- Email: [email protected]
Importance of Sebaceous Glands in Cutaneous Penetration of an Antiandrogen: Target Effect of Liposomes
Importance of Sebaceous Glands in Cutaneous Penetration of an Antiandrogen: Target Effect of Liposomes EMMANUELLE BERNARD*, JEAN-LUC DUBOIS‡,
AND JACQUES
WEPIERRE*X
Received September 18, 1996, from the *Laboratoire de dermopharmacologie, Faculte´ de Pharmacie, 5 rue J. B. Cle´ment, 92290 Chaˆtenay-Malabry, France, and the ‡Laboratoire de recherche Gale´nique, ROUSSEL UCLAF, 102 route de Noisy, 93235 Romainville, France. Final revised manuscript received December 12, 1996. Accepted for publication January 17, 1997X. Abstract 0 The significance of the sebaceous gland pathway in the cutaneous permeation of an antiandrogen, 4-[3-(4-hydroxybutyl)-4,4dimethyl-2,5-dioxo-1-imidazolidinyl]-2-(trifluoromethyl)benzonitrile (RU 58841), was studied with normal hairless rat skin and an induced scar hairless rat skin without sebaceous glands. RU 58841 was dissolved in an alcoholic solution and encapsulated in liposomes for comparison. After 24 h, the cumulative percentage of RU 58841 absorbed in vitro was 3−4-fold higher in the normal skin than in the scar skin; in the case of liposomes, the accumulation of the drug in the normal dermis was significantly higher than in the scar one. In the in vivo cutaneous distribution, the epidermis and dermis of the normal skin contained higher amounts of RU 58841 than the scar skin (ninefold with the solution and 16-fold with liposomes). An autoradiography study showed that with the solution, the drug was mainly localized in the stratum corneum/epidermis, and with the liposomes, the drug was mainly localized in the sebaceous glands. We concluded that the sebaceous glands constituted the main pathway for RU 58841. The alcoholic solution encouraged the localization of the drug into the stratum corneum, whereas liposomes targeted the sebaceous glands.
Introduction It is now accepted that hair follicles and sebaceous glands may participate in skin penetration for a wide range of compounds.1 Several investigations were carried out to evaluate the importance of the transfollicular route and to develop models permitting the assessment of the pathways in permeation.2 Another reason for these recent reports is that sebaceous glands and hair follicles are the site of dermatological disorders such as acne and androgenetic alopecia. RU 58841 (4-[3-(4-hydroxybutyl)-4,4-dimethyl-2,5dioxo-1-imidazolidinyl]-2-(trifluoromethyl)benzonitrile) is a topical antiandrogen that is being developed for the treatment of androgen-dependent diseases. RU 58841 is a benzonitrile derivative, slightly soluble in water (0.44 mg/mL), freely soluble in ethanol (250 mg/mL), and has a log octanol/water partition coefficient of 1.76. The drug has a molecular weight of 369.4, and appears as a white powder with a melting point around 101 °C. RU 58841 has selective binding to the androgen receptors,3 which are localized especially in the sebaceous glands and hair follicles of human skin.4 Thus, the interest in the pilosebaceous unit as a target for RU 58841 is widely justified. The aim of the present study was to determine the routes of penetration of this molecule, and particularly to assess the importance of the sebaceous glands. At the present time, no perfect animal model exists to study transepidermal and transfollicular routes distinctly. We used a technique adapted by Illel and Schaefer5 consisting of inducing a truly follicle-free skin of hairless rats. The method allows the cutaneous permeation between the follicle and X
Abstract published in Advance ACS Abstracts, March 15, 1997.
© 1997, American Chemical Society and American Pharmaceutical Association
sebaceous gland-free skin and the normal hairless rat skin to be compared. We quantified the contribution of the sebaceous glands in the percutaneous absorption in vitro and in the distribution in the skin strata in vivo. Finally, we completed these investigations by assessing the localization of RU 58841 within the skin by a qualitative autoradiography. The effect of vehicle on the penetration pathways was also studied, with a comparison between the permeation of RU 58841 dissolved in an alcoholic solution, suitable to scalp in the treatment of androgenetic alopecia, or encapsulated in liposomes, suitable to skin in the treatment of acne. A previous investigation6 indicated that liposomes, compared with the solution, reduce percutaneous absorption and increase retention of the drug in the dermis.
Experimental Section ChemicalssRU 58841 (4-[3-(4-hydroxybutyl)-4,4-dimethyl-2,5dioxo-1-imidazolidinyl]-2-(trifluoromethyl)-benzonitrile) is a novel topical antiandrogen that is being developed by Roussel Uclaf. RU 58841 and [14C]RU 58841, synthesized in the Roussel Uclaf laboratories, were mixed to obtain specific radioactivities of 355 KBq/mg and 1.2 MBq/mg. Radiochemical purity was >98%. Lipoid E 100s35 (Lipoid KG, Ludwigshafen, Germany); L-3phosphatidyl[N-methyl-3H]choline 1,2-dipalmitoyl (Amersham), specific activity 3.11 GBq/mg; R-tocopherol (Fluka AG, Buchs, Switzerland); and a phosphate buffer were used for the liposomes preparation. Ethanol and propylene glycol were used for the solution.
Liposome FormulationsSmall unilamellar liposomes (SUV) containing 0.5% weight of [14C]RU 58841 were prepared with Lipoid E 100s35 (egg phosphatidylcholine > 94%) and R-tocopherol in a 0.05 M phosphate buffer pH 7. Briefly, powders of [14C]RU 58841, R-tocopherol, and Lipoid E 100s35 were mixed at 60 °C. Then, the appropriate volume of phosphate buffer was added at 60 °C with gentle stirring. Finally, the mixture was homogenized with a microfluidizer M110S (Sodexim, Muizon, France). Liposome size was measured with a N4 MD Coulter counter (Coultronics, Margency, France), and the mean volume diameter was 84 ( 12 nm. In the in vitro percutaneous absorption studies, the final concentration for [14C]RU 58841 was 4.3 mg/g, with an activity of 1.6 MBq/g of preparation, and the final concentration for [3H]phospholipids was 98 mg/g, with an activity of 2.45 MBq/g of preparation. In the in vivo quantitative cutaneous distribution and qualitative autoradiography, the final concentration for [14C]RU 58841 was 4.3 mg/g, with an activity of 5.2 MBq/g of preparation. SolutionsAlcoholic solution was also studied as it is a suitable formulation for application on scalp. Ethanol and propylene glycol are good solvents for RU 58841, and 10% in propylene glycol limits skin irritation. The solutions of [14C]RU 58841 were prepared with ethanol:propylene glycol:water (40:10:50, w/w) and contained the same specific radioactivities, final concentrations, and the pH 7 as liposomes.
S0022-3549(96)00394-2 CCC: $14.00
Journal of Pharmaceutical Sciences / 573 Vol. 86, No. 5, May 1997
Radioactivity AssayssScintillation cocktails used for the sample analysis were Picofluor 40, Soluene 350, Hionic Fluor, and Toluene Scintillator (Packard, Rungis, France). Radioactivity was quantified by liquid scintillation counting with a Tri-Carb 4530 counter (Packard) and corrected for quenching by the external standard method. Induced Follicle-Free Rat SkinsThe hairless rat skin was chosen because the sebaceous gland density and size are closely related to those of human forehead skin (100-200 per cm2). After inducing anesthesia with 6% sodium pentobarbital [0.1 mL/100 g, intraperitoneal (ip) route] the 7-week-old hairless male rats (Iffa Credo), were treated by immersing ∼10 cm2 of dorsal skin in 60 °C water for exactly 1 min. The epidermis was then removed and allowed to redevelop over several weeks into a truly follicle-free skin. Within 3 months, the transepidermal water loss, which is an indication of the integrity of the stratum corneum, was similar in scar (7.1 ( 2.3 g/m2/h) and normal (5.2 ( 1.6 g/m2/h) adjacent skins. A detailed description of this methodology and a detailed histological investigation of the regenerated skin have been described by Schaefer et al.7 Briefly, epidermal damage and restructuring of the dermal matrix are minimal. Epidermal lipid analysis showed comparable concentrations after 15 weeks, with the exception of sebaceous lipids. A normal distribution of ceramides was observed. In Vitro Percutaneous Absorption of RU 58841sAnimals were sacrificed by chloroform inhalation. Dorsal scar skin and normal adjacent skin samples of the hairless rats were immediately excised, the subcutaneous fat was removed, and full-thickness skin samples were mounted in static diffusion cells with a 0.685-cm2 surface area. The dermal side was in contact with a 6 mL of aqueous receptor phase (0.9% sodium chloride, 1.5% bovine serum albumin, pH 6.9), continuously stirred, and thermostatically controlled at 37 °C. The receptor phase was totally removed from the cells, which were filled with fresh solution, manually every 2 h up to 10 h and then at 24 h. Ten microliters of the solution or the liposomes (43 µg of [14C]RU 58841, 16 kBq; 980 µg of [3H]phospholipids, 24 kBq) were applied to normal and scar skin samples. After 24 h, the excess of the formulation on the skin surface was wiped off (with cetavlon 1/10 and water). The whole stratum corneum/epidermis was separated from the dermis by rubbing with a scalpel blade until the white surface of the dermis appeared. Stratum corneum/epidermis and dermis samples were digested with, respectively, 1 and 3 mL of soluene 350 (Packard). The radioactivity of all samples (drug excess, epidermis, dermis, and receptor fluid samples) was measured by liquid scintillation counting. The results are expressed in terms of percentages of the applied dose and represent the mean value of nine determinations, together with the standard error of the mean (SEM). In Vivo Cutaneous DistributionsRats with scar skin (∼25 weeks old, 500 g) were anesthetized by ip injection of 6% sodium pentobarbital (0.1 mL/100 g). Two 1.54-cm2 skin areas (one on scar skin and one on adjacent normal skin) were marked out with a glass cylinder and protected with a metallic device. Then, [14C]RU 58841 (8 µL of solution or liposomes, 38 µg of RU 58841, 48 kBq) was applied for 24 h. Three rats were used for each preparation. At the end of the experiment, animals were sacrificed by chloroform inhalation, and the skin was excised and washed with cetavlon 1/10 and water and lightly dried with a cotton swab. The stratum corneum was removed by 15 successive strippings with adhesive tape (3M invisible). Then, the skin was frozen and stored at -20 °C. Three biopsies (6 mm in diameter) were taken from the application area with a biopsy punch (Stiefel, Nanterre, France) and cut parallel to the skin surface (10 slices of 20 µm, 10 slices of 40 µm, 5 slices of 80 µm) with a freezing microtome (Leica Instruments GmbH, Nussloch, Germany) as described by Schaefer and Stuttgen.8 Each strip and each slice was transferred separately into vials with, respectively, 15 and 5 mL of toluene scintillator. The radioactivity of the strippings and the skin slices was measured by liquid scintillation counting 24 h later. The cutaneous distribution of RU 58841 was evaluated on intact and scar skin, in the stratum corneum (µg/cm2), in the epidermis, and dermis (µg/cm3). Cutaneous Localization of RU 58841 by Qualitative AutoradiographysRU 58841 TreatmentsThe solution or liposomes were applied in vivo to the dorsal region of hairless rat skin. A 5-µL/ cm2 sample of the preparation containing 25 µg of [14C]RU 58841 (28 kBq/cm2) was allowed to penetrate for 3, 6, 12, and 24 h. Three animals were treated per time and per preparation. After drug application, a collar was mounted around the neck to avoid licking. At the end of the penetration time, the animals were sacrificed by
574 / Journal of Pharmaceutical Sciences Vol. 86, No. 5, May 1997
Figure 1sAbsorption profiles of RU 58841 and phospholipids through normal and scar hairless rat skin. Results are expressed as percentages of the applied dose (43 µg of RU 58841, 980 µg of phospholipids) and are given as means ± SEM (n ) 9). Comparison between normal and scar skin by Student’s t test, **p < 0.01 chloroform inhalation. The treated skin was excised and wiped gently with cotton swabs soaked in cetavlon 1/10 and water to remove remaining drug. The excised skin was immediatly frozen at -20 °C. Then, 10-µm transversal sections were cut with a freezing microtome (Leica Instruments GmbH, Nussloch, Germany), were collected on glass slides coated with poly-L-lysine (Sigma), and were dried at 4 °C. Preparation of the AutoradiographiessThe slides were placed onto Biomax MR film (Kodak), and they were introduced into an envelope at -20 °C under slight pressure. After exposure times of 2 to 12 days, the autoradiograms were developed and fixed with Industrex Developer and Fixer (Kodak). Then, the slides were stained with Fe trioxyhematein, eosin, phosphomolybdic acid, and green coloring for the histological study. The autoradiograms and the corresponding skin slides were observed under a microscope and photographed.
Results Percutaneous Absorption StudiessAs shown in Figure 1, the percutaneous absorption of RU 58841 was significantly higher in the normal skin than in the scar skin. In the case of the solution, the cumulative percentage of RU 58841 absorbed was threefold higher through normal skin after 24 h. In the case of liposomes, the cumulative percentage of RU 58841 absorbed after 24 h was fourfold higher through normal skin. The percutaneous absorption of the phospholipids was not significantly different between normal and scar skins. The percentages of RU 58841 and phospholipids accumulated in the epidermis and dermis at 24 h are presented in the Table
Table 1sDistribution of RU 58841 and Phospholipids in Normal and Scar Hairless Rat Skin 24 h After In Vitro Applicationa
Stratum corneum/ epidermis Dermis Absorption Total recoveredc
Sample
RU 58841 in Alcoholic Solution
RU 58841 in Liposomes
Phospholipids in Liposomes
Normal Scar
6.49 ± 0.87 14.92*** ± 1.25b
3.37 ± 0.45 3.92 ± 0.47
3.33 ± 0.52 4.07 ± 0.53
Normal Scar Normal Scar
6.38 ± 0.66 5.22 ± 1.35 36 ± 3.27 11.68 ± 1.94
2.74 ± 0.53 1.17* ± 0.22b 6.03 ± 0.57 1.46 ± 0.35
1.34 ± 0.39 0.92 ± 0.25 0.43 ± 0.07 0.22 ± 0.08
Normal Scar
100.32 ± 1.89 100.24 ± 1.48
104 ± 1.59 99 ± 1.7
102.44 ± 1.59 99.8 ± 1.7
a Results are expressed as percentages of the applied dose (43 µg of RU 58841, 980 µg of phospholipids), means ± SEM (n ) 9). b Comparison between normal and scar skin by t test: (*) p < 0.05; (***) p < 0.001. c Included the drug excess on skin surface.
Table 2sIn Vivo Distribution of RU 58841 in the Skin Strata of Normal and Scar Hairless Rat Skina
Stratum corneum Epidermis/Dermis
Sample
RU 58841 in Alcoholic Solution
RU 58841 in Liposomes
Normal Scar Normal Scar
18.5 ± 2.2 30.3* ± 2.2b 1.17 ± 0.13 0.13*** ± 0.05b
3.05 ± 0.49 5.27 ± 0.86 1.18 ± 0.18 0.074*** ± 0.015b
Figure 2sDistribution profiles of RU 58841 in normal and scar skin epidermis and dermis of hairless rat 24 h after in vivo application. Results are expressed as µg/cm3 (mean ± SEM, n ) 9).
a Results are the percentages of the applied dose (38 µg of RU 58841) at 24 h (mean ± SEM of 3 rats). b Comparison between normal and scar skin by t test: (*) p < 0.05, (***) p < 0.001.
1. In the case of the solution, the accumulation of RU 58841 in the stratum corneum/epidermis of the scar skin (14.9%) was significantly higher (p < 0.001) than in normal skin (6.5%), whereas no difference in the accumulation in the dermis of the two skin samples is apparent (5.2 and 6.4%). In the case of liposomes, the accumulation of RU 58841 in the stratum corneum/epidermis was similar in both kinds of skin samples, but the accumulation in the dermis of the normal skin (2.74%) was significantly higher (p < 0.05) than that in the scar skin (1.17%). The accumulation in the stratum corneum/epidermis and the dermis of the phospholipids was not significantly different between normal and scar skin samples. With regard to the simultaneous permeations of RU 58841 and phospholipids (Table 1), it should be noticed that the percentages of the applied doses accumulated in the stratum corneum/epidermis were equivalent (3.37% and 3.3% in the normal stratum corneum/epidermis); on the contrary, a slight difference was observed in the dermis (2.74% and 1.34% in the normal dermis), and the cumulative percentages of RU 58841 absorbed were significantly higher than those of phospholipids (6 and 0.43% after 24 h through normal skin). Quantitative Cutaneous DistributionsMean percentages of the dose of RU 58841 applied to normal and scar skin samples and found in skin layers at 24 h in the in vivo experiment are presented in Table 2. In the stratum corneum, the values were higher in scar skin than in normal skin. This difference was observed with both formulations but the variability in results in terms of standard errors was only significant with the solution. In the epidermis and dermis, the percentages of drugs found in normal skin were significantly higher than in scar skin. The distribution profiles (Figure 2) differed between normal and scar skin. In normal skin, RU 58841 was widely located in the first 500-µm depth, then the concentrations decreased in the deep dermis. In scar skin, the concentrations present in the epidermis (20-µm depth) decreased at 40 µm and were negligible in the dermis.
Figure 3sRatio between the amount of RU 58841 in normal and scar skin as a function of skin depth.
The difference between the distribution profiles could be displayed by calculating the normal skin/scar skin ratio (Figure 3). The difference was maximal at a depth of 200500 µm, the area where the sebaceous glands are localized.9 In this region, the difference was constant and corresponds to a normal skin/scar skin ratio of 20 with the solution, and the ratio values increased up to 40-50 with the liposomes. Qualitative AutoradiographysThe localization of RU 58841 in the cutaneous structures was studied by autoradiography and histology of each skin section. The study was performed at 3, 6, 12, and 24 h, and the results of the 3- and 24-h times are presented in Figure 4. Radioactivity was mainly localized in the stratum corneum and epidermis 3 h after application of the solution, and a diffused radioactivity was observed in the dermis and particularly in the sebaceous ducts. With increasing application time, radioactivity was still highly present in the stratum corneum and epidermis, and radioactivity became more pronounced and appeared deeper in the sebaceous glands. The same penetration routes and sites of deposit were observed with the liposomes, but in different proportions. After 3 h of application, radioactivity was especially localized and concentrated in the sebaceous glands, and slightly present in the stratum corneum and epidermis. With increasing application time, the concentration of the radioactivity in the sebaceous glands became more pronounced, and many sebaceous glands marked in depth were observed. The radioactivity present in the stratum corneum and epidermis remained slight. No radioactivity was concentrated in the hypodermis.
Journal of Pharmaceutical Sciences / 575 Vol. 86, No. 5, May 1997
Figure 4sHistology (A) of the slices of hairless rat skin and the corresponding autoradiography (B), after a topical application of the [14C]RU 58841:4.1, solution, 3 h; 4.2, solution, 24 h; 4.3, liposomes, 3 h; 4.4, liposomes, 24 h. Key: SC, stratum corneum; E, epidermis; D, dermis; H, hypodermis; SD, sebaceous duct; SG, sebaceous gland; DE, drug excess on the skin surface. bar ) 300 µm.
576 / Journal of Pharmaceutical Sciences Vol. 86, No. 5, May 1997
Figure 5sEstimation of the importance of the sebaceous route, in percentage, for RU 58841 as a function of application time in vitro in hairless rat skin.
Discussion To estimate the importance of the sebaceous route of the antiandrogen, we carried out complementary quantitative and qualitative studies. We created an animal model with a follicle-free skin and we compared the drug permeation through the scar sample with that through the adjacent normal skin. The percutaneous absorption studies demonstrated that a higher rate of skin diffusion of RU 58841 was observed with intact skin compared with appendage-free skin. The absorption through the sebaceous glands can be estimated as follows: (normal skin - scar skin)/normal skin (using the percentages of absorption of each sample, noncumulative values; Figure 5). The main route of RU 58841 permeation through the hairless rat skin was the sebaceous route, which represented greater than 50% of the total absorption under all conditions. This observation is corroborated by investigations using the same model with different compounds on human skin10 and hairless rat skin.11,12 With the solution, the sebaceous route decreased over time from 89% at 2 h to 51% at 24 h, whereas it appeared stable during the entire application time (67 to 76%) with liposomes. The sebaceous route was the most substantial one with the solution during the first 6 h, then it was higher with liposomes between 8 h and 24 h. A study13 of the vehicle influence on the relative importance of the sebaceous route showed that it is possible to modulate the relative importance of the transepidermal and the transfollicular pathway. RU 58841 absorption was proportional to time with liposomes; whereas, a decrease of slope was observed between 10 and 24 h with the solution (Figure 1). We applied finite doses (10 µL/0.685 cm2) to be close to real conditions of application, thus Fick’s law conditions were not respected and the absorption kinetic was complex. In the case of solution, ethanol induced a high rate of diffusion by the sebaceous route during the first hours (a high diffusion could explain the slight accumulation of RU 58841 in the sebaceous glands observed on the autoradiographies). Then, evaporation occured, the ethanol effect disappeared, the concentration of RU 58841 increased at the level of stratum corneum and favored epidermal absorption, and finally the important absorption could induce a drug depletion. With liposomes, the composition of the donor compartment was more stable, RU 58841 removed progressively, and diffusion of encapsulated RU 58841 was delayed and regular. With regard to drug amount in skin samples, a higher accumulation of RU 58841 in the dermis of intact skin compared with scar skin was observed only with liposomes. This observation indicated that liposomes favored the localization of RU 58841 in the sebaceous glands. The comparison between the simultaneous permeations of RU 58841 and phospholipids showed a progressive dissociation of the two compounds. The sebaceous route of the phospholipids was not predominant, confirming the separation of the two compounds. In the in vivo cutaneous distribution study, the quantities of drug in the stratum corneum were almost two times higher
in scar skin than in normal skin. This accumulation in scar skin was also observed by Hueber et al.14 The important result was that the RU 58841 present in the dermis was mainly localized in the sebaceous ducts and glands because the epidermis/dermis without sebaceous units contained nine times less than the normal skin with the solution and 16 times less with liposomes. This result was confirmed by the normal skin/scar skin ratio, which displayed an accumulation of the drug in the sebaceous area (Figure 3). This accumulation was markedly more important with liposomes. Lieb et al.15 showed that the sebaceous concentration of cimetidine in the hamster ear was higher with liposomes than with alcoholic solution. The autoradiography study showed an accumulation of RU 58841 in the stratum corneum and a localization in the sebaceous glands. With the solution, the accumulation in the stratum corneum was more substantial. The sebaceous glands appeared not to have clear contours, indicating that the drug also diffused through the dermis cells. With liposomes, the accumulation in the stratum corneum was slight compared with the solution and compared with the accumulation in the sebaceous glands. The sebaceous glands appeared highly and perfectly marked, demonstrating the target effect of the liposomes. These observations agreed with the results of the quantitative studies. Several studies demonstrated, by autoradiography,16-18 the specific drug delivery to the pilosebaceous unit and the influence of time and vehicles. The physicochemical properties of the drug influence the transport within the skin. RU 58841 is a slightly hydrophobic drug (water solubility, 0.44 mg/mL) with a low molecular weight (369.4) and an octanol-water partition coefficient (log Koctanol-water of 1.76. Therefore, RU 58841 is a good candidate for a transdermal delivery and it appears that the lipid regions (intercellular and sebaceous lipids) would constitute the main pathways for the cutaneous transport of RU 58841, as confirmed by our results. Ethanol and propylene glycol are enhancers for the skin permeation of lipophilic drugs.19,20 These solvents may favor RU 58841 dissolving in the sebum and open a passageway within the sebaceous glands. In spite of the favorable action of the alcoholic solution to transport RU 58841 into the lipidic areas such as the sebaceous glands, liposomes targeted the sebaceous glands best. Several investigations demonstrated that liposomes allow a higher accumulation in the sebaceous glands compared with nonliposomal formulations.21-26 Our results displayed a dissociation between RU 58841 and phospholipids in the dermis, indicating that intact liposomes did not penetrate intact into the sebaceous glands, as it is still maintained by Mezei.27
Conclusions With the aim of determining the route of penetration of the antiandrogen into the skin, we studied the percutaneous absorption in vitro throught normal and scar follicle-free hairless rat skin, the quantitative cutaneous distribution in vivo and finally the cutaneous distribution by autoradiography. The results of these complementary experiments concur that the main transdermal route of RU 58841 is the sebaceous gland route and that RU 58841 tends to accumulate in the stratum corneum and the sebaceous glands. The solution favors localization in the stratum corneum and liposomes in the sebaceous glands. The simultaneous permeations of RU 58841 and phospholipids show a progressive dissociation of the two compounds. The sebaceous route is not predominant for phospholipids, confirming the separation of the molecules and indicating that the liposomes do not penetrate intact into the sebaceous glands.
Journal of Pharmaceutical Sciences / 577 Vol. 86, No. 5, May 1997
References and Notes 1. Lauer, A. C.; Lieb, L. M.; Ramachandran, C.; Flynn, G. L.; Weiner, N. D. Pharm. Res. 1995, 12, 179-186. 2. Lauer, A. C.; Ramachandran, C.; Lieb, L. M.; Niemec, S.; Weiner, N. D. Adv. Drug Deliv. Rev. 1996, 18, 311-324. 3. Battmann, T.; Bonfils, A.; Branche, C.; Humbert, J.; Goubet, F.; Teutsch, G.; Philibert, D. J. Steroid Biochem. Molec. Biol. 1994, 48, 55-60. 4. Choudhry, R.; Hodgin, M. B.; Van der Kwast, T. H.; Brinkmann, A. O.; Boersma, W. J. A. J. Endrocrinol. 1992, 133, 467-475. 5. Illel, B.; Schaefer, H. Acta Derm. Venerol. 1988, 68, 427-430. 6. Bernard, E.; Dubois, J. L.; Wepierre, J. Int. J. Pharm. 1995, 126, 235-243. 7. Schaefer, H.; Watts, F.; Brod, J.; Illel, B. In Prediction of Percutaneous Penetration: Methods, Measurements and Modelling; Scott, R. C.; Guy, R. H.; Hadgraft, J., Eds.; IBC Technical Services: London, 1990; pp 163-173. 8. Schaefer, H.; Stuttgen, G. Arzneim.-Forsch. 1976, 26, 432-435. 9. Roguet, R. et al. Arch. Dermatol. Res. 1986, 278, 503-506. 10. Hueber, F.; Schaefer, H.; Wepierre, J. Skin Pharmacol. 1994, 7, 237-244. 11. Hueber, F.; Besnard, M.; Schaefer, H.; Wepierre, J. Skin Pharmacol. 1994, 7, 245-256. 12. Illel, B.; Schaefer, H.; Wepierre, J.; Doucet, O. J. Pharm. Sci. 1991, 80, 424-427. 13. Bamba, F. L.; Wepierre, J. Eur. J. Drug Metab. Pharmacokinet. 1993, 18, 339-348. 14. Hueber, F.; Wepierre, J.; Schaefer, H. Skin Pharmacol. 1992, 5, 99-107.
578 / Journal of Pharmaceutical Sciences Vol. 86, No. 5, May 1997
15. Lieb, L. M.; Flynn, G.; Weiner, N. Pharm. Res. 1994, 11, 14191423. 16. Rolland, A.; Wagner, N.; Chatelus, A.; Shroot, B.; Schaefer, H. Pharm. Res. 1993, 10, 1738-1744. 17. Fabin, F.; Touitou, E. Int. J. Pharm. 1991, 74, 59-65. 18. Bidmon, H. J.; Pitts, J. D.; Solomon, H. F.; Bondi, J. V.; Stumpf, W. E. Histochemistry 1990, 95, 43-54. 19. Berner, B.; Liu, P. In Percutaneous Penetration Enhancers; Smith, E. W.; Maibach, H. I., Eds.; CRC: Boca Raton, FL, 1995; pp 45-60. 20. Bendas, B.; Neubert, R.; Wohlrab, W. In Percutaneous Penetration Enhancers; Smith, E. W.; Maibach, H. I., Eds.; CRC: Boca Raton, 1995; pp 61-77. 21. Lieb, L. M.; Ramachandran, C.; Egbaria, K.; Weiner, N. Follic. Deliv. Liposomes 1992, 99, 108-113. 22. Li, L.; Margolis, L. B.; Lishko, V. K.; Hoffman, R. M. In Vitro Cell. Dev. Biol. 1992, 28A, 679-681. 23. Li, L.; Lishko, V. K.; Hoffman, R. M. In Vitro Cell. Dev. Biol. 1993, 29A, 192-194. 24. Li, L.; Lishko, V. K.; Hoffman, R. M. In Vitro Cell. Dev. Biol. 1993, 29A, 258-260. 25. Du Plessis, J.; Egbaria, K.; Ramachandran, C.; Weiner, N. Antiviral Res. 1992, 18, 259-265. 26. Niemiec, S.; Ramachandran, C.; Weiner, N. Pharm. Res. 1995, 12, 1184-1188. 27. Mezei, M. In Drug Permeation Enhancement: Theory and Applications; Hsieh, D. S., Ed.; M. Dekker: Basel, 1994; pp 171-198.
JS960394L
AND JACQUES
WEPIERRE*X
Received September 18, 1996, from the *Laboratoire de dermopharmacologie, Faculte´ de Pharmacie, 5 rue J. B. Cle´ment, 92290 Chaˆtenay-Malabry, France, and the ‡Laboratoire de recherche Gale´nique, ROUSSEL UCLAF, 102 route de Noisy, 93235 Romainville, France. Final revised manuscript received December 12, 1996. Accepted for publication January 17, 1997X. Abstract 0 The significance of the sebaceous gland pathway in the cutaneous permeation of an antiandrogen, 4-[3-(4-hydroxybutyl)-4,4dimethyl-2,5-dioxo-1-imidazolidinyl]-2-(trifluoromethyl)benzonitrile (RU 58841), was studied with normal hairless rat skin and an induced scar hairless rat skin without sebaceous glands. RU 58841 was dissolved in an alcoholic solution and encapsulated in liposomes for comparison. After 24 h, the cumulative percentage of RU 58841 absorbed in vitro was 3−4-fold higher in the normal skin than in the scar skin; in the case of liposomes, the accumulation of the drug in the normal dermis was significantly higher than in the scar one. In the in vivo cutaneous distribution, the epidermis and dermis of the normal skin contained higher amounts of RU 58841 than the scar skin (ninefold with the solution and 16-fold with liposomes). An autoradiography study showed that with the solution, the drug was mainly localized in the stratum corneum/epidermis, and with the liposomes, the drug was mainly localized in the sebaceous glands. We concluded that the sebaceous glands constituted the main pathway for RU 58841. The alcoholic solution encouraged the localization of the drug into the stratum corneum, whereas liposomes targeted the sebaceous glands.
Introduction It is now accepted that hair follicles and sebaceous glands may participate in skin penetration for a wide range of compounds.1 Several investigations were carried out to evaluate the importance of the transfollicular route and to develop models permitting the assessment of the pathways in permeation.2 Another reason for these recent reports is that sebaceous glands and hair follicles are the site of dermatological disorders such as acne and androgenetic alopecia. RU 58841 (4-[3-(4-hydroxybutyl)-4,4-dimethyl-2,5dioxo-1-imidazolidinyl]-2-(trifluoromethyl)benzonitrile) is a topical antiandrogen that is being developed for the treatment of androgen-dependent diseases. RU 58841 is a benzonitrile derivative, slightly soluble in water (0.44 mg/mL), freely soluble in ethanol (250 mg/mL), and has a log octanol/water partition coefficient of 1.76. The drug has a molecular weight of 369.4, and appears as a white powder with a melting point around 101 °C. RU 58841 has selective binding to the androgen receptors,3 which are localized especially in the sebaceous glands and hair follicles of human skin.4 Thus, the interest in the pilosebaceous unit as a target for RU 58841 is widely justified. The aim of the present study was to determine the routes of penetration of this molecule, and particularly to assess the importance of the sebaceous glands. At the present time, no perfect animal model exists to study transepidermal and transfollicular routes distinctly. We used a technique adapted by Illel and Schaefer5 consisting of inducing a truly follicle-free skin of hairless rats. The method allows the cutaneous permeation between the follicle and X
Abstract published in Advance ACS Abstracts, March 15, 1997.
© 1997, American Chemical Society and American Pharmaceutical Association
sebaceous gland-free skin and the normal hairless rat skin to be compared. We quantified the contribution of the sebaceous glands in the percutaneous absorption in vitro and in the distribution in the skin strata in vivo. Finally, we completed these investigations by assessing the localization of RU 58841 within the skin by a qualitative autoradiography. The effect of vehicle on the penetration pathways was also studied, with a comparison between the permeation of RU 58841 dissolved in an alcoholic solution, suitable to scalp in the treatment of androgenetic alopecia, or encapsulated in liposomes, suitable to skin in the treatment of acne. A previous investigation6 indicated that liposomes, compared with the solution, reduce percutaneous absorption and increase retention of the drug in the dermis.
Experimental Section ChemicalssRU 58841 (4-[3-(4-hydroxybutyl)-4,4-dimethyl-2,5dioxo-1-imidazolidinyl]-2-(trifluoromethyl)-benzonitrile) is a novel topical antiandrogen that is being developed by Roussel Uclaf. RU 58841 and [14C]RU 58841, synthesized in the Roussel Uclaf laboratories, were mixed to obtain specific radioactivities of 355 KBq/mg and 1.2 MBq/mg. Radiochemical purity was >98%. Lipoid E 100s35 (Lipoid KG, Ludwigshafen, Germany); L-3phosphatidyl[N-methyl-3H]choline 1,2-dipalmitoyl (Amersham), specific activity 3.11 GBq/mg; R-tocopherol (Fluka AG, Buchs, Switzerland); and a phosphate buffer were used for the liposomes preparation. Ethanol and propylene glycol were used for the solution.
Liposome FormulationsSmall unilamellar liposomes (SUV) containing 0.5% weight of [14C]RU 58841 were prepared with Lipoid E 100s35 (egg phosphatidylcholine > 94%) and R-tocopherol in a 0.05 M phosphate buffer pH 7. Briefly, powders of [14C]RU 58841, R-tocopherol, and Lipoid E 100s35 were mixed at 60 °C. Then, the appropriate volume of phosphate buffer was added at 60 °C with gentle stirring. Finally, the mixture was homogenized with a microfluidizer M110S (Sodexim, Muizon, France). Liposome size was measured with a N4 MD Coulter counter (Coultronics, Margency, France), and the mean volume diameter was 84 ( 12 nm. In the in vitro percutaneous absorption studies, the final concentration for [14C]RU 58841 was 4.3 mg/g, with an activity of 1.6 MBq/g of preparation, and the final concentration for [3H]phospholipids was 98 mg/g, with an activity of 2.45 MBq/g of preparation. In the in vivo quantitative cutaneous distribution and qualitative autoradiography, the final concentration for [14C]RU 58841 was 4.3 mg/g, with an activity of 5.2 MBq/g of preparation. SolutionsAlcoholic solution was also studied as it is a suitable formulation for application on scalp. Ethanol and propylene glycol are good solvents for RU 58841, and 10% in propylene glycol limits skin irritation. The solutions of [14C]RU 58841 were prepared with ethanol:propylene glycol:water (40:10:50, w/w) and contained the same specific radioactivities, final concentrations, and the pH 7 as liposomes.
S0022-3549(96)00394-2 CCC: $14.00
Journal of Pharmaceutical Sciences / 573 Vol. 86, No. 5, May 1997
Radioactivity AssayssScintillation cocktails used for the sample analysis were Picofluor 40, Soluene 350, Hionic Fluor, and Toluene Scintillator (Packard, Rungis, France). Radioactivity was quantified by liquid scintillation counting with a Tri-Carb 4530 counter (Packard) and corrected for quenching by the external standard method. Induced Follicle-Free Rat SkinsThe hairless rat skin was chosen because the sebaceous gland density and size are closely related to those of human forehead skin (100-200 per cm2). After inducing anesthesia with 6% sodium pentobarbital [0.1 mL/100 g, intraperitoneal (ip) route] the 7-week-old hairless male rats (Iffa Credo), were treated by immersing ∼10 cm2 of dorsal skin in 60 °C water for exactly 1 min. The epidermis was then removed and allowed to redevelop over several weeks into a truly follicle-free skin. Within 3 months, the transepidermal water loss, which is an indication of the integrity of the stratum corneum, was similar in scar (7.1 ( 2.3 g/m2/h) and normal (5.2 ( 1.6 g/m2/h) adjacent skins. A detailed description of this methodology and a detailed histological investigation of the regenerated skin have been described by Schaefer et al.7 Briefly, epidermal damage and restructuring of the dermal matrix are minimal. Epidermal lipid analysis showed comparable concentrations after 15 weeks, with the exception of sebaceous lipids. A normal distribution of ceramides was observed. In Vitro Percutaneous Absorption of RU 58841sAnimals were sacrificed by chloroform inhalation. Dorsal scar skin and normal adjacent skin samples of the hairless rats were immediately excised, the subcutaneous fat was removed, and full-thickness skin samples were mounted in static diffusion cells with a 0.685-cm2 surface area. The dermal side was in contact with a 6 mL of aqueous receptor phase (0.9% sodium chloride, 1.5% bovine serum albumin, pH 6.9), continuously stirred, and thermostatically controlled at 37 °C. The receptor phase was totally removed from the cells, which were filled with fresh solution, manually every 2 h up to 10 h and then at 24 h. Ten microliters of the solution or the liposomes (43 µg of [14C]RU 58841, 16 kBq; 980 µg of [3H]phospholipids, 24 kBq) were applied to normal and scar skin samples. After 24 h, the excess of the formulation on the skin surface was wiped off (with cetavlon 1/10 and water). The whole stratum corneum/epidermis was separated from the dermis by rubbing with a scalpel blade until the white surface of the dermis appeared. Stratum corneum/epidermis and dermis samples were digested with, respectively, 1 and 3 mL of soluene 350 (Packard). The radioactivity of all samples (drug excess, epidermis, dermis, and receptor fluid samples) was measured by liquid scintillation counting. The results are expressed in terms of percentages of the applied dose and represent the mean value of nine determinations, together with the standard error of the mean (SEM). In Vivo Cutaneous DistributionsRats with scar skin (∼25 weeks old, 500 g) were anesthetized by ip injection of 6% sodium pentobarbital (0.1 mL/100 g). Two 1.54-cm2 skin areas (one on scar skin and one on adjacent normal skin) were marked out with a glass cylinder and protected with a metallic device. Then, [14C]RU 58841 (8 µL of solution or liposomes, 38 µg of RU 58841, 48 kBq) was applied for 24 h. Three rats were used for each preparation. At the end of the experiment, animals were sacrificed by chloroform inhalation, and the skin was excised and washed with cetavlon 1/10 and water and lightly dried with a cotton swab. The stratum corneum was removed by 15 successive strippings with adhesive tape (3M invisible). Then, the skin was frozen and stored at -20 °C. Three biopsies (6 mm in diameter) were taken from the application area with a biopsy punch (Stiefel, Nanterre, France) and cut parallel to the skin surface (10 slices of 20 µm, 10 slices of 40 µm, 5 slices of 80 µm) with a freezing microtome (Leica Instruments GmbH, Nussloch, Germany) as described by Schaefer and Stuttgen.8 Each strip and each slice was transferred separately into vials with, respectively, 15 and 5 mL of toluene scintillator. The radioactivity of the strippings and the skin slices was measured by liquid scintillation counting 24 h later. The cutaneous distribution of RU 58841 was evaluated on intact and scar skin, in the stratum corneum (µg/cm2), in the epidermis, and dermis (µg/cm3). Cutaneous Localization of RU 58841 by Qualitative AutoradiographysRU 58841 TreatmentsThe solution or liposomes were applied in vivo to the dorsal region of hairless rat skin. A 5-µL/ cm2 sample of the preparation containing 25 µg of [14C]RU 58841 (28 kBq/cm2) was allowed to penetrate for 3, 6, 12, and 24 h. Three animals were treated per time and per preparation. After drug application, a collar was mounted around the neck to avoid licking. At the end of the penetration time, the animals were sacrificed by
574 / Journal of Pharmaceutical Sciences Vol. 86, No. 5, May 1997
Figure 1sAbsorption profiles of RU 58841 and phospholipids through normal and scar hairless rat skin. Results are expressed as percentages of the applied dose (43 µg of RU 58841, 980 µg of phospholipids) and are given as means ± SEM (n ) 9). Comparison between normal and scar skin by Student’s t test, **p < 0.01 chloroform inhalation. The treated skin was excised and wiped gently with cotton swabs soaked in cetavlon 1/10 and water to remove remaining drug. The excised skin was immediatly frozen at -20 °C. Then, 10-µm transversal sections were cut with a freezing microtome (Leica Instruments GmbH, Nussloch, Germany), were collected on glass slides coated with poly-L-lysine (Sigma), and were dried at 4 °C. Preparation of the AutoradiographiessThe slides were placed onto Biomax MR film (Kodak), and they were introduced into an envelope at -20 °C under slight pressure. After exposure times of 2 to 12 days, the autoradiograms were developed and fixed with Industrex Developer and Fixer (Kodak). Then, the slides were stained with Fe trioxyhematein, eosin, phosphomolybdic acid, and green coloring for the histological study. The autoradiograms and the corresponding skin slides were observed under a microscope and photographed.
Results Percutaneous Absorption StudiessAs shown in Figure 1, the percutaneous absorption of RU 58841 was significantly higher in the normal skin than in the scar skin. In the case of the solution, the cumulative percentage of RU 58841 absorbed was threefold higher through normal skin after 24 h. In the case of liposomes, the cumulative percentage of RU 58841 absorbed after 24 h was fourfold higher through normal skin. The percutaneous absorption of the phospholipids was not significantly different between normal and scar skins. The percentages of RU 58841 and phospholipids accumulated in the epidermis and dermis at 24 h are presented in the Table
Table 1sDistribution of RU 58841 and Phospholipids in Normal and Scar Hairless Rat Skin 24 h After In Vitro Applicationa
Stratum corneum/ epidermis Dermis Absorption Total recoveredc
Sample
RU 58841 in Alcoholic Solution
RU 58841 in Liposomes
Phospholipids in Liposomes
Normal Scar
6.49 ± 0.87 14.92*** ± 1.25b
3.37 ± 0.45 3.92 ± 0.47
3.33 ± 0.52 4.07 ± 0.53
Normal Scar Normal Scar
6.38 ± 0.66 5.22 ± 1.35 36 ± 3.27 11.68 ± 1.94
2.74 ± 0.53 1.17* ± 0.22b 6.03 ± 0.57 1.46 ± 0.35
1.34 ± 0.39 0.92 ± 0.25 0.43 ± 0.07 0.22 ± 0.08
Normal Scar
100.32 ± 1.89 100.24 ± 1.48
104 ± 1.59 99 ± 1.7
102.44 ± 1.59 99.8 ± 1.7
a Results are expressed as percentages of the applied dose (43 µg of RU 58841, 980 µg of phospholipids), means ± SEM (n ) 9). b Comparison between normal and scar skin by t test: (*) p < 0.05; (***) p < 0.001. c Included the drug excess on skin surface.
Table 2sIn Vivo Distribution of RU 58841 in the Skin Strata of Normal and Scar Hairless Rat Skina
Stratum corneum Epidermis/Dermis
Sample
RU 58841 in Alcoholic Solution
RU 58841 in Liposomes
Normal Scar Normal Scar
18.5 ± 2.2 30.3* ± 2.2b 1.17 ± 0.13 0.13*** ± 0.05b
3.05 ± 0.49 5.27 ± 0.86 1.18 ± 0.18 0.074*** ± 0.015b
Figure 2sDistribution profiles of RU 58841 in normal and scar skin epidermis and dermis of hairless rat 24 h after in vivo application. Results are expressed as µg/cm3 (mean ± SEM, n ) 9).
a Results are the percentages of the applied dose (38 µg of RU 58841) at 24 h (mean ± SEM of 3 rats). b Comparison between normal and scar skin by t test: (*) p < 0.05, (***) p < 0.001.
1. In the case of the solution, the accumulation of RU 58841 in the stratum corneum/epidermis of the scar skin (14.9%) was significantly higher (p < 0.001) than in normal skin (6.5%), whereas no difference in the accumulation in the dermis of the two skin samples is apparent (5.2 and 6.4%). In the case of liposomes, the accumulation of RU 58841 in the stratum corneum/epidermis was similar in both kinds of skin samples, but the accumulation in the dermis of the normal skin (2.74%) was significantly higher (p < 0.05) than that in the scar skin (1.17%). The accumulation in the stratum corneum/epidermis and the dermis of the phospholipids was not significantly different between normal and scar skin samples. With regard to the simultaneous permeations of RU 58841 and phospholipids (Table 1), it should be noticed that the percentages of the applied doses accumulated in the stratum corneum/epidermis were equivalent (3.37% and 3.3% in the normal stratum corneum/epidermis); on the contrary, a slight difference was observed in the dermis (2.74% and 1.34% in the normal dermis), and the cumulative percentages of RU 58841 absorbed were significantly higher than those of phospholipids (6 and 0.43% after 24 h through normal skin). Quantitative Cutaneous DistributionsMean percentages of the dose of RU 58841 applied to normal and scar skin samples and found in skin layers at 24 h in the in vivo experiment are presented in Table 2. In the stratum corneum, the values were higher in scar skin than in normal skin. This difference was observed with both formulations but the variability in results in terms of standard errors was only significant with the solution. In the epidermis and dermis, the percentages of drugs found in normal skin were significantly higher than in scar skin. The distribution profiles (Figure 2) differed between normal and scar skin. In normal skin, RU 58841 was widely located in the first 500-µm depth, then the concentrations decreased in the deep dermis. In scar skin, the concentrations present in the epidermis (20-µm depth) decreased at 40 µm and were negligible in the dermis.
Figure 3sRatio between the amount of RU 58841 in normal and scar skin as a function of skin depth.
The difference between the distribution profiles could be displayed by calculating the normal skin/scar skin ratio (Figure 3). The difference was maximal at a depth of 200500 µm, the area where the sebaceous glands are localized.9 In this region, the difference was constant and corresponds to a normal skin/scar skin ratio of 20 with the solution, and the ratio values increased up to 40-50 with the liposomes. Qualitative AutoradiographysThe localization of RU 58841 in the cutaneous structures was studied by autoradiography and histology of each skin section. The study was performed at 3, 6, 12, and 24 h, and the results of the 3- and 24-h times are presented in Figure 4. Radioactivity was mainly localized in the stratum corneum and epidermis 3 h after application of the solution, and a diffused radioactivity was observed in the dermis and particularly in the sebaceous ducts. With increasing application time, radioactivity was still highly present in the stratum corneum and epidermis, and radioactivity became more pronounced and appeared deeper in the sebaceous glands. The same penetration routes and sites of deposit were observed with the liposomes, but in different proportions. After 3 h of application, radioactivity was especially localized and concentrated in the sebaceous glands, and slightly present in the stratum corneum and epidermis. With increasing application time, the concentration of the radioactivity in the sebaceous glands became more pronounced, and many sebaceous glands marked in depth were observed. The radioactivity present in the stratum corneum and epidermis remained slight. No radioactivity was concentrated in the hypodermis.
Journal of Pharmaceutical Sciences / 575 Vol. 86, No. 5, May 1997
Figure 4sHistology (A) of the slices of hairless rat skin and the corresponding autoradiography (B), after a topical application of the [14C]RU 58841:4.1, solution, 3 h; 4.2, solution, 24 h; 4.3, liposomes, 3 h; 4.4, liposomes, 24 h. Key: SC, stratum corneum; E, epidermis; D, dermis; H, hypodermis; SD, sebaceous duct; SG, sebaceous gland; DE, drug excess on the skin surface. bar ) 300 µm.
576 / Journal of Pharmaceutical Sciences Vol. 86, No. 5, May 1997
Figure 5sEstimation of the importance of the sebaceous route, in percentage, for RU 58841 as a function of application time in vitro in hairless rat skin.
Discussion To estimate the importance of the sebaceous route of the antiandrogen, we carried out complementary quantitative and qualitative studies. We created an animal model with a follicle-free skin and we compared the drug permeation through the scar sample with that through the adjacent normal skin. The percutaneous absorption studies demonstrated that a higher rate of skin diffusion of RU 58841 was observed with intact skin compared with appendage-free skin. The absorption through the sebaceous glands can be estimated as follows: (normal skin - scar skin)/normal skin (using the percentages of absorption of each sample, noncumulative values; Figure 5). The main route of RU 58841 permeation through the hairless rat skin was the sebaceous route, which represented greater than 50% of the total absorption under all conditions. This observation is corroborated by investigations using the same model with different compounds on human skin10 and hairless rat skin.11,12 With the solution, the sebaceous route decreased over time from 89% at 2 h to 51% at 24 h, whereas it appeared stable during the entire application time (67 to 76%) with liposomes. The sebaceous route was the most substantial one with the solution during the first 6 h, then it was higher with liposomes between 8 h and 24 h. A study13 of the vehicle influence on the relative importance of the sebaceous route showed that it is possible to modulate the relative importance of the transepidermal and the transfollicular pathway. RU 58841 absorption was proportional to time with liposomes; whereas, a decrease of slope was observed between 10 and 24 h with the solution (Figure 1). We applied finite doses (10 µL/0.685 cm2) to be close to real conditions of application, thus Fick’s law conditions were not respected and the absorption kinetic was complex. In the case of solution, ethanol induced a high rate of diffusion by the sebaceous route during the first hours (a high diffusion could explain the slight accumulation of RU 58841 in the sebaceous glands observed on the autoradiographies). Then, evaporation occured, the ethanol effect disappeared, the concentration of RU 58841 increased at the level of stratum corneum and favored epidermal absorption, and finally the important absorption could induce a drug depletion. With liposomes, the composition of the donor compartment was more stable, RU 58841 removed progressively, and diffusion of encapsulated RU 58841 was delayed and regular. With regard to drug amount in skin samples, a higher accumulation of RU 58841 in the dermis of intact skin compared with scar skin was observed only with liposomes. This observation indicated that liposomes favored the localization of RU 58841 in the sebaceous glands. The comparison between the simultaneous permeations of RU 58841 and phospholipids showed a progressive dissociation of the two compounds. The sebaceous route of the phospholipids was not predominant, confirming the separation of the two compounds. In the in vivo cutaneous distribution study, the quantities of drug in the stratum corneum were almost two times higher
in scar skin than in normal skin. This accumulation in scar skin was also observed by Hueber et al.14 The important result was that the RU 58841 present in the dermis was mainly localized in the sebaceous ducts and glands because the epidermis/dermis without sebaceous units contained nine times less than the normal skin with the solution and 16 times less with liposomes. This result was confirmed by the normal skin/scar skin ratio, which displayed an accumulation of the drug in the sebaceous area (Figure 3). This accumulation was markedly more important with liposomes. Lieb et al.15 showed that the sebaceous concentration of cimetidine in the hamster ear was higher with liposomes than with alcoholic solution. The autoradiography study showed an accumulation of RU 58841 in the stratum corneum and a localization in the sebaceous glands. With the solution, the accumulation in the stratum corneum was more substantial. The sebaceous glands appeared not to have clear contours, indicating that the drug also diffused through the dermis cells. With liposomes, the accumulation in the stratum corneum was slight compared with the solution and compared with the accumulation in the sebaceous glands. The sebaceous glands appeared highly and perfectly marked, demonstrating the target effect of the liposomes. These observations agreed with the results of the quantitative studies. Several studies demonstrated, by autoradiography,16-18 the specific drug delivery to the pilosebaceous unit and the influence of time and vehicles. The physicochemical properties of the drug influence the transport within the skin. RU 58841 is a slightly hydrophobic drug (water solubility, 0.44 mg/mL) with a low molecular weight (369.4) and an octanol-water partition coefficient (log Koctanol-water of 1.76. Therefore, RU 58841 is a good candidate for a transdermal delivery and it appears that the lipid regions (intercellular and sebaceous lipids) would constitute the main pathways for the cutaneous transport of RU 58841, as confirmed by our results. Ethanol and propylene glycol are enhancers for the skin permeation of lipophilic drugs.19,20 These solvents may favor RU 58841 dissolving in the sebum and open a passageway within the sebaceous glands. In spite of the favorable action of the alcoholic solution to transport RU 58841 into the lipidic areas such as the sebaceous glands, liposomes targeted the sebaceous glands best. Several investigations demonstrated that liposomes allow a higher accumulation in the sebaceous glands compared with nonliposomal formulations.21-26 Our results displayed a dissociation between RU 58841 and phospholipids in the dermis, indicating that intact liposomes did not penetrate intact into the sebaceous glands, as it is still maintained by Mezei.27
Conclusions With the aim of determining the route of penetration of the antiandrogen into the skin, we studied the percutaneous absorption in vitro throught normal and scar follicle-free hairless rat skin, the quantitative cutaneous distribution in vivo and finally the cutaneous distribution by autoradiography. The results of these complementary experiments concur that the main transdermal route of RU 58841 is the sebaceous gland route and that RU 58841 tends to accumulate in the stratum corneum and the sebaceous glands. The solution favors localization in the stratum corneum and liposomes in the sebaceous glands. The simultaneous permeations of RU 58841 and phospholipids show a progressive dissociation of the two compounds. The sebaceous route is not predominant for phospholipids, confirming the separation of the molecules and indicating that the liposomes do not penetrate intact into the sebaceous glands.
Journal of Pharmaceutical Sciences / 577 Vol. 86, No. 5, May 1997
References and Notes 1. Lauer, A. C.; Lieb, L. M.; Ramachandran, C.; Flynn, G. L.; Weiner, N. D. Pharm. Res. 1995, 12, 179-186. 2. Lauer, A. C.; Ramachandran, C.; Lieb, L. M.; Niemec, S.; Weiner, N. D. Adv. Drug Deliv. Rev. 1996, 18, 311-324. 3. Battmann, T.; Bonfils, A.; Branche, C.; Humbert, J.; Goubet, F.; Teutsch, G.; Philibert, D. J. Steroid Biochem. Molec. Biol. 1994, 48, 55-60. 4. Choudhry, R.; Hodgin, M. B.; Van der Kwast, T. H.; Brinkmann, A. O.; Boersma, W. J. A. J. Endrocrinol. 1992, 133, 467-475. 5. Illel, B.; Schaefer, H. Acta Derm. Venerol. 1988, 68, 427-430. 6. Bernard, E.; Dubois, J. L.; Wepierre, J. Int. J. Pharm. 1995, 126, 235-243. 7. Schaefer, H.; Watts, F.; Brod, J.; Illel, B. In Prediction of Percutaneous Penetration: Methods, Measurements and Modelling; Scott, R. C.; Guy, R. H.; Hadgraft, J., Eds.; IBC Technical Services: London, 1990; pp 163-173. 8. Schaefer, H.; Stuttgen, G. Arzneim.-Forsch. 1976, 26, 432-435. 9. Roguet, R. et al. Arch. Dermatol. Res. 1986, 278, 503-506. 10. Hueber, F.; Schaefer, H.; Wepierre, J. Skin Pharmacol. 1994, 7, 237-244. 11. Hueber, F.; Besnard, M.; Schaefer, H.; Wepierre, J. Skin Pharmacol. 1994, 7, 245-256. 12. Illel, B.; Schaefer, H.; Wepierre, J.; Doucet, O. J. Pharm. Sci. 1991, 80, 424-427. 13. Bamba, F. L.; Wepierre, J. Eur. J. Drug Metab. Pharmacokinet. 1993, 18, 339-348. 14. Hueber, F.; Wepierre, J.; Schaefer, H. Skin Pharmacol. 1992, 5, 99-107.
578 / Journal of Pharmaceutical Sciences Vol. 86, No. 5, May 1997
15. Lieb, L. M.; Flynn, G.; Weiner, N. Pharm. Res. 1994, 11, 14191423. 16. Rolland, A.; Wagner, N.; Chatelus, A.; Shroot, B.; Schaefer, H. Pharm. Res. 1993, 10, 1738-1744. 17. Fabin, F.; Touitou, E. Int. J. Pharm. 1991, 74, 59-65. 18. Bidmon, H. J.; Pitts, J. D.; Solomon, H. F.; Bondi, J. V.; Stumpf, W. E. Histochemistry 1990, 95, 43-54. 19. Berner, B.; Liu, P. In Percutaneous Penetration Enhancers; Smith, E. W.; Maibach, H. I., Eds.; CRC: Boca Raton, FL, 1995; pp 45-60. 20. Bendas, B.; Neubert, R.; Wohlrab, W. In Percutaneous Penetration Enhancers; Smith, E. W.; Maibach, H. I., Eds.; CRC: Boca Raton, 1995; pp 61-77. 21. Lieb, L. M.; Ramachandran, C.; Egbaria, K.; Weiner, N. Follic. Deliv. Liposomes 1992, 99, 108-113. 22. Li, L.; Margolis, L. B.; Lishko, V. K.; Hoffman, R. M. In Vitro Cell. Dev. Biol. 1992, 28A, 679-681. 23. Li, L.; Lishko, V. K.; Hoffman, R. M. In Vitro Cell. Dev. Biol. 1993, 29A, 192-194. 24. Li, L.; Lishko, V. K.; Hoffman, R. M. In Vitro Cell. Dev. Biol. 1993, 29A, 258-260. 25. Du Plessis, J.; Egbaria, K.; Ramachandran, C.; Weiner, N. Antiviral Res. 1992, 18, 259-265. 26. Niemiec, S.; Ramachandran, C.; Weiner, N. Pharm. Res. 1995, 12, 1184-1188. 27. Mezei, M. In Drug Permeation Enhancement: Theory and Applications; Hsieh, D. S., Ed.; M. Dekker: Basel, 1994; pp 171-198.
JS960394L