Bioavailability of griseofulvin from tablets in humans and the correlation with its dissolution rate


Aoyagi, N.; Ogata, H.; Kaniwa, N.; Koibuchi, M.; Shibazaki, T.; Ejima, A.

Journal of Pharmaceutical Sciences 71(10): 1165-1169

1982


Dissolution rates of 10 commercial microsize griseofulvin tablets and one ultramicrosize griseofulvin tablet were preliminarily determined in 18 liters of pH 7.2 phosphate buffer and in 900 ml of 40% dimethylformamide as test media. Addition of dimethylformamide affected the dissolution behavior of the formulations. The products, three microsize and one ultramicrosize, were selected for further studies on the bioavailability in humans and dissolution. Significant differences among the formulations were found in serum levels Cmax, and AUC47.5 hr, but not in AUC infinity and tmax. The maximum difference of Cmax was approximately 40%. The ultramicrosize product showed lower Cmax and serum levels at earlier sampling times than two microsize products. The dissolution rates determined under sink and nonsink conditions without pretreatment significantly correlated with the serum level at 1 hr but not with the other in vivo parameters. Only the dissolution rate determined by the sink method with pretreatment with a small quantity of water (1.0 ml) and plastic beads significantly correlated with serum levels at 3 and 5 hr, Cmax, and AUC 47.5 hr.

Bioavailability
of
Griseofulvin
from
Tablets
in
Humans
and
the
Correlation
with
its
Dissolution
Rate
NOBUO
AOYAGI
x,
HIROYASU
OGATA,
NAHOKO
KANIWA,
MASANOBU
KOIBUCHI
t,
TOSHIO
SHIBAZAKI,
and
AKIRA
EJIMA
Received
July
31,
1981,
from
the
Division
of
Drugs,
National
Institute
of
Hygienic
Sciences,
18-1,
Kamiyoga
1-chome,
Setagaya-ku,
Tokyo
158,
Japan.
tDeceased.
Accepted
for
publication
December
30,
1981.
Abstract
13
Dissolution
rates
of
10
commercial
microsize
griseofulvin
tablets
and
one
ultramicrosize
griseofulvin
tablet
were
preliminarily
determined
in
18
liters
of
pH
7.2
phosphate
buffer
and
in
900
ml
of
40%
dimethylformamide
as
test
media.
Addition
of
dimethylformamide
af-
fected
the
dissolution
behavior
of
the
formulations.
The
products,
three
microsize
and
one
ultramicrosize,
were
selected
for
further
studies
on
the
bioavailability
in
humans
and
dissolution.
Significant
differences
among
the
formulations
were
found
in
serum
levels,
Cmax,
and
AUC47.5
hr,
but
not
in
AUG.
and
Cmax.
The
maximum
difference
of
C.
was
"40%.
The
ultramicrosize
product
showed
lower
C
max
and
serum
levels
at
earlier
sampling
times
than
two
microsize
products.
The
dissolution
rates
de-
termined
under
sink
and
nonsink
conditions
without
pretreatment
sig-
nificantly
correlated
with
the
serum
level
at
1
hr
but
not
with
the
other
in
vivo
parameters.
Only
the
dissolution
rate
determined
by
the
sink
method
with
pretreatment
with
a
small
quantity
of
water
(1.0
ml)
and
plastic
beads
significantly
correlated
with
serum
levels
at
3
and
5
hr,
C.,
and
A
UC47.5
hr
Keyphrases
Bioavailability-griseofulvin
from
tablets
in
humans,
correlation
with
dissolution
rate
Dissolution,
rates
-correlation
with
bioavailability
of
griseofulvin
from
tablets
in
humans
0
Griseofulvin-
bioavailability
from
tablets
in
humans,
correlation
with
dissolution
rate
The
absorption
of
a
poorly
water
soluble
drug
is
con-
sidered
to
be
dissolution
rate
limited.
Griseofulvin
used
as
an
antifungal
agent
is
a
practically
insoluble
compound.
The
in
vivo
availability
is
enhanced
by
increasing
the
dissolution
rate
by
means
of
reduction
of
particle
size
of
the
crystals
(1-6).
A
good
correlation
between
the
bio-
availability
and
dissolution
rate
was
found
for
griseofulvin
tablets
(7,
8).
However,
fine
particles
may
not
necessarily
produce
the
expected
dissolution
rate
and
bioavailability
due
to
their
aggregation
and
agglomeration
(4, 9).
A
mi-
crosize
griseofulvin
powder
has
been
widely
used
in
com-
mercial
tablets,
and
now
a
new
formulation,
an
ultrami-
crosize
griseofulvin
tablet,
can
be
obtained,
which
has
been
formulated
with
the
drug
dispersed
in
polyethylene
glycol
6000
(10-12)
and
has
been
shown
to
have
better
bioavail-
ability
than
microsize
products
(11,
12).
However,
recently
lower
absorption
of
the
drug
from
an
ultramicrosize
than
from
a
microsize
formulation
was
shown
(13).
The
present
investigation
was
undertaken
to
study
the
dissolution
rate
and
bioavailability
in
humans
for
com-
mercial
microsize
griseofulvin
tablets
and
an
ultrami-
crosize
griseofulvin
tablet,
and
to
clarify
the
relation
be-
tween
the
in
vitro
and
in
vivo
findings.
EXPERIMENTAL
Formulation
-Ten
commercial
microsize
griseofulvin
tablets
avail-
able
in
Japan
and
one
ultramicrosize
griseofulvin
tablet'
were
used
for
the
preliminary
dissolution
test.
Each
tablet
contained
125
mg
of
griseofulvin.
Based
on
preliminary
dissolution
data,
four
formulations,
1
Dorsey
Laboratories,
Division
of
Sandoz
Inc.
Table
I
-t
5
0
for
Griseofulvin
Tablets
a
Tablet
Method
Beaker°
Paddle°
1
(A)
13.8
8.3
2
25.4
24.7
3
102.6
9.1
4
80.3
10.6
5
(B)
8.1
11.0
6
73.9
37.0
7
85.0
6.0
8
(D)
211.0
38.8
9
(C)
107.8
9.0
10
86.8
29.5
11
66.0
°
Time
(t
50
)
in
minutes.
b
18
liters
of
pH
7.2
phosphate
buffer.
40%
Dimethyl-
formamide
(900
ml).
including
an
ultramicrosize
formulation,
were
selected
to
provide
a
broad
range
of
dissolution
rates
with
the
expectation
that
those
formulations
would
show
a
wide
variation
in
their
bioavailabilities.
Solubility
-The
solubility
of
griseofulvin
in
water
[pH
7.2
sodium
phosphate
buffer
(0.01
M)
and
40%
dimethylformamide]
was
determined
at
37°
spectrophotometrically
after
filtration
of
the
equilibrated
solution
of
griseofulvin
through
a
1.0
-am
membrane
filter.
Disintegration
Time
-Disintegration
times
of
the
griseofulvin
for-
mulations
were
determined
with
six
tablets
in
pH
7.2
sodium
phosphate
buffer
(0.01
M)
according
to
Japanese
Pharmacopeia
IX
specifications
(30
strokes/min,
37°).
The
time
when
there
were
no
particles
of
tablets
or
only
a
trace
amount
of
soft
residue
on
the
screen
was
selected
as
the
disintegration
time.
Dissolution
Rate
-The
dissolution
rate
of
the
drug
from
each dosage
form
was
determined
at
37°
in
pH
7.2
sodium
phosphate
buffer
(0.01
M)
unless
otherwise
specified.
Hydrochloric
acid
solution
(pH
1.2)
and
40%
dimethylformamide
were
also
used
as
solvents.
The
amount
of
the
drug
dissolved
was
monitored
spectrophotometrically
by
passing
the
solution
through
a
glass
filter
stick
(porosity
G-3)
to
a
fl
ow
cell
and
was
expressed
as
a
percentage
of
the
labeled
amount.
An
average
dissolution
rate
was
obtained
after
three
dissolution
runs.
The
dissolution
rate
was
repre-
sented
as
t5,
t30,
and
t
50
,
which
indicate,
respectively,
the
time
required
for
5,
30,
and
50%
of
the
drug
to
be
dissolved.
Polysorbate
80
2
and
diastase
(Japanese
Pharmacopeia
IX
grade)
were
used
to
investigate
their
effects
on
the
dissolution.
80
_J
60
0
40
(")
20
CC
Q.
100
MINUTES
200
Figure
1
-Dissolution
curves
of
griseofulvin
tablets
with
the
beaker
method
at
pH
7.2.
Key
(•)
Tablet
A;
(0)
Tablet
B;
A)
Tablet
C;
(❑)
Tablet
D;
(-
-
-)
dissolution
curves
of
the
other
tablets.
2
Tween
80,
Wako
Pure
Chemical
Industries
Ltd.,
Osaka,
Japan.
0022-3549/
82/
01000-1165$01.00/
0
Journal
of
Pharmaceutical
Sciences/
1165
©
1982,
American
Pharmaceutical
Association
Vol.
71,
No.
10,
October
1982
Table
II
-Dissolution
Rates
of
Griseofulvin
from
Four
Different
Tablets
with
Sink
Methods
In
Vitro
Test
t
30
,
min
t
50
,
min
Method
Condition
A
D
A
Beaker
8.7
4.8
32.8
63.3
14.1
6.5
80.6
157.5
0.1%
Polysorbate
80
9.1
4.8
27.4
44.3
13.6
6.6
70.2
105.3
0.014%
Diastase
8.0
4.8
34.0
31.5
14.5
6.3
76.0
71.7
848
rpm
6.5
4.1
28.7
48.0
9.5
5.6
65.3
119.0
pH
1.2
10.0
3.0
31.5
85.0
pH
1.2,
0.1%
Polysorbate
80
7.5
3.5
16.0
30.0
Basket
6.5
5.0
26.0
21.5
9.0
7.0 75.0
57.0
Method
I
3.1
1.7
8.3
2.7
6.1
2.5
26.7
4.6
Method
II
1.8
1.9
6.6
3.3
2.7
3.2
17.3
6.4
Sink
Method-Nonpretreatment
Method
-The
beaker
method
consisted
of
agitating
the
solution
with
a
three
-bladed
screw
-type
impeller
(5.0
-cm
i.d.,
stirring
rate:
512
rpm)
in
the
middle
of
the
solvent
(18
liters)
in
a
20
-liter
flat
-bottom
beaker
(29.0
-cm
i.d.).
The
basket
method
consisted
of
placing
a
tablet
in
a
cylindrical
basket
(1.7
-cm
i.d.
X
2.3
cm)
of
80
-mesh
stainless
steel
cloth.
This
basket
was
placed
in
a
basket
(Apparatus
1)
as
used
in
USP
XX.
The
basket
was
held
-10
cm
below
the
surface
of
the
solvent
and
rotated
at
400
rpm,
-7
cm
from
the
wall
of
the
beaker.
The
dissolution
medium
was
agitated
with
a
three
-bladed
impeller
as
described
for
the
18
-liter
beaker
method.
With
the
paddle
method
(USP
XX),
900
ml
of
40%
dimethylformamide
was
used
as
a
solvent
and
was
stirred
at
120
rpm.
Pretreatment
Method
-Method
I:
A
tablet
was
put
into
a
100
-ml
round
bottle
containing
1.0
ml
of
water
and
was
gently
shaken
for
1
min.
After
standing
for
5
min,
20
g
of
plastic
beads
(8
-mm
i.d.)
3
was
added
to
the
bottle.
The
bottle
was
fixed
at
an
angle
of
and
was
rotated
at
3.8
rpm
in
a
water
bath
(37°)
for
15
min.
The
contents
then
were
poured
into
18
liters
of
the
solvent
with
100
ml
of
water
for
washing
through
a
sieve
to
remove
the
plastic
beads.
Then
the
dissolution
rate
was
determined
according
to
the
18
-liter
beaker
method
procedure.
Method
II:
A
tablet
was
gently
shaken
in
20
ml
of
water
contained
in
a
50-m1
round
bottle
for
5
min,
and
then
27
g
of
the
plastic
beads
used
in
Method
I
was
added.
The
subsequent
procedure
was
the
same
as
method
I.
Nonsink
Method
-A
900
ml
volume
of
the
solvent
was
used
in
each
experiment.
The
following
procedures
were
used:
1.
Rotating
basket
method
(USP
XX):
The
basket
was
rotated
at
120
rpm.
2.
Paddle
method
(USP
XX):
The
stirring
rate
was
120
rpm.
3.
Beaker
method:
A
three
-bladed
screw
-type
impeller
(5.0
-cm
i.d.)
used
as
a
stirring
device
was
rotated
at
2.5
cm
from
the
bottom
of
the
beaker
which
was
used
for
the
paddle
method
at
120
rpm.
Bioavailability-Twelve
healthy
male
volunteers
who
participated
in
a
Latin
square
crossover
study
were
randomly
assigned
to
one
of
four
groups
of
equal
size.
The
subjects
ranged
in
age
from
22
to
51
years
(22,
22,
22, 22, 23,
24,
30,
32,
37, 37,
50,
and
51;
mean
31),
in
height
from
160
to
180
cm
(mean
170),
and
in
weight
from
53
to
69
kg
(mean
60).
Before
the
studies
the
participants
were
given
a
clinical
examination
to
ensure
they
were
healthy.
All
of
the
subjects
were
prohibited
from
taking
med-
icines
and
alcoholic
beverges
from
3
days
before
the
drug
administration
to
the
end
of
the
test.
Each
subject
took
a
test
tablet
orally
with
200
ml
of
water
at
9:00
am
after
fasting
overnight.
No
foods
or
liquids
were
al
-
-E
0.4
3.
f.1
0.2
E
(.1
62.5
125
62.5
125
DOSE,
mg
DOSE,
mg
Figure
2
-Relations
of
griseofulvin
dose
with
AUC,
and
peak
con-
centration
after
oral
administration
of
the
drug.
Key:
(--
-)
responses
of
the
individual
subjects;
(-)
the
average
response.
3
Sartorius-Membranfilter
GmbH.
lowed
until
4
hr
after
ingestion
of
the
tablet.
Blood
samples
(5
ml)
were
obtained
at
1,
3,
5,
8,
23.5,
33,
and
47.5
hr
after
administration
of
a
tablet
and
the
serum
samples
were
kept
frozen
at
-15°
until
assayed.
The
ex-
periments
were
repeated
every
2
weeks
according
to
the
dosage
schedule.
The
following
parameters
were
statistically
subjected
to
ANOVA,
and
the
differences
among
the
treatments
were
examined
by
Tukey's
test:
1.
Serum
levels
at
each
sampling
time.
2.
Peak
serum
level
(Cmax)•
3.
Time
to
reach
peak
serum
level
(t
n
.).
4.
Area
under
serum
concentration
-time
curves
from
zero
to
the
time
t
(AUC
t
)
calculated
by
the
trapezoidal
rule
The
AUG.
was
calculated
by
the
method
of
Wagner
(14).
Dose-Bioavailability
Relation
-A
single
tablet
and
one-half
tablet
of
Formulation
B,
corresponding
to
125
and
62.5
mg
of
griseofulvin,
re-
spectively,
were
given
orally
to
three
subjects
according
to
a
crossover
design.
The
sampling
times
were
the
same
as
described
in
the
bioavail-
ability
study.
Assay
-The
serum
griseofulvin
concentration
was
determined
by
the
GLC
method
described
previously
(15).
To
a
0.5
-ml
aliquot
of
serum
were
added
0.5
ml
of
saturated
sodium
chloride
solution
and
5
ml
of
ether.
After
shaking
for
10
min,
a
4
-ml
aliquot
of
the
ether
phase
was
taken
and
evaporated
to
dryness
in
vacuo.
The
residue
was
dissolved
with
0.4
ml
of
benzene
containing
1.1
ug
of
clothiapine
as
internal
standard,
and
5
µI
of
the
solution
was
taken
for
the
GLC
assay.
GLC
conditions:
column,
glass
column
(3
-mm
i.d.
X
70
cm)
packed
with
5%
OV-17
on
60-80
mesh
Chromosorb
W;
injection
temperature,
260°;
detector
temperature,
260°.
RESULTS
Dissolution
Rate
-The
solubilities
of
griseofulvin
at
37°
in
water,
pH
7.2
buffer,
and
40%
dimethylformamide
were
31.8, 32.2,
and
482.1
µg/ml,
respectively;
therefore,
the
preliminary
dissolution
tests
were
carried
out
with
18
liters
of
the
aqueous
medium
(18
-liter
beaker
method)
and
900
ml
of
40%
dimethylformamide
(paddle
method)
in
order
to
get
a
sink
condition.
Figure
1
shows
the
dissolution
profiles
with
the
18
-liter
beaker
method.
Relatively
large
differences
were
observed
among
the
microsize
griseofulvin
formulations.
The
ultramicrosize
formulation
(A)
showed
rapid
dissolution.
Table
I
lists
t
5
0
in
18
liters
of
aqueous
solvent
and
in
40%
dimethylformamide.
Addition
of
the
organic
solvent
to
aqueous
medium
altered
the
dissolution
behavior
of
griseofulvin;
how-
ever,
the
use
of
a
partially
alcoholic
medium
was
described
for
the
dis-
solution
test
of
water
insoluble
drugs
(16).
On
the
basis
of
the
in
vitro
data,
four
formulations
(A,
B,
C,
D),
including
an
ultramicrosize
formu-
lation,
were
selected
for
further
investigations
concerning
their
dissolu-
tion
rates
and
bioavailabilities.
Table
II
lists
the
dissolution
rates
(t30
and
too)
of
these
formulations
determined
under
the
sink
condition.
With
the
18
-liter
beaker
method
Table
III
-Dissolution
Rate
of
Griseofulvin
from
Tablets
with
Nonsink
Methods
a
Method
Polysorbate
80
t
5
,
min
A
Rotating
basket
0
6.4
5.0
21.0 24.0
0.1%
2.7
3.0
15.0
18.5
Paddle
0
4.5
4.7
9.9
18.6
0.1%
4.2
4.0
9.0
11.0
Beaker
0
3.4
3.6
6.5
14.8
0.1%
3.0
2.8
4.6
8.2
°
At
pH
7.2.
1166
/
Journal
of
Pharmaceutical
Sciences
Vol.
71,
No.
10,
October
1982
Table
IV
-Serum
Levels,
Cmax,
-nuop
and
A
UC
After
Oral
Administration
of
125
-mg
Griseofulvin
Tablets
In
Vivo
Parameter
Time,
hr
Formulation
a
A
B
C
D
Serum
levels,
µg/ml
1.0
0.250
±
0.058
0.331
±
0.048
0.154
±
0.025
0.126
±
0.023
3.0
0.446
±
0.065
0.593
±
0.052
0.340
±
0.041
0.462
±
0.074
5.0
0.408
±
0.036
0.576
±
0.045
0.365
±
0.038
0.486
±
0.063
8.0
0.326
±
0.035
0.462
±
0.039
0.320
±
0.035
0.439
±
0.041
23.5
0.234
±
0.021
0.199
±
0.022
0.192
±
0.019
0.243
±
0.025
33.0
0.152
±
0.018
0.109
±
0.015
0.146
±
0.020
0.133
±
0.024
47.5
0.071
±
0.011
0.046
±
0.006
0.072
±
0.010
0.062
±
0.011
C.,
µg/m1
0.502
±
0.049
0.660
±
0.046
0.395
f
0.036
0.546
±
0.070
tmax,
hr
5.3
±
1.7
3.6
±
0.3
6.0
±
1.7
7.9
±
2.1
A
UC47.5,
a
ug
hr/ml
10.57
±
0.86
11.53
±
0.76
9.46
±
0.73
11.47
f
0.67
A
UC.,
µg
hr/ml
11.83
±
0.95
12.30
±
0.86
11.27
±
0.91
12.76
±
0.81
Tukey's'
ANOVA
5
Test
p
<
0.01
B>A>C>D
p
<
0.01
B>D>A>C
p
<
0.01
B>D>A>C
p
<
0.01
B>D>A>C
p
<
0.01
D>A>B>C
p
<
0.05
A>C>D>B
p
<
0.05
C>A>D>B
p
<
0.01
B>D>A>C
NS
p
<
0.01
B>D>A>C
NS
The
figures
indicate
means
±
standard
errors.
b
NS:
not
significant.
Formulations
underlined
by
a
common
line
did
not
differ
significantly;
p
<
0.05.
0.6
E
a)
2
w
0.4
w
2
D
0.2
CC
w
10
20
30
40
HOURS
Figure
3
-Mean
serum
griseofulvin
concentration
after
oral
admin-
istration
of
125
-mg
griseofulvin
tablets
to
humans.
Key:
(S)
Tablet
A;
(0)
Tablet
B;
A)
Tablet
C;
(o)
Tablet
D.
The
vertical
lines
indicate
standard
errors.
at
either
pH
7.2
or
1.2
without
additives,
the
drug
rapidly
dissolved
from
Formulations
A
and
B
due
to
their
rapid
disintegration
into
fine
particles.
It
slowly
dissolved,
however,
from
Formulations
C
and
D
due
to
the
large
particles
resulting
from
disintegration.
Polysorbate
80
and
diastase,
which
was
used
to
investigate
the
digestive
action
on
the
starch
widely
employed
in
tablets
as
the
disintegrant
or
excipient,
enhanced
the
dissolution
from
Formulation
D,
and
in
the
presence
of
diastase
its
dissolution
rate
was
faster
than
that
from
Formulation
C.
This
can
be
attributed
to
a
wetting
action
of
the
surfactant
and
a
digestive
one
of
the
diastase.
The
dissolu-
tion
rates
were
not
enhanced
by
an
increase
of
agitation
intensity,
but
the
dissolution
from
Formulation
D
was
specifically
facilitated
with
the
basket
method
in
which
its
deaggregation
into
fine
particles
seemed
to
be
promoted
due
to
being
rubbed
against
the
sieve
wall
of
the
basket
as
it
was
rotated.
This
suggests
that
deaggregation,
especially
for
Formu-
lation
D,
must
be
more
accelerated
by
mechanical
forces
than
by
the
fl
ow
of
the
dissolution
medium.
This
was
further
ascertained
by
the
pre-
treatment
methods
in
which
the
dissolution
of
the
drug,
especially
from
Formulation
D,
was
greatly
enhanced.
This
is
probably
due
to
the
strong
deaggregation
effects
of
the
plastic
beads.
The
volume
of
water
used
in
the
pretreatment
affected
the
dissolution,
especially
from
the
ultramicrosize
formulation
(A).
It
formed
a
pastelike
Table
V
-Power
Analysis
with
a
=
0.05
and
13
=
0.2
Parameter
Time,
hr
Subjects
for
20%
difference
Minimum
Detectable
Difference,
%
Serum
level
1
76
53
3
28
30
5
24
28
8
20
26
23.5
20
25
33
44
39
47.5
60
47
Cmax
20
24
t
max
216
90
A
UC
47.5
12
17
AUC..
16
21
-
0.4
>
E
0.3
Lu
2u
;
0.2
cr
F
_
co
LLI
<
0.1
5
10
50
100
t,„,
min
Figure
4
-Log
-log
plots
of
serum
level
of
griseofulvin
at
1
hr
against
t
30
determined
by
the
beaker
method
at
pH
7.2
without
an
additive
(0)
and
in
the
presence
of
polysorbate
80
(s).
agglomerate
and
showed
slower
dissolution
when
a
small
quantity
of
water
(1.0
ml)
was
used
(Method
I)
than
when
20
ml
of
water
was
em-
ployed.
In
the
latter
case
the
formulation
rapidly
disintegrated
into
fine
particles.
The
initial
dissolution
rates
(t
5
)
determined
under
the
nonsink
con-
dition
are
shown
in
Table
III.
The
results
were
similar
to
those
deter-
mined
with
nonpretreatment
methods
under
the
sink
condition,
which
suggests
that
by
using
the
initial
dissolution
rates,
the
dissolution
under
the
sink
condition
can
be
predicted.
Disintegration
Time
-The
mean
disintegration
times
of
Formula-
tions
A,
B,
C,
and
D
were
9.3,
12.5,
0.4,
and
2.4
min,
respectively.
Contrary
to
the
disintegration
findings,
the
slow
dissolution
of
the
drug
from
Formulations
C
and
D
led
to
the
conclusion
that
the
resultant
particulate
state
after
disintegration
is
more
important
for
the
dissolution
of
griseofulvin
than
the
disintegration
time.
Bioavailability-The
relations
of
griseofulvin
dose
with
AUG,.
and
C
max
are
shown
in
Fig.
2.
The
nearly
linear
relationships
of
dose
-A
UC,.
(r
=
0.829)
and
dose-Cmax
(r
=
0.944)
indicate
that
linear
pharmacoki-
netics
can
be
applied
to
the
serum
level
of
griseofulvin
within
the
dose
ranges
studied.
Figure
3
shows
the
mean
serum
concentration
-time
curves
of
griseo-
fulvin
in
humans
after
oral
administration
of
four
formulations.
Table
IV
lists
the
mean
values
of
each
parameter
for
these
formulations.
The
rank
order
of
serum
level
at
1
hr
was
B>A>C>D,
but
the
serum
levels
during
3-8
hr,
C.
and
AUC47.5,
were
B>D>A>C.
The
lowest
peak
0.7
0.6
cs,
a.
0.5
w
-I
0.4
2
CC
cn
0.3
12
1,
113
(.
0
3
9
2
4
6
8
t,,
min
Figure
5
-Log
-log
plots
of
serum
level
at
3
hr
(0),
C,,„
x
(s),
and
AUC47.5
(E)
against
t
30
determined
by
Method
I.
Journal
of
Pharmaceutical
Sciences
/
1167
Vol.
71,
No.
10,
October
1982
Table
VI
-Correlation
Coefficients
Between
In
Vivo
Parameters
(X)
and
t
30
(Y)
Determined
by
Sink
Methods
In
Vitro
Test
X
-
Y
-1
log
(X)
-
log
(Y)
Serum
Level
Cmax
A
UC
47.5
Serum
Level
C.
AUG47.5
Method
Condition
1
hr
3hr
5hr
1
hr
3
hr
5
hr
Beaker
0.995°
0.809
0.655
0.727
0.430
-0.998°
-0.608
-0.399
-0.508
-0.229
Polysorbate
80
0.990a
0.820
0.679
0.471
0.441
-0.999°
-0.650
-0.453
-0.554
-0.276
Diastase
0.990°
0.840
0.680
0.763
0.489
-0.975
6
-0.771
-0.564
-0.689
-0.460
848
rpm
0.998°
0.792
0.617
0.708
0.420
-0.995°
-0.624
-0.403
-0.526
-0.259
pH
1.2
0.942
0.848
0.765
0.782
0.482
-0.991°
-0.611
-0.442
-0.514
-0.217
pH
1.2
Polysorbate
80
0.976
6
0.813
0.696 0.736
0.423
-0.992°
-0.613
-0.442
-0.516
-0.220
Basket
0.978
0.796
0.598
0.716 0.466
-0.951
-0.769
-0.545
-0.690
-0.483
Method
I
0.719
0.998°
0.965
b
0.997
0.896
-0.559
-0.982
6
-0.917
-0.987
b
-0.953
6
Method
II
0.814
0.742
0.513
0.686
0.570
-0.742
-0.824
-0.601
-0.780
-0.697
°
p
<0.01.
bp<
0.05.
Table
VII
-Correlation
Coefficients
Between
In
Vivo
Parameters
(X)
and
t
5
(Y)
Determined
by
Nonsink
Methods
In
Vitro
Test
X
-
Y
-1
log
(X)
-
log
(Y)
Serum
Level
A
UC
47.5
Serum
Level
Cmax
A
UC47.5
Method
Additive
1
hr
3hr
5hr
1
hr
3hr
5hr
Rotating
basket
none
0.986
°
0.758
0.554
0.670
0.402
-0.954
0
-0.627
-0.387 -0.527
-0.260
Polysorbate
80
0.895
0.591
0.329
0.498
0.380
-0.947
-0.591
-0.325
-0.494
-0.269
Paddle
none
0.912
0.503
0.245
0.394
0.120
-0.939
-0.387
-0.135
-0.274
0.002
Polysorbate
80
0.954
a
0.641
0.402
0.544
0.282
-0.974°
-0.584
-0.335
-0.484
-0.235
Beaker
none
0.892
0.428
0.170
0.313
0.220
-0.907
-0.287
-0.044
-0.169
0.118
Polysorbate
80
0.937
0.492
0.262
0.378
0.553
-0.934
-0.345
-0.119 -0.229
0.068
p
<
0.05.
concentration
(Formulation
C)
was
only
60%
of
the
highest
one
(For-
mulation
B).
Significant
differences
were
found
among
the
formulations;
however,
there
were
no
significant
differences
in
A
UC.,
which
suggests
that
the
four
formulations
were
equivalent
in
the
extent
of
bioavail-
ability.
The
mean
peak
times
did not
show
significant
differences,
though
the
maximum
difference
(3.6-7.9
hr)
was
relatively
large.
This
can
be
at-
tributed
to
the
large
variations
in
the
parameter
which
will
be
due
in
some
degree
to
a
few
peak
times
of
23.5
hr,
which
occurred
with
all
formulations
except
B.
It
may
have
been
artificially
caused
by
the
long
interval
of
the
sampling
time
from
8
to
23.5
hr.
The
power
analysis
was
employed
to
estimate
the
minimum
detectable
difference
between
the
formulations
which
would
be
statistically
significant
(a
=
0.05,(
=
0.2)
or
the
number
of
subjects
required
for
a
20%
difference
to
be
significant
(Table
V).
The
results
of
this
analysis
indicate
considerably
low
detectability
of
tmax
compared
with
the
other
parameters;
only
90%
of
the
minimum
difference
could
be
detected
in
this
study,
and
>200
subjects
would
have
been
re-
quired
for
a
20%
difference
to
be
significant.
The
parameters
of
A
UC
47.5
and
AUC
-
provided
better
detectabilities,
which
must
be
due
to
the
small
variabilities
in
the
parameters.
The
ultramicrosize
formulation
(A)
did
not
show
good
bioavailability,
which
is
coincident
with
previous
results
(13),
though
absorption
of
the
drug
from
the
formulation
was
reported
to
be
approximately
twice
that
from
microsize
formulations
(11,
12).
Correlation
Between
Dissolution
Rate
and
Bioavailability-The
correlation
coefficients
between
the
in
vivo
and
in
vitro
parameters
(t30
and
t
5
determined
by
sink
and
nonsink
methods,
respectively)
are
shown
in
Tables
VI
and
VII.
The
t30
determined
by
sink
methods
without
pre-
treatment
correlated
significantly
with
the
serum
level
at
1
hr
in
nor-
mal
-reciprocal
and
log
-log
regressions
but
not
with
the
other
in
vivo
parameters.
Similar
relations
were
found
between
t
5
determined
by
nonsink
methods
and
the
in
vivo
parameters.
The
t3
0
determined
by
Method
I
with
pretreatment
with
1.0
ml
of
water
was
significantly
cor-
related
with
serum
levels
at
3-5
hr
and
C.
in
normal
-reciprocal
re-
gression
and
even
with
AUG
4
7.5
in
log
-log
regression.
Those
determined
by
Method
II
using
20
ml
of
water,
however,
did not
correlate
significantly
with
any
of
the
in
vivo
parameters.
Figures
4
and
5
illustrate
the
log
-log
regression
lines
that
were
significantly
correlated
between
in
vivo
and
in
vitro
findings.
DISCUSSION
The
in
vivo
and
in
vitro
findings
for
griseofulvin
formulations
suggest
the
complex
dissolution
behavior
of
the
drug
from
its
dosage
forms
in
the
GI
tract
in
humans.
The
serum
levels
1
hr
after
administration
of
the
formulation
correlated
well
with
the
dissolution
rates
determined
under
sink
and
nonsink
conditions
without
pretreatment,
which
leads
to
the
conclusion
that
initial
absorption
is
mainly
controlled
by
relatively
simple
dissolution
behavior
as
shown
artificially
in
the
dissolution
tests
without
pretreatment.
However,
the
serum
levels
during
3-5
hr
and
C
max
did not
correlate
well
with
these
dissolution
rates,
which
is
probably
due
to
the
unexpectedly
low
values
for
Formulation
A
and
high
values
for
D
for
the
in
vivo
parameters.
It
may
be
considered
that
for
a
drug
having
as
low
solubility
in
water
or
biological
fluids
as
griseofulvin,
the
in
vivo
disin-
tegration
and
dissolution
must
be
affected
more
by
physiological
factors
in
the
GI
tract
than
those
of
a
drug
having
high
solubility,
because
the
former
drug
must
stay
in
the
GI
tract
longer.
The
in
vitro
pretreatment,
which
decreases
the
sizes
of
aggregates
or
particles
available
from
dis-
solution
with
plastic
beads,
markedly
increased
the
dissolution,
especially
from
Formulation
D.
The
resulting
dissolution
rates
highly
correlated
with
the
serum
levels
at
3
and
5
hr,
C
max
,
and
A
UC
47.5.
These
findings
suggest
a
strong
deaggregation
action
on
the
aggregates
and/or
particles
while
the
drug
having
low
solubility
stays
in
the
GI
tract.
A
previous
report
on
sugar-coated
tablets
of
chloramphenicol
also
suggested
the
violent
destructive
force
on
the
formulations
in
vivo
(17).
Therefore,
the
absorption
of
a
drug
having
low
solubility
from
its
dosage
forms
seems
to
be
correlated
more
with
the
dissolution
rates
determined
under
the
conditions
of
vigorous
agitation
and
destructive
intensities
than
under
mild
conditions,
though
the
absorption
of
a
drug
having
high
sol-
ubility
such
as
aspirin
may
be
correlated
more
with
the
dissolution
rates
determined
under
the
latter
conditions
(18,
19).
Formulation
A
showed
rapid
dissolution
under
sink
and
nonsink
conditions
without
pretreatment,
which
is
probably
due
to
the
original
ultramicrosize
particulate
state
of
the
drug.
But
the
rapid
dissolution
did
not
contribute
to
the
in
vivo
absorption.
Formulation
A
formed
a
pastelike
agglomerate
when
treated
with
a
small
quantity
of
water
(1.0
ml).
Such
an
agglomerate
may
be
also
produced
in
the
GI
tract
after
rapid
gastric
emptying
and
rapid
absorption
of
water
coadministered
with
the
formulation
and
may
be
responsible
for
slower
dissolution
and
the
re-
sulting
lower
absorption
of
the
drug.
REFERENCES
(1)
R.
M.
Atkinson,
C.
Bedford,
K.
J.
Child,
and
E.
G.
Tomich,
Na-
ture
(London),
193,
588
(1962).
(2)
P.
Kabasakalian,
M.
Katz,
B.
Rosenkrantz,
and
E.
Townley,
J.
Pharm.
Sci.,
59,
595
(1970).
(3)
M.
Kraml,
J.
Dubac,
and
D.
Beall,
Can.
J.
Biochem.
Physiol.,
40,
1449
(1962).
(4)
R.
M.
Atkinson,
C.
Bedford,
K.
J.
Child,
and
E.
G.
Tomich,
An-
tibiot.
12,
232
(1962).
1168
/
Journal
of
Pharmaceutical
Sciences
Vol.
71,
No.
10,
October
1982
(5)
M.
Kraml,
J.
Dubuc,
and
D.
Dvornik,
Arch.
Dermatol.,
87,
179
(1963).
(6)
J.
R.
Marvel,
D.
A.
Schlichting,
D.
Denton,
E.
J.
Levy,
and
M. M.
Cahn,
J.
Invest.
Dermatol.,
42,
197
(1964).
(7)
B.
Katchen
and
S.
Symchowicz,
J.
Pharm.
Sci.,
56,
1108
(1967).
(8)
S.
Symchowicz
and
B.
Katchen,
ibid.,
57,
1383
(1968).
(9)
S.
Lin,
J.
Menig,
and
L.
Lachman,
ibid.,
57,
2143
(1968).
(10)
W.
L.
Chiou
and
S.
Riegelman,
ibid.,
60,
1376 (1971).
(11)
W.
E.
Barrett,
Curr.
Ther.
Res.,
18,
501
(1975).
(12)
Ibid.,
18,
491
(1975).
(13)
A.
B.
Straughn,
M.
C.
Meyer,
G.
Raghow,
and
K.
Rotenberg,
J.
Pharmacokinet.
Biopharm.,
8,
347
(1980).
(14)
J.
G.
Wagner,
"Fundamentals
of
Clinical
Pharmacokinetics,"
Drug
Intelligence
Publications,
Hamilton,
Ill.,
1975,
p.
344.
(15)
H.
J.
Schwarz,
B.
A.
Waldman,
and
V.
Madrid,
J.
Pharm.
Sci.,
65,
370
(1976).
(16)
"USP
Pharmacopeial
Forum,"
3,
22-25
(1977).
(17)
H.
Ogata,
T.
Shibazaki,
T.
Inoue,
and
A.
Ejima,
J.
Pharm.
Sci.,
68,
712
(1979).
(18)
G.
Levy,
ibid.,
50,
388
(1961).
(19)
Ibid.,
52,
1039
(1963).
Bioavailability
of
Griseofulvin
from
Tablets
in
Beagle
Dogs
and
Correlation
with
Dissolution
Rate
and
Bioavailability
in
Humans
NOBUO
AOYAGI
*x,
HIROYASU
OGATA
*,
NAHOKO
KANIWA
*,
MASANOBU
KOIBUCHI
*t,
TOSHIO
SHIBAZAKI
*,
AKIRA
EJIMA
*,
NORIYASU
TAMAKI
t,
HIDETAKA
KAMIMURA
§,
YOSHIO
KATOUGI
t,
and
YUKIO
OMI
Received
July
31,
1981,
from
the
*Division
of
Drugs,
National
Institute
of
Hygienic
Sciences,
18-1,
Kamiyoga
1-chome
,
Setagaya-ku,
Tokyo
158,
Japan;
the
t
Yaizu
Plant,
Yamanouchi
Pharmaceutical
Co.
Ltd.,
Ozumi-180,
Yaizu-shi,
Shizuoka
-ken
425,
Japan;
and
the
§Institute
of
Research
and
Development,
Yamanouchi
Pharmaceutical
Co.
Ltd.,
1-8,
Azusawa
1-chome,
Itabashi-ku,
Tokyo
174,
Japan.
/Deceased.
Accepted
for
publication
December
30,
1981.
Abstract
0
The
bioavailability
of
four
griseofulvin
tablets
in
beagle
dogs,
including
an
ultramicrosize
tablet
used
previously
in
a
human
bioavail-
ability
study,
was
investigated
on
the
basis
of
the
plasma
6-demethyl-
griseofulvin
concentration.
The
relations
with
the
in
vivo
findings
in
humans
and
the
in
vitro
dissolution
rates
also
were
examined.
Contrary
to
the
lower
bioavailability
of
the
ultramicrosize
formulation
in
humans,
it
provided
the
best
bioavailability
in
beagles.
The
microsize
griseofulvin
formulations
showed
similar
in
vivo
results
to
those
in
humans.
Poor
correlation
of
in
vivo
parameters
between
humans
and
beagles
was
at-
tributed
to
the
discrepancy
of
the
availability
of
the
ultramicrosize
for-
mulation
between
the
two
species.
The
dissolution
rates
determined
by
the
pretreatment
method
using
plastic
beads
were
correlated
more
with
the
in
vivo
findings
than
those
determined
by
the
other
methods.
Beagles
were
a
useful
animal
model
for
bioavailability
studies
of
certain
griseof-
ulvin
formulations
but
not
ultramicrosize
ones.
Keyphrases
0
Bioavailability—griseofulvin
from
tablets
in
beagle
dogs,
correlation
with
dissolution
rate
and
bioavailability
in
humans
0
Dis-
solution
rates—bioavailability
of
griseofulvin
from
tablets
in
beagle
dogs,
bioavailability
in
humans
0
Griseofulvin—bioavailability
from
tablets
in
beagle
dogs,
dissolution
rate
and
bioavailability
in
humans
The
bioavailabilities
for
four
lots
of
griseofulvin
tablets
in
humans
have
been
reported
previously,
and
the
relations
with
in
vitro
dissolution
rates
have
been
discussed
(1).
Beagle
dogs
are
often
used
as
an
animal
model
for
bio-
availability
studies,
but
their
suitability
has
not
been
clarified
sufficiently.
A
good
relation
of
penicillin
bio-
availability
between
humans
and
dogs
was
reported
(2).
Previous
studies
on
bioavailability
of
diazepam
formula-
tions
in
humans
and
beagles
revealed
no
good
relations
between
the
results
from
both
species.
The
discrepancy
was
considered
to
be
due
to
the
differences
of
physiological
states
of
the
GI
tract,
especially
of
gastric
emptying
rate
and
GI
transition
time
(3).
In
the
present
study
the
bioavailability
of
griseofulvin
from
tablets
in
beagles
was
studied,
and
the
relations
with
in
vivo
results
in
humans
and
in
vitro
dissolution
rates
were
investigated.
EXPERIMENTAL
Formulations
—Four
lots
of
tablets
containing
125
mg
of
griseofulvin
employed
in
the
human
bioavailability
study
(1)
were
used.
One
formu-
30
10
I
le
/
/
/
/
/0
--o
125
250
DOSE,
mg
Figure
1
—Relation
between
griseofulvin
dose
and
AUC.,
of
6-dem-
ethylgriseofulvin. Key:
(-
-
-)
responses
of
the
individual
dogs;
(—)
average
response.
0022-3549/
82/
01000-1169$01.00/
0
©
1982,
American
Pharmaceutical
Association
Journal
of
Pharmaceutical
Sciences
/
1169
Vol.
71,
No.
10,
October
1982