Evaluation of β-blocker gel and effect of dosing volume for topical delivery.

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Evaluation of ␤-Blocker Gel and Effect of Dosing Volume for Topical Delivery QIAN ZHANG,1 DOUNGDAW CHANTASART,2 S. KEVIN LI1 1 2

Division of Pharmaceutical Sciences, James L. Winkle College of Pharmacy, University of Cincinnati, Cincinnati, Ohio 45267 Department of Pharmacy, Faculty of Pharmacy, Mahidol University, Bangkok 10400, Thailand

Received 9 September 2014; revised 13 January 2015; accepted 20 January 2015 Published online 18 February 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.24390 ABSTRACT: Although topical administration of ␤-blockers is desired because of the improved therapeutic efficacy and reduced systemic adverse effects compared with systemic administration in the treatment of infantile hemangioma, the permeation of ␤-blockers across skin under finite dose conditions has not been systematically studied and an effective topical ␤-blocker formulation for skin application is not available. The present study evaluated the permeation of ␤-blockers propranolol, betaxolol, and timolol across human epidermal membrane (HEM) from a topical gel in Franz diffusion cells in vitro under various dosing conditions. The effects of occlusion and dosing volume on percutaneous absorption of ␤-blockers from the gel were studied. The permeation data were compared with those of finite dose diffusion theory. The results showed that skin permeation of ␤-blockers generally could be enhanced two to three times by skin occlusion. The cumulative amounts of ␤-blockers permeated across HEM increased with increasing dosing volume. An adequate fit was obtained between the theoretical curve and experimental permeation data, indicating that the experimental results of the gel are consistent with finite dose diffusion theory. In conclusion, the findings suggest the feasibility of using topical gels of ␤-blockers for infantile hemangioma treatment C 2015 Wiley Periodicals, Inc. and the American Pharmacists Association J Pharm and topical application with skin occlusion is preferred. Sci 104:1721–1731, 2015 Keywords: ␤-blocker; topical; percutaneous; hemangioma; finite dose; occlusion; skin; diffusion; partition coefficient; stability

INTRODUCTION Infantile hemangiomas are benign tumors in children occurring in 5%–10% infants. Most of the infantile hemangiomas resolve naturally without any need of treatment, but approximately 10% of the affected infants require therapeutic intervention because of complications.1,2 Systemic propranolol, an adrenergic $-blocker, has recently become the main treatment for infantile hemangioma.3–5 However, because of the concerns of potential side effects associated with systemic $-blockers, a safer and more convenient therapy is needed.6,7 Topical $blockers are a promising alternative in the treatment of skin hemangiomas that can improve therapeutic efficacy and reduce systemic adverse effects. Currently, topical skin formulations of $-blockers are not commercially available, and topical $-blocker treatments have been carried out using commercial topical eye drops of timolol, a $-blocker, that are not optimized for skin delivery.8,9 Topical gels are appealing in the development of a topical $-blocker product because they are convenient to use and have a broad range of applications in cosmetics, pharmaceuticals, and medicine.10,11 Because of their high water content, gels also allow easier migration of the drugs within the vehicle and hydrate the skin that can promote skin permeation compared with other semisolid formulations.12 Previously, skin permeation of four $-blockers, propranolol, betaxolol, timolol, and atenolol was investigated for their potential in topical delivery for the treatment of infantile hemangiomas using side-by-side diffusion cells in vitro under different pH conditions in the donor solution.13 The experiments in this Correspondence to: S. Kevin Li (Telephone: +513-558-0977; Fax: +513-5584372; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 104, 1721–1731 (2015) C 2015 Wiley Periodicals, Inc. and the American Pharmacists Association

previous study were performed under the infinite dose condition to evaluate the steady-state permeability coefficients of the skin for the $-blockers. The results in this study showed that the apparent permeability coefficients of the $-blockers across human skin increased with their lipophilicity and the pH of the donor solution. This suggests that the lipoidal pathway in the stratum corneum (SC) was the major skin transport mechanism for the $-blockers. However, the increase in the permeability coefficients for the $-blockers was less than the 10-fold increase per pH unit expected from the theoretical permeability coefficient versus pH relationship, indicating that the development of an alkaline topical dosage form for effective local delivery of the $-blockers might not be necessary. Although the previous study has provided important information on the possibility of topical treatment of infantile hemangioma using $-blockers, the experiments in the study were designed for mechanistic understanding of skin permeation of $-blockers and did not mimic the in vivo situations in clinical setting. Skin permeation of the $-blockers under finite dose conditions was not studied and topical skin formulations were not developed for the $-blockers. Infinite and finite doses are commonly used in skin permeation studies to evaluate percutaneous absorption of drugs in vitro.14 In infinite dose experiments, the permeant is often applied in a comparatively large dose in which steady-state permeation is achieved and the steady-state flux of the permeant can be determined using the linear region of a cumulative amount delivered versus time plot. Although the infinite dose experimental design is capable of providing scientific information such as steady-state permeability coefficients for mechanistic interpretation, the finite dose method can better resemble the in vivo situation in practice. In the finite dose experiments, the applied dose is usually limited and the depletion of the permeant in the donor chamber over time generally results in

Zhang, Chantasart, and Li, JOURNAL OF PHARMACEUTICAL SCIENCES 104:1721–1731, 2015

1721

1722


Table 1. Physicochemical Properties of the $-Adrenergic Blocking Agents $-Blocker

Molecular Weight (g/mol)

Propranolol Betaxolol Timolol

259.3 307.4 316.4

pKa 9.5 ± 9.4b 9.2c

1.2a

log Ko/w 3.3a 2.8b 2.1c

a

Values from the literature.16 Values from the literature.17 Values from the literature.18

b c

a nonlinear cumulative amount delivered versus time profile. Because of such differences, finite dose permeation data are different from those of the steady state obtained under the infinite dose condition.15 In general, the cumulative amount of a drug delivered across skin (or the flux) is related to the concentration of the drug in the dosage form (the driving force), and the same drug delivery profile is expected in permeation experiments at the same drug concentration in the donor chamber under the infinite dose condition. When different total amounts of the drug are applied on the skin in the donor chamber (i.e., at same concentration but different dosing volumes) under the finite dose condition, this will result in different extent of drug depletion and drug permeation profiles (i.e., cumulative amount delivered versus time or flux versus time profiles) leading to a dosing volume effect. The objectives of the present study were to (a) compare the permeation profiles of $-blockers in skin delivery from topical gels under occlusive and nonocclusive conditions at finite doses, (b) investigate the effect of dosing volumes of the gels on the permeation of the $-blockers, (c) evaluate skin permeation of the $-blockers under the finite dose condition based on a finite dose diffusion model, and (d) examine the stability of the $-blocker gels during storage. Three $-blockers propranolol, betaxolol, and timolol were selected as the model drugs in the present study. Table 1 summarizes the physicochemical properties of the $-blockers. Permeation experiments with human epidermal membrane (HEM) were performed in Franz diffusion cells in vitro with the $-blocker gels at different dosing volumes to mimic the in vivo setting in practice. The following questions for the topical gel formulations were to be addressed. (a) What are the amounts of $-blockers that can be delivered with the topical gel formulations under the finite dose conditions? (b) What are the effects of dosing volume and drug lipophilicity upon skin permeation of the $-blockers from the gels? (c) Does skin occlusion affect skin permeation of the $-blockers and what is the extent of this effect? The present study would address these questions and provide information on the feasibility of topical skin delivery of the $-blockers for potential use in the treatment of infantile hemangiomas. These findings could also assist in future development of a $-blocker dermatological gel formulation as well as the identification of an effective dosing protocol for topical $-blocker delivery.

EXPERIMENTAL Materials D,L-Propranolol

hydrochloride and betaxolol hydrochloride at purity 98% and 1-octanol at purity 99% were purchased from Sigma–Aldrich (St. Louis, Missouri). Timolol maleate

and hydroxyethyl cellulose (5000 cps) NF (HEC) were purchased from Letco Medical (Decatur, Alabama). Betaxolol hydrochloride ophthalmic solution (0.5%) was purchased from Falcon Pharmaceuticals (Fort Worth, Texas). Sodium azide (NaN3 ) was obtained from Acros Organics (Morris Plains, New Jersey). Methyl paraben and propyl paraben were obtained from Professional Compounding Centers of America (PCCA) (Houston, Texas). HPLC grade methanol and ethyl alcohol were purchased from Pharmaco-AAPER (Shelbyville, Kentucky). HPLC grade glacial acetic acid was obtained from EMD Chemicals (Gibbstown, New Jersey). Triethylamine, sodium hydroxide (NaOH), and hydrochloric acid (HCl) were purchased from Fisher Scientific (Pittsburgh, Pennsylvania). Phosphate-buffered saline (PBS), pH 7.4, consisting of 0.01 M phosphate buffer, 0.0027 M potassium chloride, and 0.137 M sodium chloride, was prepared by PBS tablets (MP Biomedicals, LLC, Solon, Ohio) and deionized water. PBS was preserved with 0.02% (w/v) NaN3 except when PBS was used in the $-blocker gel preparation. The pH of all the solutions was checked with a pH meter (Oakton Instruments, Vernon Hills, Illinois) and adjusted to pH 7.4 with 10% NaOH or 10% HCl when necessary. Preparation of HEM Excised split-thickness human cadaver skin from posterior torso of male aged between 19 and 69 years was obtained from the New York Firefighters Skin Bank (New York, New York). The thickness of split-thickness skin before heat separation was measured by a vernier caliper and found to be 0.05– 0.06 cm. HEM, composed of SC and viable epidermis, was separated from the dermis by heat separation.19 Briefly, the cadaver skin was immersed in PBS at 60°C for 1 min. After heat treatment, the dermis was gently peeled off from HEM by a pair of forceps under immersion in PBS. The HEM was then patted dry with Kimwipe, wrapped in aluminum foil, and stored in a freezer at –20°C for later use. Preparation of ␤-Blocker Gel Topical gels of 4 mg/mL propranolol hydrochloride, 5 mg/mL betaxolol hydrochloride, and 5 mg/mL timolol maleate (equivalent propranolol, betaxolol, and timolol concentrations of 3.5 mg/mL, 4.5 mg/mL, and 3.7 mg/mL, respectively) were prepared with HEC and PBS. Briefly, the drug solution was prepared by dissolving the required amount of the drug in PBS (without NaN3 ) in a vial. Then, an appropriate amount of HEC (2% w/v) was dispersed gradually in the drug solution. The dispersion was mixed using a magnetic stirrer until a clear transparent gel was obtained. The gel was then stored in a refrigerator (4°C) to provide maximum hydration and clarity and for later use. HEM Permeation Study The $-blocker gels were evaluated in skin permeation experiments using Franz diffusion cells. Prior to the permeation studies, the HEM samples were cut into the desired sizes (1.5 × 1.5 cm2 ) and allowed to hydrate in PBS at 4°C in a refrigerator overnight. Each fully hydrated HEM sample was supported by a Millipore filter (0.45 :m nitrocellulose) and mounted on the Franz diffusion cells with a rubber gasket placed between the SC and donor chamber to provide better sealing of the edges between the diffusion half cells. The diffusion cell setup provided


DOI 10.1002/jps.24390


an effective diffusional area of around 1.4 cm2 . The receiver and donor chambers were then filled with 5 and 0.5 mL PBS, respectively. The diffusion cells were placed in a thermostated heating and stirring module that maintained the receiver chamber at 37 ± 1°C and the donor chamber was exposed to room temperature, this resulting in donor chamber temperature of 30 ± 2°C. The integrity of HEM was checked before each permeation experiment using membrane electrical resistance. In the integrity prescreening, the electrical resistance of HEM was measured with an electrical system constructed using a 1.5-V battery and a fixed resistor. The system applied a low voltage (95% amount of drug in the octanol. The octanol/PBS partition coefficient (Ko/w ) was calculated by the ratio of the concentrations of $-blockers in equilibrium in the octanol and PBS phases: K o/w

Coctanol 1 + 10pKa−pH = CPBS

(3)

where CPBS is the drug concentration in the aqueous phase, Coctanol is the drug concentration in the octanol phase, and pH is the pH of the aqueous phase in the octanol/PBS partition experiments (i.e., pH 7.4). Experiments for each $-blocker were carried out in triplicate (n = 3). Stability Test of ␤-Blocker Gel Topical gels of 5 mg/mL propranolol hydrochloride, betaxolol hydrochloride, and timolol maleate (equivalent propranolol, betaxolol, and timolol concentrations of 4.4, 4.5, and 3.7 mg/mL, respectively) were prepared with HEC and PBS as described in the section “Preparation of $-Blocker Gel” with the following exceptions. For the propranolol hydrochloride gel, a mixture of methyl paraben and propyl paraben in propylene glycol was added to the propranolol solution to produce final concentrations of 0.04% and 0.01% (w/w) methyl paraben and propyl paraben and 0.5% (w/w) propylene glycol, respectively, as preservatives before mixing the propranolol solution with HEC, and sterile water (Baxter Healthcare, Deerfield, Illinois) instead of deionized water was used. For the preparation of timolol maleate gel, timolol maleate ophthalmic solution 0.5% (Falcon Pharmaceuticals) instead of timolol maleate powder (Letco Medical) was used. The final gels of propranolol hydrochloride, betaxolol hydrochloride, and timolol maleate were divided into 1 mL portions and stored in plastic syringes (containers) either at room temperature (20–25°C) or in the refrigerator (4°C) for 9 months after the preparation. The syringes at room temperature were exposed to fluorescent room light during daytime storage. At each sampling time point (0 day, 7 days, 3 months, 6 months, and 9 months), duplicate samples of 0.1 mL were withdrawn from each syringe and collected into scintillation vials. The samples were inspected visually for microbial growth, mixed with 10 mL methanol, and sonicated for 15 min to obtain complete dissolution of the drugs from the vehicle. Each sample was then diluted five times with the HPLC mobile phase. The concentrations of $-blockers in the samples were determined by the HPLC assay. HPLC Analysis Propranolol, betaxolol, and timolol were assayed using a Shimadzu HPLC system (Shimadzu Scientific Instruments, Inc., Addison, Illinois) that consisted of two pumps (LC-20 AT), a variable wavelength UV absorbance detector (SPD-20A), and an autoinjector (SIL-20A) at room temperature. A MicrosorbMV100-5 C18 column was used (15 cm × 4.6 mm, 4.6 :m, Varian, Lake Forest, California) and the mobile phase was methanol/water/glacial acetic acid/triethylamine 500:500:3.5:1 by volume ratio. The flow rate was 1.0 mL/min, and the injection volume was 50 :L. The detection wavelengths for propranolol, betaxolol, and timolol were 295, 273, and 295 nm, respectively.

Standard solutions of 0.15–50 :g/mL for propranolol hydrochloride, 0.36–60 :g/mL for betaxolol hydrochloride, and 0.25– 65 :g/mL for timolol maleate were prepared in the mobile phase to construct the calibration curves. The concentration was determined based on peak area measurement. Statistical Analysis In the analyses of the experimental data, the mean ± SD of the data was determined. Student’s t-test and ANOVA were performed using Microsoft Excel (Redmond, Washington) and GraphPad Prism (La Jolla, California), respectively. Differences were considered to be statistically significant at p < 0.05. Finite Dose Permeation Model To analyze the relationship between dosing volumes and changes in fluxes during topical drug delivery, a finite dose diffusion model for skin permeation was employed in the analysis of the permeation data.22 In this model, the flux (J) of a drug leaving the skin at time t can be described as: J(h, t) = Jss × 2

∞ n=1

2 "2n −"nDt · exp cos "n($ + $2 + "2n) h2

(4)

where D is the diffusion coefficient in skin and Jss is the steadystate flux (see Eq. (11)). The eigenvalues, "n , are roots of the transcendental equation: "n tan "n = $

(5)

and $ is a dose-related parameter: $ = K mv

h hv

(6)

where Kmv is the partition coefficient at the membrane–vehicle interface, h is the thickness of the membrane, and hv is the thickness of the gel applied in the donor chamber. For an ionapp izable drug such as a weak base, Kmv is replaced by K mv in the model (see Eq. (9)). The thickness of the gel is related to the dosing volume (V) and the effective diffusion area in the transport experiments: hv =

V A

(7)

Equation (4) assumes negligible transport barrier in the viable epidermis and within the gel reservoir relative to the SC barrier (i.e., the SC is the only major barrier). Previous permeability data of the viable epidermis for the $-blockers13 and release experiments with the gel in Franz diffusion cells without HEM (unpublished data) are consistent with this assumption. Sink condition is also assumed in the receiver chamber in the derivation of Eq. (4), which is a reasonable assumption as the highest drug concentrations found in the receiver for the $-blockers in the present experiments were 70 :g/mL or less than 2% drug concentrations in the donor chamber. Computer Model Simulation The flux of finite dose transport was simulated using COMSOL Multiphysics software (version 4.3, Comsol, Inc., Burlington, Massachusetts) with a time-dependent diffusion


DOI 10.1002/jps.24390


1725

gels behave similarly as an aqueous solution. In the heterogeneous membrane model, the model from Hansen et al.24 was used to calculate the partition coefficient Kmv that describes the SC lipid compartment as: K mv = Figure 1. Schematic diagram of the gel in the donor diffusion chamber and the membrane represented in one-dimensional geometry (center arrows) of Comsol computer simulation.

model and parameters such as diffusion coefficient, dosing volume, and membrane thickness as the inputs. The software provided a convenient platform to model finite dose drug delivery as an alternative to Eqs. (4) and (5) and allowed analyses taking into account of the effect of drug diffusion within the gel. The membrane parameters in the model were calculated as described in the next section. In the computer simulation, the concentration profiles in the membrane at different time points were generated within a one-dimensional geometry representing the thickness (length) of the gel in the diffusion cell donor chamber and the length of the permeation pathway in HEM as shown in Figure 1. The fluxes of the drug across the membrane/receiver interface over time were calculated. The transient flux versus time profile was the output under conditions of different $ values. These flux profiles were then analyzed and compared with the experimental data of the $-blockers determined in the present study.

> Dlip T · g K o/w Dw

(10)

where T is the weight fraction of the lipids (T = 0.15) and g and > are parameters obtained previously using a linear regression of partition coefficients into SC lipids versus octanol/water partition coefficients (g = 1.32 and > = 0.67), and Dlip and Dw are the densities of the lipid and water phases, respectively. The apparent distribution coefficient of the drug was then calculated using Kmv and Eq. (9). In both models, the SC thickness parameter h was set as a variable ranging from 0.002 to 1.0 cm that would fit the experimental data in the computer simulation. This thickness parameter was then compared with the length of the tortuous permeation pathway related to the SC intercellular lipids and the thickness of SC. The thickness of the vehicle in the donor chamber, hv ,was determined from the dosing volume applied in the experiments. The $ value was then calculated using Eq. (6) for each set of membrane thickness and hv . For the membrane diffusion coefficient parameters, the diffusion coefficients were calculated from the permeability app coefficients, membrane thickness, and K mv values: app

J ss = PCD =

DK mv CD h

(11)

Data Analysis Using Computer Simulated Results For the comparison of the experimental flux and computer simulation data, the $ and Kmv values of the experiments were determined as follows. Two sets of parameters representing two skin transport models were used: (a) a homogenous uniform membrane for simplicity and (b) a heterogeneous membrane with the intercellular lipids as a single tortuous permeation pathway and impermeable corneocytes. For Kmv under the assumption of a uniform membrane, the Ko/w values of the $-blockers were first determined in the octanol/PBS partition experiments at pH 7.4. Using the octanol/PBS partition coefficients, the partition coefficients of the $-blockers for skin permeation were calculated using the skin permeation model of Potts and Guy23 : K mv = [K o/w ]f

(8)

where f is a coefficient (f = 0.71) that accounts for the differences between the lipophilic environment as well as the anisotropic nature of octanol and the lipid domains in the SC. Because the $-blockers investigated in the present study are weak bases, Kmv in Eq. (8) was further corrected to account for the fraction of ionized $-blockers in the aqueous phase: K app mv = K mv · app

1 1 + 10pKa−pH

(9)

where K mv is the apparent distribution coefficient of the drug from aqueous solution to skin and the second term on the righthand side denotes the fraction of unionized drugs (funion ) in the aqueous solution. The apparent distribution coefficient of the drug was used here under the assumption that the $-blocker DOI 10.1002/jps.24390

where P is the permeability coefficient, and the steady-state fluxes of the $-blockers across skin and their permeability coefficients were obtained from a previous study.13 The transient fluxes of the $-blockers in the present study in the form of ln (J/Jss ) were plotted against Dt/h2 . The negative values of the slopes of these ln (J/Jss ) versus Dt/h2 profiles were then plotted against the $ values in the analysis.

RESULTS AND DISCUSSION Effect of Occlusion on Topical ␤-Blocker Delivery To investigate the effect of skin occlusion on topical delivery of $-blockers, permeation studies were performed under both occlusive and nonocclusive conditions. Figure 2 presents the percentage of drug transported across HEM with dosing volume of 0.15 mL topical gel under both occlusive and nonocclusive conditions. Significant higher amounts of propranolol and betaxolol were delivered across HEM under the occlusive condition (p < 0.05, Student’s t-test): after 24 h of gel application, approximately two to three times larger amounts of propranolol and betaxolol were delivered with occlusion. The occlusion enhanced and sustained drug delivery for a longer period of time than that without occlusion. Although the $-blocker gels in the donor chamber under occlusion were still wet at the end of the experiment, the gels in the donor chamber without occlusion became completely dry around 16 h after dosing. The dried vehicles hindered the release of drug into the SC, and therefore, decreased the delivery efficiencies after the initial period. In addition, it is generally known that skin occlusion can increase SC hydration, resulting in enhanced percutaneous absorption of topically applied drugs.25,26 It should be pointed out that


1726


In clinical practice, the vast majority of topical formulations are applied under nonocclusive condition exposing the formulations to the surrounding environment. The present study has shown that, for propranolol and betaxolol gels applied on the skin surface, more efficient drug delivery can be achieved with skin occlusion. This suggests the advantage of applying a semipermeable or impermeable film (or materials) to cover the administration site to provide skin occlusion for topical $blocker delivery—an approach that can prevent water evaporation from the drug vehicle and enhance skin penetration of the $-blockers. Based on this, dosing volume experiments in the present study were conducted under the occlusive conditions. Effect of Dosing Volume on Topical ␤-Blocker Delivery

Figure 2. Percent of (a) propranolol, (b) betaxolol, and (c) timolol delivered across skin from the $-blocker gels under both occlusive (diamonds) and nonocclusive conditions (squares) when the dosing volume was 0.15 mL. Data are presented as mean ± SD (n 3).

large variability was observed for timolol data under the occlusive condition, possibly a result of skin-to-skin variability from the different skin sources. The large skin-to-skin variability could also be related to the hydrophilic nature of timolol at pH 7.4. Because of this large skin-to-skin variability, the percentage of drug delivered across the HEM under occlusion was not statistically different from that under nonocclusive condition for timolol (p > 0.05, Student’s t-test).

Figure 3 presents the cumulative amounts of $-blockers delivered across HEM with dosing volumes varying from 0.03 to 0.5 mL. The data illustrate the effects of dosing volume upon the permeation of $-blockers across HEM. For propranolol and betaxolol, the cumulative amounts of drug permeated through the skin increased with increasing dosing volume. The cumulative amounts of propranolol and betaxolol were significantly different between each dosing volume (e.g., cumulative amounts of propranolol at 0.15 mL > 0.03 mL, propranolol at 0.5 mL > 0.03 mL, propranolol at 0.5 mL > 0.07 mL, propranolol at 0.5 mL > 0.15 mL, p < 0.05, ANOVA). However, no significant differences were observed for the cumulative amounts of timolol between each dosing volume, which might be attributed to its low skin permeability and large skin-to-skin variability as discussed in the section “Effect of Occlusion on Topical $-Blocker Delivery.” In the mass balance evaluation, the total recovery of propranolol, betaxolol, and timolol was 81 ± 15%, 87 ± 9%, and 59 ± 11%, respectively, at the end of the permeation experiments. The permeation profiles also display a trend of decreasing flux with time: the $-blockers were delivered across HEM at a faster rate initially and then followed by slower rates over time. Different characteristic shapes of the cumulative amount delivered versus time profiles were observed at different dosing volumes for the $-blockers. Smaller dosing volume led to a faster rate of flux decrease. This observation is consistent with the depletion of $-blockers in the gels in the donor chamber. For the largest dosing volume studied (propranolol of 0.5 mL), the cumulative amount delivered versus time plot approaches a linear profile similar to those observed in steady-state transport under the infinite dose condition. In this case, the apparent permeability coefficients can be determined using the method commonly employed in steady-state transport experiments (Eq. (2)). The apparent permeability coefficient calculated with this method for propranolol was 3.2 ± 1.3 × 10−7 cm/s (mean ± SD), which is approximately three times smaller than that observed under steady-state condition (mean ± SD = 9.1 ± 1.2 × 10−7 cm/s) in a previous study.13 This comparison, together with the shapes of the cumulative amount delivered versus time plots, indicates that even at the dosing volume of 0.5 mL, the permeation behavior of propranolol still cannot be assumed as steady state (i.e., infinite dose condition). Effect of ␤-Blocker Lipophilicity on Topical Delivery As drug permeation across skin can be influenced by the physicochemical properties of the drugs, the effects of drug lipophilicity upon skin permeation and drug depletion were


DOI 10.1002/jps.24390


1727

olol are more lipophilic than timolol. Under the dosing volume conditions of 0.03, 0.07, and 0.15 mL, the range of cumulative amounts transported across HEM at the end of the experiments was 40–250 :g for propranolol, 90–370 :g for betaxolol, and 50–85 :g for timolol. For drug depletion in the donor chamber, as timolol is more polar and has lower permeability coefficient than propranolol and betaxolol, skin permeation of timolol from the gel is expected to decrease at a slower rate than those of propranolol and betaxolol when the dosing volume decreased from 0.15 to 0.03 mL. In general, the SC intercellular lipid is the main pathway for skin permeation of lipophilic drugs. The skin appendages, which are considered as the shunt pathway, can also contribute to percutaneous absorption after topical application but are mainly for hydrophilic drugs. For lipophilic drugs, the contribution of skin appendages to drug permeation is minor.27,28 For the $-blockers propranolol and betaxolol, a previous study has shown that the contribution of HEM polar pathway to drug permeation was not significant at pH 7.4.13 In addition, the correlation between skin permeability coefficient and lipophilicity of the $-blockers observed in this previous study supports the SC lipoidal pathway as the main transport mechanism for the $-blockers. Although the investigated $-blockers are mostly ionized under physiological pH, the lipoidal pathway remained the major route of skin permeation of the free bases of propranolol and betaxolol. For timolol, the polar pathway can contribute to skin permeation but its effect is not expected to be significant as the steady-state permeability coefficient of skin for timolol was still larger than the estimated permeability coefficient of the polar pathway at pH 7.4.13 Finite Dose Diffusion Model

Figure 3. Cumulative amounts of (a) propranolol, (b) betaxolol, and (c) timolol permeated across skin from the $-blocker gels of dosing volumes 0.03–0.5 mL. Symbols: diamonds, 0.03 mL; squares, 0.07 mL; triangles, 0.15 mL; circles, 0.5 mL. There were only seven sampling time points (16, 20, 24, 40, 44, 48, and 72 h) for propranolol at dosing volume of 0.15 mL. Data are presented as mean ± SD (n 3).

investigated. Comparing the results among the $-blockers, the cumulative amounts of propranolol and betaxolol transported across HEM were significantly larger than those of timolol at the same dosing volumes after 1 day of gel application (p < 0.05, ANOVA). The larger amounts of propranolol and betaxolol than timolol delivered across HEM are consistent with the higher skin permeability for propranolol and betaxolol than timolol because of their lipophilicity: propranolol and betaxDOI 10.1002/jps.24390

The finite dose diffusion model describes the flux of a drug delivered across a membrane over time from a constant volume applied on the membrane at finite dose. The condition of skin permeation from the $-blocker gels applied on skin under occlusion with relatively constant gel volume over the duration of the application in the present study is consistent with this model. Before the analysis of the skin permeation data, the data obtained from the computer simulation of the time-dependent diffusion model were first compared with the results of the analytical solution of the finite dose diffusion model in a previous study.22 The J/Jss versus Dt/h2 profiles in the present computer simulation of varying membrane thicknesses and diffusion coefficient parameters were found to overlap with the results from the previous study of the same parameters (data not shown). The overlapping of the present and previous data suggests good agreement between the numerical solution of the computer simulation and the analytical solution of the diffusion model, this validating the computer simulation approach in the present study. According to the results of the analytical solution of Eq. (4) in the previous study22 and the present computer simulation, the finite dose flux versus time relationships show a general trend of increasing flux [or ln (J/Jss ) values] initially and then decreasing flux after the flux reaches its maximum at around Dt/h2 = 0.1–0.4 in the ln (J/Jss ) versus Dt/h2 profile. For $ < 1 or relatively large dosing volume, the peak of the finite dose flux approaches Jss . When the dosing volume decreases that $ > 1, the peak flux is significantly smaller than Jss . In both cases, when the drug depletes in the donor chamber, the


1728


flux decreases in a logarithmic manner, this showing a linear decrease in ln (J/Jss ) with time. To explain this relationship mathematically using the diffusion model, the first term in the summation expression (n = 1) in Eq. (4) is significantly greater than the other terms (n = 2 to ) in the expression when Dt/h2 > 0.5, and the slope of the plot of ln (J/Jss ) versus Dt/h2 is related to "1 : ln (J/Jss ) = −

"21 Dt +B h2

(12)

where B is a constant. According to the conventional firstorder kinetic model for drug depletion in the donor chamber that assumes membrane transport as an instantaneous process with a rate constant, k (i.e., simple first-order depletion kinetics):29,30 ln (Ct/C0 ) = −kt

(13)

where Ct and C0 are the concentration in the donor chamber at time t and the initial donor concentration, respectively. This concentration ratio is proportional to the flux ratio J/Jss , and the rate constant is proportional to the permeability coefficient of the membrane in the first-order kinetic model: ln (Ct/C0 ) = ln (J/Jss ) = −

Pt hv

$Dt K mv Dt =− 2 hv h h

$-Blocker Propranolol Betaxolol Timolol a b

log Ko/w a

log Ko/w b

Gel pH

funion

3.3 2.8 2.1

3.48 ± 0.02 2.80 ± 0.02 1.79 ± 0.02

7.4 7.4 7.4

0.0079 0.0108 0.0153

Values from the literature.16–18 Values calculated in the present study.

Using the Ko/w values in the present study, the distribution coefficients of the $-blockers from aqueous solution to skin for skin permeation were calculated using Eqs. (8)–(10). Second, the SC thickness was a variable that would provide an adequate fit between the experimental permeation data and the results of the computer simulation. Third, the parameters of diffusion coefficient and steady-state flux were calculated from the $-blocker permeability coefficients reported previously13 under the assumption that skin permeability coefficients for the $-blockers were not significantly affected by the HEC gel. Because of the sampling interval and duration of the permeation experiments in the present study, only simulated time points beyond Dt/h2 > 0.5 were used to analyze the transient flux data with the ln (J/Jss ) versus Dt/h2 relationships over a range of $ values.

(14) Analysis of Permeation Data Using Finite Dose Diffusion Model

To compare the ln (J/Jss ) versus t relationship of the firstorder depletion model (Eq. (14)) and that of the finite dose diffusion model (Eq. (12)), Eq. (14) can be rewritten as: ln (J/Jss ) = −

Table 3. Octanol/PBS Partition Coefficients of the $-Blockers Determined and Fraction of Unionized Species at pH 7.4 Calculated in the Present Study

(15)

From this comparison (Eqs. (15) and (12)), when the slope of the ln (J/Jss ) versus Dt/h2 plot of the first-order depletion model (−$) is equal to that of the finite dose diffusion model (−"1 2 ), the finite dose depletion process can be described approximately by the simple first-order depletion kinetics.

Figure 4 presents the theoretical relationship between the negative value of the slope of the ln (J/Jss ) versus Dt/h2 plot (in the decreasing flux phase) and $, where the slope of the ln (J/Jss ) versus Dt/h2 plot is equal to −"1 2 . The theoretical curve describes the rates of the decrease in drug flux at different $ values. When $ increases (i.e., smaller dosing volume), a higher rate of flux decrease (higher "1 2 value) is expected. This rate then approaches a plateau at high $ values. At high dosing

Finite Dose Diffusion Model Parameters The relationship between dosing volumes and changes in fluxes in the permeation experiments was examined using the results of the computer simulation. In this analysis, the appropriate $ and Kmv values for the permeation experiment results were estimated as described in the section “Data Analysis Using Computer Simulated Results.” First, the Ko/w values of the $blockers were determined in the experiments with octanol and PBS at pH 7.4. These results are shown in Table 3. Within data scatter, there is general agreement between the present partition coefficient results and those reported in the literature. It should be noted that although the partition coefficient of betaxolol in the present study was consistent with a previously reported value (log Ko/w = 2.8),16 the present betaxolol partition coefficient was significantly lower than another reported value (log Ko/w = 3.5).31 Possible reasons for the discrepancy of Ko/w in these studies can be attributed to the different experimental methods in the partitioning studies and the uncertainties in the pKa values of the drugs leading to errors in the calculation of the partition coefficients from the distribution coefficients.

Figure 4. Relationship between the rates of decrease in fluxes and dosing volumes and the comparison of the experimental and theoretical data for the $-blockers. Symbols: squares, propranolol in the uniform membrane model; diamonds, betaxolol in the uniform membrane model; triangles, propranolol in the heterogeneous membrane model; crosses, betaxolol in the heterogeneous membrane model. The curve was generated by computer simulation using diffusion theory.


DOI 10.1002/jps.24390

1729


volume (i.e., small $ values), the rate of change in ln (J/Jss ) is proportional to the $ value. This relationship at high dosing volume can be described by the first-order depletion of the drug in the donor chamber (i.e., flux following the simple first-order depletion model, Eqs. (14) and (15)). At small dosing volume, a deviation from this proportional relationship occurs, as expected, as the amount of drug in the membrane becomes comparable to the amount of depleting drug in the donor chamber. This deviation occurs at $ 1, and a decrease in dosing volume beyond this point would result in flux decrease at a slower rate than anticipated from the rate constant of the first-order depletion model. The peak flux in the J versus t profile is also significantly smaller than Jss under this condition. The theoretical curve in this figure shows one of the features of finite dose absorption, and this was used in the analysis of the experimental data in the present study. The experimental data of timolol were excluded in this analysis because of the large uncertainties of permeation data for timolol (Fig. 3C). In a preliminary examination, the timolol data showed large data scatter and deviation from the theoretical curve in the analysis as expected. For propranolol and betaxolol, the experimental data were found to be in agreement with the theoretical curve when the membrane thickness was 0.1 cm for the uniform membrane model and 0.7 cm for the heterogeneous membrane model (Fig. 4, Table 4). In the figure, the experimental data of both propranolol and betaxolol show similar trends that approach the plateau when the dosing volume decreases, indicating that the experimental results follow the feature of finite dose absorption according to the diffusion theory. However, the SC barrier thicknesses that provided the adequate fit between experimental data and computer simulation result were larger than the values expected from the morphology and physical thickness of SC. For example, a model derived from SC morphology suggested that the effective length of the SC permeation barrier was 0.005 cm assuming that the physical SC thickness was 0.0013 cm and tortuosity was 3.9.32 For comparison, the physical thickness of the epidermis was shown to be approximately 0.008 ± 0.002 cm.33 In another study, the tortuosity value was suggested to be 3.7 and 12.7 for hydrated “expanded” SC and “unexpanded” SC, respectively, and the hydrated thickness of SC could range from 0.004 to 0.01 cm.34,35 The thickness of 0.1 cm used in the uniform membrane model fitting was at least 10 times larger than the SC hydrated thickness. This could be attributed to the uniform membrane assumption for skin transport and the uncertainties of the effective length of SC permeation pathway as described above. In the heterogeneous model, a significantly longer SC permeation pathway (SC thickness parameter of 0.7 cm) than that in the uniform membrane model was required to fit the experimental data, which was also signif-

icantly longer than the permeation pathway length according to SC morphology. This was to compensate for the smaller SC loading capacity, that is, the lower partition coefficient Kmv , in the heterogeneous model compared with that in the uniform membrane model. The larger than expected SC thickness parameters for the $-blockers could be a result of drug partitioning and/or binding to the SC nonlipid domains. Drug partitioning and/or binding to these domains would lead to a higher than predicted SC loading capacity in the experiments and hence the larger apparent SC thickness in the model fitting because the models did not account for these SC domains. For example, the $-blockers could permeate into these nonlipid domains that effectively increase SC loading and transport lag time. As a result, a thicker donor gel was required in finite dose delivery to meet the $ < 1 condition (the transition to first-order kinetics of Eq. (15)) than those predicted by the skin transport models. It was noted that the diffusion coefficients obtained in the heterogeneous model were also unreasonably high because of the long permeation pathway required to fit the data. Other possible explanations of the large SC barrier thickness for the $-blockers were the uncertainties of the experimental partition coefficients and the finite dose permeation results in the present study and of the diffusion coefficients calculated from the permeability coefficients in the previous study13 that assumed the gel did not have any significant effects on skin permeability coefficient. It should be pointed out that the assumption of negligible transport barrier within the gel in the donor chamber did not affect the results as shown in Figure 4; the influence of drug diffusion coefficient in the gel was examined in the computer simulation and decreasing the diffusion coefficient by an order of magnitude was found to have little effect on the theoretical curve in the figure. Future studies are required to investigate the discrepancy between the apparent SC thickness obtained in the models and physical SC thickness and the effect of SC nonlipid partitioning upon the transport behavior of $-blockers in finite dose delivery across SC. Stability of Topical ␤-Blocker Gel The stability of 5 mg/mL propranolol hydrochloride, betaxolol hydrochloride, and timolol maleate gel formulations was examined over a 9-month period at room temperature and under refrigeration. The average concentrations (percentage of drug concentration remaining) were between 84% and 108% of the initial concentrations of propranolol and betaxolol in all samples stored at room temperature and 4°C in the refrigerator over 9 months. This suggests that the $-blockers are chemically stable (e.g., within 80%–125% of the label) for up to 9 months under those conditions. For timolol, the average

Table 4. Parameters Used in the Model Fitting Uniform Membrane $-Blocker Propranolol Betaxolol Timololb a b

P

(10−7

cm/s)

9.0 ± 1.2 5.1 ± 1.3 0.98 ± 0.36

Jss

(10−11

mol/cm2 /s)

2.00 1.02 0.20

app K mv

2.33 ± 0.08 1.06 ± 0.04 0.28 ± 0.01

D

(10−8

cm2 /s)

4.29 4.81 –

Heterogeneous Membrane h

(cm)a 0.1 0.1 –

app K mv

D (10−8 cm2 /s)

h (cm)a

0.30 ± 0.01 0.15 ± 0.01 0.04 ± 0.01

232 238 –

0.7 0.7 –

A single SC thickness parameter was used to fit the experimental data of propranolol and betaxolol in each membrane model. Timolol was not evaluated in the model analysis.

DOI 10.1002/jps.24390


1730


concentrations were between 95% and 180%. The increase in concentrations over time for some samples may be related to water loss in the formulations and cannot be explained at this time. No visible evidence of microbial growth was observed throughout the duration of the study. The present stability results are consistent with a previous study of propranolol hydrochloride in a suspending agent that propranolol was found to be stable even at 70°C.36 The stability information of $-blocker gel formulations in the present study would assist decision making in product labeling in compounding pharmacy after the preparation of the $-blocker gels. Topical Delivery of ␤-Blockers in Practical Application Topical application of $-blockers provides the advantage to avoid adverse effects related to systemic administration in the treatment of hemangiomas, especially when the drugs are used in infants. The use of an efficacious $-blocker topical formulation also does not require close monitoring of the patients under treatment for adverse effects that is a common practice with the systemic administration of propranolol on infants. Although topical application of timolol has been suggested as an effective and safe therapy for treating infantile hemangiomas, a commercial topical $-blocker product is currently not available. Particularly, ophthalmic timolol products used in the treatment of glaucoma, such as timolol ophthalmic solution or gel forming solution (GFS),37,38 have become the “de facto” standard of care in such topical treatment. Ophthalmic solution is generally not considered to be an effective dermatological drug delivery system. The use of ophthalmic GFS as a topical skin product is also misleading because this “gel forming solution” does not gel on skin; the GFS formulation contains an anionic derivative of gellan gum and becomes a gel in the presence of cations such as those in the precorneal tear film (not on the skin). For the effective development of a topical $-blocker dermatological product, the present study is the first to evaluate skin permeation of $blockers from topical gels under two influencing factors: dosing volume and skin occlusion. The present study compared the permeation of $-blockers propranolol, betaxolol, and timolol in topical skin delivery under occlusive and nonocclusive conditions and determined the effect of dosing volumes on the permeation of the $-blockers, which is an important factor to be investigated during early formulation development. These $-blockers were selected because of recent investigations of their use in infantile hemangioma treatment and their physicochemical properties. Propranolol was chosen because of the current use of systemic oral propranolol in infantile hemangioma treatment. Timolol was selected because it was investigated in case studies for the treatment of infantile hemangiomas.39,40 Betaxolol was investigated because its lipophilicity falls between those of propranolol and timolol. The results in the present study could provide information for physicians to optimize the dosing regimen and method of topically applied $-blockers in 2% HEC gel. It should be emphasized that only HEC gel was examined in the present study, so the conclusion of this study might not be applicable to other $-blocker topical formulations. According to the data in the dosing volume study of 0.15 mL, approximately 0.1–0.2 mg of propranolol and betaxolol and 0.04 mg timolol could be delivered across the skin in 1 day using the topical gels under occlusion. Because timolol was suggested to be effective in topical treatment of infantile hemangiomas as timolol ophthalmic solution,

the higher amounts of propranolol and betaxolol than timolol in topical skin delivery suggest that propranolol and betaxolol could have great potential to be drug candidates of topical hemangioma products. The present results suggest that a lower concentration of propranolol or betaxolol in topical gel than timolol is required to achieve the same drug concentration in the hemangioma tissues when applied topically. In the occlusion versus nonocclusion permeation study, the significant improvement in the delivery of $-blockers across skin under occlusion suggests that skin occlusion is preferred during topical treatment. This can be accomplished by, for example, using adhesive film dressings over the application site. Assuming that the gel can stay on the skin surface under occlusion, the present results suggest that the delivery of $-blockers could be sustained up to 3 days. In addition, larger amounts of $-blockers can be delivered when larger dosing volume is applied. Using the results in the present study, the dosing conditions such as total drug application time and dosing volume can be modified to achieve the desired drug delivery effects. Future formulation development such as the incorporation of chemical permeation enhancers can further improve topical $-blocker delivery. Other vehicles such as creams and ointments can also be used for topical delivery of the $-blockers and should be investigated.

CONCLUSIONS In order to develop an effective formulation of $-blockers for topical skin treatment of infantile hemangioma, the present study compared the effects of dosing volume of $-blocker gels upon skin permeation. The effect of skin occlusion upon skin permeation of the $-blockers was also evaluated. The results showed that skin permeation under occlusion was significantly higher than that under nonocclusive condition with the gel application: approximately two to three times higher amounts of propranolol and betaxolol were found to be delivered across HEM under occlusion at 24–72 h after application. This suggests that more efficient drug delivery could be achieved by the application of an occlusive dressing on the administration site in infantile hemangioma treatment. In addition to the occlusion effect, dosing volume was observed to affect the cumulative amounts of $-blockers delivered across skin. When the dosing volume increased, the amounts of propranolol and betaxolol delivered across HEM increased. The effects of occlusion and dosing volume were not statistically significant for timolol, probably because of the large skin-to-skin variability reflected in its permeation profile data. Skin was found to be more permeable to propranolol and betaxolol than timolol under the conditions in the present study, consistent with the lipophilicity of the $-blockers. Therefore, topical propranolol and betaxolol have the potential to be more effective topical drug treatments than topically administered timolol ophthalmic solution (the drug currently used in infantile hemangioma treatment). The experimental results of skin permeation of $-blockers are consistent with finite dose diffusion theory: when the dosing volume increased, the flux decreased at a slower rate after topical application. An adequate fit was observed between the experimental permeation data and the theoretical values of the finite dose models when the skin thickness was assumed to be 0.1–0.7 cm. The analysis suggests that percutaneous absorption from the topical gel under the studied conditions followed the transport behavior described by the diffusion theory. The larger than expected skin thickness parameters could be


DOI 10.1002/jps.24390


attributed to drug partitioning and/or binding to the SC nonlipid domains and the limitations of the models. Finally, the stability experiments showed that there was no significant degradation of the $-blockers in the gel formulations when stored either at room temperature or in the refrigerator over a period of 9 months. In summary, the findings in the present study suggest the feasibility of topical propranolol and betaxolol gels for skin treatment of infantile hemangiomas preferentially with skin occlusion applied concomitantly.

ACKNOWLEDGMENTS The authors thank Drs. Gerald B. Kasting and Jinsong Hao for helpful discussion, and Drs. Anusua Dasgupta, Beth Rymeski, and Denise Lagory for their help in the stability study.

REFERENCES 1. Drolet BA, Esterly NB, Frieden IJ. 1999. Hemangiomas in children. N Engl J Med 341(3):173–181. 2. Leaute-Labreze C, Prey S, Ezzedine K. 2011. Infantile haemangioma: Part II. Risks, complications and treatment. J Eur Acad Dermatol Venereol 25(11):1254–1260. 3. Leaute-Labreze C, Dumas de la Roque E, Hubiche T, Boralevi F, Thambo JB, Taieb A. 2008. Propranolol for severe hemangiomas of infancy. N Engl J Med 358(24):2649–2651. 4. Blatt J, Morrell DS, Buck S, Zdanski C, Gold S, Stavas J, Powell C, Burkhart CN. 2011. $-Blockers for infantile hemangiomas: A singleinstitution experience. Clin Pediatr 50(8):757–763. 5. Zimmermann AP, Wiegand S, Werner JA, Eivazi B. 2010. Propranolol therapy for infantile haemangiomas: Review of the literature. Int J Pediatr Otorhinolaryngol 74(4):338–342. 6. Starkey E, Shahidullah H. 2011. Propranolol for infantile haemangiomas: A review. Arch Dis Child 96(9):890–893. 7. Sans V, de la Roque ED, Berge J, Grenier N, Boralevi F, Mazereeuw-Hautier J, Lipsker D, Dupuis E, Ezzedine K, Vergnes P, Taieb A, Leaute-Labreze C. 2009. Propranolol for severe infantile hemangiomas: Follow-up report. Pediatrics 124(3):e423–e431. 8. Guo S, Ni N. 2010. Topical treatment for capillary hemangioma of the eyelid using $-blocker solution. Arch Ophthalmol 128(2):255–256. 9. Moehrle M, Leaute-Labreze C, Schmidt V, Rocken M, Poets CF, Goelz R. 2013. Topical timolol for small hemangiomas of infancy. Pediatr Dermatol 30(2):245–249. 10. Vintiloiu A, Leroux JC. 2008. Organogels and their use in drug delivery—A review. J Control Release 125(3):179–192. 11. Hoffman AS. 2002. Hydrogels for biomedical applications. Adv Drug Deliv Rev 54(1):3–12. 12. Chang RK, Raw A, Lionberger R, Yu L. 2013. Generic development of topical dermatologic products: Formulation development, process development, and testing of topical dermatologic products. AAPS J 15(1):41–52. 13. Chantasart D, Hao J, Li SK. 2013. Evaluation of skin permeation of $-blockers for topical drug delivery. Pharm Res 30(3):866–877. 14. Franz TJ, Lehman PA, Raney SG. 2009. Use of excised human skin to assess the bioequivalence of topical products. Skin Pharmacol Physiol 22(5):276–286. 15. Selzer D, Abdel-Mottaleb MM, Hahn T, Schaefer UF, Neumann D. 2013. Finite and infinite dosing: Difficulties in measurements, evaluations and predictions. Adv Drug Deliv Rev 65(2):278–294. 16. Modamio P, Lastra CF, Marino EL. 2000. A comparative in vitro study of percutaneous penetration of $-blockers in human skin. Int J Pharm 194(2):249–259. 17. Wiczling P, Kawczak P, Nasal A, Kaliszan R. 2006. Simultaneous determination of pKa and lipophilicity by gradient RP HPLC. Anal Chem 78(1):239–249.

DOI 10.1002/jps.24390

1731

18. Denet AR, Ucakar B, Preat V. 2003. Transdermal delivery of timolol and atenolol using electroporation and iontophoresis in combination: A mechanistic approach. Pharm Res 20(12):1946–1951. 19. Kligman AM, Christophers E. 1963. Preparation of isolated sheets of human stratum corneum. Arch Dermatol 88:702–705. 20. Kasting GB, Bowman LA. 1990. DC electrical properties of frozen, excised human skin. Pharm Res 7(2):134–143. 21. Peck KD, Ghanem AH, Higuchi WI. 1995. The effect of temperature upon the permeation of polar and ionic solutes through human epidermal membrane. J Pharm Sci 84(8):975–982. 22. Kasting GB. 2001. Kinetics of finite dose absorption through skin 1. Vanillylnonanamide. J Pharm Sci 90(2):202–212. 23. Potts RO, Guy RH. 1992. Predicting skin permeability. Pharm Res 9(5):663–669. 24. Hansen S, Lehr CM, Schaefer UF. 2013. Improved input parameters for diffusion models of skin absorption. Adv Drug Deliv Rev 65(2):251–264. 25. Zhai H, Maibach HI. 2001. Effects of skin occlusion on percutaneous absorption: An overview. Skin Pharmacol Appl Skin Physiol 14(1):1– 10. 26. Treffel P, Muret P, Muret-D’Aniello P, Coumes-Marquet S, Agache P. 1992. Effect of occlusion on in vitro percutaneous absorption of two compounds with different physicochemical properties. Skin Pharmacol 5(2):108–113. 27. Frum Y, Bonner MC, Eccleston GM, Meidan VM. 2007. The influence of drug partition coefficient on follicular penetration: In vitro human skin studies. Eur J Pharm Sci 30(3–4):280–287. 28. Essa EA, Bonner MC, Barry BW. 2002. Human skin sandwich for assessing shunt route penetration during passive and iontophoretic drug and liposome delivery. J Pharm Pharmacol 54(11):1481–1490. 29. Bourne DW. 2002. Pharmacokinetics. In Banker GS, Rhodes CT, editors. Modern Pharmaceutics, 4th ed. New York: Marcel Dekker, pp. 67–92. 30. Mayer PT, Xiang TX, Niemi R, Anderson BD. 2003. A hydrophobicity scale for the lipid bilayer barrier domain from peptide permeabilities: Nonadditivities in residue contributions. Biochemistry 42(6):1624–1636. R 31. Alcon Canada Inc. 1986. BETOPTIC S, betaxolol hydrochloride 0.25% (as base) ophthalmic suspension. Product monograph, control #049005. 32. Johnson ME, Blankschtein D, Langer R. 1997. Evaluation of solute permeation through the stratum corneum: Lateral bilayer diffusion as the primary transport mechanism. J Pharm Sci 86(10):1162–1172. 33. Sandby-Moller J, Poulsen T, Wulf HC. 2003. Epidermal thickness at different body sites: Relationship to age, gender, pigmentation, blood content, skin type and smoking habits. Acta Derm Venereol 83(6):410– 413. 34. Talreja P, Kleene NK, Pickens WL, Wang TF, Kasting GB. 2001. Visualization of the lipid barrier and measurement of lipid pathlength in human stratum corneum. AAPS PharmSci 3(2):E13. 35. Barbero AM, Frasch HF. 2006. Transcellular route of diffusion through stratum corneum: Results from finite element models. J Pharm Sci 95(10):2186–2194. 36. Geiger CM, Voudrie MA, Sorenson B. 2012. Stability of propranolol hydrochloride in SyrSpend SF. Int J Pharm Compd 16(6):513–515. 37. McMahon P, Oza V, Frieden IJ. 2012. Topical timolol for infantile hemangiomas: Putting a note of caution in “cautiously optimistic”. Pediatr Dermatol 29(1):127–130. 38. Sorrell J, Chamlin SL. 2013. Topical timolol 0.5% gel-forming solution for small deep facial infantile hemangiomas. Pediatr Dermatol 30(5):592–594. 39. Khunger N, Pahwa M. 2011. Dramatic response to topical timolol lotion of a large hemifacial infantile haemangioma associated with PHACE syndrome. Br J Dermatol 164(4):886–888. 40. Ni N, Langer P, Wagner R, Guo S. 2011. Topical timolol for periocular hemangioma: Report of further study. Arch Ophthalmol 129(3):377– 379.


Formulation and evaluation of optimized oxybenzone microsponge gel for topical delivery.

Nanoemulsion gel-based topical delivery of an antifungal drug: in vitro activity and in vivo evaluation.

Cavamax W7 composite psoralen ethosomal gel versus cavamax W7 psoralen solid complex gel for topical delivery: A comparative evaluation.

Development and characterization of niosomal gel for topical delivery of benzoyl peroxide.

Investigation of ethyl cellulose microsponge gel for topical delivery of eberconazole nitrate for fungal therapy.

Mechanical properties, skin permeation and in vivo evaluations of dexibuprofen-loaded emulsion gel for topical delivery.

Serratiopeptidase Niosomal Gel with Potential in Topical Delivery.

Etodolac Containing Topical Niosomal Gel: Formulation Development and Evaluation.

Comparison of in vitro dialysis release methods of loperamide-encapsulated liposomal gel for topical drug delivery.

Development and evaluation of nanostructured lipid carrier-based hydrogel for topical delivery of 5-fluorouracil.

Development and evaluation of solid lipid nanoparticles of mometasone furoate for topical delivery.

Design, characterization, and biological evaluation of curcumin-loaded surfactant-based systems for topical drug delivery.

Development of in situ gel for nasal delivery: design, optimization, in vitro and in vivo evaluation.

Management and outcome of topical beta-blocker-induced atrioventricular block.

Erratum to: Evaluation of Skin Permeation of β-Blockers for Topical Drug Delivery.

Formulation and evaluation of a topical niosomal gel containing a combination of benzoyl peroxide and tretinoin for antiacne activity.

Development of oxybutynin chloride topical gel for overactive bladder.

Ropivacaine loaded microemulsion and microemulsion-based gel for transdermal delivery: preparation, optimization, and evaluation.

Nanomiemgel--a novel drug delivery system for topical application--in vitro and in vivo evaluation.

Evaluation of a topical gel containing a novel combination of essential oils and antioxidants for reducing oral malodor in dogs.

Preclinical evaluation of a ricinoleic acid poloxamer gel system for transdermal eyelid delivery.

Correction: Nanomiemgel--a novel drug delivery system for topical application--in vitro and in vivo evaluation.

Lipid nanoparticles for the topical delivery of retinoids and derivatives.

Evaluation of hyaluronan gel (Gengigel(®) ) as a topical applicant in the treatment of gingivitis.