Page
1
of
14
UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
OFFICE
OF
PREVENTION,
PESTICIDES
AND
TOXIC
SUBSTANCES
MEMORANDUM
DATE:
July
12,
2006
SUBJECT:
Dichlorvos
Human
Health
Risk
Assessment
­
Response
to
AMVAC's
Error
only
Correction
Comments
DP
Barcode
D290595
Reregistration
Case#:
0310
PC
Code:
084001
MRID
No.:
None
TO:
Dayton
Eckerson
and
Eric
Olson,
CRM
#
61
Special
Review
Branch
Special
Review
and
Reregistration
Division
(
7508C)

FROM:
William
Dykstra,
Ph.
D.,
Toxicologist
Rebecca
Daiss,
Environmental
Health
Scientist
David
Jaquith,
Environmental
Scientist
Reregistration
Branch
4
Health
Effects
Division
(
7509C)

THRU:
Susan
Hummel,
Branch
Senior
Scientist
Reregistration
Branch
4
Health
Effects
Division
(
7509C)

Background
and
Request:

The
June
14,
2005
HED
Dichlorvos
Human
Health
Risk
Assessment
for
Phase
5
Reregistration
(
DP
Barcode:
D302294)
was
critiqued
by
the
registrant,
AMVAC,
in
their
July
29,
2005
submission
to
the
Agency.
Reregistration
Branch
4
has
reviewed
and
addressed
the
AMVAC
comments.

First
we
will
respond
to
AMVACs
generic
comments,
followed
by
their
questions
on
the
BMD
analysis.
Lastly,
we
will
respond
to
their
page
by
page
comments.
Page
2
of
14
GENERIC
COMMENTS
 
EPA
Must
Provide
AMVAC
with
Critical
Data
Relied
Upon
in
the
Assessment:
It
is
impossible
to
comment
adequately
on
the
factual
errors
without
the
data
on
which
EPA's
conclusions
are
based.
A
significant
number
of
references
on
which
EPA
relies
in
the
Risk
Assessment
are
either
or
both
unknown
or
unavailable
to
AMVAC.
AMVAC
cannot,
therefore,
assess
the
accuracy
of
the
Risk
Assessment
without
having
had
an
opportunity
to
review
the
documents
on
which
it
relies.

Response:
No
comment.
HED
will
provide
any
data
we
have
in
our
possession
that
is
considered
appropriate
to
this
comment.

 
EPA
Must
Consider
Additional
Data
and
Information
Not
Referenced
in
the
Risk
Assessment:
It
would
be
a
significant
error
to
release
a
risk
assessment
that
does
not
reflect
a
consideration
of
all
relevant
information.

Response:
We
have
included
a
discussion
of
all
data
we
believe
is
relevant.
Many
of
the
human
studies
cited
by
AMVAC
are
on
metrifonate
(
aka
trichlorfon)
used
experimentally
for
Alzheimer's
treatment.
Although
somewhat
efficacious
for
treatment
of
Alzheimer's,
it
was
withdrawn
by
Bayer
in
1998
because
of
adverse
side
effects.
It
does
in
fact
form
dichlorvos
in
the
body
by
a
non­
enzymatic
process,
but
at
an
unknown
rate,
so
that
metrifonate
data
are
not
useful
for
dichlorvos
assessment.
AMVAC
also
cited
many
studies
where
dichlorvos
is
formulated
as
a
resin.
The
rate
of
release
of
dichlorvos
from
the
resin
is
unknown,
but
comparative
LD50
data
in
the
pig1
indicate
that
most
of
the
dichlorvos
formulated
in
plastic
beads
is
unavailable
for
uptake
upon
acute
oral
administration.
We
therefore
cannot
use
these
resin
studies
for
risk
assessment
purposes,
and
we
are
not
aware
of
any
human
pharmacology
studies
where
dichlorvos
per
se
is
administered
as
a
drug
under
controlled
conditions.

1The
oral
LD50
of
dichlorvos
in
the
pig
is
157
mg/
kg/
day
and
the
oral
LD50
of
dichlorvos
administered
in
plastic
beads
(
Atgard)
in
the
pig
is
>
1000
mg/
kg/
day
The
properties
of
the
slow
release
formulation
differ
considerably
from
those
of
a
typical
pesticide
formulation.
The
oral
LD50,
the
dose
administered
all
at
once
by
the
oral
route
that
kills
50%
of
the
test
species,
for
technical
dichlorvos
in
rats
is
157
mg/
kg,
whereas
no
deaths
occurred
in
pigs
administered
1000
mg/
kg
of
the
slow­
release
formulation
(
Stanton
et
al.
1979).
Wrathall
et
al.
(
1980)
determined
that
the
slow­
release
formulation
of
dichlorvos
was
not
developmentally
toxic
Page
3
of
14
to
pigs
at
a
dose
of
8.5
mg/
kg.
The
authors,
in
comparing
the
teratogenic
response
of
Dichlorvos
and
trichlorfon,
speculated
that
"...
because
a
slow
release
preparation
(
PVC­
Dichlorvos)
was
used,
the
amount
of
active
organophosphorus
compound
which
reached
the
fetuses
may
have
been
negligible."

There
also
seems
to
be
the
perception
on
the
part
of
the
registrant/
consultant
that
whenever
FDA
makes
a
determination
that
a
drug
is
safe
for
its
intended
use
(
as
a
drug),
it
must
be
safe
for
all
uses.
In
OPP,
we
are
dealing
with
a
different
risk/
benefit
equation
and
an
FDA
determination
of
a
drug
use
does
not
apply
to
a
pesticidal
use.

 
EPA
Erroneously
Set
DDVP's
Interspecies
Uncertainty
Factor:
EPA
erroneously
does
not
consider
in
the
Risk
Assessment
two
previous
AMVAC
submissions
that
addressed
the
issue
of
DDVP's
interspecies
uncertainty
factor.
AMVAC
urges
EPA
to
review
these
documents
and
revise
the
Risk
Assessment
to
indicate
that
healthy
humans
appear
no
more
sensitive
than
laboratory
animals
to
DDVP's
effects
on
red
blood
cell
(
RBC)
cholinesterase
and
an
interspecies
uncertainty
factor
of
1­
fold
is
appropriate
and
scientifically
warranted
for
use
in
the
Risk
Assessment.

Response:
An
interspecies
extrapolation
factor
of
1
is
used
when
human
data
are
used
for
endpoint
setting.
In
one
scenario
where
animal
studies
are
used
for
endpoint
selection
­
acute
exposure,
a
number
of
studies
were
cited
by
AMVAC,
but
there
are
no
acute
human
dosing
studies
that
we
consider
useful
for
risk
assessment
purposes,
including
informing
the
interspecies
factor.
(
AMVAC
also
claims
elsewhere
that
human
RBC
AChe
is
not
like
rat
AChe,
since
it
doesn't
recover
as
quickly
­
see
response
below)

 
EPA
Erroneously
Set
DDVP's
Intraspecies
Uncertainty
Factor:
EPA
erroneously
issued
a
Risk
Assessment
that
omits
any
review,
citation,
mention,
or
consideration
of
the
substantial
body
of
clinical
human
data
relevant
to
assessment
of
intraspecies
uncertainty.
AMVAC
urges
EPA
to
correct
the
Risk
Assessment
so
that
it
considers
and
discusses
this
database
and
applies
an
intraspecies
uncertainty
factor
of
no
more
than
3­
fold.

Response:
As
discussed
above,
the
clinical
data
are
not
relevant
to
dichlorvos
risk
assessment.
Ordinarily,
the
intraspecies
factor
is
not
adjusted
unless
a
very
large
number
of
people
have
been
included
in
a
study,
e.
g.,
an
epidemiology
study,
or
if
a
group
known
to
be
sensitive
has
been
included
in
a
study.
For
dichlorvos,
such
a
group
might
be
one
with
an
inability
to
detoxify
dichlorvos.
The
number
of
human
subjects
exposed
to
dichlorvos
under
controlled
conditions
is
too
small
and
no
populations
sensitive
to
dichlorvos
have
been
defined.
There
is
no
basis
to
decrease
the
intraspecies
factor.

 
EPA
Erroneously
Set
DDVP's
FQPA
Factor:
EPA's
Food
Quality
Protection
Act
(
FQPA)
uncertainty
factor
assessment
erroneously
fails
to
mention
numerous
Page
4
of
14
published
studies
showing
a
lack
of
differential
sensitivity
due
to
age
in
animals
and
humans,
and
erroneously
fails
to
use
the
corrected
report
showing
no
increased
sensitivity
between
the
young
and
adults
for
its
FQPA
uncertainty
factor.

AMVAC
erroneously
refers
to
a
"
corrected"
report.
The
report
needed
no
correction.
AMVAC's
toxicology
consultant
identified
two
other
repeated
dose
comparative
cholinesterase
studies
that
were
performed
in
roughly
the
same
time
frame
(­
1
month
and
+
6
months)
and
pooled
the
controls
from
all
three
studies.
With
respect
to
the
pooled
controls,
there
were
no
statistically
significant
changes
at
0.1
mg/
kg/
day,
whereas
there
is
with
the
concurrent
controls.

AMVAC's
issue
is
that
the
cholinesterase
activity
in
the
dichlorvos
control
groups
in
the
repeated
dose
studies
was
too
high.
One
of
the
first
tenets
of
toxicology
review
is
the
primacy
of
the
concurrent
controls.
When
a
study
is
conducted,
the
test
animals
are
randomly
assigned
to
dose
groups,
and
it
is
assumed
that
responses
of
all
groups
would
be
similar
were
it
not
for
the
effect
of
a
treatment
regimen.
In
other
words,
we
have
every
reason
to
believe
that
the
0.1
mg/
kg/
day
groups
would
have
responded
like
the
controls
were
it
not
for
dichlorvos
treatment.
It
should
be
noted
that
the
standard
deviation
of
the
control
groups
in
the
dichlorvos
study
were
comparable
to
the
treated
groups
(
i.
e.,
the
mean
cholinesterase
activity
in
the
concurrent
control
groups
is
not
driven
by
an
outlier).

HED
toxicologists
recommend
that
the
repeated
dose
comparative
cholinesterase
studies
be
repeated
to
address
the
concerns
that
have
been
raised.

 
EPA
Erroneously
Set
a
LOAEL
in
the
Arizona
II
Study
Based
on
Headaches
and
RBC
Cholinesterase
Inhibition:
EPA
erroneously
determines
that
there
is
a
correlation
between
headaches
and
DDVP.
EPA
also
erroneously
determines
that
a
mean
RBC
cholinesterase
inhibition
less
than
20
percent
and
lacking
statistical
effect
is
an
adverse
effect.

Response:
With
respect
to
the
headaches,
AMVAC
may
be
right,
although
the
comment
is
substantive
and
not
an
error.
The
Arizona
II
study
has
been
reviewed
by
the
HSRB,
and
cannot
be
used
for
risk
assessment
 
EPA
Erroneously
Determined
Recovery
Kinetics
of
RBC
Cholinesterase
Inhibition:
EPA
erroneously
assessed
critical
endpoints
in
the
Risk
Assessment
by
erroneously
assuming
that
human
peak
RBC
cholinesterase
inhibition
responses
mirrors
RBC
cholinesterase
recovery
kinetics
in
rats
and
that
recovery
kinetics
after
single
and
repeat
doses
cannot
be
adequately
assessed
at
a
24­
hour
sampling
point
in
human
studies.

Response:
The
basis
for
this
assertion
was
not
adequately
explained.
We
will
look
at
this
issue
further
if
there
are
relevant
human
data
available,
but
it
is
not
an
error
correction.
The
comment
Page
5
of
14
is
at
odds
with
the
earlier
comment
that
there
are
no
species
differences
that
would
warrant
an
interspecies
factor.

 
EPA
Erroneously
Concluded
That
the
DNT
Studies
Are
Unacceptable:
EPA
erroneously
did
not
review
two
developmental
neurotoxicity
(
DNT)
studies
in
combination
which
together
show
that
the
guideline
requirements
have
been
met.
EPA
also
omitted
consideration
of
submissions
of
information
pertaining
to
the
assessment
of
the
DNT
studies
including
historical
control
data,
animal
husbandry,
and
other
relevant
information.

Response:
The
DNT
studies
are
now
considered
acceptable.

 
EPA
Erroneously
Assessed
Exposure
to
DDVP
in
Resin
Strips
and
Pet
Flea
Collars
Using
Air
Concentrations
Instead
of
Applied
Dose:
EPA
erroneously
assessed
DDVP
exposures
through
direct
comparison
of
air
concentrations
when
it
can
more
appropriately
and
accurately
assess
exposures
through
dose
estimates.

Response:
Use
of
RfC
methodology
for
assessment
of
inhalation
risks
has
been
EPA
policy
since
the
mid/
late
90'
s.
Although
OPP
is
late
in
adopting
the
methodology,
we
consider
air
concentration
to
be
the
proper
metric
for
inhalation
risk
assessment
for
vapors
and
gasses.

 
EPA
Erroneously
Omitted
Discussions
of
Permissible
Resin
Strip
Uses:
EPA
should
include
in
the
Risk
Assessment
a
discussion
of
resin
strip
uses
that
should
be
permissible
but
are
not
discussed
in
the
Risk
Assessment.

Response:
The
small
closet
size
resin
strip
was
omitted
from
the
discussion
but
was
included
in
the
table
of
exposure
times
and
estimated
risks.
All
sizes
of
remaining
resin
strips
will
be
discussed
in
the
risk
assessment.

 
EPA
Erroneously
Determined
Turf
Exposure
with
Regard
to
Toddler's
Body
Weight
and
Respiratory
Volume:
EPA
erroneously
calculates
the
inhalation
exposure
to
DDVP
use
on
turf
by
assuming
a
child
is
exposed
for
2
hours
per
day,
weighs
10
kg,
and
has
a
breathing
rate
of
0.8
m3/
hour.

Response:
The
correct
toddler
exposure
on
turf
would
be
2
hours
per
day,
weight
15
kg,
and
a
breathing
rate
of
0.36
m3/
hour.
The
dichlorvos
turf
use
has
been
canceled.
This
use
is
no
longer
discussed
in
the
risk
assessment.

 
EPA
Erroneously
Double
Counted
Exposures
for
Flea
Collars:
EPA
erroneously
double­
counts
hand
to
mouth
exposures
including
both
the
oral
and
dermal
routes
to
the
same
exposure,
for
flea
collars
and
must
correct
this
error
in
the
Risk
Assessment.
Page
6
of
14
Response:
A
Draft
ExpoSAC
policy
was
used
for
estimating
dermal
and
hand
to
mouth
exposures
to
flea
collars.

 
EPA
Erroneously
Discusses
Ethical
Issues
Regarding
Two
DDVP
Human
Studies:
EPA's
ethical
discussion
of
two
DDVP
human
studies
contains
numerous
errors
and
omissions
that
must
be
corrected.

Response:
John
Carley
will
respond
to
this.

QUESTIONS
ON
BMDL10
ANALYSIS.

Which
Data
Sets
Were
Modeled?
The
only
information
provided
in
the
Risk
Assessment
is
that
the
BMDL10
of
0.8
mg/
kg
was
derived
from
data
on
Day
1
adult
female
brain
ChE
depression
and
that
it
was
selected
from
among
all
of
the
generated
BMDL10
estimates
because
it
was
"
the
lowest
value
of
all
the
studies
available
which
were
analyzed
by
BMD."
The
specific
studies
and
data
sets
that
were
utilized
in
BMD
analyses
are
not
identified
and
this
should
be
corrected.

The
modeled
data
sets
were
provided
in
the
BMD
memo
by
B.
Daiss
which
was
referenced
in
the
Dichlorvos
Human
Health
Risk
Assessment.

Were
Restrictions
Imposed
on
Polynomial
Coefficients?
Constraints
(
e.
g.,
non­
positive)
on
the
polynomial
coefficients
can
have
a
marked
impact
on
the
adequacy
of
the
model
fit
to
the
data
and
on
the
resulting
BMD10
and
BMDL10
estimates.
EPA
should
state
whether
constraints
were
imposed
or
not.

No
restrictions
were
imposed
on
polynomial
coefficients.

Were
Variances
Assumed
Constant
or
Allowed
to
Depend
on
Dose?
The
assumption
of
constant
or
dose­
dependent
variances
can
also
have
a
marked
impact
on
adequacy
of
the
model
fit
and
the
resulting
BMD10
and
BMDL10
estimates.
EPA
erred
by
not
explaining
in
the
Risk
Assessment
whether
one,
or
the
other,
or
both
of
these
options
were
explored.

Constant
variances
were
assumed
as
a
default.
Non­
constant
variances
were
selected
if
homogeneous
variance
model
was
deemed
to
be
inappropriate
based
on
p­
value.

Was
the
Polynomial
Degree
the
Same
Across
Data
Sets?
Allowing
the
polynomial
degree
to
vary
or
holding
it
constant
across
data
sets
can
also
impact
adequacy
of
model
fit
and
the
resulting
BMD10
and
BMDL10
estimates.
EPA
erred
by
not
stating
whether
the
polynomial
was
held
fixed
or
was
varied
across
the
data
set
that
were
utilized.

The
2d
degree
polynomial
provided
the
best
fit
across
data
sets.
Page
7
of
14
What
Convergence
Criteria
Were
Employed?
BMD
Software
version
1.3.2
allows
adjustments
to
be
made
in
the
parameters
that
determine
if
and
when
convergence
of
the
iteration
process
has
been
achieved.
These
parameters
can
have
an
impact
on
the
dose­
response
model
parameter
estimates,
adequacy
of
model
fit,
as
well
as
the
resulting
BMD10
and
BMDL10
estimates.
EPA
erred
by
not
providing
the
convergence
criteria
that
were
used.

No
convergence
criteria
were
used.

Was
Model
Adequacy
Considered?
Model
fit
can
have
a
large
impact
on
resulting
BMD10
and
BMDL10
estimates.
EPA
erred
by
not
stating
in
the
Risk
Assessment
whether
minimal
criteria
for
the
adequacy
of
model
fits
to
the
various
datasets
were
utilized
in
EPA's
BMD
analyses.

Minimal
criteria
were
applied
as
provided
in
the
BMD
Guidance
Document.

Were
High
Dose
Data
Ever
Dropped
Out
of
the
Modeling?
Data
from
the
highest
dose
group
can,
on
occasion,
have
undue
and
inappropriate
influence
on
BMD10
and
BMDL10
estimates.
In
cases
where
this
is
suspected,
it
is
common
practice
to
drop
the
high
dose
data
from
the
model
fitting
process.
EPA
erred
by
not
stating
whether
it
dropped
the
high
dose
for
any
of
the
BMD
analyses
that
were
undertaken.

The
high
dose
data
may
have
been
dropped
for
one
or
two
data
sets,
but
had
no
bearing
on
adequacy
of
model
fit.

Why
Was
the
Lowest
BMDL10
Estimate
Selected?
EPA
errs
by
selecting
the
lowest
BMDL10
estimate,
as
this
is
an
inappropriately
and
unnecessarily
conservative
practice.
Such
a
practice
leads
to
reliance
on
the
weakest
data
set,
namely,
the
one
that
is
least
informative
and
most
uncertain
regarding
the
location
of
the
BMD10.
An
alternative
and
more
sensible
criterion
for
BMDL10
selection
would
be
to
choose
the
one
yielding
the
smallest
BMD10
/
BMDL10
ratio.
Such
a
criterion
would
favor
reliance
on
the
strongest
data
set,
namely,
the
one
that
provides
the
tightest,
most
precise
estimate
of
the
BMD10.

There
was
not
enough
difference
in
the
results
spread
to
affect
the
outcome.

COMMENTS
ON
SPECIFIC
PAGES.

1.
Page
1,
Paragraph
4,
Line
2:
AMVAC
states
that
neurological
effects
should
be
changed
to...
cholinergic
effects
where
inhibition
of
blood
cholinesterases
is
not
an
adverse
effect
but
provides
a
marker
of
exposure.

Response:
Although
the
inhibition
of
erythrocyte
cholinesterase
(
RBC
ChE)
may
be
viewed
in
principle
as
non­
adverse,
RRB4,
in
keeping
with
OPP
policy,
uses
this
endpoint
for
the
assessment
and
protection
of
possible
cholinergic
effects
of
the
CNS
and
PNS.
In
so
doing,
HED
employs
Page
8
of
14
safety
factors
and
uncertainty
factors
in
the
regulation
of
RBC
ChE
depression.
Therefore,
HED
uses
this
non­
adverse,
marker
of
exposure
(
significant
RBC
ChE
depression
without
clinical
signs),
together
with
adequate
safety
factors,
for
the
protection
of
public
health.

2.
Page
3,
Paragraph
4,
Line
7.
AMVAC
states
that
EPA
has
erred
in
not
considering
the
results
of
the
corrected
study
(
MRID
46433201)

Response:
AMVAC
is
correct
to
note
that
HED
had
not
considered
the
MRID
46433201.
However,
a
review
of
the
MRID
has
now
been
completed
and
HED
believes
that
the
concurrent
controls,
rather
than
the
combined
controls
of
the
three
studies,
should
be
used
to
assess
the
selection
of
NOAEL/
LOAEL
for
the
Day
18
and
Day
48
test
groups.
The
standard
deviations
of
the
means
for
the
concurrent
controls,
where
individual
animal
data
were
provided,
were
comparable
between
control
and
treated
groups
and
suggest
that
the
differences
in
the
results
of
the
concurrent
controls
and
test
groups
are
not
being
driven
by
an
"
outlier"
in
the
concurrent
controls.
The
only
scientific
way
to
resolve
whether
the
concurrent
controls
are
"
unusually
high"
is
to
repeat
the
study
and
submit
the
results
to
EPA.

3.
Page
5,
paragraph
1,
line
5
Occupational
scenarios
are
described
as
short­
and
intermediateterm
but
the
notes
for
Table
9.1
(
pages
96
and
97)
list
2
re­
entry
scenarios
as
acute.

Response:
AMVAC
is
correct.
We
will
add
"
acute"
to
the
text
on
that
page.

4.
Page
5,
paragraph
3,
line
4.
Small
closet
strips
should
be
included
in
residential
scenarios
that
do
not
exceed
EPA's
level
of
concern.

Response:
Small
closet
strips
will
be
added
to
that
section.

5.
Page
30,
Paragraph
1,
Line
13.
AMVAC
states
that
EPA
has
erred
in
correlating
headaches,
even
marginally,
with
DDVP
exposure
in
the
Arizona
II
study.

Response:
AMVAC
is
correct
in
their
comment
that
headaches
are
not
related
to
RBC
cholinesterase
depression
in
the
Arizona
II
study
and
has
not
been
identified
as
a
clinical
sign
in
other
human
studies
with
DDVP.
The
Arizona
II
study
is
no
longer
used
in
the
risk
assessment.

6.
Page
32,
Table
4.1b
1st
Row,
line
8.
(
MRID
45845301)
Brain
ChE
inhibition
LOAEL
should
be
5
mg/
kg.

Response:
AMVAC
is
correct
in
assigning
the
LOAEL
of
5.0
mg/
kg
for
brain
ChE
inhibition
(
the
5
mg/
kg
dose
level
was
incorrectly
listed
as
the
NOAEL
in
the
RED).

7.
Page
32,
Table
4.1b,
2nd
Row.
Use
of
"
corrected
study"
(
MRID
46433201)
rather
than
MRID
46153304.
Page
9
of
14
Response:
This
comment
was
answered
in
EPA
Item
#
5.

8.
Page
32,
Table
4.1b,
5th
row.
AMVAC
states
that
MRID
44317901
is
acceptable
in
weight
of
the
evidence
comparison
of
human
and
animal
toxicology
studies
and
endpoints
for
DDVP
on
top
of
page
14
of
17.

Response:
Since
the
time
of
peak
effect
was
not
included
in
the
analysis
of
the
RBC
ChE
inhibition
of
the
single,
oral
dose
of
DDVP
to
male
subjects,
the
absence
of
results
in
this
study
precluded
the
determination
of
a
NOAEL/
LOAEL
for
humans.
Therefore,
this
human
study
was
not
justified
in
being
used
for
weight
of
the
evidence
determinations.

9.
Page
35,
Table
4.1b,
MRID
46153302,
46239801,
Rows1st,
2nd,
3rd,
4th,
Taken
collectively,
AMVAC
states
that
EPA
could
combine
the
results
of
both
DNT
studies
(
MRID
46153302
and
MRID
46239801)
to
arrive
at
NOAELs/
LOAELs
for
both
maternal
and
offspring
criteria.
Additionally,
EPA
errs
in
stating
that
effects
were
seen
in
offspring
in
MRID
46153302
when
these
were
not
reproducible
in
MRID
46239801.

Response:
The
DNT
Committee
met
on
September
1,
2005
to
consider
the
AMVAC
request
in
light
of
the
newly
submitted
Historical
Control
information
(
MRID
46487401).
The
DNT
Committee
determined
that
the
two
DNT
studies
combined
had
acceptable
numbers
of
total
pups
examined
in
the
controls
and
high
dose
groups
(>
35
pups/
sex
in
combined
studies)
and,
therefore,
the
developmental
results
of
the
combined
studies
could
be
evaluated
for
the
NOAEL/
LOAEL.
The
classification
of
the
studies
was
changed
from
unacceptable/
non­
guideline
to
Acceptable/
nonguideline
A
comparison
of
the
developmental
findings
showed
that
the
auditory
startle
reflex
habituation
Vmax
in
PND
23
high
dose
males
in
study
RR0886
had
statistically
significant
increases
(
39­
47%)
in
4
out
of
5
blocks
and
study
RR0988
had
increases
(
7­
15%),
although
not
statistically
significant,
in
this
same
Vmax
parameter
also
in
PND
23
high
dose
males
in
5
out
of
5
blocks
in
comparison
to
controls.
Therefore,
the
developmental/
offspring
NOAEL
was
determined
to
be
1.0
mg/
kg/
day
(
based
on
study
RR0886)
and
the
LOAEL
was
7.5
mg/
kg/
day
(
based
on
both
studies
RR0886
and
RR0988)
with
the
effect
being
increases
in
auditory
startle
amplitude
in
both
studies.

10.
Page
36,
Paragraph
2,
Line
1.
EPA
states
"
There
is
a
concern
for
neurotoxicity
resulting
from
exposure
to
DDVP".
A
sentence
should
be
added
that
neurotoxicity
is
not
observed
in
either
animals
or
humans
at
levels
of
DDVP
exposure
that
produce
inhibition
of
RBC
ChE
of
less
than
20%.

Response:
Our
statement
is
correct.

11.
Page
37,
Paragraph
4,
Lines
1­
2.
Determination
of
susceptibility
should
be
corrected
to
show
that
reanalysis
(
MRID
46433201)
does
not
demonstrate
susceptibility.
Page
10
of
14
Response:
The
data
were
reanalyzed;
the
developmental/
offspring
NOAEL
was
determined
to
be
1.0
mg/
kg/
day
(
based
on
study
RR0886)
and
the
developmental/
offspring
LOAEL
was
7.5
mg/
kg/
day
(
based
on
both
studies
RR0886
and
RR0988)
with
the
effect
being
increases
in
auditory
startle
reflex
habituation
Vmax
in
PND
23
high
dose
males
in
both
studies.

12.
Page
39,
Paragraph
5,
Line
1.
AMVAC
states
that
EPA
omitted
the
discussion
of
data
which
shows
that
the
interspecies
factor
for
DDVP
should
be
1
rather
than
10.

Response:
An
interspecies
factor
of
1
rather
than
10
was
used
when
the
endpoint
was
based
on
human
data.
The
only
instance
of
the
use
of
10
for
the
interspecies
factor
is
with
the
acute
exposure
scenario
where
the
BMD
analysis
is
used
with
single
dose
rat
studies.
The
interspecies
factor
of
10
was
used
in
the
absence
of
acceptable
single
dose
human
studies.
The
available
acute
human
studies
missed
the
time
of
peak
effect
and
were
inappropriate
to
use
to
reduce
the
interspecies
factor
or
conduct
risk
assessments.

13.
Page
39,
Paragraph
5,
Line
1.
AMVAC
states
that
EPA
omitted
the
discussion
of
data
which
shows
that
the
intraspecies
factor
for
DDVP
should
be
3
rather
than
10.

Response:
The
reduction
of
the
intraspecies
factor
from
10
to
3
is
not
warranted
(
1)
since
too
few
human
subjects
have
been
evaluated
with
DDVP
under
controlled
conditions
and
(
2)
there
are
no
known
human
subgroups
or
populations
which
are
known
to
be
more
sensitive
to
exposure
to
DDVP
(
e.
g.,
a
group
that
has
an
inability
to
detoxify
DDVP).
Therefore,
there
is
no
basis
to
decrease
the
intraspecies
factor
from
10
to
3.

14.
Page
39,
Paragraph
8,
Line
1.
EPA
has
erred
in
not
using
the
acute
human
oral
study
of
Gledhill
(
1997).

Response:
The
available
acute
human
studies
had
ChE
measurements
at
24
hours
after
dosing.
This
missed
the
time
of
peak
effect
and
the
study
was
inappropriate
to
use
to
conduct
risk
assessments.

15.
Page
40,
Paragraph
3,
Line
1;
Page
41,
Paragraph
2
and
8,
Lines1,
and
Page
43,
Paragraph
5,
Line
1.
AMVAC
states,
collectively
in
these
comments,
that
EPA
incorrectly
used
both
the
interspecies
and
intraspecies
safety
factors.

Response:
All
of
these
comments
have
been
addressed
in
Items
12
and
13.

16.
Page
42,
paragraph
2,
line
14
Under
"
Executive
Summary"
Atmospheric
is
misspelled.

Response:
Corrected.

17.
Page
43,
Paragraph
1
and
3,
Lines
10
and
1,
AMVAC
states
that
EPA
has
erred
in
calling
the
correlation
of
headaches
with
exposure
"
equivocal
or
marginal".
Page
11
of
14
Response:
These
comments
were
addressed
in
Item
#
5.

18.
Page
46
Table
4.4
6
The
Levels
of
Concern
(
3rd
column)
for
"
Short­,
Intermediate­,
and
Long­
Term
Inhalation
of
vapors"
should
be
listed
as
"
30"
and
"
10"
for
residential
and
occupational,
respectively,
rather
than
"
30x"
and
"
10x".

Response:
Corrected.

19.
Page
46
Table
4.4,
line
7.
There
is
no
explanation
for
the
choice
of
this
study
for
assessing
inhalation
of
particulates.
Additionally,
and
inhalation
of
particulates
is
not
assessed
in
the
risk
assessment,
thus
this
row
should
be
removed
from
the
table.

Response:
Instead
of
particulates,
the
appropriate
word
to
use
is
aerosols.
This
endpoint
is
used
for
the
short
and
intermediate
term
inhalation
for
occupational
exposure.

20.
Page
47,
Paragraph
5,
4th
bullet,
AMVAC
states
that
based
on
the
"
corrected
study"
(
MRID
46433201)
there
is
no
need
for
a
special
FQPA
factor
for
DDVP.

Response:
This
comment
was
addressed
in
Item
#
2.

21.
Page
48,
Paragraph
1,
Line
16.
EPA
should
acknowledge
that
since
DDVP
is
regulated
on
the
basis
of
exposure,
RBC
ChE
inhibition,
and
not
toxicity,
the
potential
for
endocrine
disruption
is
lower
than
for
chemicals
where
the
critical
endpoint
is
frank
toxicity.

Response:
The
occurrence
of
frank
toxicity
with
DDVP
(
e.
g.,
inhibition
of
brain
ChE
activity,
appearance
of
clinical
signs,
etc.)
is
avoided
by
regulating
the
human
exposure
to
DDVP
based
on
RBC
ChE
activity
through
the
use
of
safety
factors.
Therefore,
for
regulatory
purposes,
the
significant
depression
of
RBC
ChE
activity,
which
is
regulated
by
the
use
of
safety
factors,
is
considered
the
precursor
of
frank
toxicity.
The
critical
endpoints
of
RBC
ChE
inhibition
with
DDVP
and
"
frank"
toxicity
with
other
chemicals
are
treated
equally
in
assessing
endocrine
disruption
potential.

22.
Page
70,
paragraph
1,
line
10
The
Jaquith
2001a
reference
is
not
listed
in
the
reference
list
and
this
should
be
corrected
(
possibly
Jaquith
2003)
or
deleted.

Response:
Jaquith
2003
is
correct.
The
document
has
been
corrected.

23.
Page
70,
paragraph
2,
line
1
Residential
scenarios
were
evaluated
for
acute
duration.
This
sentence
should
be
changed
to
read:
"
Residential
Scenarios
which
were
evaluated
were
of
acute,
short
term
or
long
term
duration."

Response:
AMVAC
is
correct.
The
document
has
been
corrected.
Page
12
of
14
24.
Page
70,
paragraph
2,
line
3
Any
mention
of
"
inhalation
(
particulate)"
assessment
should
be
removed
from
the
document
as
there
is
no
exposure
through
this
pathway.

Response:
The
document
has
been
corrected.

25.
Page
71,
paragraph
1,
line
1
Crack
and
crevice
and
aerosol
post­
application
exposures
should
not
be
assessed
based
on
exposures
to
total
release
foggers,
as
there
are
minimal,
if
any,
postapplication
exposures
to
these
applications.
(
More
detailed
comments
were
submitted
on
the
2000
risk
assessment.

Response:
AMVAC
agreed
to
the
use
of
total
release
foggers
to
represent
all
3
uses,
to
minimize
the
need
for
additional
data.
However,
EPA
will
look
into
alternative
ways
to
conduct
this
assessment
during
the
risk
mitigation
phase.
AMVAC
should
note,
however,
that
available
data
show
that
there
is
very
little
difference
between
the
exposure
from
a
broadcast
use
and
a
crack
and
crevice
treatment.
Crack
and
crevice
treatments
have
been
canceled,
and
are
no
longer
included
in
the
risk
assessment.

26.
Page
73,
paragraph
7,
line
1
The
assessment
of
inhalation
exposure
used
in
the
RED
does
not
correspond
to
the
assessment
done
by
Jaquith
in
2003
(
Addendum
to
Residential
Turf
Assessment
for
Dichlorvos
(
DDVP))
PC
Code
084001,
Barcode
D288914,
dated
March
28,
2003).

Response:
Turf
use
of
dichlorvos
has
been
canceled,
and
there
is
no
discussion
of
this
in
the
risk
assessment.

27.
Page
73,
paragraph
7,
line
3.
The
air
concentration
of
1.4
mg/
m3
does
not
agree
with
the
air
concentration
of
1.6
mg/
m3
in
the
notes
section
of
Table
6.3
(
page
78).

Response:
The
air
concentration
of
1.6
mg/
m3
is
correct.
The
document
will
be
corrected.

28.
Page
73,
paragraph
7,
line
5
The
MOE
of
0.63
for
toddlers
does
not
correspond
to
the
MOE
listed
on
Table
6.3
(
Average
MOE
=
0.13­
0.62)
on
page
75.

Response:
The
calculations
were
correct
in
the
table
footnotes.
This
use
pattern
has
been
canceled.

29.
Page
75
Table
6.3,
line
12.
The
Exposure
Pattern
for
exposure
to
lawns,
turf
and
ornamental
plants
should
be
changed
to
Acute
rather
than
Short­
term
according
to
Note
9
(
page
78).

Page
75
Table
6.3,
line
14.
The
Exposure
Pattern
for
exposure
to
Lawns,
Trichlorfon
use
should
be
changed
to
Acute
rather
than
Short­
term
according
to
Note
9
(
page
78).
Page
13
of
14
Response:
We
believe
that
short­
term
is
more
appropriate,
since
dichlorvos
is
formed
as
trichlorfon
degrades,
forming
dichlorvos
for
perhaps
several
days.

30.
Page
75
Table
6.3,
line
14.
The
dermal
exposure
values
(
0.00038
and
0.00011
mg/
kg/
day
for
adults
and
0.00086
and
0.00024
for
toddlers)
do
not
correspond
to
the
dermal
dose
values
from
Leighton
2000
memo
that
is
referenced
in
Note
9
or
to
the
values
in
the
triclorfon
TRED
available
on
the
EPA
website.
It
is
unclear
from
a
time­
limited
review
of
the
registrant­
submitted
study
and
the
Leighton
2000
memo
based
on
the
registrant
study
which,
if
any,
are
the
correct
values
as
there
appear
to
be
errors
in
the
Leighton
2000
memo
(
for
example,
the
equation
before
Section
4.4.3.3
lists
a
Cpi
of
0.0138
µ
g/
cm2
whereas
two
paragraphs
previously
the
initial
concentration
was
listed
as
0.0829
µ
g/
cm2).

Response:
The
exposures
were
recalculated
and
the
calculation
included
in
the
table
footnotes.

31.
Page
108,
Line
27.
EPA
has
erred
in
stating
that
there
is
"
lack
of
information
on
time
of
peak
effect".
Although
this
study
by
itself
may
not
have
addressed
peak
effects,
there
are
several
published
studies
in
humans
with
this
data
that
allows
for
assessment
of
RBC
cholinesterase
enzyme
kinetics
following
DDVP
exposure.
The
data
are
consistent
in
showing
that
the
peak
effect
of
RBC
cholinesterase
inhibition,
following
either
a
single
or
repeated
doses,
is
adequately
assessed
by
monitoring
the
effect
24
hours
after
dosing.

Response:
This
comment
is
at
odds
with
the
earlier
comment
that
there
are
no
species
differences
that
would
warrant
an
interspecies
factor.
The
relevant
human
data
which
substantiates
AMVAC's
conclusion
need
to
be
submitted
before
a
correction
is
considered.

32.
Pages
109­
111,
Paragraphs
3­
4,
AMVAC
states
that
there
are
adequate
data
in
the
two
DNT
studies
when
the
results
are
combined.

Response:
These
comments
have
been
addressed
in
Item
#
9.

33.
Page
113.
EPA
should
be
using
the
Revised
Report
"
Revision
001­
Repeat
Dose
Cholinesterase
Inhibition
Study
in
Pre­
Weaning
and
Young
Adult
Rats"
MRID
46433201
which
was
issued
on
October
24,
2003,
and
submitted
to
EPA
on
December
8,
2004,
which
replaces
MRID
46153304.
This
reanalysis
of
the
study
data
established
the
NOAELs
at
postnatal
days
18
and
48
for
brain
and
RBC
cholinesterase
inhibition
at
0.1
mg/
kg/
day.
There
were
no
clear
differences
between
pre­
weaning
and
young
adult
rats.

Response:
AMVAC
is
correct.

34.
Page
118,
Paragraph
1,
Line
2.
AMVAC
states
that
there
is
a
NOEL
for
the
critical
endpoint
of
RBC
cholinesterase
of
2
mg/
kg/
day
in
rats
based
on
lack
of
statistically
significant
decreases
in
the
endpoint.
Page
14
of
14
Response:
The
NOAEL
of
2
mg/
kg/
day
was
identified
in
the
13­
week
range
finding
study
and
is
below
the
dose
levels
used
in
the
main
2­
year
rat
study.

35.
Page
118,
Paragraph
6,
Line
1.
AMVAC
states
there
is
a
NOEL
for
the
critical
endpoint
of
RBC
cholinesterase
of
10
mg/
kg/
day
in
mice
based
on
lack
of
statistically
significant
decreases
in
the
endpoint.

Response:
The
DER
for
the
mouse
carcinogenicity
study
does
not
identify
the
10
mg/
kg/
day
dose
level
as
a
NOAEL
for
RBC
ChE
depression.

36.
Page
128,
paragraph
1,
line
12
Temperature
is
misspelled.

Response:
Corrected.

37.
Page
128,
paragraph
4,
line
1
EPA
has
erred
in
not
discussing
the
large
number
of
human
studies
on
DDVP
in
humans.
These
studies
have
important
information
relevant
to
interspecies
uncertainty
factor
choices,
intraspecies
uncertainty
factor
choices,
FQPA
factor
choices,
and
enzyme
kinetics
of
DDVP.

Response:
Many
of
the
human
studies
to
which
AMVAC
is
referring
were
considered
not
relevant
to
the
Dichlorvos
pesticide
risk
assessment.

Some
of
the
human
studies
have
been
reviewed
separately,
and
were
considered
not
relevant.
See
Dannon
G
&
Dapson
S.
1999;
Sette,
W.
F.
1998b;
Stewart
J.
1993;
and
Stewart
J.
1997.
