the data were not
statistically analyzed, there was a dose-related decrease in brain ChE
activity at the 3 and 6 mg/kg/day doses.  It is notable that the degree of
inhibition observed at 3 mg/kg/day is consistent with that observed in the
comparative ChE study at the same dose.
                           17

-------
Table 2. Brain choiinesterase activities (^g/kg) from dimethoate range-finding study
Dos
tmg/kg/day)
                     0.2
                                               Benchmark Dose
                                                 (mg/fcgttay)*
                                                                                          BMDL,
Dams
12710 1333.9
12680 + 640.9
     (0)
3240 411.4
    (75)
 1580  195.6
     (88)
0.2
0.2
Male fetuses
 2150 i 562.4
2320  675.1
    (+8)
1670  564.1
    (22)
 1390*780.5
     (35)
1.0
0.3
Female fetuses
 1970  288.5
2100 871.8
1500 t 500.0
    (24)
 1140 638.7
     (42)
1.0
0.4
Male pups
10555  623.8
9942*614.1
     (6)
5839  749.6
    (45)
4720  1654.5
    (55)
0.4
0.3
Female pups
9338  2709.6
9886 t 445.2
    (+6)
5414 756.7
    (42)
 3186  827.3
     (66)
0.5
0.4
                                               18

-------
Data taken from Tables 28-32, pp. 78-82, MRID 45529701.
Number in parentheses is percent inhibition; data were not statistically analyzed
*See Appendix 9 for
details of analysis
      As discussed earlier, BMD analyses provide a way to model the
dose-response relationship for a given endpoint based on available data.
Additional analyses of dimethoate data were recently submitted by the
registrant for dimethoate (Appendix 13), using BMD modeling to refine the
hazard assessment for dimethoate. EPA has also conducted additional
analyses of the dimethoate data, using BMD modeling.

      BMD models are used to interpolate a dose at which a specific
response level of effect occurs (the benchmark response, or BMP). Thus,
use of these models in hazard assessment requires selection of a BMR,
which can then be compared across different treatment groups. A BMDi0
(using the 'extra risk' model) is the dose resulting  in a 10% change from
the estimated background level for the endpoint being evaluated.  For the
endpoint of brain ChE inhibition, BMDi0 is defined as the estimated dose
which is expected to result in a 10% reduction in enzyme activity, when
compared with control levels.  A BMDL10 is the lower 95% confidence limit
on the BMD-io estimate.  As described in detail in  EPA's Revised
Cumulative Risk Assessment for the OPs (USEPA, 2002), EPA has
previously performed a power analysis of brain ChE data extracted from
rat toxicity studies submitted to OPP for purposes of pesticide registration.
This analysis included data from more than 30 OPs and over 100 studies.
The power for each study to detect a difference of 1 %, 5%, 7.5%, 10%,
15%, and 20% in mean brain ChE activity between control and a single
treatment group was calculated. This analysis showed  that a 10% change
in mean activity was at the low end of detectability of assays for brain ChE
activity as they were conducted in those studies.  In the dimethoate
comparative ChE study, the method used to measure brain ChE and also
the number of animals in each dose group are similar to those used in the
studies considered in the power analysis. Since a 10% change in mean
brain ChE activity has been previously determined to be at the lower limit
of detectability, it is considered appropriate for use as a benchmark
response in the dimethoate evaluation.

      To better compare relative sensitivity across groups, BMD10 and
BMDLio values for brain ChE inhibition were computed  for each exposure
group in the comparative ChE and range-finding DNT studies using the
same dose-response model (i.e., exponential model) as utilized in the OP
Cumulative Risk Assessment. This method has been previously
evaluated by the FIFRA SAP (2001, 2002).  To provide an indication of the
consistency of these values across different studies, BMDip and BMDLio
values were also determined for several additional studies in the

                          19

-------
dimethoate database (a range-finding one generation reproductive toxicity
study in rats, a 28-day dietary study in rats, and two two-generation
reproductive toxicity studies in rats). Complete results of these analyses
are provided in Appendix 9.  BMDio and BMDLio values for the companion
ChE and DNT range-finding studies are included with the results for ChE
inhibition in Tables 1 and 2, above. BMD10 and BMDLio values for
additional studies are provided in Table 3, below.

      For repeated exposures in the DNT companion ChE and range-
finding studies, results of the BMD analysis demonstrate very similar
BMDio and BMDLio values for all groups (range 0.20-1.0 mg/kg/day for
BMD10 and 0.2-0.7 mg/kg/day for BMDLio). No age or sex-related
differences were seen. Similar BMD values were estimated from other
studies and thus support the findings from the comparative ChE and
range-finding DNT studies. Computed values from other studies ranged
from 0.3-1.0 mg/kg/day for BMDio, and 0.2-0.8 mg/kg/day for BMDLio, a
very similar range to those from the DNT-related studies.  Generally, the
various brain ChE BMD values (both BMDs and BMDLs) provided in the
analysis submitted  by the pesticide registrant for dimethoate,  Cheminova,
are comparable to those computed by OPP.

      Clear dose-response relationships were exhibited by a majority of
the brain ChE data from repeated dosing and adequate model fits were
attained for most of the data sets.  Although the dose-response curve
appeared to adequately model the data (by visual inspection), the
goodness-of-fit statistic for the PND 42 males from the one generation
reproductive toxicity study and the day 28 adult females from  the 28-day
dietary study resulted in highly significant p-values (i.e. p-value < 0.01)
indicating the model's lack of fit. Additionally, the goodness-of-fit statistic
for the day 218 adult females from the "new" two-generation reproductive
toxicity study has a borderline significant p-value of 0.057 indicating the fit
of that model is also questionable.
Table 3. BMD Values for brain cholinesterase inhibition for related dimethoate studies.
Study fyp*
Two-generation dietary Reproductive Toxicity



Two-generation dietary Reproductive Toxicity



MRlDNo,
4225S01



46181001



Sabpopwtetiorf
Day 224, FD adults
Day 224, F0 adults
Day 308 Ft adults
Day 308 F, adults
Day 205, Fo adults
Day 205, Fo adults
Day 218 F! adults
Day 21 8 F, adults
Sex
M
F
M
F
M
F
M
F
8**0m
0.7
0.3
0.4
0.4
0.3
0.5
0.8
0.6@
SMDi
0.7
0.3
0.3
0.3
0.2
0.3
0.4
0.5@
                          20

-------
One-generation dietary range-finding
reproductive toxicity



28-day dietary toxicity

46348201



46288001

Day 91 adults
Day 91 adults
PND42 offspring
PND42 offspring
Day 28 adults
Day 28 adults
M
F
M
F
M
F
0.4
0.5
0.3@
0.4
1.0
0.8@
0.3
0.4
0.2@
0.3
0.8
0.7@
@=poor model fit; details of analysis provided in Appendix 9
      For acute exposures, one data set for adults and one for PND11
offspring (both from the comparative ChE study) were available. The
BMD analyses of these data also support similar BMD10 and BMDL10
values across age and sex (1.5-2.6 mg/kg/day for BMDio, 1.3-2.0
mg/kg/day for BMDLio).  As expected, values are higher for acute than for
repeated exposure scenarios, because a larger dose is needed to produce
a similar amount of inhibition following single exposures at a given dose,
compared to repeated dosing.
                          21

-------
2.    Pup Mortality:  Main DNT Study, Comparative ChE Study, &
      Range-finding DNT Study
      a.    Study Results

            Complete reproduction data (i.e. litter size, pup body weight,
      mortality, etc.) from the main DNT study, comparative ChE study,
      and range-finding DNT study, are available in attached DERs
      (Appendices 2-4). There was no difference in litter size or pup
      weight at birth for any groups, with the possible exception of the 6
      mg/kg dose in the range-finding study (there was a slight decrease
      in live litter size on day 1 in that study [12.4 pups per litter
      compared with 14.4 for controls], with no difference in total litter
      size at birth [14.1 at the high dose, 14.4 for controls]).

            Pup mortality data, including both pup and litter incidence,
      are presented in Tables 4-6.  A dose-related increase in pup
      mortality was seen in the 0.5 and  3 mg/kg/day groups of the main
      dimethoate DNT study (23-24 litters/group; see Table 4). No
      increase in pup mortality was seen in the comparative ChE study
      (8-10  litters/group; see Table 5). There was an increase in pup
      mortality at the 6 mg/kg/day dose level  in the range-finding study
      but not at the 3 mg/kg/day level (8-10 litters/group; see Table 6).
      Total litter loss was seen in two litters at the 6 mg/kg dose in the
      range-finding study, and in one litter at the 0.5 mg/kg dose and
      three litters at the 3.0 mg/kg dose in the main DNT study.  Most of
      the pup  mortality in the DNT and DNT range-finding studies was
      seen during early lactation (PNDs 1-11). It is notable that there
      was no excess mortality during the period of direct dosing to pups.
Table 4. Post-natal Pup Mortality - Dimethoate DNT study a
Dose

0 (Control)
live pups
flitters bom

371/24
Days of Lactation"

1-4
10(7)
Ml
3(3)

2(2)
17*21
0(0)
1-21
15(10)
Total
Utter loss
(pups/
Utters!*
0
Meant
dead
pups/ litter
0.6
                          22

-------
0.1  (LOT)
0.5 (MDT)
343/23
360/24
 8(5)
 32* (9)
3(2)
10(3)
0(0)
1(1)
0(0)
0(0)
 11(6)
43" (10)
15/1
            0.5
1.8
3.0 (HOT)
366/24
71* (13)
15(7)
1(1)
2(2)
89* (14)
38/3
3.7
a Includes pups that were found dead, missing and presumed dead, or sacrificed due to poor condition; slightly
revised numbers recently submitted by Cheminova vary from these totals by no more than 1 death/group;
b Number of pups affected (number of litters affected)
* Number of pups humanely sacrificed; additional pups from the affected litters died prior to sacrifice
 * p<0.01, Chi Square test.
                                                 23

-------
Table 5. Post-natal Pup Mortality - Dimethoate Comparative ChE study
Dose
(mgfkjf/day)
0 (Control)
0.1 (LOT)
0.5 (MDT)
3.0 (HOT)
live pups
/litters bom
133/10
142/10
146/10
137/10
Days of Lactation4
t-4
0
2
2
1
541
0
0
0
1
141
0
2
2
2
11-21
NA
NA
NA
NA
1-31
NA
NA
NA
NA
Total
litter toss
(pups'
Ittters)
0/0
0/0
0/0
0/0
Mean*
dead pups/
litter
0
0.2
0.2
0.2
" Number of pups affected
NA=data not available.
Statistical analysis not performed
24

-------
Table 6. Post-natal Pup Mortality - Dimethoate range-finding ONT study '
Dose
(mg/hg/day)
0 (Control)
0.2 (LOT)
3.0 (MDT)
6.0 (HOT)
live pups
/litters born;
144/10
128/9
129/9
113/8
Days of Lactation"
1-4
5(4)
KD
5(4)
38(7)
541
0(0)
0(0)
1(1)
1(1)
1.11
5(4)
1(1)
6(4)
39(7)
11-21
0
0
0
2(2)
1-31
5(4}
1(1)
6(4)
41(7)
Total
litter loss
((Hips'
litters)0
0
0
0
3/2
Meant
dead pups*
litter
0.5
0.1
0.7
5.1
3 Includes pups that were found dead, missing and presumed dead, or sacrificed due to poor condition.
b Number of pups affected (number of litters affected)
c Number of pups humanely sacrificed; additional pups from the affected litters died prior to sacrifice
Statistical analysis not performed
b.    Benchmark Dose Analysis

      As noted previously, recently submitted analyses from the
registrant included BMD analyses of available pup mortality data for
dimethoate (Appendix 13). To confirm these analyses, and to
further explore the dose/response relationship for the increase in
pup mortality in the main DNT study, a separate Benchmark Dose
(BMD) analysis was performed by EPA.  Although analyses
submitted by the registrant included data from several related
studies,  EPA analyses were conducted using data from the main
DNT study only (see Appendix 8 for full details of the analysis).
                    25

-------
      The EPA analysis was conducted using EPA's Benchmark
Dose Software (BMDS; www.epa.gov/ncea/bmds.htm). The
response of pup death due to dosing was modeled using the
available BMDS nested models: NLogistic, NCTR, and RaiVR
(Technical details provided in Appendix 8). A culling event on PND
4 artificially reduced the sizes of the litters making the periods, PND
1-4 and PND 5-11 incomparable. Consequently pup death data for
the two periods were modeled separately. For both study periods,
NCTR and RaiVR produced very similar BMD values.  The minimal
Akaike Information Criterion (AIC) was used to identify the most
parsimonious model within and among model families. Although
the NCTR model resulted  in the smallest AIC for both study
periods, NLogistic had an  only slightly larger AIC, but resulted in
smaller BMD values. Therefore the NLogistic BMD values were
chosen for purposes of comparison with ChE BMD estimates.
BMDs and their corresponding lower limits (BMDLs) were
computed for benchmark responses of 5% and 10% based on the
pup death data from the PND 1-4 period. Computed values for the
BMDs and BMDLs are provided below (Table 7); complete results
of the analysis are provided in Appendix 8.
Table 7. Benchmark Dose Values for Increased Pup Mortality during PNDs 1-4 in the main DNT study
BMOUVf
BMD5
BMDio
BMO(mg/k9/day)
8M&
0.47
0.99
BMDL
0.27
0.57
BMD=Benchmark dose
BMDL=siatistical lower limit on BMD
      The goodness of fit statistic supports an adequate fit of the
NLogistic model for the PND1-4 pup death data. The BMD values
ranged from 0.47 to 0.99 mg/kg/day, depending on the specified
benchmark response. As expected, BMDL values were somewhat
lower than BMD values, ranging from 0.27 to 0.57 mg/kg/day.
BMD/BMDL6 levels for PND1-4 pup mortality are consistent with
study findings for that endpoint, falling between the study LOAEL of
0.5 mg/kg/day and the study NOAEL of 0.1 mg/kg/day. The BMD10
estimates for the PND5-11 pup death data are higher than those for
PND1-4 and outside the dose range for the main DNT study (see
Appendix 8 for details). No analysis of the sensitivity of the
benchmark dose was performed for the PND 5-11 data.

      Pup mortality in control groups from the main DNT,
companion ChE, and range finding study ranged from 0-3.5% of
live bom for PNDs1-4. In the cross-fostering study, conducted
                   26

-------
      using a similar protocol in the same laboratory (see below), pup
      mortality in the control group was very similar: 2.7% of livebom for
      PND1-4. Insufficient information was available to compute pup
      mortality as a percent of liveborn for available historical control
      data.  Based on these background levels for pup mortality, an
      increase of 5% above background (i.e. a BMD5) was considered to
      be the smallest detectable change from background and therefore
      an appropriate BMP for this effect.  Additional support for this
      selection is provided by several analyses in the literature
      (Faustman et al, 1994; Allen et al.,  1994a; Allen et al. 1994b;
      Kavlock et al., 1995), which report that the  use of a BMDL5 for
      developmental endpoints results in values similar to available
      NOAELs within the same studies. Available EPA guidance also
      indicates that a BMR of 5% has typically been used for
      developmental studies (EPA, 2000).

3.    Dimethoate Cross-fostering Study

      In an  attempt to determine whether the increased pup mortality
seen in the dimethoate DNT study could be attributed specifically to pre-
or post-natal exposure to dams, a cross-fostering  study was conducted by
the dimethoate registrant (MRID 46214501, Appendix 6). Dimethoate was
administered by gavage to dams from GD 6 through PND 10, at doses of
0, 3, or 6 mg/kg/day. Pups were cross-fostered on PND 1, to create
groups with no exposure (control), maternal pre-natal exposure only (3
and 6 mg/kg), maternal post-natal exposure only (3 and 6 mg/kg), or
maternal pre and post-natal exposure (6 mg/kg group only). In addition to
pup survival  and reproductive outcome data, detailed observations were
conducted on both pups and dams to evaluate possible treatment-related
toxicity that might contribute to increased mortality. Clinical chemistry and
hematological parameters were also evaluated in  pups on PND11, but
ChE activity  was not evaluated.

      At the 6 mg/kg dose, the timing of deaths appeared related to the
timing of exposure: 24/25 deaths in the 'pre-natal  only' exposure group
occurred during PND 1-4, 22/31 deaths in the 'post-natal' only group
occurred during PND 4-11. In the 'pre- and post-natal' group, 28/38
deaths occurred during PND1-4 and 10/38 deaths occurred during PND 4-
11 (see Table 8).

      In the 3 mg/kg/day 'pre-natal only' and 'post-natal only' groups,
there was a  slight increase in the total number of pup deaths (12 deaths in
controls, 16  deaths in each of the 3 mg/kg groups; see Table 8). There
was no increase in the total number of litters with  pup death, but there was
a small increase in the number of litters with multiple deaths (0 litters with
multiple deaths in controls, 5 in 'pre-natal only' and 3 in 'postnatal only'
                          27

-------
groups treated at 3 mg/kg/day). The increase in litters with multiple
deaths at 3 mg/kg was not as pronounced as that seen at the 6 mg/kg (5
litters in the 'prenatal only', 9 litters in 'postnatal only', and 10 litters in 'pre-
and post-natal' groups; see Table 9).  Note, interpretation of results at 3
mg/kg/day is complicated by two features of the study design: 1) a group
exposed both pre- and post-natally at that dose was not included in the
study; and 2) a control for cross-fostering was not included in the study.

      Results of the hematological evaluation are described in detail in
the DER (Appendix 6).
                           28

-------
Table 8 Pup Mortality in the Cross-fostering study [dead/missing pups (litters)] a
Post-natal Day



No,pap$/
litters born
Dayl"
PND1-4C
PND1-4"
PND 4-7
PND 7-1 1
PND 4-11
Group
1C
Dam:
Control
Pup: Own
litter
mm-  

7(7)
3(3)
10(10)
1(1)
1(1)
2(2)
1A
Dam:
Control
Pup; 3
mg/kg/day*
347/23

7(5)
2(1)
9(6)
3(3)
4(4)
7(7)
1B
Dam:
Control
Pup: 6
mg/kg/day*
382/23

12(6)
12(5)
24(9)
1(1)
0(0)
KD
2
Dam: 3
mg/kg/day
Pup:
Control*
352/23

7(4)
2(2)
9(5)
3(3)
4(4)
7(6)
3A
Dam: 6
mg/kg/day
Pp:
Confrof
ysm*'''' :

3(2)
6(6)
9(7)
6(6)
16(10)
22(12)
3B
Dam: 6
mg/kg/day
Pup: Own
litter
341(22

7(7)
21 (11)
28(14)
6(4)
4(4)
10(6)
29

-------
PND1-11C
PND1-11"
total pup death
as % live birth,
PND1-11
5(5)
12(12)

3.2

9(8)
16(10)

4.6

13(6)
25 (10)

7.1

9(8}
16(11)

4.5

28(14)
31 (15)

8.8

31 (13)
38(16)

11.1

' Data extracted from Appendix 22, pp. 281-286, MRID 46214501
b Includes stillborn and other nonviabie pups
0 Without Day 1 stillborn and other nonviabie pups
d Includes Day 1 stillborn and other nonviabie pups
* Pups from mothers treated with listed dose.
1A - dams in control group fostering pups from dams treated at 3 mg/kg/day
1B - dams in control group fostering pups from dams treated at 6 mg/kg/day
1C - dams in control group rearing own litter
2 - dams treated at 3 mg/kg/day fostering pups from a dam in the control group
3A - dams treated at 6 mg/kg/day fostering pups from a dam in the control group
38 - dams at the 6 mg/kg/day rearing own litter
                                                30

-------
Table 9. Distribution of Pup Deaths in cross-fostering study [no. of dams with dead/missing pups] a
No. of dead/
missing
pups*
No. litters
0
1
2
3
4
5
6
Group
1C
Dam:
Control
Pup; Own
Utter
25
13
12
0
0
0
0
0
1A
Dam:
Control
Pup: 9
nig/kg/day1
23
13
5
4
1
0
0
0
1B
Dam:
Control
Pup:$
mg/kg/da/
23
13
5
2
1
0
1
0
2
Dam: 3
mg/kgftlay
Pup:
Controf
23
12
8
1
2
0
0
0
3A
Dam: 6
mg/kgWay
Pup:
Control"
23
8
6
3
5
1
0
0
SB
Dam: 9
mgrttg/day
Pup:; Own
litter
22
6
6
4
3
2
0
0
31

-------
* Data obtained from Appendix 22, pages 281-286. MRID 46214501.
b Includes stillborn and other nonviable pups
c Pups from mothers treated with listed dose.
1A - dams in control group fostering pups from dams treated at 3 mg/kg/day
1B - dams in control group fostering pups from dams treated at 6 mg/kg/day
1C- dams in control group rearing own litter
2 - dams treated at 3 mg/kg/day fostering pups from a dam in the control group
3A - dams treated at 6 mg/kg/day fostering pups from a dam in the control group
3B - dams at the 6 mg/kg/day rearing own litter
                  There was no indication of treatment-related maternal toxicity or
           clinical signs during gestation; there were no treatment-related clinical
           signs and no difference in body weight/body weight gain among treatment
           groups (see DER, appendix 6). However, during the lactation period,
           there was a higher proportion of dams exposed at 3 and 6 mg/kg/day
           showing restlessness on 2 days or more, regardless of whether they were
           rearing their own litters (group 3B) or control offspring (groups 2 and 3A)
           (Table 10).  Scattering of offspring in the cage on two or more days of
           lactation was also increased in dams exposed during the study (2, 3A &
           3B) at 3 and 6 mg/kg/day.

                  The number of pups with the umbilicus still attached during the
           early perinatal period (PND  1) was increased in groups 16 (15), 3A (10),
           and 3B (13) compared to controls (4) (Table 11). However, the  number of
           pups with umbilicus attached after PND 1 was similar between control and
           treated groups and therefore did not suggest reduced maternal care of
           pups, even after cross-fostering.

                  During lactation, the incidence of pups with "no milk in the stomach"
           was increased in groups 2, 3A and 3B (15, 28, and 11, respectively)
           compared to control (group 1C: 4) (Table 11). It is noted that this finding
           is supported by pup necropsy results, specifically that the incidence of
           pups found dead after cross-fostering with no milk in the stomach was
           increased in groups 2, 3A, and 3B (14,  24, and 28 pups, respectively), as
           well as in group 1B (12 pups),  compared to control (group 1C: 7 pups)
           (see Table 19 in the DER).  It cannot be determined whether this finding is
           due to adverse effects on the dams or on the pups.
                                       32

-------
Table 10. Incidence of maternal restlessness and offspring scattering"
Observation;



Nu mber of dams or litters affected J# observations^ days]
1C (n*25)
Dam:
Control
Pup: Own
litter
1A (n23)
Dam:
Control
Pup: 3
mg/kg/da/
1B{n-23)
Dam: Control

Pup: 6
mg/kgWay6
2 (n-23)
Dam: 3
mg/kg/day
Pup: Control"
'-:?
3A (n=23)
Danes
mg/ke/day
Pupr
Con4rotb

3B(n22)
Dam: 6
mg/kg/day
PUp: Own
Utter
Wxss'r '''%''};<',',
1I:;^T^
* * "* yjKKyy$$$&$ &>&^&K3*;>Av;v;>v;><^ ;S$K*M*SKJJ% ^^^i$'>^i$f'>Kv^ v^i^A^tx}^ ixvj'K'-'S'&KCjSijSJsiy'Si.j. w ./ y" * < ^ "v "^
''"''' '^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^' '"'; '/T' "'"
ODays
1 Day
2-3 Days
4 or more
days
15
7
[7 obs/7
days]
3
[7 obs/6
days]
0
13
10
[11 obs/1 0
days]
0
0
15
7
[8 obs/6 days]
1
[4 obs/2 days]
0
8
6
[8 obs/6 days]
8
[20 obs/1 9
days]
1
[6 obs/4 days]
3
5
[6 obs/5 days]
13
[40 obs/33
days]
3
[19 obs/1 2
days]
1
6
[8 obs/5 days]
12
[41 obs/32
days]
3
[24 obs/1 4
days]
'fj^W^^^f;/^ isr:)$;;y;: 3? -t V i::f if^^ltB'^H ; '-; ^
; ' ',M'-' / v- "''** ', J ' / '' ; ; ; 5 r "/ v' ' '' s " i , '' -' "!v rr " / ", ;;  i;: '- / ^ ; >"$, ', ;, 2r lx/v? ';';'"'"'' , , ,'
'ft,-*',, ' f *, f J ' ' , * , ', *Jf-.f V '*',, "'f^, 'ffl, ", ,-St'lt > 1 , '' ' ft t '
. * > * ' ' *, t f \ ' 'f ''!:''' "'/' " "' ' "'f ' "', ' ' ' / ' * " ' $""' " '* ^' '''''>>.''}'$ r ' '' ' '"' "<
33

-------
ODays
1 Day
2-3 Days
4 or more
days
13
6
[6 obs/6
days]
6
[17obs/14
days]
0
13
3
[3 obs/3
days]
4
[10obs/10
days]
3
[19obs/13
days]
9
6
[9 obs/7 days]
6
[14obs/12
days]
1
[4 obs/4 days]
6
4
[6 obs/5 days]
8
[29obs/19
days]
5
[36 obs/28
days]
1
3
[4 obs/3 days]
12
[53 obs/33
days]
7
[37 obs/30
days]
5
4
[6 obs/5 days]
7
[23obs/17
days]
6
[37 obs/29
days]
1 Data obtained from Table 19, page 84, MRIO 46214501.
b Pups from mothers treated with listed dose
1 A - dams in control group fostering pups from dams treated at 3 mg/kg/day
1 B - dams in control group fostering pups from dams treated at 6 mg/kg/day
1C- dams in control group rearing own litter
2 - dams treated at 3 mg/kg/day fostering pups from a dam in the control group
3A - dams treated at 6 mg/kg/day fostering pups from a dam in the control group
3B - dams at the 6 mg/kg/day rearing own litter
34

-------
Table 11. Observations of umbilicus attached and no milk in the stomach of pups (PND 1-11)a
Clinical Sign
                     Group
                        Dam:  -:
                       Control

                      Pup: Own
                        litter
    1A
   Dam:
  Control

  Pup: 3
mgftg/day"
Control
 Pup;{f
    2
  Dam: 3
mgfltfl/day

   Pup:
 Control"
   3A
 Dam: 6
mg/kgfttey

  Pup;
 ConW
   38
 Dam: 6
mg/kg/day

Pup: Own
  litter
PND1
                15
                          10
                           13
PNDs 2-4
after PND 4
Incidences of no milk in
stomach"
                       15
                      28
                       11
Number of litters
                                   13
                                            35

-------
Specific days of
observation
2, 4, 5, 7
4,5,8
1,5,7
2,3,4,5,6,7
1 ,3,4,5,6,7,
8,9
3,4,6,7
' Data extracted from Appendix 20, page 245-265,and Appendix 21, pp 270-279, MRID 46214501
b Pups from mothers treated with listed dosa
c Number of observations of umbilicus still attached after cross-fostering of litters. Observations of offspring not
allocated to cross-fostering include umbilicus attached for 15 control pups on PNO 1, none for 3 mg/kg/day litters,
and 5 pups for 6 mg/kg/day litters pre-PND 1.
tf Observation of one or more pups in a litter with the finding at a scheduled observation time.
1A - dams in control group fostering pups from dams treated at 3 mg/kg/day
1B - dams in control group fostering pups from dams treated at 6 mg/kg/day
1C- dams in control group rearing own litter
2 - dams treated at 3 mg/kg/day fostering pups from a dam in the control group
3A - dams treated at 6 mg/kg/day fostering pups from a dam in the control group
3B - dams at the 6 mg/kg/day rearing own litter
                          In summary, pup mortality in the cross-fostering study was
             increased after cross-fostering, days 1-4, and again on post-natal days 4-
             11, suggesting that pup mortality increased regardless of pre- or post-
             natal maternal exposure to dimethoate at 3 or 6 mg/kg/day.  Post-natal
             deaths appeared to be correlated to some extent with  the incidences of
             maternal restlessness and litter scattering for groups 2, 3A, and 36.
             However, the cause of maternal behaviors and  their contribution to pup
             mortality cannot be determined.  A combination of pre- and post-natal
             exposure to pups and/or dams contributed to the  observed pup mortality.
                                          36

-------
4.    Standard Reproductive Toxicity Studies in Dimethoate and
      Omethoate

      EPA and PMRA have evaluated the available data related to pup
survival from three reproductive toxicity studies in dimethoate (2 full
studies and one range-finding study) and two reproductive toxicity studies
in omethoate.  Results from dimethoate studies are presented below;
results from omethoate studies are presented in Appendix 12. The
purpose of this evaluation is to further characterize available pup
mortality/survival data and related ChE inhibition in pup/offspring and/or
maternal animals. In available dimethoate reproductive toxicity studies,
dimethoate was administered in the feed {Appendices 7,10,11). Results
of the BMD analyses of the ChE data from these studies have been
previously discussed (see above).  Results related to pup mortality are
presented in Tables 12, 13, and 14, below.
                           37

-------
Table 12. Pup Death Data from Ditnethoate first dietary reproductive toxicity study (MRID 42251501).

live pttpa? litters
bom
teatti
PNDO-4

-------
Table 13, Pup Death Data from Dimethoate second dietary reproductive toxicity study (MRID
46181001).
:v-;gg}^;'^v:i::^^
ii^l^i^;';;..'
;MiiP]^t
;-pm$4:^
.' '^ '':-':. ':::':" :;"'::: :":;:':
;il!|l ;;:;.:.;:
$r$iifii ]
mmm*::
B!$ilf&:;v:
FO/first mating
control
0.2 mg/kg
1.0mg/kg
6. 5 mg/kg
206/21
244/24
259/25
234**/23
0/0
1/1
2/2
3/3
1/1
0/0
0/0
0/0
1/1
0/0
0/0
0/0**
1/1
1/1
2/2
3/3
FO/second mating
control
0.2 mg/kg
1.0 mg/kg
6.5 mg/kg
262/24
232/22
244/21
251/25
2/2
1/1
1/1
10/4
0/0
0/0
0/0
0/0
1/1
0/0
0/0
0/0
3/3
1/1
1/1
10/4
F1 /first mating
control
0.2 mg/kg
1.0 mg/kg
6.5 mg/kg
247/22
251/24
281/24
261/25
8/3
8/6
5/4
1/1
1/1
0/0
1/1
0/0
1/1
0/0
1/1
0/0
9/4
8/6
6/5
1/1
F1 /second mating
control
0.2 mg/kg
1 .0 mg/kg
6.5 mg/kg
256/22
285/25
248/22
265/25
9/7
4/4
3/2
6/5
0/0
2/2
1/1
11/3
0/0
3/3
1/1
12/4
9/7
7/6
4/2
18/6
a Data were obtained from pages 627 - 634 and 1005 - 1012 of the study report.
'values represent total pup deaths/total litters with pup deaths. Pup deaths included cannibalized pups.
** Pup which suffered from accidental death on Day 17 is omitted.
      In a one-generation range-finding reproductive toxicity study,
treatment-related effects were seen on several reproductive parameters,
including decreased number of implantation sites, increased post-
implantation loss, and decreased litter size at birth.  These dose-related
effects were seen in all treatment groups, at doses ranging from 3.9-7.5
mg/kg/day (50, 75,  or 100 ppm in the diet). Although there was no
increase in pup mortality between PND1-4, there was an increase in
mortality from PND4-21 at doses of 75 and 100 ppm (5.8 or 7.5
mg/kg/day, respectively); this finding was also reflected in a lower lactation
index at these doses, as well as a continued decrease in mean litter size.
                           39

-------
            Pup mortality findings from this study are summarized in Table 14; the
            complete study review (DER) is provided in Appendix 11.
Table 14. Pup Death Data from Dimethoate dietary one-generation range-finding reproductive toxicity
study.{MRID 46348201)
                         litters born
                                   deatns
                                   PNDS-21
                      totaKfeaths
                      PNDO-21
FO
control
153/10
50 ppm (3.9 mg/kg)
137/9
75 ppm (5.8 mg/kg)
140/10
13
13
14
100 ppm (7.5 mg/kg)
132/10
                       10
* values represent total pup deaths/total litters with pup deaths
                  Dose-related decreases in brain ChE activity were seen in both of
            the dimethoate two-generation reproductive toxicity studies (see Tables 15
            and 16) and in the one-generation range-finding reproductive toxicity study
            (see Table 17). BMD analyses conducted using these data have been
            presented above, and result in BMD values similar to those seen in the
            DNT companion ChE study. Due to differences in exposure (lack of direct
            exposure to pups during lactation due to dietary route of administration),
            data suitable for comparing young and adult dose/response curves are not
                                       40

-------
available from these studies. However, based on the absence of a clear
increase in pup mortality at doses up to 6 mg/kg/day (doses resulting in
approximately 60% brain ChE inhibition),  it is clear that the increase in pup
mortality seen in the DNT study is not solely due to maternal or fetal ChE
inhibition.

      Relevant data from available reproductive toxicity studies for
omethoate (the active metabolite of dimethoate) were also examined (see
Appendix 12). Two multi-generation reproductive studies  are available:
the drinking water study found pup mortality at the highest dose tested (1
mg/kg/day), most notably in the second generation. In a feeding study
conducted at doses up to 0.5 mg/kg/day of omethoate, small increases in
pup mortality were noted in the second generation (note: significant
deficiencies were noted in the study protocol of the feeding study).  The
increases in pup mortality seen in these studies provide support for the
findings in the main DNT study, the range-finding DNT study, and the
cross-fostering study, but differences in study  design preclude any  direct
comparison of dose/response relationships.
                           41

-------
      In summary, there is some indication in available dimethoate
reproductive toxicity studies of offspring toxicity, manifested as decreases
in pups and/or litters born or as increased pup mortality. However, effects
seen in these studies generally occurred at higher doses than similar
effects in the DNT study. Increased pup mortality was also seen in
available reproductive toxicity studies for omethoate (see Appendix 12).
Taken together, these results support the determination that treatment-
related increases in offspring mortality can occur following exposure to
dimethoate, but provide conflicting information regarding the doses at
which that effect occurs.
                           42

-------
    I
f
IS
.2-

I
hoale
I

S
       f'l
                     U.Z


                     CO CN
                     cn 10

                     ci r^
                     h- >


                     CN CN
U. S


etc.


!
Ci T-


EP
id co
                               -2-
cp co

CN CN
                             ss
                             H
                                     U. S



                                     ..
                                     U.S



                                     CN CO
        mm
        O cp
                                             U.Z
0 Ci


J^ I

CN rt
CNCN
                      i jff ^
                        Sfff

                      . " '
ion
                                                                                 .

                                                                                 1
                                                                                 if

-------


1
?I
           5

          I
                 u. S

                 s
                 ci d
                 u.2


                 SS
                                u.2


                                #?
U_ S


88

ci o
                           *- CN
                           & Ci



                           ii
                           10 K
U_5


(D n
^- o>

T^ O

 H  

to (o
55 r)

CN N
                           SS


                           Sfe
                           o o




                           ID <7)
                           f^- O
                                           8
               |-

               |
                             .S

                              CN
                                                  as;
                                                  o o
                                            g

                                            ft
                                                                   I
                                                                   I

-------
                                                                                   u.2
                                                                                                                                                            in
.1
E
                                                                                   2
                                                                                   .a
                                                                                                                    g-
                                                                                                                    I

                                                                                                                    d.

-------
01
S
                             u_5
                             O 
-------
5.    Discussion

      Increased pup mortality was seen in several recently conducted
studies following maternal exposure to dimethoate during gestation and
early lactation. However, the dose at which the pup mortality was
observed varies among the studies. The doses at which increased
mortality is seen in the main DNT study (0.5 and 3 mg/kg/day) are lower
than that at which the effect is seen in other dimethoate studies conducted
using similar exposure regimes (6 mg/kg/day). No increase in pup
mortality was seen at the 0.5 mg/kg/day in the comparative ChE study and
at 3.0 mg/kg/day in the comparative ChE or range-finding studies. The
reason for this difference in results is  unknown, but may be related to the
smaller number of litters/group used in the comparative ChE and range-
finding studies (8-10/dose, compared with 23-24 in the main study) and,
therefore, less statistical power to detect the effect. The effect seen at 6
mg/kg in the DNT range-finding study was found at that same dose in the
cross-fostering study. In the cross-fostering study, although a 3 mg/kg
dose was administered, no pups  were exposed at that dose both in utero
and during lactation, as were the pups in the main DNT study,
comparative ChE, and range-finding studies.

      a.    Relationship Between Maternal Toxicity and Pup
            Mortality

            It has been suggested that the increase in pup death seen
      during early lactation following maternal exposure to dimethoate is
      attributable to maternal toxicity (see discussion in trip audit memo,
      3/22/2001, Appendix 5). Although there was no indication of overt
      maternal toxicity in the main dimethoate DNT study, there is
      significant ChE inhibition at both the 0.5 and 3.0 mg/kg/day doses
      in both dams and fetuses in the comparative ChE inhibition study.
      Since both pups and dams have been exposed in the DNT study
      design, it is not possible to separate effects on pups from effects on
      dams.  Consequently EPA and PMRA conclude that data from the
      main DNT, range-finding, and comparative  ChE inhibition studies
      do not support a determination regarding the dams as  the exclusive
      cause of the pup mortality seen in the main DNT study.

            To obtain further information regarding a possible link
      between maternal toxicity  and pup mortality, the registrant
      conducted the cross-fostering study described above.  In that study,
      detailed behavioral observations were conducted to look for
      symptoms of maternal toxicity that may not have been apparent in
      the cageside evaluations conducted in the main DNT study. In
      addition, separate groups  of pups were exposed exclusively pre-
      natal ly (in utero only) or post-natally (via maternal dosing), in order
                          47

-------
to characterize what factor(s) may lead to pup mortality. Although
ChE inhibition was not measured in the cross-fostering study, the
exposure and dosing regime were similar to that used in the DNT
range-finding study.

      Results from the cross-fostering study showed that
increases in pup mortality could result from either pre- or post-natal
exposure to the dams. The number and timing of the pup deaths
appeared to vary depending on the timing of the dam's exposure.
At the 6 mg/kg level: in the 'pre-natal1 group, there were 25 deaths,
24 of 25 deaths during PND1-4; in the 'post-natal' group; 31 deaths,
22 of 31 during PND 4-11; and in the 'pre- and post-natal' group, 38
deaths, spread throughout PND1-11.  The increased incidence of
death during the period immediately following birth (i.e. on day 1,
prior to cross-fostering) is consistent with the pattern seen in prior
studies. No clear-cut increase in pup mortality was seen at the 3.0
mg/kg/day dose in the cross-fostering study, possibly because of
the lack of combined pre- and post-natal exposure to pups.

      Although cholinergic signs were not observed in the dams
during the main DNT study at any dose, based on the results of the
comparative ChE  inhibition study significant cholinergic inhibition in
the maternal brains is expected and could have impacted maternal
behavior.  In order to address possible maternal toxicity not
detected by routine clinical observations, detailed observations of
maternal care were conducted for dams in the cross-fostering study
(note: behavioral observations were performed by individuals aware
of the treatment received by animals they were observing, raising
the possibility of observer bias). There was an increase in
incidence of several observations in treated dams and pups
(maternal restlessness and scattering of offspring, no milk in
stomach of pups,  and umbilicus attached in pups).

            Careful examination of the data indicated no clear
relationship between these observations and increased pup
mortality within the affected litters. First, there was an increase in
pup death in the 6 mg/kg 'pre-natal only' group, but no increase in
scattering when compared to controls. In addition, the increase in
restlessness/scattering in dams treated with 3.0  mg/kg was similar
in magnitude to the increase in symptoms in dams treated with 6.0
mg/kg, with a much smaller increase in pup death in the 3 mg/kg
'post-natal only' group.  Evaluation of in-life observations of 'no milk
in the stomach of  pups' leads to similar conclusions; the incidence
is increased in pups of all dams receiving post-natal treatment (on a
pup or litter basis), regardless of dose, and thus cannot be
specifically related to the increase in pup death (Tables 10  &  11).
                     48

-------
An increase in the number of pups with umbilicus attached on PND
1 was seen in pups treated at 6 mg/kg, either pre- or post-natally,
but in no other group. However, the number of pups with umbilicus
attached after PND 1 was similar between control  and treated
groups and therefore did not suggest reduced maternal care of
pups, even after cross-fostering.

      Decreases in maternal care may originate from effects in the
pups or the dams (i.e. ill mothers neglect their pups, or mothers
may neglect their ill pups).  Actual exposure of pups to dimethoate
or its metabolites via lactation, following maternal exposure to
dimethoate, has not been quantified. Available data from the
companion ChE inhibition study indicate similar levels of ChE
inhibition in fetuses and dams following in utero exposure; lower
levels of inhibition found in  PND 4 pups could suggest less
exposure via lactation, but the actual exposure levels are unknown.
Detailed pup observations and clinical chemistry/ hematology
evaluations conducted in the cross-fostering study provide some
indication of pup toxicity in addition to the increase in  mortality.
There was an apparent delay in development of the surface righting
reflex (the righting reflex took more than 3 seconds) in male and
female pups reared by 6 mg/kg treated dams (3A) and in pups of
dams dosed at 6 mg/kg (3B). Increases in several hematologicai
parameters (most consistently hematocrit and mean red cell
volume) were seen in pups reared  by 6 mg/kg treated dams, and
increases in urea were seen in pups from dams treated post-natally
at 3 or 6 mg/kg (Appendix 6). These results reinforce the finding of
pup toxicity in dams treated with dimethoate, but do not provide
information regarding the cause of that toxicity. At this time,
available data  regarding the role of maternal toxicity in the pup
mortality seen  following dimethoate exposure do not allow for a
definitive conclusion on this issue,  but support a contribution of both
pre-  and post-natal exposure.

b.    Litters vs. Pups as the Unit of Analysis

      Most sources regard the litter as the most appropriate unit of
statistical analysis for developmental toxicity studies and related
studies, e.g. dominant lethal studies, where exposures are purely
pre-natal and the dose to all fetuses in the litter are largely
determined by the maternal variables, and also where the fetus
depends on the mother for life.  Pups from a single dam are
considered to be closely related in  terms of their exposure in utero,
and thus lack statistical independence.
                     49

-------
      Following weaning in a reproduction study, when the pups
are directly exposed after lactation, it seems clear that the
individual pup again becomes the appropriate unit of analysis.
Between birth and lactation, independent living and exposure
depend not only on the dose to the lactating dam and maternal
behavior and health, but also on the behavior/health of each pup in
successfully suckling and on the amount the pup consumes.  Thus,
determination of the most appropriate statistical evaluation during
that time period is more complex.

      In the dimethoate DNT study, there was an increase in the
number of litters with pup death at 3.0 mg/kg (14 litters with at least
one pup death, as opposed to 10 control litters); at 0.5 mg/kg, the
number of affected litters (10 with at least one death) is the same
as for controls. When the total number of pup deaths is compared
across groups, there is an increase at 0.5 mg/kg (43 deaths), when
compared with control (15 deaths) or low dose (11 deaths) groups.
The increase in number of deaths is seen during both the PND 1-4
and PND 5-11 intervals, at  both 0.5 and 3.0 mg/kg doses.
Evaluation of pup death as mean number of dead pups per litter or
as percent of livebom also shows a dose-related increase. The
increase in pup death is statistically significant at the mid-dose (0.5
mg/kg/day) when evaluated as the total number of deaths, but not
when evaluated as a litter effect.  A BMD analysis of pup mortality
from the main DNT study resulted in a BMDL5 of 0.27 mg/kg/day,
falling between the mid and low doses, and consistent with the
NOAEL/LOAELs  determined for the study.  All of the candidate
BMDS models considered for estimating BMD values for pup death
provide options for modeling pup death data at the individual level
or the litter level.  In the model selected for use in this analysis, no
litter specific covariates were specified, however inter!itter
correlation (i.e. a litter effect) was modeled.

      In the cross-fostering study, there is also an increase in the
number of litters with multiple pups dying, as detailed in Table 8
(distribution of number of deaths  by litters):

Q    No control litters have more than one death

Q    There are  5 litters with multiple deaths in pups exposed to
      3.0 mg/kg  only pre-natally

Q    3 litters with multiple deaths in pups exposed to 3.0 mg/kg
      only post-natally
                    50

-------
Q    5 litters with multiple deaths in pups exposed to 6.0 mg/kg
      only pre-natally

Q    9 litters with multiple deaths in pups exposed to 6.0 mg/kg
      only post-natally

Q    16 litters with multiple deaths in pups exposed to 6.0 mg/kg
      both pre- and post-natally.

Thus, although  the number of litters with at least one pup death
was increased only in the two groups of pups exposed post-natally
at 6 mg/kg  (the 'post-natal only' and 'pre-and post-natal' 6 mg/kg
groups had at least one death in 15-16 litters, compared to 12 litters
for controls and 10-11 litters in other treatment groups), the number
of litters with multiple deaths was increased in all treatment groups.

      Another issue raised with respect to the analysis of the pup
mortality seen in the dimethoate DNT study was the inclusion in the
analysis of pups from litters that were humanely sacrificed. In the
main dimethoate DNT study,  a substantial portion of the total pup
deaths at the mid dose can be attributed to a single incident of total
litter loss occurring at that dose (15 of 43 pup deaths at 0.5 mg/kg
occurred due to litter sacrifice; two additional pups from that litter
died prior to the litter sacrifice). Three dams with total litter loss
were also observed at the high dose (3.0 mg/kg), with 38 of 89 pup
deaths at that dose attributed to humane sacrifice. There were no
instances of total litter loss in the control or low dose groups.
Although total litter loss is sometimes excluded from evaluation of
pup survival, the EPA and PMRA included them in this instance.

      Total litter loss is uncommon in the relevant historical control
data base for the performing laboratory: for 11 studies, conducted
between September, 1996 and August, 1999, one study showed
total litter loss in 2/25 litters; in the other 236 litters, there were no
instances of total litter loss. Historical control data recently
submitted from  5 additional studies conducted from  October, 2000
to September, 2002 show total litter loss in 1/24 litters for each of
two studies (both conducted in 2002), with no litter loss in the other
3 studies. Thus, the  loss of 3 litters at the high dose in the DNT
main study (conducted in 2000/2001) is outside the  control range
for this species and strain, and exceeds that seen in 16 available
historical control studies.

      The sacrificed pups were moribund and were from litters
where other pups had already died (two pups in the mid-dose litter
had died prior to sacrifice; at the high dose 11/14 pups from one
                     51

-------
litter, and 2/16 pups from a second litter had died prior to sacrifice).
Observations conducted during the study indicate that all pups from
the sacrificed litters (both mid-dose and high dose) were cold to
touch, underactive, and had little food in their stomachs (Appendix
5).  EPA and PMRA must rely on the judgement of the investigator
that the pups were moribund and that the sacrifice of these pups
was a humane procedure to prevent additional suffering. It is,
therefore, appropriate to consider these pups as having been
rendered fatally ill, assume they would have died nevertheless, and
include them in the total count of pup deaths. While use of such a
procedure does introduce a measure of uncertainty about what
otherwise might have been the fate of these pups, if such
procedure was, as EPA and PMRA believe, uniformly applied, there
is no reason to expect that it has biased the results.

      c.    Pup Mortality

      Since developmental effects, including increased mortality,
have been shown to occur as a result of single exposures during
development (USEPA, 1991), this endpoint (increased pup
mortality) is considered to be appropriate for use in risk assessment
for single dose exposures. Results from the cross-fostering study
at the 6 mg/kg dose, which demonstrated an increase in pup death
following pre-natal only, post-natal only, or combined pre- and post-
natal exposure, provide no data to contradict the assumption that a
single exposure to dimethoate at 6 mg/kg could result in increased
pup mortality. However, a clear effect on pup mortality was not
seen in  the cross-fostering study following pre- or post-natal
exposure to 10 or 15 doses of 3.0 mg/kg.  These results suggest
that the increases in pup death seen at the 0.5 or 3.0 mg/kg/day
doses in the main DNT study were not due to a single exposure
during a critical window, but were more likely a result of continuing
exposure throughout gestation and early lactation. Thus, the
assumption that the increase in pup mortality seen at the 0.5
mg/kg/day dose could have been the result of a single maternal
exposure is not supported by the results of the cross-fostering
study.

      d.    Brain Cholinesterase Inhibition: Protective for
      Pup Death Endpoint?

      In the main DNT study, increases in pup mortality were seen
at doses that also caused inhibition of brain ChE following repeated
dosing.  BMD analyses for pup mortality and brain ChE were
conducted to provide more information regarding the relationship
between brain ChE inhibition and pup mortality, as well as to
                    52

-------
evaluate the relative sensitivity of young and adult animals to brain
ChE inhibition following dimethoate exposure.

      Evaluation of results from the BMD analysis for the
repeated-dose studies reveals no age-related difference in
susceptibility to dimethoate-induced brain ChE inhibition. BMD10
and BMDL10 values for brain ChE inhibition following repeated
dosing are very similar across age groups (range 0.20-1.0
mg/kg/day for BMD10 and 0.2-0.7 mg/kg/day for BMDL10). Low end
BMDLio values for brain ChE inhibition (0.2 mg/kg/day) are lower
than the BMDLio values for pup mortality (0.6 mg/kg/day) and are
comparable to BMDLs  values for pup mortality (0.3 mg/kg/day).

      BMDLio estimates for brain ChE  inhibition following  single
doses of dimethoate are  also comparable for pups and adults (1.3-
2.0 mg/kg). Although these doses are higher than the BMDLs and
BMDLio for pup mortality in the main DNT study, they are lower
than the dose producing  minimal effects in relation to pup death in
the cross-fostering study (3.0 mg/kg/day following isolated  pre-natal
or post-natal exposure).

      Although these  results provide no information to support
brain ChE inhibition as a cause of pup mortality, comparison of the
results of the BMD analyses for pup mortality and brain ChE
inhibition provide a reasonable basis for the conclusion that brain
ChE inhibition occurs at doses similar to or lower than those
causing increases in pup mortality, given similar exposure
scenarios.

These analyses lead to the following conclusions:

1)    The BMD analyses indicate that dose/response relationships
      for brain ChE inhibition are similar across age groups;

2)    Brain ChE inhibition occurs at doses similar to or slightly
      lower than those causing increases in pup mortality;

3)    The BMDLio for brain ChE inhibition is similar to the BMDL5
      for increased pup  mortality following repeated dosing.

4)    The BMDL10 for brain ChE inhibition is lower than the dose
      assumed to be associated with pup mortality following acute
      dosing

Based on this analysis: (1) dimethoate risk assessments based on
a lack of inhibition of brain ChE will be protective against pup
                    53

-------
mortality for acute and repeated-dose scenarios; and (2) for
dimethoate, use of adult ChE inhibition endpoints will be protective
of effects in the young, for exposure routes where age-specific data
are not available.
                     54

-------
4ft
                 D.    Weight of the Evidence and Summary

                       Dimethoate is an OP that exerts its neurotoxicity by binding to, and
                 phosphorylation of, the enzyme acetylcholinesterase in the central (brain) and
                 peripheral nervous systems. Dimethoate undergoes oxidative desulfuration to its
                 more potent ChE inhibiting oxon (omethoate).  In the rat, approximately 5% of
                 dimethoate is converted to omethoate.

                       The critical effect1 for many OPs is the inhibition of ChE (in the brain or
                 blood compartment). However, in the case of dimethoate, both ChE inhibition
                 and pup mortality are its critical effects. This is a unique finding for this OP
                 because pup mortality has not been found to be the critical effect or the lowest
                 observed adverse effect, for other OPs based on two-generation reproductive rat
                 studies  and on the rat DNT studies submitted and reviewed to date.

                       Pup mortality as a critical effect for dimethoate was first observed in a rat
                 DNT gavage study.  In the main DNT study, there is a statistically  significant and
                 dose-related increase in total pup mortality at the 0.5 and 3.0 mg/kg/day dose
                 groups when pups are evaluated as individuals. Similarly, although not
                 statistically evaluated, a dose-related increase in mean pup mortality/litter was
                 observed. No effects on pup mortality are found at 0.1 mg/kg/day. Most of the
                 deaths occur on PND  1-4.

                       The litter is typically considered the appropriate unit of analysis for
                 developmental and reproductive toxicity studies. For study designs such as the
                 DNT, however, where post-natal pup survival is dependant on the behavior and
                 health of both the dam and the pup, it is important to consider both the litter and
                 the individual pup separately as units of analysis.  As seen in the 0.5 and 3.0
                 mg/kg/day dose groups of the main DNT study, when the whole litter(s) from one
                 or two dams die or are humanely sacrificed, it is important to evaluate the impact
                 of whole litter loss on the total pup mortality and to consider the degree to which
                 this entire litter loss influences the total count.  In the 0.5 and 3.0 mg/kg/day
                 groups, the whole litter from one and three dams,  respectively, died or were
                 humanely sacrificed. These losses accounted for 15 of 43 pups and for  38 of 89
                 pups in  the 0.5 and 3.0 mg/kg/day dose groups, respectively. This represents
                 mean pup mortality/litter of 1.8 for the 0.5 mg/kg/day dose group and 3.7 for the
                 3.0 mg/kg/day dose level, whereas the control group only shows that 0.6
                 pups/litter died. Removal of whole litter losses from the mean/litter calculations
                 result in 1.2  and 2.4 for the 0.5 and 3 mg/kg/day dose groups, respectively.
                 Thus, on recalculation of the mean pup mortality/litter without the pups from
                 whole litter losses, a dose-related  increase  in pup mortality is still observed.
           1A critical effect is one considered the most sensitive endpoint from the most appropriate species.

                                                  55

-------
      It is important to point out that total litter loss appears to be an infrequent
occurrence in the DNT study.  In the historical control data, which included 11
studies conducted between September, 1996 and August, 1999, only one study
showed total litter loss in 2/25 litters; in the other 236 litters from these 11
studies, there were no instances of total litter loss.  In more recent studies
conducted from October, 2000 to September, 2002, however, total litter loss was
observed in 1/24 litters for each of two studies (both conducted in 2002), no litter
loss was observed in the additional three studies.  Although two of these five
recent studies show whole litter loss in untreated animals, overall the historical
control data provide evidence for low incidence of whole litter losses in  untreated
rats. The BMD analysis includes both the individual pups and the litter  as part of
the model, taking into account both these variables.

      Although the pup mortality observed at both the 0.5 and 3.0 mg/kg/day
dose levels in the DNT study appears to be dose- related, and thus treatment-
related, this finding is not supported by other studies that had similar exposure
regimes (repeated gavage dosing at similar dose levels). In the comparative
ChE inhibition study,  no pup mortality was observed at any dose (i.e., 0.1, 0.5
and 3 mg/kg/day). In addition, the range-finding study showed no increased pup
mortality at 0.2 or 3.0 mg/kg/day. An increase in pup mortality was observed
(total litter loss = 2 of litters) at the highest dose tested (6.0 mg/kg/day)  in this
study.  In the cross-fostering gavage study of dimethoate, although a slight
increase in total number of pup deaths was observed at 3.0 mg/kg/day  following
either pre-natal only or post-natal only exposure, the results at this level are
difficult to interpret. Lastly, pup mortality was found at 6 mg/kg/day following pre-
natal only, post-natal only, and combined pre- and postnatal exposure in the
cross-fostering study.

      The rat multi-generation reproductive studies on dimethoate and
omethoate are important to evaluate given that exposure extends over  the entire
period of development up to sexual maturation, and viability is evaluated.
Although doses are not comparable for dimethoate and omethoate, similar
effects on pup survival are seen for both chemicals. Both two-generation
reproductive toxicity studies on dimethoate are dietary studies; similar high doses
were used, of approximately 6 mg/kg/day.  No clear increase in pup death was
seen in either study; however a reduction in live births was seen in one study at
the 6 mg/kg/day dose level. In a one-generation range-finding reproductive
toxicity study with dimethoate, dose-related changes in reproductive parameters
were seen starting at 3.9 mg/kg/day (decreases in implantation rate and litter size
at birth, increases in post-implantation loss), and increases in pup mortality were
seen at doses of 5.8 and 7.5 mg/kg/day. Two multi-generation reproductive
toxicity studies are also available for omethoate (see Appendix 12).  A drinking
water study found pup mortality at the highest dose tested (1 mg/kg/day), most
notably in the second generation.  In a feeding study conducted at doses up to
0.5 mg/kg/day of omethoate, small increases in pup mortality were noted in the
                                 56

-------
second generation (note: significant deficiencies were noted in the study protocol
of the feeding study).

      The association between pup mortality observed in the DNT study and
brain ChE inhibition is unclear. Following treatment with the lowest dose
producing pup mortality (0.5 mg/kg/day) in the DNT study, only minimal brain
ChE inhibition was found in the GD 20 dams (10%), fetus (10%) and the PND 4
pups (8%) for males. At the next highest dose (3 mg/kg/day), there was more
pronounced brain ChE inhibition (dams 60%, fetus 33%). The small amount of
brain ChE inhibition (7-13%) in PND 4 pups does not support a link between pup
mortality nor a "burst" of exposure to dimethoate via lactation. The association
between brain ChE inhibition as a causative factor in the pup deaths is also
called into question by the results of the comparative  ChE and range finding
studies. In the comparative ChE study no pup mortality was observed, but the
highest dose tested (3 mg/kg/day) produced pronounced brain ChE inhibition in
the dams (60%) and fetuses (33%) (albeit, minimal inhibition (13%) was found in
the PND 4 pups). In the range-finding study, no pup mortality was found at 3
mg/kg/day of dimethoate although greater than 70% brain ChE inhibition was
found in the dams and 22-24% inhibition in the fetus.  In  addition,  no increase in
post-natal pup deaths was found in the multi-generation  reproductive study with
dimethoate, where greater than 60% brain ChE inhibition was found in dams
(albeit, little brain ChE inhibition found in PND 4 pups) at the highest dose tested
(6 mg/kg/day).

      Although the underlying basis of the pup mortality is unclear, maternal
toxicity does not appear to be the only determining factor. In some studies where
significant maternal brain ChE inhibition was observed, increases in pup mortality
were not observed. With the exception of the special observations made in the
cross-fostering study, no clinical signs of overt toxicity were observed in dams at
any dose even where pup death occurred.  Lastly, in the cross fostering study,
which was designed to address this issue, no dear correlation could be drawn
between maternal behavior and pup death.

      Although no clear association can be made between a specific level of
brain ChE inhibition and an increase in pup death, results of the BMD analyses
would appear to support the conclusion that protection against brain ChE
inhibition will also result in protection against increased pup mortality following
repeated dosing.  No consistent age-related differences were seen in calculated
BMDio or BMDLio values for brain ChE inhibition; these values were similar to or
tower than those calculated for increases in pup mortality from the main DNT
study (the most sensitive study for that effect). Although the BMDio /BMDLio is
higher for ChE inhibition following a single dose, results of the recent cross-
fostering study may support a conclusion that increased  mortality  is not seen
following a single exposure at doses up to 3.0 mg/kg. Thus, use of the acute
BMDLio values for brain ChE inhibition (1.3-2.0 mg/kg) could also be protective
for increased pup mortality which might be seen after a single exposure at doses
                                57

-------
greater than 3.0 mg/kg (for example, at 6.0 mg/kg/day in the cross-fostering
study).

      In conclusion, several studies (i.e., the main DNT, range-finding study,
and cross fostering studies, the one-generation range-finding reproductive
toxicity study, and the omethoate reproductive toxicity studies) demonstrate
increased pup mortality following maternal exposure to dimethoate.  In addition,
there were reduced live births in a rat multi-generation reproductive toxicity study
and in the one-generation range-finding reproductive toxicity study with
dimethoate.  The important issues to address in this hazard assessment are the
characterization of the dose response and also the underlying basis for the pup
deaths.  Although the comparative ChE inhibition, range-finding, and cross-
fostering studies are not consistent with the findings of pup mortality in the main
DNT study, it is concluded that the pup mortality observed at both the 0.5 and 3.0
mg/kg/day dose levels cannot be discounted as treatment related. This
conclusion is based on the statistically significant response at both the 0.5 and 3
mg/kg/day doses and the dose-related nature of the response. The underlying
basis of pup mortality is not understood.  The available data do not support
maternal toxicity as being the only determinant of pup mortality.

      Based on the overall weight of the evidence which shows a dose-related
increase in pup death in the mid- and high-dose pups, this finding is considered
to be treatment-related and adverse at both the mid- and high doses. The key
evidence includes: pups were reported to be cold to the touch and unresponsive;
low incidence of total litter loss in performing laboratory; similar effect observed in
other studies-although dose levels differed; qualitative increased pup death/litter
(although does not reach statistical significance until 3 mg/kg/day); and
quantitative increase in pup death when evaluated as individuals. It is
appropriate to consider the increase in pup mortality as a treatment-related effect
following dimethoate administration at 0.5 mg/kg in the main DNT study.

      In conclusion, the current analysis supports the use of brain ChE inhibition
as an appropriate endpoint for acute or repeated-dose risk assessment
scenarios, based on the following:

Q    Brain ChE inhibition occurs at doses similar to or lower than those causing
      ChE inhibition in other compartments

Q    BMD analyses results indicate a very robust dose-response curve for
      brain ChE inhibition, with similar BMDio values from studies with varying
      modes of administration (dietary or gavage) and durations (short term for
      DNT studies and longer term for reproduction studies);

Q    BMD analyses results indicate similar dose-response curves at all ages,
      with no difference in BMD10 values for different age groups following
      similar exposure durations;
                                 58

-------
 4ft
                  Q    Comparison of BMP dose levels for brain ChE inhibition and pup mortality
                        following repeated dosing indicates that ChE inhibition occurs at doses
                        similar to those associated with increases in pup mortality;

                  Q    Evaluation of pup mortality data from the cross-fostering study reveals
                        clear increases in mortality only at the highest dose following short-term
                        exposure, indicating that increased mortality at lower doses occurs only
                        with repeated dosing;

                  Q    Comparison of the NOAEL for increased pup  mortality from limited dosing
                        with the BMDio for brain ChE inhibition following a single dose indicates
                        that brain ChE inhibition occurs at doses below those causing a clear
                        increase in pup mortality.

                  Therefore, regulation of dimethoate exposure at levels below those causing brain
                  ChE inhibition in adults will also  protect against brain ChE inhibition and
                  increased mortality in pups.
O
                                                   59

-------
III.    REFERENCES

Allen, B.C., R.J. Kavlock, C.A. Kimmel, and E.M. Faustman. (1994a).  Dose-Response
Assessment for Developmental Toxicity.  III. Statistical Models. Fund. Appl. Toxicol.
23:496-509.

Allen, B.C., R.J. Kavlock, C. A. Kimmel, and E.A. Faustman. (1994b).  Dose-Response
Assessment for Developmental Toxicity.  II. Comparison of generic benchmark dose
estimates with no observed adverse effect levels. Fund. Appl. Toxicol. 23:487-495.

Faustman, E.M., B.C. Allen, R.J. Kavlock, and C.A. Kimmel (1994) Dose-Response
Assessment for Developmental Toxicity.  I.  Characterization of database and
determination of no observed adverse effect levels.  Fund. Appl. Toxicol. 23:478-486.

Health Effects Test Guidelines OPPTS 870.6300 Developmental Neurotoxicity Study

Kavlock, R.J., B.C. Allen, E.M. Faustman, and C.A. Kimmel. (1995). Dose-Response
Assessments for Developmental Toxicity. IV. Benchmark Doses for fetal weight
changes. Fund. Appl. Toxicol. 26:221-222.

USEPA (2002). Revised Organophosphorus Pesticide Cumulative Risk Assessment
Office of Pesticide Programs, U.S. Environmental Protection Agency. Washington, DC.
June 10, 2002.  http://www.epa.gov/pesticides/cumulative/rra-op/

USEPA (2000)  Benchmark Dose Technical Guidance Document.  Risk Assessment
Forum, U.S. Environmental Protection Agency. Washington, DC.  October, 2000

USEPA (1991)  Guidelines for Developmental Toxicity Risk Assessment.  Risk
Assessment Forum, U.S. Environmental  Protection Agency. Washington, DC
December 1991.
                                     60

-------