Skin Permeability of Various Drugs with Different Lipophilicity

Skin Permeability of Various Drugs with Different Lipophilicity CHEON Koo LEE,TAKAHIRO UCHIDA,KAZUHISA KITAGAWA, AKIRAYAGI, NAK-SEO KIM, AND SHIGERU GO TO^ Received April 9, 1993, from the faculty of Pharmaceutical Sciences, Kyushu University, Maidashi 3- 1- 1, Higashi-ku, Accepted for publication September 9, 1993@.

fukuoka 812, Japan.

Abstract 0 The in vitro and in vivo skin permeability of 16 drugs with a wide span of lipophilicity (log P ranging from -0.95 to 4.40) was evaluated with an ethanol1panasate 800 (tricaprylin)(40160) system as a lipophilic vehicle. The ethanollpanasate800 (40160) binary vehicle remarkably improved the in vitro skin permeability of the drugs across excised hairless mouse skin compared with ethanol or panasate 800 as single vehicles. The in vivo skin permeability of the large majority of the drugs across abdominal rat skin also showed high permeation

rates and short lag times. The relationship between lipophilicity and skin permeability of the drugs from the ethanol/panasate800 (40160) binary vehicle indicated parabolic shapes with their peaks at much greater hydrophilic range (log R -0.88 for in vitro, -0.83 for in vivo) compared with other past references (log/? 2-3). The results suggest that the lipophilicity of a drug is a main factor for prediction of the skin permeability of the drug and that the ethanollpanasate 800 (40160)

lipophilic binary vehicle would be a good candidate as a vehicle for future clinical application of hydrophilic drugs.

The advantages of transdermal administration are avoiding hepatic first-pass effect, minimizing inter- and intrapatient variation, maintaining steady-state plasma level to provide longterm therapy from a singledose, and allowing a shift termination of drug input can be realized. For the potential use of a drug as a transdermal delivery system, two limiting factors (along lag time and a low steady-state flux) must be overcome. One method for overcoming these obstacles is to select an optimal vehicle system. Most researchers have paid attention to hydrophilic vehicles,lJ such as propylene glycol, ethanol, water, or their mixtures, but the number of reports on lipophilic vehicle3 is limited. We developed a lipophilic vehicle for skin permeation enhancement of a drug. The in uitro skin permeation results of ketoprofen*showed that ethanol plays a role in a long lag time and a high permeation rate, and that panasate 800 (caprylic triglyceride) plays a role in reducing lag time. Panasate 800 is a neutral oil that is not miscible with water and is stable toward oxidation compared with natural vegetable oils. It has been also suggestedthat ethanollpanasate 800 (40/60) lipophilicbinary vehicle improved both the permeation rate and the lag time, and the mutual effect was concluded to be due to the increase of diffusivityof a drug by panasate 800 and the increase of partition of a drug into skin tissue by ethan01.~ Generally,the skin can be considered as a trilaminate structure, consisting of the outer hydrophobic stratum corneum, the underlying viable hydrophilic epidermis, and dermis. The amphiphilic nature of the skin dictates that its permeability will be highly dependent on the lipophilicity of a penetrant. In the present study, we evaluated the effect of the ethanol/ panasate 800 binary lipophilic vehicle on skin permeation of various drugs having a wide range of lipophilicity as indicated by the n-octanollwater partition coefficient P (log P, -0.954.40). We selected 16 drugs, including anti-cancer drugs, xanthines, local anesthetics, and anti-inflammatory drugs, as model drugs with different log P values. The relationships between lipophilicities of the drugs and in uitro and in uiuo skin @

Abstract published in Advance ACS Abstracts, January 1, 1994.

562 /Journal of Pharmaceutical Sciences Vol. 83, No. 4, April 1994

permeabilities were studied. A correlation between the in uiuo and the in uitro permeabilities of the drugs was also investigated.

Experimental Section Chemicals-Tegafur was obtained from Taiho Pharmaceutical Company (Tokushima, Japan). Alclofenac, ketoprofen and ibuprofen were gifts from Hiaamitsu Company (Saga, Japan). Propentofylline and pentoxifylline were kindly supplied from Hoechst-Japan Company (Tokyo,Japan). The followingdrugs were purchased from the companies described in parentheses: 5-fluorouracil and caffeine (Wako Pure Chemicals, Japan); lidocaineHC1 and salicyluric acid (Sigma Company, St. Louis, MO); and theophylline, antipyrine, procaineHC1, tetracaineHC1, dibucaine-HC1,and salicylic acid (Nacalai Tesque Company, Kyoto, Japan). Panasate 800as tricaprylin was kindly supplied by Nihon Yushi Company (Tokyo, Japan). All other chemicals were of reagent grade. Zn VitroS k i n Permeation Studies-The in uitro experiments were performed accordingto the method given in our previous r e p ~ r t Briefly, .~ the excised skin of hairless female mouse was obtained from 8-9-weeksold, 27-33-g animals. Vertically assembled diffusion cells with an effective diffusional area of -0.785 cm2 and downstream volume of 5 mL were used. The above skin was mounted on the diffusion cell, and the receiver compartment was filled with 5 mL of 50 mM phosphate buffer saline (PBS) a t pH 7.4 maintained a t 37 "C. The donor compartment was charged with 0.5 mL of 0.5% (wiw) drug preparation at 37 "C and capped. Five hundred-microliter aliquots were withdrawn from the receiver compartment periodically for 26 h and replaced with equal volume of fresh PBS maintained at 37 "C. The sample solution was filtered through a membrane filter (pore size, 0.2 pm; Tosoh W-13-2; Tosoh Company, Japan) and injected onto the HPLC. The HPLC consisted of a solvent delivery pump (LC-SA, Shimadzu Company, Kyoto, Japan), a column (Shimpack CLS-ODS, 0.6 X 15 cm, Shimadzu Company), a UV detector (SPD-6A, Shimadzu Company), and an integrator (C-RGA,Shimadzu Company). In all cases, the flow rate and column temperature were kept a t 1.5 mL/min and 40 "C, respectively. Mobile phases and UV conditions for each drug are listed in Table 1. The correction of concentration against each sample point was also undertaken. ZII Vivo Skin Permeation Studies-The abdominal hair of a rat (Wistar strain, 7 weeks old, 180-220 g, Nihon SLC, Shizuoka, Japan) was removed with an electric clipper on the day before the experiments. Pentobarbital sodium was administered intraperitoneally a t a dose of 50 mg/kg to induce continuous anesthesia during the experiment, and the animal was restrained in the supine position on a 37 "C surface. We used a specially made glass cell with an effective diffusional area of 3.14 cm2. The glass cell was applied to the hair-free region of the rat with a cyanoacrylate adhesive (Aron Alpha, Toa Gosei Chemical Company, Ltd., Tokyo, Japan), and then the cell was charged with 0.5 mL of 0.5% (wiw) drug preparation a t 37 "C and capped. Two hundred fifty microliters of blood were collected periodically from the jugular vein. The blood sample was centrifuged a t 12000 rpm for 5 min, and a 0.1-mL plasma sample was used for analysis. The experiment was performed for 26 h. The concentrations of drugs in plasma samples were determined by HPLC. Mobile phases, UV conditions, internal standards, and extraction solvents for each drug are summarized in Table 1. To 0.1 mL of plasma, 0.1 mL of internal standard solution (20 pg/mL) and 8.0 mL of dichloromethane or ethylacetate as extraction solvents were added. The mixture was extracted by shaking for 20 min and then centrifuged at 2500 rpm for 20 min. Seven milliliters of organic layer was collected and evaporated under nitrogen gas or a t 60 "C, a water bath to remove the solvent. The residue was dissolved in 0.2 mL of the mobile phase, and 20 p L of the solution was injected onto the HPLC. When the last collection of blood sample was finished (at 26 h), the glass cell was removed from the animal a t the site it was attached to the

0022-3549/94/ 1200-562$04.50/0

0 1994, American

Chemical Society and American Pharmaceutical Association

Methods for Determination of Concentration of Drugs in in Vltro and in Vivo Experiments

Table 1-HPLC

~~

I n Vitro

Drug

I n Vivo

Mobile Phases

Mobile Phasea

Anticancer drugs 5-Fluorouracil (FU) Tegafur (TG)

Internal Standard

A:C (1:99) A:C (17:83)

A:D (0:lOO) A:D (9.5:90.5)

5-Bromouracil 5-Bromouracil

A:C (18:82) A:C (12:88) A:C (30:70) A:C (45:55)

A:D (10:90) A:D (10:90) A:B:D (20:12:68) A:B:D (21:20:59)

TP CF

A:B:D (10:387) A:B:D (19:3:78) A:B:D (31:8:61) A:B:D (38:8:54)

A:B:D A:B:D A:B:D A:B:D

A:D (28:72)

A:D (12:88) A:D (30:70) A:D (30:70) A:B:D (30:30:40) A:B:D (30:30:40) A:B:D (30:30:40)

Extraction Solvent

UV Detection, nm

E

280 280

E

Xanthines Caffeine (CF) Theophylline (TP) Pentoxifylline (PT) Propentofylline(PP) Local anesthetics Procaine-HCI (PC) Lidocaine-HCI (LC) Tetracaine-HCI (TC) Dibucaine-HCI (DC) Anti-inflammatory drugs Antipyrine (AP) Salicyluric acid (SU) Salicylic acid (SD) Alclofenac (AF) Ketoprofen (KP) Ibuprofen (IP) a Expressed

A:C (15:85) A:C (17533) A:C (27:75) A:C (30:70) A:C (34:66)

273 273 273 273

TC TC

(62591.5) (7.5:3:89.5) (23:20:57) (25:20:55)

CF CF PP

Dexamethasone

F F F F

Cefadroxil SD Anisic acid IP IP AF

290 210 290 240 254 300 300 220 258 220

as volume fraction. A, acetonitrile;B, methanol; C, pH 7.0, 10 mM phosphate buffer: D, 0.5 mM phosphoric acid: E, ethyl acetate;

F, dichloromethane. Table 2-Lipophliiclty and in VNro Skln Permeability of Various Drugs from EthanoVPanasate 800 (40/60) Binary Vehiclea

Permeation Permeation

Lag Time,

LogP

%fh

Percent, % of dose at 12 h

-0.95b -0.48'

7.2 (0.6) 8.0 (0.7)

54.3 (5.2) 56.8 (5.4)

2.2 (0.5) 2.2 (0.7)

Caffeine (CF) Theophylline (TP) Pentoxifylline (PT) Propentofylline (PP) Local anesthetics

-0.07b -0.02b 0.72d 1.72e

5.2(0.4) 5.9 (0.5) 4.1 (0.4) 2.1 (0.3)

47.2(3.4) 47.7 (4.5) 32.7 (4.0) 18.8 (5.5)

3.1 (0.3) 2.9 (0.4) 3.8 (0.5) 3.5 (0.5)

Procaine-HCI (PC) Lidocaine-HCI (LC) Tetracaine-HCI (TC)

1.87b 2.26b 3.73b 4.40b

5.0 (0.4) 2.3 (0.3) 1.0 (0.2) 0.5 (0.1)

41.5 (1.8) 19.4 (4.6) 5.1 (1.3) 4.7 (1.3)

4.4 (0.7) 3.8 (0.5) 4.0 (0.7) 4.2 (0.4)

0.23' 1.138 2.258 2.489 2.948 3.5V

7.0(0.6) 6.4 (0.7) 3.3 (0.4) 2.8(0.2) l.g(O.2) 1.2(0.2)

47.7(4.5) 57.5 (8.3) 33.2 (4.6) 25.4(3.3) 16.1 (2.3) 7.7(0.6)

3.5(0.6) 2.8 (0.3) 1.5 (0.3) 3.0(0.5) 3.7(0.4) 2.2(0.4)

Drug Anticancer drugs 5-Fluorouracil (FU) Tegafur (TG) Xanthines

Dibucaine-HCI (DC) Anti-inflammatorydrugs Antipyrine (AP) Salicyluric acid (SU) Salicylic acid (SD) Alclofenac (AF) Ketoprofen (KP) Ibuprofen (IP)

Rate,

h

a Each value represents the mean (SD) of four experiments. Reference 6 (calculated). Reference 7 (apparent). Log P = log P(caffeine) r(-C=O) i-4r(CH3) = (-0.07) (-1.21) (2.00) = 0.72. "Log P = log P (pentoxifylline) 2 r (CH3) = 0.72 + 1.00 = 1.72. Reference 8 (calculated). g Reference 9 (calculated).

+

+

+

+

skin and then rinsed five times with 10 mL of ethanol to collect unabsorbedresidual drug from the cell and the skin surface. The residual amounts of the drugs were measured by the HPLC method described in the in uitro skin permeation studies. Determination of o-Octanol/ Water Partition Coefficient-The logarithms of n-octanol-water partition coefficient (logP)were obtained either from references69 or with the equation shown in Table 2 (the calculation was made with Hansch ?r values8). Calculation of Permeation Parameters-For each diffusion cell, the cumulative permeation percentage (%) of drugs versus time (h) was

plotted. The in uitro permeation rate (%/h) and the lag time (h) could be calculated from the slopes and the intercepts, respectively, on the time axis of the linear portion of the plots. In in uiuo studies, elimination rate constants of the drugs were determined by intravenous administrations of the drugs. The cumulative absorption percentage (% ) of the drugs was obtained by the deconvolution method with the plasma concentrations and elimination rate constants of the drugs. I n uiuo permeation rate ( % / h ) and lag time (h) were calculated from the cumulative absorption percentage ( % ) of the drugs. Four experiments per group were performed. All data were expressed as the mean (SD).

Results and Discussion In Vitro Skin Permeability of Drugs-The in uitro permeabilities of 16 drugs through excised hairless mouse skin were examined with ethanol, panasate 800, and the ethanol/ panasate 800 (40/60) binary system as vehicles. The in uitro permeation profiles of 5-fluorouracil, procaineHC1, propentofylline, and ibuprofen are shown in Figure 1(A-D, respectively) as typical examples. The permeation profiles of 16 drugs with a wide span of log P values [two anti-cancer drugs (A), four xanthines (B), four local anesthetics (C), and six anti-inflammatory drugs (D)] from the ethanol/panasate 800 (40/60) binary vehicle are shown in Figure 2. The corresponding in uitro permeation parameters and log P of the drugs are summarized in Table 2. The in uitro skin permeabilities of all drugs were remarkably enhanced by the ethanol/panasate (40160) binary vehicle compared with ethanol or panasate 800 as single vehicles. These results are in agreement with our previous results that skin permeation of ketoprofen4 or tegafur5 was enhanced by the combination vehicle of ethanol and panasate 800 compared with each single vehicle. In the present study, it was suggested that the ethanollpanasate 800 (40/60) binary vehicle is also effective at improving the permeability of a wide span of drugs with different lipophilicity. In Vivo Skin Permeability of Drugs-Figure 3 represents the in uiuo permeation percentage versus time curves after percutaneous administration of ethanol/panasate 800 (40/60) binary vehicle containing 0.5% (w/w) of each drug. The corresponding in uiuo permeation parameters are listed in Table 3. Journal of Pharmaceutical Sciences / 563 Vol. 83, No. 4, April 1994

100

r

0 100

loo

4

8

r

0

12 16 20 24 28

(C)

r

100

r

80

-

4

8

0

12 16 20 24 28

4

(C)

100

(D)

-8

0

8 12 16 20 24 28

I

4

8 12 16 20 24 28

100

(D)

I

T

80

0

60

-

40

-

e

2

60

h 0 0

4

8

12 16 20 24 28

0

Tirneh)

4

8

12 16 20 24 28

4

The lag times could be reduced by almost half to one-third for all drugs. The in uiuo permeation rates quickly attained steady state and exhibited high values in the case of hydrophilic drugs. However, a low i n uiuo permeation rate was observed with drugs with high degrees of lipophilicity (dibucaine.HC1: 1.4 % /h, ibuprofen: 2.4 % /h). Short lag times below 2.4 h, which may be useful for clinical use, were observed in these experiments. The residual amount of each drug in the donor cell after 26 h in in uiuo absorption experiments with the ethanoVpanasate 800 (40/60) binary vehicle decreased with an increase of hydrophilicity of the drugs in each group (Table 3). Relationship between Lipophilicity and Skin Permeability of Drugs-Figure 4 represents the relationship between lipophilicity of drugs and in uitro permeation rate (A) and i n uiuo permeation rate (B) of all drugs tested in this study with ethanol/panasate (40/60) lipophilic binary vehicle. The inuitro permeation rate of the drugs across excisedhairless mouse skin increased with an increase in the hydrophilicity of the drugs; this was also observed with the i n uiuo permeation studies across abdominal rat living skin. The relationship between log Pand the permeation rate of the drugs yielded good parabolic shapes in both the i n uitro and the i n uiuo models, which are described in the following regression equations: log( %/h);,, "itm = 0.81 - 0.07 (log P) - 0.04 (log P)2 (r = 0.916, n = 16) (1) log(%/h)i,, "iu0 = 0.91 - 0.05 (log P) - 0.03 (log P)2 (r = 0.816, n = 16) (2) Maximum values of permeation rates were obtained with a log P of -0.88 for the in uitro and a log P of -0.83 for the i n uiuo studies with eqs 1 and 2, respectively. Therefore, it was considered that the ethanolipanasate 800 (40/60) binary lipo-

8

12 16 20 24 28

0

4

Tirne(h)

Tirne(h)

Flgure 1-In vitroskin permeation profiles of 5-fluorouracil(A), procaineHCl (B), propentofylline (C),and ibuprofen (D) across excised hairless mouse skin from ethanol,panasate 800,and ethanollpanasate 800 (40l60) binary ethanol; (A)panasate 800;(0)ethanollpanasate800 vehicle. Key: (0) (40160).

564 / Journal of Pharmaceutical Sciences Vol. 83, No. 4, April 1994

0

8

12 16 20 24 28 Tirne(h)

Figure 2-Zn vitro skin permeation percentage profiles of various drugs across excised hairless mouse skin from ethanollpanasate 800 (40l60) binary vehicle. Key: (A) anticancer drugs: (0)FU; ( 0 )T G (B) xanthines: (0)CF; (0)TP; (A)PT; (A)PP; (C) local anesthetics: (0)PC; (0)LC; (A) TC; (A)DC; (D) anti-inflammatory drugs: (0) AP; (W) SU; (0)SD; (0)AF; (A) KP; (A)IP.

philic vehicle would be successful at enhancing the skin permeation of relatively hydrophilic drugs. Generally, the skin is considered as a heterogeneous structure, composed of a comparatively lipophilic stratum corneum and hydrophilic viable skin (epidermis and dermis). Therefore, for hydrophilic penetrants, partitioning into the stratum corneum becomes the rate-determining step of skin permeation. By contrast, for penetrants with a high lipophilicity, partitioning out of the stratum corneum into the viable epidermis becomes important. The heterogeneous skin, composed of alternative aqueous and lipidic phases, leads one to assume that a parabolic relationship may be obtained between lipophilicity and skin permeability of drugs if drugs with a wide range of lipophilic character are examined. Such a parabolic relationship was first recognized by Hansch and Claytonlo as the correlation between the pharmacological activity of drugs and their representative lipophilic parameters. It has already been reported that skin permeabilities of homologous salicylate and nonsteroidal antiinflammatory drug series show a parabolic relationship to log P value, with peaks at log P of 2.24 and 2.43.9 Many authorsg.11-15 also have reported that the relationship between skin permeability and lipophilicity showed a characteristic parabolic shape with the maximum occurring at a log P of 2-3. Their parabolic shapes, withmaximum in amphiphilic drugs of log P2-3, indicate that large skin permeations for clinical use cannot be expected for relatively hydrophilic drugs, and the stratum corneum is a permeation barrier for hydrophilic drugs. In the present study, however, we could obtain parabolic relationships, with maximum peaks of log P at -0.88 or -0.83, which are inclinedtoward a much greater hydrophilic range when compared with the previous reports (log P, 2-3). It was concluded that the lipophilicity of a drug is the main factor for prediction of the skin permeability of the drug from the ethanol/panasate

-

loo

r

T T

-

1.0

100

0.8 0.6

2 '5

-

0.4

f

-8 f

0.2

8

- 0.0 -0.2 -0.4

0

4

8

0

4

8 12 16 20 24 28

12 16 20 24 28

0

4

8

4

8

12 16 20 24 28

-1

0

1

-1

0

1

~~

0

Tirne(h)

Time(h)

Table 3-111 Vlvo Skin Permeablllty of Various Drugs from EthanoVPanasate 800 (40180) Binary Vehicle* Permeation Percent, % of Permeation dose Lag Rate, %/h at 12 h Time, h

Drug ~~

Residual Percent, % of dose at 26 h

3

4

5

2

3

4

5

log P

12 16 20 24 28

Flgure 3-In vivo skin absorption percentage profiles of various drugs acrossabdominalrat skin from ethanoVpanasate800 (40/60) binaryvehicle. Key: (A) anticancer drugs: (0)FU; (0)TG (B) xanthines: (0)CF; (0) TP; (A)PT; (A)PP; (C) local anesthetics: (0)PC; (0)LC; (A)TC; (A)DC; (D) anti-inflammatory drugs: (0) AP (M) S U (0)S D (0)AF; (A)KP; (A) IP.

2

Flgure 4-Relationship between lipophilicity(log P) and in vifro permeation rate (A) or in vivo permeation rate (B) of various drugs from ethanol/ panasate 800 (40160) binary vehicle. Key: log (% /h) vtbo = 0.81 - 0.07 (log P) - 0.04 (log P)* (r = 0.916, n = 16). log (%/h),,, = 0.91 - 0.05 (log P) - 0.03 (log P)* (r = 0.816, n = 16).

,,

panasate 800 (40/60) binary lipophilic vehicle; steady-state flux was increased and lag time was decreased. The relationship between lipophilicity and skin permeability of t h e drugs from the ethanol/panasate 800 (40/60)binary vehicle showed a parabolic shape w i t h ape& a t amore hydrophilic range compared w i t h other past references. We suggest t h a t the ethanol/panasate 800 (40160)binary vehicle could be considered as a potential candidate for future clinical transdermal application o f a hydrophilic drug.

~~

Anticancer drugs 5-Fluorouracil (FU) Tegafur (TG) Xanthines Caffeine (CF) Theophylline (TP) Pentoxifylline (PT) Propentofylline(PP) Local anesthetics Procaine-HCI(PC) Lidocaine-HCI(LC) Tetracaine-HCI (TC) Dibucaine-HCI(DC) Anti-inflammatory drugs Antipyrine (AP) Salicylic acid (SU) Salicylic acid (SD) Alclofenac (AF) Ketoprofen (KP) Ibuprofen (IP) a

9.3 (1.1) 9.3 (1.1)

69.4 (7.9) 68.2 (7.4)

2.4 (0.4) 0.0 (0.0) 1.7 (0.3) 11.4 (2.1)

7.6 (0.7) 8.4 (1.0) 5.6 (0.6) 3.4 (0.5)

60.5 (7.5) 55.9 (7.5) 46.1 (3.7) 32.0 (5.4)

0.8 (0.1) 3.3 (0.3) 1.6 (0.4) 11.1 (1.5) 1.7 (0.3) 22.4 (5.3) 1.9 (0.2) 38.0 (7.6)

6.4 (0.3) 3.2 (0.4) 2.0 (0.2) 1.4 (0.2)

55.2 (4.5) 0.4 (0.1) 1.4 (0.3) 33.1 (5.1) 0.5 (0.1) 7.8 (2.5) 21.6 (1.6) 0.4 (0.1) 12.6 (3.4) 15.3 (1.1) 0.9 (0.3) 24.9 (2.5)

7.0 (0.9) 9.2 (1.3) 7.9 (0.7) 5.6 (0.5) 2.9 (0.3) 2.4 (0.3)

61.0 71.4 68.9 59.9 28.9 23.3

(7.2) (8.4) (5.5) (7.6) (3.4) (3.2)

1.9 (0.5) 2.5 (1.1) 1.9 (0.2) 5.0(2.3) 0.8 (0.1) 4.2 (1.0) 1.1 (0.5) 0.8 (0.4) 2.3 (0.4) 26.9 (5.6) 2.1 (0.3) 33.5 (6.2)

Each value represents the mean (SD) of four experiments.

800 (40160)lipophilic binary vehicle, and this vehicle more effectively enhances t h e skin permeability o f a hydrophilic drug. In conclusion, the skin permeation of various drugs covering a wide span o f lipophilicity was enhanced by the use of an ethanoV

References and Notes 1. Ostrenga, J.; Steinmitz, C.; Poulsen, B. J . Pharm. Sci. 1971, 60, 1175-1179. 2. K-Bergstrom,T.;Knutson,K.; NeNoble, L. J.;Goates,C. Y. Pharm. Res. 1990, 7, 762-766. 3. Takahashi, K.; Tamagawa,S.;Katagi, T.; Yoshitomi, H.; Kamada,

H.; Rytting, J. H.; Nishihata, T.; Mizuno, N. Chem. Pharm. Bull.

1991,39,154-158. 4. Goto,S.;Uchida,T.;Lee,C.K.;Yasutake,T.;Zhang,J.B. J.Pharm. S C ~1993, . 82, 959-963. 5. Lee, C. K.; Uchida, T.;Noguchi, E.; Kim, N.-S.;Goto, S. J. Pharm. S C ~1993.82.1155-1159. . - , -, 6. Leo, A.; Hansch, C.FElkins, D. Chem. Rev. 1971, 71, 525-616. 7. Yamashita, S.; Suda, Y.; Masada, M.; Nadai, T.; Sumi, M. Chem. Pharm. Bull. 1989,37,2861-2863. 8. Hansch, C.; Anderson, S. J. Org. Chem. 1967, 32, 2583. 9. Yano, T.; Nakagawa, A.; Tsuji, M.; Noda, K. Life Sci. 1986, 39, 1043-1050. 10. Hansch. C.: Clavton. J. M. J . Pharm. Sci. 1973.62.1-21. 11. Dies, I.; Coiom, H.; Obach, R.;Peraire, C.; Domenech, J. J. Pharm. Sci. 1991,80, 931-934. 12. Pozzo, A. D.; Liggeri, E.; Delucca, C.; Calabrese,G. Znt. J. Pharm. 1991, 70. 219-223. 13. Bonina,.F. P.; Montenegro, L.; De Capraris, P.; Bousquet, E.; Tirendi, S. Znt. J . Pharm. 1991, 77, 21-29. 14. Houston, J. B.; Upsall, D. G.; Bridges, J. W. J . Pharm. Erp. Ther. 1974,189,244-254. 15. Hadgraft,J.; Guy, R. H. Drugs and The Pharmaceutical Science: volume 35, Transdermal Drug Delivery; Marcel Dekker: NewYork, 1989; pp 62-67. ~~

Journal of Pharmaceutical Sciences / 585 Vol. 83, No. 4, April 1994