Changes in Plasma -Glutamyl Transpeptidase Activity Associated with Alterations in Drug Metabolism in Man

Abstract

A significant rise in plasma gamma-glutamyl transpeptidase activity (GGT) was observed on 13 out of 14 occasions on which patients on long-term treatment with the oral anticoagulant warfarin were given amylobarbitone, quinalbarbitone, or phenazone (antipyrine) for 30 days. In 13 of these 14 studies there was evidence that drug administration had stimulated the rate of warfarin metabolism. One patient showed no increase in plasma GGT activity, yet a significantly increased rate of warfarin metabolism, and another patient showed an increase in plasma GGT activity without a change in warfarin metabolism. When alterations in both plasma GGT activity and plasma warfarin concentration occurred together in response to drug administration the changes followed a similar time course, occurring after about one week of drug administration with maximal changes at about 10 or 15 days. Administration of chlordiazepoxide, diazepam, nitrazepam, and methaqualone did not stimulate the rate of warfarin metabolism in four patients studied, but plasma GGT activity increased significantly in two of these four instances. The implications of these observations in the interpretation of plasma GGT activities are discussed.

Full text

Changes
in
Plasma
T-Glutamyl
Transpeptidase
Activity
Associated
with
Alterations
in
Drug
Metabolism
in
Man
J.
B.
WHITFIELD,
D.
W.
MOSS,
G.
NEALE,
M.
ORME,
A.
BRECKENRIDGE
British
Medical
Journal,
1973,
1,
316-318
Summary
A
significant
rise
in
plasma
y-glutamyl
transpeptidase
activity
(GGT)
was
observed
on
13
out
of
14
occasions
on
which
patients
on
long-term
treatment
with
the
oral
anticoagulant
warfarin
were
given
amylobarbitone,
quinalbarbitone,
or
phenazone
(antipyrine)
for
30
days.
In
13
of
these
14
studies
there
was
evidence
that
drug
administration
had
stimulated
the
rate
of
warfarin
metabolism.
One
patient
showed
no
increase
in
plasma
GGT
activity,
yet
a
significantly
increased
rate
ofwarfarin
metabolism,
and
another
patient
showed
an
increase
in
plasma
GGT
activity
without
a
change
in
warfarin
metabolism.
When
alterations
in
both
plasma
GGT
activity
and
plasma
warfarin
concentration
occurred
together
in
response
to
drug
administration
the
changes
followed
a
similar
time
course,
occurring
after
about
one
week
of
drug
administration
with
maximal
changes
at
about
10
or
15
days.
Administration
of
chlordiazepoxide,
diazepam,
nitrazepam,
and
methaqualone
did
not
stimulate
the
rate
of
warfarin
metabolism
in
four
patients
studied,
but
plasma
GGT
activity
increased
significantly
in
two
of
these
four
instances.
The
impli-
cations
of
these
observations
in
the
interpretation
of
plasma
GGT
activities
are
discussed.
Introduction
Many
studies
have
shown
that
enzyme
activity
in
the
endo-
plasmic
reticulum
of
the
liver
is
markedly
increased
by
adminis-
tration
of
various
drugs,
hormones,
or
other
chemicals
(Conney,
1967).
The
greater
activity
appears
to
reflect
an
increased
concentration
of
enzyme
protein
and
is
referred
to
as
enzyme
induction.
This
is
of
clinical
importance
because
administration
of
inducing
agents,
such
as
barbiturates,
can
alter
the
duration
and
intensity
of
action
of
many
drugs,
including
the
oral
anticoagulant
warfarin,
which
are
inactivated
by
microsomal
enzymes.
Direct
proof
of
liver
microsomal
enzyme
induction
depends
on
showing
that
administration
of
an
inducing
agent
results
in
an
increase
of
enzyme
activity
in
the
liver.
Indirect
evidence
comes
from
the
finding
of
a
shortening
in
plasma
half-lives
and
lowered
steady-state
concentrations
in
the
plasma
of
drugs
metabolized
by
these
enzymes,
or
from
an
increased
rate
of
production
of
drug
metabolites.
Recently,
increases
in
the
activity
of
the
enzyme
y-glutamyl
transpeptidase
(GGT)
in
human
serum
after
administration
of
phenobarbitone
and
phenytoin
have
been
reported
(RosaLki
et
al.,
1971;
Whitfield
et
al.,
1972)
while
administration
of
phenobarbitone
to
rats
has
been
shown
to
be
accompanied
by
an
increased
activity
of
this
enzyme
in
liver
tissue
(Ideo
et
al.,
1972).
The
possibility
of
changes
in
serum
GGT
activity
resulting
from
drug
administration
in
man
introduces
a
complicating
Royal
Postgraduate
Medical
School,
London
W12
OHS
J.
B.
WHITFIELD,
PH.D.,
M.R.C.PATH.,
Senior
Scientific
Officer,
Depart-
ment
of
Chemical
Pathology
D.
W.
MOSS,
PH.D.,
Reader
in
Enzymology,
Department
of
Chemical
Pathology
G.
NEALE,
M.B.,
F.R.C.P.,
Lecturer
in
Medicine
M.
ORME,
M.B.,
M.R.C.P.,
Senior
Registrar
in
Clinical
Pharmacology
A.
BRECKENRIDGE,
M.B.,
M.R.C.P.,
Lecturer
in
Clinical
Pharmacology
factor
into
the
use
of
this
enzyme
test
in
the
investigation
of
liver
function
(Whitfield
et
al.,
1972).
On
the
other
hand,
alterations
in
serum
GGT
may
provide
information
regarding
the
activity
of
drug-oxidizing
enzymes
in
the
liver.
Correct
interpretation
of
serum
GGT
activities
will
depend
on
a
knowledge
of
such
factors
as
the
type
of
compounds
cavable
of
altering
serum
GGT,
possible
dependence
on
dose,
and
tte
time-
course
of
drug-induced
changes
in
serum
enzyme
activity.
We
have
therefore
investigated
the
relation
between
changes
m
plasma
GGT
and
changes
in
the
rate
of
metabolism
of
warfarin
in
patients
given
a
series
of
hypnotic
and
sedative
drugs.
Patients
and
Methods
Altogether
11
patients
were
investigated,
four
on
more
than
one
occasion.
The
ages
of
these
patients,
the
reason
for
adminis-
tration
of
warfarin,
and
the
hypnotic
or
sedative
studied
are
shown
in
the
table.
No
patient
was
taking
any
other
drug
that
would
alter
warfarin
metabolism.
Age,
Reason
for
Administration
of
Warfarin,
and
Sedative
or
Hypnotic
Drug
Given
in
the
11
Patients
Studied
Case
Age
and
Clinical
Indication
for
Sedatives
and
Hypnotics
No.
Sex
Warfarin
Therapy
Administered
1
40
M.
Deep
vein
thrombosis
Amylobarbitone,
metha-
qualone,
phenzone
(antipyrine)
2
52
M.
Deep
vin
thrombosis
Phenazone
3
55
M.
Rheumatic
heart
disease
Chlordiazepoxide,
quinalbarbitone
4
57
M.
Deep
vein
thrombosis
Diazepam
amylobarbitone
5
60
M.
Deep
vein
thrombosis
Amylobarbitone
6
54
F.
Deep
vein
thrombosis
Phenazone
7
54
F.
Rheumatic
heart
disease
Dichloralphenazone
8
55
F.
Deep
vein
thrombosis
Quinalbarbitone
9
79
F.
Rheumatic
heart
disease
Dichloralphenazone
10
54
M.
Deep
vein
thrombosis
Nitrazepam
11
55
F.
Deep
vein
thrombosis
Amylobarbitone
Patients
on
long-term
warfarin
were
studied
for
a
control
period
of
at
least
30
days.
During
this
time
twice-weekly
blood
samples
were
taken
with
plastic
syringes
into
sodium
citrate
(10%
v/v).
Plasma
was
stored
at
-20°C
until
the
analyses
were
carried
out.
Plasma
warfarin
concentrations
were
measured
by
the
method
of
Lewis
et
al.,
(1970)
and
plasma
GGT
by
the
method
of
Szasz
(1969).
While
continuing
the
same
dose
of
warfarin
the
patient
was
given
the
drug
under
study,
usually
for
a
period
of
30
days,
and
twice-weekly
measurements
were
continued.
Also,
when
the
hypnotic
was
stopped,
measurements
were
continued
for
at
least
four
weeks.
Informed
consent
was
obtained
from
each
patient
for
these
investigations.
In
two
patients
the
period
of
drug
administration
was
increased
to
60
days.
For
both
the
plasma
warfarin
concentration
and
the
plasma
GGT
activity
the
mean
of
four
results
before
drug
administra-
tion
was
compared
with
the
mean
of
four
results
obtained
during
the
last
two
weeks
of
drug
therapy.
These
results
were
analysed
by
Student's
t
test.
Results
A
significant
fall
in
plasma
warfarin
concentration
was
accom-
panied
by
a
significant
rise
in
plasma
GGT
activity
(P
<
0.01)
on
13
out
of
14
occasions
(fig.
1).
In
these
studies
the
fall
in
316
BmisH
mEDicAL
jouRNAL
10
r
RuARY
1973

BRITISH
MEDICAL
JOURNAL
10
FEBRUARY
1973
'
C
4
4)
E
E
aX
a
v
V
.-
,
or
V
.3
_
-0
0
,
u
u
D
a
a
E
E
a
a
ElE
0
O
4)
.0_
00
EE
5.-
4.-
3.-
2-
0
A
S
A
A
a
-
A
A
S.'
A
0
0.;
0
0
-0o
2
0.4
0.
0Q8
10
2
Ratio
Plasma
Warfarin
during
treatment
Plasma
Warfarin
before
treatment
FIG.
1-Relation
between
changes
in
steady-state
plasma
warfarin
concen-
tration
and
plasma
Y-glutamyl
transpeptidase
(GGT)
in
11
patients
given
hypnotic
or
sedative
drugs
on
18
occasions.
o
=
Diazepam,
nitrazepam,
chiordiazepoxide,
and
methaqualone.
*=
Barbiturates.
A
=
Dichlora-
phenazone
and
phenazone
(antipyrine).
plasma
warfarin
and
the
rise
in
plasma
GGT
were
roughly
proportional.
In
seven
out
of
10
patients
whose
plasma
GGT
was
normal
before
drug
administration
the
rise
in
enzyme
activity
was
great
enough
to
cause
the
normal
range
to
be
exceeded.
The
enzyme
activity
was
initially
abnormal
on
the
three
other
occasions
on
which
drug
administration
produced
a
significant
rise
in
plasma
GGT.
The
time
scale
of
both
changes
was
very
similar.
Little
change
was
seen
in
either
value
during
the
first
week
of
drug
administration
and
a
maximum
change
occurred
in
about
10-15
days
(fig.
2).
After
withdrawal
of
the
drug,
pretreatment
levels
were
reached
in
15-20
days.
The
drugs
producing
these
changes
were
amylobarbitone
(three
studies),
dichloralphenazone
(one
study),
phenazone
(four
studies),
quinalbarbitone
(four
studies),
and
phenazone
with
amylobarbitone
(one
study).
There
was
a
single
exception
to
the
usual
finding
that
administration
of
enzyme-inducing
drugs
led
to
a
rise
in
circulating
GGT
activity.
In
one
patient
(case
9)
a
four-fold
change
in
plasma
warfarin
concentration
during
dichloralphenazone
therapy
was
accompanied
by
a
slight
change
in
plasma
GGT
activity
which
was
not
statistically
significant
(P
>
0-1).
On
four
occasions
nitrazepam,
diazepam,
chlordiazepoxide,
or
methaqualone
was
given
and
in
all
these
studies
there
was
no
significant
change
in
plasma
warfarin
concentration,
but
in
two
instances
a
significant
rise
in
plasma
GGT
activity
was
seen.
In
one
of
these
a
patient
(case
10)
treated
with
nitrazepam
showed
a
rise
from
an
initially
abnormal
level
of
30-9
+
1-6
to
40-3
±
7-0
IU/1.
(P
<
0-05)
while
in
the
other
patient
(case
1)
given
methaqualone
plasma
GGT
rose
from
14-4
±
1-4
to
21-5
±
3-6
IU/1.
(P
<
0-01).
(This
patient
had
40-
Plasma
GGT
20
-
0-
Phenazone|
5
Plasma
Warfarin
O.0
0
20
Time
in
days
40
60
80
FIG.
2-Time-course
of
changes
in
plasma
warfarin
concentration
and
GGT
activity
in
Case 6
during
and
after
administration
of
phenazone
(antipyrine).
earlier
shown
a
typical
rise
in
plasma
GGT
on
administration
of
phenazone
and
amylobarbitone.)
In
the
two
remaining
studies
with
diazepam
and
chlordiazepoxide
there
was
no
significant
change
in
either
plasma
GGT
activity
or
plasma
warfarin
concentration.
Although
most
patients
received
inducing
drugs
for
not
longer
than
30
days,
there
was
evidence
that
both
the
GGT
and
the
plasma
warfarin
concentration
reached
new
steady
state
levels
during
this
time.
In
two
patients
(cases
1
and
2)
the
period
of
drug
administration
was
increased
to
60
days.
In
the
first
of
these
patients,
initially
given
phenazone
300
mg
twice
a
day,
the
plasma
warfarin
concentration
fell
from
1-42
+
0*06
to
0-98
+
0-07
jig/ml
(P
<
0-01)
while
the
plasma
GGT
activity
rose
from
19-7
±
1-2
to
40-0
i
3-1
IU/1.
(P
<
0-01).
For
the
second
30
days
amylobarbitone
200
mg
at
night
in
addition
to
the
phenazone
caused
both
a
further
significant
rise
in
plasma
GGT
activity
(to
54-2
±
5-1
IU/1.,
P
<
0-01)
and
a
further
fall
in
plasma
warfarin
concentration
(to
0-50
:1
0-06
t±g/ml,
P
<
0-01).
In
the
second
patient
the
dose
of
phenazone
was
increased
from
300
mg
twice
daily
to
600
mg
twice
daily
for
the
last
30
days.
There
was
no
further
significant
fall
in
plasma
warfarin
concentration
nor
an
increase
in
GGT
activity
after
the
first
30
days
of
phenazone
therapy.
In
one
patient
(case
3)
who
was
given
quinalbarbitone
at
three
dose
levels
on
separate
occasions,
both
the
plasma
warfarin
concentration
and
the
plasma
GGT
activity
showed
some
evidence
of
a
dose-dependent
change
(fig.
3).
Quinalbarbitone
100
mg
at
night
caused
no
significant
change
in
plasma
warfarin
concentration.
However,
a
significant
rise
in
GGT
from
59.0
±
6-9
to
72-5
±
3.7
IU/1.
was
observed
(P
<
0-05).
During
administration
of
quinalbarbitone
200
mg
at
night
a
further
rise
of
plasma
GGT
activity
to
100-0
±
10-9
IU/1.
(P
<
0-01)
was
accompanied
by
a
fall
in
plasma
warfarin
concentrationfrom
2-97
i
0-25
to
lA41
±
0-22
jig/ml
(P
<
0-01).
However,
a
subsequent
increase
in
the
dose
of
quinalbarbitone
to
300
mg
produced
no
significant
further
fall
in
plasma
warfarin
concentration
or
further
significant
change
in
plasma
GGT
(P
>
0.1).
140
120
100
C)
80C)
4-40
E
2
0
-20
0
10
0Q0m
200mg
Q
300mg
Er
-
0
~
~
~ ~ ~ ~ ~ ~ ~ ~
~~4
20
0~~~~~~~~~0~~-
0
20
40
O0
20
40
60
0
20
40
60
Time
in
days
FIG.
3-Changes
in
plasma
GGT
and
plasma
warfarin
in
Case
3
given
quinalbarbitone
at
three
dose
levels
for
30
days.
0
=
Plasma
GGT.
o
Plasma
warfarin.
Discussion
The
day-to-day
variation
in
plasma
GGT
activity
in
any
one
subject
was
found
to
be
of
the
order
of
10%,
and
thus
slight
increases
in
activity
associated
with
drug
administration
could
be
detected
without
difficulty.
Increases
in
enzyme
activity
were
found
both
in
patients
with
normal
plasma
GGT
activity
and
in
those
who
had
raised
levels
in
the
control
period
in
13
out
of
the
14
occasions
on
which
drugs
known
to
induce
hepatic
microsomal
enzymes
were
given.
I
I
I
I
I
317
m
0

318
BRITISH
MEDICAL
JOURNAL
10
FERUARY
1973
It
has
been
shown
previously
(Breckenridge
et
al.,
1971;
Breckenridge
and
Orme,
1971;
Orme
et
al.,
1972)
that
of
the
drugs
studied
dichloralphenazone,
phenazone,
quinalbarbitone,
and
amylobarbitone
will
increase
the
rate
of
warfarin
metabolism
while
chlordiazepoxide,
diazepam,nitrazepam,
and
methaqualone
will
not.
This
property
was
used
as
a
marker
in
this
study
so
that
the
time-course
and
extent
of
changes
in
plasma
GGT
activity
could
be
compared
with
some
changes
in
values
known
to
reflect
hepatic
microsomal
enzyme
activity.
It
was
found
that
the
timing
of
the
changes
in
plasma
warfarin
concentration
was
closely
similar
to
that
of
changes
in
GGT
activity.
Both
changes
were
observable
by
seven
days,
in
most
cases
a
plateau
was
reached
after
10-15
days,
and
a
return
towards
baseline
values
began
immediately
drug
administration
was
stopped.
It
has
been
suggested
previously
that
changes
in
serum
GGT
are
found
only
after
three
months'
treatment
with
inducing
agents
(Rosalki
et
al.,
1971)
but
this
was
not
confirmed
in
the
present
investigation.
Some
evidence
of
a
relation
between
drug
dose
and
the
increase
in
enzyme
activity
was
also
apparent.
In
those
instances
where
a
rise
in
GGT
activity
occurred,
in
over
half
the
cases
the
increase
took
the
GGT
activity
above
the
upper
limit
of
normal.
Thus
the
possibility
of
drug-induced
changes
in
plasma
GGT
must
influence
the
interpretation
of
raised
GGT
activities
in
suspected
liver
disease,
especially
when
such
raised
levels
occur
unaccompanied
by
abnormalities
in
other
tests
of
liver
function
(Whitfield
et
al.,
1972).
Administration
of
three
benzodiazepines,
diazepam,
nitra-
zepam,
and
chlordiazepoxide
or
methaqualone
produced
no
change
in
steady
state
plasma
warfarin
concentration
in
four
patients,
but
two
of
these
patients
showed
a
significant
rise
in
plasma
GGT
activity.
If
these
changes
in
plasma
GGT
activity
reflect
changes
in
the
liver
enzymes
then
these
drugs
may
have
caused
an
increase
in
liver
GGT
without
an
increase
in
the
activity
of
enzymes
concerned
with
drug
oxidation.
Alternatively,
other
factors,
such
as
the
effects
of
alcohol
consumption,
may
have
influenced
GGT
activity
(Rosalki
and
Rau,
1972).
In
view
of
these
discrepancies
between
changes
in
rates
of
drug
oxidation
and
plasma
GGT
activity,
it
appears
that
increases
in
the
plasma
GGT
activity
cannot
always
be
used
as
an
index
of
changes
in
the
activity
of
liver
microsomal
enzymes
concerned
with
drug
oxidation
since
it
might
lead
to
both
false-positive
and
false-negative
results.
It
remains
to
be proved
that
the
increase
in
plasma
GGT
concentration
produced
by
inducing
agents
is
accompanied
by
an
increase
in
hepatic
GGT
in
man.
Other
possibilities
are
that
other
tissues
such
as
the
kidney
or
gut
are
the
source
of
the
increased
plasma
GGT
or
that
the
rate
of
catabolism
of
circulating
GGT
is
altered
or
that
these
drugs
produce
hepatic
damage
of
the
type
which
is
accompanied
by
an
increase
in
plasma
GGT
activity
in
hepatobiliary
disease.
However,
RosaLki
et
al.
(1972)
have
reported
that
a
wide
range
of
serum
enzymes
other
than
GGT
remained
normal
in
patients
in
whom
a
raised
serum
GGT
accompanied
drug
treatment.
This
last
explanation,
therefore,
appears
improbable.
Ideo
et
al.
(1972)
have
shown
an
increase
in
liver
GGT
content
in
rats
treated
with
phenobarbitone,
and
our
own
preliminary
observa-
tions
support
this.
Therefore,
the
most
likely
sequence
of
events
seems
to
be
that
the
induction
of
hepatic
microsomal
drug-metabolizing
enzymes
is
accompanied
by
induction
of
hepatic
GGT,
which
in
most
cases
is
released
into
the
circulation.
There
may
be
differences
between
species,
and
individuals
within
a
species,
in
the
extent
to
which
the
GGT
from
the
liver
appears
in
the
serum.
The
role
of
GGT
in
the
liver
and
the
effects
of
increased
activity
of
this
enzyme
there
are
at
present
obscure.
References
Breckenridge,
A.,
Orme,
M.
L'E.,
Thorgeirsson,
S.,
Davies,
D.
S.,
and
Brooks,
R.
V.
(1971).
Clinical
Science,
40,
351.
Breckenridge,
A.,
and
Orme,
M.
(1971).
Annals
of
the
New
York
Academy
of
Sciences,
179,
421.
Conney,
A.
H.
(1967).
Pharmacological
Reviews,
19,
317.
Ideo,
G.,
Morganti,
A.,
and
Dioguardi,
N.
(1972).
Digestion,
5,
326.
Lewis,
R.
J.,
Ilnicki,
L.
P.,
and
Carlstrom,
M.
(1970).
Biochemical
Medicine
4,
376.
Orme,
M.,
Breckenridge,
A.,
and
Brooks,
R. V.
(1972).
British
Medical
Journal,
3,
611.
Rosalki,
S.
B.,
and
Rau,
D.
(1972).
Clinica
Chimica
Acta,
39,
41.
Rosalki,
S.
B.,
Rau,
D.,
Tarlow,
D.,
and
Baylis,
E.
M.
(1972).
Proceedings
of
the
8th
International
Congress
on
Clinical
Chemistry,
Copenhagen.
Rosalki,
S.
B.,
Tarlow,
D.,
and
Rau,
D.
(1971).
Lancet,
2,
376.
Szasz,
G.
(1969).
Clinical
Chemistry,
15,
124.
Whitfield,
J.
B.,
Pounder,
R.,
Neale,
G.,
and
Moss,
D.
W.
(1972).
Gut,
13,
702.
Morbidity
from
Acute
Carbon
Monoxide
Poisoning
at
Three-year
Follow-up
J.
SIDNEY
SMITH,
S.
BRANDON
British
Medical
Journal,
1973,
1,
318-321
Summary
Seventy-four
survivors
of
acute
carbon
monoxide
poisoning
were
followed
up
for
an
average
of
three
years.
In
eight
patients
gross
neuropsychiatric
damage
was
directly
attributable
to
the
poisoning.
Three
patients
had
committed
suicide
and
eight
had
died
from
other
causes.
Morbidity
and
mortality
in
those
deliberately
and
accidentally
poisoned
was
approximately
equal.
University
Deparment
of
Psychological
Medicine,
Royal
Victoria
Infirmary,
ewcstle
upon
Tyne
1
J.
SYDNEY
SMITH,
M.A.N.Z.C.P.,
D.P.M.,
Research
Senior
Registrar
(Now
Lecturer
in
Psychiatry,
University
of
New
South
Wales,
Sydney,
Australia)
S.
BRANDON,
M.D.,
M.R.C.
Psym.,
Nuffield
Foundation
Felow
in
Psychiatry
and
Hononary
Consultant
Psychiatrist
(Now
Reader
in
Psychiatry,
Uni-
versity
of
Manchester)
Of
63
patients
alive
at
follow-up
eight
showed
an
im-
provement
and
21
(33
3%)
a
deterioration
of
personality
after
poisoning,
and
27
(43%)
reported
a
subsequent
impairment
of
memory.
Deterioration
of
personality
and
memory
impairment
were
highly
correlated.
The
level
of
consciousness
on
admission
to
hospital
in
the
acute
phase
of
poisoning
correlated
significantly
with
the
development
of
gross
neuropsychiatric
sequelae.
These
findings
emphasize
the
importance
of
prompt
and
efficient
treatment
of
carbon
monoxide
poisoning
and
the
need
to
foliow-up
all
cases
in
the
anticipation
of
a
relapsing
course
or
the
development
of
sequelae.
Introduction
Carbon
monoxide
(CO)
acts
pathogenically
by
displacing
oxygen
from
the
haemoglobin
molecule,
shifting
the
oxyhaemo-