In-vitro transdermal penetration of cytarabine and its N4-alkylamide derivatives

Article abstract of DOI:10.1211/jpp.62.06.0012

Objectives: The aim of this study was to synthesise and determine the transdermal penetration of cytarabine alkylamide derivatives and assess the correlation of flux with physicochemical properties. Methods: The alkylamide derivatives of cytarabine were synthesised by acylation at the N4-amino group by the mixed anhydride method. The in-vitro permeation studies were performed using the Franz diffusion cell methodology. Furthermore, partition coefficients (n-octanol-water) and aqueous solubility of the N4-alkylamide derivatives of cytarabine were determined in order to obtain information about their lipophilicity and hydrophilicity. Key findings: The N4-alkylamides of cytarabine (acetyl, butanoyl, hexanoyl, octanoyl, and decanoyl derivatives) showed decreased hydrophilicity and increased lipophilicity. The log D values of the alkylamides were higher than that of the parent compound and increased linearly as the alkyl chain lengthened. N4-hexanoyl-4-amino-1-[(2R,3S,4R,5R)-3,4- dihydroxy-5-(hydroxymethyl)oxolan-2-yl] pyrimidin-2-one) showed the highest median steady-state flux (J_ss) of 89.0 nmol/cm² per h in the series, which shows a high statistical difference with the parent compound flux value (3.70 nmol/cm² per h). Conclusions The prodrug approach appears to be a promising strategy for the enhancement of transdermal penetration of cytarabine.

Full text of DOI:10.1211/jpp.62.06.0012

Research Paper
JPP 2010, 62: 756–761
ß 2010 The Authors
Journal compilation ß 2010
Royal Pharmaceutical Society
of Great Britain
In-vitro transdermal penetration of cytarabine and its
N4-alkylamide derivatives
Received September 30, 2009
Accepted February 12, 2010
DOI 10.1211/jpp/62.06.0012
ISSN 0022-3573
Lesetja J. Legoabea, Jaco C. Breytenbachb, David D. N’Daa and
J. Wilma Breytenbachc
aUnit for Drug Research and Development, bPharmaceutical Chemistry, School of Pharmacy and cStatistical
Consultation Services, North-West University, Potchefstroom, South Africa
Abstract
Objectives The aim of this study was to synthesise and determine the transdermal
penetration of cytarabine alkylamide derivatives and assess the correlation of flux with
physicochemical properties.
Methods The alkylamide derivatives of cytarabine were synthesised by acylation at the
N4-amino group by the mixed anhydride method. The in-vitro permeation studies were
performed using the Franz diffusion cell methodology. Furthermore, partition coefficients
(n-octanol–water) and aqueous solubility of the N4-alkylamide derivatives of cytarabine
were determined in order to obtain information about their lipophilicity and hydrophilicity.
Key findings The N4-alkylamides of cytarabine (acetyl, butanoyl, hexanoyl, octanoyl,
and decanoyl derivatives) showed decreased hydrophilicity and increased lipophilicity. The
log D values of the alkylamides were higher than that of the parent compound and
increased linearly as the alkyl chain lengthened. N4-hexanoyl-4-amino-1-[(2R,3S,4R,5R)-
3,4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl] pyrimidin-2-one) showed the highest med-
ian steady-state flux (Jss) of 89.0 nmol/cm2 per h in the series, which shows a high
statistical difference with the parent compound flux value (3.70 nmol/cm2 per h).
Conclusions The prodrug approach appears to be a promising strategy for the
enhancement of transdermal penetration of cytarabine.
Keywords alkylamides; cytarabine; in vitro; transdermal penetration
Introduction
Cytarabine (Figure 1) is a nucleoside analogue of deoxycytidine that is extensively used in the
treatment of both acute and chronic myeloblastic leukaemias. It is a geometric isomer of
cytidine and differs in that the 20-hydroxyl group is oriented in the trans position.[1] The major
impediments to a broad clinical use of cytarabine include the rapid metabolism of the drug in
plasma to its inactive metabolite uracil arabinoside (ara-U) by the enzyme deoxycytidine
deaminase and its cell cycle (S-phase) specificity.[2] A prolonged exposure of cells to
cytarabine’s cytotoxic concentrations is essential to achieve maximum activity since it is a
cell-cycle-specific drug. In practice it is administered by repetitive schedules or continuous
intravenous infusion in order to achieve a sustained supply.[1] These modes of administration are
inconvenient and invasive, and could therefore contribute to increased patient non-compliance.
Transdermal drug delivery (TDD) offers several advantages over more traditional
dosage forms such as oral and intravenous infusion. These include the potential for
sustained release, which is useful for drugs with short biological half-lives requiring
frequent oral or parenteral administration, and controlled input kinetics, which are
indispensable for drugs with narrow therapeutic indices.[3] TDD can achieve consistent
plasma concentrations similar to intravenous infusion without the inconvenience. Due to its
unfavourable physicochemical properties, cytarabine will not easily permeate the skin
without chemical modification.
Correspondence: Lesetja J.
Legoabe, Unit for Drug Research
and Development, North-West
University, Private Bag X6001,
Potchefstroom Campus,
Potchefstroom 2520,
South Africa.
E-mail: lesetja.legoabe@
nwu.ac.za
Many approaches have been explored to protect cytarabine from deamination, including
chemical modifications. Among others, N4-acylation has been shown to prevent
inactivation by cytidine deaminase.[4,5] N4-acylation of cytarabine with long chain fatty
acids such as stearic acid and behenic acid lead to lipophilic amide derivatives of
cytarabine, which exhibit greater antileukaemia activity than native cytarabine.[6]
756

Transdermal penetration of cytarabine
Lesetja J. Legoabe et al.
757
O
O
1²
O
O
O
Cl
O
NH
4
n
Cytarabine
DMF
OH
n
3 N
O
O
5
n
THF, TEA
2
1
6
N
O
4¢
O
HO
(2): nϭ0
(3): nϭ2
(4): nϭ4
(5): nϭ6
(6): nϭ8
1¢
5¢
HO
2¢
HO 3¢
Figure 1 Synthesis of N4-alkylamide derivatives of cytarabine
In studies performed on a structurally related compound CA, USA). A Phenomenex (Luna C-18, 150 ꢀ 4.60 mm,
(gemcitabine), the amide linkage was shown to have extreme 5 μm) column was used with a Securityguard pre-column
chemical stability in pH range 4–9 and yet is bioreversible.[7] (C-18, 4 ꢀ 3 mm) insert (Phenomenex, Torrance, CA, USA)
Therefore amide is a plausible linker for N4-prodrugs of in order to prolong column life. The wavelength of detection
cytarabine.
was fixed at 247 nm. Elution was accomplished with gradient
The objective of this study was to synthesise and at a flow rate of 1 ml/min, with a mobile phase consisting of
investigate the transdermal flux of N4-alkylamides of methanol (solvent A) and 0.005 M heptane sulfonic acid-Na in
cytarabine. Physicochemical properties were determined water adjusted to pH 3.5 with orthophosphoric acid (B). The
and assessed for correlations to transdermal flux values.
gradient was started with 30% A, then increased linearly to
95% A in 8 min and held until 11 min, after which the column
was re-equilibrated at the starting conditions. A calibration
plot of peak area versus drug concentration was constructed
for each compound. The plots showed excellent linearity over
the concentration range (0.6–200 mg/ml) employed for the
assays. The correlation coefficient values 0.998 < r2 ≤ 1 were
found for all calibration plots.
Material and Methods
Materials
Cytarabine (98% purity) was purchased from Jingma
Chemicals Ltd, China. The carboxylic acid anhydrides and
ethylchlorocarbonate were purchased from Sigma-Aldrich
South Africa Ltd. HPLC grade methanol was obtained from
Labchem South Africa Ltd. All other reagents were of
analytical grade and were used without further purification.
Chemical synthesis
The N4-alkylamide derivatives of cytarabine were synthe-
sised by a mixed anhydride method.[8] Aliphatic carboxylic
anhydrides were linked with amino group of cytarabine in a
two-step process: by reacting ethylchlorocarbonate (chloro-
formic acid ethyl ester) and the carboxylic acid in the
presence of triethylamine (TEA) to obtain the correspond-
ing very reactive mixed anhydride; and then adding
cytarabine solution to the reactive mixture (without
isolating the mixed anhydrides) to give the N4-alkylamide
derivatives.
Ethylchlorocarbonate (0.56 mmol, 60.8 mg) was added
dropwise to an anhydrous tetrahydrofuran (THF) solution
of the lipophilic acid (0.56 mmol) and TEA (0.56 mmol,
56.7 mg). The reaction was maintained under stirring at –15°C
for 20 min. A solution of cytarabine (0.56 mmol, 136.2 mg)
in anhydrous dimethylformamide was added under stirring at
–15°C. The reaction mixture was maintained at 5–10°C with
stirring until the end of the reaction. The reaction mixture was
concentrated to dryness under high vacuum and the residue was
purified by silica column chromatography using a mixture of
organic solvents as eluting agents (see Figure 2).
General procedures
The 1H, 13C, COSY (correlation spectroscopy), HMQC
(heteronuclear multiple quantum coherence) and HMBC
(heteronuclear multiple bond coherence) spectra were
recorded on a Bruker 600 spectrometer, using deuterated
dimethylsulfoxide (DMSO) as solvent. The 1H and 13C spectra
were recorded at frequencies of 600.17 and 151.92 MHz,
respectively. All the chemical shifts are reported in parts per
million (ppm) relative to tetramethylsilane (δ = 0). The
splitting pattern abbreviations are as follows: s (singlet), d
(doublet), t (triplet), q (quartet), bs (broad singlet) and m
(multiplet). The melting points of solid products were
determined with a Shimadzu DSC-60A using TA60 (Version
2.11) software. MS spectra were recorded on an analytical VG
7070E mass spectrometer using fast atom bombardment
(FAB) as the ionisation technique. Thin-layer chromato-
graphy was performed using silica gel plates (60F254 Merck)
and flash column chromatography on silica gel (70–240 mesh,
G60 Merck).
High performance liquid chromatography
N4-acetyl-4-amino-1-[(2R,3S,4R,5R)-3,
4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]
The high performance liquid chromatography (HPLC) system
consisted of a Hewlett Packard (HP) Agilent 1100 series auto pyrimidin-2-one (2)
sampler, HP Agilent 1100 series variable wavelength detector Compound (2) was purified by column chromatography using
(VWD) and HP Agilent 1100 series pump (Agilent, Palo Alto, DCM:MeOH (8:1,v/v) as a mobile phase to obtain 1.7 g

758
Journal of Pharmacy and Pharmacology 2010; 62: 756–761
1
(36%) of white powder. m.p. 183°C. H NMR (600 MHz,
DMSO) δ 2.10 (s, 3H, –HNCOCH3), 3.63 (t, J = 5.3, 2H, 50),
3.80–3.86 (m, 1H, H-40), 3.94 (s, 1H, H-30), 4.07 (dd, J = 7.6,
4.1, 1H, H-20), 4.92 (t, J = 5.5, 1H, 50), 5.33 (d, J = 4.3, 1H,
OH-30), 5.35 (d, J = 5.6, 1H, OH-20), 6.06 (d, J = 4.0, 1H,
H-10), 7.13 (d, J = 7.4, 1H, H-5), 8.03 (d, J = 7.5, 1H, H-6),
10.63 (s, 1H, –HNCO). 13C NMR (151 MHz, DMSO) δ 24.07
(–HNCOCH3), 60.92 (C-50), 74.59 (C-30), 76.19 (C-20), 85.52
(C-40), 86.72 (C-10), 93.98 (C-5), 146.35 (C-6), 154.27 (C-4),
161.94 (C-2), 170.57 (C-100). MS FAB 244.0 (M + H+), 175.9,
153.8, 122.2, 102.0, 74.0, 57.9.
N4-decanoyl-4-amino-1-[(2R,3S,4R,5R)-3,
4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]
pyrimidin-2-one (6)
Amide (6) was purified by column chromatography using
DCM:EtOAc:MeOH (6:2:1) as a mobile phase to obtain 1.8 g
1
(28%) of white powder. m.p. 153°C. H NMR (600 MHz,
DMSO) δ 0.86 (t, J = 7.0, 3H, –CH2CH3), 1.21–1.33
(m, 14H, (–CH2–)6CH3), 1.52–1.60 (m, 2H, COCH2CH2–),
2.39 (t, J = 7.4, 2H, HNCOCH2–), 3.64 (s, 2H, H-50), 3.81–
3.88 (m, 1H, H-40), 3.91– 3.99 (m, 1H, H-30), 4.09 (s, 1H,
H-20), 4.93 (s, 1H, OH-50), 5.35 (s, 2H, OH-2'&30), 6.07 (d, J =
4.0, 1H, H-10), 7.18 (d, J = 7.5, 1H, H-5), 8.04 (d, J = 7.5, 1H,
H-6), 10.60 (s, 1H, HNCO). MS FAB 397.9 (M+), 390.9, 288.1,
265.8, 265.8, 243.9, 112.3, 60.3.
N4-butyryl-4-amino-1-[(2R,3S,4R,5R)-3,
4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]
pyrimidin-2-one (3)
Physicochemical properties
Amide (3) was purified by column chromatography using
DCM:MeOH (8:1) as a mobile phase to obtain 1.2 g (23%)
of white powder. m.p. 193°C. H NMR (600 MHz, DMSO)
δ 0.90 (t, J = 7.4, 3H, –CH2CH3), 1.59 (h, J = 7.3, 2H,
-CH2CH3), 2.39 (t, J = 7.3, 2H, HNCOCH2–), 3.64 (s, 2H,
H-50), 3.85 (td, J = 5.2, 3.1, 1H, H-40), 3.95 (d, J = 1.8, 1H,
H-30), 4.09 (s, 1H, H-20), 4.94 (s, 1H, OH-50), 5.35 (s, 2H,
OH-20&30), 6.07 (t, J = 4.9, 1H, H-10), 7.18 (d, J = 7.4, 1H,
H-5), 8.05 (d, J = 7.5, 1H, H-6), 10.60 (s, 1H, –HNCO). MS
FAB (M+) 313.9, (M + Na)+ 335.9.
Solubility determination
The aqueous solubility of the alkylamides was determined by
preparing saturated solutions in phosphate buffered saline
(PBS) (10–2 M, pH 7.4). The slurries were stirred with
magnetic bars in a water bath at 32°C for 24 h. It was ensured
that an excess of solute was present at all times to provide
saturation. The solutions were filtered through 0.2 mm
acrodisc filters, diluted appropriately in PBS (pH 7.4) and
analysed by HPLC to determine the concentration of dissolved
solutes in the PBS.
1
Experimental log D
The partition coefficients were determined by a method
reported by Taylor and Sloan.[9] Equal volumes of n-octanol
and phosphate buffer solution of pH 7.4 were saturated with
each other under vigorous stirring for at least 24 h. An
accurately weighed quantity of 30 mg of the amides (2)–(6)
was dissolved in 3 ml of pre-saturated n-octanol, stoppered
and agitated for 10 min in a 10 ml graduated tube (0.5 ml
division). Subsequently 3 ml of pre-saturated buffer was
transferred to the tubes containing the previously mentioned
solutions. The tubes were stoppered and agitated for 45 min
then centrifuged at 4000 rev/min for 30 min. The volume
ratio (octanol:buffer) was not discernably different from 1.
The octanol and buffer phases were each diluted with
methanol and the concentrations of compounds (2)–(6) were
measured by HPLC. The log D values were calculated as
logs of the concentration ratios in the two phases. These
experiments were done in triplicate. The results expressed as
means are listed in Table 1.
N4-hexanoyl-4-amino-1-[(2R,3S,4R,5R)-3,
4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]
pyrimidin-2-one (4)
Compound (4) was purified by column chromatography
using DCM:MeOH (12:1) as a mobile phase to obtain 3.5 g
1
(63%) of white powder. m.p. 130°C. H NMR (600 MHz,
DMSO) δ 0.88 (t, J = 7.0, 3H, –CH2CH3), 1.19–1.38 (m, 4H,
–CH2CH2CH3), 1.50–1.65 (m, 2H, –COCH2CH2–), 2.40
(t, J = 7.4, 2H, –HNCOCH2–), 3.64 (t, J = 4.5, 2H, H-50),
3.82–3.87 (m, 1H, H-40), 3.95 (s, 1H, H-30), 4.08 (s, 1H,
H-20), 4.93 (s, 1H, OH-50), 5.35 (t, J = 4.8, 2H, OH-20&30),
6.07 (d, J = 4.0, 1H, H-10), 7.18 (d, J = 7.4, 1H, H-5), 8.04
(d, J = 7.5, 1H, H-6), 10.60 (s, 1H, HNCO). MS FAB 341.7
(M+), 231.7, 210.0, 148.7, 112.0.
N4-octanoyl-4-amino-1-[(2R,3S,4R,5R)-3,
4-dihydroxy-5-(hydroxymethyl)oxolan-2-yl]
pyrimidin-2-one (5)
In-vitro skin permeation experiment
Amide (5) was purified by column chromatography using
DCM:MeOH (10:1) as a mobile phase to obtain 2.8 g (43%)
of white powder. m.p. 145°C. 1H NMR (600 MHz, DMSO) δ
0.87 (t, J = 7.0, 3H, –CH2CH3), 1.28 (dd, J = 7.5, 3.7, 8H,
(–CH2–)4CH3, 1.48–1.67 (m, 2H, –COCH2CH2–), 2.40
(t, J = 7.4, 2H, HNCOCH2–), 3.64 (t, J = 4.6, 2H, H-50),
The project ‘In-vitro transdermal delivery of drugs through
human skin’ was approved by the Ethics Committee of the
North-West University (Potchefstroom Campus, South
Africa).
3.81–3.87 (m, 1H, H-40), 3.91–4.00 (m, 2H, H-30), 4.09 Preparation of donor phase
(s, 1H, H-20), 4.93 (s, 1H, OH-50), 5.35 (t, J = 4.5, 2H, Donor solutions were obtained by preparing saturated
OH-20& 30), 6.07 (d, J = 4.0, 1H, H-10), 7.18 (d, J = 7.5, 1H, solutions of compounds (1)–(6) in PBS (10–2 M, pH 7.4) at
H-5), 8.04 (d, J = 7.5, 1H, H-6), 10.60 (s, 1H, HNCO). MS 32ºC. To ensure saturation, the slurries were stirred in a water
FAB 370.2 (M + H+), 260.0, 238.0, 112.1.
bath at 32ºC for 24 h.

Transdermal penetration of cytarabine
Lesetja J. Legoabe et al.
759
Skin preparation
The female Caucasian human abdominal skin used for these
permeation studies was obtained from Sunwardpark Clinic
(Boksburg, South Africa) where it had been removed during
a cosmetic procedure. The donor was approximately 40 years
old. The skin was stored in a cooler box during transportation
and stored in a freezer at –20ºC until time of use (less than
3 months). A scalpel was used to separate the skin from the
fat layer. Afterwards, the epidermis was removed by first
immersing the skin in 60ºC HPLC water for 60 s.[10] The
epidermis was then gently teased away from the skin with
forceps. The epidermis was placed in a bath filled with
HPLC water, with the outer side facing up, and was then
carefully set on Whatman filter paper, left to dry at room
temperature and wrapped in foil. The foil containing the
epidermis was stored in a freezer at –20ºC and was used
within 3 months. Prior to use, the epidermis was thawed and
examined by light microscope for any defects, before
mounting on the Franz diffusion cells.
Skin permeation determination
Vertical Franz diffusion cells with 2.0 ml receptor compart-
ments and 1.0751 cm2 effective diffusion area was used for the
permeation studies. The epidermis skin layer prepared above
was mounted on the lower half of the Franz cell with the stratum
corneum facing upwards. Subsequent to that, the upper half of
the Franz cell was mounted and vacuum grease applied around
the junction of the upper and lower components to prevent
possible leakage. A clamp was used to fasten the upper and
lower parts of the Franz cell together, with the epidermis
separating the donor and receptor compartments. The receptor
compartment was filled with PBS (pH 7.4). Special care was
taken to ensure that no air bubbles came between the receptor
vehicle and epidermis, as this would reduce the effective
diffusion area. The donor compartment was filled with 1.0 ml
PBS and equilibrated at 32ºC for 1 h in a water bath. Only the
receptor compartments were submerged in the water bath and
were equipped with magnetic stirring bars. After a period of
1 hour, 1 ml of freshly prepared saturated solution was added
to each donor compartment, which was immediately covered
with Parafilm to prevent evaporation of any constituent within
the solution for the duration of the experiment. An excess of
solutes was present in the donor phase at all times during the
experimental procedure to ensure that the donor solutions
remained saturated. After 2, 4, 6, 8, 10 and 12 h the receptor
phase was withdrawn and replaced by fresh buffer (32ºC) to
mimic sink conditions as they occur in the human body. The
withdrawn samples were immediately analysed by HPLC to
determine the concentrations of the compounds which
permeated through the epidermis. These experiments were
performed in triplicate. The steady-state flux (Jss) was
determined by plotting the cumulative drug quantity permeat-
ing per unit area versus time and determining the slope of the
steady state.
Statistical methods
The following statistical procedures were used to test if there
were significant statistical differences between the parent
compound0s median flux value and that of the newly
synthesised compounds. The reason for comparing the

760
Journal of Pharmacy and Pharmacology 2010; 62: 756–761
medians of the flux values instead of the mean was that non-
parametric tests were done because normality of the
distribution of the data of this study could not be assumed.
The usual parametric test to compare means in the case of
samples less than 30 can only be applied when normality is
assumed.[11]
The non-parametric Kruskal–Wallis test was therefore
performed with Statistica[12] to test the statistical significance
of differences between the medians of different compounds
at a 5% level. Multiple comparisons on the mean ranks of
individual groups were performed to determine where the
differences occurred. Note that the compounds where no
fluxes were detected were not included in the statistical
analysis because the conclusion that they were much less
efficient than the parent compound is trivial.
4.00
3.00
2.00
1.00
0.00
1.00
2.00
4.00
3.00
2.00
1.00
0.00
log Sw (PBS 7.4)
log D (7.4)
2
4
6
8
10
Ϫ
Ϫ
Ϫ
1.00
2.00
Ϫ
Number of carbon atoms on alkyl substituents
Figure 3 Relationship between number of carbon atoms on N4-
alkylamide of cytarabine, log D (7.4) and log Sw
place at the N4-amino group. The deshielding of the signal
from 7.1 ppm, assignable to the N4-amino group, to around
10.7 ppm suggests that the N4-amino group is adjacent to an
electron withdrawing carbonyl group. Mass spectra con-
firmed the molecular weights of (2), (3), (4), (5) and (6) as
244.0, 313.3, 341.7, 370.2 and 397.9, respectively.
Results
Synthesis
The N4-alkylamide derivatives of cytarabine (2)–(6) with
molecular weight ranging from 285.3 to 397.47 were
synthesised by the mixed anhydride method in 23–63% yield.
Physicochemical properties
Physicochemical parameters such as aqueous solubility and
lipophilicity have been reported to influence the membrane
permeation, therapeutic activity and pharmacokinetic profile of
medicines.[13] The physicochemical data of cytarabine amide
derivatives are presented in Table 1. In this homologous series,
octanol solubility values increased while aqueous solubility
decreased as the alkyl chain lengthened (see Figure 2).
As a consequence, the log D increases as the chain lengthens
(Figure 3). This corroborates previous findings in the literature
that increasing the non-polar portion of a molecule by
extending the length of the chain produces certain character-
istic features, such as elevation of boiling point, decreased
aqueous solubility and increased partition coefficient.[14] It is
noteworthy that compound (2) exhibited lower solubility in
both water and lipid than the parent compound. Moreover,
amide (2), the first member of the homologous series, exhibits
lesser solubility in octanol than does cytarabine, which is
contrary to previous reports that, generally, the first member of
a homologous series exhibits increased lipid solubility, thought
to be due the masking of hydrogen bond donor.[9] All amides
showed lower melting points than the parent compound.
Generally, compounds with lower melting points are asso-
ciated with higher solubility and are known to permeate the
skin better than those with higher melting points.[15]
Hydrophilicity and lipophilicity
The physicochemical data of cytarabine amide derivatives
are presented in Table 1. The interrelation between octanol
solubility, aqueous solubility and partition coefficient are
presented in Figures 2 and 3.
Skin permeation
The transdermal penetration parameters of cytarabine and its
N4-alkylamide derivatives were determined in vitro and are
summarised in Table 1.
Discussion
Synthesis of cytarabine derivatives
The 13C NMR spectra of amides exhibit the resonance of
carbonyl carbons at around 170 ppm. Due to the electron-
withdrawing capacity of the amide linker, H-5 and H-6
signals were deshielded from 5.6 and 7.5 ppm to 7.1 and
8.0 ppm, respectively. This confirms that the acylation took
3.50
3.00
3.50
Log Sw (µmol/l)
Log Soct (µmol/l)
3.00
2.50
2.50
2.00
2.00
In-vitro skin permeation study
1.50
1.50
The diffusion experiments showed that there is a varying
degree of transdermal permeation between compound (1) and
its N4-alkylamide (Table 1). There is no clear trend between
flux values and molecular weights of compounds. Compound
(4) penetrated the skin much better than (2) and (3), despite
having a higher molecular weight. This corroborates previous
findings in which some members of homologous series with
higher molecular weights penetrated the skin better than
members with lower molecular weights.[16,17]
1.00
1.00
0.50
0.50
0.00
0.00
0
2
4
6
8
10
Ϫ0.50
Ϫ1.00
Ϫ1.50
Ϫ0.50
Ϫ1.00
Ϫ1.50
Number of carbon atoms
Compound (4) exhibited a significant enhancement of
transdermal penetration compared to cytarabine. It showed a
Figure 2 Relationship between aqueous and lipid solubility versus
number of carbon atoms on the alkyl substituent

Transdermal penetration of cytarabine
Lesetja J. Legoabe et al.
761
median steady-state flux of 89.0 nmol/cm2 per h, which is a
significant enhancement. This enhancement may be attribu-
table to the solubility of (4) in both aqueous and lipid media
(Figure 2). A good balance between lipid and aqueous solubility
values is reported to be essential to optimise transdermal
flux.[18] The other compounds, (3), (5) and (6), were not
detected in the receptor phase, which means they did not
permeate the epidermis. The epidermis, especially its outermost
layer, is the major source of resistance to the permeation of the
skin by drug molecules.[3] It has been reported that the optimal
log octanol/water partition coefficient for a drug to penetrate
the stratum corneum is approximately 2.[19] However, in this
study compound (5), with log D of 2.15, exhibited poor
percutaneous penetration. This could be due to inter alia the
hydrogen-bonding phenomenon. The hydrogen bonding func-
tional groups in the permeant tend to slow down transdermal
diffusion to due to their interaction with polar head groups of
the intercellular lipids present in the skin.[20,21]
3. Naik A et al. Transdermal drug delivery: Overcoming the
skin’s barrier function. Pharmaceut Sci Tech Today 2000; 3:
318–326.
4. Hadfield AF, Sartorelli AC. The pharmacology of prodrugs of
5-fluorouracil and 1-b-D-arabinofuranosylcytosine. Adv Phar-
macol Chemother 1984; 20: 21–67.
5. Wipf P, Li W. Prodrugs of ara-C. Drugs Future 1994; 19: 49–54.
6. Aoshima M et al. N4-behenoyl-1-beta-D-arabinofuranosylcyto-
sine as a potential new antitumor agent. Cancer Res 1977; 37:
2481–2486.
7. Immordino ML et al. Preparation, characterization, cytotoxicity
and pharmacokinetics of liposomes containing lipophilic
gemcitabine prodrugs. J Control Release 2004; 100: 331–346.
8. Tokunaga Y et al. Liposomal sustained-release delivery systems
for intravenous injection. IV. antitumor activity of newly
synthesized lipophilic 1-b-D-arabinofuranosylcytosine prodrug-
bearing liposomes. Chem Pharm Bull 1988; 36: 3574–3583.
9. Taylor HE, Sloan KB. 1-alkylcarbonyloxymethyl prodrugs of
5-fluorouracil (5-FU): synthesis, physicochemical properties,
and topical delivery of 5-FU. J Pharm Sci 1998; 87: 15–20.
10. Kligman AM, Christophers E. Preparation of isolated sheets of
human stratum corneum. Arch Dermatol 1963; 88: 702–705.
11. Steyn AGW et al. Modern Statistics in Practice, 2nd edn.
Pretoria: J.L. van Schaik, 1996.
Conclusions
The alkylamide derivatives of cytarabine were successfully
prepared and their structures were confirmed by NMR and
MS techniques. The transdermal data showed no trend in
melting point, lipophilicity and hydrophilicity. The amide
(4), characterised by a good balance of lipophilicity and
hydrophilicity, exhibited the highest enhancement of trans-
dermal penetration of cytarabine.
12. StatSoft Inc STATISTICA (data analysis software system).
2007; 8.1.
13. Goosen C et al. Correlation between physicochemical char-
acteristics, pharmacokinetic properties and transdermal absorp-
tion of NSAIDs. Int J Pharm 1998; 163: 203–209.
14. Abdul HM. Dissolution, bioavailability and bioequivalence,
Easton: Mack, 1989.
15. Cleary GW. Biological factors in absorption and permeation. In:
Zatz, JL, ed. Skin Permeation: Fundamentals and Applications.
Allured: Wheaton, 1993: 300.
Declarations
16. Bonina FP et al. In vitro and in vivo evaluation of polyoxy-
ethylene indomethacin esters as dermal prodrugs. J Control
Release 1995; 34: 223–232.
Conflict of interest
The Author(s) declare(s) that they have no conflicts of
interest to disclose.
17. Bonina FP et al. In vitro and in vivo evaluation of
polyoxyethylene esters as dermal prodrugs of ketoprofen,
naproxen and diclofenac. Eur J Pharm Sci 2001; 14: 123–134.
18. Sloan KB, Wasdo S. Designing for topical delivery: prodrugs
can make the difference. Med Res Rev 2003; 23: 763–793.
19. Riviere JE, Papich MG. Potential and problems of developing
transdermal patches for veterinary applications. Adv Drug Deliv
Rev 2001; 50: 175–203.
Funding
This project was financially supported by National Research
Foundation (NRF), North-West University and Medical
Research Council (MRC) of South Africa.
20. Du Plessis J et al. Physico-chemical determinants of dermal
drug delivery: effects of the number and substitution pattern of
polar groups. Eur J Pharm Sci 2002; 16: 107.
21. Pugh WJ et al. Epidermal permeability–penetrant structure
relationships: 4, QSAR of permeant diffusion across human
stratum corneum in terms of molecular weight, H-bonding and
electronic charge. Int J Pharm 2000; 197: 203–211.
References
1. Hamada A et al. Clinical pharmacokinetics of cytarabine
formulations. Clin Pharmacokinet 2002; 41: 705–718.
2. Rustum YM, Raymakers RA. 1-b-arabinofuranosylcytosine in
therapy of leukemia: preclinical and clinical overview.
Pharmacol Ther 1992; 56: 307–321.