Formulation and processing of solid self-emulsifying drug delivery systems (HME S-SEDDS): A single-step manufacturing process via hot-melt extrusion technology through response surface methodology

Abstract

The objective of the current study is the formulation development and manufacturing of solid self-emulsifying drug delivery systems (HME S-SEDDS) via a single-step continuous hot-melt extrusion (HME) process. For this study, poorly soluble fenofibrate was selected as a model drug. From the results of pre-formulation studies, Compritol^® HD5 ATO, Gelucire^® 48/16, and Capmul^® GMO-50 were selected as oil, surfactant and co-surfactant respectively for manufacturing of HME S-SEDDS. Neusilin^® US2 was selected as a solid carrier. The design of experiments (response surface methodology) was employed to prepare formulations via a continuous HME process. The formulations were evaluated for emulsifying properties, crystallinity, stability, flow properties and drug release characteristics. The prepared HME S-SEDDS showed excellent flow properties, and the resultant emulsions were stable. The globule size of the optimized formulation was 269.6 nm. The DSC and XRD studies revealed the amorphous nature of the formulation and FTIR studies showed no significant interaction between fenofibrate and excipients. The drug release studies showed significant (p < 0.05) improvement in solubility compared to the pure drug (${\text{DE}}_{15}=45.04$ for the optimized formulation), as >90% of drug release was observed within 15 min. The stability studies for the optimized formulation were conducted for 3 months at 40 °C/75% RH.

1. Introduction

Most new chemical entities (70–90%) and 30% of commercially available products exhibit poor water solubility and poor bioavailability (Nikolakakis and Partheniadis, 2017). The low solubility and reduced poor gastrointestinal (GI)-stability leads to poor bioavailability, high intra-subject/inter-subject variability and deficiency of dose proportionality (Bernkop-Schnürch, 2013; Lipinski, 2002). Thus, the poor solubility of the drug causes hindered dissolution, which is a rate-limiting step in absorption (Palmer, 2003). Various strategies such as complexation with cyclodextrin (Ammar et al., 2006), solid dispersions (Farmoudeh et al., 2020; Mande et al., 2017; Weuts et al., 2004), and lipid-based drug delivery systems (Odeberg et al., 2003) have been reported in the literature to improve the solubility to overcome the challenge of the poor solubility of the drugs. Among the lipid-based drug delivery systems (LBDDS), self-emulsifying drug delivery systems (SEDDS) acquired special attention as they have been efficient in improving the oral bioavailability of many poorly water-soluble drugs and improving the pharmacokinetic profiles (Nielsen et al., 2008).

SEDDS improve the solubility of poorly soluble drugs by preserving the drug in the solubilized state and in the form of finely dispersed droplets in the transit through the gastrointestinal tract (GIT), thus improving the bioavailability (Pouton, 2000). SEDDS typically consists of an isotropic mixture of oils (vehicle) and surfactants, optionally co-surfactants capable of forming fine oil in water (o/w) emulsions with the help of mild agitation from peristaltic movements in the GIT following dilution with the aqueous phase (Oh et al., 2011). Thus, SEDDS present the drug in the form of tiny emulsion droplets ($<5\text{μ}$), which contributes a large interfacial area for partitioning of the drug between the dispersed droplets and aqueous GI fluids, thereby improving the drug dissolution and absorption (Joshi et al., 2008; Nerurkar et al., 1996).

The lipids within the GIT undergo emulsification, enabling the drug’s dissolution via enhancing the effective interfacial surface area for drug release and promoting absorption (Abdalla et al., 2008). Additionally, certain self-emulsifying excipients and lipids enhance the drug permeability through the intestinal epithelium and lymphatic transport of the drugs into the systemic circulation (Joyce et al., 2019). The physicochemical properties and in-vivo performance of LBDDS can be significantly different based on the formulation aspects, drug load, and mechanism of solubilization of the lipids. Self-emulsifying drug delivery systems (SEDDS) have the potential for the commercial scale due to the flexibility in manufacturing and capacity to improve the pharmacokinetic profiles of poorly water-soluble drugs immensely. Many reports of promising in-vivo data highlight the efficiency of the SEDDS and vast commercial potential with successfully marketed products (I Jethara et al., 2014; Leonaviciute and Bernkop-Schnürch, 2015). The commercial SEDDS products are either soft gelatin capsule (Norvir^®, Sandimmune^®, Neoral^®, Agenerase^®, Depakene^®, Fortovase^®, Aptivus^®, Rocaltrol^®, Vesanoid^®, Accutane^®, Targretin^®) or liquids encapsulated in a hard gelatin capsule (Lipirex^®, Gengraf^®) (Park et al., 2020). One of the mechanisms through which LBDDS improves drug solubilization and absorption is stimulating the digestive juice secretion and thereby increasing the concentration of phospholipids, cholesterol and bile salts in the intestinal lumen (Kossena et al., 2007). The increased concentration of these endogenous substances results in complex colloidal structures (like vesicles, mixed micelles and liquid crystalline structures) by interacting with lipids and digested lipid products. The resulting colloidal structures affect the solubilization and distribution of drugs in the in-vivo digestive environment (Kossena et al., 2005). Therefore, the intermediate colloidal species resulting from lipid digestion play a vital role in the solubilization of poorly soluble drugs and possibly oral absorption.

The traditional liquid SEDDS has disadvantages such as low drug loading capacity, drug leakage, limited dosage forms, incompatibility of some excipients with capsules, stability issues and the likelihood of irreversible drug or excipient precipitation (Prajapati and Patel, 2007). The liquid SEDDS are converted into solid form to troubleshoot these issues by adsorbing them onto the inert solid carriers. In addition, this strategy of producing solid SEDDS (S-SEDDS) provides advantages such as good flow, manufacturing feasibility, dose accuracy and patient compliance (Tan et al., 2013). Thus, converting liquid SEDDS into solid SEDDS integrates the advantages of lipid-based drug delivery and solid dosage forms. Furthermore, solid SEDDS affords many advantages compared to the liquid precursor form (liquid SEDDS), such as enhanced safety, controlled drug release applications, improved dissolution, improved gastric residence and permeability, and feasibility of industrial manufacturing for commercial production. (Joyce et al., 2019; Sermkaew et al., 2013).

Hot-melt extrusion (HME) has been a widely used technology in the solubility improvement of poorly soluble materials and in producing amorphous solid dispersions. In HME technology, the active pharmaceutical ingredient (API) and excipients are subjected to specific thermal conditions associated with shear, which promotes the physicochemical interaction between the components of the formulation. As a result, the mobility of the drug molecules in the amorphous solid dispersions is inhibited, preventing the recrystallization of the API and thus improving the stability of the formulation (Baghel et al., 2016). Traditionally SEDDS are prepared as liquid dosage forms. However, there are reports in the literature where the liquid SEDDS are converted into solid SEDDS by adsorbing them onto solid carriers in a two-step process (Bajwa et al., 2022; Kanuganti et al., 2012; Lee et al., 2022). The current study focuses on the systematic development of solid SEDDS via HME technology and continuous manufacturing of solid self-emulsifying drug delivery systems (HME S-SEDDS). In this process, the simultaneous solubility of the API in the molten excipients occurs due to heat produced in the barrel in addition to the applied shear. The temporarily formed molten liquid SEDDS are distributed onto the porous solid carrier Neusilin^® US2 and instantaneously converted into stable solid SEDDS, collected at the discharge zone for downstream processing. Fenofibrate, a poorly soluble drug (log P = 5.3), was selected as the model drug, and excipients were chosen through solubility studies. Preformulation studies focused on excipient selection, ratio of solid carrier and lipid content, and flow properties. The final formulations were manufactured via a DoE (design of experiments) employing response surface methodology. The HME S-SEDDS were characterized for emulsifying properties, flow properties, crystallinity and dissolution. The study’s novelty is that it is designed to produce HME S-SEDDS in a single step by utilizing the heat and shear in the HME process. The physical mixture was directly introduced into the hot-melt extruder and the product was directly collected from the discharge zone. As residence time is short, the process is cost-effective and viable for further scale-up. In hot-melt extrusion technology, systematic development and continuous manufacturing of solid SEDDS (HME S-SEDDS) is an unexplored area. The formulation development and analysis were performed systematically, and the statistical approach (response surface methodology) was employed to optimize the formulations. This novel exploratory study can provide valuable insights for formulation development and continuous manufacturing of solid SEDDS via HME technology. Per our knowledge, there are no reports of systematic formulation development and continuous manufacturing of solid SEDDS in a single step via HME technology in literature till now. As a continuous single-step process, this improves the ease of manufacturing and provides a more economical solution for commercial production.

2. Materials and methods

2.1. Materials

Fenofibrate (API) was donated by Ashland Inc. (Lexington, KY, USA). Compritol ^® HD5 ATO (Behenoyl polyoxyl-8 glycerides), Gelucire^® 48/16 (Polyoxyl-32 stearate (type 1), Gelucire^® 50/13 (Stearoyl polyoxyl-32 glycerides), Geleol^™ (Glycerol Monostearate 40–55 (Type I)), Labrafil^® 2130CS (Lauroyl polyoxyl-6 glycerides) were gift samples from Gattefosse (NJ, USA). Capmul^® GMO-50 (Glyceryl Monooleate with di and triglycerides) was a gift sample from Abitec (OH, USA). Kolliphor^® 407 (co-polymer of polyethylene oxide and polypropylene oxide), Cremophor RH^® 40 (PEG-40 Hydrogenated Castor Oil) were purchased from BASF (NJ, USA). Sodium lauryl sulfate, Sodium carboxymethyl cellulose, Polyethylene glycol 3350, Polyethylene glycol 6000, and TPGS (D-$\alpha$-Tocopherol polyethylene glycol 1000 succinate) were purchased from Sigma Aldrich (MO, USA). Neusilin^® US2 (Magnesium aluminometasilicate) was generously donated by Fuji chemicals (Japan). Gelatin capsules were purchased from Total Pharmacy Supply (Arlington, TX, USA). All the other chemicals employed in this study were of analytical grade.

2.2. Pre-formulation studies

2.2.1. Selection of excipients through solubility studies

The solubility studies were conducted for binary mixtures of API and excipients using a method which is a slightly modified method developed by Gattefosse, U.S.A. The binary mixtures of excipients with fenofibrate were prepared at a different drug: excipient ratios, represented in Table 1. First, the excipient to be tested was taken in a glass vial and heated to a temperature of 5–10 °C above the melting point (M. P) of the excipient to melt the excipient. Then fenofibrate was added to the vial and shaken on a mechanical shaker with a temperature to maintain the excipient in molten condition. The binary mixture was shaken for 1 h, and then the binary mixture was kept at ambient temperature. In the next step, the binary mixtures were analyzed via DSC studies to determine the saturation solubility of the drug in the respective individual excipient. Based on the results of solubility studies the final excipients of the formulation were selected for the development studies.

Table 1.

Details of solubility studies (mean ± SD; n = 3).

Excipient	Melting point of excipient (°C)	Range of solubility tested (mg (API)/g of excipient)	Saturation solubility(experimental value (mg of API)/g of excipient))	Enthalpy valueat maximum solubility range (J/g)
Compritol® HD5 ATO	60–69	Up to 100 mg	50–75	103.14–119.64
Labrafil® L2130CS	33–38	Up to 100 mg	N/A	N/A
Geleol™	54–63	Up to 100 mg	–	N/A
Gelucire® 50/13	50–55	Up to 100 mg	–	N/A
Cremophor RH® 40	N/A	Up to 200 mg	N/A	N/A
Gelucire® 48/16	46–50	Up to 200 mg	125–150	77.79–119.07
TPGS	75	Up to 150 mg	10–25	92.01–105.52
PEG 6000	64	Up to 150 mg	–	N/A
PEG 3350	64	Up to 150 mg	–	N/A
Kolliphor® 407	63	Up to 150 mg	–	N/A
Capmul® GMO 50	25–30	Up to 350 mg	100–150	3.16–5.38

2.2.2. Preparation of blank SEDDS

The excipients Compritol^® HD5 ATO, Gelucire^® 48/16 were taken in a glass beaker and heated to 65 °C–70 °C to melt the excipients. To the mixture molten Capmul^® GMO-50 was added with the aid of stirring and stirred until the mixture becomes monophasic. (Oil, surfactant, and co-surfactant in 1:1:1 ratio). The monophasic molten mixture of excipients was added to solid carrier (Neusilin^® US2) in a mortar (lipid portion and solid carrier ratio of 1:1) and triturated for 2–3 min. Finally, the resultant mixture was passed through a #30 mesh ASTM and evaluated for flow properties by checking the angle of repose and Carr’s index.

2.2.3. Preparation of solid SEDDS (S-SEDDS) via conventional (2-step) method

A series of S-SEDDS were prepared in the conventional 2-step method with a range drug load of 1.0%−5.0%. The % of oil tested was 20–50%, and the surfactant/ co-surfactant ratios tried were 2:1 and 1:1. The molten excipient mixture of Compritol^® HD5 ATO, Gelucire^® 48/16 and Capmul^® GMO-50 was prepared as per the procedure in section 2.2.2. Once the excipients were molten and became monophasic, the API (fenofibrate) was added to the beaker and kept under vortex mixing until the API was completely dissolved. Then the prepared liquid SEDDS was adsorbed onto the solid carrier (Neusilin^® US2) at different proportions (lipid portion and solid carrier ratio of 1:1 and 2:1) by triturating for 2–3 min with a pestle in a mortar. Finally, the resultant mixture was passed through a #30 mesh ASTM and evaluated for flow properties by checking the angle of repose and Carr’s index. The prepared S-SEDDS was subjected to dilution (1000X-30000X) using 0.14% sodium lauryl sulfate (SLS) as the medium for dilution and tested for globule size and physical stability.

2.3. Preparation of HME S-SEDDS

2.3.1. Preparation of physical mixture

A physical mixture of HME S-SEDDS was prepared before the execution of the HME process. First, Gelucire^® 48/16 was milled using a lab mixer and sifted through 30# mesh ASTM. Next, Compritol^® HD5 ATO, fenofibrate, and Neusilin^® US2 were sifted through 30# mesh ASTM. Quantities of excipients and API were precisely weighed from the sifted material. Next, the semi-solid component of the formulation (Capmul^® GMO-50) was weighed in a porcelain dish and molten at a temperature of 35 °C −40 °C. In a mortar, previously weighed Gelucire^® 48/16 was transferred, followed by Compritol^® HD5 ATO and fenofibrate and mixed well with the help of a spatula for 1–2 min. Next, the molten Capmul^® GMO-50 was slowly added to the mixture in the mortar prepared by mixing with a spatula for 1–2 min. Then, Neusilin^® US2 was added to the prepared mixture (lipid portion and solid carrier ratio of 2:1) in the mortar and mixed well for 1–2 min to break if there were any lumps. Next, the physical mixture was passed through #30 mesh ASTM and blended for 15 min at 20 RPM using a V-shell blender (Maxiblend^®).

2.3.2. HME process for preparation of HME S-SEDDS

About 100 g of the physical mixture is prepared and introduced into the hot-melt extruder through a feeder. The HME process was carried out on an 11 mm co-rotating twin-screw extruder (Thermo Scientific^™ Pharma 11). The screw configuration used was Thermo Fisher Scientific standard configuration, which consists of 3 mixing zones to facilitate the dispersive action between the components of the molten components of the formulation and the solid carrier Neusilin^® US2. The first mixing zone had mixing elements making an angle of 30° to 90° with each other. The second mixing zone was relatively short in length, with the elements making an angle of 60° with each other, and the third mixing zone had mixing elements at an angle of 60° and 90° with each other. During the extrusion process, the feed rate of the physical mixture was 4.0–5.0 g/min through the feeder. Each time before the experiment, the feed rate was calibrated (n = 3) to ensure the feed rate was consistent. The residence time of the HME process was determined by introducing a colorant (blue food color) into zone 1 (feeding zone). Once the barrel was occupied entirely with the physical mixture, a part of physical mixture which was already blended with blue food color was freshly introduced into the feeder. The duration between the introduction of the colorant into the feeding zone and the first appearance of color in the product at the discharge zone was noted as residence time.

The extrusion process was carried out at a process temperature of 90 °C in all zones except the feeding zone (zone 1), which was maintained at ambient temperature. The basis for selecting the process temperature was the M.P. of fenofibrate, as API was the component of the formulation with the highest M.P. in the formulation. The extruder screw speed for the process was maintained at 50 RPM. To facilitate the product collection from the extruder, the die plate was detached, and the product was collected from the barrel and placed in receivers. The product (HME S-SEDDS) was allowed to cool to room temperature for approximately 5–10 min. The collected HME S-SEDDS was milled using a laboratory mixer and further sifted through an #30 mesh ASTM. The milled HME S-SEDDS was filled into hard gelatin capsules and stored in an airtight HDPE container at ambient conditions until further use (Fig. 1).

Fig. 1.

Fenofibrate loaded HME S-SEDDS A) at discharge zone during hot-melt extrusion, B) collected from hot-melt extruder (before milling), and C) milled HME S-SEDDS.

2.4. Experimental design

Response surface methodology using Central Composite Design (CCD) was employed in the experimental design where two independent variables were applied to determine the optimal levels for the average globule size (GS, nm, ${\text{Y}}_1$), polydispersity index (PDI, ${\text{Y}}_2$) and zeta potential (ZP, mV, ${\text{Y}}_3$) at 2 levels as assigned by design. The selected factors were the oil concentration (Compritol^® HD5 ATO, %, ${\text{X}}_1$) and Surfactant: Co-surfactant ratio (S/Co-S) (Gelucire^® 48/16: Capmul^® GMO-50, %, ${\text{X}}_2$), which were studied at five different levels: one central point (${\text{X}}_1$: 37.5% and ${\text{X}}_2$: 2.0), level + 1 (${\text{X}}_1$: 50% and ${\text{X}}_2$: 3.0), level −1 (${\text{X}}_1$: 25% and ${\text{X}}_2$: 1.0), level + $\alpha$ (${\text{X}}_1$: 55.2% and ${\text{X}}_2$: 3.41421), and level − $\alpha$ (${\text{X}}_1$: 19.82% and ${\text{X}}_2$: 0.585786). The $\alpha$ value (1.41421) was obtained by the equation $\pm \sqrt{2}$ for k = 2 (two independent variables). Table 2 provides the details on the CCD employed in this study. According to the CCD generated by the Design-Expert^® software (StatEase^®, version 13.0), a total of 13 experiments, including 4 factorial points (levels, ±1), 4 axial points (levels, $\pm \alpha$), and 5 replicates at the central point for estimation of pure error, were performed. All experiments were performed randomly to minimize the effects of variability of the observed responses due to systematic errors.

Table 2.

Independent variables with their coded levels of Central Composite Design for fenofibrate loaded HME solid self-emulsifying delivery systems (HME S-SEDDS).

Independent variables (% w/w)	Symbol	Coded levels
		−$\alpha$	−1	0	+1	+$\alpha$
Compritol® HD5 ATO	${\text{X}}_1$	19.82	25	37.5	50	55.2
Gelucire® 48/16: Capmul GMO® −50	${\text{X}}_2$	0.585786	1.0	2.0	3.0	3.41421

2.5. Evaluation of self-emulsification properties

2.5.1. Visual assessment of self-emulsification properties and stability

The prepared HME S-SEDDS was introduced into 0.14% SLS (at 1000X dilution (w/w)) at a mild agitation of 50 RPM with the help of a small magnetic stirrer. The emulsification time was recorded, and the emulsions were kept at ambient temperature for 24 h to observe the phase separation or precipitation (Bachynsky et al., 1997; Kallakunta et al., 2012). Furthermore, the emulsions were filtered through 0.22μ Whatman^® filter paper and the precipitate was dried at 45–50 °C overnight. Finally, the amount of precipitate (Neusilin^® US2) was determined by gravimetric method. Ternary Phase diagram was constructed using Tri plot v1–4 software and area of self-emulsification was represented in Fig. 2.

Fig. 2.

Ternary phase diagram of fenofibrate HME S-SEDDS formulations prepared by central composite design.

2.5.2. Droplet size analysis

The emulsions formed from introducing HME S-SEDDS into 0.14% SLS (1000X dilution w/w) (step 2.5.1) were analyzed for globule size using Malvern Zeta sizer (Nano ZS 90, Malvern Instruments, UK). The emulsions were centrifuged at 4000 for 15 min to remove traces of solid particles if any, and average globule size, PDI and ZP were determined. The size analysis was conducted at 25 °C at an angle of 90°.

2.5.3. Cloud point determination

The optimized HME S-SEDDS formulation was subjected to thermal treatment, and the stability of the formulation was evaluated. First, the HME-S SEDDS was introduced into 0.14% SLS (1000X dilution) at a temperature of 25 °C at a stirring rate of 50 RPM with the help of a small magnetic stirrer. After the emulsion formation, the glass beaker with the emulsion was placed on a hot plate, and the temperature was increased at a rate of 5 °C/minute. The initial sign of turbidity or phase separation was recorded as cloud point (Zhang et al., 2008).

2.6. High-performance liquid chromatography (HPLC) analysis

Quantitative HPLC analysis was performed as reported in USP 43 on an isocratic HPLC (Waters Corp., Milford, MA, USA) equipped with an auto sampler, UV/VIS detector (Waters 2489) and Empower software. The analytical column used was a Phenomenex Luna ${\text{C}}_{18}$ (5 μm, 250 mm × 4.6 mm) at a detection wavelength of 286 nm for fenofibrate. The mobile phase composition was acetonitrile and acidified purified water (adjusted to pH 2.5 ± 0.1 using phosphoric acid) at 70:30 (v/v). A calibration curve plotted between 2 and 50 μg/mL concentrations was found to be linear with a correlation coefficient (${\text{R}}^2$) of 0.999. The mobile phase was maintained at a flow rate of 1 mL/min from the solvent reservoir to the column. The samples collected were filtered through a 0.22 μm filter (Durapore^® PVDF) before being injected into the column and the injection volume was 10 μL.

2.7. Particle size distribution (PSD)

The PSD of the S-SEDDS was determined by sieve analysis using a vibratory sieve shaker (Performer III SS-3, Gilson Inc., USA). The analysis was performed using a combination of sieve set consisting of #30, 40, 60, 100 and 200 ASTM mesh series. First, an accurately weighed quantity of a 5 g sample of milled S-SEDDS was transferred into the sieve set. Then the analysis was conducted for 5 min at an amplitude of 5 and tapping frequency of 60 taps/min. After the analysis, the amount of milled S-SEDDS retained on each sieve was accurately weighed, and the weight fraction was calculated.

2.8. Flow characteristics

The flow properties of the blank SEDDS, S-SEDDS, and HME S-SEDDS were evaluated by determining the angle of repose and Carr’s index. The angle of repose test was done by placing a funnel on an even stable platform with a funnel height of 2–4 cm from the top of the product pile. The angle of repose was calculated by using the following equation:

$$\text{Angle of repose :}tan(\theta )=\frac{h}{r}$$ (1)

h = height of the pile

r = radius of the pile at the base.

Carr’s index evaluation was performed until there was no change in the tapped volume of the S-SEDDS. The number of taps performed for the test was in the order of 500,750,1000, 1250 and 1500. Then, the change in the final volume of the product was recorded. The following equation calculated Carr’s index:

$$\text{Carr’s index}(\text{CI}):CI=\left(\frac{D_T-D_B}{D_T}\right)\times 100$$ (2)

${\text{D}}_{\text{T}}$ = tapped density ${\text{D}}_{\text{B}}$ = bulk density

2.9. Fourier transform infrared spectroscopy (FTIR)

FTIR studies were performed for fenofibrate, physical mixture, HME S-SEDDS, Neusilin^® US2 (solid carrier) and Gelucire^® 48/16 (surfactant) with an Agilent Cary 660 FTIR Spectrometer (Agilent Technologies, Santa Clara, CA). The FTIR study was performed in the range of 4000 – 650 cm⁻¹ to assess the interaction between the fenofibrate and the rest of the formulation components. The bench consists of ATR (Pike Technologies MIRacle ATR, Madison, WI) equipped with a single-bounce, diamond-coated ZnSe internal reflection element.

2.10. Differential scanning calorimetry (DSC)

DSC analysis was performed for fenofibrate, physical mixture, and HME S-SEDDS (Discovery DSC25; TA Instruments, New Castle, DE, USA). The crystallinity of fenofibrate, physical mixture and HME S-SMEDDS was determined with a temperature ramp of 10 °C/min. In the solubility studies, the enthalpy of the physical mixtures was recorded and the range of drug solubility was calculated based on the drop in the enthalpy value.

2.11. X-ray diffraction (XRD) analysis

XRD analysis was carried out for fenofibrate and HME S-SEDDS (sample size of 500 mg) using Bruker D2 phaser SSD 160 diffractometer (Bruker AXS, Madison, WI) equipped with LYNXEYE scintillation detector and Cu K $\alpha$ radiation $(\lambda =1.541\text{Å})$. The step width was 0.02°/s over a range of 5° to 50° on a 2θ scale at room temperature. The operating conditions for the experiment were 30 kV and a current of 10 mA.

2.12. In-vitro permeation studies

For in-vitro permeation studies, optimized HME S-SEDDS at a concentration of 1 mg/ml in 0.14% SLS solution were prepared. For fenofibrate suspension, 1 mg/ml of samples of pure API was prepared in 0.14% SLS solution using 0.5% of Sodium carboxymethyl cellulose. Initially, the dialysis membrane (M.W of 10000) was stabilized for 45 min by adding 0.5 mL of 0.14% SLS solution. After 45 min, the SLS solution was replaced with 300 μL of the prepared sample in the donor chamber. The receiving chamber (scintillation vial) was filled with 20 mL of 0.14% SLS solution. The setup was maintained at 37.0 ± 0.5℃ on a multi-stationed magnetic stirrer (IKA, USA) at 200 rpm. 1 mL of sample was collected at predetermined time points of 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12, 16, 20, and 24 h and further analyzed for drug content using the HPLC method described in section 2.6. Each time the sample was collected, the volume was compensated with fresh media.

2.13. In-vitro drug release studies

The in-vitro drug release (dissolution) studies were performed for fenofibrate (40 mg of pure API filled in gelatin capsule) and HME-SEDDS (equivalent to 40 mg of fenofibrate, filled in gelatin capsule) using a USP apparatus type II (SR8-plus^™, Hanson) at a temperature of 37 ± 0.5 °C and paddle speed of 50 RPM (n = 6). The study was conducted with a medium, 1000 mL phosphate buffer at a pH 6.8 + 0.14% SLS. The capsules were placed in stainless steel sinker and dropped into the dissolution medium. The Aliquots of 2 mL were collected at predetermined time intervals and replenished with an equivalent volume of fresh medium. The collected samples were filtered through a 0.45 μm membrane filter and analyzed for fenofibrate content using the HPLC method described in section 2.6.

2.14. Treatment of dissolution data

From the in-vitro release profiles, various dissolution parameters like dissolution efficiency (DE), initial dissolution rate (IDR), and mean dissolution rate (MDR) were calculated. The dissolution efficiency (DE) is defined as the area under the dissolution curve up to a particular time (t), expressed as a percentage of the area of the rectangle described by 100% dissolution at the same time. DE at 15 and 60 min was calculated by using the following equation:

$$DE=\frac{\left({\int }_0^t\text{y}\times \text{dt}\right)}{\text{y}100\times \text{t}}\times 100\text{\%}$$ (3)

The initial dissolution rate (IDR) was calculated for the first 15 min of dissolution using the following equation:

$$IDR=\frac{\%dissolved}{min}$$ (4)

The mean dissolution rate (MDR) is calculated using the following equation:

$$\text{MDR}=\frac{{\sum }_{j=1}^n\text{ΔMj}/\text{Δt}}{n}$$ (5)

Where j is the dissolution sample number, n is the number of dissolution sampling times, $\text{Δt}$ is the time at the midpoint between two consecutive time points (${\text{t}}_{\text{j}}$ and ${\text{t}}_{\text{j}-1}$), and ${\text{ΔM}}_{\text{j}}$ is the additional amount of drug released between two consecutive time points (${\text{t}}_{\text{j}}$ and ${\text{t}}_{\text{j}-1}$).

2.15. Stability studies

The stability studies were carried out at 40 °C and 75% RH for 3 months for optimized HME-SEDDS. The product was stored in a tightly closed HDPE container and placed in a stability chamber (Caron, 6030) for the study period. At specific time point, the samples were collected from the chamber. The droplet size, emulsification time, crystallinity, drug content and in-vitro release profiles were evaluated. For the in-vitro drug release studies, the similarity factor (${\text{f}}_2$) was calculated per the implementations of the U.S. Food and Drug Administration by comparing the in-vitro drug release profiles of the fresh and stability samples (Pillay and Fassihi, 1998). The ${\text{f}}_2$ value is a logarithmic transformation of the sum-squared error of the differences between the fresh and stability samples. The ${\text{f}}_2$ value is calculated using the following equation:

$$f_2=50\mathit{log}\left\{{\left[1+\left(\frac{1}{n}\right)\sum _{t=1}^n{\left(R_t-T_t\right)}^2\right]}^{-1/2}\times 100\right\}$$ (6)

Where ${\text{T}}_{\text{t}}$ is the cumulative release rate of the stability sample, and ${\text{R}}_{\text{t}}$ represents the cumulative drug release rate of the fresh sample at predetermined time points. The term n denotes the number of time points. A ${\text{f}}_2$ value ≥50 indicates the similarity between the in-vitro drug release profiles of the fresh and stability samples.

3. Results and discussion

3.1. Pre-formulation studies

3.1.1. Selection of excipients from solubility studies

The solubility studies were conducted to determine the solubility of fenofibrate in different excipients (vehicles/surfactants). The SEDDS consists of oil, surfactants/ co-surfactants, and the API should be presented in a monophasic state when introduced into the aqueous medium. In addition, the SEDDS should have good solvent properties to present the drug in the solubilized state (Quan et al., 2007). Thus, in the formulation of SEDDS, choosing suitable excipients plays a vital role in the stability of the dosage form and optimizing the drug loading. The components in the formulation of SEDDS should have good solubility of API and good miscibility to produce a stable formulation and maximize the drug loading (Balakrishnan et al., 2009).

The details of solubility studies are presented in Table 1. The solubility measurement method followed in this work is a slight modification of the proposed procedure by Gattefosse USA, Paramus, NJ. The principle involved in measuring the solubility of fenofibrate in solid/semi-solid excipients is the depression of melting enthalpy as the solute (fenofibrate) gets dissolved in it. As the proportion of the solute dissolved increases, the energy required to melt the sample (solute + solvent) decreases until the saturation solubility is reached. Above the saturation solubility concentration, the depression of enthalpy will not be observed. Instead, the increment of enthalpy will be observed due to the excess solute (which is in undissolved form) added beyond the saturation solubility range. Solubility studies showed that the API (fenofibrate) has solubility in Compritol^® HD5 ATO, Gelucire^® 48/16, and Capmul^® GMO-50, TPGS. Based on the solubility studies, excipients with maximum solubility were selected for the final formulations. Compritol ^® HD5 ATO (Behenoyl polyoxyl-8 glycerides) was selected as the oil/vehicle for the API in the formulation. Gelucire^® 48/16 (Polyoxylethylene stearates) was selected as the surfactant that can reduce the interfacial tension between the oil and aqueous phases and aid in the dispersion of emulsion droplets in the aqueous phase. Capmul^® GMO-50 (Glyceryl Monooleate) was selected as a co-surfactant to further reduce the interfacial tension between oil and aqueous phases.

3.1.2. Selection of % drug load and proportion of carrier

The proportion of solid carrier was preliminary evaluated by preparing preliminary formulations (blank SEDDS) with equal proportions (1:1:1 ratio) of oil, surfactant, and co-surfactant. Further, a series of formulations of S-SEDDS were prepared with a % drug load of 1%−5% by conventional 2-step method (section 2.2.3). The resultant emulsions of the formulations were evaluated for globule size with a range of dilution of 1000X-30000X with 0.14% SLS solution. In addition, the proportion of solid carrier (Neusilin^® US2) was further evaluated. 1:1 and 2:1 ratio of liquid SEDDS portion (oil/S-Co-S + drug) and solid carrier (Neusilin^® US2)) produced solid formulation with good flow properties. Based on the results (data not shown), to maintain desired globule size around 300 nm and good flow properties, a drug load of 4% and 2:1 ratio of liquid SEDDS: Neusilin^® US2 were selected for further studies.

3.2. Preparation of HME- S-SEDDS

The preparation of HME S-SEDDS was a single-step continuous extrusion process performed by introducing the blend into the feeding zone. HME S-SEDDS was directly collected from the discharge zone. The process temperature of 90 °C was selected based on the melting point of fenofibrate (82 °C) and other lipid excipients. As fenofibrate is the component with the highest M.P of the formulation, the selection of process temperature of 90 °C provides a processing condition where API and excipients could have interacted in a molten state. This interaction helps to solubilize the molten fenofibrate in the excipients and simultaneously distributes the molten components onto the porous structure of carrier Neusilin^® US2. The Thermo standard screw configuration with 3 mixing zones was selected to accommodate maximum physical interaction between the molten components (lipid portion and fenofibrate) and Neusilin^® US2. The residence time of the process was <1 min for the all the experiments performed. The product from the discharge zone was like a solid waxy material. After congealing for 5–10 min at room temperature, it became a solid material that was easily milled (Fig. 1). The low % torque (6%−12%) of the process indicates the ease or feasibility of the extrusion process due to the lubrication provided by molten component and a proper balance between the proportion of molten portion and solid carrier. The product was easily milled with a lab mixer, and within 5–10 sec (for 100 g of product) of milling, a fine granular product was easily passed through the #30 mesh ASTM. The milled product was filled into gelatin capsules and stored in an HDPE bottle for further studies. The optimum ratio (2:1) of lipid portion (oil + surfactant + co-surfactant) and solid carrier is the reason for producing a solid product that can be easily milled. Excess of molten lipid portion could have resulted in a sticky product which would have been difficult for collection and downstream processing. On the other hand, an extreme proportion of a solid carrier (Neusilin^® US2) would cause an unnecessary increment in the dosage form weight and excess fine portion in the product (Qureshi et al., 2015). In addition, the excess proportion of carrier also alters the extent of adsorption between the molten liquid portion on the carrier. So the balance between the lipid and solid carrier portions is vital for producing a stable product that could be easily subjected to down-streaming.

3.3. Evaluation of HME S-SEDDS and ternary phase diagram

The observations made during the self-emulsification studies are represented in Table 3, and the ternary phase diagram (Fig. 2) represents the composition. The emulsification time of 180–300 sec shows the spontaneity of the emulsification process. The emulsions were stable after 24 h without any coalescence of globules or phase separation of emulsion. The emulsification time for different formulations was observed to be close. However, a higher oil proportion and relatively lesser surfactant proportion resulted in slightly greater emulsification times. Another reason for the closer emulsification time may be the adsorption of the lipid system onto the porous solid carrier (Neusilin^® US2). This is because the self-emulsifying portion is adsorbed on porous carrier Neusilin^® US2, upon encounter with aqueous media it gets emulsified, forming fine emulsion droplets leaving the carrier. The proportion of the solid carrier (33.3%) is similar in all the formulations, and this might have subjugated the effect of lipid composition to some extent on emulsification time.

Table 3.

Self-emulsification properties of fenofibrate-loaded HME solid self-emulsifying delivery system formulations (HME S-SEDDS) (mean ± SD; n = 3).

Formulation code	Emulsification time (s)	Appearance of emulsion	Stability after 24 h
F1	274 ± 7.11	Slightly Milky	Stable
F2	259 ± 5.97	Slightly Milky	Stable
F3	179 ± 5.68	Translucent	Stable
F4	192 ± 7.21	Translucent	Stable
F5	185 ± 7.24	Translucent	Stable
F6	194 ± 5.28	Translucent	Stable
F7	188 ± 4.86	Translucent	Stable
F8	191 ± 4.95	Translucent	Stable
F9	181 ± 9.25	Translucent	Stable
F10	199 ± 3.65	Translucent	Stable
F11	269 ± 4.67	Slightly Milky	Stable
F12	179 ± 5.68	Translucent	Stable
F13	291 ± 6.92	Slightly Milky	Stable

During the cloud point determination for the optimized formulation, the emulsions were stable up to 80 °C. At the temperature range of 85–95 °C, the emulsions turned cloudy, and phase separation was observed. This might be due to the dehydration of polyoxyethylene chains of Gelucire^® 48/16 and alkyl chains of Compritol ^® HD5 ATO and Capmul^® GMO-50. The thermodynamic stability of the emulsions indicates the stability of the dispersed systems. The cloud point is the temperature above which the emulsion turns cloudy, indicating the emulsion’s instability and phase separation. As the drug solubilization and stability of the emulsion decrease above the cloud point and the cloud point should be above 37 °C. (Elnaggar et al., 2009). The high cloud point range of the emulsions (>37 °C) indicates the solvent capacity of the excipients for API and the appropriate proportion of the oil: S/Co-S.

3.4. Evaluation of physical properties of HME S-SEDDS

The prepared milled HME S-SEDDS and S-SEDDS were evaluated for particle size analysis and flow properties per the method described in sections 2.7 and 2.8. The size evaluation results showed that 70–80% of the product was retained between the ASTM mesh # 30 and #60 (250μ−600μ). The flow evaluation showed that the blank SEDDS and S-SEDDS showed good flow properties with an angle of repose values of 31.54–33.15 and Carr’s index values of 17.28–20.24. HME S-SEDDS formulations had excellent flow with angle of repose values of 26.3–29.1 and Carr’s index values of 13.5–15.3. The shear produced by HME process might have helped in efficient distribution of molten lipid portion on to the porous Neusilin^® US2. The reason for excellent flow is the proportional balance between the lipid portion and solid carrier Neusilin^® US2 and carrier’s adsorption capacity (1.3 to 3.1 mL/g) for greater proportions of liquid portion into the pores of silicate particles (average pore size of 15 nm) rather than the surface (Gumaste et al., 2013).

3.5. DSC analysis

DSC analysis was performed for fenofibrate, physical mixture (optimized formulation), and HME S-SEDDS (optimized formulation) (Fig. 3). DSC analysis of pure fenofibrate showed a melting peak at 82.4 °C, indicating the crystalline nature of API. The DSC analysis of the physical mixture shows broad peaks at 43.02 °C and 69.35 °C corresponding to Gelucire^® 48/16 and Compritol^® HD5 ATO. The peak at 81.45 °C in the physical mixture indicates the crystalline nature of fenofibrate. As Capmul^® GMO-50 is semi-solid at room temperature did not show any crystalline peak in the sample of the physical mixture. The fresh samples of HME S-SEDDS (optimized formulation) showed melting peaks at 41.28 °C and 62.82 °C indicating the recrystallization of lipid components Gelucire^® 48/16 and Compritol ^® HD5 ATO after HME S-SEDDS was congealed to room temperature. The stability samples (HME S-SEDDS) showed peaks at 41.37 °C and 62.51 for Gelucire^® 48/16 and Compritol ^® HD5 ATO, indicating no shift in the melting peaks of the excipients compared to fresh samples. The single peak of each lipid component indicates that the recrystallized lipids are stable (Bhalekar et al., 2009; Hamdani et al., 2003). The initial (fresh) HME S-SEDDS samples showed no melting peaks for fenofibrate, indicating the amorphous conversion of fenofibrate. The absence of melting peaks in stability samples shows the preservation of the amorphous nature of fenofibrate, and no significant shift in the melting peaks of recrystallized lipids was observed. These observations indicate the stability of the prepared HME S-SEDDS.

Fig. 3.

DSC thermograms of A) fenofibrate B) Physical mixture (optimized formulation) C) optimized HME S-SEDDS (3 Month stability at 40 °C /75%RH) D) optimized HME S-SEDDS (fresh formulation).

3.6. XRD analysis

In XRD analysis, fenofibrate showed characteristic peaks at 12.15°,14.51°,16.87°, 21.05°, 22.37°, 24.17°, and 24.89°, indicating the crystallinity of the API (Fig. 4). The initial and stability samples of optimized formulation showed the absence of characteristic peaks of fenofibrate, confirming the amorphous nature and stability of HME S-SEDDS. The DSC and XRD results indicate the solubility of fenofibrate in the excipients Compritol ^® HD5 ATO, Gelucire^® 48/16, and Capmul^® GMO-50 (determined from solubility studies) and amorphous conversion of API contributing to the drug distribution in the amorphous form onto a solid carrier.

Fig. 4.

XRD diffractograms of fenofibrate, and optimized HME S-SEDDS formulation (fresh and stability).

3.7. FTIR

The FTIR spectra of the API (fenofibrate), Gelucire^® 48/16, Neusilin^® US2, physical mixtures, and HME S-SEDDS are presented in Fig. 5. The FTIR spectra of fenofibrate showed characteristic peaks at 1725 cm⁻¹ (represents a $\text{C}=\text{O}$ stretching of the ester group), 2980 cm⁻¹ (indicates the benzene ring; aromatic stretching), 1649 cm⁻¹ (represents $\text{C}=\text{O}$ of the ketone group), 1597 cm⁻¹ (lactone carbonyl functional group) (Ige et al., 2013; Karolewicz et al., 2016). Gelucire^® 48/16 showed characteristic peaks at 2885 cm⁻¹, 2160 cm⁻¹,1467 cm⁻¹ 1341 cm⁻¹ and 1103 cm⁻¹ (Aldosari, 2018). Neusilin^® US2 showed characteristic peaks at 2160 cm⁻¹ and 1647 cm⁻¹. The physical mixture showed characteristic peaks only at 1737 cm⁻¹, 1464 cm⁻¹ and 1104 cm⁻¹ due to the low concentration of API in the formulation. The same peaks were observed in HME S-SEDDS (optimized formulation), indicating no significant interaction between fenofibrate and the excipients.

Fig. 5.

FTIR spectra of A) Gelucire^® 48/16, B) HME S-SEDDS (optimized formulation), C) physical mixture (optimized formulation) D) fenofibrate and E) Neusilin^® US2.

3.8. Optimization using CCD

CCD has been successfully adopted to optimize the compositions for various drug delivery systems (Liu et al., 2009). The oil concentration in the formulation, the ratio of surfactant to co-surfactant, and drug content were reported to affect the properties of SEDDS (Wu et al., 2006). However, to meet the therapeutic requirement, the drug content in the formulation was kept at a fixed concentration (equivalent to a 40 mg dose of fenofibrate) in this study. The compositions of HME S-SEDDS formulations are presented in Table 4. The CCD was applied to investigate the main and interaction effects of independent variables selected for the study on GS, PDI, and ZP. The randomized experimental runs with the actual composition of each variable in every single run and the corresponding responses obtained are represented in Table 4. F-test was applied to evaluate the lack of fit in each model and identify the model fitting. Response surface delineation was also performed accordingly. Graphs of surface responses were plotted for the relationship between the response and the tested two factors.

Table 4.

Globule size, polydispersity index and zeta potential of fenofibrate-loaded HME solid self-emulsifying delivery systems (HME S-SEDDS) prepared by Central composite design.

Formulation Code*	Run	Compritol® HD5 ATO (%w/w, ${\text{X}}_1$)	Surfactant: Co-surfactant (${\text{X}}_2$)	Surfactant (%w/w)	Co-surfactant (%w/w)	GS (nm, ${\text{Y}}_1$)	PDI (${\text{Y}}_2$)	ZP (mV, ${\text{Y}}_3$)
F1	1	50.0	1	25.0	25.0	378	0.49	−24.3
F2	2	37.5	0.585786	23.1	39.4	341	0.44	−20.5
F3	3	37.5	2	41.67	20.83	324.5	0.40	−31.1
F4	4	37.5	2	41.67	20.83	329.5	0.41	−29.6
F5	5	25	3	56.25	18.75	265	0.29	−24.2
F6	6	37.5	2	41.67	20.83	319	0.39	−30.8
F7	7	37.5	2	41.67	20.83	320	0.40	–32.0
F8	8	37.5	2	41.67	20.83	324	0.39	−30.4
F9	9	25	1	37.5	37.5	272	0.33	−25.6
F10	10	37.5	3.41421	48.3	14.2	293.6	0.36	–33.8
F11	11	50	3	37.5	12.5	370	0.45	−19.9
F12	12	19.82	2	53.45	26.73	250	0.25	−20.5
F13	13	55.2	2	29.9	14.9	386	0.52	−16.7

GS – globule size, PDI – poly dispersity index, ZP – Zeta potential.

^*All formulations contain fenofibrate (4 % w/w) which is equivalent to 40 mg of dose.

3.8.1. Effect of independent variables on globule size (${\text{Y}}_1$, GS) and polydispersity index (Y2, PDI)

Globule size is a critical aspect of evaluating SEDDS. The smaller globule size provides a larger surface area for drug absorption and contributes a faster release rate (Kang et al., 2004). The least-square second-order polynomial model equation that describes the relationship between the independent variables and the mean GS obtained from CCD at a 95% confidence level is given below.

$${\text{Y}}_1=32.97+50.42\ast X_1-10.25\ast X_2-0.25\ast X_1{X}_2$$ (7)

$${\text{Y}}_2=0.393846+0.0877297\ast X_1-0.0241421\ast X_2+1.11456{\text{e}}^{-16\ast }{X}_1{X}_2$$ (8)

The Model F-value of 120.75 for GS and 156.10 for PDI obtained from ANOVA testing (Table 5) implies that the model for both responses was significant, with only a 0.01% chance that the large F-value could occur due to noise. While the observed lack-of-fit was insignificant (GS; F-value = 5.24 and p-value = 0.0669, PDI; F-value = 2.84 and p-value = 0.1671) with a 6.69% and 16.71% chance that the considerable lack of Fit F-value could occur due to noise for GS and PDI, respectively. Further, the value of predicted regression coefficients (${\text{R}}^2$) was in reasonable agreement with the adjusted ${\text{R}}^2$ because the difference is<0.2, which implies that the predicted GS and PDI values from the model will be within a 95% confidence interval (CI) of the observed/experimental values. Out of the 13 experimental runs suggested by the Design-Expert^® software, the smallest GS and PDI (250 nm, 0.25) values were observed for run # 12, while the largest GS and PDI (386 nm, 0.52) values were observed for run # 13 (Table 4).

Table 5.

Statistical results of central composite design for optimization of dependent variables and independent variable effects on fenofibrate-loaded HME solid self-emulsifying delivery systems (HME S-SEDDS).

Source	Globule size (${\text{Y}}_1$, nm)				Polydispersity index (${\text{Y}}_2$)				Zeta potential (${\text{Y}}_3$, mV)
	Sum of squares	Degree of freedom	F-value	P-value	Sum of squares	Degree of freedom	F-value	P-value	Sum of squares	Degree of freedom	F-value	P-value
Model	21176.13	3	120.75	<0.0001	0.0662	3	156.10	<0.0001	346.28	5	38.11	<0.0001
${\text{X}}_1$	20334.69	1	347.86	<0.0001	0.0616	1	435.33	<0.0001	122.15	1	67.22	<0.0001
${\text{X}}_2$	841.19	1	14.39	0.0043	0.0047	1	32.97	0.0003	122.53	1	67.43	<0.0001
${\text{X}}_1$ ${\text{X}}_2$	0.2500	1	0.0043	0.9493	0.0000	1	0.0000	1.0000	12.72	7	1.00	0.3499
${\text{X}}_1^2$	―	―	―	―	―	―	―	―	91.22	1	91.22	50.20
${\text{X}}_2^2$	―	―	―	―	―	―	―	―	17.17	1	17.17	9.45
Residuals	526.11	9	―	―	0.0013	9	―	―	9.59	3	―	―
Lack of Fit	456.41	5	5.24	0.0669	0.0010	5	2.84	0.1671	3.13	4	4.09	0.1036
Pure Error	69.70	4	―	―	0.0003	4	―	―	12.72	7	―	―
Total	21702.25	12	―	―	0.0675	12	―	―	9.59	3	―	―
Adjusted ${\text{R}}^2$	0.9677				0.9749				0.9393
Predicted ${\text{R}}^2$	0.9214				0.9456				0.7964
Adequate precision	33.6234				37.6141				17.5503

A positive sign before a variable (${\text{X}}_1$; Compritol^® HD5 ATO) in the polynomial equations signifies that the response (GS and PDI) increases with increasing the level of the factor, while a negative sign (${\text{X}}_2$; Gelucire^® 48/16: Capmul^® GMO-50) indicates that the response and the factor have a reciprocal relationship as the response is decreasing with increasing levels. As shown in Fig. 6 (A-D) and equations (7) and (8), an increase in the amount of oil (Compritol^® HD5 ATO) resulted in a significant increase in GS and PDI due to an increase in lipid (oil) density (Patil et al., 2018). In addition, the emulsifying efficiency of surfactants is reduced with increasing % of Compritol^® HD5 ATO, a long chain (${\text{C}}_{22}$) derivative resulting in higher interfacial tension and thus, large GS and PDI. However, increasing the surfactant: Co-surfactant ratio reduced GS and PDI significantly due to the significant reduction in interfacial tension between the oil and aqueous phases (Garg et al., 2017; Khames et al., 2019).

Fig. 6.

Surface response curves and contour plots of globule size (A, B), polydispersity index (C, D) and zeta potential (E, F) of HME S-SEDDS formulations prepared by central composite design.

3.8.2. Effect of independent variables on zeta potential (Y3, ZP)

ZP is an essential factor that plays a crucial role in the physical stability of colloidal dispersions, and it is the determinant of the surface charge of dispersed droplets. It was obvious that the independent variable significantly affected the ZP of the HME S-SEDDS. ZP obtained from the 13 runs ranged from −16.7 to –33.8 mV. ANOVA revealed that the proposed design was significant (p < 0.0001) for the statistical analysis of zeta potential. The various correlation data for fitting the selected factorial model of ZP response are shown in Table 5. The model F value of 38.11 indicated that the selected model was significant, with only a 0.01% chance that the large F-value could occur due to noise. The predicted ${\text{R}}^2$ (0.9393) value was in reasonable agreement with the adjusted ${\text{R}}^2$ (0.7964) value (the difference was<0.2), which is an indication of a good predictive model. The least-square second-order polynomial model equation that describes the relationship between the independent variables and the mean ZP obtained from CCD at a 95% confidence level is given below.

$$\begin{gathered} Y_3=-30.78+3.90756\ast {\text{X}}_1-3.91363\ast {\text{X}}_2+0.675\ast {\text{X}}_1{\text{X}}_2 \\ +3.62125\ast {\text{X}}_1^2+1.57125\ast {\text{X}}_2^2 \end{gathered}$$ (9)

The observed lack-of-fit was insignificant (F-value = 4.09, p-value = 0.1036). As shown in Fig. 6 (E and F) and Equation (9), an increase in the surfactant co-surfactant ratio resulted in a significant increase in ZP due to reduced interfacial tension. The ZP sign and magnitude are related to the net charge type and density displayed on the oil globule surface. Further, the accumulation of more negative charges on the surface of the globules with increase in surfactant and co-surfactant ratio was observed. Zeta potential of −20mv to −30mv is required for electrostatic stabilization of the dispersed systems. Similar outcomes have been reported in many earlier published investigations (Kundu et al., 2015; Onaizi, 2022).

3.8.3. Optimization and validation

All tested responses are vital factors for a successful SEDDS preparation because they can affect the self-emulsifying properties of the formulation and there by influencing the rate of drug release and extent of absorption (Tarr and Yalkowsky, 1989). Therefore, SEDDS composition was optimized after analyzing responses and building up good regression models by setting the goals of the independent variables and responses and applying the global desirability function (D). Based on these criteria, the desirability plot was generated with a D value of 0.93. In order to achieve the desired responses with 95% confidence intervals (CI), the software selected one formulation out of 9 proposed solutions with 25% w/w of Compritol^® HD5 ATO and a ratio of 3 for surfactant: Co-surfactant to fulfill the optimum formulation requirements. These respective levels can result in a formulation with a GS of ~260.5 nm, PDI of ~0.28, and ZP of ~−34.1 mV. To check the validity of the selected model, a HME S-SEDDS formulation (n = 6) was prepared to compare the observed response values against the software-predicted values. The results of the validation trial were observed as GS of ~269.6 nm, PDI of ~0.27, and ZP of ~–32.1 mV, in agreement with the predicted values. The mean of observed response values was within a 95% CI of the predicted software values.

3.9. In-vitro permeation studies

After oral administration, self-emulsifying systems encounter the aqueous medium, and API may be present in the form of emulsion or solubilized form in the micelle or free molecular state. In order to get released from emulsion droplets, API should go through the interfacial transport across the surfactant layer around the droplet. Further, the API enters the surrounding aqueous medium via diffusion and convective transport (Zhang et al., 2008). Thus, the drug diffusion from fine emulsion droplets into the aqueous medium is not instantaneous. Considering this, the free drug molecules should be separated from the drug entrapped in the micelle or emulsion droplets to assess the real drug release pattern difference between different formulations (Wu et al., 2006). Therefore, the dialysis bag with a molecular weight cut-off of 10,000 was utilized to study the drug release studies.

Based on the DoE study results, the in-vitro permeation studies were performed for optimized formulation and fenofibrate suspension (Fig. 7). The in-vitro permeation studies revealed the improved drug release of optimized formulation compared to the fenofibrate suspension. The drug permeated from fenofibrate suspension was about 16% in the first hour and remained plateau for the rest of the study period, indicating that the drug diffusion occurred in the initial hour, probably due to the sink condition provided by the media (0.14% SLS solution). Further plateau indicates the poor solubility of the fenofibrate. The drug permeation from emulsion droplets of HME S-SEDDS throughout the study indicates gradual diffusion of API from the dispersed emulsion droplets (Kallakunta et al., 2012; Wu et al., 2006). The significantly improved drug release (p < 0.05) might be due to the appropriate proportion of the oil and surfactant system, which presented the drug in the emulsified form, and the reduced interfacial tension facilitating the drug release across the membrane. This infers the improvement in the solubility of fenofibrate by formulating into HME S-SEDDS, which is further confirmed by the improved % drug release compared to pure drug suspension. Though the in-vitro permeation studies cannot provide an accurate picture of drug release from SEDDS, from the results it is evident that there was shift in partitioning of drug between oil phase and aqueous phase. Additional studies are needed for further investigation of drug uptake mechanism in future studies.

Fig. 7.

In-vitro drug permeation profiles of fenofibrate suspension and optimized HME S-SEDDS (fresh and stability) formulation (mean ± SD; n = 6).

3.10. In-vitro drug release studies

SEDDS on contact with the medium of dispersion form o/w emulsion. The free energy required for emulsification is very low, thus thereby allowing spontaneous emulsification. This forms an interface between the emulsion droplets and water resulting in the initial formation of the liquid crystalline phases at the surface of droplets (Borovicka et al., 1997). Further, the oil/surfactant system and water phases would swell effectively, reducing the emulsion droplet size and eventually increasing the drug release rate into the medium (Wakerly et al., 1986). The assay values of the formulations were found to be in the range of 97.21 ± 0.71–102.31 ± 1.24. In the dissolution studies, the drug release of all the formulations (compositions mentioned in Table 4) was significantly (p < 0.05) improved compared to the pure drug, indicating a significant improvement in the solubility of fenofibrate (data not shown).

In-vitro dissolution studies of optimized HME S-SEDDS and pure drug are represented in Fig. 8. The dissolution studies were carried out for 1 h to observe if the phase separation phenomenon occurs. The significant (p < 0.05) improvement of dissolution might be due to the excellent solvent capacity of the excipients for API, which kept the API in amorphous form, selection of process conditions, and appropriate proportions of oil and surfactant systems. Gelucire^®48/16 (hydrophilic surfactant with HLB > 12) is a medium-chain surfactant that can significantly reduce the interfacial tension between the dispersed phase and aqueous phase facilitating the drug release. The dissolution parameters are represented in Table 6. For the optimized HME S-SEDDS, 90.08 ± 0.13% of drug release was observed in the first 15 min, and ${\text{DE}}_{15}$ of optimized formulation (${\text{DE}}_{15}=45.04$) was significantly (p < 0.05) improved compared to pure API (${\text{DE}}_{15}=4.09$), and these values are coordinating with the formulation aspects. Another factor is the porous Neusilin^® US2, on which the solubilized fenofibrate was distributed along with other formulation ingredients. The larger surface area (110 to 300 m²/g), small particle size ($112\text{μ}$), pore size (15 nm) and good adsorption capacity of 1.3 to 3.1 (ml/g) of Neusilin^® US2 help in the distribution of API in amorphous form (confirmed by DSC and XRD studies) and helped in avoiding the recrystallization of fenofibrate. This was evident as no phase separation or precipitation was observed at the end of the dissolution study, and accelerated stability studies proved these formulation aspects. Based on some previous reports, the interactions between the solid carrier and SEDDS portion might hinder drug release. This depends on the carrier’s surface chemistry and its interaction with lipids and drug molecule (Krupa et al., 2015; Williams et al., 2014). The drug release studies show that the solid carrier (Neusilin^® US2) is not interfering with drug emulsification and subsequent drug release.

Fig. 8.

In-vitro drug release profiles of fenofibrate (pure API) and optimized HME S-SEDDS (fresh and stability) formulation (mean ± SD; n = 6).

Table 6.

Summary of dissolution parameters of API and optimized HME S-SEDDS (mean ± SD; n = 6).

Formulation	${\text{Q}}_{15(\%)}$	${\text{Q}}_{60(\%)}$	${\text{DE}}_{15}$	${\text{DE}}_{60}$	IDR	MDR	${\text{t}}_{90\%}$ (minutes)
Fenofibrate (pure API)	8.18 ± 0.181	8.445 ± 0.011	4.09	7.29	0.545	0.165	15
Optimized formulation	90.08 ± 0.13	99.15 ± 0.09	45.04	84.69	6.005	1.850	N/A

${\text{Q}}_{15}$ and ${\text{Q}}_{60}$ - drug released in 15 and 60 min; ${\text{DE}}_{15}$ and ${\text{DE}}_{60}$ are dissolution efficiency at 15 and 60 min; IDR and MDR are the initial dissolution rate at 15 min and mean dissolution rate respectively: ${\text{t}}_{90\%}$ is time at which 90% drug release is achieved.

3.11. Stability studies

The optimized HME S-SEDDS was subjected to accelerated stability studies. The stability studies confirmed the preservation of the amorphous nature of the optimized formulation (Fig. 3 and Fig. 4). The stability studies showed that there is no significant difference (p < 0.05) between droplet size, emulsification time, in-vitro permeation studies, and dissolution studies compared to fresh samples. The stability of the formulation might be due to proper ratio of the emulsifying agents and the proportion of solid carriers. In addition, the distribution of the molten HME S-SEDDS on the solid carrier with the aid of shear might have helped the uniform distribution of the lipid excipient on the high surface area of the Neusilin^® US2, provided additional stability inhibiting the recrystallization of API and precipitation of lipid excipients. These aspects collectively provided stability to HME S-SEDDS and aided in preserving the emulsifying properties, further reflected in almost superimposable in-vitro permeation profiles and similar in-vitro drug release profiles. The in-vitro release profiles of fresh and stability samples were similar (${\text{f}}_2=89$), showing the amorphous nature and stability of the formulation.

4. Conclusion

The HME S-SEDDS were successfully prepared via single-step continuous HME manufacturing. The prepared HME S-SEDDS were evaluated for emulsifying properties, crystallinity, stability and drug release. The prepared HME S-SEDDS showed spontaneous emulsification properties, and the emulsions were stable for the tested duration. The prepared HME S-SEDDS exhibited amorphous nature, and subsequently, in-vitro dissolution studies showed significant improvement in the solubility of the fenofibrate. The formulations were statistically treated using the response surface method, and the optimized formulation was in good agreement with the data predicted by the software. The stability studies conducted for 3 months at 40 °C/75%RH demonstrated preservation of the amorphous nature of the formulation indicating the proper selection of excipients and processing conditions. This systematic preparation of SEDDS via a single-step continuous HME process has not been explored. Thus, this novel study can guide future studies on the continuous manufacturing of solid SEDDS.

Data availability

Data will be made available on request.