Comfrey: assessing the low-dose health risk


Abbott, P.J.

Medical Journal of Australia 149(11-12): 678-682

1988


The regular use of comfrey as part of the diet or for medicinal purposes may be a potential health risk as a result of the presence of naturally occurring pyrrolizidine alkaloids. The majority of these alkaloids are hepatotoxic in both animals and humans, and some have been shown to induce tumours in experimental animals. In this article, the toxic properties of pyrrolizidine alkaloids are reviewed briefly, with particular reference to their presence in comfrey. The acute and long-term health risks at the normally low levels of comfrey consumption are evaluated and discussed. On the basis of the data that are available currently, the small but significant longterm risk that is associated with the consumption of comfrey justifies the need to limit its intake. This is being achieved by controls under various state Poisons Acts, but also requires further education on the potential dangers of naturally-occurring chemicals of plant origin.

678
Therapeutics
December
5/19,
1988
Vol.
149
THE
MEDICAL
JOURNAL
OF
AUSTRALIA
Comfrey:
assessing
the
low-dose
health
risk*
(for
editorial
comment,
see
page
572)
ABSTRACT
The
regular
use
of
comfrey
as
part
of
the
diet
or
for
medicinal
purposes
may
be
a
potential
health
risk
as
a
result
of
the
presence
of
naturally-
occurring
pyrrolizidine
alkaloids.
The
majority
of
these
alkaloids
are
hepato-
toxic
in
both
animals
and
humans,
and
some
have
been
shown
to
induce
tumours
in
experimental
animals.
In
this
article,
the
toxic
properties
of
pyrrolizidine
alkaloids
are
reviewed
briefly,
with
particular
reference
to
their
presence
in
comfrey.
The
acute
and
long-term
health
risks
at
the
normally-
low
levels
of
comfrey
consumption
are
evaluated
and
discussed.
On
the
basis
of
the
data
that
are
available
currently,
the
small
but
significant
long-
term
risk
that
is
associated
with
the
consumption
of
comfrey
justifies
the
need
to
limit
its
intake.
This
is
being
achieved
by
controls
under
various
state
Poisons
Acts,
but
also
requires
further
education
on
the
potential
dangers
of
naturally-occurring
chemicals
of
plant
origin.
(Med
J
Aust
1988;
149:
678-682)
Peter
J.
Abbott
Me
me
Me
Me
COO
OCO
Me
OH
OH
Symphytine
Me
/
Me
COO
Me
Me
OH
OCO
Me
OH
OH
C
omfrey
(Symphyturn)
species
are
among
a
large
number
of
plant
species
that
contain
pyrrolizidine
alkaloids'
and
represent
a
major
source
of
these
alkaloids
in
the
Australian
diet."
Pyrrolizidine
alkaloids
have
been
shown
to
be
hepatotoxic
and,
in
some
cases,
carcinogenic
in
animals.'
-
'
They
also
are
associated
with
human
liver
disease
in
various
countries
as
a
result
of
the
contamination
of
staple
food
crops
or
by
the
deliberate
inges-
tion
of
herbal
medicines.""
The
variety
and
variable
amount
of
pyrrolizidine
alkaloids
that
are
present
in
any
particular
plant
species,
together
with
the
difficulty
of
their
chemical
identification,
has
made
the
establishment
of
a
causal
association
with
toxic
effects
a
complex
task.
However,
because
many
pyrrolizidine
alkaloids
are
metabolized
in
the
body
by
a
common
pathway,
it
is
possible
to
generalize
about
the
likely
properties
of
the
class
as
a
whole.
Thus,
while
this
article
focuses
on
comfrey,
the
toxicity
data
of
pyrrolizidine
alkaloids
from
other
sources
have
been
considered
in
assessing
any
potential
hazard.
An
extensive,
although
still
incom-
plete,
body
of
toxicity
data
has
accumulated
slowly
over
many
years,
and
has
improved
greatly
our
understanding
of
pyrrolizidine
alkaloid-induced
toxicity
and,
by
extrapolation,
the
effects
of
comfrey
on
human
health.
After
consideration
of
the
toxicological
data
on
comfrey,
and
pyrrolizidine
alkaloids
in
general,
in
relation
to
their
effects
on
human
health,
the
National
Health
and
Medical
Research
Council
recommended
in
1984
that
comfrey
be
placed
in
Schedule
1
of
the
Standard
for
Uniform
Scheduling
of
Drugs
and
Poisons;
thus,
it
recommended
that
comfrey
be
available
only
from
pharmacists,
medical
practitioners
or,
in
isolated
communities,
other
licensed
individuals.
Criticism
of
the
decision
was
vigorous
and
generally
fell
into
one
of
four
categories:
disbelief
that
a
natural
herb could
be
anything
but
safe,
in
view
of
its
long
history
of
use;
attempts
to
discredit
the
scientific
evidence
on
the
toxicity
of
pyrrolizidine
alkaloids
in
general;
attempts
to
show
that
corn
frey
(particularly
Symphytum
officinale)
was
less
toxic
than
were
other
pyrrolizidine
alkaloid-containing
plants;
and
concerns
that
the
professional
integrity
of
herbalists
was
in
question.
This
controversy
serves
to
highlight
the
difficulty
of
arriving
at
a
consensus
of
opinion
about
the
assessment
of
risk
to
human
health,
as
well
as
to
question
the
role
of
governments
in
regulating
naturally-occurring
toxic
chemicals.
The
toxicology
of
pyrrolizidine
alkaloids
has
been
the
subject
of
a
number
of
excellent
reviews"'
and
only
a
brief
résumé
will
be
presented
here.
The
purpose
of
this
article
is
to
discuss
the
avail-
able
toxicological
data
on
comfrey
and
the
potential
of
low-dose
exposure
to
cause
health
effects
in
humans.
'The
views
that
are
expressed
in
this
paper
do
not
necessarily
reflect
the
views
of
the
National
Health
and
Medical
Research
Council.
Toxicology
Unit,
Department
of
Community
Services
and
Health,
PO
Box
9848,
Canberra,
ACT
2601.
Peter
I.
Abbott,
PhD,
Senior
Toxicologist.
Reprints:
Dr
P.J.
Abbott.
Echimidine
Me
Me
AcO
OCO
Me
OH
OH
7-Acetyllycopsamine
Me
Me
N,../
OH
HO
OCO
OH
Intermidine
Me
Me
Me
OH
Me
COO
OCO
Me
Me
FIGURE
1:
The
major
pyrrolizidine
alkaloids
that
are
Lasiocarpine
found
in
corn
frey.
Human
exposure
Symphytum
officinale
(common
comfrey)
is
the
traditional
Euro-
pean
dietary
and
medicinal
herb
and
is
available
as
such
in
Australia.
An
additional
and
probably
a
greater
exposure
to
pyrrolizidine
alkaloids
in
Australia
has
been
through
the
consumption
of
Russian
comfrey
(Symphytum
X
uplandicum
Nyman)
either
as
a
salad
item
(leaves)
or
as
a
tea
(infused
powdered
root).
The
fresh
leaf
contains
0.01%-0.15%
of
alkaloids,
the
major
ones
being
echimidine
and
7-acetyllycopsamine
(24%
and
32%
of
total
alkaloid,
respectively;
Figure
1).'"
A
consumption
rate
of
up
to
five
leaves
per
day
would
represent
an
intake
of
approximately
5
mg
of
alkaloid
per
day
(approximately
70
µg/kg
a
day)."
The
intake
from
root-tea
is
variable
but
has
been
estimated
at
between
8
and
26
mg
per
cup.
2
°
Herbal
preparations
also
have
been
available
as
S.
officinale
extracts,
tinctures
or
tablets.
Creams
that
contain
S.
officinale
extract
are
not
considered
to
be
a
significant
source
of
the
intake
of
comfrey
in
view
of
the
low
dermal
absorp-
tion
of
the
pyrrolizidine
alkaloids.
2
'
While
comfrey
is
the
most
widely-recognized
source
of
dietary
pyrrolizidine
alkaloids
in
developed
countries,
other
herbal
prepa-
rations
that
contain
pyrrolizidine
alkaloids
also
have
been
implicated
in
liver
disease
in
humans,"
-10
although
little
information
is
avail-
able
on
their
alkaloid
content.
The
most-serious
poisoning
incidents
to
involve
pyrrolizidine
alkaloids
have
occurred
in
developing
countries.
Major
outbreaks
M
e
THE
MEDICAL
JOURNAL
OF
AUSTRALIA
Vol.
149
December
5/19,
1988
Therapeutics
679
of
poisoning
have
occurred
in
the
West
Indies,"
in
Afghanistan"
and
in
India"
as
a
result
of
the
contamination
of
cereal
grains
with
seeds
of
pyrrolizidine
alkaloid-containing
plants,
principally
Senecio,
Heliotropium
or
Crotalaria.
In
the
Afghan
outbreak,
which
involved
some
7000
persons,
an
estimated
daily
dose
of
approximately
30
µg/kg
led
to
acute
symptoms
and,
in
many
cases,
death.
Apart
from
these
major
outbreaks
of
poisoning,
there
have
been
large
numbers
of
case
reports
from
many
countries
of
acute
veno-
occlusive
disease
that
has
arisen
from
the
use
of
herbal
preparations
that
were
known
or
were
suspected
to
contain
pyrrolizidine
alkaloids.'
In
most
cases,
the
outcome
of
the
acute
symptoms
was
either
death
or
cirrhosis
of
the
liver.
A
list
of
plants
which
are
known
or
are
suspected
to
contain
pyrrolizidine
alkaloids
and
which
have
been
used
as
herbal
medicines
or
as
food
has
been
compiled
by
Mattocks.'
Two
cases
of
poisoning
(one
in
the
United
States
and
one
in
the
United
Kingdom)
that
specifically
were
associated
with
the
consump-
tion
of
comfrey
have
been
published
recently.
These
cases
reflect
potential
exposure
situations
in
Australia.
In
the
first
case,
a
49-year-
old
woman
with
progressive
swelling
of
the
abdomen
and
extremi-
ties
over
four
months
was
diagnosed
as
having
veno-occlusive
disease
on
the
basis
of
a
liver
biopsy."
The
patient
was
a
heavy
consumer
of
herbs,
vitamins,
and
food
supplements
including
herbal
tea
and
"comfrey-pepsin"
pills,
both
of
which
contained
pyrrolizidine
alkaloids.
Her
consumption
of
pyrrolizidine
alkaloids
was
estimated
to
be
15
µg/kg
a
day
over
a
period
of
several
months,
with
a
minimal
consumption
of
85
mg
of
pyrrolizidine
alkaloids.
The
second
case
involved
a
13-year-old
boy
who
had
hepatomegaly
and
ascites,
who
also
was
diagnosed
as
having
veno-occlusive
disease
on
the
basis
of
a
liver
biopsy."
Over
a
period
of
two
to
three
years,
he
regularly
had
been
administered
a
herbal
tea
that
contained
comfrey
leaf,
although
the
exact
quantity
and
the
frequency
of
consumption
were
unknown.
The
continuing
trend
towards
the
widespread
use
of
herbal
remedies
in
Western
countries
suggests
that
the
potential
for
individual
misuse
and
poisoning
remains
high.
There
has
been
a
documented
case
of
veno-occlusive
disease
in
a
recent
New
Zealand
Coroner's
Report,
which
implicated
the
use
of
comfrey
as
being
associated
directly
with
the
death
of
a
young
New
Zealander.
The
comfrey
allegedly
was
consumed
while
he
was
living
on
a
farming
community
near
Bredbo
in
New
South
Wales.
Since
there
is
some
controversy
surrounding
this
case,
it
will
not
be
discussed
further
here.
The
lack
of
accurate
information
on
the
levels
of
human
exposure
to
pyrrolizidine
alkaloids
both
the
amount
and
the
duration
of
exposure
has
been
a
factor
in
our
inability
to
assess
the
potential
risk.
The
outbreaks
of
poisoning
and
the
case-report
studies
have
provided
some
indication
of
the
dangerous
dosage
levels
for
short-
term
exposure;
however,
they
provide
no
guarantee
of
long-term
safety
in
view
of
the
evidence
from
animal
studies
for
cumulative
effects
and
carcinogenicity.
A
detailed
hazard
evaluation
is
necessary
as
well
as
are
exposure
data
to
ensure
both
long-term
and
short-
term
safety.
Liver
toxicity
Clinical
symptoms
The
most
common
disease
to
be
associated
with
the
consumption
of
pyrrolizidine
alkaloids
in
humans
is
veno-occlusive
disease
(a
form
of
Budd-Chiari
syndrome),
which
is
characterized
by
the
occlusion
of
the
small
branches
of
the
hepatic
vein
in
the
liver.
The
clinical
symptoms
include
a
dull
ache
in
the
upper
half
of
the
abdomen,
hepatomegaly,
ascites
and
often
pedal
oedema
and
a
fall
in
urinary
output.
Nausea
and
vomiting
may
be
present,
while
jaundice
and
fever
are
rare.
The
disease
usually
begins
with
an
acute
phase
which,
in
severe
cases,
can
lead
to
death
as
a
result
of
hepatocellular
failure.
In
less-
severe
cases,
the
disease
may
enter
a
subacute
and
eventually
a
chronic
phase.
In
both
acute
and
subacute
phases,
only
slight
changes
in
liver
function
test-results
are
observed.
Children
appear
to
be
particularly
vulnerable,
and
in
a
clinical
study
in
Jamaica,"
children
of
up
to
six
years
of
age
accounted
for
65%
of
the
cases.
After
a
subacute
presentation,
patients
either
may
recover
or
may
enter
the
chronic
phase
which
is
characterized
by
cirrhosis.
While
veno-occlusive
disease
is
the
primary
liver
disease
that
is
induced
by
pyrrolizidine
alkaloids,
progression
to
cirrhosis
has
occurred
in
many
cases.
Pyrrolizidine
alkaloids
are
thought
to
be
one
of
the
major
causes
of
childhood
cirrhosis
in
the
West
Indies,'
and
also
may
be
a
major
factor
in
adult
cirrhosis
in
a
number
of
countries."
The
possibility
that
subclinical
hepatic
injury
eventu-
ally
may
lead
to
adult
cirrhosis
cannot
be
excluded.
The
rarity
of
veno-occlusive
disease
in
Western
countries
may
result
in
its
mis-
diagnosis
and
a
failure
to
recognize
an
association
with
an
intake
of
pyrrolizidine
alkaloids.
Pathological
features
The
pathological
changes
that
occur
in
the
liver
after
the
ingestion
of
pyrrolizidine
alkaloids
have
been
studied
both
in
laboratory
and
in
farm
animals
and
case-study
reports
of
human
poisonings
have
been
reported.
In
animals,
these
changes
have
been
described
exten-
sively
by
Bull
et
al.
4
and
by
McLean."
In
both
humans
and
other
animals,
haemorrhagic
hepatic
necrosis
(either
zonal
or
focal)
is
a
common
feature
of
acute
high-dose
exposure.
4
.
22
The
other
major
toxic
effect
is
that
on
the
vascular
endothelium
of
the
central
and
sublobular
veins.
Further
pathological
features
include
hepatocyte
swelling,
biliary
hyperplasia
and
marked
fibrosis.
After
necrosis
of
the
vascular
endothelial
cells,
the
endothelium
gradually
is
replaced
by
fibrous
connective
tissue.
In
humans
and
in
monkeys,
the
fibrosis
of
the
veins
appears
to
occur
earlier
than
in
other
species,
and
the
cellular
debris
from
the
necrotic
hepato-
cytes,
together
with
the
fibrosis,
causes
venous
occlusion.
Attempts
to
produce
venous
occlusion
in
rats
largely
have
been
unsuccessful.
In
rats,
the
formation
of
giant
abnormal
hepatic
parenchymal
cells
(megalohepatocytes)
is
a
common
feature
of
the
administration
of
pyrrolizidine
alkaloids,"
whereas,
in
humans,
these
cells
have
not
been
observed."
Although
hepatocyte
swelling
is
present
in
humans,
this
is
reversible
and
can
be
distinguished
from
the
giant-cell
forma-
tion
that
is
observed
in
rats
which
can
persist
throughout
the
lifetime
of
the
animal.
The
fact
that
these
giant
cells
have
been
observed
also
in
mice,
sheep,
horses
and
pigs
suggests
that
this
is
a
feature
that
is
common
to
most
species,
although
monkeys
and
humans
appear
to
have
a
limited
resistance
to
the
antimitotic
effect
which
causes
giant-cell
formation.
Metabolism
The
development
of
veno-occlusive
disease
in
humans
after
exposure
to
any
one
of
a
wide
range
of
pyrrolizidine
alkaloids
implies
a
common
mechanism
of
action.
An
analysis
of
their
metabolism
has
provided
a
common
link.
The
pyrrolizidine
alkaloids
themselves
are
not
toxic
and
the
observed
effects
are
mediated
through
their
metabo-
lites.
Data
from
animal
studies
suggest
that
pyrrolizidine
alkaloids
are
absorbed
extensively
from
the
gastrointestinal
tract;
this
is
followed
by
metabolism
in
the
liver
and
rapid
excretion
mainly
in
the
urine.
Deactivation
can
occur
by
either
N-oxidation
or
by
hydrolysis
of
the
ester
bonds.
Activation
by
liver
microsomal
P-450
enzymes
leads
to
the
formation
of
pyrrolic
metabolites
(Figure
2);
the
primary
metabolite
is
a
dehydroalkaloid
(or
pyrrolic
ester)
which
is
metabo-
lized
further
to
a
secondary
metabolite,
a
dehydroamino
alcohol
(or
pyrrolic
alcohol).
Formation
of
these
pyrrolic
metabolites
has
been
shown
to
occur
in
the
liver
cells
of
many
species
as
well
as
in
human
embryo
primary
liver
cultures."
Individual
pyrrolizidine
alkaloids
are
metabolized
at
different
rates
and,
therefore,
exhibit
somewhat
differing
toxicities.
The
major
pyrrolizidine
alkaloids
in
comfrey
symphytine;
the
isomeric
alkaloids,
lycopsamine/intermidine;
7-acetyllycopsamine;
echimidine;
and
lasiocarpine
all
contain
structural
features
which
favour
activation
by
oxidative
dehydrogenation
(Figure
1).
The
cases
of
veno-occlusive
disease
that
have
been
induced
by
comfrey
14
•"
suggest
that
humans,
as
are
other
species,
are
susceptible
to
the
toxic
680
Therapeutics
December
5/19,
1988
Vol.
149
THE
MEDICAL
JOURNAL
OF
AUSTRALIA
RO
CH
2
OCOR'
Esler
hydrolysis
Pyrrolizidine
alkaloid
N-oxides
P450
Excretion
RO
CH
2
OCOR'
Reaction
with
glutathione
Dehydroalkaloid
(Pyrrolic
ester)
Alkylation
of
protein
and
DNA
P450
Carcinogenesis
RO
CH
2
OH
Dehydroaminoalcohol
FIGURE
2:
Activation
and
deactivation
pathways
of
(Pyrrolic
alcohol)
pyrrolizidine
alkaloids.
effects
of
these
particular
pyrrolizidine
alkaloids.
Susceptibility
may
be
enhanced
in
individuals
when
oxidative
metabolism
has
been
induced.
The
primary
metabolites,
the
pyrrolic
esters,
are
highly
reactive
and
undergo
either
hydrolysis
to
the
secondary
metabolite
or
reac-
tion
with
cell
constituents.
The
secondary
metabolites,
the
pyrrolic
alcohols,
are
relatively-stable
and
can
be
distributed
throughout
the
body
or
excreted.
Acute
hepatotoxicity
is
caused
by
the
pyrrolic
ester
which
leads
to
tissue
degeneration.
Chronic
hepatotoxicity,
which
is
characterized
in
experimental
animals
by
the
formation
of
greatly-
enlarged
hepatocytes,
is
the
result
of
the
antimitotic
effects
of
both
the
pyrrolic
alcohol
and
pyrrolic
ester,
together
with
a
stimulus
to
cell
division,
which
is
provided
by
either
regenerative
hyperplasia
in
damaged
tissue
or
normal
hyperplasia
in
the
immature
liver."
While
the
majority
of
pyrrolizidine
alkaloids
have
a
similar
route
of
metabolism,
the
wide
species
variation
in
the
rates
of
metabolism
(which
possibly
is
a
result
of
the
balance
between
activation
and
deac-
tivation
pathways")
makes
prediction
of
the
likely
human
toxicity
difficult.'
However,
given
that
the
major
acute
and
chronic
effects
are
seen
in
the
liver
and
are
caused
by
the
same
metabolites,
pyrrolizi-
dine
alkaloids
which
are
acutely
toxic
in
a
particular
species
are
likely
also
to
produce
chronic
toxicity
in
that
species.
Carcinogenicity
While
there
are
no
reported
cases
of
cancer
in
humans
that
have
arisen
as
a
result
of
the
ingestion
of
comfrey
or,
indeed,
of
any
other
source
of
pyrrolizidine
alkaloids
this
is
not
very
surprising
in
view
of
the
difficulty
of
determining
the
aetiology
of
cancer
and
the
relatively-small
number
of
persons
who
use
comfrey.
Therefore,
indirect
evidence
for
human
carcinogenicity
must
rely
on
data
from
experimental
animals;
in-vitro
genotoxicity
data;
an
analysis
of
the
pathways
of
pyrrolizidine
alkaloid
metabolism;
and,
if
available,
a
comparison
between
animal
and
human
pathology
reports.
A
large
number
of
studies
has
shown
that
purified
pyrrolizidine
alkaloids
that
are
administered
to
animals
in
the
diet
can
produce
tumours
at
a
variety
of
sites.'
Similarly,
dried
plants
or
plant
extracts
that
contain
pyrrolizidine
alkaloids
also
have
produced
an
increased
incidence
of
tumours
in
rats.
Hirono
et
al.
fed
dried
and
milled
S.
officinale
(common
comfrey)
leaves
and
roots
to
rats
for
up
to
20
months
at
doses
that
varied
from
1%
to
16%
(wt/wt)
of
the
total
diet."
Animals
in
all
treated
groups
had
an
increased
incidence
of
tumours,
particularly
of
liver
adenomas
and
urinary
bladder
papil-
lomas
and
carcinomas,
and
haemangioendothelial
sarcomas.
However,
the
relatively-high
doses
that
were
used
in
these
studies
leave
open
the
possibility
of
secondary
effects
that
influenced
the
development
of
cancers.
An
alternative
explanation
for
the
apparent-
low
oncogenic
potency
of
pyrrolizidine
alkaloids
in
rats
is
the
clear
antimitotic
effect
of
pyrrolizidine
alkaloids
in
this
species,
which
also
may
inhibit
tumour
formation.
Supporting
evidence
for
the
carcinogenicity
of
pyrrolizidine
alkaloids
has
come
from
an
analysis
of
their
genotoxicity."
All
the
organic
chemicals
which
are
known
human
carcinogens
have
produced
positive
results
in
appropriate
short-term
mutagenicit\
assays.
While
a
positive
result
for
mutagenicity
does
not
always
indi-
cate
carcinogenicity,
chemicals
that
give
a
positive
result
in
a
range
of
short-term
tests,
particularly
in-vivo
tests,
would
be
regarded
as
possible
animal
and
human
carcinogens.
Positive
results
for
mutagenicity
have
been
obtained
for
a
range
of
pyrrolizidine
alkaloids
in
assays
in
Drosophila
melanogaster
and
in
Salmonella
typhimurium,
as
well
as
in
mammalian
cells
in
culture.'
Pyrrolizidine
alkaloids
have
been
shown
to
induce
both
point
muta-
tions
and
chromosomal
aberrations,
as
well
as
being
able
to
induce
sister
chromatid
exchange
and
"unscheduled"
DNA
synthesis
in
mammalian
cells
in
culture.'
Positive
results
in
cell-transformation
assays
also
have
been
obtained.'
While
the
genotoxicity
data
still
are
limited,
the
data
support
the
view
that
pyrrolizidine
alkaloids
may
be
carcinogenic
by
way
of
a
genotoxic
mechanism.'
If
this
is
the
case,
the
threshold
dose
for
the
development
of
cancer,
if
it
exists,
is
likely
to
be
low.
Culvenor
has
examined
the
dosage
levels
of
pyrrolizidine
alkaloids
that
were
encountered
in
known
instances
of
their
intake
by
humans
and
compared
them
with
the
dosage
levels
which
have
led
to
a
signifi-
cant
increase
in
the
incidence
of
tumours
in
experimental
animals
:
9
Such
a
comparison
is
difficult
with
the
number
of
different
pyrrolizi-
dine
alkaloids
and
the
variety
of
dosing
schedules.
The
known
rates
of
intake
by
humans
vary
from
0.01
to
50
mg/kg
a
day,
while
in
rats,
tumours
have
been
produced
at
dosage
levels
of
0.2
to
3
mg/kg
a
day
for
a
12-month
period,
which
sometimes
were
preceded
by
an
initial
four-week
period
at
a
higher
dosage
range
of
2
to
6
mg/kg
a
day.'
Unfortunately,
the
major
pyrrolizidine
alkaloids
to
be
implicated
in
outbreaks
of
poisoning
in
Afghanistan
(heliotrine)'
2
and
India
(crotananine
and
cronaburmine)"
have
not
been
tested
for
their
carcinogenicity
in
rats.
However,
the
relatively-low
oncogenic
doses
that
are
noted
above
are
close
to
or
are
within
the
range
of
estimated
intake
levels
in
the
human
cases
of
poisoning.
Culvenor
has
suggested
that
by
monitoring
the
survivors
of
the
outbreaks
of
poisoning
in
Afghanistan
and
India,
it
eventually
may
be
possible
to
obtain
epidemiological
evidence
for
or
against
an
association
between
the
consumption
of
pyrrolizidine
alkaloids
and
cancer
in
humans."
Low-dose
exposure
There
can
be
little
doubt
that
the
excessive
consumption
of
pyrrolizi
-
dine
alkaloids,
either
as
a
single
high
dose
or
as
multiple
moderate
doses,
can
be
harmful.
While
some
variation
in
the
degree
of
toxicity
may
occur,
the
majority
of
pyrrolizidine
alkaloids,
including
those
that
are
found
in
comfrey,
have
been
associated
with
toxic
symptoms
in
either
animals
or
humans.
At
low
doses,
such
an
association
is
much
more
difficult
to
establish,
and
requires
an
extrapolation
to
dosage
levels
where
the
toxic
effect
either
is
not
measurable
easily
or
may
take
a
long
time
to
be
expressed.
Over
a
lifetime,
efficient
detoxification
or
repair
may
be
able
to
minimize
cellular
damage
or,
on
the
other
hand,
small
cumulative
damage
eventually
may
cause
serious
illness.
The
question
of
whether
permanent
and/or
cumulative
damage
can
occur
in
humans
after
Necrosis
Antirnitotic
activity
THE
MEDICAL
JOURNAL
OF
AUSTRALIA
Vol.
149
December
5/19,
1988
Therapeutics
681
moderate
exposure
to
cornfrey
and
can
result
either
in
veno-occlusive
disease
or
in
mutations
and
cancer
is
the
basis
for
the
current
debate.
The
likelihood
of
low-dose
acute
and
long-term
effects
are
considered
below.
The
most
notable
acute
effect
in
laboratory
animals
of
poisoning
with
pyrrolizidine
alkaloids
is
haemorrhagic
necrosis
of
the
liver.
The
high
doses
that
are
required
to
produce
this
lesion
make
it
very
unlikely
to
be
found
in
humans
at
normal
levels
of
comfrey
consump-
tion.
The
estimated
daily
intake
of
pyrrolizidine
alkaloids
from
the
use
of
Russian
comfrey
leaves
(5
mg/person
a
day'
9
)
is
far
below
the
estimated
acute
LD,0
value
in
rats
for
the
mixture
of
pyrrolizi-
dine
alkaloids
in
comfrey
(550
mg/kg").
The
daily
intake
from
root-tea
preparations
or
tablets
is
much
more
variable
but
still
is
unlikely
to
cause
acute
effects
unless
these
are
ingested
to
excess.
While
S.
officinale
(common
comfrey)
contains
less
of
the
more
toxic
pyrrolizidine
alkaloids,
echimidine
and
symphytine,
than
does
Russian
comfrey,
the
variability
in
both
the
composition
and
the
strength
of
herbal
preparations
means
that
the
differences
in
over-
all
acute
toxicity
between
the
two
plant
species
is
likely
to
be
small.
Also,
extrapolation
of
rat
L13
5
°
data
to
a
risk
in
humans
is
not
always
reliable
and
can
be
influenced
by
species
differences,
as
well
as
by
individual
differences
in
sex,
age,
diet
and
concurrent
medication.
From
the
small
amount
of
human
data
that
is
available,
Mattocks
has
concluded
that
humans
may
be
slightly-more
susceptible
to
the
acute
effects
of
pyrrolizidine
alkaloids
than
are
rats.
6
However,
considering
all
the
above
factors,
the
risk
to
humans
of
acute
liver
necrosis
as
a
result
of
the
moderate
consumption
of
comfrey
appears
to
be
very
low.
One
of
the
major
concerns
of
the
long-term
use
of
comfrey
is
veno-
occlusive
disease,
although
the
data
that
are
available
to
assess
the
possible
occurrence
of
veno-occlusive
disease
after
low-dose
exposure
to
pyrrolizidine
alkaloids
are
limited.
Pathological
evidence
of
cumulative
liver
damage
in
rats
indicates
progressive
liver
degener-
ation
over
a
lifetime.'"
In
humans,
the
pathological
changes
that
are
observed
in
the
liver
appear
to
be
somewhat
different,
although
case
studies
suggest
that
after
the
initial
insult,
liver
damage
is
progressive."
-
"
At
present,
there
is
no
way
to
define
the
dose
at
which
liver
effects
may
become
significant
and
thus
lead
to
progressive
damage,
in
either
rats
or
humans.
The
estimated
doses
of
pyrrolizidine
alkaloids
from
the
Afghan
poisoning
outbreak
(30
1.4g/kg
a
day)"
and
from
one
of
the
cases
of
comfrey
poisoning
(15
µg/kg
a
day)"
suggest
that
veno-occlusive
disease
may
develop
after
a
relatively-modest
intake.
On
the
basis
of
the
present
data,
it
is
not
possible
to
determine
a
"safe"
level
of
long-term
exposure
to
avoid
veno-occlusive
disease.
An
additional
concern
is
the
greater
susceptibility
of
children
to
pyrrolizidine
alkaloid-induced
liver
damage,
as
indicated
by
the
relatively-high
number
of
cases
of
cirrhosis
to
be
reported
in
this
age-group
in
exposed
communities."
The
"safe"
dosage
level
in
this
age-group
is
likely
to
be
much
lower.
The
other
potential
concerns
of
long-term
exposure
are
mutation
and
cancer.
The
ability
of
pyrrolizidine
alkaloids
to
produce
genetic
Mutations
both
point
mutations
and
chromosomal
aberrations
in
in-vitro
mammalian
cells
is
well
established.
6
However,
their
inutagenic
potential
in
whole
animals,
particularly
at
low-dosage
levels,
is
not
known
but,
in
view
of
the
relative
stability
of
the
pyrrolic
alcohol
metabolite,
mutations
in
the
germ
cells
must,
at
least,
be
Considered
possible,
until
further
testing
is
carried
out.
With
regard
to
cancer,
the
estimation
of
the
risk
to
humans
at
low
levels
of
exposure
to
a
chemical
is
notoriously
difficult,
and
requires
both
species
and
dose
extrapolations
for
which
there
is
little
or
no
information.
Numerous
variables
may
influence
both
of
these
extrapolations,
and
in
the
case
of
pyrrolizidine
alkaloids
and
comfrey,
more
information
is
required
before
the
risk
to
humans
can
be
determined
with
any
confidence.
Nevertheless,
there
are
a
number
of
reasons
why
the
risk
of
cancer
from
the
moderate
use
of
comfrey,
particularly
herbal
preparations,
may
be
low.
First,
there
is
the
evidence
of
animal
studies
to
suggest
that
pyrrolizidine
alkaloids
are,
at
best,
weak
carcinogens,
and
produce
liver
tumours
at
only
a
relatively-low
incidence
after
high-
dose
exposure.
This
may
indicate
a
threshold
effect
and/or
a
slow
rate
of
cumulative
liver
damage.
Secondly,
the
incidence
of
human
liver
cancer
is
low
even
in
populations
where
the
use
of
herbal
medicines
is
widespread.
Finally,
comfrey
herbal
treatments
gener-
ally
are
prescribed
by
herbalists
for
periods
that
do
not
exceed
three
months.
The
use
of
comfrey
herbal
preparations
or
comfrey
leaves
in
large
amounts
over
a
long
period
of
time
may,
of
course,
increase
the
risk
of
cancer
significantly.
Over
and
above
the
scientific
arguments
that
advocate
a
reduced
intake
of
comfrey,
the
question
of
an
individual's
freedom
of
choice
to
determine
an
acceptable
risk
must
be
considered.
Unfortunately,
public
perceptions
of
acceptable
risk
rarely
coincide
with
the
real
risk,
and
can
be
influenced
greatly
by
the
perceived
benefit.
An
example
of
this
is
the
widespread
misconception
that
naturally-
occurring
chemicals
intrinsically
are
less
harmful
than
are
the
synthetic
chemicals
that
are
used
in
food.
In
some
cases,
the
limited
toxicological
data
on
naturally-occurring
chemicals
can
hide
the
potential
hazards,
and
the
health
risks
in
these
cases
cannot
be
ignored.
Conclusions
The
restriction
of
the
distribution
of
comfrey
by
controls
through
state
Poisons
Acts
has
served
to
raise
the
awareness
of
consumers
of
comfrey
to
the
potential
hazard,
and
may
help
towards
reducing
the
level
of
unintentional
overdosing.
While
minimal
direct
evidence
exists
of
liver
toxicity
in
humans
as
a
result
of
the
consumption
of
comfrey,
this
may
be
due
to
a
long
latency
period
and
subclinical
symptoms,
or
to
a
lack
of
recognition
of
the
symptoms
of
veno-
occlusive
disease,
particularly
if
these
are
mild.
From the
data
that
are
available,
it
must
be
concluded
that
the
potential
exists
for
pyrrolizidine
alkaloid-induced
liver
toxicity
after
long-term
exposure,
albeit
with
mild
effects,
even
at
modest
levels
of
intake.
With
regard
to
carcinogenicity,
although
the
evidence
at
present
suggests
that
comfrey
should
be
regarded
as
a
potential
human
carcinogen,
it
seems
likely
that
at
the
normally-low
levels
of
exposure,
the
risk
would
be
very
small.
More
definite
conclusions
regarding
the
potential
toxicity
of
comfrey
in
humans
have
been
hindered
by
a
lack
of
information
on
exposure
levels,
by
limited
toxicokinetic
data
in
humans
or
animals,
and
by
the
absence
of
a
suitable
laboratory
animal
in
which
to
study
veno-occlusive
disease.
The
numbers
of
reported
and
poten-
tial
cases
of
pyrrolizidine-alkaloid
poisoning
world-wide
justifies
further
research
in
these
areas.
While
complete
toxicological
data
for
comfrey
are
not
yet
avail-
able,
there
is
sufficient
concern
to
advocate
a
restricted
intake.
This
can
be
achieved
either
voluntarily
through
an
educational
programme
and
the
use
of
warning
labels,
or
by
restrictions
on
the
availability
and
distribution
of
comfrey.
The
second
of
these
two
options
has
been
taken,
and
controls
under
Poisons
Acts
have
been
used
to
minimize
the
exposure
of
the
public
to
comfrey.
However,
in
view
of
the
diversity
of
plant
products
that
contain
pyrrolizidine
alkaloids,
the
public
also
requires
education
about
the
potential
dangers
of
such
naturally-occurring
chemicals
which
are
found
in
plants.
References
I.
Smith
LW,
Culvenor
CC.
Plant
sources
of
hepatotoxic
pyrrolizidine
alkaloids.
J
Nat
Prod
1981;
44:
129-152.
2.
Culvenor
CC,
Edgar
JA,
Frahn
JL,
Smith
LW.
The
alkaloids
of
Symphytuto
x
uplandicum
(Russian
comfrey).
Aust
J
Chew
1980;
33:
1105-1113.
3.
Mattocks
AR.
Toxic
pyrrolizidine
alkaloids
in
comfrey.
Lancet
1980;
2:
1136-1137.
4.
Bull
1.B,
Culvenor
CC,
Dick
AT.
The
pyrrolizidine
alkaloids.
Amsterdam:
North
Holland
Publishing
Co,
1968.
5.
Peterson
JE,
Culvenor
CC.
Hepatotoxic
pyrrolizidine
alkaloids.
In:
Keeler
RI,
Tu
AT,
eds.
Hand-
book
of
natural
toxins.
Vol
I.
New
York:
Marcel
Dekker,
1983:
637-671.
6.
Mattocks
AR.
Chemistry
and
toxicology
of
pyrrolizidine
alkaloids.
London:
Academic
Press,
1986.
7.
WHO
International
Program
on
Chemical
Safety.
Environmental
health
criteria
for
pyrrolizi-
dine
alkaloids.
Geneva:
WHO
1988
(in
press).
8.
Huxtablc
RJ.
Herbal
teas
and
toxins:
novel
aspects
of
pyrrolizidine
poisoning
in
the
United
States.
Perspect
Rio!
Med
1980;
24:
1-14.
9.
Kumana
CR,
Ng
M,
Lin
Hi,
el
al.
Herbal
tea
induced
hepatic
veno-occlusive
disease:
quantifi-
cation
of
toxic
alkaloid
exposure
in
adult.
Gut
1985;
26:
101-104.
10.
McGee
J,
Patrick
RS,
Wood
CB,
Blumgart
1.H.
A
case
of
veno-occlusive
disease
of
the
liver
in
Britain
associated
with
herbal
tea
consumption.
J
Clin
Porno/
1976;
29:
788-794.
682
Therapeutics
December
5/19,
1988
Vol.
149
THE
MEDICAL
JOURNAL
OF
AUSTRALIA
11.
Stuart
KL,
Bras
G.
Veno-occlusive
disease
of
the
liver.
Q
J
Med
1957;
26:
291-315.
12.
Mohabbat
0,
Srivastava
RN,
Younos
MS,
et
al.
An
outbreak
of
hepatic
veno-occlusive
disease
in
north-western
Afghanistan.
Lancet
1976;
2:
269-271.
13.
Tandon
BN,
Tandon
RK,
Tandon
HO,
et
al.
An
epidemic
of
veno-occlusive
disease
of
liver
in
central
India.
Lancet
1976;
2:
271-272.
14.
Ridker
PM,
Ohkuma
S,
McDermott
WV,
et
al.
Hepatic
venocclusive
disease
associated
with
the
consumption
of
pyrrolizidine-containing
dietary
supplements.
Gastroenterology
1985;
88:
1050-1054.
15.
Weston
CF,
Cooper
BT,
Davies
JD,
Levine
DF.
Veno-occlusive
disease
of
the
liver
secondary
to
ingestion
of
comfrey.
Br
Med
J
1987;
295:
183.
16.
Bras
G,
Brooks
SE,
Walter
DC.
Cirrhosis
of
the
liver
in
Jamaica.
J
Pathol
Bact
1961;
82:
503-512.
17.
Gupta
PS,
Gupta
GD,
Sharma
ML.
Veno-occlusive
disease
of
liver.
Br
Med
J
1963;
1:
1184-1186.
18.
Culvenor
CC,
Clarke
M,
Edgar
JA,
et
al.
Structure
and
toxicity
of
the
alkaloids
of
Russian
comfrey
(Symphytum
X
uplandicum
Nyman),
a
medicinal
herb
and
item
of
human
diet.
Experientia
1980;
36:
377-379.
19.
Culvenor
CC.
Estimated
intakes
of
pyrrolizidine
alkaloids
by
humans.
A
comparison
with
dose
rates
causing
tumors
in
rats.
J
Toxicol
Environ
Health
1983;
11:
625-635.
20.
Roitman
1N.
Comfrey
and
liver
damage.
Lancet
1981;
I:
944.
The
calcium
antagonist
drugs*
21.
Brauchli
J,
Liithy
J,
Zweifel
U,
Schlatter
C.
Pyrrolizidine
alkaloids
from
Symphytum
officinale
and
their
percutaneous
absorption
in
rats.
Experientia
1982;
38:
1085-1087.
22.
McLean
EK.
The
toxic
actions
of
pyrrolizidine
(senecio)
alkaloids.
Pharmacol
Rev
1970;
22:
429-483.
23.
Tandon
BN,
Tandon
HD,
Mattock
AR.
Study
of
an
epidemic
of
veno-occlusive
disease
in
Afghanistan.
Indian
J
Med
Res
1978;
68:
84-90.
24.
Armstrong
SJ,
Zuckerman
AJ.
Production
of
pyrroles
from
pyrrolizidine
alkaloids
by
human
embryo
tissue.
Nature
1970;
228:
569-570.
25.
Jago
MV.
The
development
of
hepatic
megalocytosis
of
chronic
pyrrolizidine
alkaloid
poisoning.
Am
J
Pathol
1969;
56:
405-422.
26.
Peterson
JE,
Jago
MV.
Toxicity
of
Echium
plantagineum
(Paterson's
curse):
pyrrolizidine
alkaloid
poisoning
in
rats.
Aust
J
Agric
Res
1984;
35:
305-316.
27.
Hirono
I,
Mori
H.
Haga
M.
Carcinogenic
activity
of
Symphytum
officinale.
JNCI
1978;
61:
865-869.
28.
Schoental
R,
Magee
PN.
Chronic
liver
changes
in
rats
after
a
single
dose
of
lasiocarpine,
a
pyrrolizi-
dine
(senecio)
alkaloid.
J
Pathol
Bart
1957;
74:
305-319.
29.
Schoental
R,
Magee
PN.
Further
observation
on
the
subacute
and
chronic
liver
changes
in
rats
after
a
single
dose
of
various
pyrrolizidine
(Senecio)
alkaloids.
J
Pathol
Baer
1959;
78:
471-482.
T
he
last
few
years
have
seen
an
explosion
of
interest
in
the
field
of
"calcium
antagonism".'
New,
long-acting
(see
box)
and,
in
some
cases,
tissue-selective
drugs
have
been
developed,
the
therapeutic
indications
for
their
use
have
expanded,
their
modes
of
action
have
been
clarified,
and
their
receptors
have
been
identified.
In
addition,
drugs
which
act
in
the
opposite
manner,
and
enhance
calcium-ion
influx
through
the
conducting
channels,
have
been
identified.'
These
drugs
are
known
as
"calcium
agonists".
In
spite
of
this
exacerbation
of
interest,
the
predominant
property
of
the
calcium
antagonist
agents
remains
that
of
impeding
the
movement
of
calcium
ions
through
the
voltage-activated,
calcium-
selective
channels
which
traverse
the
membranes
of
most
excitable
cells,
including
those
of
the
heart
and
vasculature.'
This
is
an
important
property,
because
the
calcium
ions
which
gain
access
to
the
cytosol
in
this
way
are
not
simply
charge
carriers.
They
also
act
as
chemical
transmitters
for
processes
as
diverse
as
excitation-
contraction
coupling,
impulse
conduction,
neuronal
activity
and
excitation-secretion
coupling.
As
yet,
no
one
has
succeeded
in
seeing
a
calcium
channel,
but
recent
technological
advances
have
enabled
detailed
electrophysio-
logical
and
biochemical
studies
to
be
undertaken
on
them.
As
a
result,
their
size
has
been
determined,
the
electrical
activity
that
is
associated
with
their
activation
and
closure
has
been
recorded
on
a
single-channel
basis,"
and
the
biochemistry
of
their
associated
binding
sites
for
calcium
antagonist
(and
agonist)
drugs
has
been
elucidated.'
The
channels
have
even
been
harvested,
solubilized,
and
reconstituted
in
artificial
lipid
bilayers
with
such
precision
that
they
retain
their
electrophysiological
and
pharmacological
properties.
6
.
7
Therefore,
a
parallel
development
has
occurred
in
our
understanding
of
the
basic
physiological
properties
of
the
calcium
channels,
and
the
mechanisms
whereby
calcium
antagonist
and
agonist
drugs
modulate
their
activity.
This
article
considers
some
of
the
recent
developments
in
this
field.
In
particular,
it
concentrates
on:
calcium-channel
heterogeneity;
the
biochemistry
of
the
calcium-antagonist
binding
sites;
their
tissue
selectivity;
calcium
agonism;
and
the
pathophysiology
of
calcium-
antagonist
binding
sites.
Calcium-channel
heterogeneity
Voltage-sensitive
calcium
channels
The
characteristic
which
is
shared
by
all
calcium
antagonist
agents,
irrespective
of
their
chemistry,
is
their
ability
to
attenuate
the
movement
of
calcium
ions
through
the
voltage-activated
calcium
channels.'
These
channels
are
about
0.6
nm
wide
at
their
narrowest
point,'
and
although
the
ionic
radius
of
a
non-hydrated
sodium
ion
is
less
than
that
of
a
calcium
ion
(0.97
nm
for
sodium,
and
0.13
nm
for
calcium),
they
do
not
admit
sodium
ions,
provided
that
the
extracellular
fluid
contains
calcium
ions
and
its
pH
is
not
less
than
'First
of
two
articles
on
the
calcium
antagonist
agents
and
their
clinical
use.
University
of
Melbourne
Department
of
Medicine,
Austin
Hospital,
Heidelberg,
VIC
3084.
Winifred
G.
Nayler,
DSc,
Principal
Research
Investigator.
Reprints:
Dr
W.G.
Nayler.
Winifred
G.
Nayler
Some
calcium
antagonist
agents
Inorganic
Recently-developed
drugs
Cobalt
Dihydropyridines
Nickel
Nisoldipine
Lanthanum
Niludipine
Nitrendipine
Organic
Nicardipine
Prototypes
Felodipine'
Verapamil'
(phenylalkylamine)
Amlodipinet
Nifedipine•
(dihydropyridine)
Phenylalkylamines
Diltiazem•
(benzothiazepine)
Anipamilt
Gallopamil
Tiapamil
Available
in
Australia.
tLong-acting
drug.
This
can
be
a
result
of
slowed
metabolism
or
persistent
binding
to
the
receptor.
6.0
(that
is,
it
is
not
severely
acidotic).
The
calcium
ions
move
through
the
channels
in
single
file
along
their
concentration
gradient
(1000
to
one;
outside
relative
to
inside);
as
many
as
10
million
ions
can
diffuse
through
each
channel
per
second.
The
channels
are
gated,
pore-like
structures
and
since
their
activation
is
voltage-dependent
they
must
contain
a
"voltage
sensor".
This
sensor
must
in
turn
be
linked
to
the
"gates"
which
regulate
the
opening
and
closing
of
the
channels.
Each
channel
has
an
aqueous-filled
central
pore,
or
lumen."
The
channels
occur
in
all
excitable
tissues
and
in
all
animals
-
from
primitive
protozoans
to
humans.
This,
together
with
the
fact
that
they
develop
early
during
embryology,
before
any
sodium-conducting
channels
appear,
probably
mean
that
they
were
an
early
evolutionary
development.
Much
of
the
recent
data
which
have
been
obtained
concerning
the
functioning
of
these
channels
has
come
from
"patch-clamp"
studies.
This
technique
involves
pressing
the
polished
end
of
a
glass
micropipette
(Figure
1)
against
the
surface
of
a
cell
and
applying
a
negative
pressure,
so
that
a
small
patch
of
membrane
inverts
into
the
lumen
of
the
pipette
and
forms
a
tight
seal
against
the
wall
of
the
pipette.
The
lumen
of
the
pipette
then
is
filled
with
a
conducting
solution
and
is
connected
to
a
recording
amplifier.
Because
the
"patch"
of
membrane
remains
attached
to
the
cell
(Figure
1),
the
potential
difference
across
it
can
be
modified
by
stimulating
the
cell
in
the
traditional
manner.
The
special
feature
of
this
technique
is
that
because
the
"patch"
area
is
extremely
small,
it
usually
contains
only
one
or
two
functional
channels.
Therefore,
the
technique
has
allowed
the
electrical
activity
of
single
channels
to
be
studied.'
Some
unexpected
results
have
been
obtained.
In
particular,
it
seems
that
the
gates
which
control
the
opening
and
closing
of
each
channel
can
function
in
one
of
three
states,
or
"modes",
which
are
designated
as
modes
"0",
"1"
and
"2".
9
Mode
0
refers
to
a
state
in
which
the
channel
is
unavailable
for
calcium
flux,
even
though
the
transmembrane
potential
difference
is
in
the
range
that
is
required
for
activation.
Mode
I
refers
to
a
"flickering"
type
of
activity
in
which
the
gates,
and
hence
the
channels,
seem
to
open
and
shut
at
a
rapid
rate.
Therefore,
this
mode
is
charac-