International Journal of Pharmaceutics 447 (2013) 231240

Contents lists available at SciVerse ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsevier.com/locate/ijpharm

Enhancement of skin permeation of bufalin by limonene via reservoir type

transdermal patch: Formulation design and biopharmaceutical evaluation
Zhen Yang a,b, , Yang Teng b , Hao Wang b , Huimin Hou b
a
Department of Pharmacological and Pharmaceutical Science, College of Pharmacy, University of Houston, 77030, USA
b
National Pharmaceutical Research Center, Shanghai Institute of Pharmaceutical Industry, Shanghai 201203, China

a r t i c l e i n f o a b s t r a c t

Article history: A reservoir-type transdermal delivery system (TDS) of bufalin was designed and evaluated for various for-
Received 2 December 2012 mulation variables like different penetration enhancers, formulation matrix, rate controlling membranes
Received in revised form 12 February 2013 as well as biopharmaceutical characteristics. Hairless mouse skin was used in permeation experiments
Accepted 22 February 2013
with Franz diffusion cells. In vitro skin permeation study showed that terpenes, especially d-limonene
Available online 1 March 2013
was the most effective enhancer when ethanol and PG were used as the vehicle with a synergistic effect.
Among different rate controlling membranes, ethylene vinyl acetate (EVA) membrane containing 19%
Keywords:
vinyl acetate demonstrated a more suitable release rate for bufalin than the other membranes. In vivo
Transdermal delivery system
Bufalin
pharmacokinetic study of the bufalin patch in rat showed steady-state of bufalin from 3 h to 12 h. In
d-Limonene vivo release rate and cumulative amount analyzed by deconvolution method demonstrated the sus-
Pharmacokinetics tained release of bufalin as long as the patch remained on the animal for at least 12 h. The MRT increased
Deconvolution from 1 h of IV administration to 9 h of transdermal administration. In vitro permeation across mouse
Biopharmaceutics skin was found to have biphasic correlation with plasma AUC in the in vivo pharmacokinetic study.
Current in vitroin vivo correlation (IVIVC) enabled the prediction of pharmacokinetic prole of bufalin
from in vitro permeation results. In conclusion, current reservoir transdermal patch containing 10% d-
limonene as a permeation enhancer, 40% ethanol, 30% PG and 15% carbopolwater gel complex provided
an improved sustained release of bufalin through transdermal administration. The bufalin patch was suc-
cessfully applied to biopharmaceutical study in rats and demonstrated the feasibility of this transdermal
formulation for future development and clinical trials.
2013 Elsevier B.V. All rights reserved.

1. Introduction
Bufalin is a major component of bufadienolides extracted from
Chinese traditional medicine Chansu (toad venom). It is widely
used as one of the key components in many traditional Chinese
medicines, i.e., Liu Shen Wan (Hong et al., 1992) and Shexiang
Baoxin Wan (Song et al., 2000). Bufalin exhibits potent anti-tumor
activities against liver cancer (Zhang et al., 2012), lung cancer (Zhu
et al., 2012), gastric cancer (Li et al., 2009), as well as prostate can-
cer (Yu et al., 2008). Bufalin has been reported to induce apoptosis
in various cancer cells through both intrinsic and extrinsic path-
ways as elaborated in several publications (Hong and Choi, 2012;
Yan et al., 2012; Zhu et al., 2012) and enhance the cytotoxicity of
several anticancer drugs (Hashimoto et al., 1997). Because of dual
Corresponding author at: Department of Pharmacological and Pharmaceutical
Science, College of Pharmacy, University of Houston, 77030, USA.
Tel.: +1 832 403 6299.
E-mail address: zyang9@uh.edu (Z. Yang).
effects of anti-cancer and pain relief activity, bufalin was approved
as anticancer drugs in China and was used to treat late-stage of
liver and lung cancers (Meng et al., 2009; Qin et al., 2008). The
clinical study of bufalin was also conducted in the United States
for the treatment of pancreatic cancer and showed potential ther-
apeutic effect (Meng et al., 2009; Wang et al., 2011). Considering
low survival rate of liver and lung cancer in late-stage, alternative
medicines obtained from herb materials and animals may provide
valuable benets.
However, the clinical application of bufalin on cancer treat-
ment was greatly limited by its unfavorable biopharmaceutical
properties and cardiac toxicity. It has been reported to have poor
solubility, short half-life and narrow therapeutic window upon
intravenous administration (Yin et al., 2012). The short half-life lim-
its its capability to maintain minimal plasma drug concentration to
achieve therapeutical effect and demands multiple times admin-
istrations. Furthermore, bufalin belongs to toxic steroids, which
produce digoxin-like cardiac toxicity (Wang et al., 1997). Bufalin
has side effect of cardiac arrhythmia, breathlessness, convulsion
and coma at high dosage, which may cause safety issues (Panesar,
1992).

0378-5173/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ijpharm.2013.02.048
232 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240
Fig. 1. The chemical structure of bufalin.
Currently developed formulation of bufalin include wheat
germ agglutinin-grafted lipid nanoparticles (Liu et al., 2011), self-
microemulsing lipid-based formulation (Liu et al., 2010) and
PEGPLGA loaded nanoparticles (Yin et al., 2012). Although they
have achieved some successes on improving solubility and overall
bioavailability compared to traditional tablet and injections (Yang
et al., 2008), the major issues including high dose derived car-
diotoxicity and rapid decline of exposure still exist. Transdermal
formulation was considered as an enabling strategy for this sce-
nario by offering both controlled and sustained release patterns. A
number of FDA approved transdermal products, like Transderm-
NitroTM , EstradermTM , DuragesicTM , D-TRANSTM , AndrodermTM
and Catapres-TTSTM had demonstrated the successful application
of transdermal administration.
To the best of our knowledge, there is no transdermal formu-
lation of bufalin that has been developed and evaluated. Bufalin
is a small molecular compound (molecular weight = 386.5) with a
suitable lipophilicity (log P = 2.78). However, its poor aqueous sol-
ubility (33 g/ml) may limit permeation force through dermis (Liu
and Feng, 2008). The chemical structure of bufalin is shown in Fig. 1.
For drugs with a short biological half-life and/or a narrow thera-
peutical range, transdermal administration is one of the attracting
alternative approaches to decrease toxicity and prolong therapeu-
tical action. Therefore, the skin has become an important route for
drug delivery for topical, regional or systemic effects for the last
two decade (Paudel et al., 2010). The major challenges of develop-
ing transdermal drug delivery lie on the poor permeation of API and
its potential skin irritation. Using chemical enhancers is the most
common strategy to improve the permeation of drugs through the
skin barrier.
Therefore, the objective of this study can be divided into
three aims, (1) develop a transdermal delivery system for bufalin
administration, (2) evaluate its biopharmaceutical characteristics
to ensure sustained release of bufalin in vivo, (3) perform in vitroin
vivo correlation and use in vitro data to predict and guide pharma-
cokinetics in future.
2. Materials and methods
2.1. Materials
Bufalin was extracted and puried in our lab from Traditional
Chinese Medicine, Chansu. The puried bufalin powder had been
identied by NMR and Mass Spectrometry and quantied by HPLC
with a purity >98% in our previous study (Yang et al., 2008).
Fig. 2. The schematic preparation of reservoir-type bufalin patch.
3MTM CoTranTM 9728, 3MTM CoTranTM 9705, 3MTM CoTranTM 9715
and other types of ethylene vinyl acetate (EVA) membrane with dif-
ferent vinyl acetate content 2%, 9%, 19%, 28% were obtained from
3M Pharmaceuticals (St. Paul, MN, USA). Diphenhydramine, oleic
acid, lauric acid, sodium lauryl sulphate (SLS), azone, ethyl oleate,
ethyl lauric acid, isopropyl myristate (IPM), Tween 80, N-methyl-
2-pyrrolidone (NMP), ethanol, propylene glycol (PG), 1,8-cineole,
d-limonene and l-menthol were purchased from Sigma (St. Louis,
MO, USA). Carbopol 934 P (pharmaceutical grade) was purchased
from Goodrich Co. Ltd., (Cleveland, OH, USA). All other chemicals
were obtained from Shanghai Chemical Reagents Co. (Shanghai,
China). Aluminum foilpolyethylene backing layer was purchased
from Shanghai Package Material Company (Shanghai, China). Heat-
sealing machine was purchased from Yiheng Inc. (Shanghai, China).
Transdermal permeation diffusion instrument was purchased from
Jiekai Inc. (Shanghai, China).
2.2. Preparation of transdermal delivery system (TDS)
Reservoir-type TDS was prepared with an active membrane
release area of 15 cm2 . The schematic of TDS preparation in the
present study is shown in Fig. 2.
2.2.1. Preparation of transdermal gel of bufalin
40 mg bufalin was rstly fully dissolved in 4 ml ethanol, and then
3 ml PG added. 1 ml d-limonene (10%, w/v of mixture) was added
as a penetration enhancer to drug solutions, and then 0.1 ml Tween
80 added as a surfactant to form a uniform phase. 1 g carbopol used
as gelling agent was swelled in 50 ml double distilled water (form
2% carbopol gel) overnight while pH value was adjusted to 7.4. 1.5 g
carbopolwater gel (2%) was added to drug solutions (15%, w/v) and
stirred well to form a uniform gel. The uniformity of drug content
(4 mg/ml) was tested for all batches of gels by HPLC.
2.2.2. Preparation of reservoir transdermal patch
A procedure for drug-loaded-membrane laminate was divided
into three main phases. Phase 1 involved lling of drug reservoir.
1 ml bufalin gel was squeezed to the EVA membrane from peri-
staltic pump and then the backing layer was used to cover the
rate-controlling membrane on the top of the druggel complex. The
 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240
drug content uniformity was ensured to be within 5% of total dose
(4 mg/ml). Phase 2 involved heat-sealing of the backing layer using
the heating-sealing machine. The backing layer was placed on the
top of the reservoir, and the heater which was preheated at 90 C
was forced down pneumatically for 10 s to form a reservoir patch
with 15 cm2 (3 cm 5 cm). Phase 3 involved completing patch with
the adhesive layer. The reservoir was placed on the adhesive layer
with EVA membrane upside and covered with anti-stick liner. An
entire bufalin patch is shown at the end of Fig. 2.
2.3. In vitro skin permeation studies
The dorsal skins of hairless mice (2025 g) were excised after
sacrice by cervical dislocation. Fresh prepared skins were stored in
a refrigerator at 20 C without repeatable freeze and thaw cycles.
The storage of skin was limited within 1 month otherwise new
batch of skin would be prepared. Prior to permeation experiments,
skin was thawed and subcutaneous fat, tissue and capillaries of
skin were carefully removed. After cutting into pieces, skin was
mounted between the donor and receptor compartments of the
Franz diffusion cells with the stratum corneum facing the donor
compartment.
The permeation area of Franz diffusion cells was 2.84 cm2 and
a receiver volume was 7.0 ml. 40% PEG and 30% PG aqueous solu-
tion were used as the receiver medium and 50% PG solution was
used as donor solution with 4 mg/ml bufalin. Assembled diffusion
cells in triplicate were placed in a transdermal permeation diffu-
sion instrument and maintained isothermally at 37 C. The receptor
compartment was stirred with a magnetic stirrer at 250 rpm. Care
was taken to ensure that no air bubbles remained in the receiver
side. Samples (0.1 ml) were withdrawn at 0.5, 1, 2, 4, 6, 8 and 10 h
for HPLC analysis and were replaced with an equivalent volume. To
minimize evaporation, the donor compartment was covered with
paralm and aluminum foil.
In vitro permeation prole of transdermal patch was performed
to correlate with in vivo pharmacokinetic behavior. Intact bufalin
patches were mounted on Franz diffusion cells across hairless
mouse skin and the experiment procedure is similar to in vitro skin
permeation study.
2.4. Data analysis
The cumulative amount (Q, g/cm2 ) of bufalin permeated
through skin was calculated by the following equation.
n1
Cn V + 0.1 i=1
Ci
Q =
S
where S is the effective area, V is the volume of receptor cell, Cn
is the bufalin concentration at time point n, and Ci is the bufalin
concentration at time point i.
The cumulative amount (g/cm2 ) of bufalin permeated through
skin was plotted versus time (h). Each data is the mean SD of three
determinations. JSS represents the best-t slopes of the apparent
linear portions, usually several data points in the beginning were
removed during tting. KP is the apparent permeability coefcient
and is calculated according to Ficks rst law of diffusion,
JSS
KP =
Co
where JSS is steady state ux and Co is the initial drug concentration
in donor side. Tlag was determined by extrapolating the straight-
line portion of each steady-state permeation curve to the time axis.
The ux data were subjected to student T-test and one-way analysis
of variance (ANOVA) to determine the level of signicance. The data
was considered to be signicant at P < 0.05.
233
2.5. In vivo pharmacokinetic study in rats
The animal protocols used in this study were approved by Insti-
tutional Animal Care and Uses Committee from Shanghai Institute
of Pharmaceutical Industry. Male Sprague-Dawley rats weighing
250 20 g were used for pharmacokinetic studies. Each pharma-
cokinetic study was performed on three rats and blood samples
(200250 l) were collected via retro-orbital plexus using a ster-
ilized glass capillary tube. For i.v. bolus pharmacokinetic study,
a dose of 100 g/ml bufalin was prepared in 10% alcohol saline
solution. The injection solution was ltered through the 0.45 m
membrane and sterilized at 115 C for 30 min in a sealed ampoule.
Blood samples were drawn at 0.083, 0.167, 0.333, 0.667, 1, 2, 4, 6
and 8 h following the intravenous administration. For transdermal
administration, bufalin patches (3 cm 5 cm, 4 mg) were applied to
the shaved abdomen of rats. Rats were carefully shaved by electric
shaver one day before transdermal administration. The blood sam-
ples were collected at 0.5, 1, 2, 4, 6, 8, 10, 12, 13, 14 and 24 h while
the patches remained on the animal up to 12 h. The blood samples
were immediately centrifuged at 10,000 rpm for 10 min, and the
plasma was separated and stored at 20 C until analysis.
2.6. Pharmacokinetic and deconvolution analysis
WinNonlin 5.3 (Pharsight Corporation, Mountain View, CA)
was used for bufalin pharmacokinetic analysis. Both compartment
and noncompartmental model were applied for the pharmacoki-
netic analysis of bufalin proles. IV pharmacokinetic analysis was
performed on compartmental model and transdermal administra-
tion was conducted on noncompartmental model. Pharmacokinetic
parameters, including Cmax , Tmax , half-life, mean residence time
(MRT), CL, Vss , AUC and AUMC were derived directly from WinNon-
lin. Deconvolution analysis was also performed to estimate in vivo
bufalin release from transdermal batch (Sun et al., 2012). IV does
pharmacokinetics (100 g) was used as a reference formulation
for an impulse-response function. Pharmacokinetics of transder-
mal patch was analyzed by deconvolution methodology to obtain
input-response function, i.e., in vivo release rate and cumulative
release amount. The number of predicted values was 200 and iter-
ation number was 100. Maximum number of UIR (Unit Impulse
Response) exponentials was set at 3 and model selection was based
on Akaike Information Criteria (AIC) (Bozdogan, 2000).
2.7. In vitroin vivo correlation
The cumulative amount of bufalin permeated across mouse skin
(1) in vitro from bufalin batch was compared against the extent of
absorption and cumulative AUC values to establish in vitroin vivo
correlation (IVIVC). This correlation represents a point-to-point
relationship with the same time point (Type A).
2.8. Sample preparation and quantitative determination of
bufalin
Analysis of bufalin in rat plasma was performed by API 3000
triple-quadrupole mass spectrometer coupled with Shimazdu LC-
20 AD HPLC system. The quantication was performed in a
positive ion mode using multiple reaction monitoring (MRM) with
(2) 387.4/351.2 (m/z) for bufalin and 256.2/167.0 (m/z) for diphen-
hydramine (Internal standard). The other instrument dependent
parameters were as following: ionization gas, 11 L/min; curtain
gas, 10 L/min; collision gas, 7 L/min; ion voltage, 4500 V; ion tem-
perature, 400 C. The HPLC conditions for analyzing bufalin in rats
plasma: analytical column, C18 Diamonsil (50 mm 2.1 mm I.D.
5 m), and the mobile phase was isocratic using methanolwater
(70:30, 5 mM ammonium acetate) with a ow rate of 0.2 ml/min.
234 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240
Cumulative amounts of permeated bufalin

Cumulative amounts of permeated bufalin

4.0 oleic acid D-li monene no enhancer 40
PG/water (50/50)
lauric acid 1,8-cineole
3.5 PG/PEG400/water (10/40/50)
IPM/NMP L-menthol
3.0 azone ethyl lauric acid 30 ethanol/PG/water (40/30/30)

2.5 SLS ethyl oleate

(g/cm2)

(g/cm 2)
2.0 20
1.5
1.0 10
0.5
0.0
0 2 4 6 8 10 12 0
0 2 4 6 8 10 12
Time (h) Time (h)
Fig. 3. Effect of skin permeation enhancers (5%, w/v) on the permeation of bufalin
Fig. 4. Permeation proles of bufalin in different vehicle solutions. Data were pre-
in PG/water cosolvent system (n = 3).
sented as the mean SD (n = 3).

The chromatography was performed at 30 C and the injection

signicantly increased the permeation rate of bufalin from 0.021
volume was 5 l. Analysis of bufalin in permeation samples was
to 0.437 g/cm2 /h (22-fold). Permeation parameters of bufalin
performed by reversed phase HPLC using a Shimazdu LC-10ADvp
with 5% (w/v) of various enhancers across hairless mouse skin at
HPLC system, which was validated in our previous publication
37 C are shown in Table 1.
(Yang et al., 2008). LCMS/MS method validation including intra-
and inter-day variation, stability, linearity and matrix effect were
performed with the same methodology shown before (Yang et al., 3.2. Effect of vehicle selection on the skin permeation of bufalin
2010).
A volume of 25 l of methanol (containing 200 ng/ml diphenhy- In order to investigate the synergistic effect of different
dramine) was added to 100 l aliquot of rat plasma samples. 1 ml vehicles on the bufalin permeation with d-limonene, the accu-
ether was added to extract and vertex for 5 min. After centrifuge at mulated amount of bufalin permeated across hairless mouse skin
3000 rpm for 10 min, the supernatant was evaporated to dryness at from saturated solutions with PG/PEG 400/water (10/40/50, v/v),
40 C under nitrogen and reconstituted in 100 l of mobile phase. A ethanol/PG/water (40/30/30, v/v) and PG/water (50/50, v/v) is
5 l portion of each sample was injected into the LCMS/MS system shown in Fig. 4. Ethanol/PG/water (40/30/30, v/v) showed 9 times
for analysis. higher accumulative permeated amount of bufalin compared to
PG/water (50/50, v/v) (P < 0.05). Permeation parameters of bufalin
2.9. Skin irritation across hairless mouse skin with different vehicle solutions at 37 C
are shown in Table 1.
Evaluation of skin irritation was performed by visual obser-
vation after pharmacokinetic and pharmacodynamic studies in 3.3. Effect of enhancer concentration on the skin permeation of
Sprague-Dawley rats. Histological evaluations will be performed bufalin
if obvious redness and swelling are observed.
To estimate the enhancement of the different concentrations
3. Results of d-limonene, various concentrations (5%, 7%, 10%, 15%) were
added to the ethanol/PG/water (40/30/30) cosolvent system. Per-
3.1. Effect of permeation enhancers on the skin permeation of meation of bufalin increased proportionally as the concentration
bufalin of d-limonene increased from 5% up to 10% (Fig. 5). The results

The control study using nicotine as a model compound was

Cumulative amounts of permeated bufalin

75 5%D-limonene
performed and ux results were similar to those published in liter-
ature (Zorin et al., 1999). Except for a short lag time in the beginning 7%D-limonene
of permeation study, the cumulative permeation amount of bufalin 10%D-limonene
proportionally increased with time up to 10 h, indicating a good
50 15%D-limonene
integrity of skin during permeation study. Bufalin showed an
(g/cm2)

intrinsic ux of 21 ng/cm2 /h across hairless mouse skin when 50%

PG was used in donor side. The low permeation rate indicated
that chemical enhancers were needed for transdermal admin-
25
istration. The feasibility of using oleic acid, lauric acid, sodium
lauryl sulphate, ethyl oleate, ethyl lauric acid, isopropyl myristate,
N-methyl-2-pyrrolidone, azone, 1,8-cineole, d-limonene and
l-menthol to reduce permeation barrier was investigated. The
0
enhancement on the skin permeation rate of bufalin by adding 5%
0 2 4 6 8 10 12
(w/v) of various enhancers to the cosolvent system is shown in
Fig. 3. Compared to terpenes including d-limonene, l-menthol and Time (h)
1,8-cineole, other types of chemical enhancers increased the ux
Fig. 5. Effect of d-limonene concentration (%, w/v) on the permeation of bufalin
of bufalin within 5-fold. Terpenes, especially d-limonene, were saturated in ethanol/PG/water (40/30/30) cosolvent system. Data were presented
the most effective among ten different chemical enhancers, which as the mean SD (n = 3).
 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240 235

Table 1
Permeation parameters of bufalin through excised hairless mouse skin from 4 mg/ml bufalin solution with different formulation variables at 37 C (n = 3).

Formulation # Membrane Carbopol Matrix Enhancer JSS (g/cm2 /h) KP (103 , cm/h) Tlag (h) ER
gel

1 PG/water (50/50) Control 0.02 0.01 0.01 0 1.92 0.42

2 PG/water (50/50) 5% oleic acid 0.10 0.01* 0.02 0.0* 2.32 0.35 5.1
3 PG/water (50/50) 5% lauric acid 0.11 0.04* 0.04 0.01* 2.89 0.07 5.2
4 PG/water (50/50) 5% sodium lauryl sulfate 0.05 0.01 0.01 0.0 3.97 0.21* 2.5
5 PG/water (50/50) 5% azone 0.05 0.01 0.02 0.0 2.76 0.19 2.4
6 PG/water (50/50) 5% ethyl oleate 0.03 0.0 0.01 0 2.27 0.23 1.4
7 PG/water (50/50) 5% ethyl lauric acid 0.11 0.04* 0.03 0.01* 2.25 0.03 5.3
8 PG/water (50/50) 5% IPM/NMP 0.04 0.01 0.01 0.0 2.63 0.26 2.0
9 PG/water (50/50) 5% 1,8-cineole 0.35 0.04* 0.08 0.01* 4.08 0.25* 17.1
10 Ethanol/PG/water (40/30/30) 5% d-limonene 3.86 0.53* 0.99 0.06* 2.86 0.21 183.5
11 PG/PEG400/water (10/40/50) 5% d-limonene 0.59 0.12* 0.15 0.01 1.94 0.32 28.1
12 PG/water (50/50) 5% d-limonene 0.44 0.09** 0.11 0.01* 4.11 0.46* 22.2
13 Ethanol/PG/water (40/30/30) 7% d-limonene 5.82 0.54** 1.32 0.09* 2.95 0.32 277.6
14 Ethanol/PG/water (40/30/30) 10% d-limonene 8.73 1.35** 1.57 0.17* 2.04 0.43 415.7
15 Ethanol/PG/water (40/30/30) 15% d-limonene 8.92 1.54** 1.64 0.21* 2.32 0.37 424.7
16 PG/water (50/50) 5% l-menthol 0.32 0.06* 0.08 0.0* 4.16 0.38* 15.3
17 15% Ethanol/PG/water (40/30/30) 10% d-limonene 6.84 1.14** 1.44 0.25* 4.02 0.27* 325.7
18 25% Ethanol/PG/water (40/30/30) 10% d-limonene 3.62 0.56** 0.99 0.16* 4.16 0.31* 172.3
19 40% Ethanol/PG/water (40/30/30) 10% d-limonene 0.48 0.14* 0.12 0.03* 4.24 0.32* 24.0
20 19% EVA 15% Ethanol/PG/water (40/30/30) 10% d-limonene 5.66 1.01** 1.33 0.20* 4.80 0.42* 269.3

Note: Donor side concentration of bufalin was assumed as 4 mg/ml in various donor solutions to facilitate calculation and they could be supersaturated solutions.
Data are the means SD of three independent experiments.
ER is the enhancement ratio calculated by JSS of treatment versus control.
*
P < 0.05 compared to control group.
**
P < 0.01 compared to control group.

showed the permeation enhancement is concentration-dependent, 3.5. Effect of gelling agent

and achieved the highest permeation effect at 10%. As no signicant
difference on the effects of permeation enhancement between 10% In order to increase the viscosity of the reservoir solution to facil-
and 15%, 10% d-limonene was chosen as the appropriate concentra- itate manufacture, Carbopol 934 P was added in the solution since
tion of enhancer. Permeation parameters of bufalin across excised it is known as a thickening agent or viscosity modier, which has
hairless mouse skin with various concentration of d-limonene at been mainly used in gels, suspensions and emulsions. After dissolv-
37 C are shown in Table 1. ing bufalin in ethanol/PG (40/30, v/v) solution with 10% d-limonene
and 1% Tween 80, the effect of different volumes of carbopol gel on
3.4. Effect of membrane on the controlled release of bufalin the permeation proles across the EVA membrane/skin compos-
ite was studied. As shown in Fig. 7, the permeation rate of bufalin
To control the release of bufalin from the reservoir com- decreased from 6.84 to 0.48 g/cm2 /h when the concentration of
partment, EVA copolymer membrane was selected as the Carbopol 934 P increased from 15% to 40% (w/v). When the content
rate-controlling membrane. EVA copolymer membranes used in of gelling polymer is less than 15%, the viscosity of bufalin solution
current study contained various content of vinyl acetate [2%, 9%, is not sufcient to remain in the center of backing layer during heat
19%, 28% (w/w)] with the uniform thickness of 76.20 m. When sealing. The results showed the addition of 15% (v/v) of carbopol
bufalin was dissolved in ethanol/PG/water (40/30/30) containing gel provided desirable rheological properties with the insignicant
10% (w/v) d-limonene, the permeation proles of bufalin across reduction in bufalin permeation rate.
the EVA copolymer membrane are shown in Fig. 6. The membrane
permeation rate of bufalin increased as the content of vinyl acetate
increased from 2% up to 28%.
Cumulative amounts of permeated bufalin

200
Cumulative amounts of permeated bufalin

2%vinyl acetate 19%vinyl acetate

45 15% carbopo l gel
40 25% carbopo l gel
9%vinyl acetate 28%vinyl acetate
150 35
40% carbopo l gel
30
(g/cm2)
(g/cm2)

25
100
20
15
50 10
5
0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Time (h) Time (h)

Fig. 6. Permeation proles of bufalin across the EVA membrane with various content Fig. 7. Permeation proles of bufalin through membrane/skin composite with var-
of vinyl acetate. Data were expressed as the mean SD (n = 3). ious volume of carbopol gel. Data were expressed as the mean SD (n = 3).
236 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240

Table 3
Cumulative amounts of permeated bufalin
45 Pharmacokinetic parameters of bufalin after IV and transdermal administration in
no adhesive layer rats (n = 3).
40
with adhesive layer Parameters Transdermal IV
35
30 Cmax (ng/ml) 174.77 45.24 657.81 125.43
Tmax (h) 8.0 2.0
(g/cm2)

25 MRT (h) 9.17 1.05 1.09 0.09

T1/2 (h) 3.15 0.63 0.76 0.06
20 Cl (ml/h) 143.63 35.78
15 Vss (ml) 156.02 31.62
AUC0 (h ng/ml) 2093.66 566.46 722.38 157.41
10 AUMC0 (h h ng/ml) 19,190.66 5848.01 796.28 210.79
5 Data are the means SD of three rats.

0
0 2 4 6 8 10 12
Time (h) aqueous samples. There was no interference in rat plasma regarding
to determination of bufalin, and the chromatograms are shown in
Fig. 8. Permeation proles of bufalin through membrane/skin composite with poly- Supplemental Fig. 1.
acrylic acid pressure sensitive adhesive. Data were presented as the mean SD
(n = 3).

3.8. Pharmacokinetic and deconvolution analysis of bufalin patch

3.6. Effect of pressure sensitive adhesive on the permeation of
bufalin
The EVA membrane containing 19% vinyl acetate was chosen as
controlled release membrane, and pressure sensitive adhesive con-
taining polyacrylic acid (total contents = 55%) was coated with the
thickness of 5.0 m. Permeation proles of bufalin through the EVA
membrane coated with and without pressure sensitive adhesive are
shown in Fig. 8. Pressure sensitive adhesive with polyacrylic acid
was usually used as the adhesive agent to keep the close contact
between the patch and skin. The permeation rate of bufalin through
the adhesive layer was much lower than that without the adhesive
layer (only 15% of the original rate). This may be due to the interac-
tion between the carbopol gel and polyacrylic acid, which resulted
in the barrier between the adhesive layer and skin. Therefore, pres-
sure sensitive adhesive (polyacrylic acid type) was coated on the
peripheral region of the patch to avoid decreasing the skin per-
meation of bufalin. The nal formulation composition is shown in
Table 2.
3.7. Analytical method validation
The method validation was conducted in blank rat plasma. The
standard curves of bufalin were linear in the concentration range
from 2.5 to 1000 ng/ml, and the lower limit of quantitation was
2.5 ng/ml in rat plasma. 1/x2 was used a weighting factor to quantify
bufalin plasma concentration. Linear regression showed R2 value
equals to 0.999. The intraday and interday precisions of QC samples
at three concentration levels (20, 200, and 1000 ng/ml) are shown
in Supplemental Table 1. Rat plasma samples containing bufalin
showed within 15% CV in short-term (8 h at 20 C), long-term (4
weeks at 20 C) and three freezethaw cycle stability study. No
obvious matrix effect was detected in plasma samples compared to
in rats
The pharmacokinetic studies were conducted in rats for intra-
venous (100 g/ml) and transdermal routes (4 mg/ml). The mean
plasma concentrationtime of bufalin after IV and transdermal
administration is presented in Fig. 9. A summary of the pharmacoki-
netic parameters is shown in Table 3. Pharmacokinetic parameters
for IV administration were tted by both one and two compartment
models. As seen in Fig. 9A, bufalin did not differentiate distribution
phase and elimination phase obviously after IV administration and
the concentration at 8 h was below detection limit. Based on AIC
value, one compartment model showed smaller AIC values (60)
than two compartment models (90), therefore one compartment
model was used to estimate parameters for IV administration.
IV dose of bufalin showed that bufalin has a very short half life
(0.8 h) and MRT (1 h). The clearance was estimated at 143.63 ml/h
and steady-state volume distribution was 156 ml. For transder-
mal administration, non-compartment model was used. In vivo
pharmacokinetic study showed steady state around 3 h. MRT of
transdermal administration (9.2 h) is much longer than IV admin-
istration (1.1 h). The plasma concentrations of bufalin remained
above 100 ng/ml from 3 h until patch was removed at 12 h.
Deconvolution analysis showed that patch was able to provide
sustained release of bufalin from 2 h up to 12 h. The estimated
release rate achieved peak at 4 h and it was consistent with the
lag time shown at in vitro permeation study (lag time = 4 h). After
8 h, the estimated release rate gradually declined but still main-
tained at a relatively stable level (Fig. 9B). The predicted in vivo
release amounts showed almost linear increase after 2 h transder-
mal administration (Fig. 9C), which was also consistent with both
in vitro and in vivo behaviors of the bufalin batch. Both pharmacoki-
netic and deconvolution results demonstrated that reservoir type
transdermal patch of bufalin signicantly improve the sustained
release of bufalin in vivo and the decreased dose frequency.

Table 2
Formulation composition of bufalin transdermal patch (%, w/v). 3.9. In vitroin vivo correlation
Components Contents
IVIVC between the cumulative % of bufalin permeated across
Bufalin 4 mg
mouse dorsal skin and plasma AUC at the same time point showed
Carbopol gel 15%
Ethanol 40%
a biphasic curve pattern as shown in Fig. 10, which can be clearly
PG 30% distinguished into two stages. Good linear correlation was observed
d-Limonene 10% with correlation coefcients, R2 = 0.999 during the initial phase
Tween-80 1% (Fig. 10B) and R2 = 0.972 during steady-state (Fig. 10C). The mean
Water 4%
AUC of bufalin in vivo was in agreement with the observed perme-
The total volume of bufalin transdermal patch is 1 ml. ation prole in vitro.
 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240 237

B
40
35

In vivo release rate (g/h)

30
A I.V. 25
2000 Patch 20
Plasma Concentraons (ng/ml) 15
Remove the
patch at 12 h 10
200 5
0
0 2 4 6 8 10 12
20 Time (h)

C 250

In vivo cumulave release

2 200
0 4 8 12 16 20 24

amounts (g)
Time (h)
150

100

50

0
0 2 4 6 8 10 12
Time (h)

Fig. 9. Plasma concentrationtime proles of bufalin after IV and transdermal administration. Data are presented as mean SD. The doses for intravenous and transdermal
routes were 100 g and 4 mg, respectively.

3.10. Skin irritation Sprague-Dawley rats. No obvious redness and swelling were found
on skin after 12 h patch administration. Bufalin was not thought
Evaluation of skin irritation was performed by visual obser- to be a skin irritant based on present transdermal patch study in
vation after pharmacokinetic and pharmacodynamic studies in animal model.

A 1600

1200
AUC (ng/ml*hr)

800

400

0
0 200 400 600 800

Amount permeated (g) in vitro

B 120 C 2000

100 y = 2.481x + 1.200 y = 2.030x + 307.6

1600
R = 0.998 R = 0.972
AUC (ng/ml*hr)

AUC (ng/ml*hr)

80
1200
60
800
40

20 400

0 0
0 10 20 30 40 0 200 400 600 800
Amount permeated (ug) in vitro Amount permeated (ug) in vitro

Fig. 10. In vitroin vivo correlation of cumulative amount permeated in vitro versus plasma AUC in Sprague-Dawley rats (n = 3). It is a type A IVIVC (point to point).
238 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240
4. Discussion
This is the rst systematic study of development and biophar-
maceutical evaluation of bufalin via transdermal administration.
Transdermal formulation as an alternative delivery route for
bufalin is proposed based on the following perspectives. Firstly,
bufalin is well known for its cardiotoxicity, and current formula-
tions, i.e., tablet and injection, usually cause serious side effect.
Transdermal formulation provides constant and controlled release
of drug which may solve this issue and keep plasma concentration
in the desired therapeutic window. Secondly, physicochemical
properties of bufalin indicate that it might be suitable for transder-
mal administration when its molecular weight is less than 500 and
log P value is around 3. Thirdly, bufalin is a potent therapeutical
agent which makes transdermal formulation viable due to low
dose requirement. It showed strong induction effect on differenti-
ation of human leukemia at as low as 10 nM (Kawazoe et al., 1999;
Zhang et al., 1992). Other reports showed that bufalin effectively
inhibits proliferation of gastric, prostate cancer cells and other
cancer cells within the range of several hundred nanomolar (Dong
et al., 2011; Li et al., 2009; Yeh et al., 2003). Our pharmacodynamic
study also showed no signicant cardiotoxicity of bufalin up to
1500 nM of plasma concentration in rodents (data not shown).
Therefore, the current bufalin plasma concentration is perfectly
tted in the therapeutic window.
Although other chemical enhancers also increased permeation
of bufalin, the chemical enhancer screening result implied that d-
limonene is the most potent enhancer by increasing 22 times higher
permeation rate of bufalin than control. As a result of the synergy
between d-limonene and ethanol, the enhancement was greater
than 400 times. d-Limonene is included in the list of Generally
Recognized As Safe (GRAS) agents listed by US FDA (FDA) and 10%
d-limonene was not found to cause any signicant pharmacological
effect (Fox et al., 2011).
d-Limonene is known to have a potent skin permeation-
enhancing effect when used alone or in combination with PG
(Williams and Barry, 1991). It can enhance both hydrophilic drugs
including propranolol, sumatriptan succinate and lipophilic drugs
such as testosterone, ketoprofen, indomethacin and estradiol
(Amnuaikit et al., 2005; Williams and Barry, 2004). A proposed
mechanism to improve the skin permeation of drugs is mainly the
increase in drug diffusivity in the skin by modifying the intercellu-
lar packing and disrupting highly ordered structure of lipids (Vaddi
et al., 2002; Williams and Barry, 1991). These effects appear to
involve the modication of stratum corneum by solvent, improving
drug partitioning into the tissue. Some authors have established
certain drug-specic structureactivity relationship for these ter-
penes (Drakulic et al., 2008; Ghafourian et al., 2004) based on the
physicochemical properties and conformations of the complexes
(drug + enhancer). According to this hypothesis, that hydrocarbon
or the non-polar group containing terpenes, such as limonene pro-
vide better enhancement for lipophilic permeates, and conversely,
the polar group containing terpenes (such as menthol, 1,8-cineole)
provide better enhancement for hydrophilic permeates (Williams
and Barry, 2004). In our experiments, d-limonene showed the high-
est enhancing activity on the transdermal absorption of bufalin,
which may be explained by the above mechanism. In order to better
homogenize limonene in the aqueous matrix, we added 1% Tween
80 to the formulation. We did not nd any signicant permeation
difference between the formulation with and w/o Tween 80.
Ethanol is one of the most widely used solvents in many
transdermal formulations. The synergistic effect of ethanol with
d-limonene is better than PG can be interpreted by several
mechanisms that have been proposed for ethanol permeation
enhancing activity (Panchagnula et al., 2001). Firstly, it may
increase the solubility of bufalin in the vehicle compared with
PG, thus increasing concentration gradient of permeation between
donor compartment and receiver compartment. Secondly, per-
meation of ethanol into the stratum corneum can alter the
solubility properties of the tissue with a consequent improve-
ment for drug partitioning into the membrane. Additionally,
ethanol may extract some of the lipid fraction within stratum
corneum when used at high concentration for prolonged times,
and thus improve drug ux through skin (Krishnaiah et al.,
2002).
When the analyte is lipophilic which has low solubility in the
receiver side, modifying of the receptor uid is recommended in
order to achieve accurate determination of permeated amounts.
In our current study, PEG/PG/water (40/30/30, w/v) was used as
receiver medium and permeated bufalin was completely dissolved
in the receiver side for up to 10 h. In order to keep sink conditions
to maintain a favorable driving force for permeation, 4 mg/ml was
chosen as the donor side concentration, the same dose as in vivo
pharmacokinetic study. The permeated bufalin was less than 20%
of total amount in the donor side after 12 h permeation study and
the concentration of bufalin in the receptor uid was always below
5% of its concentration in the donor medium.
The permeation rate of bufalin across 28% vinyl acetate EVA
membrane was higher than its permeation rate across skin
(Figs. 5 and 6). In order to design a device controlled transdermal
patch formulation, EVA membrane with 19% of vinyl acetate was
chosen as the controlled release membrane.
In the present study, hairless mouse skin was used as in vitro
permeation model and rats were used at in vivo pharmacokinetic
study. These models were chosen because other cardiotoxicity and
anti-cancer activity studies of bufalin were performed on rodents,
which enabled pharmacokinetic and pharmacodynamic correla-
tion (data not published). Rodents have a thinner stratum corneum
and higher hair follicles density than human skin, so it may overes-
timate the permeability of drugs in human when using rodents skin
as model. However, the recent research indicated that Sprague-
Dawley rat is a useful model for predicting human skin permeability
with low inter-individual variations and similar permeation rate
(within twofold difference) (Takeuchi et al., 2011).
Considering the free-volatilized property of terpenes and
ethanol which are not suitable for matrix type transdermal sys-
tem, a reservoir type transdermal delivery system was developed.
Reservoir-type TDS usually consists of a drug reservoir (as a solution
or gel) in between the impermeable backing layer and a rate con-
trolling membrane. The rate controlling membrane can be either a
microporous (e.g., polypropylene) or a nonporous polymeric mem-
brane (e.g., ethylene vinyl acetate copolymer with a specic drug
permeability). Pressure sensitive adhesive polymer was usually
applied on the external surface of the polymer membrane to pro-
vide close contact of patch with skin surface (Babu and Pandit,
2005). In the present study, in order to eliminate the inuence of the
pressure sensitive adhesive on the permeation of drug, the pressure
sensitive adhesive is coated on the peripheral region of the patch.
Transdermal patch of bufalin provided sustained release of
bufalin and the MRT was 9.2 h, which is much longer than pre-
viously developed lipid (2.4 h) and nano (7.6 h) formulations (Liu
et al., 2011; Yin et al., 2012). The optimization for future design of
bufalin is to improve its low bioavailability (8%) through increas-
ing the patch size or reduce of the dose. From deconvolution
analysis, both in vivo release rate and accumulative permeation
amounts of bufalin from transdermal patch indicated that bufalin
can maintain sustained release up to 12 h. The release rate of
bufalin from batch slightly declined with time after 8 h, the pos-
sible reasons may due to decreased concentrations of ethanol and
d-limonene in transdermal reservoir upon their fast permeation to
skin. Loose contact of patch with skin surface in the later stage of
PK study due to movement of animal may also contribute to it.
 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240
In the current study, the highest level IVIVC (type A), which is
a point to point correlation of in vitro skin permeation to in vivo
plasma AUC was established. We realized the potential limitations
caused by different animal models (mice skin in vitro and rats PK
in vivo) used for IVIVC. However, our consistent IVIVC and decon-
volution results further demonstrated that these two models have
good correlations for the release of bufalin. The good correlation
indicated that the permeation study using mice skin could differ-
entiate the rate order of formulations and help with formulation
screening and development. On the basis of in vitro permeation
results, Gannu et al. (2010) successfully established ex vivoin vivo
correlation to estimate systemic concentration of lacidipine fol-
lowing transdermal gels administration in rabbits. It also showed
similar biphasic prole between in vitro permeation across mouse
skin and plasma. The possible reason for the initial fast absorption
may due to fast transport and removal of bufalin by rich capillary of
skin in live animal than excised skin. The real mechanism remains
to be explored. After achieving steady state, in vitro permeation
showed excellent IVIVC with correlation coefcient equals to 0.97.
This excellent IVIVC relationship demonstrated that in vitro exper-
iments can be used to screen formulations and predict preclinical
and clinical pharmacokinetics in further studies.
5. Conclusions
Clinical application of bufalin was mainly limited due to its nar-
row therapeutical window and short half life. Transdermal drug
delivery of bufalin provides a sustained release rate and avoids
the peak concentration, which prolong the duration of action and
decrease the risk of side effect. The present transdermal patch of
bufalin maintained therapeutical plasma concentration as long as
patch remained on the animal for at least 12 h. Deconvolution anal-
ysis also proved its sustained release rate in vivo. IVIVC enabled
the prediction of pharmacokinetic prole of bufalin from in vitro
permeation results. Our results demonstrated that transdermal
administration is an enabling formation strategy to deliver ade-
quate bufalin across skin on rodent models. Both in vitro and in vivo
studies showed consistent results that with optimized chemical
enhancer and formulation components, bufalin can achieve sus-
tained and sufcient delivery across skin.
Acknowledgements
This work was supported by research grants from Shanghai Sci-
ence Foundation (Shanghai, China). We thank Dr. Qinghua Ge and
her team for analytical assistance and Jin Lan for her value com-
ments on this manuscript.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.ijpharm.2013.02.048.
References
Amnuaikit, C., Ikeuchi, I., Ogawara, K., Higaki, K., Kimura, T., 2005. Skin permeation of
propranolol from polymeric lm containing terpene enhancers for transdermal
use? Int. J. Pharm. 289, 167178.
Babu, R.J., Pandit, J.K., 2005. Effect of penetration enhancers on the release and skin
permeation of bupranolol from reservoir-type transdermal delivery systems.
Int. J. Pharm. 288, 325334.
Bozdogan, H., 2000. Akaikes information criterion and recent developments in infor-
mation complexity. J. Math. Psychol. 44, 6291.
Dong, Y., Yin, S., Li, J., Jiang, C., Ye, M., Hu, H., 2011. Bufadienolide compounds sen-
sitize human breast cancer cells to TRAIL-induced apoptosis via inhibition of
STAT3/Mcl-1 pathway. Apoptosis 16, 394403.
239
Drakulic, B.J., Juranic, I.O., Eric, S., Zloh, M., 2008. Role of complexes formation
between drugs and penetration enhancers in transdermal delivery? Int. J. Pharm.
363, 4049.
FDA Part 182 SUBSTANCES GENERALLY RECOGNIZED AS SAFE. In: www.fda.gov/
Food/../GenerallyRecognizedasSafeGRAS/default.htm
Fox, L., Gerber, M., Plessis, J., Hamman, J., 2011. Transdermal drug delivery enhance-
ment by compounds of natural origin. Molecules 16, 1050710540.
Gannu, R., Palem, C.R., Yamsani, V.V., Yamsani, S.K., Yamsani, M.R., 2010. Enhanced
bioavailability of lacidipine via microemulsion based transdermal gels: formu-
lation optimization, ex vivo and in vivo characterization. Int. J. Pharm. 388,
231241.
Ghafourian, T., Zandasrar, P., Hamishekar, H., Nokhodchi, A., 2004. The effect of pene-
tration enhancers on drug delivery through skin: a QSAR study. J. Control. Release
99, 113125.
Hashimoto, S., Jing, Y., Kawazoe, N., Masuda, Y., Nakajo, S., Yoshida, T., et al., 1997.
Bufalin reduces the level of topoisomerase II in human leukemia cells and affects
the cytotoxicity of anticancer drugs. Leuk. Res. 21, 875883.
Hong, S.H., Choi, Y.H., 2012. Bufalin induces apoptosis through activation of both the
intrinsic and extrinsic pathways in human bladder cancer cells. Oncol. Rep. 27,
114120.
Hong, Z., Chan, K., Yeung, H.W., 1992. Simultaneous determination of bufadieno-
lides in the traditional Chinese medicine preparation, liu-shen-wan, by liquid
chromatography. J. Pharm. Pharmacol. 44, 10231026.
Kawazoe, N., Watabe, M., Masuda, Y., Nakajo, S., Nakaya, K., 1999. Tiam1 is involved
in the regulation of bufalin-induced apoptosis in human leukemia cells. Onco-
gene 18, 24132421.
Krishnaiah, Y.S., Satyanarayana, V., Bhaskar, P., 2002. Effect of limonene on the
in vitro permeation of nicardipine hydrochloride across the excised rat abdom-
inal skin. Pharmazie 57, 842847.
Li, D., Qu, X., Hou, K., Zhang, Y., Dong, Q., Teng, Y., et al., 2009. PI3K/Akt is involved in
bufalin-induced apoptosis in gastric cancer cells. Anticancer Drugs 20, 5964.
Liu, D., Feng, J.F., 2008. Determination of apparent aqueous solubility and apparent
oilwater partition coefcients of three bufadienolides composition. Zhongguo
Zhong Yao Za Zhi 33, 12561258.
Liu, Y., Chen, Z.Q., Zhang, X., Feng, N.P., Zhao, J.H., Wu, S., et al., 2010. An improved
formulation screening and optimization method applied to the development of
a self-microemulsifying drug delivery system. Chem. Pharm. Bull. (Tokyo) 58,
1622.
Liu, Y., Wang, P., Sun, C., Zhao, J., Du, Y., Shi, F., et al., 2011. Bioadhesion and
enhanced bioavailability by wheat germ agglutinin-grafted lipid nanoparti-
cles for oral delivery of poorly water-soluble drug bufalin. Int. J. Pharm. 419,
260265.
Meng, Z., Yang, P., Shen, Y., Bei, W., Zhang, Y., Ge, Y., et al., 2009. Pilot study of
huachansu in patients with hepatocellular carcinoma, nonsmall-cell lung can-
cer, or pancreatic cancer. Cancer 115, 53095318.
Panchagnula, R., Salve, P.S., Thomas, N.S., Jain, A.K., Ramarao, P., 2001. Trans-
dermal delivery of naloxone: effect of water, propylene glycol, ethanol and
their binary combinations on permeation through rat skin. Int. J. Pharm. 219,
95105.
Panesar, N.S., 1992. Bufalin and unidentied substance(s) in traditional Chi-
nese medicine cross-react in commercial digoxin assay. Clin. Chem. 38,
21552156.
Paudel, K.S., Milewski, M., Swadley, C.L., Brogden, N.K., Ghosh, P., Stinchcomb, A.L.,
2010. Challenges and opportunities in dermal/transdermal delivery. Ther. Deliv.
1, 109131.
Qin, T.J., Zhao, X.H., Yun, J., Zhang, L.X., Ruan, Z.P., Pan, B.R., 2008. Efcacy and safety of
gemcitabineoxaliplatin combined with huachansu in patients with advanced
gallbladder carcinoma. World J. Gastroenterol. 14, 52105216.
Song, H., Guo, T., Bi, K., Wang, H., Zhang, R., 2000. Determination of resibufogenin
and cinobufagin in heart-protecting musk pill by HPLC. Biomed. Chromatogr.
14, 130132.
Sun, L., Cun, D., Yuan, B., Cui, H., Xi, H., Mu, L., et al., 2012. Formulation and in vitro/in
vivo correlation of a drug-in-adhesive transdermal patch containing azasetron.
J. Pharm. Sci. 101, 45404548.
Takeuchi, H., Mano, Y., Terasaka, S., Sakurai, T., Furuya, A., Urano, H., et al., 2011. Use-
fulness of rat skin as a substitute for human skin in the in vitro skin permeation
study. Exp. Anim. 60, 373384.
Vaddi, H.K., Ho, P.C., Chan, Y.W., Chan, S.Y., 2002. Terpenes in ethanol: haloperidol
permeation and partition through human skin and stratum corneum changes?
J. Control. Release 81, 121133.
Wang, L., Raju, U., Milas, L., Molkentine, D., Zhang, Z., Yang, P., et al., 2011. Huachansu,
containing cardiac glycosides, enhances radiosensitivity of human lung cancer
cells. Anticancer Res. 31, 21412148.
Wang, S.W., Chiao, Y.C., Tsai, S.C., Lu, C.C., Chen, J.J., Shih, H.C., et al., 1997. Inhibition
of bufalin on pituitary and testicular function in rats. J. Pharmacol. Exp. Ther.
283, 528532.
Williams, A.C., Barry, B.W., 2004. Penetration enhancers. Adv. Drug Deliv. Rev. 56,
603618.
Williams, A.C., Barry, B.W., 1991. Terpenes and the lipidprotein-partitioning theory
of skin penetration enhancement. Pharm. Res. 8, 1724.
Yan, S., Qu, X., Xu, C., Zhu, Z., Zhang, L., Xu, L., et al., 2012. Down-regulation
of Cbl-b by bufalin results in up-regulation of DR4/DR5 and sensitization of
TRAIL-induced apoptosis in breast cancer cells. J. Cancer Res. Clin. Oncol. 138,
12791289.
Yang, Z., Luo, H., Wang, H., Hou, H., 2008. Preparative isolation of bufalin and cinob-
ufagin from Chinese traditional medicine ChanSu. J. Chromatogr. Sci. 46, 8185.
240 Z. Yang et al. / International Journal of Pharmaceutics 447 (2013) 231240
Yang, Z., Zhu, W., Gao, S., Xu, H., Wu, B., Kulkarni, K., et al., 2010. Simul-
taneous determination of genistein and its four phase II metabolites in
blood by a sensitive and robust UPLCMS/MS method: application to an
oral bioavailability study of genistein in mice. J. Pharm. Biomed. Anal. 53,
8189.
Yeh, J.Y., Huang, W.J., Kan, S.F., Wang, P.S., 2003. Effects of bufalin and cinobufagin
on the proliferation of androgen dependent and independent prostate cancer
cells. Prostate 54, 112124.
Yin, P., Wang, Y., Qiu, Y., Hou, L., Liu, X., Qin, J., et al., 2012. Bufalin-loaded
mPEGPLGAPLLcRGD nanoparticles: preparation, cellular uptake, tissue dis-
tribution, and anticancer activity. Int. J. Nanomed. 7, 39613969.
Yu, C.H., Kan, S.F., Pu, H.F., Jea Chien, E., Wang, P.S., 2008. Apoptotic signaling
in bufalin- and cinobufagin-treated androgen-dependent and -independent
human prostate cancer cells. Cancer Sci. 99, 24672476.
Zhang, D.M., Liu, J.S., Tang, M.K., Yiu, A., Cao, H.H., Jiang, L., et al., 2012. Bufotalin from
Venenum Bufonis inhibits growth of multidrug resistant HepG2 cells through
G(2)/M cell cycle arrest and apoptosis. Eur. J. Pharmacol. 692, 1928.
Zhang, L., Nakaya, K., Yoshida, T., Kuroiwa, Y., 1992. Induction by bufalin
of differentiation of human leukemia cells HL60, U937, and ML1 toward
macrophage/monocyte-like cells and its potent synergistic effect on the differ-
entiation of human leukemia cells in combination with other inducers. Cancer
Res. 52, 46344641.
Zhu, Z., Sun, H., Ma, G., Wang, Z., Li, E., Liu, Y., et al., 2012. Bufalin induces lung
cancer cell apoptosis via the inhibition of PI3K/Akt pathway. Int. J. Mol. Sci. 13,
20252035.
Zorin, S., Kuylenstierna, F., Thulin, H., 1999. In vitro test of nicotines permeability
through human skin. Risk evaluation and safety aspects. Ann. Occup. Hyg. 43,
405413.