In vitro evaluation of acyloxyalkyl esters as dermal prodrugs of ketoprofen and naproxen

Article abstract of DOI:10.1021/js970465w

A series of acyloxyalkyl esters of ketoprofen and naproxen were synthesized and investigated as topical prodrugs with the aim of improving the dermal delivery of the drugs. In addition, some hydroxyalkyl esters of ketoprofen and naproxen were synthesized as possible intermediates of acyloxyalkyl prodrugs. All of the prodrugs were more lipophilic than their parent molecules, as evaluated by drug partitioning between 1-octanol and phosphate buffer at pH 7.4 (log P(app)). However, their solubilities in aqueous solutions decreased markedly compared with the parent molecules. The prodrugs were stable toward chemical hydrolysis in aqueous solutions (pH 7.4), but were hydrolyzed to the parent drug both in 80% human serum and in human skin homogenate, with half-lives ranging from 4 to 137 min and from 13 to 403 min, respectively. The abilities of the selected naproxen acyloxyalkyl prodrugs to deliver naproxen through excised human skin were evaluated. Generally, the prodrugs showed similar dermal delivery as the patent drug through cadaver skin. In the present series of lipophilic prodrugs of naproxen, the prodrug with the highest aqueous solubility was the most effective prodrug to deliver naproxen through the skin.

Full text of DOI:10.1021/js970465w

In Vitro Evaluation of Acyloxyalkyl Esters as Dermal Prodrugs of
Ketoprofen and Naproxen
,†
JARKKO RAUTIO,* HANNU TAIPALE,† JUKKA GYNTHER,† JOUKO VEPSA¨LA¨INEN,‡ TAPIO NEVALAINEN,† AND
†
TOMI JA¨RVINEN
Contribution from Department of Pharmaceutical Chemistry, and Department of Chemistry, University of Kuopio,
P.O. Box 1627 FIN-70211, Kuopio, Finland.
Received December 9, 1997.
Final revised manuscript received July 21, 1998.
Accepted for publication July 24, 1998.
etration have been studied. Penetration enhancers have
been extensively used to achieve this goal. They are
chemical compounds that are themselves pharmacologically
inactive, but can partition into skin, and interact with the
constituents of the superficial skin layer and reduce the
resistance of skin to drug diffusion.7 The selection of
formulations including different vehicle compositions is one
strategy because properties of the vehicle lead to changes
in drug solubility, diffusion, and partition into skin and,
therefore, influence the permeability of a drug.8
The prodrug approach represents an alternative method
of enhancing the skin permeability of drugs. The prodrugs
are bioreversible pharmacologically inactive derivatives of
a drug molecule that require a chemical or enzymatic
transformation to release the active parent drug in situ.9
Because the epidermis is relatively rich in nonspecific
esterases and other enzymatic activity,10,11 the prodrug
approach has been increasingly used to improve delivery
of a drug through the skin and/or to localize drug action
within the skin.12-15 However, high prodrug concentration
in skin may lead to enzyme saturation kinetics and as a
result, limited conversion of prodrug to the parent drug. A
recent study with ketorolac acid esters suggested that a
large fraction of intact ester may transport into the
cutaneous microcirculation because of low biotransforma-
tion in human epidermis, and convert to the parent drug
by the serum esterases.15
A successful dermal prodrug of ketoprofen and/or naprox-
en should exhibit optimum lipophilicity (partition coef-
ficient) and should be stable toward chemical degradation
prior to hydrolysis within the skin by enzymes. The
temporary masking of the carboxylic group of ketoprofen
and naproxen via simple esterification has been proposed
as a promising means of reducing GI irritation16 and it is
also a useful means of modifying the lipophilicity of parent
molecules to optimize partitioning into the skin and to
maximize percutaneous penetration.17,18 Simple naproxen
esters release naproxen very slowly in human serum and
skin-serum homogenate, but a few prodrug esters pen-
etrate the skin markedly better than naproxen. However,
conclusions cannot be drawn because drug penetration was
calculated as the sum of the prodrug and naproxen.18
Abstract 0 A series of acyloxyalkyl esters of ketoprofen and naproxen
were synthesized and investigated as topical prodrugs with the aim
of improving the dermal delivery of the drugs. In addition, some
hydroxyalkyl esters of ketoprofen and naproxen were synthesized as
possible intermediates of acyloxyalkyl prodrugs. All of the prodrugs
were more lipophilic than their parent molecules, as evaluated by drug
partitioning between 1-octanol and phosphate buffer at pH 7.4 (log
P ). However, their solubilities in aqueous solutions decreased
app
markedly compared with the parent molecules. The prodrugs were
stable toward chemical hydrolysis in aqueous solutions (pH 7.4), but
were hydrolyzed to the parent drug both in 80% human serum and in
human skin homogenate, with half-lives ranging from 4 to 137 min
and from 13 to 403 min, respectively. The abilities of the selected
naproxen acyloxyalkyl prodrugs to deliver naproxen through excised
human skin were evaluated. Generally, the prodrugs showed similar
dermal delivery as the parent drug through cadaver skin. In the
present series of lipophilic prodrugs of naproxen, the prodrug with
the highest aqueous solubility was the most effective prodrug to deliver
naproxen through the skin.
Introduction
Ketoprofen and naproxen are nonsteroidal antiinflam-
matory drugs (NSAIDs) and are used through the oral or
suppository routes for the treatment of pain and inflam-
mation. Given orally, gastrointestinal (GI) side-effects
constitute the most frequent of all the adverse reactions
of NSAIDs.1 Therefore, percutaneous administration of
NSAIDs, like ketoprofen and naproxen, has been studied
to minimize GI side-effects and other possible systemic
side-effects due to high plasma peak levels after oral
administration.2-4 Recent studies have shown that GI side-
effects of NSAIDs are, at least partly, due to inhibition of
the COX-1 enzyme, and thus it may not be possible to
eliminate GI side-effects of NSAIDs by percutaneous
administration.5 However, a topically administered drug
would be more suitable for treatment of local inflammatory
and pain processes because of higher local drug concentra-
tion.6
An essential prerequisite for the development of a dermal
drug delivery system is that the drug be capable of
penetrating the skin at a sufficiently high rate to obtain
the desired pharmacologic activity. Because most drugs
such as ketoprofen and naproxen show unsuitable physi-
cochemical properties, various strategies to aid skin pen-
In the present study, a number of acyloxyalkyl esters of
ketoprofen and naproxen have been synthesized for topical
drug delivery and their aqueous solubilities, lipophilicities,
and hydrolysis rates in buffer, human serum, and skin
homogenate were investigated. The permeation of selected
naproxen prodrugs across the excised human skin was
studied in vitro. In addition, some hydroxyalkyl ester
derivatives of ketoprofen and naproxen were synthesized
and investigated as possible intermediates of acyloxyalkyl
prodrugs.
* Author to whom all correspondence should be sent. Telephone:
358-17-163445. Fax.: 358-17-162456. E-mail: J arkko.Rautio@uku.fi.
† Department of Pharmaceutical Sciences.
‡ Department of Chemistry.
1622 / Journal of Pharmaceutical Sciences
10.1021/js970465w CCC: $15.00
Published on Web 09/12/1998
© 1998, American Chemical Society and
Vol. 87, No. 12, December 1998
American Pharmaceutical Association

Experimental Section
Ch em ica lssNaproxen was kindly donated by Orion Pharma
(Espoo, Finland) and ketoprofen was purchased from Sigma (St.
Louis, MO). 2-Bromoethanol, 3-bromo-1-propanol, bromomethyl
acetate, 2-bromoethyl acetate, 3-chloropropyl acetate, 4-bromobu-
tyl acetate, chloromethyl pivalate, and isopropyl myristate were
purchased from Aldrich (Steinheim, Germany). 4-Bromo-1-bu-
tanol was synthesized according to previously described method.19
Acetonitrile used in the HPLC procedures was of HPLC grade and
was purchased from Rathburn (Walkerburn, UK), and the bulk
solvents bought were from Merck (Darmstadt, Germany). Bovine
albumin was purchased from Pierce (Rockford, IL).
Gen er a l Meth od s of Syn th esis of P r od r u gssKetoprofen (I-
VIII) and naproxen (IX-XVI) prodrugs (Figure 1) were synthe-
sized according to the general procedure described for nalidixic
acid esters.20 Briefly, a mixture of either ketoprofen sodium salt
or naproxen sodium salt and one equivalent of appropriate
hydroxyalkyl or acyloxyalkyl reagent in N,N-dimethylformamide
(10 mL) was stirred at 60 °C for 24 h. Water (50 mL) was added
to the reaction mixture, and the mixture was extracted with ethyl
acetate (2 × 50 mL). The combined extracts were washed with a
5% aqueous solution of sodium carbonate (2 × 25 mL) and water
(2 × 25 mL), dried over anhydrous calcium sulfate, and evaporated
under reduced pressure. The residue obtained was purified by
the semipreparative HPLC method described later (method 3). The
abbreviations of 1H chemical shifts are as following: tm is triplet
of multiplets; ddd is doublet of doublets of doublets; dm is doublet
of multiplets; bs is broad singlet; qui is quinted; td is triplet of
doublets.
Ketoprofen Hydroxyethyl Ester (I)s82% yield; 1H NMR (CDCl3,
δ): 7.80 (2H, m), 7.79 (1H, tm), 7.67 (1H, ddd), 7.60 (1H, m), 7.55
(1H, dm), 7.49 (2H, m), 7.44 (1H, tm), 4.23 (2H, m), 3.85 (1H, q),
3.78 (2H, bs), 1.98 (OH, bs), 1.56 (3H, d); high-resolution mass
spectrometry (HRMS): data not available.
Ketoprofen Hydroxypropyl Ester (II)s48% yield; 1H NMR
(CDCl3, δ): 4.25 (2H, m), 3.58 (2H, m), 1.90 (OH, bs), 1.83 (2H,
qui); HRMS: calcd for C19H20O5: 312.136; found, 312.139.
Ketoprofen Hydroxybutyl Ester (III)s22% yield; 1H NMR (CDCl3,
δ): 4.12 (2H, m), 3.58 (2H, t), 2.01 (OH, bs), 1.70 (2H, m), 1.68
(2H, m); HRMS: calcd for C20H22O4: 326.152; found, 326.148.
Ketoprofen Acetyloxymethyl Ester (IV)s64% yield; 1H NMR
(CDCl3, δ): 7.79 (2H, m), 7.74 (1H, tm), 7.68 (1H, ddd), 7.59 (1H,
m), 7.53 (1H, dm), 7.49 (2H, m), 7.44 (1H, td), 5.74 (1H, d)/5.72
(1H, d), 3.84 (1H, q), 2.04 (3H, s), 1.55 (3H, d); HRMS: calcd for
Figure 1sChemical structures of ketoprofen and naproxen prodrugs (IV−VIII
and XII−XVI) and their intermediates (I−III and IX−XI).
Naproxen Acetyloxypropyl Ester (XIV)s85% yield; 1H NMR
(CDCl3, δ): 4.15 (2H, m), 4.04 (2H, m), 1.91 (3H, s), 1.89 (2H, m);
HRMS: calcd for C19H22O5: 330.147; found, 330.145.
Naproxen Acetyloxybutyl Ester (XV)s73% yield; 1H NMR (CDCl3,
δ): 4.09 (2H, t), 3.98 (2H, t), 1.98 (3H, s), 1.62 (2H, m), 1.57 (2H,
m); HRMS: calcd for C20H24O5: 344.162; found, 344.162.
Naproxen Pivaloyloxymethyl Ester (XVI)s65% yield; 1H NMR
(CDCl3, δ): 5.76 (1H, d), 5.73 (1H, d), 1.05 (9H, s); HRMS: calcd
for C20H24O5: 344.162; found, 344.163.
C
19H18O5: 326.115; found, 326.114.
Ketoprofen Acetyloxyethyl Ester (V)s49% yield; 1H NMR (CDCl3,
δ): 4.28 (2H, m), 4.22 (2H, m), 1.97 (3H, s); HRMS: calcd for
C
20H20O5: 340.131; found, 340.131.
Ketoprofen Acetyloxypropyl Ester (VI)s90% yield; 1H NMR
NMR Sp ectr oscop ys1H and 13C NMR spectra were recorded
on a Bruker AM 400 WB operating at 400.1 and 100.6 MHz,
respectively, using tetramethylsilane (TMS) as a reference and
CDCl3 as the solvent. The spectra were acquired using 32 kW
data points with zero filling to a point resolution of >0.1 Hz. 1H-
1H correlation spectroscopy (COSY) and 1H-1H nuclear Over-
hauser enhancement spectroscopy (NOESY) spectra (mixing time
0.7 s) were measured using 256*512 matrixes and 1H-13C cor-
related spectra were measured using 128*2048 matrixes with zero
filling. On request, the 13C chemical shifts are available from
authors.
Electr on -Im p a ct Ion iza tion Ma ss Sp ectr om etr ysElectron
impact (EI) mass spectra of the prodrugs were recorded on a VG
70-250SE magnetic sector mass spectrometer (VG Analytical,
Manchester, UK). The resolution of the instrument was adjusted
to 10.000. The electron energy was 70 eV, the ionization current
was 500 µA, and the ion source temperature was 200 °C. Samples
were introduced to the mass spectrometer in a glass sample holder
with a direct insertion probe. The probe temperature was raised
from 50 to 300 °C in 5 min.
(CDCl3, δ): 4.16 (2H, m), 4.05 (2H, m), 2.02 (3H, s), 1.92 (2H, m);
HRMS: calcd for C21H22O5: 354.147; found, 354.149.
Ketoprofen Acetyloxybutyl Ester (VII)s82% yield; 1H NMR
(CDCl3, δ): 4.11 (2H, t), 4.02 (2H, t), 2.02 (3H, s), 1.66 (2H, m),
1.61 (2H, m); HRMS: calcd for C22H24O5: 368.162; found, 368.164.
Ketoprofen Pivaloyloxymethyl Ester (VIII)s51% yield; 1H NMR
(CDCl3, δ): 5.76 (1H, d), 5.73 (1H, d), 1.11 (9H, s); HRMS: calcd
for C22H24O5: 368.162; found, 368.162.
Naproxen Hydroxyethyl Ester (IX)s96% yield; 1H NMR (CDCl3,
δ): 7.65 (2H, d), 7.63 (1H, m), 7.38 (1H, dd), 7.11 (1H, dd), 7.07
(1H, d), 4.14 (2H, m), 3.86 (1H, q), 3.84 (3H, s), 3.66 (2H, t), 2.60
(OH, bs), 1.55 (3H, d); HRMS: calcd for C16H18O4: 274.121; found,
274.120.
Naproxen Hydroxypropyl Ester (X)s87% yield; 1H NMR (CDCl3,
δ): 4.19 (2H, m), 3.50 (2H, t), 2.25 (OH, bs), 1.75 (2H, qui);
HRMS: calcd for C17H20O4: 288.136; found, 288.135.
Naproxen Hydroxybutyl Ester (XI)s21% yield; 1H NMR (CDCl3,
δ): 4.10 (2H, t), 3.53 (2H, t), 1.80 (OH, bs), 1.64 (2H, m), 1.50 (2H,
m); HRMS: calcd for C18H22O4: 302.152; found, 302.151.
Naproxen Acetyloxymethyl Ester (XII)s87% yield; 1H NMR
(CDCl3, δ): 7.70 (1H, m), 7.69 (1H, d), 7.69 (1H, d), 7.37 (1H, dd),
7.14 (1H, dd), 7.10 (1H, d), 5.73 (1H, d), 5.70 (1H, d), 3.89 (1H, s),
3.88 (1H, q), 1.97 (3H, s), 1.58 (3H, q); HRMS: calcd for
Liqu id Ch r om a togr a p h ysMethod 1sThe analytical HPLC
system for determination of physicochemical properties of prodrugs
consisted of a Beckman model 116 pump with a model 166 UV
detector, a Marathon automatic sample injector, and a Osborne
MD 700 computer. The column was a Kromasil 100-C8 (150 ×
4.6 mm, 5 µm) reversed-phase column. A mixed mobile phase of
acetonitrile, water, and acetic acid (55:44:1, v/v) at a flow rate of
C
17H18O5: 302.115; found, 302.115.
Naproxen Acetyloxyethyl Ester (XIII)s93% yield; 1H NMR
(CDCl3, δ): 4.27 (2H, m), 4.21 (2H, m), 1.91 (3H, s); HRMS: calcd
for C18H20O5: 316.131; found, 316.133.
Journal of Pharmaceutical Sciences / 1623
Vol. 87, No. 12, December 1998

1.2 mL/min was used. The column effluent was monitored at 254
nm for ketoprofen prodrugs and at 230 nm for naproxen prodrugs.
Method 2sThe analytical HPLC system for determination of
drug in skin permeation samples consisted a Merck Hitachi
L-6200A intelligent pump, Hewlett-Packard HP1046A program-
mable fluorescence detector (exitation 226 nm; emission 368 nm),
a Merck Hitachi D-6000A interface module, a Merck Hitachi AS-
2000 autosampler, and a Merck LaChrom column oven L-7350.
Separations were performed with a Purospher RP-18 endcapped
reversed-phase column (125 × 4.0 mm, 5 µm). A mobile phase
mixture of methanol and a 0.02 M phosphate buffer solution of
pH 5.5 at a flow rate of 1.2 mL/min were used. The proportion of
methanol in the mobile phase was increased linearly from 55 to
80% over 10 min, maintained for 3 min, then returned to the initial
conditions over 4 min. Quantitation of the compounds was made
from measurements of the peak areas in relation to those of
standards chromatographed under the same conditions.
Method 3sThe semipreparative HPLC system consisted of a
Shimadzu model LC-6A with a model SPD-6A detector (350 nm),
a Rheodyne 480 injector module with a 2-mL loop, and a Shimadzu
model SCL-6B system controller. The column was a Kromasil
KR100-5-C8 (250 × 10 mm, 5 µm) and the isocratic solvent system
was acetonitrile and water (55:45, v/v). The flow rate was 5.0 mL/
min.
hydrolysis of the prodrugs were calculated from the linear slopes
of plots of the logarithm of residual prodrugs against time. The
pseudo-first-order times at which 50% of total parent compound
had been formed (f50%) were determined from the linear slopes of
plots of the logarithm of unformed parent compound (log(parent
compoundmax - parent compoundt)) against time.21 Duplicate
determinations were made for each prodrug. The protein concen-
tration in the serum (47 mg/mL) was determined by a dye-binding
assay, with bovine serum albumin (BSA) as the standard.22
Hyd r olysis in Hu m a n Sk in Hom ogen a tesThe rates of
hydrolysis of ketoprofen and naproxen prodrugs (IV-VIII, XII-
XVI) in human skin homogenate (see Skin Preparation section)
were investigated by placing ∼5 µmol of prodrugs in 10.0 mL of
phosphate buffer (0.05 M, pH 7.4). The solutions were either
filtered (IV-XIII) or centrifuged (XIV-XV), and 1.0 mL of
preheated skin homogenate (37 °C) was added to 9.0 mL of clear
prodrug solution. The solutions were placed in a water bath and,
at suitable intervals, 0.5-mL samples were withdrawn and each
mixed with 1.0 mL of ethanol. The mixture was centrifugated and
analyzed by HPLC using the method described earlier (method
1). The t1/2 values for the hydrolysis of the prodrugs were
calculated from the linear slopes of plots of the logarithm of
residual prodrug concentration versus time. The formation of
parent compound (f50%) was determined by the method described
earlier.21 A single determination was made for each prodrug.
Sk in P r ep a r a tion sSamples of human skin, not >5 days old,
were obtained from adult human cadavers. The fresh skin was
immersed in water at 60 °C for 2 min, after which stratum
corneum and epidermis were removed from the dermis.23 The skin
specimens were placed on a Teflon plate, dried for 2-4 days at
room temperature, and frozen.
For the skin homogenate, a sufficient amount of dried human
skin was homogenized (Ystral 10/25, dispensing tool 10G, Dottin-
gen, Germany), in 50 mM phosphate buffer (pH 7.4) at 22 000 rpm,
to obtain a final concentration of 360 mg/mL. The supernatant
was separated after centrifugation at 14000g (Eppendorf 5415C,
Hinz GmbH., Hamburg, Germany), and lipid drops were removed
from the surface of the clear mixture and then stored at -20 °C
until use. The protein concentration in the skin homogenate was
determined to be 560 µg/mL by a dye-binding assay, with BSA as
the standard.22
In Vitr o Sk in P er m ea tion Stu d ysIn permeation studies, the
dried and frozen skin specimens were thawed, rehydrated, and
mounted in Franz-type diffusion cells (PermeGear, Inc., Riegels-
ville, PA) with the stratum corneum facing the donor chamber.
The surface area of the diffusion cell was 0.71 cm2. The receptor
chamber was filled with isotonic phosphate buffer (0.05 M, pH 7.4)
containing 0.02% sodium azide as a preservative. Samples of
naproxen and its prodrugs (5.0 mM) were applied as suspensions
in this buffer solution. The receptor phase was stirred magneti-
cally and kept at a constant temperature of 37 °C with a circulating
water bath throughout the study. Samples were taken from the
receptor compartment at appropriate time intervals, with replace-
ment with fresh buffer solution. The drug concentrations were
assayed by HPLC as described earlier (method 2) and corrected
for dilution attributable to the sampling procedure. The skin flux
values for rate of delivery of naproxen and its prodrugs XIII, XIV,
and XVI were determined from Fick′s law of diffusion (eq 1):
Deter m in a tion of Solu bilitiessThe solubilities of ketoprofen,
naproxen, and their acyloxyalkyl prodrugs (IV-VIII, XII-XVI)
were determined in phosphate buffer at pH 5.0 and 7.4, and the
solubilities of the prodrugs (VII, XII-XVI) only were determined
in isopropyl myristate (IPM). These solubilities were measured
at room temperature by placing excess amounts of each compound
in 1-5 mL of one of these solvents. The slurries in phosphate
buffers were placed in an ultrasonic bath for 3 min and then stirred
for 3-4 h and filtered (Millipore 0.45 µm). The filter did not absorb
compounds after its saturation with 1 mL of solution. Triplicate
determinations were made for each compound. The slurries in
IPM were stirred for 1 h and, after filtration, the solution was
diluted with HPLC eluent. Determinations were made once. The
concentrations of the compounds in their saturated solutions were
determined by the HPLC method described earlier (method 1). The
hydrolysis of esters were negligible during the solubility studies.
Deter m in a tion of P a r tition Coefficien tsThe apparent par-
tition coefficients (Papp) of ketoprofen and naproxen and their
acyloxyalkyl prodrugs (IV-VIII, XII-XVI) were determined at
room temperature in
a 1-octanol-phosphate buffer (pH 7.4)
system. Before use, the 1-octanol was saturated with phosphate
buffer for 24 h by stirring vigorously. A known concentration of
prodrug in pH 7.4 phosphate buffer (0.16 M) was shaken for 60
min with a suitable volume of 1-octanol to achieve equilibrium.
After shaking, the phases were separated by centrifugation at 4000
rpm for 10 min. All experiments were performed in triplicate. The
concentrations of the compounds in the buffer phase before and
after partitioning were determined by the HPLC method described
earlier (method 1).
Hyd r olysis in Aqu eou s Solu tion sThe rates of chemical
hydrolysis of prodrug derivatives (I-XIV) were determined in 0.16
M phosphate buffer (pH 7.4, ionic strength 0.5) at 37 °C. Solutions
of prodrugs were prepared by adding an appropriate amount of
compound to 2.5 mL ethanol (prodrugs IV-VII, XII-XV) or to
5.0 mL ethanol (prodrugs VIII and XVI). Ethanol was used to
effect solution. After vortexing and sonicating the solutions for
10 min, preheated phosphate buffer was added (25 mL). The
solutions were kept in a water bath at 37 °C, and, at appropriate
intervals, samples were taken and were analyzed for remaining
prodrug and any ketoprofen and/or naproxen formed, by using the
HPLC method described earlier (method 1). A single determina-
tion was made for each prodrug.
Hyd r olysis in Hu m a n Ser u m sThe rate of hydrolysis for
ketoprofen and naproxen prodrugs (I-XVI) was determined in
human serum (Institute of Public Health, University of Kuopio)
diluted to 80% with 0.16 M phosphate buffer of pH 7.4 at 37 °C.
The prodrugs (∼5 µmol) were added to phosphate buffer. After
vortexing, preheated human serum was added. The solutions were
kept in a water bath at 37 °C. At suitable intervals, 1.0-mL
samples of serum/buffer mixture were withdrawn and added to
2.0 mL of ethanol to precipitate protein from the serum. After
mixing and centrifugation, the resulting clear supernatant was
analyzed for remaining prodrug and ketoprofen and/or naproxen
by HPLC (method 1). Pseudo-first-order half-times (t1/2) for the
J ss ) V/A(dC/dt)
(1)
where J ss is the steady-state skin flux (nmol/cm2/h), V is the
receptor volume (mL), A is the surface area of the diffusion cell
(cm2), C is the receptor concentration (nmol/mL), and t is time (h).
The steady-state skin flux was determined from the slope of
the linear portion of the cumulative amounts (in nmol) of the
parent drugs, intermediates, and intact prodrugs measured in the
receptor phase versus time plot. The lag time (tlag) was determined
by extrapolating the linear portion of the curve to the abscissa.
The experiments were performed at least in triplicate.
Results and Discussion
Syn th esis a n d Ch a r a cter iza tion of P r od r u gssThe
acyloxyalkyl prodrugs of ketoprofen (IV-VIII) and naprox-
en (XII-XVI), found in Figure 1, have not been reported
1624 / Journal of Pharmaceutical Sciences
Vol. 87, No. 12, December 1998

Table 1sApparent Partition Coefficient (mean ± SD; n ) 3), Aqueous
Solubility (mean ± SD; n ) 3), and Lipid Solubility (One
Determination) of Ketoprofen or Naproxen and their Various
Acyloxyalkyl Prodrugs
Table 2sHalf-Lives (t1/2) of Hydroxyalkyl Esters of Ketoprofen (I−III)
and Naproxen (IX−XI) in Buffer Solution (pH 7.4) and in Human
Serum (pH 7.4) at 37 °C
t1/2 (days)
t1/2 (min)
aqueous solubility (µM)
compound
phosphate buffer 80% human serum
lipid solubility
a
b
compound
log P
pH 7.4
pH 5.0
(mM) IPM
app
I ketoprofen hydroxyethyl ester
II ketoprofen hydroxypropyl ester
III ketoprofen hydroxybutyl ester
15
47
38
124
69
35
ketoprofen −0.2 ± 0.01 1 400 000
3250
46 ± 3
sc
IV
V
VI
VII
VIII
3.4 ± 0.03
3.1 ± 0.01
3.6 ± 0.1
3.6 ± 0.1
3.5 ± 0.1
72 ± 6
89 ± 4
47 ± 6
22 ± 3
2 ± 0.1
64± 3
111 ± 2
56± 1
18± 2
2 ± 0.1
400
7± 2
52± 5
3± 0.4
13± 0.5
0.3 ± 0.03
sc
IX naproxen hydroxyethyl ester
X naproxen hydroxypropyl ester
XI naproxen hydroxybutyl ester
25
52
65
224
147
61
sc
>280d
sc
naproxen
XII
XIII
XIV
XV
0.3 ± 0.03
3.4 ± 0.1
3.5 ± 0.03
3.3 ± 0.1
3.3 ± 0.03
2.8 ± 0.04
102 000
22 ± 3
Table 3sRate of Hydrolysis of Ketoprofen Acyloxyalkyl Prodrugs
(IV−VIII) and Naproxen Acyloxyalkyl Prodrugs (XII−XVI) In Buffer
Solution, In Human Serum, and In Human Skin Homogenate at pH 7.4
and 37 °C
8 ± 1
60 ± 3
4 ± 1
14 ± 1
0.4 ± 0.1
>90
>300d
sc
>320d
>440d
t1/2 (h)
t1/2 (min)
f50% (min)a
t1/2(min)
f50% (min)a
XVI
com- phosphate human serum human serum human skin human skin
a P is an apparent partition coefficient between phosphate buffer (pH
app
pound
buffer
(n ) 2)d
(n ) 2)d
homogenate homogenate
7.4) and 1-octanol at room temperature. b Isopropylmyristate. c Solubility was
not determined because of low yield of prodrug. d Solubility study was
terminated before the saturated condition because of very high lipid solubility
of prodrug.
IV
V
VI
VII
VIII
10
150
385
575
30
4 (3, 5)
12 (10, 14)
29 (28, 29)
61 (59, 62)
35 (32, 37)
11 (11, 12)
13
344
403
66
13
sc
sc
62
26
25 (24, 26)
44 (42, 45)
11 (10, 12)
10 (9, 11)
17
earlier. The hydroxyalkyl esters I-III and IX-XI were
synthesized as possible intermediates of acyloxyalkyl pro-
drugs. The HRMS and NMR investigations verified the
structures of the prodrug substances. The purities were
determinated by HPLC and NMR and they were >95%
(mol %) for each prodrug.
Solu bility a n d Ap p a r en t P a r tition Coefficien ts
Drug solubility is regarded as a key, but not exclusive,
parameter that controls the skin permeation of a drug. One
of the main aims in dermal prodrug delivery is to obtain a
prodrug that not only exhibits increased lipid solubility but
also maintains significant water solubility compared with
parent drug.14,24,25 The log Papp and aqueous solubilities
of ketoprofen and naproxen acyloxyalkyl prodrugs (IV-
VIII, XII-XVI) were determined at physiological pH 7.4
and at pH 5.0 due to acidic conditions (pH ∼5)26 on the
outer surface of the skin.
XII
20
400
485
1900
100
33 (32, 33)
82 (77, 86)
137 (136, 137) 149 (145, 153)
68 (62, 74)
9 (7, 11)
61 (60, 62)
74 (62, 85)
30
158
139
144
sb
50
XIII
XIV
XV
XVI
sc
sc
sc
sb
70 (65, 75)
25 (22, 27)
a f50% is the time by which 50% of total ketoprofen and/or naproxen has
been formed. b Not determined because of low aqueous solubility. c Not
determined because of slow enzymatic hydrolysis of these prodrugs. d Ranges
are shown in parantheses.
The solubilities of ketoprofen, naproxen, and their pro-
drugs (IV-VIII, XII-XVI) in phosphate buffer at pH 5.0
and pH 7.4, the solubilities of prodrugs (IV, VI-X) in IPM,
and the log Papp values of the compounds are shown in
Table 1.
All the prodrugs are markedly more lipophilic than the
parent drugs, as illustrated by the log Papp values. Due to
the acid character of ketoprofen (pKa 4.60) and naproxen
(pKa 4.15), they are more water soluble at pH 7.4 than at
pH 5.0. However, due to a lack of ionizable groups, the
prodrugs showed similar aqueous solubilities at both pH
values, but all were much lower than that of their respec-
tive parent drugs. The acetyloxyethyl esters (V and XIII)
showed the highest aqueous solubility, whereas pivaloyl-
oxymethyl esters (VIII and XVI) showed the lowest aque-
ous solubility.
Ch em ica l Hyd r olysis in Aqu eou s Solu tion sThe
degradations of ketoprofen and naproxen prodrugs (I-XIV)
were studied in buffer solution (pH 7.4) at 37 °C. The
degradation of each prodrug followed first-order kinetics.
The t1/2 values for degradation of hydroxyalkyl prodrugs
(I-III, IX-XI) and acyloxyalkyl prodrugs (IV-VIII, XII-
XVI) are shown in Tables 2 and 3, respectively.
Scheme 1
prodrug metabolism (Scheme 1).14,27 The present results
indicate that hydroxyethyl, -propyl, and -butyl esters of
ketoprofen and naproxen do not easily release the parent
drug by chemical hydrolysis. The results are in good
agreement with earlier reported results for hydroxyalkyl
esters of penicillin,28 mefenamic acid, and diclofenac,29 all
of which exhibited slow hydrolytic degradation in aqueous
solutions. On the other hand, the first step in the hydroly-
ses of the acyloxymethyl prodrug is the hydrolysis of its
terminal ester bond with formation of a hydroxymethyl
ester, which rapidly hydrolyzes to the parent drug and
formaldehyde (Scheme 1). Unfortunately, in this study,
no comparison could be made between the respective
hydroxymethyl esters and other hydroxyalkyl esters be-
cause attempts to synthesize the very unstable hydroxy-
methyl ester derivatives failed. Acyloxyalkyl prodrugs also
proved to be stable in aqueous solutions at pH 7.4.
Prodrugs VI, VII, XIV, and XV were the most stable
prodrugs, whereas acetyloxymethyl esters (IV and XII)
were the most labile prodrugs.
All hydroxyalkyl prodrugs (I-III, IX-XI) were highly
resistant to chemical hydrolysis. The t1/2s (37 °C) of all
compounds were >15 days at pH 7.4 and much longer yet
at pH 5.0 (data not shown). It is generally assumed that
the hydroxyalkyl esters are intermediates in acyloxyalkyl
En zym a tic Hyd r olysissAn essential prerequisite for
the successful topical use of prodrugs is bioconversion of
the prodrug in the skin. In the present study, the rates of
enzymatic prodrug hydrolysis were determined both in
human skin homogenates and in human serum. This
Journal of Pharmaceutical Sciences / 1625
Vol. 87, No. 12, December 1998

Figure 2sPseudo-first-order plots for the hydrolysis of ketoprofen prodrug
VIII (0) and naproxen prodrug XV (9) in 80% human serum (pH 7.4) and
prodrugs VIII (O) and XV (b) in human skin homogenate (pH 7.4) at 37 °C.
Figure 4sTime courses for ketoprofen acetyloxymethyl ester (IV) (O) and
ketoprofen (b) during hydrolysis of the prodrug in 10% human skin homogenate
(pH 7.4) at 37 °C.
Figure 5sHPLC chromatogram of a sample (30 min) from enzymatic
hydrolysis studies of naproxen acetyloxyethyl ester (XIII).
in each medium. Previously reported results for the
hydrolysis of highly lipophilic naproxen esters showed a
similar considerable decline in the rate of hydrolysis of the
esters in skin homogenate compared with human serum.17
However, a true comparison of enzymatic degradation rate
of the prodrug between different vehicles is not reasonable
because the total esterase activity varies for each.
Figure 3sTime courses for ketoprofen acetyloxymethyl ester (IV) (O) and
ketoprofen (b) during hydrolysis of the prodrug in 80% human serum (pH
7.4) at 37 °C.
method has been applied in assessing bioconversion of
prodrugs for dermal delivery.17,20,30
Table 3 includes f50% values for ketoprofen and naproxen,
which represent the time at which 50% of total parent drug
has been formed in human serum and in human skin
homogenate.21 The f50% values are useful in the evaluation
of prodrugs that require many kinetic steps before the
parent drug is formed. Therefore, f50% can be considered a
more relevant indicator of the formation of the parent drug
than the prodrug degradation rate. In the case of esters
V and XIII, the formations of ketoprofen and naproxen
from their acyloxyalkyl esters in human serum were faster
than degradation of corresponding hydroxyalkyl esters in
human serum or in aqueous solutions. Thus, esters V and
XIII may hydrolyze to the parent drug without formation
of the hydroxyalkyl ester as intermediate. However,
formation of both ketoprofen/naproxen and corresponding
hydroxyalkyl esters was detected in hydrolysis studies on
some prodrugs (Figure 5). This result suggests that both
hydrolysis mechanisms (via intermediate and direct hy-
drolysis to the parent drug) are included in the degradation
of the acyloxyalkyl prodrugs of ketoprofen and naproxen
(Scheme 1). The determined f50% values for prodrugs IV,
VII, VIII, and XII in human skin homogenates indicated
that formation of the parent drug take place at the same
rate as the loss of these prodrugs from solution (Table 3).
Acyloxyalkyl prodrugs (IV-VIII, XII-XVI) exhibited
pseudo-first-order degradation kinetics for 3-4 half-lives
in both serum and skin homogenate (Figure 2). Half-lives
(t1/2) for the degradation of prodrugs in 80% human serum
and in 10% human skin homogenate at 37 °C are displayed
in Table 3 with the half-lives for degradation of the
prodrugs in pH 7.4 phosphate buffer at 37 °C listed for
comparison. Figures 3 and 4 show a typical time course
for the degradation of ketoprofen acetoxymethyl ester (IV)
and the concurrent formation of ketoprofen in human
serum and skin homogenate, respectively.
Hydroxyalkyl ester derivatives of both ketoprofen and
naproxen (I-III, IX-XI) showed bioconversion half-lives
in human serum ranging from 35 to 224 min (Table 2). A
longer hydroxyalkyl chain resulted in faster hydrolyses.
However, all the ketoprofen and naproxen acyloxyalkyl
prodrugs were also highly susceptible to hydrolysis in
human serum (Table 3). The half-lives of ketoprofen and
naproxen acyloxyalkyl prodrugs ranged from 4 to 44 min
and from 9 to 137 min, respectively. The hydrolyses of
prodrugs in human skin homogenate were generally slower
compared with hydrolyses of the same compounds in
serum. Only prodrugs XII and XIV hydrolyzed similarly
1626 / Journal of Pharmaceutical Sciences
Vol. 87, No. 12, December 1998

and hydroxypropyl esters in the receptor phase may be
explained by a high stability of these intermediates in skin
specimens during permeation, which may be at least partly
due to decreased enzymatic activity in the skin specimens
after their preparation.36
Naproxen acetyloxypropyl (XIV) and pivaloyloxymethyl
(XVI) prodrugs have a similar or much lower flux than
naproxen, respectively. However, the acetyloxyethyl (XIII)
prodrug, which has the highest aqueous solubility among
the acyloxyalkyl prodrugs, resulted in 2-fold greater in vitro
skin permeation (determined from the first 30 h) than
naproxen itself. The poor skin permeation obtained for
most of these prodrugs is most probably due to high
partition coefficient of the prodrugs associated with a lower
aqueous solubility compared with naproxen. An optimal
log P values have been found to be between 2 and 3 for
maximal fluxes of various NSAIDs,37,38 whereas all acyl-
oxyalkyl prodrugs, except XIV, were above this range. The
permeation of naproxen was also determined from the
buffer solution of pH 5.0. An almost 8-fold increase in flux
of naproxen, at pH 5.0 compared with pH 7.4, was observed
and is due to the reduced ionization and higher partition
coefficient of naproxen at pH 5.0 compared with that at
pH 7.4. The present results confirm the earlier conclusion
that an effective dermal prodrug must possess both in-
creased lipophilicity and a moderate aqueous solubility.24,25
In conclusion, all acyloxyalkyl ester prodrugs were much
more lipophilic than their parent molecules, proved to be
highly stable in aqueous solutions, and hydrolyzed readily
to the parent drug both in human serum and human skin
homogenate. However, the fluxes through excised human
skin in vitro were still low, most probably due to poor
aqueous solubility of the prodrugs and to high partition
coefficients that were above the optimal range for skin
permeation. Only acetyloxyethyl ester prodrug of naproxen
(XIII), which is the most hydrophilic member of the series
synthesized, exhibited a slight enhancement of in vitro skin
permeability compared with naproxen itself.
Figure 6sPermeation profiles (mean ± SE, n ) 3−5) for naproxen (b) and
prodrugs XIII (O), XIV (0), and XVI (4) through human skin from 0.05 M
phosphate buffer (pH 7.4) vehicle. The data represent the sum of naproxen,
the intermediate, and its prodrug in the case of prodrug application.
Table 4sSteady-State Fluxes (mean ± SE, n ) 3−5) for Delivery of
Total Naproxen Species Through Excised Human Skin In Vitro from
Isotonic Phosphate Buffer (0.05 M, pH 7.4)
flux
2
compound
(nmol/cm /h)
tlag (h)
naproxen
0.23 ± 0.03a
0.44 ± 0.05
0.22 ± 0.03
0.02 ± 0.00
3.4 ± 0.35
1.2 ± 0.49
2.7 ± 0.21
2.4 ± 0.06
XIII naproxen acetyloxyethyl ester
XIV naproxen acetyloxypropyl ester
XVI naproxen pivaloyloxymethyl ester
a The steady-state flux for naproxen from isotonic phosphate buffer (0.05
2
M, pH 5.0) was 1.80 ± 0.41 nmol/cm /h.
The rate of enzymatic hydrolysis of acyloxyalkyl pro-
drugs can be adjusted by changes in the (CH2)n moiety,
with the methylene group producing the fastest hydrolysis
both in human plasma and skin homogenate. Change of
the methyl group in the R1 position to the pivaloyl group
did not have a significant effect on the rate of enzymatic
hydrolysis of the prodrugs. This result is not in agreement
with earlier studies that have shown increased branching
of the acyl portion of esters to decrease the rate of
enzymatic hydrolysis.31-35
In Vitr o Sk in P er m ea tion Stu d ysPermeation of
naproxen and prodrugs XIII, XIV, and XVI through excised
human skin was evaluated from isotonic phosphate buffer
(0.05 M, pH 7.4). The cumulative permeated amount (sum
of intact prodrug, intermediate, and naproxen) of each
compound usually displayed a linear relationship with
time. Steady-state fluxes (J ss) were obtained from slopes
of the linear portions of these plots. Because overall drug
permeation across the skin was very low, sink conditions
were essentially maintained throughout these experiments.
Figure 6 shows the cumulative amounts (in nmol) of total
naproxen species permeated through cadaver skin divided
by the surface area of the diffusion cell as a function of
time. The steady-state fluxes and the lag times of naprox-
en and its prodrugs are summarized in Table 4. During
permeation, the naproxen pivaloyloxymethyl prodrug (XVI)
is converted totally to naproxen. However, in the cases of
prodrugs XIII and XIV, mostly the intact prodrug and
corresponding hydroxyalkyl ester of the prodrug were
found. These differing abilities of prodrugs to undergo
hydrolysis during permeation are in good accordance with
relative stabilities of prodrugs in human serum and in
human skin homogenates. The presence of hydroxyethyl
References and Notes
1. Price, A. H.; Fletcher, M. Mechanisms of NSAID-Induced
Gastroenteropathy. Drugs 1990, 40, S5, 1-11.
2. Baixauli, F.; Ingle´s, F.; Alcantara, P.; Navarrete´, R.; Puchol,
E.; Vidal, F. Percutaneous Treatment of Acute Soft Tissue
Lesions with Naproxen Gel and Ketoprofen Gel. J . Int. Med.
Res. 1990, 18, 372-378.
3. Flouvat, B.; Roux, A.; Delhotal-Landes, B. Pharmacokinetics
of Ketoprofen in Man After Repeated Percutaneous Admin-
istration. Arzneim.-Forsch./ Drug Res. 1989, 39, 812-815.
4. van den Ouweland, F. A.; Eenhoorn, P. C.; Tan, Y.; Gribnau,
F. W. J . Transcutaneous Absorption of Naproxen Gel. Eur.
J . Clin. Pharmacol. 1989, 36, 209-211.
5. Vane, J . R.; Botting, R. M. New Insights into the Mode of
Action of Antiinflammatory Drugs. Inflamm. Res. 1995, 44,
1-10.
6. Suh, H.; J un, H. W.; Dzimianski, M. T.; Lu, G. W. Pharma-
cokinetic and Local Tissue Disposition Studies of Naproxen
Following Topical and Systemic Administration in Dogs and
Rats. Biopharm. Drug Dispos. 1997, 18, 623-633.
7. Barry, B. W. Mode of Action of Penetration Enhancers in
Human Skin. J . Controlled Rel. 1987, 6, 85-97.
8. Henmi, T.; Fujii, M.; Kikuchi, K.; Yamanobe, N.; Matsumoto,
M. Application of an Oily Gel Formed by Hydrogenated
Soybean Phospholipids as a Percutaneous Absorption-Type
Ointment Base. Chem. Pharm. Bull. 1994, 42, 651-655.
9. Stella, V. J .; Charman, W. N. A.; Naringrekar, V. H.
Prodrugs: Do They Have Advantages in Clinical Practice?
Drugs 1985, 29, 455-473.
10. Pannatier, A.; J enner, P.; Testa, B.; Etter, C. The Skin as a
Drug-Metabolizing Organ. Drug Metab. Rev. 1978, 8, 319-
343.
11. Ta¨uber, U. Metabolism of Drugs on and in the Skin. In
Dermal and Transdermal Absorption; Brandau, R., Lippold,
B. H., Eds.; Wissenschaftliche Vertagsgesellschaft mbH:
Stuttgart, 1982; pp 133-151.
Journal of Pharmaceutical Sciences / 1627
Vol. 87, No. 12, December 1998

12. Higuchi, W. I.; Yu, C. D. Prodrugs in Transdermal Delivery.
In Transdermal Delivery of Drugs, CRC Series Vol. III;
Kydonieus, A. F., Berner, B., Eds.; 1987; pp 43-83.
13. Chan, S. Y.; Li Wan Po, A. Prodrugs for Dermal Delivery.
Int. J . Pharm. 1989, 55, 1-16.
thesis and Properties of Some Dihydropyridine and Dihy-
droiosquinoline Derivatives of Benzylpenicillin. J . Med.
Chem. 1989, 32, 1774-1781.
29. J ilani, J . A.; Pillai, G. K.; Salem, M. S.; Najib, N. M.
Evaluation of Hydroxyethyl Esters of Mefenamic Acid and
Diclofenac as Prodrugs. Drug Dev. Ind. Pharm. 1997, 23,
319-323.
30. J ohansen, M.; Mollgaard, B.; Wotton, P. K.; Larsen, C.;
Hoelgaard, A. In Vitro Evaluation of Dermal Prodrug De-
livery - Transport and Bioconversion of a Series of Aliphatic
Esters of Metronidazole. Int. J . Pharm. 1986, 32, 199-206.
31. J a¨rvinen, T.; Poikolainen, M.; Suhonen, P.; Vepsa¨la¨inen, J .;
Alaranta, S.; Urtti, A. Comparison of Enzymatic Hydrolysis
of Pilocarpine Prodrugs in Human Plasma, Rabbit Cornea,
and Butyrylcholinesterase Solutions J . Pharm. Sci. 1995, 84,
656-660.
32. Bundgaard, H.; Falch, E.; Larsen, C.; Mosher, G. L.; Mikkel-
son, T. J . Pilocarpine Prodrugs II. Synthesis, Stability,
Bioconversion, and Physicochemical Properties of Sequen-
tially Labile Pilocarpine Acid Diesters. J . Pharm. Sci. 1986,
75, 775-783.
33. Chien, D. S.; Sasaki, H.; Bundgaard, H.; Buur, A.; Lee, V.
H. L. Role of Enzymatic Lability in the Corneal and Con-
junctival Penetration of Timolol Ester Prodrugs in the
Pigmented Rabbit. Pharm. Res. 1991, 8, 728-733.
34. Huang, H. P.; Ayres, J . W. Dyphylline Prodrugs: Plasma
Hydrolysis and Dyphylline Release in Rabbits. J . Pharm. Sci.
1988, 77, 104-109.
35. Shao, Z.; Park, G. B.; Krishnamoorthy, R.; Mitra, A. K. The
Physicochemical Properties, Plasma Enzymatic Hydrolysis,
and Nasal Absorption of Acyclovir and Its 2′-Ester Prodrugs.
Pharm. Res. 1994, 11, 236-242.
36. Wester, R. C.; Christoffel, J .; Hartway, T.; Poblete, N.;
Maibach, H. I.; Forsell, J . Human Cadaver Skin Viability
for In Vitro Percutaneous Absorption: Storage and Detri-
mental Effects of Heat-Separation and Freezing. Pharm. Res.
1998, 15, 82-84.
37. Yano, T.; Nakagawa, A.; Masayoshi, T.; Noda, K. Skin
Permeability of Various Non-Steroidal Antiinflammatory
Drugs in Man. Life Sci. 1986, 39, 1043-1050.
38. Singh, P.; Roberts, M. S. Skin Permeability and Local Tissue
Concentrations of Nonsteroidal Antiinflammatory Drugs
after Topical Application. J . Pharmacol. Exp. Ther. 1994,
268, 144-151.
14. Sloan, K. B. Prodrugs for Dermal Delivery. Adv. Drug Del.
Rev. 1989, 3, 67-101.
15. Roy, S. D.; Manoukian, E. Permeability of Ketorolac Acid and
Its Ester Analogues (Prodrugs) Through Human Cadaver
Skin. J . Pharm. Sci. 1994, 83, 1548-1553.
16. Whitehouse, M. W.; Rainsford, K. D. Side-Effects of Antiin-
flammatory Drugs: Are They Essential or Can They Be
Circumvented? Agents Actions 1977, 3, 171-187.
17. Kasting, G. B.; Smith, R. L.; Anderson, B. D. Prodrugs for
Dermal Delivery: Solubility, Molecular Size, and Functional
Group Effects. In Prodrugs, Topical and Ocular Drug
Delivery; Sloan K. B., Ed.; Marcel Dekker: New York, 1992;
pp 117-161.
18. Mueller, L. G. Novel Antiinflammatory Agents, Pharmaceu-
tical Compositions and Methods for Reducing Inflammation.
U.S. Patent 4, 912, 248, 1990.
19. Vedejs, E.; Arnost, M. J .; Hagen, J . P. Conformational Effects
on the 2,3-Sigmatropic Shift of Sulfur-Stabilized Ylides and
Enolates Involving Bicyclic Transition States. J . Org. Chem.
1979, 44, 3230-3238.
20. Bundgaard, H.; Mork, N.; Hoelgaard, A. Enhanced Delivery
of Nalidixic Acid Through Human Skin via Acyloxymethyl
Ester Prodrugs. Int. J . Pharm. 1989, 55, 91-97.
21. J a¨rvinen, T.; Suhonen, P.; Urtti, A.; Peura, P. O,O′-(1,4-
Xylene) Bispilocarpic Acid Esters as New Potential Double
Prodrugs of Pilocarpine for Improved Ocular Delivery. II.
Physicochemical Properties, Stability, Solubility and Enzy-
matic Hydrolysis. Int. J . Pharm. 1991, 75, 259-269.
22. Bradford, M. M. A Rapid and Sensitive Method for the
Quantitation of Microgram Quantities of Protein Utilizing
the Principle of Protein-Dye Binding. Anal. Chem. 1976, 48,
248-254.
23. Kligman, A. M.; Christophers, E. Preparation of Isolated
Sheets of Human Skin. Arch. Dermatol. 1963, 88, 702-705.
24. Guy, R. H.; Hadgraft, J . Percutaneous Penetration Enhance-
ment: Physicochemical Considerations and Implication for
Prodrug Design. In Prodrugs; Topical and Ocular Drug
Delivery; Sloan, K. B., Ed.; Marcel Dekker: New York, 1992;
pp 1-16.
25. Waranis, R. P.; Sloan, K. B. Effects of Vehicles and Prodrug
Properties and Their Interactions on the Delivery of 6-Mer-
captopurine Through Skin: Bisacyloxymethyl-6-Mercaptopu-
rine Prodrugs. J . Pharm. Sci. 1987, 76, 587-595.
26. Flynn, G. L. Topical Drug Absorption and Topical Pharma-
ceutical Systems. In Drugs and Pharmaceutical Sciences, Vol.
40, Modern Pharmaceutics; Banker, G. S., Rhodes, C. T.,
Eds.; Marcel Dekker: New York, 1990; pp 263-325.
27. Bundgaard, H. The Double Prodrug Concept and Its Ap-
plications. Adv. Drug Deliv. Rev. 1989, 3, 39-65.
28. Pop, E.; Wu, W.-M.; Shek, E.; Bodor, N. Brain-Specific
Chemical Delivery Systems for â-Lactam Antibiotics. Syn-
Acknowledgments
The authors thank the Academy of Finland and Technology
Development Centre (Finland) for financial support. We are also
grateful to Mr. J ukka Knuutinen and Mrs. Helly Rissanen for their
skillful technical assistance.
J S970465W
1628 / Journal of Pharmaceutical Sciences
Vol. 87, No. 12, December 1998