- Email: [email protected]
Risk assessment of skin lightening cosmetics containing hydroquinone
Accepted Manuscript Risk assessment of skin lightening cosmetics containing hydroquinone Mariko Matsumoto, Hiroaki Todo, Takumi Akiyama, Mutsuko Hirata-Koizumi, Kenji Sugibayashi, Yoshiaki Ikarashi, Atsushi Ono, Akihiko Hirose, Kazuhito Yokoyama PII:
S0273-2300(16)30226-4
DOI:
10.1016/j.yrtph.2016.08.005
Reference:
YRTPH 3647
To appear in:
Regulatory Toxicology and Pharmacology
Received Date: 29 March 2016 Revised Date:
4 August 2016
Accepted Date: 8 August 2016
Please cite this article as: Matsumoto, M., Todo, H., Akiyama, T., Hirata-Koizumi, M., Sugibayashi, K., Ikarashi, Y., Ono, A., Hirose, A., Yokoyama, K., Risk assessment of skin lightening cosmetics containing hydroquinone, Regulatory Toxicology and Pharmacology (2016), doi: 10.1016/j.yrtph.2016.08.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Risk assessment of skin lightening cosmetics containing hydroquinone
2
Mariko Matsumoto1,4, Hiroaki Todo2, Takumi Akiyama3, Mutsuko Hirata-Koizumi1,
4
Kenji Sugibayashi2, Yoshiaki Ikarashi3, Atsushi Ono1, Akihiko Hirose1 Kazuhito
5
Yokoyama4
SC
RI PT
3
7
1
8
2
9
3
M AN U
6
Division of Risk Assessment, National Institute of Health Sciences, Tokyo, Japan
Josai University Faculty of Pharmaceutical Sciences, Saitama, Japan Division of Environmental Chemistry, National Institute of Health Sciences, Tokyo,
Japan
11
4
12
of Medicine, Tokyo, Japan
EP
Department of Epidemiology and Environmental Health, Juntendo University Faculty
AC C
13
TE D
10
14
Corresponding author: Mariko Matsumoto, Division of Risk Assessment, Biological
15
Safety Center, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-ku,
16
Tokyo 158-8501, Japan, tel: +81-3-3700-9878, fax: +81-3-3700-1408,
17
e-mail: [email protected]
18
1
ACCEPTED MANUSCRIPT
19
Abstract Following reports on potential risks of hydroquinone (HQ), HQ for skin
21
lightening has been banned or restricted in Europe and the US. In contrast, HQ is not
22
listed as a prohibited or limited ingredient for cosmetic use in Japan, and many HQ
23
cosmetics are sold without restriction. To assess the risk of systemic effects of HQ, we
24
examined the rat skin permeation rates of four HQ (0.3%, 1.0%, 2.6%, and 3.3%)
25
cosmetics. The permeation coefficients ranged from 1.2 × 10−9 to 3.1 × 10−7 cm/s, with
26
the highest value superior than the HQ aqueous solution (1.6 × 10−7 cm/s). After dermal
27
application of the HQ cosmetics to rats, HQ in plasma was detected only in the
28
treatment by highest coefficient cosmetic. Absorbed HQ levels treated with this highest
29
coefficient cosmetic in humans were estimated by numerical methods, and we
30
calculated the margin of exposure (MOE) for the estimated dose (0.017 mg/kg-bw/day
31
in proper use) to a benchmark dose for rat renal tubule adenomas. The MOE of 559 is
32
judged to be in a range safe for the consumer. However, further consideration may be
33
required for regulation of cosmetic ingredients.
AC C
EP
TE D
M AN U
SC
RI PT
20
34 35
Keywords: Hydroquinone, dermal absorption, permeation coefficients, benchmark dose,
36
margin of exposure, risk assessment
2
ACCEPTED MANUSCRIPT
Abbreviations
38
HQ: hydroquinone; i.v.: intravenous; LOAEL: lowest observed adverse effect level;
39
AUC: area under curve, BMD: benchmark dose; BMDL: lower confidence limit BMD;
40
MOE: margin of exposure; POD: point of departure
AC C
EP
TE D
M AN U
SC
RI PT
37
3
ACCEPTED MANUSCRIPT
41
1. Introduction
Hydroquinone (HQ) is used in skin bleaching agents, hair dyes, and finger nail
43
treatments (FDA, 2009). The WHO reported that a 1% HQ aqueous solution or a 5%
44
HQ cream caused dermal irritation in humans (WHO, 1996). Prolonged use of HQ
45
products (1–2%) is associated with exogenous ochronosis (Findlay et al., 1975), and a
46
worldwide total of 789 cases of exogenous ochronosis had been reported by 2007
47
(Levitt, 2007). In addition to these topical local effects, concerns have been raised
48
regarding the carcinogenic potential of HQ due to its carcinogenicity concerns reported
49
by animal studies (NTP, 1989; Shibata et al., 1991). Therefore, cosmetic use of HQ for
50
skin lightening has been banned in the UK and EU (EC, 2009). In the US, only
51
prescription skin lightening products can contain from greater than 2 to 4% HQ, and 2%
52
or less is allowed for cosmetic use (FDA, 2009). Recently, the US Cosmetic Ingredient
53
Review (CIR, 2014) concluded that HQ is safe at concentrations less than 1% in
54
cosmetic formulations designed for discontinuous, brief use followed by rinsing from
55
the skin and hair. On the other hand, HQ is not listed as a prohibited or limited
56
ingredient for cosmetic use in Japan (MHW Japan, 2000), and many kinds of skin
57
lightening cosmetics containing HQ are sold in Japan, some of which contain up to 10%
AC C
EP
TE D
M AN U
SC
RI PT
42
4
ACCEPTED MANUSCRIPT
HQ. Therefore, further information is required to evaluate whether the current use of
59
HQ cosmetics needs to be reconsidered in Japan.
RI PT
58
Dermal absorption of HQ was previously studied in 14 humans, and HQ was
61
dermally absorbed in humans with a bioavailability of 45.3 ± 11.2% for a 24-h
62
application (Wester et al., 1998). The blood elimination half-lives of HQ in a male
63
volunteer was reported to be 16.6 min for oral administration (NDMA, 1994), and that
64
for rats was 18.7 min for i.v. administration (Fox et al., 1986). Barber et al. (1995)
65
indicated that permeation of HQ in the human stratum corneum was slower than that in
66
fully thick rat skins. The permeability constant (K) values of
67
estimated to be 2.6 × 10−9 cm/s for human skin and 6.3 × 10−9 cm/s for rat skin (Barber
68
et al., 1995). Bucks et al. (1988) showed that penetration of HQ was reduced in
69
presence of a sunscreen and increased with a penetration enhancer in humans (Bucks et
70
al., 1988). Dermal absorption of HQ from cosmetics can vary among products because
71
the absorption rate varies depending on the base formulations (Ratna, 2004). In this
72
study, we examined rat skin permeation rates for four commercially available HQ
73
cosmetic products (HQ-1, HQ-2, HQ-3, and HQ-4) using a side-by-side diffusion cell
74
system to predict plasma HQ concentrations in humans after dermal absorption.
75
14
C-HQ solution were
AC C
EP
TE D
M AN U
SC
60
Because HQ is dermally absorbed, risk assessment for systemic effects, such as 5
ACCEPTED MANUSCRIPT
general repeated toxicity, carcinogenicity, and reproductive/developmental toxicity of
77
HQ are required. HQ induced sister chromatid exchanged, chromosome aberrations,
78
and/or gene mutation in vitro (Galloway et al., 1987; Tsutsui et al., 1997). Mutagenic
79
carcinogens are generally considered to have irreversible effects. If HQ carcinogenesis
80
is related to mutagenic events, the no-threshold concept should be applied for risk
81
assessment. A transgenic mouse mutation assay in the target organs of carcinogenicity is
82
useful to find if carcinogenesis is related to mutagenic events. We recently reported that
83
HQ is not mutagenic in transgenic MutaTM mice (Matsumoto et al., 2014), suggesting
84
that the mutagenic mechanism is not responsible for HQ induced carcinogenesis.
85
McGregor (2007) also suggested that the renal tumors are exacerbated non-genotoxic
86
spontaneous rodent renal disease that has no relevance to humans. In this study, we
87
estimate human absorbed levels of HQ after dermal application of the HQ cosmetic with
88
the highest permeation rate. Using the estimated human absorbed levels and
89
toxicological data of laboratory animals in the literature, a margin of exposure (MOE)
90
was calculated. In this paper, a risk assessment was conducted for the current use of HQ
91
cosmetics.
AC C
EP
TE D
M AN U
SC
RI PT
76
92 93
2. Materials and Methods
6
ACCEPTED MANUSCRIPT
94
2.1 Chemicals Four commercially available cosmetics (HQ-1, HQ-2, HQ-3, and HQ4) from
96
different manufacturers with different concentrations of HQ (unknown concentration,
97
3%, 10%, and 1%, respectively) were purchased. The measured HQ concentrations in
98
the products (HQ-1, HQ-2, HQ-3, and HQ-4) were 1.0%, 3.3%, 2.6%, and 0.3%,
99
respectively. From our survey, 10% HQ was the highest concentration available in the
100
market. Special grade HQ (CAS: 123-31-9; >99.0%) was purchased from Wako Pure
101
Chemical Industries, Ltd.
103
2.2. In vivo studies
104
2.2.1 Animals
TE D
102
M AN U
SC
RI PT
95
Male hairless rats (WBN/ILA-Ht; eight weeks old) were purchased from Life
106
Science Research Center, Josai University. This species was chosen because of its wide
107
use in toxicity and toxicokinetic studies and most commonly used for in vivo
108
permeation studies. Rats were reared on a basal diet (oriental yeast) and water ad
109
libitum. Animals were treated according to the ethical committee guidelines of Josai
110
University. Hair on the abdomen was shaved about 30 min after anesthesia of sodium
111
pentobarbital (50.0 mg/kg bw, intraperitoneal). After 12 h of fasting, in vivo
AC C
EP
105
7
ACCEPTED MANUSCRIPT
112
examination of rats was conducted. During the examination, body temperature of the
113
rats was kept using an electrical carpet.
RI PT
114
2.2.2 Blood concentration of HQ by dermal application of HQ aqueous solution or HQ
116
products
SC
115
The male hairless rats described in 2.2.1 were used. A diffusion cell (effective
118
surface area: 1.77 cm2) was glued with a biological glue (Aron Alpha) on the abdominal
119
intact skins, and 1 mL of 2% HQ aqueous solution (vehicle: saline) was applied.
120
Stripped skins were prepared by repeating the stripping 20 times using adhesive tape
121
(Nichiban Co., Ltd.). HQ aqueous solution was applied in the same manner as the intact
122
skins. HQ products (HQ-1, HQ-2, HQ-3, and HQ-4) were also applied to the stripped
123
skins as follows: The HQ products (about 100 mg) were applied on the skins, and
124
diffusion cells filled with the HQ products were applied on the skins. The blood samples
125
(0.2 mL) were taken from the cervical vein at 0, 30, 60, 120, 180, and 240 min after
126
application.
AC C
EP
TE D
M AN U
117
127 128 129
2.2.3 Blood concentration of HQ by intravenous administration Three rats were intraperitoneally anesthetized with urethane at 1.0 mg/kg. HQ
8
ACCEPTED MANUSCRIPT
aqueous solution at 2% (10 mg/kg) was administrated intravenously (i.v.), and blood
131
samples (0.2 mL) were taken at 0, 15, 30, 60, 90, and 180 min after administration. The
132
blood samples were centrifuged (15,000 rpm, 4°C, 5 min), and each 0.1 mL of blood
133
plasma was collected into a 1.5 mL microtube and kept in a freezer at −30°C.
134
2.2.4 Preparation of HQ cosmetics for UPLC
M AN U
135
SC
RI PT
130
Each 50 µL of HQ cosmetics was taken into a microtube, and the weight of the
137
cosmetics was precisely measured. The sample was completely dissolved in
138
tetrahydrofuran (4.5 times in weight) with vigorous mixing. After adding water (4.5
139
times in weight), the mixture was sonicated for 5 min. and centrifuged. The lower layer
140
of supernatant was taken and precisely diluted 10 times with water.
EP
141
TE D
136
2.2.5 Determination of HQ concentration
143
Concentrations of HQ were determined by UPLC (ACQUITY UPLC system H-CLASS;
144
Waters Company) equipped with a photodiode array detector (PDA eλ detector; Waters
145
Company). Limit of detection (LOD) of HQ was determined as signal-to-noise ratio of
146
3 and was 0.06 mg/mL. An ACQUITY BEH C18 column (size: 2.1 mm i.d. × 50 mm;
147
particle size: 1.7 µm; Waters Company) was used at 40℃. Five microliter of test
AC C
142
9
ACCEPTED MANUSCRIPT
solution was injected. Acetonitrile (5% or 10%) was used as a mobile phase and a flow
149
rate was set to 0.1 mL/min. The scanning range of wavelength was set to 210–400 nm.
150
HQ was detected and quantitated at 288–289 nm. The absolute calibration curve method
151
was used for measurement of HQ. Water solution containing special grade HQ was used
152
as a standard solution.
SC
RI PT
148
M AN U
153 154
2.3 In vitro studies
155
2.3.1 Determination of HQ concentration for release and permeation studies HQ release and skin permeation from the studied products were examined by a
157
side-by-side diffusion cell system. For the release study, a cellulose dialysis membrane
158
(UC24-32-100, EIDIA Co., Ltd., Tokyo, Japan) was placed in between the diffusion
159
cells. The cellulose dialysis membrane was chosen because it is widely used for drug
160
release studies. For the permeation study, abdominal skin pieces were excised from rats
161
under anesthesia of sodium pentobarbital (50.0 mg/kg bw, intraperitoneal) and one piece
162
of intact skin was set in between the diffusion cells (effective surface area: 0.95 cm2).
163
The intact rat skin data were used for a realistic exposure assessment because
164
HQ-containing products are generally applied to intact human skin. Full-thickness skins
165
were used as defined by OECD TG 428 in this study (OECD, 2004). Permeability of
AC C
EP
TE D
156
10
ACCEPTED MANUSCRIPT
hydrophilic chemicals like HQ would not be affected by skin thickness in contrast to
167
lipophilic chemicals (Yamaguchi et al., 2008). The thickness of the skin was measured
168
according to a method reported previously (Watanabe et al., 2001). The
169
biotransformation was not inhibited in this system. Each HQ product (HQ-1, HQ-2,
170
HQ-3, or HQ-4 at 250 mg) was applied on one side of the cellulose membrane or the
171
stratum corneum surface in a donor compartment. The other side of the compartment, a
172
receiver compartment, was filled with distilled water (3.0 mL). The temperature of skin
173
surface in the diffusion cell system was kept at 32°C, and the receiver compartment was
174
agitated with a magnetic stirrer using a star-head magnetic bar. Samples (0.5 mL each)
175
were withdrawn at appropriate intervals (every 5-30 min for first one or two hours, and
176
hourly afterward) from the receiver compartment and immediately replaced with 0.5 mL
177
fresh distilled water. Each sample was added to a same volume of methanol and
178
centrifuged (15,000 × g for 5 min at 4°C). The supernatant contents were determined by
179
HPLC equipped with a photodiode array detector (Shimadzu LC-20AD, Shimadzu
180
SPDM-20A, Kyoto, Japan). It was validated by linearity, and the lowest concentration
181
was set as 0.1 µg/mL. Conditions of the HPLC were as follows: mobile phase was water
182
and methanol (75:25); column was Inertsil ODS-3, 5 µm, 4.6 × 150 mm; column
183
temperature was 40°C; detection wavelengths were 280 nm (excitation) and 310 nm
AC C
EP
TE D
M AN U
SC
RI PT
166
11
ACCEPTED MANUSCRIPT
184
(emission); flow rate was 1.0 mL/min; injection volume was 5 µL.
185
2.4 Theoretical approach
187
2.4.1 Determination of permeation parameters (in vitro data)
188
In this study, human absorption levels were estimated with in vitro permeation data of
189
rats. Nakamura et al., (2012) indicated that the predicted absorption profiles obtained
190
from rat skins in vitro were well agreed with human clinical values.
191
Skin permeation of HQ can be determined by Fick’s second law of diffusion:
SC
=
(1),
M AN U
192
RI PT
186
where Dskin is the diffusion coefficient of HQ across the skin, and Cskin is the skin
194
concentration of HQ at a time t and a position of the skin . Cskin was calculated using
195
Eq. 1 with the following initial and boundary conditions:
196
t = 0, 0 <
197
t > 0, when = 0,
EP
AC C
198
TE D
193
Cskin = Kskin・Cdonor and
=
when = Lskin, Cskin = 0 (Boundary conditions),
199
where Lskin is the thickness of the skin, Kskin is the partition coefficient of HQ from the
200
donor to the skin, and Cdonor is the HQ concentration in the donor.
201
The skin permeation rate to the receiver, Jskin is expressed by Fick’s first law (Eq. 2),
12
ACCEPTED MANUSCRIPT
204
205 206 207
(2).
The cumulative amount of drug permeated per unit area, Q is expressed by Eq. 3
Q = − "
! (3).
RI PT
203
= −
Differential terms in Fick’s second law (Eq. 1) are approximated as follows: #,% #
,% #
&
= ∆ () ,*+& − ) ,* , (4), and =
&
∆
() -&,* − 2) ,* + ) +&,* , (5),
SC
202
where Ci,j is HQ concentration at i-th position and j-th time, ∆ is
209
tj+1-tj. By substation of the above equations (Eq. 4 and Eq.5) to Fick’s second law (Eq.1),
210
the following equation is obtained. 0
0
i+1- i,
and ∆t is
0
) ,*+& = 0 ) ,*-&,* + 1 − 2 0 ) ,* + 0 ) +&,* (6).
TE D
211
M AN U
208
The above differential equations (Eq.2 and Eq.3) can be approximated to the following
213
difference equations:
214
* = −
23,% ∆
(7), and
AC C
215
EP
212
4* = 4*-& + * ・∆! (8),
216
where n is the number of divisions of skin. Then, Jj was calculated by setting n=10. In
217
this calculation, ∆t was set to be less than 0.5 for ∆t/∆2 Dskin. Qj was obtained by the Eq
218
(8) with a value of Jj. The effective diffusion coefficient, Dskin and the partition
219
coefficient, Kskin were obtained with the least-square method by fitting observed values.
13
ACCEPTED MANUSCRIPT
The calculation was conducted by a pseudo-Newtonian method of the Microsoft Excel
221
(Office 2013) solver-function (Max time: 100 seconds, Iterations: 100, Precision:
222
0.000001, Tolerance: 5%, Convergence: 0.001).
RI PT
220
223
2.4.2 Determination of parameters (in vivo data)
225
The amount of HQ in the body at any particular time after i.v. administration is given by
226
the following equation: = " 5 - 67 (9),
M AN U
227
SC
224
where is the amount remaining to be eliminated, 0 is the dose administered (initial
229
amount), kel is an elimination rate constant, and t is time. In vivo blood concentration
230
data (i.v. administration of 2% HQ aqueous solution) were used for data fitting. The
231
non-linear least squares Damping Gauss–Newton method was used for parameter
232
estimation and the elimination rate constant kel was calculated. The Microsoft Excel
233
(Office 2013) was used for calculation.
235
EP
AC C
234
TE D
228
2.4.3 Estimation of in vivo blood concentration-time profile
236
Time course of plasma HQ concentrations after dermal absorption was
237
predicted from in vitro permeation data and in vivo i.v. data by a numerical convolution
14
ACCEPTED MANUSCRIPT
method with the following assumption: the elimination rate of HQ is linear and the
239
permeation rate of HQ is identical in in vitro and in vivo. Using the two parameters
240
(Dskin and Kskin) determined in 2.4.1, cumulative amounts of HQ and skin permeation
241
rates (Jskin) of HQ at each sampling time were calculated on the assumption that 1.5 g
242
HQ-2 was applied on a whole face, a décolleté and hands (1440 cm2). HQ-2 was
243
selected because it showed the highest permeation coefficient among the four cosmetic
244
products. The blood concentration of HQ after a single dermal application of 1.5 g
245
HQ-2 on a whole face, a décolleté and hands (left for 8 hours overnight and washed off),
246
and after 2-time dermal applications of 1.5 g/time HQ-2 on a whole face, a décolleté
247
and hands (left for 8 hours overnight and washed off, and left for 16 hours during
248
day-time and washed off) were estimated. The Microsoft Excel (Office 2013) was used
249
for calculation.
251 252
SC
M AN U
TE D
EP AC C
250
RI PT
238
2.5 Determination of benchmark dose In the literature, the lowest LOAEL (lowest observed adverse effect level) of a
253
repeated dose of HQ was reported to be 17.9 mg/kg bw/day (25 mg/kg bw, 5 days/week
254
for 103 weeks) for general toxicity due to lowered body weight and for carcinogenicity
255
due to renal tubule adenomas in rats given HQ by gavage (NTP, 1989). To find out a
15
ACCEPTED MANUSCRIPT
better point of departure (POD), a benchmark dose approach was performed using the
257
benchmark dose software (BMDS) 2.5.0 developed by US EPA (US EPA, 2014). In the
258
103-week study, numbers of adenoma found were 0/55, 4/55, and 8/55 at 0, 25, and 50
259
mg/kg bw/day, respectively. Using these data, a lower 95% confidence limit of BMD at
260
10% response level (BMDL10) was determined. The following models were used for the
261
analysis: Gamma (restricted or unrestricted), Logistic, LogLogistic (restricted or
262
unrestricted), LogProbit (restricted or unrestricted), Multistage (restricted or
263
unrestricted), Probit, Weibull (restricted or unrestricted), and Quantal-Linear. Model fit
264
was judged by P-values for goodness of fit (P > 0.1 is considered to be a good fit),
265
visual inspection of the dose response curve, and scaled residuals at the data point
266
(absolute value of scaled residuals < 2.0 is considered to be a good fit).
267
269
3. Results
AC C
268
EP
TE D
M AN U
SC
RI PT
256
Table 1 presents HQ concentration in rat plasma after application of 2% HQ
270
aqueous solution to the tape-stripped or intact skins of the abdomen. HQ was not
271
detected for the intact skin. The highest level was detected at 60 min after treatment of
272
the abdominal tape-stripped skins.
273
HQ concentrations in the rat plasma after application of the four HQ cosmetics
16
ACCEPTED MANUSCRIPT
to the tape-stripped abdomen are shown in Table 2. The highest concentration of 4.64
275
mg/L HQ was observed at 60 min after application of HQ-2. HQ was not detected in the
276
rat plasma treated by HQ-1, HQ-3, and HQ-4.
RI PT
274
The measured HQ concentrations in the products (HQ-1, HQ-2, HQ-3, and
278
HQ-4) were 1.0%, 3.3%, 2.6%, and 0.3%, respectively. The measurements were
279
conducted as described in 2.2.4 and 2.2.5. The elimination rate constant of HQ, Kel was
280
calculated to be 2.1/h using i.v administration data (Fig. 1).
M AN U
SC
277
Fig. 2 shows the cumulative amount of HQ permeated through the hairless rat
282
skins from a 2% aqueous solution. The permeation coefficient of the HQ aqueous
283
solution was calculated to be 1.6 × 10−7 cm/s for the intact hairless rat skin.
TE D
281
Fig. 3 shows the cumulative amount (%) of HQ permeated from the HQ
285
products through the intact hairless rat skin. The permeation coefficients of HQ were
286
estimated to be HQ-2 (3.1 × 10−7 cm/s) > HQ-4 (1.1 × 10−8 cm/s) > HQ-1 (7.0 × 10−9
287
cm/s) > HQ-3 (1.2 × 10−9 cm/s) for the rat skin. The permeation coefficients of HQ for
288
HQ-4, HQ-1, and HQ-3 were lower than that of the HQ aqueous solution.
AC C
EP
284
289
Fig. 4 shows the cumulative amount (%) of HQ released from HQ products.
290
The release amount of HQ was the highest in HQ-2, followed by HQ-3, HQ-4, and HQ1.
291
The effective diffusion coefficient, Dskin of HQ for HQ-2 was 2.2 × 10−4 cm/s, and the
17
ACCEPTED MANUSCRIPT
292
partition coefficient of HQ for HQ-2, Kskin was 0.32.
293
RI PT
294
The estimated blood concentrations after a single dermal application of HQ-2
296
on the whole face, a décolleté and hands is shown in Fig. 5, and after the 2-time dermal
297
application of HQ-2 on the whole face, a décolleté and hands in Fig. 6.
M AN U
SC
295
The areas under the curves (AUCs) for 24 h after the single and 2-time
299
applications (1.5 g/time) on a whole face, a décolleté and hands (1440 cm2) were
300
calculated to be 0.183 mg h/L and 0.310 mg h/L, respectively. The blood concentrations
301
of HQ were calculated on the following two assumptions. Single application: left for 8
302
hours overnight and washed off. Two-time application: left for 8 hours overnight and
303
washed off, and left for 16 hours during day-time and washed off. After the single
304
application, the blood concentration of HQ was considered to reach zero at 30 h, and the
305
AUC (30 h) was calculated to be 0.186 mg h/L. The cumulative permeation level of HQ
306
at 24 h after the 2-time application of HQ-2 was estimated to be 60.9 µg/cm2 (Fig. 7).
307
Using this value, the dermal permeation of HQ for 24 h (one day) for a 50 kg adult
308
person was estimated to be 1.75 mg/kg bw/day (60.9 × 10−3 mg/cm2 × 1440 cm2 per 50
309
kg bw) for an adult person.
AC C
EP
TE D
298
18
ACCEPTED MANUSCRIPT
For the BMD analysis, fitting was successfully conducted in all the models, but
311
BMDL10 values were not obtained by the unrestricted LogLogistic, LogProbit, or
312
Weibull models. The estimated BMD10 and BMDL10 for the 103-week study are shown
313
in Table 3. The BMDL10 ranged from 13.7266 mg/kg bw to 34.7432 mg/kg bw. The
314
unrestricted models for Gamma, LogLogistic, LogProbit, and Weibull did not provide
315
an acceptable BMDL10. The Akaike’s Information Criterion (AIC) ranged from 76.2961
316
to 80.476. P-values for global goodness of fit were all acceptably large (>0.1). Absolute
317
values of scaled residuals for local measurement were all acceptably small (<2.0).
M AN U
SC
RI PT
310
319
4. Discussion
TE D
318
When the 2% HQ aqueous solution was applied to the tape-stripped skin under
321
anesthesia for 4 hours, HQ was detected in the plasma of rats, with the highest level
322
detected at 60 min. On the other hand, HQ was not detected in the plasma after the 2%
323
aqueous solution was applied to the intact skins of rats, indicating that the stratum
324
corneum worked as a barrier for the permeation of HQ. Because the rate controlling
325
factor of permeation of chemicals is permeability of stratum corneum, anesthesia
326
condition would not affect skin absorption as long as temperature of skin surface was
327
kept 32 ± 1 °C (OECD, 2004). In this study, skin surface was kept at 32°C to assure skin
AC C
EP
320
19
ACCEPTED MANUSCRIPT
temperature was not lowered by anesthesia. HQ was detected in the plasma after
329
application of HQ-2, but was not detected after application of HQ-1, HQ-3, and HQ-4
330
even for the tape-striped skin. 1,2-Hexanediol included in HQ-2 could be acted as a
331
permeation enhancer (Warner et al., 2001). However, a known enhancer such as ethanol,
332
stearic acid, or ascorbic acid was included in other products. Permeability differs by
333
enhancer concentration (Warner et al., 2001), and there were so many ingredients
334
included in the products; therefore, it is difficult to conclude what enhancer affected
335
permeability. In addition, many ingredients included in the products can affect
336
thermodynamic activity of HQ. Because the thermodynamic activity is known to affect
337
skin permeation of drugs (Ishii et al., 2010), this result may also be related to the
338
thermodynamic activities of HQ in the cosmetic formulations.
TE D
M AN U
SC
RI PT
328
The plasma levels of HQ were rapidly decreased after i.v. administration of the
340
2% HQ aqueous solution in rats and Kel was calculated to be 2.1/h, which indicates that
341
the metabolism of HQ is rapid. This is in good agreement with the short elimination
342
half-lives of HQ radiolabel; 16.6 min for a male volunteer after ingestion (NDMA,
343
1994) and 18.7 and 14.8 min for rats after i.v. and oral administrations (Fox et al., 1986).
344
After oral administration of HQ in rats, <1% of the detected radiolabel in blood was
345
associated with the parent compound, indicating rapid metabolism of HQ (Fox et al.,
AC C
EP
339
20
ACCEPTED MANUSCRIPT
1986). In general, animal skin absorption is greater than human skins; however, no
347
concrete conclusion on human skins would be derived from findings of in vivo data in
348
this study. We could not detect HQ in the blood of rats after dermal application in the
349
intact skin study. It cannot conclude that HQ was not absorbed, because elimination of
350
HQ from the blood is rapid. It was suggested that the HQ levels under the LOD in the
351
blood could be a target of this study. In fact, the highest estimated blood concentration
352
of HQ for application of HQ-2 was lower than the LOD (Fig. 5 and Fig. 6). Therefore,
353
we further evaluated HQ cosmetics by an in vitro method, which is useful for predicting
354
the pharmacokinetic profile of chemicals in human (Nakamura et al., 2012).
M AN U
SC
RI PT
346
The permeation coefficient of the 2% HQ aqueous solution was calculated to
356
be 1.6 × 10−7 cm/s and was higher than the previously reported value of 6.3 × 10-9 cm/s
357
for the
358
cells (Barber et al., 1995). The permeation coefficients of HQ were estimated to be
359
HQ-2 (3.1 × 10−7 cm/s) > HQ-4 (1.1 × 10−8 cm/s) > HQ-1 (7.0 × 10−9 cm/s) > HQ-3 (1.2
360
× 10−9 cm/s) for the rat skin. The previously reported permeation coefficient value was
361
within the range of three HQ products with lower permeation coefficients. Factors that
362
might contribute for the difference are unclear because Barber et al., (1995) did not
363
clearly state the concentration or the vehicle of the HQ solution. When relatively higher
C-HQ solution to full-thickness Fisher 344 rat skin in Franz-type diffusion
AC C
EP
14
TE D
355
21
ACCEPTED MANUSCRIPT
permeation coefficient and absorption of HQ through the rat skins are taken into
365
account for HQ-2, HQ is expected to be absorbed and distributed in the body after
366
application of HQ-2 in humans. This indicates that not only local effects of HQ but also
367
systemic effects have to be considered for dermal application of HQ. The permeation
368
coefficients of HQ for HQ-4, HQ-1, and HQ-3 were lower than that of the HQ aqueous
369
solution. This result is consistent with the in vivo tape-stripped study. The release rate of
370
HQ was the highest in HQ-2, followed by HQ-3, HQ-4 and HQ-1. This order did not
371
exactly match the order of the permeation coefficient, but the high release rate may
372
partially contribute to the permeability of the HQ cosmetics.
M AN U
SC
RI PT
364
The measured concentrations of HQ in HQ-1, HQ-2, HQ-3, and HQ-4 were
374
1.0%, 3.3%, 2.6%, and 0.3%, respectively. The values of HQ-3 and HQ-4 were lower
375
than those of labelling on the products. These differences may be related to the process
376
of manufacturing or degradation of HQ. In any case, we confirmed that two products
377
(HQ-2 and HQ-3) contained HQ at a level greater than 2%. HQ-2 showed a higher
378
permeation coefficient of HQ than that of the HQ aqueous solution, and dermal
379
absorption was confirmed. On the other hand, HQ was not detected in the plasma of rats
380
after dermal application of HQ-3 in spite of its relatively high concentration of HQ
381
(2.6%). Because a number of products examined was relatively small, it can be difficult
AC C
EP
TE D
373
22
ACCEPTED MANUSCRIPT
to draw solid conclusions for effects of formulation or concentration of HQ. However,
383
our findings suggest that regulating based on the concentration of a target substance
384
may not be safe enough for dermally applied cosmetics. Therefore, especially high
385
concentration products may need to be tested for their permeability for a safer use.
RI PT
382
In the literature, the lowest LOAEL of a repeated oral dose of HQ was reported
387
to be 17.9 mg/kg bw/day for general toxicity due to decreased body weights and for
388
carcinogenicity due to renal tubule adenomas in rats given HQ by gavage for 103 weeks
389
(NTP, 1989). As for a reproductive/developmental endpoint, a study by Murphy et al.,
390
(1992) showed the lowest NOAEL of 75 mg/kg bw/day, in which marginal increases of
391
skeletal and internal malformations were observed at a maternal toxic dose of 150
392
mg/kg bw/day. Therefore, reproductive/developmental toxicity was not considered a
393
specifically sensitive endpoint. HQ carcinogenesis is not considered to be related to
394
mutagenic events (McGregor 2007, Matsumoto et al., 2014); therefore, the threshold
395
concept was applied for risk assessment.
M AN U
TE D
EP
AC C
396
SC
386
The range of the BMDL10s for the NTP 103-week study (the LOAEL = 25
397
mg/kg bw) calculated from the fitted models were between 13.7 mg/kg bw and 34.7
398
mg/kg bw. As a conservative risk assessment, the lowest BMDL10 (13.7 mg/kg bw) was
399
chosen for further quantitative assessment. After adjusting 5 days per week
23
ACCEPTED MANUSCRIPT
administration to 7 days per week administration, the POD for calculating MOE was
401
estimated as 9.78 mg/kg bw/day. The MOE was calculated by the POD (9.78 mg/kg
402
bw/day: oral dose) divided by the estimated human dose level (1.75 mg/kg bw/day) by
403
dermal application and was 5.59 for the worst case exposure scenario. By using the
404
BMDL value instead of the LOAEL for calculation of the MOE, uncertainty was
405
considered to be reduced. Orally administrated HQ is considered to get well absorbed in
406
humans (Deisinger et al., 1996), and >90% absorption was confirmed in rats
407
(Divincenzo et al., 1984). Therefore, for the calculation of the MOE, we assumed that
408
orally dosed HQ (9.78 mg/kg bw/day) to rats in the NTP study was 100% absorbed. If
409
the MOE is less than one, an immediate regulatory action might have been required for
410
HQ cosmetics in Japan. The MOE of 5.59 is less than an acceptable MOE of 25 for the
411
TK-based safety margins (WHO, 2005); however, this is for a worst case exposure
412
scenario.
SC
M AN U
TE D
EP
AC C
413
RI PT
400
According to the instructions of HQ-2, this cosmetic is recommended to be
414
used to spot areas and to avoid applying to skin areas without pigmented spots. If the
415
application was limited to spot areas, e.g., 1/100 of a face, a décolleté and hands area,
416
the MOE is judged to be in a range safe (MOE：559 >100: intra- and inter-species
417
differences) for a nonmutagenic carcinogen. HQ-containing products are generally
24
ACCEPTED MANUSCRIPT
applied to intact human skin, and HQ were not detected after application of HQ
419
cosmetics to intact rat skin. The permeation rate of HQ is expected to be slower in
420
humans compared to rats as stated before (Barber et al., 1995). Taken into account of
421
this, the MOE in reality could be improved. However, Wester et al. (1998) reported a
422
human in vivo absorption rate of 45%, and elimination of HQ in humans seemed a little
423
slower than in rats according to half-lives of HQ (14.8 min in rats and 16.6 min in
424
humans) for oral administration (NDMA, 1994; Fox et al., 1986). This could result
425
higher blood levels than our calculation based on elimination rate estimated from in vivo
426
rat data. Therefore uncertainties remain for our exposure estimation based on species
427
differences.
TE D
M AN U
SC
RI PT
418
Although HQ-1, HQ-2, and HQ-3 are recommended for topical use for
429
pigmented spots, HQ-4 is recommended to be used for the whole face. However, the
430
concentration of HQ-4 was very low. Besides, from the permeation coefficient of HQ
431
for HQ-4, its permeability into the skin was found to be “slow” according to the
432
qualitative ranking system by Marzulli et al., (1969), and the permeation rate is
433
expected to be slower in humans compared to rats (Barber et al., 1995). Based on this,
434
as long as users follow the instructions of the products, the risk of these HQ cosmetics is
435
considered to be low.
AC C
EP
428
25
ACCEPTED MANUSCRIPT
In the Japanese market, there are several other products containing HQ at >2%
437
and there are many products containing unknown levels of HQ that we haven’t assessed.
438
There is a possibility that these other products show a higher permeability than HQ-2.
439
We have to also note that MOE of >1000 could be preferable given the toxic property of
440
HQ (carcinogenesis). In addition, cosmetic users include pregnant women and women
441
planning a pregnancy in near future. HQ as a prescription product is assigned to
442
“pregnancy category C: the benefits from the use of the drug in pregnant women may be
443
acceptable despite its potential risks” by the FDA based on “no adequate studies in
444
humans, but adverse effects were observed on the fetus in animal reproduction studies”
445
(Drugs.com, 2014). It is still controversial whether Japanese consumers have enough
446
information for the safe use of cosmetics. Cosmetics are usually marketed based on only
447
safety evaluation tests for acute or local toxicity under the company’s self-imposed
448
control in Japan. Since the possibility that dermally absorbed cosmetic ingredients like
449
HQ may cause systemic adverse effects cannot be ruled out, further consideration may
450
be required for the regulation of cosmetic ingredients.
AC C
EP
TE D
M AN U
SC
RI PT
436
451 452 453
5. Acknowledgement This study was supported by the Ministry of Health, Labour and Welfare,
26
ACCEPTED MANUSCRIPT
454
Japan.
455
RI PT
456
AC C
EP
TE D
M AN U
SC
457
27
ACCEPTED MANUSCRIPT
458
References
459
461
Barber, E. D., et al., 1995. The percutaneous absorption of hydroquinone (HQ) through rat and human skin in vitro. Toxicol Lett. 80, 167-72.
RI PT
460
Bucks, D. A., et al., 1988. Percutaneous absorption of hydroquinone in humans: effect
463
of 1-dodecylazacycloheptan-2-one (azone) and the 2-ethylhexyl ester of
464
4-(dimethylamino)benzoic acid (Escalol 507). J Toxicol Environ Health. 24,
465
279-89.
M AN U
466
SC
462
CIR (2014) Amended Safety Assessment of Hydroquinone as Used in Cosmetics available
468
http://online.personalcarecouncil.org/ctfa-static/online/lists/cir-pdfs/FR647.pdf
471 472
Deisinger, P. J., et al., 1996. Human exposure to naturally occurring hydroquinone. J
EP
470
at
Toxicol Environ Health. 47, 31-46.
AC C
469
TE D
467
Divincenzo, G. D., et al., 1984. Metabolic fate and disposition of [14C]hydroquinone given orally to Sprague-Dawley rats. Toxicology. 33, 9-18.
473
Drugs.com, Hydroquinone cream. Vol. 2014, 2014.
474
EC, REGULATION (EC) No 1223/2009 OF THE EUROPEAN PARLIAMENT AND
475
OF THE COUNCIL of 30 November 2009 on cosmetic products. 2009.
28
ACCEPTED MANUSCRIPT
477 478 479
FDA, 2009. Hydroquinone [CAS 123-31-9] Supporting Information for Toxicological Evaluation by the National Toxicology Program 21 May 2009. Findlay, G. H., et al., 1975. Exogenous ochronosis and pigmented colloid milium from
RI PT
476
hydroquinone bleaching creams. Br J Dermatol. 93, 613-22.
Fox, J. A., et al., 1986. Blood Elimination Kinetics of [U-14C]Hydroquinone
481
Administered by Intragastric Intubation, Intratracheal Instillation or Intravenous
482
Injection to Male Fischer 344 Rats,. Rochester, NY, Health and Environment
483
Laboratories; Eastman Kodak Company. Report No. TX-86-1 (cited in
484
DeCaprio 1999; Crit Rev Toxicol: 29, 283-330).
M AN U
SC
480
Galloway, S. M., et al., 1987. Chromosome aberrations and sister chromatid exchanges
486
in Chinese hamster ovary cells: evaluations of 108 chemicals. Environ Mol
487
Mutagen. 10 Suppl 10, 1-175.
489 490 491
EP
Iimura, N., et al., 2005. Development of new whitening agents with hydroquinone
AC C
488
TE D
485
stabilized by the complex formation with surfactants and the evaluation for melanogenesis inhibitory effect and skin stimulus. J. Jpn. Cosmet. Sci. Soc. 29, 301-313.
492
Ishii, H., et al., 2010. Effect of thermodynamic activity on skin permeation and skin
493
concentration of triamcinolone acetonide. Chem Pharm Bull (Tokyo). 58,
29
ACCEPTED MANUSCRIPT
497 498 499 500
Federal Register. J Am Acad Dermatol. 57, 854-72.
RI PT
496
Levitt, J., 2007. The safety of hydroquinone: a dermatologist's response to the 2006
MHW Japan, 2000. Standards for Cosmetics Notification No. 331 of 2000 (in Japanese).
SC
495
556-61.
Marzulli, F. N., et al., 1969. Techniques for studying skin penetration. Toxicology and
M AN U
494
Applied Pharmacology. 14, Supplement 3, 76-83.
Matsumoto, M., et al., 2014. Evaluation of in vivo mutagenicity of hydroquinone in
502
Muta™ mice. Mutation Research/Genetic Toxicology and Environmental
503
Mutagenesis. 775–776, 94-98.
506 507 508 509
carcinogenic and mutagenic properties. Crit Rev Toxicol. 37, 887-914.
EP
505
McGregor, D., 2007. Hydroquinone: an evaluation of the human risks from its
Murphy, S. J., et al., 1992. A study of developmental toxicity of hydroquinone in the
AC C
504
TE D
501
rabbit. Fundam Appl Toxicol. 19, 214-21.
NDMA, 1994. Chronic Health Effects Testing for Hydroquinone (Nonprescription Drug Manufactures Association).
510
NTP, 1989. Toxicology and carcinogenesis studies of hydroquinone (CAS No.
511
123-31-9) in F344/N rats and B6C3F1 mice (gavage studies). National
30
ACCEPTED MANUSCRIPT
Toxicology Program Technical Report Series, No.366, U.S. Department of
513
Health and Human Surivice, Public Health Service, National Institutes of
514
Health.
RI PT
512
Nakamura, A., et al., 2012. Evaluation of the predicted time-concentration profile of
516
serum tulobuterol in human after transdermal application. Chem Pharm Bull
517
(Tokyo). 60, 300-5.
M AN U
SC
515
518
OECD, 2004. Test Guideline 428. Skin absorption: in vitro method, OECD, Paris
519
Ratna, M., 2004. Topical and Transdermal Drug Delivery: What a Pharmacist Needs to
520
Know. ACPE ID number 221-146-04-054-H01.
Shibata, M. A., et al., 1991. Induction of renal cell tumors in rats and mice, and
522
enhancement of hepatocellular tumor development in mice after long-term
523
hydroquinone treatment. Jpn J Cancer Res. 82, 1211-9.
525
EP
Smiles, K. A., et al., 2007. A hydroquinone formulation with increased stability and
AC C
524
TE D
521
decreased potential for irritation. J Cosmet Dermatol. 6, 83-8.
526
Tsutsui, T., et al., 1997. Benzene-, catechol-, hydroquinone- and phenol-induced cell
527
transformation, gene mutations, chromosome aberrations, aneuploidy, sister
528
chromatid exchanges and unscheduled DNA synthesis in Syrian hamster embryo
529
cells. Mutat Res. 373, 113-23.
31
ACCEPTED MANUSCRIPT
530
US EPA, Benchmark Dose Software (BMDS) version 2.5.0. 2014.
531
WHO, 1996. IPCS INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY
533
Health and Safety Guide No. 101.
RI PT
532
WHO, 2005. Harmonization Project Document No.2: Chemical-specific adjustment
factors for interspecies differences and human variability: guidance document
535
for use of data in dose/concentration – response assessment. World Health
536
Organization, Geneva, 2005.
M AN U
SC
534
Warner, K. S., et al., 2001. Influences of alkyl group chain length and polar head group
538
on chemical skin permeation enhancement. J Pharm Sci. 90, 1143-53.
539
Watanabe, T., 2001. Utility of the three-dimensional cultured human skin model as a
540
TE D
537
tool to evaluate skin permeation of drugs. Altern. Animal Test. Exp. 7, 1-14. Wester, R. C., et al., 1998. Human in vivo and in vitro hydroquinone topical
542
bioavailability, metabolism, and disposition. J Toxicol Environ Health A. 54,
544
AC C
543
EP
541
301-17.
Yamaguchi, K., et al., 2008. Structure-permeability relationship analysis of the
545
permeation
546
epidermis/dermis of rat skin. J Pharm Sci. 97, 4391-403.
barrier
properties
of
547
32
the
stratum
corneum
and
viable
ACCEPTED MANUSCRIPT
Table 1 Hydroquinone concentration (mg/L) in the hairless rat plasma (n=3) after
549
abdominal administration of 2% hydroquinone aqueous solution (effective diffusion
550
area; 1.77 cm2)
551
___________________________________________ Time after Tape-stripped Intact
553 554 555 556 557
administration (min) (mg/L) (mg/L) ___________________________________________ 0 0.21 0.00 30 1.78 0.00
SC
552
RI PT
548
562
LOD: 0.06 mg/L
558 559 560
M AN U
561
60 2.17 0.00 120 0.90 0.00 180 0.61 0.00 240 0.29 0.00 _____________________________________________
563
TE D
564
Table 2 Hydroquinone concentration (mg/L) in the hairless rat plasma (n=3) after
566
application of the four HQ products to the tape-stripped abdomen (effective diffusion
567
area; 1.77 cm2)
568
______________________________________________________________________ Time after HQ-1 HQ-2 HQ-3 HQ-4 application (min) (mg/L) (mg/L) (mg/L) (mg/L) ______________________________________________________________________
570 571 572 573 574 575 576 577
AC C
569
EP
565
0 30 60
0.00 0.00 0.00
0.00 3.70 4.64
0.00 0.00 0.00
0.00 0.00 0.00
120 180 240
0.00 0.00 0.00
0.31 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.01
33
ACCEPTED MANUSCRIPT
578
______________________________________________________________________
579
HQ-1: 1.0%, HQ-2: 3.3%, HQ-3: 2.6%, and HQ-4: 0.3% LOD: 0.06 mg/L
580 581
Table 3 Estimated BMD10 and BMDL10 levels for adenoma in rats exposed to
583
hydroquinone for 103 weeks (NTP, 1989)
Gamma f) (Gamma Logistic
78.2918 78.2918 80.476
ｇ)
LogLogistic f) LogLogistic LogProbit f)
78.2918 78.2918
ｇ)
77.0747 78.2918
ｇ)
LogProbit Multistage (2-degree) f) Multistage (2-degree) Multistage (1-degree) f)
Weibull ｇ) Quantal-Linear 585 586 587 588 589 590 591
80.1013 78.2918 78.2918 76.2961
Scaled residual e)
34.2434 34.2434 42.4421
21.876 8.04E-12 34.7432
0 0) -0.26
1 1 0.659 1
34.1399 34.1399 35.7302 33.8199
20.9038
0 0 0.747 0
1 1 0.9979 0.9979
34.3586 34.3586 33.9716 33.9716
21.876 13.7266 21.8689 21.8689
0 0 -0.053 -0.053
0.2444 1
41.0694 34.2551
33.064 21.876
-0.298 0
1 0.9979
34.2551 33.9716
21.8689
0 -0.053
27.5444
AIC (Akaike’s Information Criterion) b) P-value from the chi-square test for global goodness of fit c)
AC C
584
a)
ｇ)
EP
Multistage (1-degree) Probit Weibull f)
78.2918 78.2918 76.2961 76.2961
BMDL10 d) (mg/kg bw)
1 1 0.2076
TE D
ｇ)
P-value
BMD10 c) (mg/kg bw)
SC
AIC
b)
M AN U
Model Name
a)
RI PT
582
BMD10 (Benchmark Dose): maximum likelihood estimate of the exposure level associated with a 10% extra risk d) BMDL10: 95% lower confidence limit on the BMD10 e) Scaled residuals for local measurement f) Restricted models ｇ)
Unrestricted models
592
34
ACCEPTED MANUSCRIPT
1.400
RI PT
1.200 1.000 0.800 0.600
0.200 0.000 0
30
120
Fig. 1
596
TE D
597
(a) 120
150
180
EP
100 80
40 20 0
AC C
60
0
60
120 180 240 300 360 420 480
Time (min)
(b) 0.6
Cumlative amount of HQ (%/cm2)
595
Cumulative amount of HQ (µg/cm2)
90
(min)
594
598
60
SC
0.400
M AN U
HQ concentration in rat plasma (mg/L)
593
0.5 0.4 0.3 0.2 0.1 0 0
60
120 180 240 300 360 420 480
Time (min)
Fig. 2
35
ACCEPTED MANUSCRIPT
HQ-4
RI PT
HQ-3
0.1
0.01
SC
HQ-2
1
0.001
0.0001
0.00001 0
2
M AN U
HQ-1
Comulative amount of HQ permeated (%/cm2)
10
4
Time (h)
EP
Fig. 3
AC C
600
TE D
599
36
6
8
ACCEPTED MANUSCRIPT
HQ-4
RI PT
HQ-3
SC
HQ-2
10
1
M AN U
HQ-1
Comulative amount released (%/cm2)
100
0.1 0
60
120
180
240
300
Time (min)
601
Fig. 4
TE D
602
EP
0.02
0.015
AC C
Blood concentration of HQ (mg/L)
0.025
0.01
0.005
0
0
5
10
HQ-2 leave on for 8 h and wash off (night-time)
603 604
15
20
Time (h)
Fig. 5
37
25
30
360
ACCEPTED MANUSCRIPT
0.03
Blood concentration of HQ (mg/L)
0.025
RI PT
0.02 0.015 0.01
0 4
8
HQ-2 leave on for 8 h and wash off (night-time)
16
20
24
HQ-2 leave on for 16 h and wash off (day-time)
Time (h)
605 606
12
M AN U
0
SC
0.005
Fig. 6
TE D
60 50
EP
40 30 20
AC C
Comulative pereation of HQ (µg/cm2)
70
10
0
0
4
8
12
16
Time (h)
607 608
Fig 7
609
38
20
24
ACCEPTED MANUSCRIPT
610
612
Fig. 1 Hydroquinone (HQ) concentration in hairless rat plasma (n=3) after i.v. administration of 2% HQ aqueous solution
RI PT
611
613
Fig. 2 Cumulative amount as (a) a mass or (b) a percentage of hydroquinone (HQ)
615
permeated from 1 mL of 2% hydroquinone aqueous solution through the intact hairless
616
rat skin (mean and SE, n = 3)
M AN U
SC
614
617
Fig. 3 Cumulative amount (%) of hydroquinone (HQ) permeated from the four
619
cosmetics (HQ-1, HQ-2, HQ-3, and HQ-4) through the in vitro intact hairless rat skin
620
(mean and SE, n = 3)
EP
621
TE D
618
Fig. 4 Comulative amount (%) of hydroquinone (HQ) released from the four cosmetics
623
(HQ-1, HQ-2, HQ-3, and HQ-4) through the cellulose dialysis membrane (mean
624 625
AC C
622
and SE, n = 3)
626
Fig. 5 Estimated blood concentrations of hydroquinone (HQ) after a single dermal
627
application of 1.5 g HQ-2 on a whole face, a décolleté and hands of 1440 cm2. Time
628
course of plasma HQ concentrations was predicted from the skin permeation rate (Jskin) 39
ACCEPTED MANUSCRIPT
629
obtained from in vitro intact rat skin data and elimination rate constant (Kel) by a
630
convolution method.
RI PT
631
Fig. 6 Estimated blood concentrations of hydroquinone (HQ) after a 2-time dermal
633
application of each 1.5 g HQ-2 on a whole face, a décolleté and hands of 1440 cm2.
634
Time course of plasma HQ concentrations was predicted from the skin permeation rate
635
(Jskin) obtained form in vitro intact rat skin data and elimination rate constant (Kel) by a
636
convolution method.
M AN U
SC
632
637
Fig 7. Estimated cumulative permeation levels of hydroquinone (HQ) after a 2-time
639
dermal application of HQ-2 (1.5 g/time) on a whole face, a décolleté and hands of 1440
640
cm2
EP AC C
641
TE D
638
40
ACCEPTED MANUSCRIPT
Highlights
Rat skin permeation rates of hydroquinone (HQ) and HQ cosmetics were determined. One cosmetic showed a higher permeation rate than that of HQ aqueous solution. Plasma HQ levels after application of this cosmetic in humans were predicted. The margin of exposure (MOE) was estimated with a human dose level and a
benchmark dose. The MOE is judged to be in a range safe for the consumer if HQ cosmetics are used properly.
AC C
EP
TE D
M AN U
SC
RI PT
S0273-2300(16)30226-4
DOI:
10.1016/j.yrtph.2016.08.005
Reference:
YRTPH 3647
To appear in:
Regulatory Toxicology and Pharmacology
Received Date: 29 March 2016 Revised Date:
4 August 2016
Accepted Date: 8 August 2016
Please cite this article as: Matsumoto, M., Todo, H., Akiyama, T., Hirata-Koizumi, M., Sugibayashi, K., Ikarashi, Y., Ono, A., Hirose, A., Yokoyama, K., Risk assessment of skin lightening cosmetics containing hydroquinone, Regulatory Toxicology and Pharmacology (2016), doi: 10.1016/j.yrtph.2016.08.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Risk assessment of skin lightening cosmetics containing hydroquinone
2
Mariko Matsumoto1,4, Hiroaki Todo2, Takumi Akiyama3, Mutsuko Hirata-Koizumi1,
4
Kenji Sugibayashi2, Yoshiaki Ikarashi3, Atsushi Ono1, Akihiko Hirose1 Kazuhito
5
Yokoyama4
SC
RI PT
3
7
1
8
2
9
3
M AN U
6
Division of Risk Assessment, National Institute of Health Sciences, Tokyo, Japan
Josai University Faculty of Pharmaceutical Sciences, Saitama, Japan Division of Environmental Chemistry, National Institute of Health Sciences, Tokyo,
Japan
11
4
12
of Medicine, Tokyo, Japan
EP
Department of Epidemiology and Environmental Health, Juntendo University Faculty
AC C
13
TE D
10
14
Corresponding author: Mariko Matsumoto, Division of Risk Assessment, Biological
15
Safety Center, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-ku,
16
Tokyo 158-8501, Japan, tel: +81-3-3700-9878, fax: +81-3-3700-1408,
17
e-mail: [email protected]
18
1
ACCEPTED MANUSCRIPT
19
Abstract Following reports on potential risks of hydroquinone (HQ), HQ for skin
21
lightening has been banned or restricted in Europe and the US. In contrast, HQ is not
22
listed as a prohibited or limited ingredient for cosmetic use in Japan, and many HQ
23
cosmetics are sold without restriction. To assess the risk of systemic effects of HQ, we
24
examined the rat skin permeation rates of four HQ (0.3%, 1.0%, 2.6%, and 3.3%)
25
cosmetics. The permeation coefficients ranged from 1.2 × 10−9 to 3.1 × 10−7 cm/s, with
26
the highest value superior than the HQ aqueous solution (1.6 × 10−7 cm/s). After dermal
27
application of the HQ cosmetics to rats, HQ in plasma was detected only in the
28
treatment by highest coefficient cosmetic. Absorbed HQ levels treated with this highest
29
coefficient cosmetic in humans were estimated by numerical methods, and we
30
calculated the margin of exposure (MOE) for the estimated dose (0.017 mg/kg-bw/day
31
in proper use) to a benchmark dose for rat renal tubule adenomas. The MOE of 559 is
32
judged to be in a range safe for the consumer. However, further consideration may be
33
required for regulation of cosmetic ingredients.
AC C
EP
TE D
M AN U
SC
RI PT
20
34 35
Keywords: Hydroquinone, dermal absorption, permeation coefficients, benchmark dose,
36
margin of exposure, risk assessment
2
ACCEPTED MANUSCRIPT
Abbreviations
38
HQ: hydroquinone; i.v.: intravenous; LOAEL: lowest observed adverse effect level;
39
AUC: area under curve, BMD: benchmark dose; BMDL: lower confidence limit BMD;
40
MOE: margin of exposure; POD: point of departure
AC C
EP
TE D
M AN U
SC
RI PT
37
3
ACCEPTED MANUSCRIPT
41
1. Introduction
Hydroquinone (HQ) is used in skin bleaching agents, hair dyes, and finger nail
43
treatments (FDA, 2009). The WHO reported that a 1% HQ aqueous solution or a 5%
44
HQ cream caused dermal irritation in humans (WHO, 1996). Prolonged use of HQ
45
products (1–2%) is associated with exogenous ochronosis (Findlay et al., 1975), and a
46
worldwide total of 789 cases of exogenous ochronosis had been reported by 2007
47
(Levitt, 2007). In addition to these topical local effects, concerns have been raised
48
regarding the carcinogenic potential of HQ due to its carcinogenicity concerns reported
49
by animal studies (NTP, 1989; Shibata et al., 1991). Therefore, cosmetic use of HQ for
50
skin lightening has been banned in the UK and EU (EC, 2009). In the US, only
51
prescription skin lightening products can contain from greater than 2 to 4% HQ, and 2%
52
or less is allowed for cosmetic use (FDA, 2009). Recently, the US Cosmetic Ingredient
53
Review (CIR, 2014) concluded that HQ is safe at concentrations less than 1% in
54
cosmetic formulations designed for discontinuous, brief use followed by rinsing from
55
the skin and hair. On the other hand, HQ is not listed as a prohibited or limited
56
ingredient for cosmetic use in Japan (MHW Japan, 2000), and many kinds of skin
57
lightening cosmetics containing HQ are sold in Japan, some of which contain up to 10%
AC C
EP
TE D
M AN U
SC
RI PT
42
4
ACCEPTED MANUSCRIPT
HQ. Therefore, further information is required to evaluate whether the current use of
59
HQ cosmetics needs to be reconsidered in Japan.
RI PT
58
Dermal absorption of HQ was previously studied in 14 humans, and HQ was
61
dermally absorbed in humans with a bioavailability of 45.3 ± 11.2% for a 24-h
62
application (Wester et al., 1998). The blood elimination half-lives of HQ in a male
63
volunteer was reported to be 16.6 min for oral administration (NDMA, 1994), and that
64
for rats was 18.7 min for i.v. administration (Fox et al., 1986). Barber et al. (1995)
65
indicated that permeation of HQ in the human stratum corneum was slower than that in
66
fully thick rat skins. The permeability constant (K) values of
67
estimated to be 2.6 × 10−9 cm/s for human skin and 6.3 × 10−9 cm/s for rat skin (Barber
68
et al., 1995). Bucks et al. (1988) showed that penetration of HQ was reduced in
69
presence of a sunscreen and increased with a penetration enhancer in humans (Bucks et
70
al., 1988). Dermal absorption of HQ from cosmetics can vary among products because
71
the absorption rate varies depending on the base formulations (Ratna, 2004). In this
72
study, we examined rat skin permeation rates for four commercially available HQ
73
cosmetic products (HQ-1, HQ-2, HQ-3, and HQ-4) using a side-by-side diffusion cell
74
system to predict plasma HQ concentrations in humans after dermal absorption.
75
14
C-HQ solution were
AC C
EP
TE D
M AN U
SC
60
Because HQ is dermally absorbed, risk assessment for systemic effects, such as 5
ACCEPTED MANUSCRIPT
general repeated toxicity, carcinogenicity, and reproductive/developmental toxicity of
77
HQ are required. HQ induced sister chromatid exchanged, chromosome aberrations,
78
and/or gene mutation in vitro (Galloway et al., 1987; Tsutsui et al., 1997). Mutagenic
79
carcinogens are generally considered to have irreversible effects. If HQ carcinogenesis
80
is related to mutagenic events, the no-threshold concept should be applied for risk
81
assessment. A transgenic mouse mutation assay in the target organs of carcinogenicity is
82
useful to find if carcinogenesis is related to mutagenic events. We recently reported that
83
HQ is not mutagenic in transgenic MutaTM mice (Matsumoto et al., 2014), suggesting
84
that the mutagenic mechanism is not responsible for HQ induced carcinogenesis.
85
McGregor (2007) also suggested that the renal tumors are exacerbated non-genotoxic
86
spontaneous rodent renal disease that has no relevance to humans. In this study, we
87
estimate human absorbed levels of HQ after dermal application of the HQ cosmetic with
88
the highest permeation rate. Using the estimated human absorbed levels and
89
toxicological data of laboratory animals in the literature, a margin of exposure (MOE)
90
was calculated. In this paper, a risk assessment was conducted for the current use of HQ
91
cosmetics.
AC C
EP
TE D
M AN U
SC
RI PT
76
92 93
2. Materials and Methods
6
ACCEPTED MANUSCRIPT
94
2.1 Chemicals Four commercially available cosmetics (HQ-1, HQ-2, HQ-3, and HQ4) from
96
different manufacturers with different concentrations of HQ (unknown concentration,
97
3%, 10%, and 1%, respectively) were purchased. The measured HQ concentrations in
98
the products (HQ-1, HQ-2, HQ-3, and HQ-4) were 1.0%, 3.3%, 2.6%, and 0.3%,
99
respectively. From our survey, 10% HQ was the highest concentration available in the
100
market. Special grade HQ (CAS: 123-31-9; >99.0%) was purchased from Wako Pure
101
Chemical Industries, Ltd.
103
2.2. In vivo studies
104
2.2.1 Animals
TE D
102
M AN U
SC
RI PT
95
Male hairless rats (WBN/ILA-Ht; eight weeks old) were purchased from Life
106
Science Research Center, Josai University. This species was chosen because of its wide
107
use in toxicity and toxicokinetic studies and most commonly used for in vivo
108
permeation studies. Rats were reared on a basal diet (oriental yeast) and water ad
109
libitum. Animals were treated according to the ethical committee guidelines of Josai
110
University. Hair on the abdomen was shaved about 30 min after anesthesia of sodium
111
pentobarbital (50.0 mg/kg bw, intraperitoneal). After 12 h of fasting, in vivo
AC C
EP
105
7
ACCEPTED MANUSCRIPT
112
examination of rats was conducted. During the examination, body temperature of the
113
rats was kept using an electrical carpet.
RI PT
114
2.2.2 Blood concentration of HQ by dermal application of HQ aqueous solution or HQ
116
products
SC
115
The male hairless rats described in 2.2.1 were used. A diffusion cell (effective
118
surface area: 1.77 cm2) was glued with a biological glue (Aron Alpha) on the abdominal
119
intact skins, and 1 mL of 2% HQ aqueous solution (vehicle: saline) was applied.
120
Stripped skins were prepared by repeating the stripping 20 times using adhesive tape
121
(Nichiban Co., Ltd.). HQ aqueous solution was applied in the same manner as the intact
122
skins. HQ products (HQ-1, HQ-2, HQ-3, and HQ-4) were also applied to the stripped
123
skins as follows: The HQ products (about 100 mg) were applied on the skins, and
124
diffusion cells filled with the HQ products were applied on the skins. The blood samples
125
(0.2 mL) were taken from the cervical vein at 0, 30, 60, 120, 180, and 240 min after
126
application.
AC C
EP
TE D
M AN U
117
127 128 129
2.2.3 Blood concentration of HQ by intravenous administration Three rats were intraperitoneally anesthetized with urethane at 1.0 mg/kg. HQ
8
ACCEPTED MANUSCRIPT
aqueous solution at 2% (10 mg/kg) was administrated intravenously (i.v.), and blood
131
samples (0.2 mL) were taken at 0, 15, 30, 60, 90, and 180 min after administration. The
132
blood samples were centrifuged (15,000 rpm, 4°C, 5 min), and each 0.1 mL of blood
133
plasma was collected into a 1.5 mL microtube and kept in a freezer at −30°C.
134
2.2.4 Preparation of HQ cosmetics for UPLC
M AN U
135
SC
RI PT
130
Each 50 µL of HQ cosmetics was taken into a microtube, and the weight of the
137
cosmetics was precisely measured. The sample was completely dissolved in
138
tetrahydrofuran (4.5 times in weight) with vigorous mixing. After adding water (4.5
139
times in weight), the mixture was sonicated for 5 min. and centrifuged. The lower layer
140
of supernatant was taken and precisely diluted 10 times with water.
EP
141
TE D
136
2.2.5 Determination of HQ concentration
143
Concentrations of HQ were determined by UPLC (ACQUITY UPLC system H-CLASS;
144
Waters Company) equipped with a photodiode array detector (PDA eλ detector; Waters
145
Company). Limit of detection (LOD) of HQ was determined as signal-to-noise ratio of
146
3 and was 0.06 mg/mL. An ACQUITY BEH C18 column (size: 2.1 mm i.d. × 50 mm;
147
particle size: 1.7 µm; Waters Company) was used at 40℃. Five microliter of test
AC C
142
9
ACCEPTED MANUSCRIPT
solution was injected. Acetonitrile (5% or 10%) was used as a mobile phase and a flow
149
rate was set to 0.1 mL/min. The scanning range of wavelength was set to 210–400 nm.
150
HQ was detected and quantitated at 288–289 nm. The absolute calibration curve method
151
was used for measurement of HQ. Water solution containing special grade HQ was used
152
as a standard solution.
SC
RI PT
148
M AN U
153 154
2.3 In vitro studies
155
2.3.1 Determination of HQ concentration for release and permeation studies HQ release and skin permeation from the studied products were examined by a
157
side-by-side diffusion cell system. For the release study, a cellulose dialysis membrane
158
(UC24-32-100, EIDIA Co., Ltd., Tokyo, Japan) was placed in between the diffusion
159
cells. The cellulose dialysis membrane was chosen because it is widely used for drug
160
release studies. For the permeation study, abdominal skin pieces were excised from rats
161
under anesthesia of sodium pentobarbital (50.0 mg/kg bw, intraperitoneal) and one piece
162
of intact skin was set in between the diffusion cells (effective surface area: 0.95 cm2).
163
The intact rat skin data were used for a realistic exposure assessment because
164
HQ-containing products are generally applied to intact human skin. Full-thickness skins
165
were used as defined by OECD TG 428 in this study (OECD, 2004). Permeability of
AC C
EP
TE D
156
10
ACCEPTED MANUSCRIPT
hydrophilic chemicals like HQ would not be affected by skin thickness in contrast to
167
lipophilic chemicals (Yamaguchi et al., 2008). The thickness of the skin was measured
168
according to a method reported previously (Watanabe et al., 2001). The
169
biotransformation was not inhibited in this system. Each HQ product (HQ-1, HQ-2,
170
HQ-3, or HQ-4 at 250 mg) was applied on one side of the cellulose membrane or the
171
stratum corneum surface in a donor compartment. The other side of the compartment, a
172
receiver compartment, was filled with distilled water (3.0 mL). The temperature of skin
173
surface in the diffusion cell system was kept at 32°C, and the receiver compartment was
174
agitated with a magnetic stirrer using a star-head magnetic bar. Samples (0.5 mL each)
175
were withdrawn at appropriate intervals (every 5-30 min for first one or two hours, and
176
hourly afterward) from the receiver compartment and immediately replaced with 0.5 mL
177
fresh distilled water. Each sample was added to a same volume of methanol and
178
centrifuged (15,000 × g for 5 min at 4°C). The supernatant contents were determined by
179
HPLC equipped with a photodiode array detector (Shimadzu LC-20AD, Shimadzu
180
SPDM-20A, Kyoto, Japan). It was validated by linearity, and the lowest concentration
181
was set as 0.1 µg/mL. Conditions of the HPLC were as follows: mobile phase was water
182
and methanol (75:25); column was Inertsil ODS-3, 5 µm, 4.6 × 150 mm; column
183
temperature was 40°C; detection wavelengths were 280 nm (excitation) and 310 nm
AC C
EP
TE D
M AN U
SC
RI PT
166
11
ACCEPTED MANUSCRIPT
184
(emission); flow rate was 1.0 mL/min; injection volume was 5 µL.
185
2.4 Theoretical approach
187
2.4.1 Determination of permeation parameters (in vitro data)
188
In this study, human absorption levels were estimated with in vitro permeation data of
189
rats. Nakamura et al., (2012) indicated that the predicted absorption profiles obtained
190
from rat skins in vitro were well agreed with human clinical values.
191
Skin permeation of HQ can be determined by Fick’s second law of diffusion:
SC
=
(1),
M AN U
192
RI PT
186
where Dskin is the diffusion coefficient of HQ across the skin, and Cskin is the skin
194
concentration of HQ at a time t and a position of the skin . Cskin was calculated using
195
Eq. 1 with the following initial and boundary conditions:
196
t = 0, 0 <
197
t > 0, when = 0,
EP
AC C
198
TE D
193
Cskin = Kskin・Cdonor and
=
when = Lskin, Cskin = 0 (Boundary conditions),
199
where Lskin is the thickness of the skin, Kskin is the partition coefficient of HQ from the
200
donor to the skin, and Cdonor is the HQ concentration in the donor.
201
The skin permeation rate to the receiver, Jskin is expressed by Fick’s first law (Eq. 2),
12
ACCEPTED MANUSCRIPT
204
205 206 207
(2).
The cumulative amount of drug permeated per unit area, Q is expressed by Eq. 3
Q = − "
! (3).
RI PT
203
= −
Differential terms in Fick’s second law (Eq. 1) are approximated as follows: #,% #
,% #
&
= ∆ () ,*+& − ) ,* , (4), and =
&
∆
() -&,* − 2) ,* + ) +&,* , (5),
SC
202
where Ci,j is HQ concentration at i-th position and j-th time, ∆ is
209
tj+1-tj. By substation of the above equations (Eq. 4 and Eq.5) to Fick’s second law (Eq.1),
210
the following equation is obtained. 0
0
i+1- i,
and ∆t is
0
) ,*+& = 0 ) ,*-&,* + 1 − 2 0 ) ,* + 0 ) +&,* (6).
TE D
211
M AN U
208
The above differential equations (Eq.2 and Eq.3) can be approximated to the following
213
difference equations:
214
* = −
23,% ∆
(7), and
AC C
215
EP
212
4* = 4*-& + * ・∆! (8),
216
where n is the number of divisions of skin. Then, Jj was calculated by setting n=10. In
217
this calculation, ∆t was set to be less than 0.5 for ∆t/∆2 Dskin. Qj was obtained by the Eq
218
(8) with a value of Jj. The effective diffusion coefficient, Dskin and the partition
219
coefficient, Kskin were obtained with the least-square method by fitting observed values.
13
ACCEPTED MANUSCRIPT
The calculation was conducted by a pseudo-Newtonian method of the Microsoft Excel
221
(Office 2013) solver-function (Max time: 100 seconds, Iterations: 100, Precision:
222
0.000001, Tolerance: 5%, Convergence: 0.001).
RI PT
220
223
2.4.2 Determination of parameters (in vivo data)
225
The amount of HQ in the body at any particular time after i.v. administration is given by
226
the following equation: = " 5 - 67 (9),
M AN U
227
SC
224
where is the amount remaining to be eliminated, 0 is the dose administered (initial
229
amount), kel is an elimination rate constant, and t is time. In vivo blood concentration
230
data (i.v. administration of 2% HQ aqueous solution) were used for data fitting. The
231
non-linear least squares Damping Gauss–Newton method was used for parameter
232
estimation and the elimination rate constant kel was calculated. The Microsoft Excel
233
(Office 2013) was used for calculation.
235
EP
AC C
234
TE D
228
2.4.3 Estimation of in vivo blood concentration-time profile
236
Time course of plasma HQ concentrations after dermal absorption was
237
predicted from in vitro permeation data and in vivo i.v. data by a numerical convolution
14
ACCEPTED MANUSCRIPT
method with the following assumption: the elimination rate of HQ is linear and the
239
permeation rate of HQ is identical in in vitro and in vivo. Using the two parameters
240
(Dskin and Kskin) determined in 2.4.1, cumulative amounts of HQ and skin permeation
241
rates (Jskin) of HQ at each sampling time were calculated on the assumption that 1.5 g
242
HQ-2 was applied on a whole face, a décolleté and hands (1440 cm2). HQ-2 was
243
selected because it showed the highest permeation coefficient among the four cosmetic
244
products. The blood concentration of HQ after a single dermal application of 1.5 g
245
HQ-2 on a whole face, a décolleté and hands (left for 8 hours overnight and washed off),
246
and after 2-time dermal applications of 1.5 g/time HQ-2 on a whole face, a décolleté
247
and hands (left for 8 hours overnight and washed off, and left for 16 hours during
248
day-time and washed off) were estimated. The Microsoft Excel (Office 2013) was used
249
for calculation.
251 252
SC
M AN U
TE D
EP AC C
250
RI PT
238
2.5 Determination of benchmark dose In the literature, the lowest LOAEL (lowest observed adverse effect level) of a
253
repeated dose of HQ was reported to be 17.9 mg/kg bw/day (25 mg/kg bw, 5 days/week
254
for 103 weeks) for general toxicity due to lowered body weight and for carcinogenicity
255
due to renal tubule adenomas in rats given HQ by gavage (NTP, 1989). To find out a
15
ACCEPTED MANUSCRIPT
better point of departure (POD), a benchmark dose approach was performed using the
257
benchmark dose software (BMDS) 2.5.0 developed by US EPA (US EPA, 2014). In the
258
103-week study, numbers of adenoma found were 0/55, 4/55, and 8/55 at 0, 25, and 50
259
mg/kg bw/day, respectively. Using these data, a lower 95% confidence limit of BMD at
260
10% response level (BMDL10) was determined. The following models were used for the
261
analysis: Gamma (restricted or unrestricted), Logistic, LogLogistic (restricted or
262
unrestricted), LogProbit (restricted or unrestricted), Multistage (restricted or
263
unrestricted), Probit, Weibull (restricted or unrestricted), and Quantal-Linear. Model fit
264
was judged by P-values for goodness of fit (P > 0.1 is considered to be a good fit),
265
visual inspection of the dose response curve, and scaled residuals at the data point
266
(absolute value of scaled residuals < 2.0 is considered to be a good fit).
267
269
3. Results
AC C
268
EP
TE D
M AN U
SC
RI PT
256
Table 1 presents HQ concentration in rat plasma after application of 2% HQ
270
aqueous solution to the tape-stripped or intact skins of the abdomen. HQ was not
271
detected for the intact skin. The highest level was detected at 60 min after treatment of
272
the abdominal tape-stripped skins.
273
HQ concentrations in the rat plasma after application of the four HQ cosmetics
16
ACCEPTED MANUSCRIPT
to the tape-stripped abdomen are shown in Table 2. The highest concentration of 4.64
275
mg/L HQ was observed at 60 min after application of HQ-2. HQ was not detected in the
276
rat plasma treated by HQ-1, HQ-3, and HQ-4.
RI PT
274
The measured HQ concentrations in the products (HQ-1, HQ-2, HQ-3, and
278
HQ-4) were 1.0%, 3.3%, 2.6%, and 0.3%, respectively. The measurements were
279
conducted as described in 2.2.4 and 2.2.5. The elimination rate constant of HQ, Kel was
280
calculated to be 2.1/h using i.v administration data (Fig. 1).
M AN U
SC
277
Fig. 2 shows the cumulative amount of HQ permeated through the hairless rat
282
skins from a 2% aqueous solution. The permeation coefficient of the HQ aqueous
283
solution was calculated to be 1.6 × 10−7 cm/s for the intact hairless rat skin.
TE D
281
Fig. 3 shows the cumulative amount (%) of HQ permeated from the HQ
285
products through the intact hairless rat skin. The permeation coefficients of HQ were
286
estimated to be HQ-2 (3.1 × 10−7 cm/s) > HQ-4 (1.1 × 10−8 cm/s) > HQ-1 (7.0 × 10−9
287
cm/s) > HQ-3 (1.2 × 10−9 cm/s) for the rat skin. The permeation coefficients of HQ for
288
HQ-4, HQ-1, and HQ-3 were lower than that of the HQ aqueous solution.
AC C
EP
284
289
Fig. 4 shows the cumulative amount (%) of HQ released from HQ products.
290
The release amount of HQ was the highest in HQ-2, followed by HQ-3, HQ-4, and HQ1.
291
The effective diffusion coefficient, Dskin of HQ for HQ-2 was 2.2 × 10−4 cm/s, and the
17
ACCEPTED MANUSCRIPT
292
partition coefficient of HQ for HQ-2, Kskin was 0.32.
293
RI PT
294
The estimated blood concentrations after a single dermal application of HQ-2
296
on the whole face, a décolleté and hands is shown in Fig. 5, and after the 2-time dermal
297
application of HQ-2 on the whole face, a décolleté and hands in Fig. 6.
M AN U
SC
295
The areas under the curves (AUCs) for 24 h after the single and 2-time
299
applications (1.5 g/time) on a whole face, a décolleté and hands (1440 cm2) were
300
calculated to be 0.183 mg h/L and 0.310 mg h/L, respectively. The blood concentrations
301
of HQ were calculated on the following two assumptions. Single application: left for 8
302
hours overnight and washed off. Two-time application: left for 8 hours overnight and
303
washed off, and left for 16 hours during day-time and washed off. After the single
304
application, the blood concentration of HQ was considered to reach zero at 30 h, and the
305
AUC (30 h) was calculated to be 0.186 mg h/L. The cumulative permeation level of HQ
306
at 24 h after the 2-time application of HQ-2 was estimated to be 60.9 µg/cm2 (Fig. 7).
307
Using this value, the dermal permeation of HQ for 24 h (one day) for a 50 kg adult
308
person was estimated to be 1.75 mg/kg bw/day (60.9 × 10−3 mg/cm2 × 1440 cm2 per 50
309
kg bw) for an adult person.
AC C
EP
TE D
298
18
ACCEPTED MANUSCRIPT
For the BMD analysis, fitting was successfully conducted in all the models, but
311
BMDL10 values were not obtained by the unrestricted LogLogistic, LogProbit, or
312
Weibull models. The estimated BMD10 and BMDL10 for the 103-week study are shown
313
in Table 3. The BMDL10 ranged from 13.7266 mg/kg bw to 34.7432 mg/kg bw. The
314
unrestricted models for Gamma, LogLogistic, LogProbit, and Weibull did not provide
315
an acceptable BMDL10. The Akaike’s Information Criterion (AIC) ranged from 76.2961
316
to 80.476. P-values for global goodness of fit were all acceptably large (>0.1). Absolute
317
values of scaled residuals for local measurement were all acceptably small (<2.0).
M AN U
SC
RI PT
310
319
4. Discussion
TE D
318
When the 2% HQ aqueous solution was applied to the tape-stripped skin under
321
anesthesia for 4 hours, HQ was detected in the plasma of rats, with the highest level
322
detected at 60 min. On the other hand, HQ was not detected in the plasma after the 2%
323
aqueous solution was applied to the intact skins of rats, indicating that the stratum
324
corneum worked as a barrier for the permeation of HQ. Because the rate controlling
325
factor of permeation of chemicals is permeability of stratum corneum, anesthesia
326
condition would not affect skin absorption as long as temperature of skin surface was
327
kept 32 ± 1 °C (OECD, 2004). In this study, skin surface was kept at 32°C to assure skin
AC C
EP
320
19
ACCEPTED MANUSCRIPT
temperature was not lowered by anesthesia. HQ was detected in the plasma after
329
application of HQ-2, but was not detected after application of HQ-1, HQ-3, and HQ-4
330
even for the tape-striped skin. 1,2-Hexanediol included in HQ-2 could be acted as a
331
permeation enhancer (Warner et al., 2001). However, a known enhancer such as ethanol,
332
stearic acid, or ascorbic acid was included in other products. Permeability differs by
333
enhancer concentration (Warner et al., 2001), and there were so many ingredients
334
included in the products; therefore, it is difficult to conclude what enhancer affected
335
permeability. In addition, many ingredients included in the products can affect
336
thermodynamic activity of HQ. Because the thermodynamic activity is known to affect
337
skin permeation of drugs (Ishii et al., 2010), this result may also be related to the
338
thermodynamic activities of HQ in the cosmetic formulations.
TE D
M AN U
SC
RI PT
328
The plasma levels of HQ were rapidly decreased after i.v. administration of the
340
2% HQ aqueous solution in rats and Kel was calculated to be 2.1/h, which indicates that
341
the metabolism of HQ is rapid. This is in good agreement with the short elimination
342
half-lives of HQ radiolabel; 16.6 min for a male volunteer after ingestion (NDMA,
343
1994) and 18.7 and 14.8 min for rats after i.v. and oral administrations (Fox et al., 1986).
344
After oral administration of HQ in rats, <1% of the detected radiolabel in blood was
345
associated with the parent compound, indicating rapid metabolism of HQ (Fox et al.,
AC C
EP
339
20
ACCEPTED MANUSCRIPT
1986). In general, animal skin absorption is greater than human skins; however, no
347
concrete conclusion on human skins would be derived from findings of in vivo data in
348
this study. We could not detect HQ in the blood of rats after dermal application in the
349
intact skin study. It cannot conclude that HQ was not absorbed, because elimination of
350
HQ from the blood is rapid. It was suggested that the HQ levels under the LOD in the
351
blood could be a target of this study. In fact, the highest estimated blood concentration
352
of HQ for application of HQ-2 was lower than the LOD (Fig. 5 and Fig. 6). Therefore,
353
we further evaluated HQ cosmetics by an in vitro method, which is useful for predicting
354
the pharmacokinetic profile of chemicals in human (Nakamura et al., 2012).
M AN U
SC
RI PT
346
The permeation coefficient of the 2% HQ aqueous solution was calculated to
356
be 1.6 × 10−7 cm/s and was higher than the previously reported value of 6.3 × 10-9 cm/s
357
for the
358
cells (Barber et al., 1995). The permeation coefficients of HQ were estimated to be
359
HQ-2 (3.1 × 10−7 cm/s) > HQ-4 (1.1 × 10−8 cm/s) > HQ-1 (7.0 × 10−9 cm/s) > HQ-3 (1.2
360
× 10−9 cm/s) for the rat skin. The previously reported permeation coefficient value was
361
within the range of three HQ products with lower permeation coefficients. Factors that
362
might contribute for the difference are unclear because Barber et al., (1995) did not
363
clearly state the concentration or the vehicle of the HQ solution. When relatively higher
C-HQ solution to full-thickness Fisher 344 rat skin in Franz-type diffusion
AC C
EP
14
TE D
355
21
ACCEPTED MANUSCRIPT
permeation coefficient and absorption of HQ through the rat skins are taken into
365
account for HQ-2, HQ is expected to be absorbed and distributed in the body after
366
application of HQ-2 in humans. This indicates that not only local effects of HQ but also
367
systemic effects have to be considered for dermal application of HQ. The permeation
368
coefficients of HQ for HQ-4, HQ-1, and HQ-3 were lower than that of the HQ aqueous
369
solution. This result is consistent with the in vivo tape-stripped study. The release rate of
370
HQ was the highest in HQ-2, followed by HQ-3, HQ-4 and HQ-1. This order did not
371
exactly match the order of the permeation coefficient, but the high release rate may
372
partially contribute to the permeability of the HQ cosmetics.
M AN U
SC
RI PT
364
The measured concentrations of HQ in HQ-1, HQ-2, HQ-3, and HQ-4 were
374
1.0%, 3.3%, 2.6%, and 0.3%, respectively. The values of HQ-3 and HQ-4 were lower
375
than those of labelling on the products. These differences may be related to the process
376
of manufacturing or degradation of HQ. In any case, we confirmed that two products
377
(HQ-2 and HQ-3) contained HQ at a level greater than 2%. HQ-2 showed a higher
378
permeation coefficient of HQ than that of the HQ aqueous solution, and dermal
379
absorption was confirmed. On the other hand, HQ was not detected in the plasma of rats
380
after dermal application of HQ-3 in spite of its relatively high concentration of HQ
381
(2.6%). Because a number of products examined was relatively small, it can be difficult
AC C
EP
TE D
373
22
ACCEPTED MANUSCRIPT
to draw solid conclusions for effects of formulation or concentration of HQ. However,
383
our findings suggest that regulating based on the concentration of a target substance
384
may not be safe enough for dermally applied cosmetics. Therefore, especially high
385
concentration products may need to be tested for their permeability for a safer use.
RI PT
382
In the literature, the lowest LOAEL of a repeated oral dose of HQ was reported
387
to be 17.9 mg/kg bw/day for general toxicity due to decreased body weights and for
388
carcinogenicity due to renal tubule adenomas in rats given HQ by gavage for 103 weeks
389
(NTP, 1989). As for a reproductive/developmental endpoint, a study by Murphy et al.,
390
(1992) showed the lowest NOAEL of 75 mg/kg bw/day, in which marginal increases of
391
skeletal and internal malformations were observed at a maternal toxic dose of 150
392
mg/kg bw/day. Therefore, reproductive/developmental toxicity was not considered a
393
specifically sensitive endpoint. HQ carcinogenesis is not considered to be related to
394
mutagenic events (McGregor 2007, Matsumoto et al., 2014); therefore, the threshold
395
concept was applied for risk assessment.
M AN U
TE D
EP
AC C
396
SC
386
The range of the BMDL10s for the NTP 103-week study (the LOAEL = 25
397
mg/kg bw) calculated from the fitted models were between 13.7 mg/kg bw and 34.7
398
mg/kg bw. As a conservative risk assessment, the lowest BMDL10 (13.7 mg/kg bw) was
399
chosen for further quantitative assessment. After adjusting 5 days per week
23
ACCEPTED MANUSCRIPT
administration to 7 days per week administration, the POD for calculating MOE was
401
estimated as 9.78 mg/kg bw/day. The MOE was calculated by the POD (9.78 mg/kg
402
bw/day: oral dose) divided by the estimated human dose level (1.75 mg/kg bw/day) by
403
dermal application and was 5.59 for the worst case exposure scenario. By using the
404
BMDL value instead of the LOAEL for calculation of the MOE, uncertainty was
405
considered to be reduced. Orally administrated HQ is considered to get well absorbed in
406
humans (Deisinger et al., 1996), and >90% absorption was confirmed in rats
407
(Divincenzo et al., 1984). Therefore, for the calculation of the MOE, we assumed that
408
orally dosed HQ (9.78 mg/kg bw/day) to rats in the NTP study was 100% absorbed. If
409
the MOE is less than one, an immediate regulatory action might have been required for
410
HQ cosmetics in Japan. The MOE of 5.59 is less than an acceptable MOE of 25 for the
411
TK-based safety margins (WHO, 2005); however, this is for a worst case exposure
412
scenario.
SC
M AN U
TE D
EP
AC C
413
RI PT
400
According to the instructions of HQ-2, this cosmetic is recommended to be
414
used to spot areas and to avoid applying to skin areas without pigmented spots. If the
415
application was limited to spot areas, e.g., 1/100 of a face, a décolleté and hands area,
416
the MOE is judged to be in a range safe (MOE：559 >100: intra- and inter-species
417
differences) for a nonmutagenic carcinogen. HQ-containing products are generally
24
ACCEPTED MANUSCRIPT
applied to intact human skin, and HQ were not detected after application of HQ
419
cosmetics to intact rat skin. The permeation rate of HQ is expected to be slower in
420
humans compared to rats as stated before (Barber et al., 1995). Taken into account of
421
this, the MOE in reality could be improved. However, Wester et al. (1998) reported a
422
human in vivo absorption rate of 45%, and elimination of HQ in humans seemed a little
423
slower than in rats according to half-lives of HQ (14.8 min in rats and 16.6 min in
424
humans) for oral administration (NDMA, 1994; Fox et al., 1986). This could result
425
higher blood levels than our calculation based on elimination rate estimated from in vivo
426
rat data. Therefore uncertainties remain for our exposure estimation based on species
427
differences.
TE D
M AN U
SC
RI PT
418
Although HQ-1, HQ-2, and HQ-3 are recommended for topical use for
429
pigmented spots, HQ-4 is recommended to be used for the whole face. However, the
430
concentration of HQ-4 was very low. Besides, from the permeation coefficient of HQ
431
for HQ-4, its permeability into the skin was found to be “slow” according to the
432
qualitative ranking system by Marzulli et al., (1969), and the permeation rate is
433
expected to be slower in humans compared to rats (Barber et al., 1995). Based on this,
434
as long as users follow the instructions of the products, the risk of these HQ cosmetics is
435
considered to be low.
AC C
EP
428
25
ACCEPTED MANUSCRIPT
In the Japanese market, there are several other products containing HQ at >2%
437
and there are many products containing unknown levels of HQ that we haven’t assessed.
438
There is a possibility that these other products show a higher permeability than HQ-2.
439
We have to also note that MOE of >1000 could be preferable given the toxic property of
440
HQ (carcinogenesis). In addition, cosmetic users include pregnant women and women
441
planning a pregnancy in near future. HQ as a prescription product is assigned to
442
“pregnancy category C: the benefits from the use of the drug in pregnant women may be
443
acceptable despite its potential risks” by the FDA based on “no adequate studies in
444
humans, but adverse effects were observed on the fetus in animal reproduction studies”
445
(Drugs.com, 2014). It is still controversial whether Japanese consumers have enough
446
information for the safe use of cosmetics. Cosmetics are usually marketed based on only
447
safety evaluation tests for acute or local toxicity under the company’s self-imposed
448
control in Japan. Since the possibility that dermally absorbed cosmetic ingredients like
449
HQ may cause systemic adverse effects cannot be ruled out, further consideration may
450
be required for the regulation of cosmetic ingredients.
AC C
EP
TE D
M AN U
SC
RI PT
436
451 452 453
5. Acknowledgement This study was supported by the Ministry of Health, Labour and Welfare,
26
ACCEPTED MANUSCRIPT
454
Japan.
455
RI PT
456
AC C
EP
TE D
M AN U
SC
457
27
ACCEPTED MANUSCRIPT
458
References
459
461
Barber, E. D., et al., 1995. The percutaneous absorption of hydroquinone (HQ) through rat and human skin in vitro. Toxicol Lett. 80, 167-72.
RI PT
460
Bucks, D. A., et al., 1988. Percutaneous absorption of hydroquinone in humans: effect
463
of 1-dodecylazacycloheptan-2-one (azone) and the 2-ethylhexyl ester of
464
4-(dimethylamino)benzoic acid (Escalol 507). J Toxicol Environ Health. 24,
465
279-89.
M AN U
466
SC
462
CIR (2014) Amended Safety Assessment of Hydroquinone as Used in Cosmetics available
468
http://online.personalcarecouncil.org/ctfa-static/online/lists/cir-pdfs/FR647.pdf
471 472
Deisinger, P. J., et al., 1996. Human exposure to naturally occurring hydroquinone. J
EP
470
at
Toxicol Environ Health. 47, 31-46.
AC C
469
TE D
467
Divincenzo, G. D., et al., 1984. Metabolic fate and disposition of [14C]hydroquinone given orally to Sprague-Dawley rats. Toxicology. 33, 9-18.
473
Drugs.com, Hydroquinone cream. Vol. 2014, 2014.
474
EC, REGULATION (EC) No 1223/2009 OF THE EUROPEAN PARLIAMENT AND
475
OF THE COUNCIL of 30 November 2009 on cosmetic products. 2009.
28
ACCEPTED MANUSCRIPT
477 478 479
FDA, 2009. Hydroquinone [CAS 123-31-9] Supporting Information for Toxicological Evaluation by the National Toxicology Program 21 May 2009. Findlay, G. H., et al., 1975. Exogenous ochronosis and pigmented colloid milium from
RI PT
476
hydroquinone bleaching creams. Br J Dermatol. 93, 613-22.
Fox, J. A., et al., 1986. Blood Elimination Kinetics of [U-14C]Hydroquinone
481
Administered by Intragastric Intubation, Intratracheal Instillation or Intravenous
482
Injection to Male Fischer 344 Rats,. Rochester, NY, Health and Environment
483
Laboratories; Eastman Kodak Company. Report No. TX-86-1 (cited in
484
DeCaprio 1999; Crit Rev Toxicol: 29, 283-330).
M AN U
SC
480
Galloway, S. M., et al., 1987. Chromosome aberrations and sister chromatid exchanges
486
in Chinese hamster ovary cells: evaluations of 108 chemicals. Environ Mol
487
Mutagen. 10 Suppl 10, 1-175.
489 490 491
EP
Iimura, N., et al., 2005. Development of new whitening agents with hydroquinone
AC C
488
TE D
485
stabilized by the complex formation with surfactants and the evaluation for melanogenesis inhibitory effect and skin stimulus. J. Jpn. Cosmet. Sci. Soc. 29, 301-313.
492
Ishii, H., et al., 2010. Effect of thermodynamic activity on skin permeation and skin
493
concentration of triamcinolone acetonide. Chem Pharm Bull (Tokyo). 58,
29
ACCEPTED MANUSCRIPT
497 498 499 500
Federal Register. J Am Acad Dermatol. 57, 854-72.
RI PT
496
Levitt, J., 2007. The safety of hydroquinone: a dermatologist's response to the 2006
MHW Japan, 2000. Standards for Cosmetics Notification No. 331 of 2000 (in Japanese).
SC
495
556-61.
Marzulli, F. N., et al., 1969. Techniques for studying skin penetration. Toxicology and
M AN U
494
Applied Pharmacology. 14, Supplement 3, 76-83.
Matsumoto, M., et al., 2014. Evaluation of in vivo mutagenicity of hydroquinone in
502
Muta™ mice. Mutation Research/Genetic Toxicology and Environmental
503
Mutagenesis. 775–776, 94-98.
506 507 508 509
carcinogenic and mutagenic properties. Crit Rev Toxicol. 37, 887-914.
EP
505
McGregor, D., 2007. Hydroquinone: an evaluation of the human risks from its
Murphy, S. J., et al., 1992. A study of developmental toxicity of hydroquinone in the
AC C
504
TE D
501
rabbit. Fundam Appl Toxicol. 19, 214-21.
NDMA, 1994. Chronic Health Effects Testing for Hydroquinone (Nonprescription Drug Manufactures Association).
510
NTP, 1989. Toxicology and carcinogenesis studies of hydroquinone (CAS No.
511
123-31-9) in F344/N rats and B6C3F1 mice (gavage studies). National
30
ACCEPTED MANUSCRIPT
Toxicology Program Technical Report Series, No.366, U.S. Department of
513
Health and Human Surivice, Public Health Service, National Institutes of
514
Health.
RI PT
512
Nakamura, A., et al., 2012. Evaluation of the predicted time-concentration profile of
516
serum tulobuterol in human after transdermal application. Chem Pharm Bull
517
(Tokyo). 60, 300-5.
M AN U
SC
515
518
OECD, 2004. Test Guideline 428. Skin absorption: in vitro method, OECD, Paris
519
Ratna, M., 2004. Topical and Transdermal Drug Delivery: What a Pharmacist Needs to
520
Know. ACPE ID number 221-146-04-054-H01.
Shibata, M. A., et al., 1991. Induction of renal cell tumors in rats and mice, and
522
enhancement of hepatocellular tumor development in mice after long-term
523
hydroquinone treatment. Jpn J Cancer Res. 82, 1211-9.
525
EP
Smiles, K. A., et al., 2007. A hydroquinone formulation with increased stability and
AC C
524
TE D
521
decreased potential for irritation. J Cosmet Dermatol. 6, 83-8.
526
Tsutsui, T., et al., 1997. Benzene-, catechol-, hydroquinone- and phenol-induced cell
527
transformation, gene mutations, chromosome aberrations, aneuploidy, sister
528
chromatid exchanges and unscheduled DNA synthesis in Syrian hamster embryo
529
cells. Mutat Res. 373, 113-23.
31
ACCEPTED MANUSCRIPT
530
US EPA, Benchmark Dose Software (BMDS) version 2.5.0. 2014.
531
WHO, 1996. IPCS INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY
533
Health and Safety Guide No. 101.
RI PT
532
WHO, 2005. Harmonization Project Document No.2: Chemical-specific adjustment
factors for interspecies differences and human variability: guidance document
535
for use of data in dose/concentration – response assessment. World Health
536
Organization, Geneva, 2005.
M AN U
SC
534
Warner, K. S., et al., 2001. Influences of alkyl group chain length and polar head group
538
on chemical skin permeation enhancement. J Pharm Sci. 90, 1143-53.
539
Watanabe, T., 2001. Utility of the three-dimensional cultured human skin model as a
540
TE D
537
tool to evaluate skin permeation of drugs. Altern. Animal Test. Exp. 7, 1-14. Wester, R. C., et al., 1998. Human in vivo and in vitro hydroquinone topical
542
bioavailability, metabolism, and disposition. J Toxicol Environ Health A. 54,
544
AC C
543
EP
541
301-17.
Yamaguchi, K., et al., 2008. Structure-permeability relationship analysis of the
545
permeation
546
epidermis/dermis of rat skin. J Pharm Sci. 97, 4391-403.
barrier
properties
of
547
32
the
stratum
corneum
and
viable
ACCEPTED MANUSCRIPT
Table 1 Hydroquinone concentration (mg/L) in the hairless rat plasma (n=3) after
549
abdominal administration of 2% hydroquinone aqueous solution (effective diffusion
550
area; 1.77 cm2)
551
___________________________________________ Time after Tape-stripped Intact
553 554 555 556 557
administration (min) (mg/L) (mg/L) ___________________________________________ 0 0.21 0.00 30 1.78 0.00
SC
552
RI PT
548
562
LOD: 0.06 mg/L
558 559 560
M AN U
561
60 2.17 0.00 120 0.90 0.00 180 0.61 0.00 240 0.29 0.00 _____________________________________________
563
TE D
564
Table 2 Hydroquinone concentration (mg/L) in the hairless rat plasma (n=3) after
566
application of the four HQ products to the tape-stripped abdomen (effective diffusion
567
area; 1.77 cm2)
568
______________________________________________________________________ Time after HQ-1 HQ-2 HQ-3 HQ-4 application (min) (mg/L) (mg/L) (mg/L) (mg/L) ______________________________________________________________________
570 571 572 573 574 575 576 577
AC C
569
EP
565
0 30 60
0.00 0.00 0.00
0.00 3.70 4.64
0.00 0.00 0.00
0.00 0.00 0.00
120 180 240
0.00 0.00 0.00
0.31 0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.01
33
ACCEPTED MANUSCRIPT
578
______________________________________________________________________
579
HQ-1: 1.0%, HQ-2: 3.3%, HQ-3: 2.6%, and HQ-4: 0.3% LOD: 0.06 mg/L
580 581
Table 3 Estimated BMD10 and BMDL10 levels for adenoma in rats exposed to
583
hydroquinone for 103 weeks (NTP, 1989)
Gamma f) (Gamma Logistic
78.2918 78.2918 80.476
ｇ)
LogLogistic f) LogLogistic LogProbit f)
78.2918 78.2918
ｇ)
77.0747 78.2918
ｇ)
LogProbit Multistage (2-degree) f) Multistage (2-degree) Multistage (1-degree) f)
Weibull ｇ) Quantal-Linear 585 586 587 588 589 590 591
80.1013 78.2918 78.2918 76.2961
Scaled residual e)
34.2434 34.2434 42.4421
21.876 8.04E-12 34.7432
0 0) -0.26
1 1 0.659 1
34.1399 34.1399 35.7302 33.8199
20.9038
0 0 0.747 0
1 1 0.9979 0.9979
34.3586 34.3586 33.9716 33.9716
21.876 13.7266 21.8689 21.8689
0 0 -0.053 -0.053
0.2444 1
41.0694 34.2551
33.064 21.876
-0.298 0
1 0.9979
34.2551 33.9716
21.8689
0 -0.053
27.5444
AIC (Akaike’s Information Criterion) b) P-value from the chi-square test for global goodness of fit c)
AC C
584
a)
ｇ)
EP
Multistage (1-degree) Probit Weibull f)
78.2918 78.2918 76.2961 76.2961
BMDL10 d) (mg/kg bw)
1 1 0.2076
TE D
ｇ)
P-value
BMD10 c) (mg/kg bw)
SC
AIC
b)
M AN U
Model Name
a)
RI PT
582
BMD10 (Benchmark Dose): maximum likelihood estimate of the exposure level associated with a 10% extra risk d) BMDL10: 95% lower confidence limit on the BMD10 e) Scaled residuals for local measurement f) Restricted models ｇ)
Unrestricted models
592
34
ACCEPTED MANUSCRIPT
1.400
RI PT
1.200 1.000 0.800 0.600
0.200 0.000 0
30
120
Fig. 1
596
TE D
597
(a) 120
150
180
EP
100 80
40 20 0
AC C
60
0
60
120 180 240 300 360 420 480
Time (min)
(b) 0.6
Cumlative amount of HQ (%/cm2)
595
Cumulative amount of HQ (µg/cm2)
90
(min)
594
598
60
SC
0.400
M AN U
HQ concentration in rat plasma (mg/L)
593
0.5 0.4 0.3 0.2 0.1 0 0
60
120 180 240 300 360 420 480
Time (min)
Fig. 2
35
ACCEPTED MANUSCRIPT
HQ-4
RI PT
HQ-3
0.1
0.01
SC
HQ-2
1
0.001
0.0001
0.00001 0
2
M AN U
HQ-1
Comulative amount of HQ permeated (%/cm2)
10
4
Time (h)
EP
Fig. 3
AC C
600
TE D
599
36
6
8
ACCEPTED MANUSCRIPT
HQ-4
RI PT
HQ-3
SC
HQ-2
10
1
M AN U
HQ-1
Comulative amount released (%/cm2)
100
0.1 0
60
120
180
240
300
Time (min)
601
Fig. 4
TE D
602
EP
0.02
0.015
AC C
Blood concentration of HQ (mg/L)
0.025
0.01
0.005
0
0
5
10
HQ-2 leave on for 8 h and wash off (night-time)
603 604
15
20
Time (h)
Fig. 5
37
25
30
360
ACCEPTED MANUSCRIPT
0.03
Blood concentration of HQ (mg/L)
0.025
RI PT
0.02 0.015 0.01
0 4
8
HQ-2 leave on for 8 h and wash off (night-time)
16
20
24
HQ-2 leave on for 16 h and wash off (day-time)
Time (h)
605 606
12
M AN U
0
SC
0.005
Fig. 6
TE D
60 50
EP
40 30 20
AC C
Comulative pereation of HQ (µg/cm2)
70
10
0
0
4
8
12
16
Time (h)
607 608
Fig 7
609
38
20
24
ACCEPTED MANUSCRIPT
610
612
Fig. 1 Hydroquinone (HQ) concentration in hairless rat plasma (n=3) after i.v. administration of 2% HQ aqueous solution
RI PT
611
613
Fig. 2 Cumulative amount as (a) a mass or (b) a percentage of hydroquinone (HQ)
615
permeated from 1 mL of 2% hydroquinone aqueous solution through the intact hairless
616
rat skin (mean and SE, n = 3)
M AN U
SC
614
617
Fig. 3 Cumulative amount (%) of hydroquinone (HQ) permeated from the four
619
cosmetics (HQ-1, HQ-2, HQ-3, and HQ-4) through the in vitro intact hairless rat skin
620
(mean and SE, n = 3)
EP
621
TE D
618
Fig. 4 Comulative amount (%) of hydroquinone (HQ) released from the four cosmetics
623
(HQ-1, HQ-2, HQ-3, and HQ-4) through the cellulose dialysis membrane (mean
624 625
AC C
622
and SE, n = 3)
626
Fig. 5 Estimated blood concentrations of hydroquinone (HQ) after a single dermal
627
application of 1.5 g HQ-2 on a whole face, a décolleté and hands of 1440 cm2. Time
628
course of plasma HQ concentrations was predicted from the skin permeation rate (Jskin) 39
ACCEPTED MANUSCRIPT
629
obtained from in vitro intact rat skin data and elimination rate constant (Kel) by a
630
convolution method.
RI PT
631
Fig. 6 Estimated blood concentrations of hydroquinone (HQ) after a 2-time dermal
633
application of each 1.5 g HQ-2 on a whole face, a décolleté and hands of 1440 cm2.
634
Time course of plasma HQ concentrations was predicted from the skin permeation rate
635
(Jskin) obtained form in vitro intact rat skin data and elimination rate constant (Kel) by a
636
convolution method.
M AN U
SC
632
637
Fig 7. Estimated cumulative permeation levels of hydroquinone (HQ) after a 2-time
639
dermal application of HQ-2 (1.5 g/time) on a whole face, a décolleté and hands of 1440
640
cm2
EP AC C
641
TE D
638
40
ACCEPTED MANUSCRIPT
Highlights
Rat skin permeation rates of hydroquinone (HQ) and HQ cosmetics were determined. One cosmetic showed a higher permeation rate than that of HQ aqueous solution. Plasma HQ levels after application of this cosmetic in humans were predicted. The margin of exposure (MOE) was estimated with a human dose level and a
benchmark dose. The MOE is judged to be in a range safe for the consumer if HQ cosmetics are used properly.
AC C
EP
TE D
M AN U
SC
RI PT