Richard
A.

Becker,

Ph.
D.,

D.
A.
B.
T
American
Chemistry
Council
Arlington,
Virginia
Comments
on
"Dose
Setting"

EDMVS 
Meeting 
J
ly 

23­
24, 

2002
Dose
Setting
Considerations
°

Dose­
setting
is
a
serious
consideration
for
validation
and
pre­

validation
exercises
and
must
be
handled
carefully
and
correctly.
°

Range
finding
assays
cannot
be
validated
or
standardized
apart
from
the
specific
assay
to
which
they
apply.
°

No
single
range
finding
approach
will
apply
across
all
assays
and
tests.

Criteria
and
endpoints
for
setting
maximum
doses
will
vary
depending
on
the
specific
assay.

For
example
10%

body
weight
decrement
is
not
an
appropriate
criteria
for
pubertal
models.
°

Attempts
should
be
made
to
minimize
the
use
of
animals
for
range­
finding
studies;
however,

it
must
also
be
recognized
that
unless
a
sufficient
number
of
animals
are
used
to
produce
a
reliable
range­
finding
study,
those
animals
that
are
used
may
be
wasted
generating
useless
data.
Dose
Setting
Considerations
°

Up/

down­
type
range
finding
protocols
are
not
appropriate
for
studies
in
which
animals
are
exposed
via
the
diet
or
drinking
water.
°

Dose­
setting
must
control
for
the
ability
of
dietary
restriction
and
resultant
body
weight
effects
to
impact
the
endpoints
used
to
infer
endocrine
activity.
°

Stress
alone
is
a
sufficient
stimulus
to
alter
endocrine
function.
Dose­
setting
protocols
should
be
specifically
designed
to
identify
doses
that
can
be
given
without
significantly
stressing
the
animals.
Dose
Setting
Considerations
°

Pre­

validation
exercises
should
include
specific
analysis
of
general
mechanisms
of
toxicity
(in
addition
to
endocrine­
specific
pathways)

­

in
order
to
ascertain
information
re:

the
specificity
of
endpoints.
Substances
for
Validation
Studies
of
Screening
Assays
1.

The 

hormonal 
a
tivity 
and 
me
hanism 
of 

hormonal 
effe
t 
of 
a 

substan
e 

should 
already 
be 

known 
from 
both 
in 

vitro 
and 
in 

vivo 

resear
h 

methods. 
There 
must 
be
suffi
ient 
and 

robust 

information 
and 
data 
from 
s
ientifi
 
reports 
on 
ea
h 

substan
e 

with 
respe
t 

to 

the 

hormonal 
mode 
of 
a
tion, 
the 

hormonal 
poten
y 

and 
spe
ifi
ity
and 

ADME2 
hara
teristi
s. 

These 
data 

enable 
a 
predi
tion 
of 

results 
for 
the 
s
reenin
 method 
and 
a 
reasonable 
assessment 
of 

proto
ol 

performan
e.
2. 

Substan
es 

sele
ted 

must 
be 

readily 
available 
throu
h 
ommer
ial 

vendors. 
These 
substan
es 

are 

likely 
to 
be 

used 
over 
a 

number 
of 

years, 
in 

several 
proto
ols 
and 
by
a 

number 
of 

laboratories 
s 

part 
of 

the 

standardization 
and 
validation 
pro
ram.
Further, 
other 
labs 
will 

have 
an 

interest 
to 

establish 
and 
demonstrate 
their 
profi
ien
y 

with 

these 
s
reenin
 methods. 
Therefore, 
it 
is 
ne
essary 
to 

sele
t 

substan
es 
whi
h 

will 
be 
readily 
available 
throu
h 

ommer
ial 

sour
es 

presently
and 
in 

the 

future.
3. 

The 
A
en
y 

must 
fo
us 
on 

substan
es 

with 

known 
estro
en, 
andro
en 

and 

thyroid (
EAT) 
a
tivity, 

onsistent 
with 
the 
A
en
y's 

EDSP 

Statement 
of 
Poli
y. 

The 

priority 
for 
the 
EDSP 
should 
be 

estro
en, 

andro
en 

and 

thyroid 
hormonal 
a
tivities 
or 

modes 
of 
a
tion. 
The 
fo
us 

should 
be 
on 

dire
t 
modes 
of 

EAT 
a
tions 
and 

should 
in
lude 
re
eptor 
a
onists/
anta
onists 
and, 
if 

appli
able, 

hormone 
synthesis
inhibitors. 
Importantly, 
the 
A
en
y 

should 
avoid 
use 
of 
substan
es 

that 
exert 
endo
rine 
effe
ts 

via 

indire
t 

modes 
or 
me
hanisms (
ex
ept 
to 

establish 
spe
ifi
ity, 
as 
des
ribed 
in 

point 
7 

below)
4. 

Substan
es 

with 
hi
h 

spe
ifi
ity (

either 
as 
a
onists 
or 
anta
onists) 
are 

preferred 
and 

should 
be 

used 
to 

the 
maximum 
extent 
pra
ti
able. 
In 

ases 

where 
the 
use 
of 
a
mixed 
a
onist/
anta
onist 
is 
ne
essary 
or 

where 
there 
are 
other 

overlappin
 spe
ifi
ities, 
EPA 
must 
sele
t 

the 
on
entrations 
and 

doses 

arefully, 
keepin
 in 

mind 
the 
effe
ts 
su
h 

mixed 
a
tivities 
may 
have 
upon 
the 

type, 
ma
nitude 
and 

nature 
of 

the 

response(
s).
5. 

Substan
es 

with 
parti
ular 
EAT 
a
tivity 
should 
be 

evaluated 
in 

the 

appropriate 
s
reenin
 method. 
While 
there 
may 
be 
some 

overlap, 
it 
is 

not 
ne
essary 
to 

use 
exa
tly 
the 

same 
set 
of 

substan
es 
in 

the 

validation 
of 
ea
h 
s
reenin
 
method. 
For 

example, 
substan
es 

with 
estro
eni
 a
tivity 
should 
be 

used 
for 

validation 
of 
the
uterotrophi
assay, 
but 
it 

would 
make 
no 

sense 
to 

use 
the 

same 

omplete 
set 
of 
substan
es 
in 

the
Hershber
er
assay 
for 

andro
ens.
6. 
In 

eneral, 

validation 
must 
over 
the 

entire 
ran
e 

of 
a
tivities 
anti
ipated 
from 
the 

population 
of 

substan
es 

that 
will 
be 
sele
ted 
to 
be 

evaluated 
with 
the 

assay. 
Little 
or 
no 
onfiden
e 

an 
be 
pla
ed 

upon 

results 
of 

substan
es 

whose 
a
tivities 
fall 

outside 
the 
a
tivities 
or 

modes 
of 
a
tion 
of 
the 
set 
of 

substan
es 

for 
whi
h 

the 

assay 
has 

been 
validated. 
Further, 
the 
set 
of 

substan
es 

used 
for 
development 
and 

standardization 
of 
an 

assay 
should 
be 
different 
from 
the 
set 
of 

substan
es 

used 
for 

validation. 
In 
the 

validation 
series, 
the 

substan
es 

sele
ted 

should 
in
lude 
materials 
with 
a 

ran
e 

of 

poten
ies; 
from 
stron
 to 

weak 
to 
ompletely 
ne
ative 
for 
the 

appropriate 
EAT 
me
hanisms.
7. 
It 
is 

essential 
to 

address 
the 

issue 
of 
spe
ifi
ity (

false 
positive 

responses) 
in 

the 

validation 
studies 
of 
ea
h 

assay. 
In 

parti
ular, 
sin
e 

the 

EDSP 
s
reenin
 assays 
and 
the 
Tier 
1 

battery 
have 
been 
sele
ted 
by 

EPA 
to 

minimize 
or 
eliminate 
false 
ne
atives, 
su
h 

hara
teristi
s 

will 

likely 
enerate 
false 

positives. 
Therefore, 
in 

the 

validation 
of 
EDSP 
s
reenin
 assays, 
it 
is 

riti
al 
to 
in
lude 

substan
es 
that 
exert 
effe
ts (
and/
or 

toxi
ity) 
by 
me
hanisms 
that 
are 
not 

primarily 
hormonal 
in 

order 
to 

establish 
the 
spe
ifi
ity 
of 

the 

assay 

endpoints (
e.

., 

evaluate 
potential 
for 
false
positive 

responses 
due 
to 
a 

non­

hormonal 
toxi
ity). 
In 

some 
ases 
it 

may 
be 

benefi
ial 
to 

establishin
 spe
ifi
ity 
by 
evaluatin
, 

for 

example, 
a 

pure 
estro
en 
a
onist 
in 
an 

assay 
desi
ned 
for 

andro
ens (
and 
vi
e 

versa).
8. 

EPA 
must 

oordinate 
its 
a
tivities 
with 
the 

OECD 
EDTA 
with 
respe
t 

to 

study 
desi
n, 

sele
tion 
of 

substan
es 

and 
dose 
levels 
for 

assay 

validation. 
OECD 
has 

initiated (
and 
for 

some 
assays, 
lar
ely 

ompleted) 
validation 
studies 
usin
 spe
ifi
hemi
al 

substan
es. 

EPA's 
a
tivities 
with 
respe
t 

to 

assay 
validation 
for 
the 

EDSP 
should 

demonstrate 
the 
A
en
y's 
stron
 support 
of 

international 
harmonization 
and 

mutual 
a
eptan
e 

of 

data.
9. 

The 

approa
h 

EPA 

adopts 
for 

standardization 
and 

validation 
should 
be 

suffi
iently 
ri
orous 
to 

omply 
with 

enerally 
re
o
nized 
s
ientifi
 prin
iples 
of 

study 
desi
n 

and 

ondu
t. 
With 
respe
t 

to 

test 
arti
les 
sele
ted 
for 

EDSP 

validation, 
this 

should 
in
lude 

knowled
e 

of 

hemi
al 

purity, 
stability 
and 

on
entration (
parti
ularly 
the 

applied 
or 

administered 
dose). 
In 

evaluatin
 substan
es 

for 

potential 
sele
tion 
for
use 
in 

parti
ular 

assays 
and 

routes 
of 

administration, 
EPA 
should 

onsider 
what 
de
ree 
of 

analyti
al 

hemistry 
would 
be 
ne
essary 
to 

meet 
these 
re
o
nized
s
ientifi
 standards.
10. 
In 

compilin
 substances 
for 

standardization 
and 

validation, 
EPA 
must 

appropriately 
qualify 
and 

characterize 
any 
and 
all 

such 
lists. 
E
STAC 
spent 
a 

reat 
deal 
of 

time 
and 

effort 

addressin
 
communications 
issues, 
and 
EPA 

should 

implement 
the 
E
STAC 
recommendations 
to 

ensure 
proper 

understandin
 by 
the 

public 
of 
such 
a 

list 
of 

substances. 
The 

Table 
must 
be 

qualified 
and 

include 
a 

disclaimer 
alon
 the 
lines 
of: 
"

Inclusion 
of 
a 

subs
ance 
in 

his 
able 
does 
no
 mean 

ha
 EPA 
has 
or 

will 

make 
a 
de
ermina
ion 
ha
 any 
of 

he 

uses 
of 

he 

chemical 
will 

pose 
a 

significan
risk. 
Fur
her, 

his 

able 

should 
no
 be 

aken 
as 
a 
lis
 of "

endocrine 
disrup
ors." 

he 

subs
ances 
lis
ed 

are 

simply 

compounds 
which 
have, 
or 

may 

prove 

o 

be, 

useful 
in 

developing, 
s
andardizing 
or
valida
ing 

screening 
and 

es
ing 
me
hods."
11. 
Ea
h 

entry 
in 
whi
h 

referen
e 
is 

made 
to 
a 

parti
ular 
hormonal 
me
hanism 
of 
a
tion 
or 
to 

poten
y 

or 
a
tivity 
must 
be 

referen
ed. 
This 
is 
ne
essary 
for 

transparen
y 

and
a
ura
y. 

This 
would 
permit 

members 
of 

the 

EDMVS (
and 
the 
publi
) 

to 

readily 
a
ess 
the 

itation 
and 
to 

review 
the 
a
tual 

study 

results (
study 
desi
n, 

dose 
levels,
endpoints 
measured 
and 

results). 
This 
is 

riti
al 

and 
is 
ne
essary 
for 
sele
tion 
of 

hemi
als 
and 
dose 
levels 
for
prevalidation
studies 9
it 

also 

important 
for 

onstru
tin
 the 
predi
tive 

models.
