Optimized permeation enhancer for topical delivery of 5-fluorouracil-loaded elastic liposome using Design Expert: part II

Abstract

Objective: To prepare and optimize the topical elastic liposome (EL)-loaded carbopol-980 gel of 5-Fluorouracil (5-FU) containing permeation enhancers (azone, propylene glycol (PG) and lauryl alcohol (LA)) and further evaluation for permeation flux of 5-FU, the activation energy and irritation in the rat skin.

Methods: EL formulations were prepared using phosphatidylcholine and varied surfactants (Span 60, Span 80 and Tween-80) by rotator evaporation method and optimized by experimental design. In vitro characterizations dictated the EL containing Span 80 (lipid:surfactant = 7:3) (EL3-S80) for further optimization of gel. Different gel formulations (5% w/w) with varying concentration (1–3%) of permeation enhancers were prepared and evaluated for viscosity, spreadability, the 5-FU permeation and deposition. The activation energy using the Franz diffusion cell and the plausible irritation using the Draize test were assessed on the albino rat and rabbit, respectively.

Results and discussion: EL3-S80 was selected as an optimized EL owing to maximum desirability (0.99) and enhanced 5-FU flux (187.86 ± 14.1 μg/cm²/h). EL3-S80 suspension loaded gels (0.5%) revealed reduced viscosity leading to higher spreadability than blank gel. EL containing 3% azone in gel, EL containing 3% LA in gel and EL containing 3% PG in gel portrayed 187.86 ± 14.1, 117.7 ± 13.4 and 106.7 ± 7.3 μg/cm²/h as enhanced 5-FU flux values, respectively as compared to drug solution (8.8 ± 0.76 μg/cm²/h). Furthermore, reduced value of activation energy (2.63-folds) and the non-irritancy of gel could be effective and safe.

Conclusion: ELA-3 gel formulation could be used as an effective and economic gel in cutaneous cancer and skin-related keratoses.

• 5-Fluorouracil
• elastic liposome
• experimental design
• permeation flux
• topical delivery

Introduction

5-Fluorouracil (5-FU) is a well-established drug used to treat several cutaneous cancers and skin conditions such as actinic and solar keratoses (Tutrone et al., 2003). Clinical efficacy of orally delivered conventional dosage form is challengeable due to enzymatic and non-enzymatic degradation at pre-enterocyte and intra-enterocyte and subsequently with the hepatic (cytochrome P450 3A4 enzyme) enzyme. Moreover, lipophilic nature of the intestinal membrane, P-gp efflux pumps and post-enterocyte process are considerable critical factors for limited oral bioavailability (BA) (Dahan & Hoffman, 2008). It was reported that the 5-FU has exhibited rapid elimination rate with a short half-life (8–20 min) following the oral administration. Therefore, high therapeutic dose are required to achieve high systemic plasma level and less than 20% of the administered dose undergoes to enzymatic degradation. It is noteworthy that the tumor cells are exposed to the rate limiting-active metabolite for a short time locally within the tissue or systemically (10–20 min) when the 5-FU was administered with 600 mg/m²/day (Tanaka et al., 2000). Thus, oral administration of 5-FU has exhibited erratic and poor oral absorption, inter-and intra-subject variation and low plasma level leading to significant variation in BA indicating clinically limited therapeutic efficacy (Diasio & Harris, 1989). In addition, high and frequent dose administration of 5-FU to get improved oral BA is challenging and responsible for precipitating several serious side effects such as gastric disturbances and myelosuppression (Fraile et al., 1980; Lai & Guo, 2011). Thus, as described earlier, variable and poor oral BA, short plasma level, rapid elimination half-life and poor patient compliance render this 5-FU as a suitable candidate for topical delivery with enhanced permeation across the skin.

These limitations could not be overlooked in formulation fabrication to develop topical drug delivery for enhanced clinical efficacies and reduced toxicity. In several research publications, elastic liposomes (ELs) (cholesterol-free)-based drug delivery has been exploited as a promising alternative carrier which holds ultra-deformability, high drug encapsulation efficiency and cost-effective formulation for enhanced permeation of acyclovir and paclitaxel as compared to conventional liposome (cholesterol-containing) (Jain et al., 2008; Utreja et al., 2011). Thus, our research laboratory utilized EL for delivery of 5-FU as reported earlier (Hussain et al., 2014a). However, in the present study, we further elaborated to identify the suitable and optimum application of skin permeation enhancers which may assist permeation of EL through the tiny microscopic pores of the skin. In general, EL is more flexible and permeable as compared to rigid liposome (due to cholesterol) leading to potential application in topical drug delivery. In last decade, extensive researches have been reported on the topical and transdermal drug delivery using several permeation enhancers such as Azone®, isopropyl myristate and lauryl alcohol (LA) for improved permeation and drug deposition (Singh et al., 2005).

The present investigation has attempted to prepare 5-FU loaded EL for topical delivery through response surface (RS) method and the significance of mathematic models were estimated by the analysis of variation (ANOVA). The most robust formulation with optimum desired outcome was selected based on the RS plots and desirability function. The selected optimum formulation was further loaded into the carbopol-980 gel (1%) and evaluated with varied concentration of permeation enhancers (azone, LA and propylene glycol (PG)). Thus, prepared gels were characterized for vesicular size, zeta potential, pH, viscosity, spreadability, in vitro skin permeation and deposition assessment, activation energy of the treated skin and Draize test.

Methods

Materials

5-FU and high purity Phospholipon®90 H (phosphatidylcholine, PC) were kindly gifted from Spectrochem Pvt. Ltd (Mumbai, India) and Lipoid-GMB (Frigenstrasse-4, D-67065, Ludwigshafen, Germany), respectively. Phospholipon®90 H is hydrogenated PC containing stearic acid (∼85%) and palmitic (∼15%) fatty acids. Tween-80, PG and polyethylene glycol 400 (PEG-400) were obtained from Merck Chemicals (Mumbai, India). Sodium hydroxide pellets (AR grade), PG, triethanolamine, Span 60 and Span 80 were purchased from S D Fine Chemicals Ltd. (Mumbai, India) and Hi Media (Mumbai, India), respectively. LA was procured from Sisco Research Laboratories Pvt. Ltd. (Mumbai, India). Azone (laurocapram) was obtained from Biochem Pharmaceutical Industries (Mumbai, India). All other reagents used were of analytical grade.

Preparation of elastic liposomes and characterizations

5-FU-loaded EL were prepared and characterized for the particle size, zeta potential, viscosity, percent drug entrapment efficiency (% EE) and permeation flux of 5-FU as reported in our preceding article (Hussain et al., 2014a). Briefly, different formulations of EL with the varied amount of surfactants (Span 60, Span 80 and Tween-80) and PC were prepared to investigate in vitro flux value across the abdominal rat skin. Span 80 containing EL (EL3-S80) was selected as an optimized formulation (PC:Span 80 = 7:3) with maximum % EE, elasticity and enhanced permeation flux without permeation enhancer. In this study, the strength (5-FU concentration) of the final formulation was 5% w/w in EL3-S80 which was stored and used to investigate the effect of permeation enhancer containing gel on permeation parameters across the rat skin under varied conditions.

Particle size and its distribution, polydispersity index and zeta potential for EL suspension as well as gel loaded with EL were subjected for size analysis after 100 times dilution with distilled water. For suspension, 1 ml sample (EL suspension) was dispersed in 100 ml distilled water to get uniformly dispersed EL vesicles. Similarly, 1 g of gel containing EL was completely dispersed in 100 ml of distilled water to get homogeneous colloidal dispersion of EL for particle size and zeta potential measurement as reported earlier (Khurana et al., 2013; Hussain et al., 2014b). The particle size and surface (zeta potential) analysis of the prepared EL suspension and gel formulations were carried out using the photon correlation spectroscopy with a Malvern Nanosizer ZS (Malvern Instruments, Worcestershire, UK). The analysis was performed at 25 ± 1°C temperature and scattering angle of 90°. The samples were analyzed in a triplicate manner to obtain mean (±standard deviation) size and zeta potential value.

Optimization using experimental design software

A two-factor and three-level of full factorial design (FFD) (3²) was used for optimization process employing the Design Expert® software (version 8.0.7.1, Stat-Ease, Inc., Minneapolis, MN). In this experimental design (FFD), total nine required formulations (EL) were obtained using independent variables against dependable variables (responses). In vitro permeation flux of 5-FU for different surfactants and % EE were selected as response variables. The concentration of PC (A) and surfactant (B) were selected as the independent variables (at three levels; (+) as maximum, (0) as intermediate and (−) as minimum). A statistical analysis was performed using the same software to obtain the most robust gel formulation.

Preparation of different gel formulations

A weighed amount of carbopol-980 powder was dispersed in cold distilled water and then basic triethanolamine (few drops) was added to initiate cross-linking which leads to increase viscosity. Final pH was adjusted to 7.4 with the addition of sodium hydroxide (2 M) solution. The pH was measured using digital electronic pH meter (Hanna Instrument HI 9321, Ann Arbor, MI). The preceding optimized EL3-S80 containing 5-FU was selected to incorporate into the freshly prepared carbopol-980 gel (1% w/v) as per method reported earlier (Hussain et al., 2014b). The final strength of 5-FU in 0.5% w/v gel was maintained at 5% w/w. Blank EL3-S80 (drug-free) incorporated in the blank gel (permeation enhancer free) was served as control gel. Furthermore, three permeation enhancers such as azone, LA and PG in the varied concentration range (1, 2 and 3% w/w) were used in gel formulations. Thus, total nine gel formulations were prepared. EL3-S80 containing drug and control gel (EL3-S80 blank gel) were prepared for comparison. EL3-S80 gels containing a varied concentration of azone (expressed in descending order of concentration) were ELA-3 (3%), ELA-2 (2%) and ELA-1 (1%). Similarly, ELL-3, ELL-2 and ELL-1 were prepared gels containing 3%, 2% and 1% LA, respectively while ELP-3, ELP-2 and ELP-1 gels were formulated by incorporating PG.

Characterization of gels

All prepared EL gels loaded with EL3-S80 were characterized for vesicle size, polydispersity index, zeta potential and pH. Vesicular size and zeta potential (mV) of EL3-S80 were determined using Zetasizer Nano ZS (Malvern Instruments) at 25 ± 1 °C after dilution with distilled water just before analysis. The samples were analyzed using an aqueous dip cell in an automatic mode.

Viscosity assessment

The viscosity of the different prepared EL gel formulations was determined using R/S CPS Plus Rheometer (Brookfield Engineering Laboratories, Middleboro, MA) using spindle # C 50-1 at different temperature of 27, 37 and 47 °C as per method reported in our preceding paper (Hussain et al., 2013). A sample of gel formulations were analyzed at different rotational speeds (0.5, 1, 2, 3, 4, 5, 10, 20, 30, 50, 60, 70, 80, 100 rpm) of the spindle in a triplicate manner. One gram of each sample was used for the assessment of viscosity for 50 min of the operation time. Applied shear rate and the diameter of the spindle were 410 min⁻¹ and was 50 mm, respectively.

Spreadability

The spreading ability of the topical formulations offers great advantages over non-spreadable formulations intended for the topical application. This property renders the topical pharmaceutical preparations as a good parameter to be assessed which indicates the extent of thin film formation left after application. This represents the thickness of the film that the formulation leaves a thin film on the applied skin and considered as an important feature in the topical formulation. The spreadability was assessed using the method described earlier with slight modification (Contreras & Sanchez, 2002). The gel formulation with known amount (0.5 ± 0.01 g) and a constant initial area (0.88 cm²) in cylindrical shape was cast using a microtome. This was then pressed between the microtome and the flat surface of the graduated glass plate (6 × 6, 25 g) upon which it was pressed with different varied (50, 100, 200 and 500 g) weights at an interval of each minute. The diameter after each time interval was expressed in term of the area (cm²) and the variation in the area was analyzed as a function of the weight. The experiment is performed in a triplicate manner at 25 °C. Gel producing a thin film was considered with maximum spreadability after topical application.

In vitro skin permeation and deposition studies of gels

The abdominal portion of skin was obtained from a healthy male albino rat (180–200 g) of about 6–8-weeks old for the experiment. In vitro passive drug permeations of prepared gel were carried out using double chamber diffusion cell after approval by ethical committee (IAEC, Faculty of Pharmacy, Hamdard University) and procedure adopted as reported earlier (Hussain et al., 2014a). Each cell has the effective diffusion area and volume of 1 cm² and 22.5 ml, respectively. The rat skin specimen was placed between the donor and receptor chamber in such way that epidermis portion faces donor chamber. The receptor chamber was filled (22.5 ml) with PBS (pH 7.4). Different gel formulations, drug-loaded EL3-S80 suspension and drug solution (DS) were placed onto the donor chamber on the skin to carry out the experiment stirred constantly at 32 ± 1 °C using a magnetic bead. Sample (1 ml) was withdrawn from the receptor compartment at 0.5, 1, 2, 4, 6, 8, 10, 12, 16, 20 and 24 h time intervals and immediately replaced with fresh PBS. The amount of 5-FU permeated across the skin was analyzed using validated HPLC method at 266 nm as reported earlier (Hussain et al., 2014a).

After completion of the permeation study, the exhausted skin specimens were carefully removed and washed with a hydroalcoholic solution (50% ethanol) to free with the adhered samples. The washed skin was cut into small pieces, homogenized with the hydroalcoholic solution and kept for 24 h. The amount of 5-FU was quantified from the supernatant using HPLC method obtained after centrifugation at 3000 rpm for 5 min (Hussain et al., 2014a).

Assessment of the activation energy

The energy of activation of 5-FU from the optimized gels (ELA-3, ELL-3 and ELP-3) containing different permeation enhancer (3%), 5-FU-loaded EL-S80 suspension and DS were determined. In vitro permeation studies of the drug across the rat skin at three varied temperature (27, 37 and 47 °C of the receptor medium) were carried out. All the experimental conditions were kept same as described under the section of the permeation studies except the temperature of the medium (receptor chamber) and receptor medium comprising of 30% PEG-400. PEG-400 used in the medium was soluble and developed sink condition for hydrophilic 5-FU. Donor chamber was loaded with gel formulations. Permeability coefficient (Kp) was then determined using the Arrhenius equation:

Ea is as the activation energy. R and T are the gas constant (1.987 kcal/mol) and the absolute temperature, respectively. P and P₀ are the Kp and the Arrhenius coefficient, respectively in the above equation (Shakeel et al., 2008).

In vivo skin irritation potential

The Draize study was performed for the assessment of irritation and hypersensitive reaction using healthy male albino rabbits (1.8–2.0 kg) under ethical conditions (Draize et al., 1944). The hairs on the back with the marked area were trimmed off just 24 h before the experiment. Three squares at the different position were made on each animal. Each rabbit had three different marked areas for application. The first, second, third and fourth groups were served as untreated, blank gel-treated, DS-treated and positive control sodium lauryl sulfate (SLS) (20% w/v), respectively for topical application. Rest, fifth, sixth and seventh groups received ELA-3 gel, ELA-2 and ELA-1 gel respectively. Similarly, remaining groups were divided for ELL (eighth, ninth, 10th) and ELP (11th, 12th and 13th) gel formulations. At varied time intervals (0, 4 and 72 h) after application, the exposed areas were scored for developed erythema and edema on a grade of 0–4.

Results

Preparation of elastic liposomes and optimization

Three levels and two responses (flux and %EE) have been portrayed in Table 1. Surfactant concentrations (5–30 mg) have shown a significant effect on flux profile of 5-FU as shown in Table 2. The polynomial equations for the flux of 5-FU and % EE responses of the formulations prepared with Span 60 are Flux-Span 60 = 32.37 + 1.01A + 26.75B and % EE-Span 60 = 53.26 + 1.05A + 24.25B, respectively (analyzed using Design Expert software).

Table 1. Factors and their corresponding levels for 3² optimization technique.

	Levels
Factors	Low (−)	Medium (0)	High (+)
Phosphatidylcholine (mg)	70	82.5	95
Surfactant (mg)	5	17.5	30
Responses	Constraints
Flux (μg/cm^2/h)	Maximum
Entrapment efficiency (%)	Maximum

Table 2. Experimental flux values (μg/cm²/h) and % EE obtained from formulations prepared of various surfactants used in 3² optimization technique after 24 h (mean ± SD; n = 3).

		Span 60		Span 80		Tween-80
*PC (mg)	Surfactant (mg)	Flux	% EE	Flux	% EE	Flux	% EE
70.0 (−1)	5.0 (−1)	8.60 ± 0.4	40.3 ± 2.1	11.0 ± 0.7	48.6 ± 3.3	12.6 ± 0.2	34.7 ± 1.1
70.0 (−1)	17.5 (0)	18.0 ± 1.2	56.4 ± 3.6	23.0 ± 0.5	59.8 ± 3.7	34.5 ± 1.0	46.8 ± 3.2
70.0 (−1)	30.0 (+1)	73.1 ± 5.01	60.5 ± 2.7	83.91 ± 5.1	77.0 ± 8.3	72.20 ± 9.5	55.2 ± 9.5
82.5 (0)	5.0 (−1)	13.0 ± 1.3	42.0 ± 5.6	25.8 ± 1.3	51.5 ± 4.5	16.8 ± 1.8	39.8 ± 5.7
82.5 (0)	17.5 (0)	21.0 ± 1.0	61.7 ± 3.5	28.0 ± 1.1	75.8 ± 7.2	45.6 ± 2.9	48.9 ± 3.8
82.5 (0)	30.0 (+1)	60.0 ± 2.8	75.8 ± 7.1	70.5 ± 4.2	79.8 ± 9.5	61.0 ± 3.8	64.9 ± 3.6
95.0 (+1)	5.0 (−1)	15.9 ± 1.1	49.1 ± 2.2	22.8 ± 1.4	58.7 ± 4.4	19.0 ± 0.5	46.1 ± 2.7
95.0 (+1)	17.5 (0)	28.0 ± 1.5	69.1 ± 5.9	35.7 ± 2.0	71.4 ± 5.6	43.0 ± 3.2	52.7 ± 5.2
95.0 (+1)	30.0 (+1)	65.0 ± 2.0	78.5 ± 6.0	78.9 ± 3.8	84.8 ± 6.8	58.0 ± 3.5	71.9 ± 7.3
				Predicted		Actual
Code	Lipid (mg, A)	Surfactants (mg, B)		Flux (μg/cm^2/h)	% EE	Flux (μg/cm^2/h)	% EE
EL3-S60	PC	Span 60		64.62	78.57	73.12 ± 5.01	60.5 ± 2.7
EL3-S80	PC	Span 80		73.35	84.6	83.91 ± 5.1	77.0 ± 8.3
EL3-T80	PC	Tween-80		64.75	68.48	72.20 ± 9.5	55.20 ± 9.5

*PC, phosphatidylcholine; EL3-S60, elastic liposome containing Span 60; EL3-S80, elastic liposome containing Span 80; EL3-T80, elastic liposome containing Tween-80.

Statistically, the responses were precisely described by the respective models evidenced with the F value of 17.12 (p < 0.05) and 9.99 for Flux-Span 60 and % EE-Span 60, respectively (Verma et al., 2014). These equations clearly indicated that the Flux-Span 60 and % EE-Span 60 were significantly affected with the concentration of B for flux (p < 0.0006) and % EE (p < 0.002), respectively. Thus, the increase in the concentration of Span 60, the flux rate of 5-FU across the stratum corneum (SC) caused significantly enhanced (8.60 ± 0.4–73.1 ± 5.01 μg/cm²/h, Table 2). Furthermore, a 3-dimensional (3-D) contour plot, 2-dimensional (2-D) contour plot and interaction curve are illustrated in Figures 1(A–C) and 2(A–C) for the Flux-Span 60 and % EE-Span 60, respectively. The mean % EE values for Span 60 containing formulations were found to be in the range from 40.3 ± 2.1 to 78.5 ± 6.0.

Figure 1. Effect of surfactants on flux value across rat skin of optimized three elastic liposomes: (A, D and G) represent 3D-contour profile, (B, E and H) represent 2D contour and (C, F and I) represent interaction curves.

Figure 2. Effect of surfactants on % EE of optimized three elastic liposomes: (A, D and G) represent 3D-contour profile, (B, E and H) represent 2D contour and (C, F and I) represent interaction curves.

These mathematical equations were Flux-Span 80 = 41.27 + 2.16A + 29.92B and % EE-Span 80 = 67.6 +6.11A + 11.7B for flux and %EE, respectively (Table 2) when the Span 80 was used in EL formulations. The Flux-Span 80 response was influenced with the concentration of B (p < 0.0014). Thus, flux value of 5-FU increases with the increase in the concentration of the Span 80. The mean flux values of nine EL formulations were ranged from 11.0 ± 0.7 to 83.91 ± 5.1 μg/cm²/h as shown in Table 2. For % EE, the equation was 67.6 + 6.11A + 11.7B and significantly affected with the concentration of B (p < 0.002) and A (p < 0.04). A 3-D contour plot, 2-D contour plot and interaction curve are presented in Figure 1(D–F) for the response Flux-Span 80 and in Figure 2(D–F) for the response % EE-Span 80.

Finally, equations for these preceding responses with Tween-80-based EL suspensions were Flux-Tween-80 = 41.27 + 2.16A + 29.92B and % EE-Tween-80 = 67.6 +6.11A + 11.7B, respectively which displayed a linear trend. ANOVA analysis recommended the suitability of the models with F value of 51.84 (p < 0.0001) and 76.81 (p < 0.0001) for Flux-Tween-80 and % EE-Tween-80, respectively. Thus, both responses (flux and % EE) were increased with the increase in the concentration of Tween-80. The mean flux range (from 12.6 ± 0.2 to 72.20 ± 9.5 μg/cm²/h), % EE (from 34.7 ± 1.1 to 71.9 ± 7.3) and 2-D plots along with 3-D plots and interaction curves are presented in Table 2 and Figures 1(G–I) and 2(G–I), respectively.

Preparation and characterizations of EL3-S80-loaded gel formulations

A detailed composition and characterizing parameters of gels such as vesicle size, size distribution, zeta potential and pH have been portrayed in Tables 3 and 4. The vesicular size was not affected with the concentration of permeation enhancers. However, the charge densities (zeta potential) were slightly changed with them which showed maximum negative value in LA-based gel formulation as compared to others. These values were in a range from −34.9 to −39.7 mV and −18.7 to −26.2 mV for LA and azone, respectively. The zeta potential values of PG-based gel were found to be comparable to LA containing gel. ELA-3, ELL-3 and ELP-3 were selected for evaluation of rheological profile such as viscosity and spreadability as demonstrated in Table 4. ELA-3 was found to have better spreadability and viscosity (3098 ± 113.2 cP) as tabulated in Table 4. Percent spread by weight of ELA-3, ELL-3 and ELP-3 was 147.1 ± 17.9, 132.3 ± 14.5 and 112.3 ± 21.7 which were observed higher than blank gel formulation (53.31 ± 7.9).

Table 3. The composition of EL3-S80-loaded gel formulations with different permeation enhancers.

Formulation code	†SPE	Amount (w/v)	TE (w/w)	5-FU (w/w)
ELA-3	Azone	3	0.5	5.0
ELA-2	Azone	2	0.5	5.0
ELA-1	Azone	1	0.5	5.0
ELL-3	Lauryl alcohol	3	0.5	5.0
ELL-2	Lauryl alcohol	2	0.5	5.0
ELL-1	Lauryl alcohol	1	0.5	5.0
ELP-3	Propylene glycol	3	0.5	5.0
ELP-2	Propylene glycol	2	0.5	5.0
ELP-1	Propylene glycol	1	0.5	5.0
EL3-S80 gel	–	–	0.5	5.0
EL3-S80 blank	–	–	0.5	–

Value represented as mean ± SD (n = 3);

† SPE, skin permeation enhancer; TE, triethanolamine; EL3S80, elastic liposome containing Span 80; ELA-3, EL3S80 incorporated into gel containing 3% Azone (the number so on); ELL-3, EL3S80 incorporated into gel containing 3% Lauryl alcohol; ELP, EL3S80 incorporated into gel containing 3% propylene glycol.

Table 4. Characterizations of various developed gel formulations and evaluation of screened gels.

Formulation code	Vesicle size (nm)*	Polydispersity index*	ζ (mV)	pH (initial)	pH (adjusted)
ELA-3	231.6 ± 15.4	0.91 ± 0.01	−18.7	6.8	7.4
ELA-2	227.3 ± 18.7	0.83 ± 0.05	−22.9	6.5	7.4
ELA-1	216.5 ± 12.8	0.80 ± 0.03	−26.2	5.9	7.4
ELL-3	219.9 ± 10.0	0.88 ± 0.01	−34.9	5.4	7.4
ELL-2	215.5 ± 19.2	0.79 ± 0.07	−38.6	5.1	7.4
ELL-1	211.2 ± 17.6	0.81 ± 0.05	−39.7	5.1	7.4
ELP-3	235.0 ± 13.8	0.83 ± 0.09	−33.9	5.4	7.4
ELP-2	221.3 ± 18.5	0.77 ± 0.07	−29.1	5.6	7.4
ELP-1	215.1 ± 16.0	0.73 ± 0.06	−32.9	5.5	7.4
EL-S80 gel	218.8 ± 16.3	0.82 ± 0.03	−32.4	5.8	7.4
EL-S80 blank	225.8 ± 17.8	0.86 ± 0.04	−30.1	5.6	7.4
Evaluation of selected gel formulations
Parameters		ELA-3	ELL-3	ELP-3	Gel (1% w/v)
% spread by weight		147.1 ± 17.9	132.3 ± 14.5	112.3 ± 21.7	53.31 ± 7.9
Viscosity (cP)		3098 ± 113.2	3167 ± 89.05	3271 ± 65.08	4794 ± 45.11
Drug deposited (μg)		532.5 ± 10.1	428.07 ± 16.3	327.5 ± 13.8	NA
Steady state flux (Jss, μg/cm^2/h)		187.86 ± 14.1	117.7 ± 13.4	106.7 ± 7.33	NA

*Value represented as mean ± SD (n = 3), ζ, zeta potential. Values represented as mean ± SD (n = 3).

Drug permeation and deposition studies

The values of flux and ER were compared with the 5-FU aqueous solution as well as EL3-S80 suspension. Maximum flux (5-FU) values obtained were 187.86 ± 14.1 and 65.6 ± 7.2 μg/cm²/h for ELA-3 and EL-S80 as compared to DS (8.8 ± 0.76 μg/cm²/h), respectively. ER¹ and ER² are the enhancement ratio of permeation flux of 5-FU from the formulation to DS and from the formulation to EL3-S80 suspension respectively. ER¹ and ER² were found to be 21.34 and 2.86, respectively for ELA-3 formulation. The permeation flux of 5-FU was approximately 21.3- and 2.8-folds greater than the 5-FU solution and EL-S80, respectively which indicated facilitated flux rate due to permeation enhancer. As reported in Figure 3, the ELA-3 was found to enhance 5-FU permeation significantly as compared to the DS across the SC barrier. Similarly, the maximum flux (steady state flux) were 117.7 ± 13.4 and 106.7 ± 7.3 μg/cm²/h for ELL-3 and ELP-3, respectively which revealed that ER¹ values were found to be 13.35 and 12.13, respectively. These two values were closely related but significantly different than ELA-3 (21.34-folds higher). Moreover, the values of ER² were found to be 1.8 and 1.6 suggesting increased the permeation flux of 5-FU as compared to EL suspension (EL3-S80).

Figure 3. The percent cumulative amount of 5-FU permeated across rat skin (24 h).

After permeation studies, the gel formulations were subjected to drug deposition study under same experimental conditions. ELA-3, ELL-3 and ELP-3 showed 5-FU deposition of 532.5 ± 10.1, 428.07 ± 16.3 and 327.5 ± 13.8 μg, respectively as compared to the free DS (84.5 ± 4.7 μg) and liposome suspension (132.0 ± 8.2 μg). Thus, the amount of 5-FU deposited was found to be 6.3, 5.0 and 3.8-folds higher in ELA-3, ELL-3 and ELP-3, respectively as compared to DS after 24 h. The image of drug deposited by all gel formulations is portrayed in Figure 4.

Figure 4. Drug deposited into the skin after 24 h from various formulations.

Assessment of the activation energy

Three selected gel formulations (ELA-3, ELL-3 and ELP-3) were subjected for determination of the magnitude of activation energy and compared with EL3-S80 suspension and DS. The study was carried out at three different temperatures (27, 37 and 47 °C). The value of this energy for drug molecule depends on its route of diffusion across the rat skin and physicochemical nature drug entity. It must be significantly changed with the change in structural and phase behavior of SC lipid after formulation application. Thus, EL encapsulated in gel might be able to change the value of this energy by the overall action of gel formulation on SC. Kp values at each respective temperature were obtained. Further, logKp were plotted against 1/T (×10³) as demonstrated in Figure 5. The plot exhibited the linear relationship in the explored temperature ranges. In previous studies, authors reported about 10.7 and 4.1 kcal/mole energy of activation after application of PC-based liposomal formulation across rat skin and human skin, respectively (Pagano & Thompson, 1968; Monti et al., 1995). In this study, the energy was 3.8 kcal/mole for ELA-3 gel indicating 2.63- and 2-fold higher as compared to the DS and EL3-S80, respectively. This indicated an augmented effect on rat SC after application of gel formulation containing a permeation enhancer.

Figure 5. Arrhenius plot of formulations and drug solution permeation across the skin.

Skin irritation potential

The irritation ability of developed gel formulations containing permeation enhancers at all explored concentration, blank gel and DS in water were carried out and compared against the positive control as well as the untreated group. The study was performed on rabbit skin on back portion prepared carefully without any injury. The results in terms of scores have been listed in Table 5 after an observation made over the period of 72 h. It is noteworthy that the acceptability and applicability of topical formulation are limited by the patient if any erythema, edema and sign of irritation are elicited. Therefore, it is required to demonstrate any possibility of these reactions after topical application. The study revealed that all developed formulations were free from serious erythematic reaction excluding positive group (SLS solution) over the experimental period of 72 h. Table 5 illustrates that test formulations scored ranged as 0–0.32 which dictated safety criteria of the developed formulation. The untreated, blank gel and DS scored zero in Table 5. There were no any redness, edema and hypersensitive inflammatory reaction after topical application of the formulation. However, maximum scores were obtained in positive control group suggesting the irritating ability of surfactant SLS after topical application. The scored values were described below the table in the footnote. Thus, the results suggested non-irritant and skin compatible of developed gel formulations containing permeation enhancer. Similar findings were reported in our preceding report using carbopol gel loaded with nanoemulsion of amphotericin B (Hussain et al., 2014a). The collated images of treated skin with higher content of the permeation enhancers, control groups and others scored other than zero value have been portrayed in Figure 6. Positive control treated with SLS solution showed marked redness owing to inflammation and severe irritation on the applied area (Figure 6B) and no any sign of irritation or inflammation occurred in the untreated groups (Figure 6A). Other treated groups did not reveal any sign of inflammation or hypersensitivity reaction as shown in the Figure 6(E–H) which was comparable to control negative blank and DS-treated group.

Figure 6. Representative images of Draize test on rabbit dorsal skin. (A) Untreated skin. (B) SLS. (C) Blank gel. (D) Drug solution. (E) ELA-3 gel. (F) ELA-2. (G) ELL-3. (H) ELP-3.

Table 5. Mean erythemal and edema scores observed at the end of 1, 24, 48 and 72 h.

	Erythema				Edema
Formulation	1 h	24 h	48 h	72 h	1 h	24 h	48 h	72 h
Control untreated	0	0	0	0	0	0	0	0
Blank gel	0	0	0	0	0	0	0	0
Drug solution	0	0	0	0	0	0	0	0
SLS (20% w/v)	1.6	2.24	3.8	4.0	0	0.5	1.37	1.9
ELA-3	0	0	0.15	0.32	0	0	0	0
ELA-2	0	0	0	0.21	0	0	0	0
ELA-1	0	0	0	0	0	0	0	0
ELL-3	0	0	0	0.28	0	0	0	0
ELL-2	0	0	0	0	0	0	0	0
ELL-1	0	0	0	0	0	0	0	0
ELP-3	0	0	0	0	0	0	0	0
ELP-2	0	0	0	0	0	0	0	0
ELP-1	0	0	0	0	0	0	0	0

Values represented as mean (n = 3). Scores defined as 0, no erythema; 0–0.1, non-irritant and safe; 1.0–1.9,  irritant; 2.0–3.0, moderate to severe erythema (light red); 4, severe erythema (extreme redness).

SLS, sodium lauryl sulfate.

Discussion

Preparation of elastic liposomes and optimization

Vesicular drug delivery using EL was selected to deliver hydrophilic 5-FU owing to several obvious advantages over conventional dosage form. In the present study, the EL was successfully prepared using PC and surfactants (Span 60, Span 80 and Tween-80). It could be suggested that the concentration of Span 80 should have to be in optimum range to get maximum responses. There were magnified values of responses with increased concentration of Span 80 which might be due to dissolving and the extracting ability of surfactant of lipid from the SC. Span 80 has maximum fluidization property of the vesicle than Tween-80 in explored amount which could be due to the preferred placement of surfactant monomer into the lipid bilayer. A similar finding was reported with Tween-80 (>15%w/w) to deliver neomycin sulfate using the EL (Darwhekar et al., 2012). Optimized EL was selected using the desirability function to get the most robust formulation with desired quality and characteristics satisfying the maximum target of all responses within the given constraints. Figure 7(A–F) illustrated copious representative images of the desirability bar and corresponding 3-D contour plot of optimized three formulations. To achieve high desirability on increasing the concentration of both PC and surfactant, the effect of an independent variable is elicited with the red color of RS (at the apex of RS) plot (Figure 7(B), (D) and (F)). EL3-S80 with maximum desirability (0.9964) was selected among them for further study which satisfied the maximum desirability of responses.

Figure 7. Desirability curve of optimized three elastic liposomes. (A) Desirability bar graph of Span 60 and PC. (B) 3D-contour profile Span 60 and PC. (C) Desirability bar graph of Span 80 and PC. (D) 3D-contour profile Span 80 and PC. (E) Desirability bar graph of Tween-80 and PC. (F) 3D-contour profile Tween-80 and PC.

Preparation and characterizations of EL3-S80-loaded gel formulations

From the extensive literature, it was evident that no interaction of hydrophilic 5-FU was observed with lipophilic lipid bilayer membrane of the liposome. However, it has been reported that an increase in % EE on increasing the PC content must be owing to the larger size of vesicles or reduced size of lamellarity (Kaiser et al., 2003). Therefore, maximum vesicle size may load the higher amount of 5-FU in the aqueous compartment for enhanced permeation. Moreover, the negative value of zeta potential might be due to the alcoholic functional group present in PC and carbopol gel (acidic groups). 5-FU is a diprotic drug with two pK_a values (8 and 13) which indicate maximum protonated form in acidic medium leading to lipophilicity of molecule with reduced % EE. Moreover, high pH (∼8.0) value is not suitable for achieving maximum EE due to limited storage stability. Chemically, phospholipid exhibits the maximum stability against hydrolysis at pH 6.5 which triggered the need of optimum pH for maximum EE. Therefore, gel formulations were adjusted to final pH of 7.4 for maximum %EE and biocompatibility with the physiological condition after topical application (Grit et al., 1993).

A topical formulation required a certain amount of yield stress to show minimum resistance to flow after application and acquires certain yield stress to hold their shape until the applied shear exceeds the yield point. It is well-known fact that on increasing the concentration of the carbopol polymer, the viscosity decreases which results into decreased permeation of deformable vesicular carrier (Darwhekar et al., 2012). Furthermore, the higher value of percent spread by weight is the indicative of ease in spreadability on topical application. An increase in this value might be owing to the incorporation of suspension of EL3-S80 in the gel (blank) leading to reduced viscosity (Table 4).

In vitro skin permeation and drug deposition studies

The crystalline nature of the SC layer of the skin is the prime barrier for topical delivery of hydrophilic drugs which needs to be modified using formulation carrier with latent ability. The permeation potential of the gel formulations improved the percutaneous transport of the drug to a significant extent as compared to free DS. The values of flux and ER were found to be higher with ELA-3 and EL-S80 as compared to the DS respectively due to facilitated permeation across the SC. A similar finding was reported in the human SC with 6.2-fold higher flux when co-administered with resveratrol (Cosco et al., 2015). Limited permeation of 5-FU solution might be owing to hydrophilicity, polarity and low intercalating ability of 5-FU in lipid bilayer of EL across the SC (Tsukada et al., 1984; Singh et al., 2005). These are considered as the reasonable possible reasons for the low topical permeation and penetration of 5-FU. Experimentally, the permeation (diffusion) coefficient of 5-FU across the human SC was approximately an order of 10⁻⁷ cm²/h, responsible for limited permeation as evidenced with the logP (SC/water, logp∼0.46) value (Williams & Barry, 1991). Thus, this has provoked us to investigate the effect of surfactant and permeation enhancer in gel formulations. Hydrophilic Tween-80 provided weak flexibility to the lipid bilayer whereas Span 80 has a significant effect by incorporation into the lipid layer. This difference leads to augmented permeation flux rate of 5-FU across the skin when ultra-deformable EL vesicles were capable to get squeezed across the microscopic pores of the SC. The SC barrier was somewhat loosened with formulation carriers which could allowed further enhanced vesicle progression into the deeper layer. Azone and LA are considered as potential skin permeation enhancer via several reported mechanisms. These are fluidization of lipid by molecular insertion into the structure of intercellular lipid layer of SC, by structural changes, charge density and changed conformation state in the SC leading to increased diffusivity. Moreover, it exerts hydration effect on SC which is the major noticeable mechanism for hydrophilic 5-FU and promotes permeation. All these mechanisms may play together to exhibit the augmented effect when the gel with azone was applied topically. The gel might be responsible to provide sufficient hydration to the SC which further facilitates permeation. Our results were in agreement with the 3% azone used for increased permeability of 5-FU loaded in a liposome (Darwhekar et al., 2012).

Chemically, Azone and LA shared a common dodecyl side chain. However, the permeation enhancement by LA is much less as compared to azone which was a similar trend as reported earlier (Touitou & Abed, 1985; Darwhekar et al., 2012). This may likely cause improved solubility in fatty matrix of the SC which in turns lead to augmented the drug partitioning in the skin strata (Bhatt et al., 1991). Similar possible mechanism due to PG may be played in the skin due to alkanol.

Enhanced drug deposition and reduced systemic availability for reduced systemic side effects of 5-FU were investigated earlier using conventional liposome and deformable liposome without permeation enhancer after topical application (El Maghraby et al., 2001). The authors reported improved permeation of 5-FU through the human SC using deformable liposome composed of PC and sodium cholate. Presently, 5-FU deposition study was performed to determine the depot effect of gel for localized and sustained delivery for a cutaneous disorder. For effective therapeutic treatment of skin, dosage forms are required to be accumulated in the skin. Therefore, the study corroborated the maximum deposition of drug using an ELA-3 formulation which might be due to enhanced permeation potential of azone as compared to LA and PG. The highest deposition achieved by the same formulation might be the results of diverse ways permeation mechanism by Azone. The accumulated EL (5-FU) has been observed with the slow and sustained release as depot which led to localized effect. A similar finding was observed using cyclodextrin as an enhancer to deliver hydrophilic colchicines in an EL (Singh et al., 2010).

Determination of activation of energy

The temperature is a potential well-known factor to influence the drug release from the topically applied vehicle. In the present investigation, the Kp of 5-FU-loaded in carbopol gel containing permeation enhancer is linearly associated with the absolute temperature (T) which was measured at varied temperature. Furthermore, the gel viscosity represents the viscosity of the entrapped vesicular carrier in the gel network is principally accountable for determining the diffusion of 5-FU in the polymeric gel. As per Arrhenius plot, the graph revealed a linear relationship between the logarithm of Kp values and reciprocal of absolute temperature as shown in Figure 5. The activation energy was found to be 2.63-folds higher reduction than DS which suggested significant structural changes in prime SC barrier after topical application of 5-FU in EL gel. This interface further facilitated drug permeation across the rat skin for augmented drug delivery as compared to conventional topical dosage forms.

Skin irritation potential

It is noteworthy that the acceptability and applicability of topical formulation are limited by the patient if any erythema, edema and sign of irritation are elicited. Therefore, it is required to demonstrate any possibility of these reactions after topical application. Thus, the results suggested non-irritant and skin compatibility of gels containing permeation enhancer. Similar findings were reported in our preceding report using carbopol gel loaded with nanoemulsion of amphotericin B (Hussain et al., 2014b).

Conclusion

Delivery of 5-FU using conventional dosage forms displayed varied pharmacokinetic behavior and undesirable side effects to treat the cancerous skin and other skin disease such as keratoses. The present study has investigated an alternative topical approach using ultra-deformable elastic liposomal gel with optimum permeation enhancer to overcome the therapeutic limitations. Topical delivery of 5-FU-loaded EL gel showed substantial flux and deposition across and into the skin, respectively. Both vesicular elastic carrier and permeation enhancer played a concurrent role in enhancing the drug permeability across the rat skin for effective and efficient therapeutic efficacy. Moreover, more reduction in the energy of activation and higher flux, as well as drug deposition in the skin strata further, corroborated these intriguing aspects of topical delivery. Furthermore, the results of the Draize study suggested that employed gel containing explored optimum concentration of the permeation enhancer was found to be non-irritant and free from any hypersensitivity reactions. Thus, considering the optimized amount of suitable permeation enhancer may be useful in formulation development of 5-FU EL gel for cost effective commercial viable topical therapeutic drug delivery system as compared to the conventional dosage form.

Declaration of interest

Authors are responsible for the content of the article and report no conflicts of interest.