Testosterone Ethosomes for Enhanced Transdermal Delivery

Abstract

Physiological decrease in testosterone levels in men with age causes various changes with clinical significance. Recent testosterone replacement therapy is based mainly on transdermal nonpatch delivery systems. These products have the drawback of application on extremely large areas to achieve required hormone blood levels. The objective of the present study was to design and test a testosterone nonpatch formulation using ethosomes for enhanced transdermal absorption. The ethosomal formulation was characterized by transmission electron microscopy and dynamic light scattering for structure and size distribution and by ultracentrifugation for entrapment capacity. To evaluate the feasibility of this delivery system to enhance testosterone permeation through the skin, first the systemic absorption in rats was compared with a currently used gel (AndroGel®). Further, theoretical estimation of testosterone blood concentration following ethosomal application in men was made. For this purpose, in vitro permeation experiments through human skin were performed to establish testosterone skin permeation values. In the design of these experiments, testosterone solubility in various solutions was measured and the effect of the receiver medium on the skin barrier function was assessed by confocal laser scanning microscopy. Theoretical estimation shows that testosterone human plasma concentration value in the upper part of the physiological range could be achieved by application of the ethosomal formulation on an area of 40 cm². This area is about 10 times smaller than required with current nonpatch formulations. Our work shows that the ethosomal formulation could enhance testosterone systemic absorption and also be used for designing new products that could solve the weaknesses of the current testosterone replacement therapies.

Keywords :

• Delivery System
• Ethosome
• Testosterone
• Transdermal

Testosterone, the principal androgen in the human body, is responsible for the growth and maturation of male sex organs, and maintenance of such secondary sex characteristics, as male hair distribution, vocal cord thickening, and alterations in body musculature, and fat distribution. Normal testosterone blood levels in young male are 300–1000 ng/dL and show circadian variation with peak concentration in the morning (Bremner, Vitiello, and Prinz 1983; Basaria and Dobs 2001). The main indication for testosterone replacement therapy is primary or secondary hypogonadism, caused by failure of the testis function or at the hypothalamus and anterior pituitary level (Bagatell and Bremner 1996).

During the last decade, a new hypogonadal population has been recognized. The increasing life span has resulted in a steady rise in the proportion of older men who now make up the bulk of hypogonadal population, as the aging process leads to the physiological lowering of testosterone (Tan and Culberson 2003; Yialamas and Hayes 2003). Testosterone levels in men start declining in the third decade of life, with the lowest levels ∼50% of the normal value, seen in men 70 years of age and older. There is an increasing understanding that this decrease in the testosterone level has clinical significance. It reduces the quality of life, changes the body composition by decreasing the muscle mass and increasing the fat mass, causes sexual dysfunction, osteoporosis, mood changes, and many other complications (Lunenfeld 2003).

The goal of testosterone replacement therapy, according to the WHO/NIH/FDA consensus of 1990, is to replace testosterone levels at as close to physiological concentrations as possible. The existing androgen replacement therapy is based on oral, intramuscular, and transdermal preparations. With the exception of transdermal testosterone application, none of the testosterone preparations available can fully achieve this goal. Following oral testosterone ingestion, the hormone undergoes first-pass metabolism in the liver (Zitzmann and Nieschlag 2000). Intramuscular injections often yield supraphysiological and subphysiological testosterone levels (Dobs et al. 1999).

The transdermal delivery of testosterone is the most efficient and convenient way of administration, because it can bypass the first-pass metabolism in the liver, is able to deliver physiological drug blood levels, and mimics a circadian pattern close to that of healthy men (Bagatell and Bremner 1996; Dobs et al. 1999). Three different application systems are available: scrotal patch, nongenital patch, and gel. Transdermal patches have several drawbacks: they often cause local skin irritation (Jordan 1997); application of scrotal patches requires adequate scrotum size that can be insufficient in patients with early onset of hypogonadism (Meikle et al. 1992); and they produce supraphysiological levels of dihydrotestosterone, which can cause enlargement in the size of the prostate and other side effects. Recently, new testosterone-containing transdermal gels were developed. These gels cause only minimal skin irritation, but must be applied over a large surface area of the skin, and thus they may not be acceptable to all patients (Steidle et al. 2003; Marbury et al. 2003; Gooren and Bunck 2003).

To achieve better testosterone skin permeation and thus higher testosterone blood levels, following smaller application surface area on the skin, an efficient permeation enhancement should be used. Ethosomal carriers, composed mainly of phospholipid, relatively high concentrations of ethanol and water, were developed in our laboratory. In these carriers, ethanol with its fluidizing effect on the phospholipids bilayers contributes to the creation of vesicles with a soft malleable structure. Our previous work shows that these carriers are effective in enhancing permeation of hydrophilic and lipophylic molecules through the skin into systemic circulation (Touitou et al. 2000a, 2000b; Godin and Touitou 2003).

A transdermal testosterone ethosomal patch was designed and tested in our laboratory versus testosterone commercial patch-Testoderm™ (Alza, USA). The testosterone transdermal permeation from ethosomal patch was significantly higher compared with the commercial patch (Touitou et al. 2000a). The aim of the present study was to design and test for skin permeation an ethosomal testosterone nonpatch formulation with enhanced delivery to allow application on a smaller skin area.

MATERIALS AND METHODS

Soybean phosphatidylcholine (Phospholipon 90) was a gift from Natterman Phospholipid (Germany). AndroGel (Unimed, USA), a 1% testosterone gel containing 68.9% ethanol, was purchased. Testosterone was acquired from Fluka (Israel), fluorescein sodium from Sigma (Israel), ethanol (EtOH) from Frutarom (Israel), and propylene glycol from Merck (Israel). Carbopol 934 was a gift from Trima (Israel).

Preparation and Physical Characterization of Ethosomal Systems

The main components of testosterone ethosomes (Touitou 1996) were testosterone 1%, alcohol 50%, propylene glycol, Phospholipon 90, and water. The preparation method was previously described (Touitou et al. 2000a). Briefly, Phospholipon 90 and testosterone were dissolved in alcohol and propylene glycol. The aqueous phase was added in small increments with constant mixing at room temperature.

Ethosomal systems were characterized by dynamic light scattering (DLS), transmission electron microscopy (TEM) and ultracentrifugation. The size distribution of vesicles was determined by DLS using a computer inspection system (Malvern Zetamaster ZEM 5002, Malvern, UK) or ALV-NIBS/HPPS-high performance particle sizer (ALV-GMBH Langen, Germany). Before each measurement, the vesicular systems were gently mixed with appropriate medium. The measurements were conducted in triplicate, of 30 sec each.

Ethosomal vesicles were visualized using TEM by negative staining method. Ethosomal solution was applied on a microscopic carbon-coated grid and stained with 1% aqueous solution of PTA for 30 sec. After drying, the specimen was viewed under Philips TEM CM 12 electron microscope (Eindhoven, The Netherlands) at a 10–100k-fold enlargement.

The entrapment ability of ethosomes was measured by ultracentrifugation. Ethosomal preparation was kept overnight at 4°C. Then 1 ml of the ethosomal solution was spun in a TL-100 ultracentrifuge (Beckman, USA), in a TLA-45 rotor, at 4°C, 40 000 rpm, for 3 hr. After centrifugation, the supernatant was removed and the amount of testosterone in it was determined by HPLC. The entrapment capacity was calculated as follows: [(T-C)/T]*100, where T is the total amount of drug in 1 ml of ethosomal solution, and C is the amount of drug detected in the supernatant.

Delivery of Testosterone in Rats and Measurement of Plasma Concentration

In vivo experiments of testosterone transdermal administration were carried out in 12 male Sprague-Dawley rats (240–270 g), divided into 3 groups: 1 control group and 2 treatment groups. The animals were treated in accordance with the institutional guidelines by a protocol approved by the institutional animal care ethical committee. After acclimation, the rats were castrated under anesthesia with 85% Ketaset and 15% Rompun (20 mg/ml). A 2-week recovery period was allowed for baseline testosterone to drop to undetectable levels.

On the day prior to the experiment, blood samples (300 μl) were withdrawn to examine the baseline levels of testosterone in these rats and to evaluate the efficiency of the castration. The dorsal rat skin was shaved using Oster electric clippers. An application area of 24 cm² was marked on the skin.

On the day of experiment, the animals were anesthetized as described above for 1 hr. The formulations (400 mg) were applied to the marked area on the animals back and covered by semiocclusive dressing in order to mimic the application of current transdermal nonpatch delivery systems in humans. The application area on the animal's dorsal skin was examined for irritation, at the beginning and at the end of the experiment.

Blood samples of 300 μ l were taken at 2, 4, 8, 12, and 24 hr after application via the retro-orbital sinus. The samples were collected into nonheparinized vials and centrifuged for serum separation. The separated serum was frozen at −20°C until analysis. The drug concentration was determined by radio-immuno assay (RIA). On the basis of testosterone plasma concentrations, the following pharmacokinetic parameters were obtained:

• Cmax is the peak plasma concentration during the dosing period.

• AUC is the area under the plasma concentration-time curve calculated by the trapezoidal rule for a period from 0 to 24 hr.

Testosterone Solubility Measurements

Testosterone solubility was measured in various hydroethanolic solutions, ethanol phosphate buffer (Samour, Krauser, and Gyurik 1999) and aqueous solutions of Brij 98, at 37°C and room temperature. Ten mg of testosterone were added to solutions, following constant mixing with Orbit shaker (Lab-Line Melrose Park, PA, USA) until stable sediment was obtained (at least for 48 hr). Samples were withdrawn from the supernatant and centrifuged 10 min at 5000 rpm (Hermle Z 160 M, Hermle Labortechnik, Herteller Spintron Inc). Testosterone concentration in the solution was determined by ultraviolet spectrophotometer (Uvikon 933, Kontron Instruments) at 242 nm.

Effect of Ethanol in the Receiver Compartment

Effect of ethanol in the receiver compartment on the depth of skin penetration was evaluated using fluorescein sodium (FS) 0.03% aqueous solution. First, 100 μl of FS solution were applied for 24 hr to dermatomed human cadaver skin in Franz diffusion cells at 37°C. The receiver medium contained either water or 50% (v/v) hydroethanolic solution. One skin served as control for skin autofluorescence (without fluorescent probe).

Then, after 24 hr, excess FS was cleaned from the skin surface with distilled water. Skin specimens were analyzed by confocal laser scanning microscopy (CLSM). Fluorescence intensity was analyzed at 10 μm increments through the z-axis using a Sarastro Phoibos 1000 confocal laser scanning microscope (Molecular Dynamics, Sunnyvale, CA, USA) attached to a universal Zeiss epifluorescence microscope with an oil-immersed plan apo 63 × 1.4 NA objective lens. Optical excitation was carried out with a 488 nm argon laser beam. The fluorescence intensity of the obtained images was quantified by Image Pro-Plus Software.

In Vitro Permeation of Testosterone Through Human Skin

Testosterone skin permeation experiments were carried out nonocclusively in Franz diffusion cells for 24 hr. For skin permeation profiles of testosterone, the ethosomal formulation and AndroGel® (100 mg) were applied on the stratum corneum side of cadaver dermatomed defrosted human skin. The receiver compartment contained 50% hydroethanolic solution. Samples were withdrawn at predetermined times during the experiment—1,2,4,5,7,16,18,20,22, and 24 hr. The sample volume was 200 μl. Testosterone concentration in the receiver samples was determined by HPLC.

Assays

Testosterone concentration in the receiver compartment from the in vitro experiments and in the supernatant following ultracentrifugation was quantified by reverse-phase HPLC using a Merck Hitachi D-7000 interface equipped with a L-7400 variable ultraviolet detector, L-7300 column oven, L-7200 auto sampler, L-7100 pump, and HSM computerized analysis program. Separations were carried out using LiChrospher® 125-4 100 RP-18 (5 μm) column, with a mobile phase containing a 1:1 mixture of Acetonitrile: Na-acetate 0.01M pH 4.0 buffer, at a flow-rate of 1.0 ml/min. Testosterone was detected at 242 nm (Figure 1).

FIG. 1. HPLC chromatogram of 25 μg/ml testosterone in ethanol: water (1:1) solution.

Validation of the method was performed by limit of detection (LOD) and limit of quantification (LOQ) determination. For determination of LOD, 6 samples of 0.020 μg/ml of testosterone were injected. The average signal-to-noise ratio was found to be 22 ± 2 (> 3). Concentration of 0.020 μg/ml was set as the LOD concentration of testosterone. The LOQ was determined by injection of 6 samples of 0.050 μg/ml of testosterone. The average signal to noise ratio was found to be 56 ± 2 (> 10). Mean peak area was 5760 ± 205 μV*sec and relative standard deviation (RSD) was 4% (< 10%). The LOQ concentration of testosterone was set at 0.050 μg/ml. The amount of testosterone in the serum of rats was determined using Coat-A-Count total testosterone RIA kit (Diagnostic Products, California, USA).

Data Analysis

The cumulative amount of testosterone that permeated the skin during in vitro experiments was calculated using the Transderm software (Touitou and Wartenfeld 1987). This software enables rapid calculation of kinetic parameters such as cumulative amount of drug permeated the skin, flux, Kp (skin permeability constant), and lag time. Statistical analysis was performed using an InStat program by unpaired nonparametric two-tailed Mann-Whitney test.

RESULTS

In this work, we designed and tested a testosterone ethosomal formulation for enhanced transdermal permeation. For this purpose, the delivery enhancing property of ethosomes was first tested for testosterone systemic absorption in rats. Further, the C_ss of testosterone in human plasma following application of testosterone ethosomal formulation was estimated by using data obtained in in vitro permeation tests through human skin. DLS analysis of the size distribution of testosterone ethosomes shows one narrow peak confirming the homogenous population of the vesicular size (Figure 2). The radius of the ethosomes was 75.6 ± 0.2 nm.

FIG. 2. Size distribution of ethosomal vesicles containing 1% testosterone measured by DLS.

Visualization by negative-stain TEM (Figure 3) demonstrates that testosterone ethosomal system contains multilamellar vesicles of homogenous size. The lamellas were extended to the core of the ethosome. The entrapment ability of the ethosomes was determined by ultracentrifugation and found to be 61%.

FIG. 3. TE micrograph of a testosterone ethosomal vesicle.

To evaluate the efficiency of the ethosomal formulation to enhance testosterone permeation through the skin, testosterone systemic absorption from this formulation was compared with AndroGel®, following a single dermal application of 400 mg formulation (containing 4 mg of testosterone) to rat shaved dorsal skin. The results of blood samples show that Cmax and AUC values for the ethosomal system were significantly higher (p < 0.05) as compared with AndroGel (1970 ± 251 ng/dL and 601 ± 88 ng/dL for the Cmax, respectively, and 9313 ± 385 ng*dL^{− 1}*h and 5678 ± 719 ng*dL^{− 1}*h for the AUC, respectively). Examination of the skin in control versus treated rats (ethosomal system application) after 24 hr of the experiment showed no visible differences in the color and appearance of the application area. Recent studies evaluated the human skin tolerability in vivo of ethosomal formulations containing 2% Phospholipon and 45% ethanol versus 45% hydroethanolic solution and saline solution. The variation of erythema index following ethosomal system application was comparable to normal saline, while application of hydroethanolic solution caused a remarkable skin erythema (Paolino et al. 2004).

Further, estimation of testosterone blood concentrations in men following application of the ethosomal nonpatch formulation was performed by measurement of testosterone flux through human skin based on in vitro testosterone permeation experiments. Testosterone is a lipophylic molecule; therefore, it is very important to select an appropriate receiver solution for in vitro permeation experiments, which will permit both the maintenance of sink conditions throughout the experiment and allow drug clearance from the skin dermis. Testosterone solubility in various solutions was evaluated at room temperature (20°C) and at 37°C, the temperature at which in vitro permeation experiments were performed.

The results presented in Table 1 indicate that the highest values of testosterone solubility were received in ethanol containing solutions. Testosterone solubility increased significantly with the increase in ethanol concentration. At room temperature (20°C), the solubility was 0.248 mg/ml in 20% (v/v) hydroethanolic solution and 6.763 mg/ml in 50% (v/v) hydroethanolic solution. Addition of Brij 98 to water did not increase the solubility of testosterone. Solubility of testosterone in 20% (v/v) ethanol in phosphate buffer was similar to the solubility in 20% (v/v) hydroethanolic solution. Based on testosterone solubility results, a hydroethanolic solution was selected as receiver medium. However, it was important to show that no artifacts were introduced by using the hydroethanolic solution as a receiver in permeation experiments. Ethanol is known for its ability to fluidize lipids in the stratum corneum of the skin and therefore could enhance skin penetration of molecules (Williams and Barry 2004).

TABLE 1 Solubility of testosterone in various receiver solutions

Solvent	Solubility at 20°C (mg/ml)	Solubility at 37°C (mg/ml)
Water	Practically insoluble (< 0.1)	—
20% hydroethanolic solution	0.248	0.344
20% ethanol in phosphate buffer	0.224	0.303
50% hydroethanolic solution	6.763	10.766
0.5% Brij 98 aqueous solution	0.057	0.084
6.0% Brij 98 aqueous solution	0.169	0.194

*Reynolds, J. 1996. Martindale The Extra Pharmacopoeia, 31st ed. London: Royal Pharmaceutical Society.

We evaluated the effect of receiver solution composition on the barrier properties of the skin. A CLSM experiment was carried out. FS was chosen to serve as a probe in this experiment due to its good water solubility. The use of a water solution of FS in the donor compartment was made to isolate the effect of ethanol in the receiver compartment on skin permeation. The results of FS permeation into human cadaver skin with receiver solutions containing water or 50% (v/v) hydroethanolic solution are presented in Figure 4.

FIG. 4. CLS micrograph skin optical sections showing fluorescein sodium penetration from aqueous solution into human skin. Receiver compartment containing (A) 50% hydroethanolic solution or (B) water.

CLS micrographs show that the depth of penetration of the fluorescent probe into the skin was similar in these different experimental conditions and independent of receiver solution composition. The fluorescence intensity of the obtained images was calculated for each skin sample depth. The results were plotted as fluorescence intensity arbitrary units (AU) versus skin depths. The area under the curve was used for evaluation of the overall intensity of FS penetration into the skin. The obtained values show that a similar fluorescence intensity was measured for 50% ethanol and water in the receiver compartments (14054 ± 264 AU*μ m and 16580 ± 192 AU*μ m, respectively).

Our results point out that ethanol in the receiver compartment has no effect on the penetration of fluorescent probe into human skin, and therefore it can be assumed that it does not influence the barrier properties of the skin. Based on these data, a 50% hydroethanolic solution was selected as receiver solution in testosterone in vitro permeation experiments, due to its ability to maintain sink conditions and allow a good release of the lipophylic drug in the receiver medium. In following in vitro experiments, testosterone skin permeation from ethosomal formulation was tested versus AndroGel. The permeation experiments performed in Franz diffusion cells through human skin during a period of 24 hr show a different permeation profile for the ethosomal formulation compared with AndroGel (Figure 5).

FIG. 5. Testosterone skin permeation profiles in vitro through human skin from ethosomal formulation versus AndroGel (mean ± S.E.M.). Each donor contains 1% testosterone.

Testosterone flux, from ethosomal formulation through human skin, of 1.237 * 10^{− 2} mg*cm^{− 2}*hr^{− 1} was calculated using the Transderm program. The amount of testosterone that permeated the skin after 24-hr from ethosomes was higher than the amount from AndroGel, 594.57 ± 39.9 μg versus 92.27 ± 2.86 μg, respectively. The results show that ethosomal formulation enhanced transdermal permeation of testosterone by sixfold.

This enhancement in testosterone permeation through the skin following ethosomal application can be explained on the basis of a proposed model for skin delivery from the ethosomal systems. Ethanol has long been known to have permeation-enhancing properties: it can increase the solubility of drug in the vehicle, alter the solubility properties of the tissue, and even extract some of the lipid fraction from within the stratum corneum (Williams and Barry 2004). However, previous studies (Dayan and Touitou 2000; Touitou et al. 2000a) that compared enhancement permeation of drugs from ethosomal systems versus hydroethanolic solutions show that the permeation enhancement from ethosomes is much greater than would be expected from ethanol alone. This suggests some kind of synergistic mechanism between ethanol, vesicles, and skin lipids. In the proposed model for skin delivery from ethosomal system, ethanol disturbs the organization of the stratum corneum lipid bilayer and enhances its lipid fluidity. It also fluidizes the lipids that compose the ethosomes, thus producing soft vesicles that can then penetrate the disturbed stratum corneum (SC) bilayers. The release of drug in the deep layers of the skin and its transdermal absorption could then be the result of fusion of ethosomes with skin lipids and drug release at various points along the penetration pathway.

Theoretical Calculation of Testosterone Plasma Concentration

The normal testosterone plasma levels in men are 300–1000 ng/dL. Using the data received in in vitro permeation experiments through human skin, it is possible to predict the C_ss testosterone blood levels in men following testosterone ethosomal formulation application. The expected testosterone blood concentration (C_ss) can be calculated by using the following equation, if an application area of 40 cm² is targeted: where J_ss is the flux, 1.237*10^{− 2} mg*cm^{− 2}*hr^{− 1}; A is the application area, and CL is the total testosterone body clearance, 54.33 L/h (Xing et al. 1998). The calculated C_ss, according to the above data, is 911 ng/dL. This value lies well within the normal testosterone plasma range in men.

It is noteworthy that the transdermal testosterone gel available today (AndroGel) requires a large application area of the formulation on the skin (> 400 cm²). Therefore, there is a possibility of testosterone transfer between subjects. Moreover, application on such a large surface is inconvenient. These calculations show that testosterone ethosomal system, due to higher skin permeation enhancement, would allow the use of a smaller application area and reduced dose, and therefore better patient compliance.

Prof. E. Touitou also is affiliated with David R. Bloom Center of Pharmacy at the School of Pharmacy of The Hebrew University of Jerusalem. The results of this work were partially submitted by D. Ainbinder to The Hebrew University of Jerusalem in fulfillment of the M.Sc. requirements.