Simultaneous Diffusion and Metabolism of Betamethasone 17-Valerate in the

Living Skin Equivalent

JOHNADEMOLA,AND HOWARD
KIYOSHIKUBOTA~, I. MAIBACH~
Received November 30, 1994, from the Department of Dermatology, University of California, San Francisco, School of Medicine,
Box 0989, Surge 110, San Francisco, CA 941434989, Accepted for publication September 27, 1995@. §Present address: Drug
Safety Research Unit, Bursledon Hall, Southampton SO31 1AA, England.

Abstract 0 Simultaneous diffusion and metabolism of betamethasone

17-valeratewas studied using betamethasone 17-valerate,betarnethasone
21-valerate, and betamethasone as permeants. These corticosteroids
were suspended in silicone adhesive and applied to an artificial living
skin equivalent (LSE) for 72 h. When betamethasone was applied, no
metabolites were detected in the receptor medium. Conversely, with
betarnethasone 21walerate application, only betamethasone but no
betamethasone21walerate was detected in the receptor medium indicating
the metabolism of the latter by skin esterases. When tested with the
theory for simultaneous diffusion and metabolism, the result is consistent
with the enzyme rate constant in the LSE homogenate measured in a
previous study. When betamethasone 17-valerate was applied to the
LSE, more than half of the total amount of corticosteroids detected in
the receptor medium was unchanged, consistent with the previously
reported chemical (as opposed to enzymatic) degradation half-life of about
8 h. This result also indicated that vety little metabolism of betamethasone
17-valerate occurred in the skin.
Theoretical Section
Fraction of the Influx Metabolized in Percutaneous
Permeation-When metabolism occurs inside the skin during
percutaneous permeation of a compound, the kinetic behavior
is described as follows
(1)
where h is the thickness of the skin layer in which the
compound is metabolized (e.g., viable epidermis), C is the
concentration, D is the diffusion coefficient, k is the first-order
enzyme rate constant, and x = 0 and x = h designate the
donor- and receptor-side boundaries of the layer, respectively.
At steady state, where aCBt = 0, eq 1 reduces to D 3zCf;tx2=
kC, giving the following general solution10-12
C(x)= C, cosh[axl + C , sinh[ajcl (2)

where u2 = k f D or a = (kfD)lIz. The constants C1 and C2 are

specified according to the boundary conditions. Flux per unit
Introduction area, J , a t steady-state is given as
Percutaneous absorption comprises ten or more pr0cesses.l
A permeant diffuses through stratum corneum, viable epi-
ac
J(x)= -D -= -Da(C, sinh[axl + C, cosh[axl) (3)
dermis, and dermis and may be metabolized in the several ibc
skin layers. Viable epidermis has been thought to be a major If no metabolism occurs, the flux a t steady-state J(x) is
site of m e t a b o l i ~ m . ~Several
-~ models have been developed constant throughout the layer. Conversely, the difference of
to study skin metabolism such as the “dynamic” culture the flux at the donor-side boundary, J(O), and that a t the
~ y s t e m ,skin-flap
~ technique,6 and living skin equivalent receptor-side boundary, J(h),accounts for metabolism inside
(LSE).7ss The LSE, constructed from cultured human epider- the layer. The fraction of the influx metabolized or “forward
mal cells and fibroblasts, consists of equivalents for stratum c o n s ~ m p t i o n ”f m, ~, ~
is given as
corneum, viable epidermis, and dermis. The LSE has both
advantages (e.g., reproducibility of experiments) and disad-
vantages (e.g., lack of skin appendages). In this study, c,
J ( o ) - J ( h )= 1 - cosh[ah] - - sinh[ah] (4)
simultaneous diffusion and metabolism of betamethasone 17- fm = J(0) c
2
valerate was studied using betamethasone 17-valerate, be-
tamethasone 21-valerate, and betamethasone as permeants. From eq 2, C1 and C2 may be specified by the concentrations
Corticosteroids were applied to the LSE, and the receptor fluid a t two boundaries, C(0) and C ( h )such that
(culture medium) is examined for 72 h to estimate degradation
and metabolism inside the skin. A silicone adhesive was used
as a vehicle. Betamethasone 17-valerate is converted to
betamethasone 21-valerate and then betamethasone. Two
first-order degradation rate constants of this serial conversion Equations 4 and 5 give
of betamethasone 17-valerate to betamethasone have been
estimated using the LSE homogenate in a previous study.g C(0) sinh2[uhl
(6)
The results of the current study are compatible with the rate
constants examined previo~sly.~ Particular attention is paid
f m = 1- C O S ~ [ U ~+
C ]( 0 )cosh[ahI - C(h)
to the finding that no betamethasone 21-valerate but only Provided that C(0) is fixed, the value of f m is the smallest
betmethasone is found in the receptor fluid when betametha- when C(h)= 0. The smallest value of f m , fm0, is given as
sone 21-valerate is applied to the LSE. This observation is
tested with a mathematical model for simultaneous diffusion
and metabolisrnl0-l2described in the theoretical section of the
paper.
~~ If the flux from the skin to the receptor fluid is constant
@ Abstract published in Advance ACS Abstracts, November 1, 1995. and chemical degradation occurs in the receptor fluid, the

1478 / Journal of Pharmaceutical Sciences 0022-3549/95/3184-1478$09.00/0 0 1995, American Chemical Society and
Vol. 84, No. 12, December 1995 American Pharmaceutical Association
fraction of the compound degradated in the receptor fluid a t
steady state, fdg, may be estimated as
In eq 8, kdg is the first-order degradation rate constant and
t is time when the receptor fluid is collected in the static cell
after it is replenished.
When both metabolism within the skin and chemical
degradation in the receptor fluid occur subsequently, the
fraction of the influx metabolized or degradated, f m t d g , may
be approximated as
where f, and fdg are given in eqs 6 and 8, respectively. The
minimum value of fmtdg, fm+d@, may be estimated as
(10)
where fmo is given in eq 7. Note that the fraction of the
compound found to have been metabolized or degradated in
the receptor fluid sample should be larger than the possible
smallest value fm+dgO because fm (eq 6) is larger than f , ~
(eq
7) when C(h) > 0 and metabolism or degradation may also
occur in other layers of skin (e.g., stratum corneum and
dermis).
Estimation of the First-Order Enzyme Rate Con-
stant-When a piece of viable skin is homogenized and
suspended in the receptor medium, the first-order enzyme rate
constant in the receptor medium, k,,, may be estimated as
in eq 11below, if the chemical degradation and metabolism
by the skin enzyme are assumed to occur linearly in the
receptor medium
(11)
where kcmf and kcm- are the first-order degradation rate
constants in the receptor medium with and without skin
homogenate, respectively. Presuming that the enzyme activ-
ity in the viable skin is the same as that in the receptor
medium and that the enzyme rate constant is proportional to
the amount of the enzyme molecules per unit volume, the first-
order enzyme rate constant in the skin layer, k, may be
estimated as
(12)
where Vcmis the volume of the receptor medium in which skin
homogenate is suspended and V,l is the volume of the skin
layer where the compound is metabolized.
Experimental Section
Living Skin Equivalent-Living skin equivalent (LSE), TEST-
SKIN, with a standard six-well plate, was a generous gift from
Organogenseis Inc. (Canton, MA). Methods for preparing the LSE
and some relevant information have been published
Corticosteroid Patch as a D o n o r Device-Betamethaosne 17-
valerate and betamethasone 21-valerate were generous gifts from
Schering Plough (Bloomfield, NJ). Betamethasone was purchased
from Sigma (St. Louis, MO). Medical grade pressure sensitive silicone
adhesive, BIOPSA, in trichlorotrifluoroethane as solvent carrier, was
purchased from Dow Corning (Midland, MI). Corticosteroids were
dissolved in the mixture of dichloromethane-BIOPSA in solvent
carrier [64:36, (w/w)l. The drug concentrations in the mixtures were
0.15% (w/v) for betamethasone 17-valerate and betamethasone and
0 24 48 72
Time (h)
Figure 1-The average (n = 6 each) cumulative amount of corticosteroids
eliminated to the receptor medium till 72 h after the application of betarnethasone
17-valerate (0),betamethasone 21-valerate (O), and betamethasone (A) to the
LSE. Total amounts of betamethasoneequivalents of three corticosteroids detected
in the receptor medium are plotted.
0.075% (w/v) for betamethasone 21-valerate. Three milliliters of each
solution was applied on Scotchpak (Type 1009 backing membrane,
3M, St. Paul, MN) and then developed into a n approximately 6 x 20
cm area by a casting device (2.5 in. in width and 0.010 in. in
thickness, designed and made by Stanford Research International
(Menlo Park, CA)). All of the adhesive containing the three corti-
costeroids turned cloudy after the solvent evaporated overnight, which
suggests a suspension was obtained. Low adhesion polyester film,
Scotchpak (Type 1022 release liner, 3M, St. Paul, MN), was then
placed on BIOPSA, and the patches were stored at room temperature.
The drug load was 38 ,ug/cm2 for betamethasone, 38 ,ug/cm2 for
betamethasone 17-valerate (31 ,ug/cm2as betamethasone equivalent),
and 19 ,ug/cm2 for betamethasone 21-valerate (16 ,ug/cm2 as be-
tamethasone equivalent) since the mixture was developed at a
thickness of 0.010 in. (0.0254 cm).
SimultaneousDiffusion and Metabolism of Betamethasone
17-Valerate, Betamethasone 21-Valerate, and Betamethasone-A
patch containing one of three corticosteroids was cut into round pieces
(0.5 in. in diameter with a n area of 1.27 cm2) and applied centrally
onto the LSE ( n = 6 each). Skin models with the patches were
incubated in a tissue incubator at 37 "C for 72 h. The receptor
medium (2 mL) provided by Organogenesis Inc. (Canton, MA) was
composed of 6.6 g L Dulbecco's Modified Eagle's medium, 5.3 g/L
Ham's F-12 powder, 0.35 g/L sodium bicarbonate, 2.6 g/L HEPES
(sodium salt), and 1.0 ,ug/mL gentamicin sulfate (pH = 7.4). The
receptor medium was removed and replenished by a fresh medium
a t 10, 24, 34, 48, 58, and 72 h.
Measurement-Corticosteroids were extracted immediately, and
the receptor medium was collected and then measured within 24 h
as described p r e v i o ~ s l y .At
~ 72 h, the patch used in the experiment
was removed from the LSE and the drug in the patch was analyzed
as previously described.14
Permeation of Corticosteroid through LSE from the
Patches Containing Betamethasone 17-Valerate, Be-
tamethasone 21-Valerate, and Betamethasone-Figure 1
shows the time courses of the average (n = 6) cumulative
amount of corticosteroids that permeated through the LSE
from BIOPSA containing betamethasone 17-valerate (circles),
betamethasone 21-valerate (squares), and betamethasone
(triangles). The amounts of the three corticosteroids detected
in the receptor fluid sample (culture medium) in the static
cell (well on the plate) are expressed as betamethasone
equivalents. When more than two corticosteroids are detected
in the receptor medium, the total amount is shown in Figure
1.
Journal of Pharmaceutical Sciences / 1479
Vol. 84, No. 12, December 7995

loo 1 the manufacturer, the thickness of viable epidermis equivalent
L h (eq 1)was estimated as h 0.017 cm assuming the same
-
X
51 60
ii
7
c
0
$
40:
20
07
1001' - - - - a 1001. - - - - '
B 4 0 1 - . ~ ~, - .
20
0 24 48 72 '0 24 48
Time (h) Time (h)
Figure 2-The average (n = 6 each) percentage of flux of betamethasone 17-
valerate (O),betamethasone 21-valerate (m), and betamethasone (A)plotted on
the midpoint in each sampling period. Plots in parts a, b, and c show those after
the application of betamethasone 17-valerate, betamethasone 21walerate, and
betamethasone to the LSE, respectively.
Figure 2a shows the average (n = 6) percentage of flux of
the three corticosteroids recovered in each sampling period
following the application of the patch containing betametha-
sone 17-valerate. Betamethasone 21-valerate (squares) in the
receptor fluid was always less than 10% of the total corti-
costeroids measured. The percentage of flux of betamethasone
17-valerate (circles) decreased while betamethasone (tri-
angles) increased with time. Figure 2b shows the average ( n
= 6) percentage of flux of betamethasone 21-valerate and
betamethasone when the patch containing betamethasone 21-
valerate was applied. Only betamethasone but no betametha-
sone 21-valerate was detected in the receptor fluid. Figure
2c shows the average (n = 6) percentage of flux of betametha-
sone when the patch containing betamethasone was applied.
No compound which was potentially a metabolite of be-
tamethasone was detected even when the chromatograph was
carefully traced until 150 min.
When the corticosteroid remaining in the patch which had
been applied to the LSE for 72 h was analyzed, only the
corticosteroid initially loaded (i.e., betamethasone 17-valerate,
betamethasone 21-valerate, or betamethasone) but no other
compound was recovered in BIOPSA.
Estimation of Parameters f o r the Metabolism of
Betamethasone 21-Valerate-The smallest fraction of the
influx metabolized or forward consumption,13fmo, in eq 7 was
estimated as follows for the patches containing betamethasone
21-valerate (Figure 2b). It is assumed that esterase is
uniformly distributed in the viable epidermis equivalent of
LSE but not in other layers. In a previous paper using the
homogenate of LSE suspended in the culture medium, kcm+
and kc,- in eq 8 were estimated as 0.797 and 0.178 h-l,
respe~tively.~ Therefore, k,, (eq 11)was estimated as 0.619
h-l. In a previous experiment, four pieces of the LSE were
suspended in 80 mL of the culture medium (one piece per 20
1480 / Journal of Pharmaceutical Sciences
Vol. 84, No. 12, December 1995
mL of the medium). The mean wet weight and diameter of
the epidermis equivalent of the LSE measured at 72 h were
approximately 60 mg and 1.8 cm, respectively. According to
is approximately 70% of the total epidermis thickness. Thus,
density one for wet stratum corneum equivalent and viable
epidermis equivalent of the LSE. The volume of viable
epidermis equivalent in one piece of LSE was therefore
estimated as x 0.04 mL. Using Vcm= 20 mL, Vsl = 0.04 mL,
and k,, = 0.619 h-l, k in eq 12 was estimated as k x 300 h-l.
The diffusion coefficient of betamethasone 21-valerate in
viable epidermis equivalent was presumed to be of the same
magnitude as that of betamethasone 17-valerate estimated
in human dermis or 0.74 x cm2/s.15 Using this value for
D, a in eq 2 was estimated as a x 340 cm-l and fmo in eq 7 as
fmo x 0.994.
The fraction of the compound degradated in the receptor
fluid, f d g , in eq 8 was estimated by assuming that the
degradation rate constant k d g is equal to kc,- or 0.178 h-l.
The value of f d g was estimated as 0.53 (t = 10 h for the samples
collected 10 h after the receptor medium was replenished) to
0.63 (t = 14 h). The minimum fraction of the influx metabo-
lized or degradated, f m + d g o , (eq lo), is estimated as 0.997 (t =
10 h) to 0.998 (t = 14 h). This indicates that when a patch
containing betamethasone 21-valerate is applied to the LSE,
a t least 99.7-99.8% of corticosteroids in the receptor fluid
72 collected every 10 to 14 h should be found as betamethasone.
Discussion
The smallest fraction of the influx metabolized or forward
consumption,13fmo in eq 7, and the fraction of the compound
degradated in the receptor fluid, f d g in eq 8, were estimated
assuming the steady state while the drug flux tended to
decrease during the experiment in the current study (Figure
1). However, it is unlikely that the average values at
nonsteady state are significantly different from those a t steady
state. When the patch containing betamethasone 21-valerate
was applied to the LSE, no betamethasone 21-valerate but
only betamethasone was detected in the receptor medium
(Figure 2b). This finding was compatible with the theoretical
value of f m + d p o (eq 10) estimated t o be more than 0.997.
Approximately half of betamethasone 21-valerate, which
might have escaped from the metabolism by the skin enzyme
and reached the receptor fluid, was predicted to be degradated
into betamethasone by the chemical degradation in the
receptor fluid because f d g in eq 8 was estimated as 0.53 to 0.63.
Nevertheless, the observation that only betamethasone but
no betamethasone 21-valerate is detected in the receptor fluid
indicates that almost complete metabolism of the compound
has occurred already inside the LSE. The LSE contains
sufficient esterase to eventually metabolize almost all the
betamethasone 21-valerate molecules that permeated from the
vehicle.
On the other hand, when betamethasone 17-valerate was
applied to the LSE, more than 50% of corticosteroids were
detected as the unchanged compound in the receptor medium
(Figure 2a). Since the receptor medium was replaced only
every 10-14 h, the fraction of betamethasone 17-valerate of
the flux from the LSE should be higher than the values
obtained in the experiment (Figure 2a). In the previous
studies, the degradation half-life from betamethasone 17-
valerate to betamethasone 21-valerate was x8 h with or
without skin homogenates or esterase.l6 Therefore, the
degradation of betamethasone 17-valerate should have oc-
curred in the receptor medium after penetrating the LSE as
well as within the LSE. The small amount of betamethasone
21-valerate (less than 10% of all corticosteroids) detected in
the receptor medium was probably formed chiefly by the
degradation of betamethasone 17-valerate that had already
reached the receptor fluid.
In conclusion, an artificial living skin equivalent was useful
in assessing the metabolism of a topical drug. Although there
may be large differences in diffusivity and enzyme activity
between the LSE and human skin in vivo, the present study
suggests that a certain fraction of betamethasone 17-valerate
applied to the skin probably reaches viable epidermis and
dermis as the unchanged compound. On the other hand, the
fraction of betamethasone 21-valerate is probably small in the
skin and it is rapidly metabolized into betamethasone by skin
esterase.
References and Notes
1. Wester, R. C.; Maibach, H. I. Drug Metab. Rev. 1983,14, 169-
205.
2. Kao, J.; Carver, M. P. Drug Metab. Rev. 1990,22, 363-410.
3. Martin, R. J.; Denyer, S. P.; Hadgraft, J. Znt. J . Pharm. 1987,
39, 23-32.
4. Pannatier, A,; Jenner, P.; Testa, B.; Etter, J . C. Drug Metab.
Rev. 1978, 8, 319-343.
5. Kao, J.; Hall, J. J . Pharmacol. Exp. Ther. 1987,241, 482-487.
6. Carver, M. P.; Levi, P. E.; Riviere, J. E. Pestic. Biochem. Physiol.
1990,38, 245-254.
7. Parenteau, N. L.; Nolte, C. M.; Bilbo, P.; Rosenberg, M.; Wilkins,
L. M.; Johnson, E. W.; Watson, S.; Mason, V. S.; Bell, E. J . Cell.
Biochem. 1991,45, 245-251.
8. Nolte, C. J. M.; Oleson, M. A.; Bilbo, P. R.; Parenteau, N. L.
Arch. Dermatol. Res. 1993,285, 466-474.
9. Kubota, K.; Ademola, J.; Maibach, H. I. Dermatology 1994,188,
13-17.
10. Ando, H. Y.; Ho, N. F. H.; Higuchi, W. I. J . Pharm. Sci. 1977,
66, 1525-1528.
11. Fox, J . L.; Yu, C. D.; Higuchi, W. I.; Ho, N. F. H. Znt. J . Pharm.
1979,2, 41-57.
12. Yu, C. D.; Fox, J. L.; Ho, N. F. H.; Higuchi, W. I. J . Pharrn. Sci.
1979, 68, 1341-1346.
13. Siegel, R. A. J . Phys. Chem. 1991, 95, 2556-2565.
14. Kubota, K.; Twizell, E. H.; Maibach, H. I. J . Pharm. Sci. 1994,
83, 1593-1599.
15. Kubota, K.; Maibach, H. I. J . Pharrn. Sci. 1993,82,1039-1045.
16. Cheung, Y. W.; Po, A. L. W.; Irwin, W. J. Znt. J. Pharm. 1985,
26,175-189.
Acknowledgments
We thank Drs. David Friend and Harold W. Nolen 111, of Stanford
Research Institute (SRI),Menlo Park, CA, for their help in developing
the patches containing corticosteroids. We also thank Ms. Nicole
Kukutsch and Mrs. E. Gogoleva for their technical assistance.
JS940655W
Journal of Pharmaceutical Sciences / 1481
Vol. 84, No. 12, December 1995