Tretinoin-based formulations-influence of concentration and vehicles on skin penetration

Tretinoin is used in the management of acne and it is part of a gold standard treatment for photoaging. It has also been reported as an agent for superficial chemical peeling in highly concentrated formulations with few considerations about skin penetration. The aim of this study was to evaluate the influence of drug concentration and vehicles currently used on skin penetration of tretinoin. In vitro permeation tests were carried out using Franz diffusion cells fitted with porcine ear skin and 10% aqueous methanol in the receptor compartment. Formulations studied, cream or hydroalcoholic dispersion, containing 0.25%, 1% and 5% of tretinoin were placed in the donor compartment for six hours. Tretinoin concentration in skin layers was measured by high performance liquid chromatography. The largest amount of tretinoin from both vehicles was detected in stratum corneum with significant differences among the three concentrations. The hydroalcoholic dispersion was the best vehicle. Significant amounts of tretinoin were found even in deep layers of epidermis. The formulation with 0.25% tretinoin showed better results when considered the amount of tretinoin on skin in terms of percentage. Finally, skin penetration of tretinoin was influenced by vehicle and concentration of this drug used in formulation.


INTRODUCTION
Tretinoin (TRE) or all-trans-retinoic acid (RA) is indicated for treatment of acne vulgaris, melasma, photoaging, follicular keratosis, striae and disorders of keratinization (Thiboutot, Gollmick, 2009).It is well known that retinoids act by binding to specific nuclear receptors, interfering in gene expression and affecting cellular proliferation and differentiation, sebum production, inflammation, immunological reactions and possibly in non-melanoma skin cancer prevention (Stratigos, Katsambas, 2005).
The nuclear retinoic acid receptor family (RARs e RXR) is activated by TRE.By interaction with RARs and RXR, TRE executes its effect by one or more mechanism(s): a) epidermal thickening due to initiation of epidermal proliferation; b) compaction of the stratum corneum; and c) biosynthesis and deposition of the glycosoaminoglycans and collagen (Raza et al., 2013).
Chemical peelings consist of topical applications of caustic substances, which result in the loss and regeneration of superficial skin layers.Classical agents are: Jessner' solution, 70% glycolic acid, 30% salicylic acid and 10 to 35% tricloroacetic acid.Their use is indicated as adjuvant treatments for various dermatosis such as photoaging, acne, melasma, hyperpigmentations, scars among others (Roy, 2005).Sequential superficial chemical peelings may result in epidermis and papillary dermis reorganization (Imayama, Ueda, Isoda, 2000).
The application of high concentration TRE formulations at the beginning of treatment in self usage or in sequential superficial peelings could be useful to produce retinoid benefits in a shorter period of time.However, to obtain these benefits it is necessary that topical drugs penetrate through follicular and sweat openings, intercellular spaces, stratum corneum (SC) or keratinocytes (Wester, Maibach, 1994).Human skin is a highly effective barrier against penetration of chemical and other agents.Thus, skin penetration depends on SC thickness, integrity and lipid composition, number of follicular openings, drug concentration and formulation vehicle (Manconi et al., 2006;Contrera et al., 2005).
For this, various in vivo and in vitro methods have been proposed to evaluate skin penetration.Among these methods, we used the Franz cell method to shed some light on the possible mechanism of skin penetration by formulations under study.
Considering that skin penetration could influence the cutaneous effects of the tretinoin-based formulations, the aim of this study was to evaluate the skin penetration of TRE vehiculated in two vehicles currently used as in three different concentrations.

Material
All material used in this study were obtained from Galena (São Paulo, Brazil).Different TRE concentrations (0.25%, 1%, and 5%) were used as cream (stable emulsion composed of cetearyl alcohol, sodium cetearyl sulfate, decyl oleate, phenoxyethanol, parabens, glycerin and distilled water) and as hydroalcoholic dispersion (equal parts of distilled water and ethanol).The experimental skin was porcine ear and analytical grade 10% methanol was used in the receptor compartment of the Franz diffusion cell (HPLC grade, Sigma-Aldrich, Germany).

Penetration tests in vitro
The receptor chamber of a Franz diffusion cell (Microette-Hanson Research, Chatsworth, CA, USA) was composed of six individual cells with porcine skin placed with dermis facing the methanol solution and 400 mg of TRE formulations were placed in the donor chamber.The entire system was kept at 37 ± 0.5 °C in a thermostated bath during the permeation test (6 hours).Removed skin samples were rinsed to eliminate excess formulation and fixed.The SC was separated from epidermis and dermis by 14 repeated tape strippings using adhesive strips (Scotch Book Tape, n.845; 3M, St. Paul, MN, USA).TRE in stripings and skin layers was extracted by 10% aqueous methanol and quantified by Bonhomme (1990) method in High Performance Liquid Chromatography (HPLC) as were the contents of the receptor chambers (SPD-10A, equipped with VIS-UV detector operating at 343 nm).TRE concentrations in SC and epidermis + dermis represent topical delivery, whereas transdermal delivery is defined by the amounts detected in the receptor chamber.

Validation of the analytical method
A dose equivalent to 5 mg of TRE was dissolved in 5 mL of methanol (1 mg/mL).Aliquots were diluted in methanol to obtain the following concentrations: 5, 10, 20, 25, 50, 60, 75 and 100 μg/mL, which were analyzed in triplicate by HPLC-UV (Bonhomme, 1990); results were inserted into a graph with three curves.The medium curve was in accordance to Brazilian Regulatory Agency (Agência Nacional de Vigilância Sanitária, ANVISA, Brazil).Linearity was evaluated through the coefficient of variation of replicate points and the coefficient of linear correlation (r 2 ).Accuracy was assessed by the coefficient of variation in determinations at each point of the calibration curve in triplicate during three consecutive days (inter-day) as well as in three points 10 times in one day (intra-day).Precision was calculated by variation between expected and observed concentrations at each point of the calibration curve when solutions were analyzed in triplicate for three consecutive days (inter-day) and 10 times in one day (intra-day).

Statistical analysis
Data analysis was carried out with the software OriginPro 7.0 (OriginLab Corporation, USA).Results are expressed by means ± standard deviation.Multiple comparisons of means (Tukey test) were used to substantiate statistical differences among each sample and skin layer (stratum corneum and epidermis + dermis).
Significance was tested at the 0.01 level of probability (p).The results for different vehicles were analyzed separately, i.e., one ANOVA for each of the two used considering three TRE concentrations (0.25, 1 and 5%).Data variation analysis was performed in a completely randomized design, considering two experiments analyzed separately (two vehicles), composed of three treatments (0.25, 1 and 5% TRE concentrations in the sample) x 4 replications per treatment x 2 skin layer.

RESULTS
In the present experimental study, the skin penetration of six different formulations was tested.The in vitro topical and transdermal drug delivery was evaluated by the Franz diffusion cell test, a standard method for these procedures (Manconi et al., 2006).TRE amounts in skin layers and receptor chamber were quantified by HPLC, which was used under strict conditions to separate all-trans from 13-cis retinoic acid and other impurities as it had already been recommended (Manconi et al., 2006).Separation of trans isomer was satisfactory but over the time a slight degradation of TRE produced another compound, which was not identified.
The linear correlation coefficient determined in the TRE standard curve analysis by HPLC-UV was 0.998 which complied with limits established by ANVISA, Brazil, i.e., ≥ 0.99 32 .Accuracy and precision values showed coefficients of variation ≤ 4%.
The greatest amounts of TRE were retained in SC regardless of the concentration in cream formulations.However, it was significantly higher when the 5% concentration was used (p<0.01).The average amounts of TRE delivered into epidermis + dermis was not significantly different for 0.25%, 1% and 5% concentrations (Table I).
The amount of TRE, which permeated epidermis + dermis and reached the receptor compartment was lower than the method quantification limit.
The average amounts of TRE in hydroalcoholic dispersion retained in skin layers was significantly different (p<0.01) and much higher in SC than epidermis + dermis when comparing the three concentrations of drug used (Table II).
Comparing vehicles with the same concentrations of TRE the hydroalcoholic dispersion allowed higher penetration (Figure 1).
The cost-to-benefit ratio of each formulation was calculated in terms of percentage.The amount of tretinoin in each skin layer was divided by the total amount of tretinoin in each formulation.The results were expressed as a percentage of the total amount in the formulation.Among cream formulations, 0.25% showed the best relation in stratum corneum and epidermis + dermis (Figure 1).However, for hydroalcoholic dispersions, 0.25% TRE formulation showed only the best cost-tobenefit ratio in epidermis + dermis (Figure 1).

DISCUSSION
The first proposal about the use of highly concentrated TRE as an agent for superficial peeling was reported by a Brazilian group in 2001 (Cucé et al., 2001).The authors sequencially used 1, 2, 3, 4 and 5% TRE concentrations in a solution with equal parts of ethanol and propilenoglycol.They suggested two applications per week, in five sessions and reported no discomfort other that the remaining solution should be washed out after 6 to 12 hours, at home.One year later, they recommended the use of 5% TRE peeling, once a week, in three applications (Cucé, Bertino, 2002).At that time, a letter was sent to the journal questioning stability and the advantage of such a high concentration (Kligman, 2001).
Retinoids are the most effective comedolytic agents.They provide multiple benefits for acne treatment such as: normalizing desquamation of the follicular epithelium, preventing formation of new microcomedo and minimizing the formation of comedones and inflammatory lesions (Manela-Azulay, Bagatin, 2009).
However, they cause irritative dermatitis during the early phase of skin retinization, which may lead to discontinuation of treatment (Kang et al., 2005).Thus, considering that the tretinoin penetration on the skin could influence the efficacy of the tretinoin-based formulation, this study can contribute for the definition of the best TRE concentration and vehicle in order to obtain the best results and a rational use of this active ingredient, as well as, avoid undesirable effects.
The skin is the largest organ and, as our primary external barrier, it is on the forefront of the battle with external causes of damaging free radicals.The most significant damage by free radicals is to biomembranes and to DNA.It is thought that topical use of vitamins and antioxidants in cosmetics and skin care products can better protect and possibly correct the damage by neutralizing these free radicals.In addition, some vitamins may be beneficial to the skin because of other actions such as suppression of pigmentation and bruising, stimulation of collagen production, refinement of keratinization, or anti-inflammatory effects (Lupo, 2001).In this sense, the highest amount of TRE found in stratum corneum likely indicates the effectiveness of TRE just for exfoliation.
The influence of concentration and vehicle in retinol formulation stability and skin penetration has already been demonstrated in vitro (Gomes, Gaspar, Maia Campos, 2007;Bagatin et al., 2010).The search for an acceptable delivery system for all-trans retinoic acid with no risk of systemic absorption and side effects, using mice abdominal skin as an experimental model, demonstrated that the encapsulated form (multilamellar large vesicles) promoted retention on skin superficial layers and prolonged drug release (Gomes, Gaspar, Maia Campos, 2007).
Skin penetration of vitamin A (retinol) in two vehicles, self-nanoemulsifying system and classic gel cream, using Franz cells method was evaluated (Gianetti et al., 2012) The authors found that both formulations presented the same particle size and the gel cream classic formulation furnished better penetration in both stratum corneum and epidermis + dermis.

FIGURE 1 -
FIGURE 1 -Percentages of TRE in skin layers delivered from three concentrations of the drug in cream and hydroalcoholic dispersion formulations.

TABLE I -
Tretinoin average concentrations retained in skin layers (μg/cm 2 ) for the three cream formulations *Means followed by the same letter are not statiscally different (p<0.01).

TABLE II -
Tretinoin average concentrations retained in skin layers (μg/cm 2 ) for the three hydroalcoholic formulations

Tretinoin concentration in hydroalcoholic formulations(%) Tretinoin retained in the stratum corneum(μg/cm 2 )
Means followed by the same letter are not statiscally different (p<0.01).