A Quality by Design approach to develop Topical Creams via Hot-Melt Extrusion Technology

Abstract

The advantages of hot-melt extrusion technology (HME) over conventional techniques to develop topical semisolids have been established. However, this technique is not widely used for semisolid production. Therefore, the aim of this novel work was to develop creams using the melt extrusion technology while applying Quality by Design (QbD) principles to study the effects of the extrusion process parameters on the product characteristics. The model drug selected was hydrocortisone acetate. A 2³ factorial design was considered for the factor influence study, which resulted in eight formulations to be extruded. Of the process parameters considered, the temperature of zone 2 had a significant influence on the work of adhesion of the creams. A similar permeation profile was obtained for all the formulations with the formulations following a diffusion based drug release mechanism. The results from the size distribution graph indicated stable cream formulations. In conclusion, this technology coupled with a design of experiments approach could be utilized to study how the extrusion process parameters could be modified to develop consistent topical creams with ideal product characteristics.

1. Introduction

In comparison to topical ointments, topical cream formulations are considered to be more acceptable to patients as they can be effortlessly applied on all regions of the skin without imparting the greasy feel that is inherently present in ointments. Also, in comparison to lotions, creams are more potent and occlusive for longer periods of time (Grisby and Bryant, 2000). Topical creams are conventionally prepared by the fusion process wherein the oil phase and the water phase components are separately heated and mixed in jacketed vessels before being combined with continuous agitation to form an emulsion but prolonged processing times, expensive equipment, consumption of large amounts of energy and scale-up issues are some of the challenges faced during the fusion process (Mendonsa et al., 2018).

The hot-melt extrusion process is a continuous process that consists of the polymers, drug and the other excipients being fed into the extruder and subjected to temperatures above the glass transition temperature of the polymers and/or over the melting temperature of the components while simultaneously being pushed forward towards the die by rotating screws to obtain a molecular type of mixing of the components involved (Mendonsa et al., 2017). In the development of topical formulations, this technique has already been established to be a cost efficient process (the end product obtained is devoid of air bubbles and hence does not require any de-aerators or cooling units), as well as short processing times. In addition, the high kneading and dispersing capacities of the mixing elements of the extruder have the ability to provide a uniform dispersion of the drug particles in the molten polymer matrix (which is crucial requirement for a scale-up process) as well as prevent the formation of agglomerates when the oil phase and the aqueous phase components are mixed together. These advantages have contributed to this process being considered as a suitable alternative over the conventional methods of manufacturing (Mendonsa et al., 2018).

The idea of Quality by Design (QbD) as a strategy to acquire an in depth understanding of the pharmaceutical manufacturing process at different stages of the initial product development process and the commercial cycle has been encouraged by the global regulatory agencies such as the United States Food and Drug Administration (US FDA) and the European Medicine Agency (EMA). Through this approach, one would acquire a good understanding of the product, its process design and further improvement, the scale-up parameters and optimize and control these steps, therefore improving the efficiency of the process and the quality of the product. Through the conventional method of Quality by Testing (QbT) approach, one would expect batch failures that would frequently occur with no possible cause wherein, in the quality by design approach, operational flexibility was possible provided the product variables and the process parameters remained within the approved design space (Kayrak-Talay et al., 2013; Patil et al., 2015; Patil et al., 2016).

The entire body is a target for side effects (glaucoma, weight gain, high blood pressure, etc.) instead of a specific area when corticosteroids are administered through the oral route. When inhaled, corticosteroids could accumulate in the mouth and in the throat regions causing irritation and fungal infections. When given through the systemic route, it can result in temporary side effects such as skin thinning and insomnia. Hence corticosteroids are preferably administered through the topical route to avoid significant side effects. Hydrocortisone Acetate (HA), a synthetic corticosteroid, is utilized for cutaneous disorders such as inflammatory dermatitis and is preferred over other low dose corticosteroids as it possesses a fast onset of action and therefore, was selected as a suitable drug for this study (Fini et al., 2008).

With the conventional methods of topical cream preparation having several disadvantages, the novel aim of this work was to develop topical cream formulations via the melt extrusion technology using a quality by design approach (using the Design of Experiment tool) to study the effects of the processing parameters of this technology on the characteristic properties of the final product.

2. Materials and Methods

2.1. Materials

Stearic acid and potassium hydroxide were a kind gift from Acros Organics (NJ, USA). Stearyl alcohol (Kolliwax SA) was a gift from BASF (Ludwigshafen, Germany). Cetyl alcohol (Crodacol C95) was purchased from Croda Inc. (NJ, USA). Hydrocortisone Acetate was a kind gift from PCCA (TX, USA). Glycerol was purchased from Fisher Scientific (NJ, USA). Propyl paraben was purchased from Spectrum (CA, USA). The other reagents used in this work were of analytical grade.

2.2. Preliminary study and Design of Experiment

The preliminary studies were conducted to identify the process variables that would have a significant impact on some of the characteristic properties of the cream formulations. Based on preliminary extrusion studies, three process variables were identified which were the temperature of zone 2, screw speed and screw configuration. Each of these factors were studied at two levels i.e. a lower and a higher level (Table 1). A factor influence study was then conducted to quantify their individual influence as well as to study how each factor is influenced by other factors. A factorial design that involves the selection of process parameters as well as response variables can be utilized to conduct a factor influence study in which all levels of one factor can be combined with all levels of other factors. Through a factorial design, one can separate the factors that are significant from those that are not as well as investigate the interactions between the selected factors, if any. Based on the process parameters selected as well as the levels considered for each factor, a 2³ factorial design was considered for this study. The response variables (characteristic properties) selected were yield stress, globule size and work of adhesion (Korakianiti et al., 2000; Singh et al., 2005). The factorial design resulted in the extrusion of 8 cream formulations (Table 1). Normality was confirmed through the results of the Box-Cox plot that suggested that there was no transformation required for the data to be used in this statistical design study.

Table 1.

A. Experimental design. B. Experimental factors with their levels

A.	Factor	Factor significance	Level (−1)	Level (+1)
	X1	Temp. of zone 2	40	55
	X2	Screw speed	50	70
	X3	Screw configuration	A	B
B.	Formulation	X1	X2	X3
	F1	55	70	A
	F2	40	50	B
	F3	40	50	A
	F4	55	50	B
	F5	55	70	B
	F6	40	70	A
	F7	40	70	B
	F8	55	50	A

2.3. Hot-melt extrusion

The composition of the cream formulation is shown in Table 2 (Pillai et al., 2001). Stearic acid, stearyl alcohol, cetyl alcohol and the drug which constituted of the oil phase were sieved through US# 35 mesh screen and blended using a V-shell blender (Maxiblend, Globe Pharma, New Brunswick, NJ). The blend was then fed to a co-rotating twin screw extruder through a volumetric feeder having an extrusion die of diameter 5.79mm (Process 11, Thermo Fisher Scientific, Karlsruhe). Glycerol, propyl paraben, potassium hydroxide and water that made up the aqueous phase were pre-heated to 70°C on a hot plate prior to feeding the mixture to the extruder. The pre-heated aqueous phase was then added into the extruder via an injection port connected to a peristaltic pump at zone 4. Figure 1 demonstrates the extrusion process carried out to develop the cream formulations. The flow rate of the oil phase and the water phase was set to 3 rpm and 4.8 ml/min respectively based on preliminary extrusion studies conducted. The temperature of the barrel was gradually decreased and set as follows: zone 2 – 40/55°C, zones 3-4 - 70°C, zone 5 - 55°C, zone 6 - 50°C, zone 7 - 45°C and zone 8 and die − 40°C. The screw configurations considered for the study are also illustrated in Figure 1.

Table 2.

Composition of the HC topical cream

	Ingredient	Weight %
Oil Phase	Stearic Acid	11
	Stearyl Alcohol	0.85
	Cetyl Alcohol	0.85
	Hydrocortisone Acetate	1
Water Phase	Glycerol	0.8
	Propyl Paraben	0.12
	Potassium Hydroxide	0.34
	Water	85.04

Figure 1.

Modified screw configurations (A & B) and schematic representation of the HME process (C).

2.4. Characterization of cream formulations

2.4.1. Texture analysis

A texture analyzer with a 5 kg load cell (Model TA.XT2i, Texture Technologies Corp. /Stable Micro Systems) was used to conduct the textural analysis for the creams. A cylindrical acrylic probe having diameter of 1in (TA-3) with a soft matter kit (TA-275) was used. The sample was placed onto the soft matter fixture and the cylindrical probe was attached to the arm of the texture analyzer. Prior to the experiment, the surface of the sample was levelled by swiping any extra cream formulation. The probe was brought down at a speed of 1mm/sec until it touched the sample surface. On touching the sample surface and detecting a trigger force of 5g, the probe produced a deformation of 2mm while moving at a speed of 2mm/sec. The probe was then lifted off the sample surface at a speed of 10mm/sec. Prior to the next analysis, the probe and the soft mater fixture were cleaned (Mendonsa et al., 2018).

2.4.2. Rheology analysis

The rheological study for the formulations was conducted on a TA instrument HR-2 rheometer. A temperature of 32°C was set using a Peltier system. A parallel plate geometry was used (20mm) with the upper and the lower plates having adhesive backed sand papers (grit number #600, Allied High Tech Products) attached to them to decrease the slippage of the sample at the sample-rheometer plate interface. The lower plate had 400mg of the sample applied onto it during which the upper plate was at a gap of 550μm. Having removed the excess sample, the upper plate was then set at a gap of 500μm for the experiment. To reduce possible water loss during the testing, the rheometer was attached to a solvent trap. The study was conducted in four steps: the first step was a time sweep study for 5 minutes for the sample to relax from the stress it acquired during loading (strain (γ₀) of 0.1% and frequency (ω) of 1Hz). The second step was the strain sweep test conducted at γ₀ range of 0.05–50% with ω of 1Hz. The third step was the time sweep study conducted prior to the steady shear test, which comprised the fourth step wherein, the shear rate was steadily increased to a maximum shear rate of 100s⁻¹ in the forward mode and then steadily decreased to a minimum shear rate of 0.002s⁻¹ in the backward direction (Krishnaiah et al., 2014). To elaborate on the fourth step methodology, the fourth step consisted of the flow sweep experiments and were conducted both in the forward and backward shear rate sweep experiments, wherein the sample was subjected to increasing and decreasing shear-rate, respectively. In the forward mode, the shear-rate was varied from 0.002 s⁻¹ to 100 s⁻¹ with 10 data points per decade. For the data collection, the equilibration time of 5 seconds and averaging time of 30 seconds were used, which together added up to 35 sec per data point. Over the range of the shear rate, 47 data points were collected in 1683 secs. Similarly, in the backward flow sweep experiment, the shear rate ranging from 100 s⁻¹ to 0.002 s⁻¹, with 47 data points in 1648 secs were collected using 35 secs per data point.

2.5. Analytical method

A Waters high performace liquid chromatography (HPLC) UV (Waters Corp, Milford, MA) system equipped with a dual λ absorbance detector with a Waters C₁₈ (4.6 × 150mm; 5μm particles) column was employed to detect the HA content. The mobile phase comprised of Methanol and Water (67:33v/v) with a set flow rate of 1ml/min. The injection volume was set to 10 μl with the UV detector set to 248nm (Pedersen et al., 2000).

2.6. In-vitro drug transport across the silicone membrane

The in-vitro release study was performed on vertical franz diffusion cells having an active diffusion area of 0.64cm² (Logan Instruments, Somerset, NJ). A silicone membrane (non-porous and hydrophobic) having a thickness of 0.005” (Speciality Manufacturing, Inc., Saginaw, MI) was considered to be a suitable membrane for the study. The cream formulation (100mg) was applied onto the membrane and was placed between the donor and the receptor compartments. The receptor compartment consisted of phosphate buffer saline solution (PBS) at a pH of 7.4 (degassed prior to the experiment) to simulate systemic conditions and the cells were constantly stirred with a temperature set to 32°C (Ueda et al., 2009). At regular time intervals until the 10^th hour, 0.2 ml of the samples were withdrawn and replaced with the same volume of Phosphate Buffer Saline (PBS). The samples were then analyzed using High Performance Liquid Chromatography (HPLC).

2.7. Ex-vivo permeation study

The ex-vivo permeation studies were performed on vertical franz diffusion cells having diffusion area of 0.64cm² (Logan Instruments, Somerset, NJ). The membrane used for this study was a freshly slaughtered porcine epidermis which was obtained from a local abattoir and was kept in 8°C until the day of the experiment. Cream formulations were uniformly applied onto the epidermis which was placed between the donor and receptor compartments. The receptor compartment consisted of PBS 7.4 (degassed before experiment). A temperature of 32°C was set with the cells being provided with constant stirring. Prior to the experiment, the cells were kept for equilibrium for 1 hour. At regular time intervals (up to 10 hrs), 0.2ml of the sample was withdrawn and was replaced with fresh PBS of the same volume. The samples were then analyzed through the HPLC system (Şenyiğit et al., 2011).

2.8. Extraction of HC from the epidermis

At the end of the permeation study, the epidermis was rinsed using methanol and water to remove the leftover formulation. The active diffusion area was then obtained using a biopsy punch and weighed. After adding appropriate amounts of 2N NaOH and placing it in an oven set to 50°C, the active epidermis was homogenized. Appropriate dilutions are made with acetonitrile after which it was subjected to vortex and centrifugation. The supernatant collected was then analyzed for drug content through the HPLC system (Maurya and Murthy, 2014).

2.9. pH measurement

pH of the cream formulations were evaluated through the Mettler Toledo InLab® Micro pH probe (Mettler Toledo, Columbus, OH).

2.10. Uniformity of drug content

Weighed amounts of the formulations were collected from three regions of the container and were dissolved using appropriate amounts of acetonitrile and sonicated. Samples were then analyzed for drug content through the HPLC system.

2.11. Globule size measurements and stability studies

The formulations were stored at room temperature and at conditions of 40°C/75% RH for time periods of 1 and 3 months. The globule size for the formulations were determined using an Olympus IX inverted research microscope. A uniform layer of the sample was obtained by applying 15mg of the creams onto a glass slide and slowly pulling the sample to form a thin layer with the aid of a second glass slide. This slide was then studied under the microscope with a 10X objective lens. The images obtained from the microscope were then analyzed using the Zen Lite software. The globules were manually counted and an average of 50 randomly selected globule diameters for each formulation were considered for the design of experiment (DOE) study. A size distribution curve (d₁₀, d₅₀, d₉₀) at room temperature and 40°C/75% RH was plotted to determine the stability of the formulations.

2.12. Statistical Data analysis

Statistical significance between formulations were evaluated through a one-way analysis of variance (ANOVA) test followed by a Tukey test. If the p-value obtained was less than 0.05, the formulations were considered to be significantly different. A mean of three readings were considered.

3. Results and Discussion

3.1. Selection of critical process parameters and critical quality attributes

Cream formulations intended for topical use have different patient tolerance levels than oral or parenteral formulations. Some of the properties of cream formulations that influence the acceptance levels include textural profile (appearance, odor and extrudability), spreadability, tackiness, greasiness after application, etc. (Krishnaiah et al., 2014). Yield stress which is defined as the minimum amount of shear stress required to initiate an irreversible plastic flow in a product has been known to have a direct influence on the spreadability of the cream formulations (Barry and Grace, 1972). Besides the spreading properties, yield stress also affects the physical stability of the product. In addition, a formulation with a smaller globule size would indicate a higher drug release and physical stability (Avachat and Patel, 2015; Patel et al., 2013). Hence, yield stress, work of adhesion and globule size were selected as the critical quality attributes (CQA) /response variables for this study. The mixing process in a twin screw extruder is generally comprised of dispersive and distributive mixing. The staggering angle and the width of the disks of the mixing elements have an impact on the mixing processes, both of which contribute to a homogenous product. The mixing process as well as the screw speed contribute to the shear produced in the extruder with a higher dispersive mixing and a higher screw speed contributing to a greater shear (Shibata et al., 2009; Van Melkebeke et al., 2008). It is critical that the oil phase components be completely melted before the addition of the aqueous phase to avoid recrystallization of the oil phase ingredients (thereby resulting in a non-homogenous product), but during preliminary extrusion studies, it was observed that when the temperature of zone 2 was set to 70°C (melting temperature of stearic acid), a molten mass at zone 2 was formed and hence the blend was unable to move forward. Therefore, lower temperatures of 40 and 55°C were used instead. The screw configuration (modified), screw speed and the temperature of zone 2 were hence selected as critical process parameters (CPP) for this study.

3.2. Experimental design

A two-level factorial design is the easiest form of orthogonal design that is often used for screening and factor influence studies. Some of the advantages factorial designs offer include being efficient in determining the main effects when interactions are not present, calculated effects and interactions are not dependent on the effects of other factors (orthogonality) and conclusions can be made over a range of conditions as factor effects are determined by changing levels of other factors (Bolton and Bon, 2009). Based on the analysis of the factors on the responses, linear equations were derived for each of the responses and are given as:

$${\mathrm{Y}}_1=42.75-5{\mathrm{X}}_1-1.50{\mathrm{X}}_2+4.50{\mathrm{X}}_3$$ (1)

$${\mathrm{Y}}_2=47.21+6{\mathrm{X}}_1-0.58{\mathrm{X}}_2+2.20{\mathrm{X}}_3$$ (2)

$${\mathrm{Y}}_3=0.922-0.22{\mathrm{X}}_1+0.23{\mathrm{X}}_2-016{\mathrm{X}}_3$$ (3)

Where Y₁, Y₂ and Y₃ stand for the yield stress, work of adhesion and the globule size and X₁, X₂ and X₃ stand for temperature of zone 2, screw configuration and screw speed respectively. The first term of each equation represents the arithmetic mean of the responses of the 8 formulations. The purpose of these equations is to describe the effect that each factor has on the response variables. The magnitude (numerical value) and the direction (signs) of the factor coefficients in equations 1-3 describe the nature of the effect (i.e. whether the response variables would increase or decrease with change in process parameters). A higher value of magnitude implies a greater impact on the response variable.

3.3. Effect of temperature of zone 2

The temperature of zone 2 was varied at two temperatures: 40 and 50°C. From Eqns. (1-3), we observe that as the temperature was increased, work of adhesion also increased as indicated by the positive sign for the coefficient X₁ in Eqn. (2). For both the yield stress and the globule size variables, a negative effect was observed Eqns. (1 and 3). Since the p-value obtained was less than 0.05, the effect was not only positive but significant as well (Table 3). Some of the key factors to consider while studying the work of adhesion for a semisolid formulation include its viscosity, rate and time of shear that occurs during spreading, temperature at the site of application, etc. (Garg et al., 2002). Yield stress is indirectly proportional to the work of adhesion of the product (higher the yield stress, lower would be the work of adhesion). A desired trend in both the yield stress and the work of adhesion was observed with increase in temperature of zone 2 as demonstrated in Figure 2. As mentioned earlier, it is mandatory that the oil phase be completely melted prior to the addition of the aqueous phase. In this study, the oil phase had two zones (zones 2 and 3) to melt and form a homogenous blend before the addition of water at zone 4. It is possible that when the temperature was increased from 40 to 50°C, it aided in the efficiency of mixing and melting of the oil phase ingredients therefore having a possible impact on the textural and the rheological properties of the cream. The effect of temperature on the consistency of a w/o cream was also demonstrated by Tamburic et al. (1996) wherein the temperature used in the preparation of the creams by the cold emulsion technique (in comparison to the hot emulsion technique) resulted in a favorable product.

Table 3.

ANNOVA table for the response variables.

Source	Sum Square	df	Mean Square	F-value	p-value Prob>F
Yield Stress (Y1)
Model	380	3	126.67	1.77	0.29
X1	200	1	200	2.80	0.16
X2	18	1	18	0.25	0.64
X3	162	1	162	2.27	0.20
Work of athesion (Y2)
Model	329.25	3	109.75	4.25	0.09
X1	287.94	1	287.94	11.14	0.02
X2	2.71	1	2.71	0.10	0.76
X3	38.60	1	38.60	1.49	0.28
Globule size (Y3)
Model	1.07	3	0.36	0.54	0.67
X1	0.40	1	0.40	0.61	0.47
X2	0.44	1	0.44	0.68	0.45
X3	0.22	1	0.22	0.34	0.58

Figure 2.

Effect of temperature of zone 2 on Work of Adhesion (WOA) (A) and Yield Stress (B).

3.4. Effect of screw configuration

Two screw configurations were used in this work: screw configuration A and screw configuration B. Based on Eqns. (1-3), it was noted that a positive impact was observed only on the globule size variable with change in screw configuration Eqn. (3) whereas a negative effect was observed on the yield stress and the work of adhesion variables Eqns. (1 and 2). But despite the positive impact, with the p-value being greater than 0.05, the effect on the globule size was not considered to be significant (Table 3). The primary difference in the two configurations was the extension of the second mixing zone as can be seen from Figure 1. Mixing zones are generally considered to possess low conveying ability, thereby extending the residence time of the blend within the extruder at which time the material is subjected to additional mixing which would further reduce the globule size of the formulations. Extension of mixing zones can give rise to high torque values but during the extrusion runs, a minimum torque of 3-4% was maintained in the barrel.

3.5. Effect of screw speed

Two screw speeds were used in this study: 50 and 70rpm. From Eqns. (1-3), it was observed that a positive impact on yield stress and work of adhesion Eqns. (1 & 2) were obtained on modifying the screw speed. In spite of the favorable impact, as the p-value was more than 0.05, therefore the effects on the yield stress and the work of adhesion were considered to be insignificant. Screw speeds are associated with residence time of the physical blend in the extruder, wherein, when the extruder is operated at lower screw speeds, due to an increase in the residence time, the physical blend would experience a higher input of thermal energy in the presence of a low shear environment and therefore be subjected to low shearing action. Having being subjected to increased thermal energy, the resulting product would undergo a good expansion. On the other hand, high screw speeds would indicate low residence times but with an increase in the rate of shear (Bhattacharya, 1997). As the screw speed is gradually increased, the magnitude of torque decreases primarily due to the decrease in the filled length within the extruder and an increase in the rate of shear that would reduce the apparent viscosity of the material inside the extruder (due to the material behaving as a pseudo plastic material which demonstrates shear thinning behavior) (Guha et al., 1997; Bhattacharya et al., 1992). The screw speed and the screw configuration together contribute to the shearing rate which is graphically demonstrated in Figure 3. Herein, it is observed that with increase in shear rate, the viscosity of the formulations decreases which would indicate ease in the spreadibility of the formulations.

Figure 3.

Viscosity vs. shear rate for the formulations. Each value represents the mean (n=3).

3.6. In-vitro drug release

The in-vitro drug release rate can demonstrate the effect of both the physical and the chemical parameters (solubility, particle size and concentration of the active pharmaceutical ingredient) together with the rheological characteristics of the dosage form (such as viscosity). From Table 4, it could be observed that F7 and F5 had slightly higher values of drug release in comparison to the other formulations. The viscosity of the formulation has a direct effect on the drug release on a microstructural level with increase in viscosity leading to either an increase or decrease of the drug release from the formulation. Several authors have established that an increase in viscosity could have a negative impact on the drug release (Adeyeye et al., 2002; Desai and Blanchard, 1998; Jones et al., 1999; Tas et al., 2003). Lower yield stress values have been known to contribute to lower drug release values as well (Desai and Blanchard, 1998; Jones et al., 1999). However, this theory could not be applied to formulations F5 and F7 as their yield stress values were similar to the values of the other formulations (data not shown). Moreover the viscosity obtained for all the formulations were similar (Figure 3). The drug release from the creams was fit into the zero order, first order and the Higuchi models. The results gave a best fit to the Higuchi model as the linear correlation coefficient value (R²) obtained was closer to 1 as is observed from Table 4 in comparison to the other models (data not shown). The Higuchi equation is given in Eqn. (4):

$$Q_t=K_h\mathbf{x}{\mathrm{t}}_{1∕2}$$ (4)

Table 4.

Amount of drug release and Kh values for formulations F1-F8. Each value represents the mean (n=3).

Formulation	Amount of drug release (μg/cm2)	R2 values	Kh value
F1	29.45	0.95	8.31 ± 0.66
F2	31.25	0.94	8.86 ± 0.98
F3	27.24	0.94	6.81 ± 0.75
F4	25.13	0.92	7.37 ± 0.62
F5	36.28	0.95	10.36 ± 0.97
F6	25.62	0.72	7.18 ± 0.72
F7	48.15	0.89	15.80 ± 0.89
F8	30.70	0.96	9.84 ± 0.02

Where, Q_t stands for the amount of drug release within time’t’ and K_h stands for the Higuchi diffusion rate constant (Mendonsa et al., 2018). From Figure 4 which denotes the plot of Q vs. √t, K_h was calculated and is represented in Table 4. The formulations were not considered to be statistically significant as the p-value obtained was 0.320, which was more than 0.05.

Figure 4.

Plot of Q vs. √t for the calculation of Higuchi diffusion rate constant. Each value represents the mean ± S.D. (n=3).

3.7. Ex-vivo permeation study

Figure 5 represents the results of the ex-vivo permeation study conducted for the eight formulations. The steady-state flux and the drug retained in the epidermis was also calculated and presented in Table 5. With the p-value (0.842) being greater than 0.05, there was no significant difference obtained in the permeation of drug from the formulations. The steady state flux could be explained through the Fickian equation which describes the passive permeation of a permeant through the strateum corneum as the amount of drug that crosses through the membrane equals to the amount of drug exiting the membrane and is given by the below equation –

$${\mathrm{J}}_{\text{ss}}={\mathrm{K}}_{\mathrm{p}}\mathbf{x}{\mathrm{C}}_{\text{veh}}$$ (5)

Figure 5.

Cumulative amount of HC permeated vs. time for the creams. Each value represents the mean ± S.D. (n=3).

Table 5.

Amount of drug permeated, flux and drug retained in the epidermis values for formulations F1-F8. Each value represents the mean (n=3).

Formulation	Amount of drug permeated (ng/cm2)	Flux (ng h-1.cm-2)	Drug retained in the epidermis (ng of drug/mg of skin)
F1	3413.22 ± 336.45	292.70 ± 60.75	116.96 ± 31.97
F2	3349.90 ± 475.64	336.96 ± 773.54	233.39 ± 28.78
F3	3413.42 ± 565.96	331.91 ± 375.97	205.04 ± 43.29
F4	4719.92 ± 1096.39	435.50 ± 423.95	186.60 ± 7.41
F5	4284.80 ± 308.04	360.84 ± 807.86	338.20 ± 48.68
F6	5098.76 ± 1081.52	494.78 ± 437.15	250.90 ± 5.27
F7	3351.79 ± 1621.84	286.04 ± 414.52	222.35 ± 61.48
F8	5039.568 ± 685.658	457.56 ± 470.01	367.08 ± 73.45

Where, J_ss stands for the steady-state flux (ngh⁻¹.cm⁻²) over a membrane having thickness of ‘h’ cm, C_veh stands for the drug concentration (mg.cm⁻³) and K_p stands for the drug permeability coefficient that is dependent on the formulation (Naik et al., 2000). From the plot of cumulative amount of drug permeated vs. time, the flux was calculated as the slope of the linear portion of the curve. Several factors affect the permeation of the drug through the skin such as the pH and viscosity of the formulation, thermodynamic activity and solubility of the drug in the vehicle, penetration enhancers (type and concentration), etc. (Benson and Watkinson, 2012). Frum et al. (2008) demonstrated that hydrocortisone gel formulations demonstrated a lower drug flux in comparison to its aqueous solution primarily due to the viscous nature of the gels that prevented its flow into the follicular orifices. The drug must first diffuse out from the vehicle in order for the drug permeation to take place. Therefore, increasing the viscosity of the formulation would not only decrease the drug release but would also negatively impact the permeation through the skin. It has also been demonstrated that the efficiency of incorporation of the drug into the matrix would depend on the manufacturing conditions (such as type of equipment and mixer speed) and this would directly impact the drug permeation (Passerini et al., 2009). Two penetration enhancers were used in the formulation – stearic acid and cetyl alcohol. In general, the permeation mechanism of fatty acids (stearic acid) and fatty alcohols (cetyl alcohol) is the disruption of intercellular lipid packing and disruption of lipids that are densely packed in the extracellular spaces of the stratum corneum respectively (Rossi et al., 2000; Sanna et al., 2010).

3.8. pH measurements and uniformity of drug content

The pH and the drug content of the cream formulations is listed in Table 6. The pH obtained was well within the accepted range for the pH of topical semisolid products (4.0 – 8.0). The dispersive and the distributive mixing together contribute to the homogenous distribution of the drug in the molten polymer matrix which is one of the major factors to be considered during the development of topical semisolid products. Low values of standard deviation obtained for the formulations demonstrate that the process parameters selected for the melt-extrusion process resulted in formulations having a uniform distribution of drug in the polymer matrix.

Table 6.

pH and drug content for the formulations F1-F8. Each value represents the mean (n=3).

Formulation	pH	Drug content (%)
F1	7.01 ± 0.10	102.41 ± 3.10
F2	6.70 ± 0.04	100.60 ± 1.08
F3	6.77 ± 0.11	101.39 ± 4.74
F4	6.87 ± 0.05	99.61 ± 0.83
F5	6.85 ± 0.04	100.69 ± 1.18
F6	6.88 ± 0.09	102.19 ± 2.51
F7	6.99 ± 0.09	100.62 ± 0.59
F8	6.04 ± 0.04	100.88 ± 0.57

3.9. Globule size measurements and stability studies

The absence of creaming and coalescence of the internal phase along with maintaining its elegance with respect to its physical properties such as appearance and color are some of the desired features of a pharmaceutical emulsion with reference to its physical stability. Physical stability is determined by studying creaming and/or coalescence that takes place over a period of time. Coalescence would result when the mean globule size would gradually increase with time. Figure 6 represents the size distribution graph of the eight formulations at room temperature and at storage conditions of 40°C/75% RH. To compare relative stabilities of similar formulations, creaming and coalescence processes must be expedited by storing the formulations at higher temperatures. Based on the graphs, it could be observed that at room temperature, the formulation F1 was relatively more stable in comparison to the other formulations and at 40°C/75% RH conditions, formulations F2, F3 and F7 were relatively more stable than the other formulations. Overall, it could be inferred that changes in the processing conditions did result in stable cream formulations at both of the storage conditions.

Figure 6.

Size distribution graph (d₁₀, d₅₀, d₉₀) at RT and 40°C/75%RH for the formulations.

Conclusions

In this study, topical creams were prepared by the hot-melt extrusion technology and a factorial design study was employed to investigate the effect of the processing parameters on the product characteristics of the formulations. Of the selected process parameters, only the temperature of zone 2 showed a statistically significant impact on the work of adhesion of the creams. Although statistically insignificant, screw speed did have a positive impact on the yield stress and work of adhesion with the change in screw configuration showing a positive impact only on the globule size. The release and the permeation profile for the formulations were similar with the drug release from the formulations following a diffusion based mechanism. From the size distribution graph, it could be inferred that despite the change in the process parameters, stable cream formulations could still be obtained. In conclusion, with the advantages that the extrusion technology has to offer over the conventional techniques, this technology can be successfully utilized to develop topical creams. However, with the technology still being widely investigated in the field of topical semisolid production, an ideal way to assess using this technique would be to combine the hot-melt extrusion technology with a quality by design approach. Through the this study, one could then identify the key quality attributes from a patient’s point of view, which would then translate to ideal product characteristics. An in depth study would be conducted to determine how process parameters could be differed to consistently develop a product that would possess ideal product characteristics, thereby fulfilling the aim of utilizing the quality by design approach, which would ensure product quality at all times.