Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
1
Chapter
IX.
Risk
Characterization
Summary
A.
Hazard
Characterization
and
Mode
of
Action
Implications
The
data
on
health
effects
of
cyanogen
chloride
are
very
limited,
and
so
there
is
significant
uncertainty
in
the
cyanogen
chloride
hazard
characterization.
The
available
studies
on
cyanogen
chloride
are
useful
primarily
for
hazard
identification,
but
are
not
sufficiently
quantitative
or
of
an
appropriate
design
to
be
useful
for
dose­
response
assessment.
A
single
study
of
the
health
effects
of
occupational
exposure
to
cyanogen
chloride
reported
dizziness,
nausea,
and
prostration
lasting
several
hours
after
high­
exposure
incidents
(
Reed,
1920).
Chronic
symptoms
included
weakness,

and
eye,
nose
and
throat
irritation.
Other
studies
of
short­
term
exposure
of
humans
reported
rapid
paralysis
of
nerve
centers,
as
well
as
eye,
throat,
and
lung
irritation
(
Prentiss,
1937;
Flury
and
Zernik,
1931).
Except
for
the
irritative
effects,
which
may
be
due
to
the
parent
compound,

the
observed
symptoms
are
consistent
with
those
seen
following
cyanide
exposure.
The
animal
data
on
cyanogen
chloride
are
limited
to
acute
studies
of
rats,
cats,
and
dogs
exposed
by
either
inhalation
or
injection
before
the
advent
of
modern
toxicology
methods
(
Flury
and
Zernik,
1931;

Reed,
1920;
Aldridge
and
Evans,
1946;
Haymaker
et
al.,
1952).
These
studies
are
unsuitable
for
risk
assessment
because
key
information
about
doses
and
study
design
were
not
reported.

However,
the
studies
do
help
to
identify
the
potential
health
effects
of
cyanogen
chloride,
which
include
neurotoxicity
(
depressed
reflexes,
muscular
tremors,
rigidity
and
weakness
of
limbs,

convulsions,
brain
necrosis),
weight
loss,
cardiac
irregularities,
and
pulmonary
congestion
and
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
2
edema.
The
animal
and
human
data
also
indicate
that
cyanogen
chloride
irritates
the
respiratory
passages
and
eyes.
Based
on
the
human
and
animal
data,
high
short­
term
exposure
to
cyanogen
chloride
vapor,
and
longer­
term
exposure
to
unknown
levels,
causes
neurological
effects,
and
respiratory
and
ocular
irritation.
The
minimum
exposure
levels
causing
either
short­
term
or
longterm
effects
are
unknown.
With
the
exception
of
the
irritative
effects
from
inhalation
of
cyanogen
chloride,
the
limited
data
on
cyanogen
chloride
toxicity
are
consistent
with
the
toxic
effects
of
its
known
primary
metabolites,
cyanide
and
thiocyanate.

Due
to
this
lack
of
toxicity
data
on
cyanogen
chloride,
surrogates
were
considered
as
the
basis
for
the
cyanogen
chloride
assessment.
Toxicokinetics
data
(
Aldridge
and
Evans,
1946;

Aldridge,
1951)
show
that
cyanogen
chloride
is
rapidly
metabolized
to
cyanide
(
CN­),
and
thiocyanate
(
SCN­),
two
chemicals
for
which
much
more
toxicity
information
is
available.

Metabolism
to
either
of
these
compounds
would
also
result
in
production
of
HCl,
although
HCl
production
has
not
been
directly
evaluated.
The
reduction
of
cyanogen
chloride
to
cyanide
appears
to
occur
via
reaction
with
the
­
SH
group
of
glutathione,
hemoglobin,
or
other
proteins.

Since
cyanide
is
metabolized
to
thiocyanate,
it
is
unknown
whether
the
thiocyanate
seen
after
cyanogen
chloride
exposure
is
formed
directly
from
cyanogen
chloride,
or
as
a
metabolite
of
cyanide.
Water
chemistry
data
also
suggest
that
cyanogen
chloride
can
react
to
form
cyanate
(
OCN­)
or
cyanamide
(
CH
2
N
2),
and
can
react
with
other
nucleophiles,
although
these
other
reactions
have
not
been
confirmed
in
animals
dosed
with
cyanogen
chloride,
or
even
in
in
vitro
studies
of
metabolism.
Since
cyanogen
chloride
is
metabolized
almost
immediately,
it
is
likely
that
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
3
the
systemic
toxicity
of
cyanogen
chloride
is
due
to
the
activity
of
metabolites
rather
than
the
parent
compound.
Unfortunately,
no
mass­
balance
toxicokinetic
studies
of
cyanogen
chloride
have
been
conducted,
so
no
quantitative
identification
of
metabolites
is
possible.
The
irritative
properties
of
cyanogen
chloride
indicate
that
ingestion
of
high
concentrations
may
directly
affect
the
gastrointestinal
tract,
but
such
effects
are
unlikely
at
environmental
exposure
levels.
Because
cyanogen
chloride
is
rapidly
metabolized
in
the
blood,
and
reacts
directly
with
glutathione,
which
is
present
in
the
liver
at
high
concentrations,
first­
pass
metabolism
may
lead
to
marked
differences
between
the
toxicokinetics
of
inhaled
and
ingested
cyanogen
chloride.
Based
on
these
considerations,
the
following
text
focuses
on
the
known
metabolites
of
cyanogen
chloride.

The
toxicity
of
thiocyanate
is
generally
well
known.
Animal
(
Nagasawa
et
al.,
1980;

Pyska,
1977;
Philbrick
et
al.,
1979;
Kanno
et
al.,
1990)
and
human
(
Dahlberg
et
al.,
1984;

Banerjee
et
al.,
1997)
data
show
that
the
primary
target
of
thiocyanate
is
the
thyroid,
although
no
toxicity
studies
have
thoroughly
evaluated
other
organs
and
tissues,
so
toxicity
to
other
targets
cannot
be
ruled
out.
Human
data
on
thiocyanate
exposures
leading
to
changes
in
thyroid
hormone
levels
(
Dahlberg
et
al.,
1984;
Banerjee
et
al.,
1997)
form
the
basis
for
the
thiocyanate
RfD.
These
studies
are
supported
by
another
human
study
(
Beamish
et
al.,
1954),
which
reported
effects
at
a
similar
concentration
of
thiocyanate
in
the
plasma,
but
did
not
provide
any
information
on
dose
levels.
The
ionic
size
of
thiocyanate
is
similar
to
that
of
iodine,
and
thiocyanate
acts
by
blocking
the
active
transport
of
iodine
into
the
thyroid
gland,
and
by
causing
iodine
already
accumulated
in
the
thyroid
to
be
discharged
(
Wolff,
1998).
Although
no
standard
multigeneration
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
4
reproduction
or
developmental
toxicity
studies
of
thiocyanate
were
located,

reproductive/
developmental
toxicity
studies
that
focused
on
the
thyroid
have
found
effects
such
as
increased
thyroid
weight
and
changes
in
the
levels
of
thyroid
hormones
(
Raghunath
and
Bala,

1998;
Bala
et
al.,
1996;
Kreutler
et
al.,
1978;
Pyska,
1977).

Potential
neurodevelopmental
toxicity
is
a
key
unresolved
issue
for
both
thiocyanate
and
cyanide
(
and,
by
implication,
for
cyanogen
chloride),
due
to
the
potential
for
these
chemicals
to
induce
hypothyroidism
in
fetuses
and
neonates.
While
hypothyroidism,
defined
by
increased
serum
levels
of
TSH
and
decreased
levels
of
T3
and
T4,
causes
goiter
in
adults,
the
effects
of
hypothyroidism
in
infants
(
stunted
growth
and
mental
development)
are
more
severe.
In
humans,

congenital
hypothyroidism,
or
cretinism,
is
characterized
by
long­
term
effects
on
behavior,

locomotor
ability,
speech,
hearing
and
cognition
(
Chan
and
Kilby,
2000).
The
LOAELs
for
decreased
thyroid
hormone
levels
in
animal
developmental­
toxicity
studies
(
Raghunath
and
Bala,

1998;
Bala
et
al.,
1996)
are
more
than
an
order
of
magnitude
higher
than
the
NOAEL/
LOAEL
boundary
for
changes
in
thyroid
hormone
levels
in
humans
(
Dahlberg
et
al.,
1984;
Banerjee
et
al.,

1997)
that
forms
the
basis
for
the
thiocyanate
RfD.
It
is
not
clear,
however,
if
there
is
a
direct
correlation
between
serum
T3
levels
and
brain
T3
levels.
The
brain
changes
in
congenital
hypothyroidism
appear
to
be
mediated
by
circulating
T4,
which
is
converted
in
the
astrocytes
to
T3;
it
is
the
brain
T3
levels
that
appear
to
be
key
for
determining
neurodevelopmental
effects
(
Koibuchi
and
Chin,
2000).
The
degree
of
decrease
in
serum
T4
that
results
in
neurodevelopmental
effects
has
not
been
determined.
In
addition,
thyroid
hormones
are
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
5
controlled
by
a
negative
feedback
loop,
and
rats
rapidly
up­
regulate
serum
thyroid
hormone
levels.
Because
of
this
rapid
feedback,
the
peak
degree
of
change
in
hormone
levels
in
the
developmental
toxicity
studies
could
have
been
much
larger
at
earlier
time
points.
Since
developmental
toxicity
depends
on
exposure
during
a
critical
developmental
window,
a
large,
but
brief,
change
in
thyroid
hormone
levels
might
result
in
marked
neurodevelopmental
effects.
This
means
that
the
ratio
between
the
free­
standing
LOAELs
identified
for
changes
in
thyroid
hormone
levels
in
rat
pups
and
the
short­
term
NOAEL
(
in
the
same
species)
for
this
effect
may
be
more
than
an
a
factor
of
ten.
Because
a
short­
term
decrease
in
T4
levels
(
of
sufficient
magnitude)
can
affect
neurodevelopment
if
the
change
occurs
during
the
window
of
sensitivity,
even
short­
term
deviations
are
of
concern.
Based
on
these
considerations,
a
neurodevelopmental
toxicity
study
would
be
important
in
determining
whether
the
brain
is
affected
at
doses
below
those
observed
to
affect
serum
hormone
levels
in
rat
pups.
If
neurodevelopmental
toxicity
occurs
at
these
lower
doses,
fetuses
and
children
would
constitute
a
sensitive
population.

People
with
hypothyroid
disorders,
and
protein
or
iodine
deficiency
may
also
be
more
sensitive
to
the
thyroid
effects
of
thiocyanate
(
Kreutler
et
al.,
1978).
Protein
deficiency
is
rare
in
Western
populations
(
U.
S.
FDA,
1999),
but
the
percentage
of
the
U.
S.
population
that
may
be
at
risk
for
hypothyroidism
resulting
from
iodine
deficiency
may
be
as
much
as
a
few
percent
(
Hollowell
et
al.,
1998).
Studies
in
human
populations
that
eat
cyanide­
containing
foods,
such
as
cassava,
also
suggest
that
an
increased
susceptibility
to
thyroid
effects
may
be
associated
with
deficiencies
of
protein,
iodine,
vitamin
B
12,
or
other
vitamins,
although
an
etiological
agent
other
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
6
than
cyanide
has
also
been
proposed
as
the
cause
of
the
tropical
neuropathies
(
reviewed
in
ATSDR,
1997).

The
central
nervous
system
is
the
primary
target
of
acute
exposure
to
cyanide,
with
symptoms
including
depression,
convulsions,
coma
and
death.
Cyanide
appears
to
exert
its
acute
toxic
effects
by
binding
with
cytochrome
c
oxidase,
which
then
becomes
unable
to
catalyze
the
reactions
that
transfer
electrons
from
reduced
cytochrome
c
to
oxygen.
These
effects
have
been
reported
in
animals
and
humans
following
high­
level
exposures.

CNS
effects
are
also
observed
following
longer­
term
exposure
to
cyanide,
with
myelin
degeneration
of
the
spinal
cord
observed
in
rats
exposed
in
food
for
1
year
(
Philbrick
et
al.,
1979).

Long­
term
exposure
to
cyanide
also
causes
thyroid
effects
due
to
the
cyanide
metabolite
thiocyanate.
Philbrick
et
al.
(
1979)
reported
decreased
T4
levels
and
increased
thyroid
weights,

but
no
histopathology
changes
in
the
thyroid.
A
recent
subchronic
study
conducted
up
to
slightly
lower
doses
(
NTP,
1993)
did
not
report
histopathology
in
the
nervous
system
or
thyroid,
but
did
observe
male
reproductive
effects
(
decreased
epididymal
weight,
testis
weight,
and
sperm
count).

Thyroid
hormone
levels
were
not
measured
in
this
study.
No
data
were
located
that
describe
the
mechanism
of
cyanide's
toxicity
on
the
male
reproductive
system,
although
NTP
(
1993)

suggested
that
the
effects
may
be
related
to
perturbations
in
hormonal
balance.
The
male
reproductive
effects
observed
by
NTP
(
1993)
are
the
critical
effects
for
the
development
of
the
cyanide
RfD.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
7
The
database
for
cyanide
includes
subchronic
studies
in
rats,
mice,
and
dogs
(
Philbrick
et
al.,
1979;
Kamalu,
1993;
NTP,
1993),
and
a
chronic
study
that
did
not
evaluate
the
thyroid
(
Howard
and
Hanzal,
1955).
No
study
on
the
effects
of
cyanide
on
reproductive
function
in
males
have
been
located,
and
no
standard
developmental
toxicity
studies
of
cyanide
are
available.

The
absence
of
a
study
evaluating
reproductive
function
and
of
a
developmental
toxicity
study
are
of
particular
concern,
in
light
of
the
histopathological
changes
to
the
male
reproductive
organs
(
NTP,
1993).
Since
exposure
to
low
levels
of
cyanide
may
potentially
cause
thyroid
hormone
changes,
and
agents
that
affect
thyroid
hormone
levels
may
affect
neurological
development,
the
potential
neurodevelopmental
toxicity
of
cyanide
is
an
important
data
gap.
If
neurodevelopmental
toxicity
occurs
at
these
lower
doses,
fetuses
and
children
would
constitute
a
sensitive
population.

As
for
thiocyanate,
people
with
hypothyroid
disorders,
or
deficiencies
in
protein,
iodine,
vitamin
B
12,
or
other
vitamins
may
also
be
more
sensitive
to
the
thyroid
effects
of
cyanide.

A
further
uncertainty
for
cyanide
is
whether
the
observed
lesions
progress
with
increasing
exposure
duration.
Although
no
data
are
available
on
progression
of
testicular
effects,
it
does
not
appear
that
liver
effects
progress
with
increasing
exposure
to
cyanide.
Based
on
the
increased
liver
weight
in
rats
exposed
to
12
mg
CN/
kg­
day
for
21
days
(
Palmer
and
Olson,
1979)
and
the
absence
of
an
effect
on
liver
weight
at
the
same
dose
in
male
rats
exposed
for
13
weeks
(
NTP,

1993),
also
via
drinking
water,
there
was
no
progression
in
liver
effects
(
actually,
a
regression)

with
increased
exposure
duration.
Similarly,
progression
would
not
be
expected
for
the
thyroid
effects,
because
the
thyroid
responds
rapidly
to
iodine­
like
ions,
and
thiocyanate
does
not
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
8
accumulate
in
the
body.
Together,
these
data
suggest
that
cyanide's
effects
on
the
male
reproductive
tract
may
also
not
progress,
but
no
data
directly
addressing
that
issue
are
available.

In
light
of
the
paucity
of
data
on
cyanogen
chloride,
there
are
a
number
of
data
gaps
for
this
assessment.
A
primary
issue
is
the
quantitative
identification
of
metabolites
to
address
whether
metabolites
in
addition
to
cyanide
and
thiocyanate
are
produced.
Ideally,
these
data
would
be
collected
from
an
in
vivo
metabolism
study
in
which
all
metabolites
are
tracked
and
quantified,
although
an
in
vitro
study
using
cell
extracts
that
investigates
and
quantifies
all
metabolites
would
help
to
address
these
issues.
A
related
issue
is
determining
the
relative
amounts
of
cyanide
and
thiocyanate
produced
from
dosing
with
cyanogen
chloride,
particularly
since
thiocyanate
can
be
converted
back
to
cyanide
in
the
body.
In
the
absence
of
any
toxicity
data
on
cyanogen
chloride
collected
according
to
modern
toxicology
methods,
a
subchronic
oraltoxicity
study
that
meets
EPA
test
guidelines
is
needed
to
meet
the
minimal
database
requirement
for
development
of
a
cyanogen
chloride
RfD
from
cyanogen
chloride
data.
Based
on
information
from
cyanide,
thiocyanate,
and
the
putative
metabolites,
particular
attention
in
such
a
study
should
be
paid
to
the
nervous
system
(
e.
g.,
histopathology
that
could
detect
demyelination),
thyroid,

male
reproductive
tract,
and
liver;
a
developmental
toxicity
study
(
particular
neurodevelopmental
toxicity)
would
also
be
useful.
In
the
absence
of
such
a
high­
quality
subchronic
study
of
cyanogen
chloride,
a
short­
term
study
that
includes
a
thorough
histopathology
analysis
could
at
least
confirm
key
targets
of
cyanogen
chloride,
and
aid
in
the
determination
of
the
appropriate
surrogate(
s).
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
9
B.
Dose­
Response
Characterization
and
Implications
in
Risk
Assessment
In
the
absence
of
adequate
data
on
cyanogen
chloride,
data
on
surrogates
were
used
to
develop
the
cyanogen
chloride
RfD.
The
data
are
insufficient
for
the
derivation
of
an
RfC
for
cyanogen
chloride,
particularly
since
cyanogen
chloride
appears
to
cause
portal­
of­
entry
effects
that
may
be
due
to
the
parent
compound.
In
the
development
of
the
RfD,
data
on
cyanide
and
thiocyanate
were
emphasized,
because
these
are
the
only
metabolites
that
have
been
identified.

In
the
absence
of
adequate
human
or
animal
data
on
cyanogen
chloride,
the
cyanogen
chloride
RfD
was
derived
from
the
RfD
for
free
cyanide,
taking
into
account
the
molecular
weights
of
cyanogen
chloride
and
free
cyanide
(
CN).
Using
the
NOAEL/
LOAEL
approach,
a
cyanide
RfD
of
0.005
mg
CN/
kg­
day
was
calculated,
corresponding
to
a
cyanogen
chloride
RfD
of
0.01
mg/
kg­
day.
Using
the
benchmark
dose
approach,
a
cyanide
RfD
of
0.0008
mg
CN/
kg­
day
was
calculated,
corresponding
to
a
cyanogen
chloride
RfD
of
0.002
mg/
kg­
day.
The
updated
cyanide
RfD
was
based
on
male
reproductive
effects
(
decreased
epididymal
weight,
testis
weight,

and
sperm
count)
in
a
subchronic
drinking­
water
study
in
rats
(
NTP,
1993).
There
is
some
uncertainty
in
the
identification
of
the
NOAEL,
since
decreases
in
left
cauda
epididymis
weight
and
sperm
motility
were
observed
at
the
NOAEL
of
4.5
mg
CN/
kg­
day,
and
at
the
next
lower
dose
of
1.4
mg
CN/
kg­
day.
NTP
(
1993)
did
not
consider
the
effects
on
sperm
motility
to
be
adverse
because
they
were
small
and
well
within
the
range
of
historical
controls.
Similarly,
the
decrease
in
left
cauda
epididymus
weight
was
small
at
1.4
and
4.5
mg/
kg­
day,
and
there
was
little
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
10
evidence
of
a
dose­
response
at
these
doses.
The
uncertainty
in
the
identification
of
the
NOAEL
is
reflected
in
the
fact
that
the
study
BMDL
of
0.79
mg/
kg­
day
(
for
decreased
left
epididymis
weight)
is
much
lower
than
the
study
NOAEL.
This
BMDL
was
considered
appropriate
as
an
alternative
basis
for
the
RfD,
because
the
effect
is
biologically
significant
and
an
adequate
fit
to
the
data
was
obtained.
In
addition,
this
BMDL
is
supported
by
a
BMDL
of
1.3
mg/
kg­
day
obtained
in
the
same
study
for
two
different
measures
of
decreased
spermatid
count.

Based
on
these
RfDs,
a
Lifetime
HA
for
cyanogen
chloride
of
0.08
mg/
kg­
day
(
based
on
the
cyanide
NOAEL)
or
0.01
mg/
kg­
day
(
based
on
the
cyanide
BMDL)
was
calculated.

A
total
uncertainty
factor
of
1000
was
used
to
derive
the
cyanide
RfD.
Full
factors
of
10
were
used
to
account
for
extrapolation
from
animals
and
for
human
variability.
A
factor
of
10
was
used
to
account
for
both
subchronic
to
chronic
extrapolation
and
for
deficiencies
in
the
database
(
combined).
No
standard
multigeneration
or
developmental­
toxicity
studies
of
cyanide
are
available.
As
noted
above,
the
absence
of
a
study
evaluating
reproductive
function
and
of
a
developmental
toxicity
study
(
particularly
a
neurodevelopmental
study)
are
of
particular
concern,

since
the
current
critical
effect
is
histopathological
changes
to
the
male
reproductive
organs
(
NTP,

1993)
and
agents
that
affect
thyroid
hormone
levels
may
also
affect
neurological
development.

The
combined
uncertainty
factor
was
based
on
an
endpoint­
by­
endpoint
evaluation
of
the
potential
for
progression
with
increased
dose
duration,
and
the
implications
for
database
deficiencies.
Factors
considered
included:
(
1)
the
rapid
response
of
the
thyroid
to
the
cyanide
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
11
metabolite
thiocyanate
and
the
lack
of
progression
of
thyroid­
related
effects
from
subchronic
to
chronic
exposure;
(
2)
the
general
absence
of
progression
seen
in
studies
of
cyanide
toxicity;
(
3)

the
uncertainty
as
to
whether
the
effects
on
the
male
reproductive
tract
are
caused
by
cyanide
or
thiocyanate;
and
(
4)
the
absence
of
information
specifically
on
progression
of
reproductive
effects
from
cyanide
toxicity.

There
is
medium
confidence
in
the
principal
study
for
the
cyanide
RfD.
The
study
(
NTP,

1993)
was
generally
well­
conducted,
using
an
appropriate
number
of
animals
for
that
study
duration
(
20/
sex/
dose),
and
examined
a
range
of
endpoints,
but
it
did
not
evaluate
thyroid
hormone
levels,
the
likely
critical
effect.
The
database
includes
systemic
toxicity
studies
in
several
species,
but
does
not
include
any
standard
reproductive
or
developmental
toxicity
studies,
or
any
evaluations
of
effects
on
neurodevelopment.
Accordingly,
there
is
medium
confidence
in
the
cyanide
database,
and
in
the
overall
cyanide
RfD.

For
the
cyanogen
chloride
RfD,
the
same
medium­
confidence
principal
study
was
used.

Confidence
in
the
database
is
low,
however,
since
extrapolation
from
a
surrogate
was
needed.

The
resulting
confidence
in
the
cyanogen
chloride
RfD
is
low.
Indeed,
in
light
of
the
lack
of
data
for
cyanogen
chloride,
the
database
for
cyanogen
chloride
is
far
less
complete
than
the
minimal
database
in
order
to
derive
a
"
low
confidence"
RfD.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
12
Derivation
of
the
cyanogen
chloride
RfD
from
the
thiocyanate
RfD
was
also
considered.

The
thiocyanate
RfD
of
0.001
mg/
kg­
day
was
calculated
based
on
two
human
studies
defining
a
NOAEL/
LOAEL
boundary
for
effects
on
the
thyroid.
Dahlberg
et
al.
(
1984)
observed
no
effects
on
serum
T3,
T4,
TSH,
or
the
T3:
T4
ratio
in
37
volunteers
administered
8
mg/
day
thiocyanate
in
milk
(
0.11
mg
SCN/
kg­
day)
for
12
weeks,
and
Banerjee
et
al.
(
1997)
observed
decreased
serum
T4
and
increased
TSH
compared
to
matched
controls
in
35
women
in
India
who
ingested
thiocyanate
in
milk
for
at
least
5
years.
There
is
considerable
uncertainty
in
the
doses
in
these
studies,
since
subject
body
weights
were
not
reported.
Dahlberg
et
al.
(
1984)
was
conducted
in
Sweden
with
men
and
women,
while
Banerjee
et
al.
(
1997)
was
conducted
in
India
with
women
only;
average
body
weights
of
both
of
these
populations
are
likely
to
be
rather
different
from
those
in
the
United
States.
[
For
this
assessment,
the
usual
default
average
body
weight
of
70
kg
was
assumed
for
the
Swedish
cohort,
based
on
the
assumption
that
this
cohort
would
be
comparable
to
the
U.
S.
population.
An
average
body
weight
of
60
kg
was
used
for
the
Indian
cohort,
because
it
included
only
women,
and
because
Indian
women
tend
to
be
smaller
than
the
average
American
woman,
for
whom
the
default
average
body
weight
is
65.4
kg
(
U.
S.
EPA,

1997b)].
In
addition,
there
was
inter­
subject
variability
in
dose
within
a
study,
since
body
weight
varied
and
all
of
the
subjects
received
the
same
dose
in
mg/
day.
It
is
also
unknown
whether
the
Indian
population
had
a
borderline
protein
intake,
which
would
mean
that
they
constitute
a
sensitive
population.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
13
A
total
uncertainty
factor
of
100
was
used
to
derive
the
thiocyanate
RfD.
A
default
uncertainty
factor
of
10
was
applied
to
protect
sensitive
subpopulations.
A
factor
of
10
was
applied
for
database
deficiencies,
because
a
developmental
NOAEL
for
effects
on
the
thyroid
has
not
been
identified.
This
is
of
particular
concern
because
neurodevelopmental
effects
are
known
to
result
from
decreased
T4
levels,
and
no
neurological
effects
were
evaluated
in
pups
exposed
to
thiocyanate
during
gestation.

There
are
several
uncertainties
in
using
a
surrogate
for
the
estimation
of
the
cyanogen
chloride
RfD
and
DWEL.
A
primary
uncertainty
is
for
the
estimation
of
the
cyanogen
chloride
RfD
is
the
choice
of
a
surrogate.
Cyanide
was
selected
as
the
surrogate
for
cyanogen
chloride
based
on
the
fact
that
it
is
a
known
metabolite
that
represents
a
major
part
of
the
cyanogen
chloride
dose.
Although
the
cyanide
RfD
is
not
the
lowest
RfD
of
the
known
or
potential
metabolites,
the
toxicokinetics
of
ingested
cyanide
and
cyanogen
chloride
are
likely
to
be
sufficiently
similar
that
cyanide
is
a
reasonable
surrogate.
In
contrast,
the
toxicokinetics
of
thiocyanate
formation
from
cyanogen
chloride
are
unknown.
Given
that
a
portion
of
an
ingested
cyanide
dose
would
react
or
be
excreted
prior
to
metabolism
to
thiocyanate,
the
derivation
of
a
thiocyanate
RfD
that
is
lower
than
the
cyanide
RfD
is
not
inconsistent
with
the
fact
that
cyanide
is
metabolized
to
thiocyanate.
As
discussed
in
Chapter
8,
it
is
appropriate
to
use
cyanide
as
a
surrogate
instead
of
thiocyanate,
even
though
some
cyanogen
chloride
may
be
metabolized
directly
to
thiocyanate
and
the
thiocyanate
RfD
is
lower
than
the
cyanide
RfD.
This
is
because
(
1)

there
would
be
considerable
uncertainty
in
using
thiocyanate
as
a
surrogate,
since
there
is
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
14
inadequate
quantitative
information
on
the
degree
of
cyanogen
chloride
metabolism
directly
to
thiocyanate,
although
a
rough
estimate
is
possible
by
comparison
with
the
amount
of
cyanogen
chloride
metabolized
to
cyanide;
(
2)
a
relatively
small
amount
of
thiocyanate
(<
20%
of
the
cyanogen
chloride
dose)
would
be
formed
directly
from
cyanogen
chloride
at
environmental
exposure
levels;
and
(
3)
the
conservatism
lost
by
not
using
thiocyanate
is
well
within
the
uncertainty
of
the
method.
Less­
than­
complete
conversion
of
cyanogen
chloride
to
cyanide
and
thiocyanate,
such
as
by
reaction
with
stomach
contents
or
formation
of
other
metabolites,
would
reduce
the
tissue
dose
of
cyanide
and
thiocyanate,
adding
additional
conservatism
to
the
assessment.
Cyanate
and
cyanamide
have
not
been
identified
after
in
vivo
or
in
vitro
metabolism
of
cyanogen
chloride,
and
would
account
for
a
relatively
small
portion
(
if
any)
of
the
ingested
cyanogen
chloride
dose,
so
they
are
less
appropriate
surrogates
than
cyanide.
An
additional
uncertainty
is
that
the
critical
effect
for
cyanide,
male
reproductive
toxicity,
was
not
evaluated
in
the
limited
studies
on
cyanogen
chloride.
However,
it
is
plausible
that
male
reproductive
toxicity
could
be
observed
following
long­
term
exposure
to
cyanogen
chloride,
since
cyanide
is
a
major
metabolite.

A
second
uncertainty
is
how
to
extrapolate
quantitatively
from
the
cyanide
RfD
to
cyanogen
chloride.
For
this
assessment,
a
simple
adjustment
was
made
based
on
molecular
weight.
However,
the
degree
of
conversion
of
cyanogen
chloride
to
cyanide
at
environmentally
relevant
doses
is
unknown,
and
it
is
not
known
whether
cyanide
and
thiocyanate
together
account
for
all
of
the
cyanogen
chloride
metabolism.
Cyanide
accounts
for
approximately
30­
40%
of
the
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
15
cyanogen
chloride
dose
at
high
doses,
and
at
least
60­
80%
at
lower
doses
(
Aldridge
and
Evans,

1946;
Midwest
Research
Institute,
1997).

The
data
on
the
toxicity,
kinetics,
and
dynamics
of
cyanide
are
sufficient
to
support
the
development
of
an
RfD
from
the
available
experimental
animal
data.
Standard
assumptions
were
used
in
their
development,
and
include:

1.
the
use
of
experimental
animal
data
as
a
surrogate
for
humans;

2.
the
use
of
reproductive
effects
in
rats
as
meaningful
for
extrapolating
to
human
disease;

3.
the
use
of
factors
based
on
a
logarithmic
scale
(
10,
3
or
1)
that
address
additional
scientific
uncertainties
in
the
overall
database;
and
4.
the
use
of
only
1
digit
of
arithmetic
precision
for
the
RfD,
because
our
understanding
of
the
underlying
biology
is
unlikely
to
be
more
precise
than
this.

The
use
of
these
and
similar
assumptions
is
common
practice
in
conducting
dose­
response
assessments
by
other
environmental
and
health
agencies
throughout
the
world.

The
Health
Advisories
for
cyanogen
chloride
were
also
based
on
extrapolation
from
cyanide.
In
the
absence
of
suitable
studies
on
cyanogen
chloride
or
a
surrogate
for
a
one­
day
HA,

the
Ten­
day
HA
for
cyanide
is
used
as
a
default.
The
Ten­
day
HA
for
cyanogen
chloride
was
0.3
mg/
L,
based
on
the
molecular­
weight
extrapolation
from
cyanide.
The
Ten­
day
HA
for
cyanide
was
0.1
mg/
L,
based
on
increased
liver
weight
in
rats
provided
cyanide
in
drinking
water
for
21
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
16
days
(
Palmer
and
Olson,
1979),
and
using
the
default
uncertainty
factors
of
10
for
extrapolation
from
animals
to
humans,
for
human
variability,
and
for
extrapolation
from
a
LOAEL
The
longer­
term
HA
for
cyanide
is
based
on
the
same
study
and
endpoint
as
the
cyanide
RfD.
For
the
cyanogen
chloride
values
based
on
the
cyanide
NOAEL,
the
longer­
term
HA
for
the
child
was
0.1
mg/
L,
and
the
longer­
term
HA
for
the
adult
was
0.4
mg/
L.
These
values
were
based
on
the
longer­
term
cyanide
HA
values
of
0.05
and
0.2
mg/
L
for
the
child
and
adult,

respectively.
For
the
cyanogen
chloride
values
based
on
the
cyanide
BMDL,
the
longer­
term
HA
for
the
child
was
0.02
mg/
L,
and
the
longer­
term
HA
for
the
adult
was
0.07
mg/
L.
These
values
were
based
on
the
longer­
term
cyanide
HA
values
of
0.008
and
0.03
mg/
L
for
the
child
and
adult,

respectively.
The
critical
effect
was
male
reproductive
toxicity.
A
composite
uncertainty
factor
of
1000
was
used,
based
on
full
factors
of
10
each
for
extrapolation
from
animals
to
humans,

human
variability,
and
database
deficiencies.
Except
for
the
issue
of
progression
of
effects
with
increasing
exposure
duration,
the
uncertainties
discussed
above
in
the
context
of
the
RfD
apply.

The
potential
for
neurodevelopmental
effects
is
of
particular
concern.
Using
an
RSC
of
20%,
a
Lifetime
HA
of
0.08
mg/
L
was
calculated
for
cyanogen
chloride,
based
on
the
NOAEL­
based
cyanide
Lifetime
HA
of
0.03
mg/
L.
Using
the
same
RSC
of
20%,
an
alternative
Lifetime
HA
of
0.01
mg/
L
was
calculated
for
cyanogen
chloride,
based
on
the
BMDL­
based
cyanide
Lifetime
HA
of
0.006
mg/
L.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
17
C.
Weight
of
Evidence
and
Implications
in
Risk
Conclusions
There
are
no
cancer
bioassays
and
no
data
on
the
potential
genotoxicity
of
cyanogen
chloride,
although
a
QSTR
analysis
predicted
that
cyanogen
chloride
was
negative
for
carcinogenicity
in
rats
and
mice
(
Moudgal
et
al.,
2000).
Therefore,
surrogates
were
used
to
evaluate
the
carcinogenic
potential
of
cyanogen
chloride.
There
are
no
cancer
bioassays
of
cyanide
and
no
well­
conducted
standard
cancer
bioassays
of
thiocyanate.
Two
oral
carcinogenicity
studies
in
rats
were
conducted
in
the
same
laboratory
(
Lijinsky
and
Reuber,
1982;

Lijinsky
and
Kovatch,
1989).
The
only
effect
in
the
first
study
was
an
increase
in
liver
tumors,

and
this
was
not
confirmed
at
the
higher
dose
tested
in
the
second
study,
although
a
nonstatistically
significant
increase
in
thyroid
tumors
was
observed.
However,
these
studies
suffer
from
a
number
of
limitations.
Only
a
single
dose
was
tested
in
each
study,
only
20/
group
were
tested,
and
the
assay
continued
until
all
of
the
animals
died
(
increasing
the
background
incidence
of
age­
related
tumors).

The
overall
data
on
the
genotoxicity
of
cyanide,
thiocyanate,
and
the
putative
metabolites
of
cyanogen
chloride
suggest
that
none
of
the
metabolites
are
genotoxic,
although
the
data
are
limited.
Cyanide
has
generally
been
negative
in
bacterial
mutagenicity
assays
(
De
Flora,
1981,
De
Flora
et
al.,
1984;
NTP,
1993)
and
assays
of
DNA
damage
and
repair
(
De
Flora
et
al.,
1984;

Painter
and
Howard,
1982),
although
a
positive
result
was
obtained
in
one
strain
of
Salmonella
with
and
without
S9
activation
(
Kushi
et
al.,
1983).
No
assays
of
mammalian
gene
mutation
or
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
18
chromosome
aberration
are
available.
Genotoxicity
data
on
thiocyanate
are
limited
to
marginal
results
in
a
S.
typhimurium
mutagenicity
assay
(
Kier,
1988,
as
reported
by
Rosenkranz
and
Klopman,
1990).
Potassium
thiocyanate
does
not
have
any
structural
alerts
for
genotoxicity
(
Rosenkranz
and
Klopman,
1990).
The
limited
genotoxicity
results
available
on
the
putative
metabolites
cyanate
and
cyanamide
were
also
generally
limited.

Based
on
these
considerations,
cyanogen
chloride,
cyanide,
and
thiocyanate
are
classified
as
Group
D,
Not
Classifiable
as
to
Human
Carcinogenicity,
using
the
U.
S.
EPA
(
1986)
guidelines.

Using
the
U.
S.
EPA
(
1999)
Draft
Guidelines
for
Cancer
Risk
Assessment,
the
data
are
inadequate
for
an
assessment
of
the
human
carcinogenic
potential
of
cyanogen
chloride,
cyanide,

and
thiocyanate.
The
primary
data
gap
is
the
absence
of
standard
carcinogenicity
bioassays
for
cyanogen
chloride,
or
appropriate
surrogates
(
cyanide,
thiocyanate).
There
are
also
gaps
in
the
standard
battery
of
genotoxicity
testing
for
the
surrogates.
For
example,
no
mammalian
gene
mutation
or
chromosome
aberration
assays
are
available
for
cyanide,
and
the
only
study
of
the
standard
battery
available
for
thiocyanate
is
the
bacterial
mutagenicity
assay.

D.
Exposure
Characterization
and
Implication
in
Relative
Source
Contribution
Since
cyanide
is
the
primary
metabolite
of
cyanogen
chloride
and
the
surrogate
used
for
the
dose­
response
assessment,
this
section
includes
exposure
characterization
for
both
cyanogen
chloride
and
cyanide.
This
characterization
could
be
enhanced
by
including
thiocyanate,
since
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
19
cyanide
is
metabolized
to
thiocyanate,
and
thiocyanate
accounts
for
at
least
some
of
the
toxic
effects
of
cyanide
exposure.
Exposure
to
cyanate
and
cyanamide
are
of
less
concern,
since
they
represent
a
small
portion
of
the
cyanogen
chloride
dose,
and
they
are
not
metabolized
through
the
cyanide/
thiocyanate
pathway.

Cyanogen
chloride.
There
are
very
limited
quantitative
data
on
the
presence
of
cyanogen
chloride
in
the
environment
and
on
cyanogen
chloride
exposure.
There
is
no
information
in
the
available
literature
on
the
concentration
of
cyanogen
chloride
in
the
air
and
no
quantitative
data
on
dietary
levels.
One
investigator
found
that
cyanogen
chloride
may
be
produced
from
the
reaction
of
instant
tea
with
water
containing
chloramine
residual.
In
addition,
there
is
no
information
in
the
available
literature
on
dermal
exposure
to
cyanogen
chloride
or
on
cyanogen
chloride
body
burden.
The
National
Occupational
Exposure
Survey
(
NOES)
conducted
by
the
National
Institute
for
Occupational
Safety
and
Health
from
1980
­
1983
estimated
that
1393
workers
were
exposed
to
HCN.
However,
no
information
on
exposure
levels
was
provided.
The
only
quantitative
data
available
are
drinking­
water
concentrations.

Based
on
information
in
the
ICR
database
(
U.
S.
EPA,
2000b),
the
mean
concentrations
of
cyanogen
chloride
in
distribution
system
surface
water
and
groundwater
were
3.02
and
1.63

g/
L,

respectively.
Since
cyanogen
chloride
is
formed
primarily
from
the
reaction
of
chlorine
with
organic
material
in
the
presence
of
ammonia,
only
drinking
water
plants
that
used
chloramine
as
a
primary
or
secondary
disinfectant
were
required
to
monitor
for
cyanogen
chloride.
The
mean
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
20
concentration
of
cyanogen
chloride
in
surface
water
was
statistically
greater
(
at
p
=
0.05)
than
the
mean
cyanogen
chloride
concentration
in
groundwater.
Therefore,
a
mean
concentration
of
3.02

g/
L
was
selected
for
later
calculations,
as
a
conservative
estimate
of
the
exposure
for
people
served
by
water­
treatment
plants
that
use
chloramine.
The
ICR
database
reported
that
35%
of
water
treatment
plants
using
surface­
water,
and
23%
of
the
plants
using
groundwater
used
chloramine
as
a
primary
or
secondary
disinfectant
and
reported
cyanogen
chloride
observations.

Therefore,
if
the
type
of
plants
sampled
in
the
ICR
survey
are
representative
of
those
throughout
the
U.
S.,
and
if
the
plants
using
chloramine
are
similar
in
size
and
numbers
of
population
served
to
plants
using
different
treatment
methods,
it
could
be
argued
that
23%
to
35%
of
the
population
would
be
exposed
to
cyanogen
chloride
drinking­
water
concentrations
of
approximately
3.02

g/
L.
However,
according
to
the
WHO
Environmental
Health
Criteria
Monograph
on
Disinfectants
and
Disinfectant
Byproducts
(
WHO,
2000),
cyanogen
chloride
may
be
formed
during
disinfection
with
other
disinfectants
when
ammonia
is
present
in
the
source
waters.

Therefore
the
percentage
of
plants
for
which
the
treated
drinking
water
contains
cyanogen
chloride
may
be
under­
represented
in
the
ICR
database.
In
addition,
the
ICR
database
includes
only
those
very
large
surface­
and
ground­
water
systems
serving
at
least
100,000
persons.
The
smaller
systems,
serving
less
than
100,000
persons,
are
not
represented.
However,
examination
of
the
NCOD
cyanide
data
using
the
Student's
t­
test
indicates
that
there
was
no
statistically
significant
difference
between
the
cyanide
concentrations
in
the
different­
size
systems.
Therefore,

if
cyanogen
chloride
concentrations
follow
a
similar
pattern,
the
size
of
the
system
may
not
make
much
difference.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
21
There
is,
however,
considerable
variability
in
the
cyanogen
chloride
concentrations
to
which
an
individual
may
be
exposed.
As
shown
in
Chapter
4,
the
10th
percentile
for
cyanogen
chloride
concentrations
in
distribution
system
surface
water
was
0

g/
L,
while
the
90th
percentile
of
cyanogen
chloride
concentrations
was
8.0

g/
L.

At
a
mean
concentration
of
3.02

g/
L
cyanogen
chloride,
a
70­
kg
adult
ingesting
2
L/
day
drinking
water
would
consume
0.09

g/
kg­
day
([
3.02

g/
L
*
2
L/
day]/
70
kg)
cyanogen
chloride.

For
a
10­
kg
child
ingesting
1
L/
day,
a
concentration
of
3.02

g/
L
cyanogen
chloride
would
result
in
a
daily
dose
of
0.3

g/
kg­
day
([
3.02

g/
L
*
1
L/
day]/
10
kg).
An
adult
and
child
ingesting
8.0

g/
L
(
the
90th
percentile
concentration)
would
receive
a
dose
of
0.23
and
0.8

g/
kg­
day,

respectively.

The
relative
source
contribution
(
RSC)
for
cyanogen
chloride
is
derived
by
application
of
the
Exposure
Decision
Tree
approach
published
in
EPA's
Methodology
for
Deriving
Ambient
Water
Quality
Criteria
for
the
Protection
of
Human
Health
(
U.
S.
EPA,
2000d).
An
RSC
of
20%

accounts
for
the
likelihood
of
exposure
to
cyanogen
chloride
or
cyanide
from
sources
other
than
tap
water,
such
as
ambient
air
and
food,
in
the
absence
of
adequate
data.
This
value
also
takes
into
account
the
potential
for
exposure
to
cyanide
(
a
key
cyanogen
chloride
metabolite)
via
food,

air,
and
smoking,
as
described
in
the
next
paragraphs.
The
data
are
not
adequate
enough
to
quantify
the
contributions
of
each
source
for
an
overall
assessment
of
exposure.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
22
Cyanide.
Compared
to
cyanogen
chloride,
there
are
more
data
on
the
presence
of
cyanide
in
the
environment
and
cyanide
exposure.

The
ICR
database
does
not
include
information
on
cyanide
levels.
However,
information
on
cyanide
levels
in
drinking
water
for
the
year
2000
were
available
from
the
NCOD
survey.
As
discussed
in
Chapter
4,
average
cyanide
concentrations
in
treated
surface
water
and
groundwater
of
systems
that
detected
cyanide
were
reported
as
2844

g/
L
and
2194

g/
L,
respectively;
these
numbers
are
not
statistically
significantly
different.
However,
cyanide
was
detected
infrequently,

in
7%
of
the
plants
using
surface
water
as
a
source,
and
in
3%
of
the
plants
using
groundwater
as
a
source.
In
addition,
in
most
water
system
size
groupings,
there
were
several
orders
of
magnitude
separating
the
minimum
and
maximum
detected
levels,
and
only
a
few
analyses
with
detects
(<
3%
samples
in
surface
water,
<
2%
samples
in
groundwater).
This
means
that
the
calculated
averages
for
these
populations
may
be
highly
influenced
by
one
or
two
systems
with
high
cyanide
levels,
and
that
the
calculated
averages
do
not
accurately
reflect
the
cyanide
concentrations
to
which
these
populations
are
exposed.
Medians
were
not
available
from
the
NCOD
survey.
Despite
these
caveats,
a
conservative
mean
concentration
of
2844

g/
L
was
selected
as
representative
of
the
concentration
of
cyanide
in
drinking
water,
in
the
absence
of
more
representative
data.
This
is
because,
if
the
type
of
plants
sampled
in
the
NCOD
survey
are
representative
of
those
throughout
the
U.
S.,
and
if
the
plants
detecting
cyanide
are
similar
to
plants
not
detecting
the
presence
of
cyanide
in
the
size
and
numbers
of
people
served,
it
could
be
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
23
argued
that
approximately
3%
to
7%
of
the
population
would
be
exposed
to
cyanide
concentrations
of
2844

g/
L
(
on
average)
in
drinking
water.

At
a
conservative
average
concentration
of
2844

g/
L
HCN,
a
70­
kg
adult
ingesting
2
L/
day
drinking
water
would
consume
81

g/
kg­
day
([
2844

g/
L
*
2
L/
day]/
70
kg)
HCN.

Although
the
10th
and
90th
percentiles
were
not
provided
in
the
NCOD
database,
there
is
considerable
variability
in
the
cyanide
concentrations,
ranging
from
a
minimum
value
of
0

g/
L
to
a
maximum
value
of
32,700

g/
L.
The
number
of
samples
at
the
high
end
is,
however,
unknown.

Ambient­
air
concentrations
of
HCN
in
the
northern
hemisphere's
non­
urban
troposphere
ranged
from
160
to
166
ppt
(
177
ng/
m3
to
184
ng/
m3).
Although
ambient­
air
monitoring
data
of
HCN
near
source
areas
(
e.
g.,
HCN­
manufacturing
industries,
coke­
production
industries,

wastedisposal
sites)
could
not
be
located
in
the
available
literature,
HCN
concentrations
in
these
areas
are
expected
to
be
higher
than
the
non­
urban
tropospheric
concentrations.
Based
on
an
atmospheric
concentration
of
170
ppt
(
188
ng/
m3)
and
a
daily
average
inhalation
rate
of
20
m3,

the
ATSDR
(
1997)
Toxicological
Profile
for
Cyanide
estimated
an
inhalation
exposure
to
the
general
U.
S.
non­
urban,
nonsmoking
population
of
3.8

g
HCN/
day.
This
would
be
equivalent
to
a
daily
dose
to
a
70­
kg
adult
of
approximately
54
ng/
kg­
day
HCN
([
3.8

g
HCN/
day]/
70
kg).

Children,
with
a
higher
inhalation­
rate
to
body­
weight
ratio,
would
receive
a
higher
dose
on
a
body­
weight
basis.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
24
Smokers
and
those
exposed
to
secondary
tobacco
smoke
make
up
a
subset
of
the
general
population
that
may
be
exposed
to
elevated
levels
of
HCN.
Smokers
could
be
exposed
to
10
to
400

g
HCN
per
cigarette,
whereas
nonsmokers
exposed
to
sidestream
smoke
could
be
exposed
to
0.06
to
108

g
HCN/
cigarette.
Depending
on
the
amount
smoked
per
day,
the
HCN
exposure
from
cigarette
smoke
could
be
substantial.

Based
on
information
presented
in
the
ATSDR
(
1997)
Toxicological
Profile
for
Cyanide,

between
1981
to
1983,
4005
workers
were
potentially
exposed
to
HCN.
NIOSH
found
HCN
in
the
workplace
air
at
concentrations
ranging
from
0.001
to
4.3
mg/
m3.
Although
the
ATSDR
document
provided
no
information
on
the
amount
of
time
and
frequency
that
these
workers
were
exposed
to
these
concentrations
of
HCN,
these
levels
were
below
the
NIOSH
recommended
15­

min
short­
term
exposure
limit
(
STEL)
of
5
mg/
m3
for
HCN.

HCN
is
also
a
metabolite
of
a
number
of
industrial
chemicals
(
acetonitrile,
propionitrile,

acrylonitrile,
n­
butyronitrile,
maleonitrile,
and
succinonitrile).
As
a
result,
occupational
or
environmental
exposure
to
these
chemicals
could
contribute
to
the
background
levels
of
HCN
in
biological
fluids.
HCN
is
also
a
metabolite
of
pharmaceuticals
such
as
Laetrile
and
a
drug
used
to
reduce
high
blood
pressure,
and
clinical
use
of
these
compounds
could
increase
internal
exposure
to
HCN.
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
25
Although
concentrations
of
HCN
in
foods
are
expected
to
be
low,
one
author
(
Fiksel
et
al.,
1981)
estimated
that
HCN
intake
from
inhalation
of
air
and
ingestion
of
drinking
water
would
be
less
than
the
intake
from
food.
The
ATSDR
document
(
ATSDR,
1997)
did
not
provide
information
on
the
cyanide
concentrations
in
air
and
water
that
Fiksel
et
al.
(
1981)
used
to
come
to
this
conclusion.
In
addition,
estimates
of
the
HCN
concentration
in
the
total
diet
were
not
located
in
the
available
literature.
Therefore,
no
independent
estimate
of
daily
HCN
intake
from
food
could
be
made,
nor
could
any
comparison
with
HCN
intake
from
other
sources
be
made.

Although
there
is
more
information
on
HCN
than
on
cyanogen
chloride,
the
data
are
not
adequate
to
quantify
the
contributions
of
each
source
for
an
overall
assessment
of
exposure.
The
data
do
indicate,
however,
that
there
is
potential
for
exposure
to
cyanide
via
food,
air,
and
smoking.
Using
the
Exposure
Decision
Tree
approach
published
in
EPA's
Methodology
for
Deriving
Ambient
Water
Quality
Criteria
for
the
Protection
of
Human
Health
(
U.
S.
EPA,

2000d),
an
RSC
of
20%
is
used
to
account
for
the
likelihood
of
exposure
to
cyanide
from
sources
other
than
tap
water.

E.
Risk
Characterization
for
Drinking
Water
Exposure
Segments
of
the
population
with
an
elevated
exposure
to
cyanogen
chloride
have
not
been
identified.
However,
there
are
a
number
of
groups
that
would
have
elevated
exposure
to
cyanide
from
sources
other
than
drinking
water.
A
major
source
of
potential
cyanide
exposure
is
in
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
26
cigarette
smoke
and
through
passive
smoking.
Smokers
may
be
exposed
to
10
to
400

g
HCN
per
cigarette,
and
passive
exposure
of
nonsmokers
to
sidestream
smoke
ranges
from
0.06
to
108

g
HCN/
cigarette.
The
actual
dose
would
depend
on
the
number
of
cigarettes
smoked
per
day.

The
dose
from
sidestream
smoke
could
be
higher
for
children
than
adults
on
a
body­
weight
basis,

because
of
the
child's
higher
inhalation­
rate
to
body­
weight
ratio.
Occupational
exposure
to
nitriles
(
e.
g.,
acetonitrile,
propionitrile,
acrylonitrile)
would
also
increase
the
internal
dose
of
cyanide,
since
these
compounds
are
metabolized
to
cyanide.
The
cyanide
dose
resulting
from
inhalation
exposure
of
the
general
U.
S.
non­
urban,
nonsmoking
population
is
approximately
54
ng/
kg­
day
HCN,
based
on
an
average
atmospheric
concentration
of
170
ppt
(
188
ng/
m3)
and
a
daily
average
inhalation
rate
of
20
m3.
As
noted
above
for
passive
smoking,
the
dose
to
children
on
a
body
weight
basis
would
be
higher.
People
living
or
working
near
source
areas
(
e.
g.,

HCNmanufacturing
industries,
coke­
production
industries,
waste­
disposal
sites)
would
be
exposed
to
higher
atmospheric
concentrations
of
HCN.
Finally,
some
pharmaceuticals,
such
as
Laetrile
and
a
drug
used
to
reduce
high
blood
pressure,
also
result
in
exposure
to
cyanide.
The
available
information
is
insufficient
to
quantify
the
variability
resulting
from
these
sources
of
higher
exposure.

Although
concentrations
of
cyanide
in
foods
are
expected
to
be
low
and
quantitative
data
on
the
cyanide
concentration
in
foods
is
sparse,
the
cyanide
intake
from
food
ingestion
is
expected
to
exceed
HCN
intake
from
inhalation
of
air
and
ingestion
of
drinking
water.
People
eating
large
amounts
of
food
containing
large
amounts
of
cyanogenic
glycosides
would
have
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
27
elevated
exposure
to
cyanide,
particularly
if
these
foods
are
not
prepared
in
a
way
that
removes
the
cyanogenic
compounds.
In
particular,
people
eating
large
amounts
of
cassava
root,
such
as
immigrants
from
Africa
or
some
South
American
countries,
may
have
elevated
exposure.
People
eating
large
amounts
of
canned
fruits
with
pits
could
also
have
elevated
exposure.

Cyanogen
chloride
was
measured
only
in
systems
using
chloramine
as
a
primary
or
secondary
disinfectant
(
35%
of
the
surface­
water
plants
and
23%
of
the
groundwater
plants).

Even
among
the
plants
using
chloramine,
cyanogen
chloride
was
not
detectable
at
the
distribution
system
maximum
for
22%
of
the
systems
using
surface
water
and
39%
of
the
systems
using
groundwater.
On
the
other
hand,
WHO
(
2000)
and
Richardson
(
1998)
reported
that
ammonial
compounds
in
the
source
water
can
react
with
chlorine
and
form
cyanogen
chloride
in
the
absence
of
chloramination.
Therefore,
the
number
of
plants
where
cyanogen
chloride
may
be
present
may
be
under­
represented
in
the
ICR
database.
Subject
to
this
uncertainty
regarding
the
conditions
under
which
cyanogen
chloride
is
formed,
it
appears
that
approximately
23%
to
35%
of
the
population
would
be
exposed
to
cyanogen
chloride
in
drinking
water.
This
estimate
also
assumes
that
the
population
distribution
in
the
systems
using
chloramine
is
representative
of
all
public
drinking­
water
systems
in
the
U.
S.
A
conservative
average
concentration
of
3.02

g/
L
can
be
obtained
from
the
ICR.
This
concentration
is
approximately
two
orders­
of­
magnitude
lower
than
the
10­
day
cyanogen
chloride
HA
of
0.3
mg/
L
(
300

g/
L).
Using
the
health
advisories
based
on
the
NOAEL
for
comparison,
the
average
cyanogen
chloride
concentration
of
3.02

g/
L
is
at
least
a
factor
of
30
lower
than
the
Longer­
term
child
HA
of
0.1
mg/
L
(
100

g/
L)
and
the
Longer­
term
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
28
adult
HA
of
0.4
mg/
L
(
400

g/
L).
It
is
also
more
than
one
order­
of­
magnitude
lower
than
the
Lifetime
HA
of
0.07
mg/
L
(
70

g/
L).
Using
the
more
conservative
health
advisories
based
on
the
BMDL
for
comparison,
the
average
cyanogen
chloride
concentration
of
3.02

g/
L
is
at
least
a
factor
of
6
lower
than
the
Longer­
term
child
HA
of
0.02
mg/
L
(
20

g/
L
)
and
more
than
an
order­
of­
magnitude
lower
than
the
Longer­
term
adult
HA
of
0.07
mg/
L
(
70

g/
L).
It
is
also
approximately
a
factor
of
3
lower
than
the
Lifetime
HA
of
0.01
mg/
L
(
10

g/
L).
Thus,
even
using
the
more
conservative
approach
of
basing
the
health
advisories
on
the
cyanide
BMDL,

average
drinking
water
concentrations
are
below
the
relevant
health
advisories.

Cyanide
was
detected
in
7%
of
the
public
water
supplies
using
surface
water
as
a
source,

and
in
only
3%
of
the
public
water
supplies
using
groundwater
as
a
source.
Therefore,
exposure
to
cyanide
in
drinking
water
for
most
of
the
population
is
minimal.
However,
at
least
some
of
this
small
percentage
of
the
population
has
exposures
well
above
advisory
levels.
For
example,
of
the
systems
with
detectable
levels
of
cyanide,
the
overall
average
concentration
of
cyanide
in
surfacewater
systems
of
all
sizes
was
2844

g/
L
(
2.8
mg/
L).
However,
as
discussed
above,
this
average
may
represent
only
a
few
water
systems,
in
light
of
the
large
difference
between
the
minimum
and
maximum
detected
levels
and
the
few
analyses
with
detects.
Additional
information
on
typical
exposure
levels
could
be
provided
by
consideration
of
median
concentrations,
but
medians
were
not
available
from
the
NCOD
survey.
In
the
absence
of
more
representative
information
about
typical
exposure
levels,
the
average
concentration
for
this
small
percentage
of
the
population
was
compared
to
the
cyanide
health
advisories.
The
average
of
2844

g/
L
is
more
than
an
order­
of­
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
29
magnitude
above
the
Ten­
day
HA
of
0.1
mg/
L.
Using
the
health
advisories
based
on
the
NOAEL,

this
concentration
is
also
more
than
an
order­
of­
magnitude
above
the
Longer­
term
child
HA
of
0.05
mg/
L,
almost
two
orders­
of­
magnitude
above
the
Lifetime
HA
of
0.03
mg/
L,
and
approximately
five
times
the
Longer­
term
adult
HA
of
0.5
mg/
L.
Using
the
health
advisories
based
on
the
BMDL,
the
drinking
water
concentration
is
approximately
two
orders­
of­
magnitude
above
the
corresponding
health
advisories.
These
comparisons
suggest
that
populations
with
elevated
cyanide
levels
in
their
drinking
water
may
be
at
risk.
Based
on
the
wide
variability
in
exposure
and
the
large
numbers
of
nondetects,
it
appears
that
relatively
few
systems
would
have
such
elevated
levels,
but
no
definitive
statement
regarding
the
size
of
the
population
at
risk
is
possible
in
the
absence
of
NCOD
data
on
median
concentrations.
In
addition,
it
should
be
noted
that
the
RfD
is
an
estimate
"
within
an
order
of
magnitude....
that
is
likely
to
be
without
an
appreciable
risk
of
deleterious
effects
during
a
lifetime."
Thus,
a
small
subset
of
the
population
that
lives
at
the
same
residence
(
served
by
the
same,
unchanged
water
treatment
facility)
for
their
entire
lifetime
(~
70
years),
may
be
at
risk.
However,
the
average
number
of
years
that
a
person
in
the
United
States
remains
in
their
current
residence
was
determined
to
be
9
years
(
50th
percentile
value).
The
upper
90th
percentile
estimate
of
residence
time
was
33
years
(
U.
S.
EPA,
1997b).

Less
than
one
percent
of
the
population
in
the
United
States
lives
at
the
same
residence
for
more
than
55
years
(
U.
S.
EPA,
1997b).

As
discussed
in
greater
detail
in
Chapter
7,
males
may
be
a
sensitive
population
for
the
effects
of
cyanide
and
cyanogen
chloride,
because
the
critical
effect
occurs
in
male
reproductive
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
30
organs.
However,
it
is
unclear
if
this
is
the
true
critical
effect
for
cyanide
exposure,
in
the
absence
of
measures
of
thyroid
hormone
levels,
and
in
the
absence
of
neurodevelopmental­
toxicity
testing.
Children
are
also
a
potential
target
of
concern,
due
to
the
potential
for
neurodevelopmental
effects
resulting
from
the
effects
on
thyroid
hormones
of
the
cyanide
metabolite
thiocyanate.

F.
Comparison
With
Other
Standards
The
World
Health
Organization
(
WHO)
does
not
have
a
drinking­
water
guideline
for
cyanogen
chloride,
but
WHO
did
establish
a
drinking­
water
guideline
of
0.07
mg/
L
for
cyanide
as
total
cyanogenic
compounds
(
WHO,
1996).
This
value
was
based
on
a
LOAEL
of
1.2
mg/
kg­
day
for
effects
on
behavioral
patterns
and
serum
biochemistry
in
a
drinking
water
study
in
pigs
(
Jackson,
1988).
Uncertainty
factors
of
10
each
were
used
for
inter­
and
intraspecies
variation;

no
additional
factor
for
a
LOAEL
was
considered
necessary
because
of
doubts
regarding
the
biological
significance
of
the
observed
changes.
A
guideline
value
of
0.07
mg/
L
is
calculated
from
this
TDI
(
similar
to
an
RfD)
using
a
20%
allocation
of
the
total
daily
intake
to
drinking
water,
a
60
kg
body
weight,
and
daily
intake
of
2
L/
day.
The
NTP
(
1993)
study
was
apparently
not
available
for
the
WHO
assessment,
which
was
largely
completed
by
1993.

Health
Canada
has
established
a
maximum
allowable
concentration
of
cyanide
in
drinking
water
of
0.2
mg/
L
(
Health
Canada,
2001),
based
on
the
study
by
Howard
and
Hanzal
(
1955).
Drinking
Water
Criteria
Document
for
Cyanogen
Chloride
and
Potential
Metabolites
EPA/
OW/
OST/
HECD
Final
draft
IX­
31
The
documentation
for
this
value
was
developed
in
1991,
prior
to
the
publication
of
the
NTP
(
1993)
subchronic
study
that
was
the
basis
for
the
cyanide
RfD
presented
here.
Health
Canada
does
not
have
a
drinking­
water
standard
for
cyanogen
chloride.
