*Article* **Development and Optimization of Nanolipid-Based Formulation of Diclofenac Sodium: In Vitro Characterization and Preclinical Evaluation**

**Ameeduzzafar Zafar 1,\* , Nabil K Alruwaili <sup>1</sup> , Syed Sarim Imam 2,\* , Mohd Yasir <sup>3</sup> , Omar Awad Alsaidan <sup>1</sup> , Ali Alquraini <sup>4</sup> , Alenazy Rawaf <sup>5</sup> , Bader Alsuwayt <sup>6</sup> , Md. Khalid Anwer <sup>7</sup> , Sultan Alshehri <sup>2</sup> and Mohammed M. Ghoneim <sup>8</sup>**


**Abstract:** In the present research study, we formulate bilosomes (BMs) of diclofenac (DC) for oral delivery for enhancement of therapeutic efficacy (anti-inflammatory disease). The BMS were prepared by thin film hydration method and optimized by Box–Behnken design (BBD) using cholesterol (A), lipid (B), surfactant (C), and bile salt (D) as formulation factors. Their effects were evaluated on vesicle size (Y<sup>1</sup> ) and entrapment efficacy (Y<sup>2</sup> ). The optimized DC-BMs-opt showed a vesicle size of 270.21 ± 3.76 nm, PDI of 0.265 ± 0.03, and entrapment efficiency of 79.01 ± 2.54%. DSC study result revealed that DC-BMs-opt exhibited complete entrapment of DC in BM matrix. It also depicted significant enhancement (*p* < 0.05) in release (91.82 ± 4.65%) as compared to pure DC (36.32 ± 4.23%) and DC-liposomes (74.54 <sup>±</sup> 4.76%). A higher apparent permeability coefficient (2.08 <sup>×</sup> <sup>10</sup>−<sup>3</sup> cm/s) was also achieved compared to pure DC (6.6 <sup>×</sup> <sup>10</sup>−<sup>4</sup> cm/s) and DC-liposomes (1.33 <sup>×</sup> <sup>10</sup>−<sup>3</sup> cm/s). A 5.21-fold and 1.43-fold enhancement in relative bioavailability was found relative to pure DC and DC liposomes (DC-LP). The anti-inflammatory activity result showed a significant (*p* < 0.05) reduction of paw edema swelling compared to pure DC and DC-LP. Our findings revealed that encapsulation of DC in BMs matrix is a good alternative for improvement of therapeutic efficacy.

**Keywords:** bilosomes; diclofenac; optimization; pharmacokinetic; pharmacodynamic study

#### **1. Introduction**

Diclofenac (DC) is a non-steroidal anti-inflammatory (NSAID) agent with anti-inflammatory and antipyretic activity. It is widely used for treatment of acute pain and various antiinflammatory diseases such as osteoarthritis and rheumatoid arthritis. It has a short half-life of 1–2 h due to extensive first-pass metabolism. It belongs to the BCS-II class drug and reported low solubility [1]. The long exposure to DC inhibits prostaglandin formation

**Citation:** Zafar, A.; Alruwaili, N.K.; Imam, S.S.; Yasir, M.; Alsaidan, O.A.; Alquraini, A.; Rawaf, A.; Alsuwayt, B.; Anwer, M.K.; Alshehri, S.; et al. Development and Optimization of Nanolipid-Based Formulation of Diclofenac Sodium: In Vitro Characterization and Preclinical Evaluation. *Pharmaceutics* **2022**, *14*, 507. https://doi.org/10.3390/ pharmaceutics14030507

Academic Editors: Vibhuti Agrahari and Prashant Kumar

Received: 24 December 2021 Accepted: 17 February 2022 Published: 25 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

which causes gastric irritation, bleeding, and ulcers [2]. These side effects can be minimized by reducing direct contact of drug with GIT [3].

Various studies have been published to overcome complications of gastric irritation and ulcers. Various types of formulation have been reported to increase therapeutic efficacy and side effects of DC. A diclofenac-loaded nanoformulation was prepared and evaluated for different parameters [4]. They reported a nanometric size with high entrapment efficiency. A significant effect was observed on pharmacokinetic and pharmacodynamic activities. The low dose depicted clinical therapeutic levels in blood for up to 120 h, with minimal drug accumulation in organs as well as better efficacy than other controls. Akbari et al. developed transdermal diclofenac niosomal gel for improvement of therapeutic activity. The prepared formulation showed nano-metric size, negative zeta-potential, and high entrapment efficiency. The biological activity result revealed significantly lower licking time than conventional formulation [5]. In another study, diclofenac sodium-loaded nanovesicles were prepared by double solvent displacement method [6]. The prepared liposomes showed nano-metric size, negative zeta potential, and high encapsulation efficiency. The permeation results revealed a higher transdermal passage of drug. Gaur et al. prepared diclofenac sodium-loaded lipid vesicles and analyzed them for physical and biological activity [7]. The prepared formulation depicted more than 90% release with an enhanced pharmacokinetic profile.

The application of lipid-based nanoformulations is rising as an effective method for drug delivery. It can enhance drug solubility as well as bioavailability and reduce side effects. There are various lipid nanoformulations such as solid lipid nanoparticles [8,9], nanostructured lipid carriers [10], liposomes [11], and bilosomes [12]. Among them, bilosomes (BMs) are the new nano-sized lipid vesicular formulation used for different therapeutic agents. They are an elastic vesicular system composed of phospholipid, surfactant, cholesterol, and bile salt [13]. The bile salt enters into the lipid bilayer and lowers the phase transition temperature and builds vesicles deformable under body temperature [14]. They have been reported to enhance bioavailability of many drugs [15,16]. They show less drug leakage and high lymphatic drug transport as compared to niosomes and liposomes. They also prevent the enzymatic degradation in gastrointestinal tract and pass from reticular endothelial system [17]. In the GIT, bile salt that breaks vesicle before reaching to target site [18]. The bile salt acts as a solubilizing and permeation enhancer agent and may improve bioavailability of poorly soluble drugs [19]. Different types of bile salts such as sodium deoxycholate, sodium glycolate, sodium taurocholate, and sodium glycolate are used to prepare BMs. Among them, sodium deoxycholate is commonly used for formulation of BMs due to its nontoxic nature [20]. Shukla et al. formulated a diphtheria toxoid-loaded BM for oral administration. It produced quantifiable anti-diphtheria toxoid response in serum as well as mucosal secretion [21]. In another study, Shukla et al. formulated BM oral delivery of hepatitis B and produced concentration level of systemic and mucosal antibodies [22]. Zakaria et al. formulated piperine-loaded BMs for antiviral and anti-inflammatory activity [23]. Piperine-loaded BMs exhibited significantly reduced oxidant markers and cytokines in MERS-Co-V infected mice compared to pure piperine. El Taweel et al. formulated zolmitriptan-loaded BMs in situ gel for nose brain delivery [24]. BMs in situ gel produced significant bioavailability (1176.9%) compared to BMs dispersion (835.7%).

Up until now, no study has been performed to evaluate efficacy of diclofenac bilosomes (DC-BMs) to improve therapeutic efficacy. The objective of study is to prepare and optimize them using experimental design software (Stat-Ease, Minneapolis, MN, USA). The selected formulation (DC-BMs-opt) was evaluated for physicochemical characterization, in vitro, ex vivo study, and pharmacokinetic and pharmacodynamic study.

#### **2. Experimental**

#### *2.1. Materials*

Diclofenac potassium, lipid (L-α-Phosphatidylcholine), pluronic F123, cholesterol, and bile salt (sodium deoxycholate) were procured from Sigma Aldrich (St. Louis, MO, USA). Dialysis bag (MWCO 12,000 kDa) was procured from HiMedia laboratory (Mumbai, India). HPLC grade water, acetonitrile, and methanol were obtained from SD-fine chemicals (Mumbai, India).

#### *2.2. Methods*

#### 2.2.1. Formulation of Bilosomes

DC-BMs were prepared by slightly modified thin-film hydration method [13]. The lipid, surfactant, and cholesterol with a fixed dose of DC were taken in different ratios and dissolved in organic solvent (methanol: chloroform) as shown in Table 1. The solution was transferred to a round bottom flask and then organic solvent was evaporated at a temperature of 50 ◦C with reduced pressure using a rotary evaporator (IKA, RV-3V, Staufen, Germany). A thin lipid film was formed on wall of flask and stored in a desiccator for 24 h to remove moisture. The film was hydrated with phosphate buffer (10 mL) containing sodium deoxycholate for 3 h. The dispersion was collected and sonicated for 15 min to reduce size. The prepared formulations were collected and stored at 4 ◦C for further study.

**Table 1.** Formulation composition and their effect on vesicle size and entrapment efficiency.


#### 2.2.2. Optimization

DC-BMs were optimized by using 4 factors at 3 level Box–Behnken design (BBD). The independent variables cholesterol (A), lipid (B), pluronic F127 (C), and bile salt (D) were taken as independent factors and their effects were assessed on VS (R1) and EE (R2). The design showed twenty-seven formulations with five center points from software. The practical value of dependent variables (VS as R<sup>1</sup> and EE as R2). The data were fitted into software and evaluated for different models, i.e., linear, 2nd order, and quadratic models to determine best fit model. The regression analysis and ANOVA of best fit model were applied. The three-dimensional plots (3D plots) were plotted to interpret the effect of each factor over each response.

#### 2.2.3. Bilosomes Evaluation

The prepared DC-BMs (F1-F27) vesicle size (VS), PDI, and zeta potential (ZP) were measured by size analyzer (Zeta sizer Nano S90, Malvern, UK) at 25 ◦C. The diluted DC-BMs were placed in quartz cuvette and their size and PDI were measured. The same samples were evaluated for Zeta potential by using cuvette with an electrode.

#### 2.2.4. Entrapment Efficiency (EE)

EE of DC in prepared BMs was analyzed by indirect method [25]. The prepared DC-BMs (2 mL) were taken in a tube and samples centrifuged at 6000 rpm. The supernatants were collected and diluted and DC content was measured using a UV spectrophotometer. EE in each sample was calculated by using formula:

$$\text{EE} = \frac{\text{Total DC} - \text{DC in compartment}}{\text{Total DC}} \times 100\tag{1}$$

#### 2.2.5. Surface Morphology

The surface morphology of an optimized bilosomes (DC-BMs-opt) was examined by transmission electron microscopy (TEM, Philips CM 10, Eindhoven, The Netherlands). One drop of diluted sample was placed over grid and stained with phosphotungistic acid. The grid was air-dried and placed into instrument, and image was captured.

#### 2.2.6. Thermal Analysis

DSC analysis of DC, lipid, cholesterol (CHO), Pluronic F127, SC, physical mixture and optimized formulation DC-BMs-opt was analyzed using DSC instrument (Mettler Toledo, South Miami, FL, USA). Each sample (5 mg) was taken, packed into an aluminum pan, and scanned between 25–400 ◦C under an inert condition. The thermograms were recorded and compared to each other.

#### *2.3. In Vitro Drug Release*

The release study was performed using a pretreated dialysis bag. The test samples of pure DC, DC-LP and DC-BMs-opt (equivalent to 3 mg DC) were filled into a dialysis bag and tied from both ends. The bag was immersed into release media (500 mL, phosphate buffer pH 6.8) and assembly fixed at a temperature of 37 ± 0.5 ◦C with stirring speed of 50 rpm. 5 mL of released content was withdrawn at a fixed time and filled with fresh release media to maintain the volume. The absorbance was measured by UV-spectrophotometer (Genesys 10S UV-Vis, Thermo-scientific, Waltham, MA, USA) at 276 nm. The release data fitted to different release kinetic models to find best fit model.

#### *2.4. Ex Vivo Permeation Study*

The ex vivo permeation study was done using rat intestine. The rats were kept fasted overnight (24 h), then sacrificed and intestines were collected. The intestine was washed with normal saline (0.9% NaCl) and DC-BMs-opt, DC-LP, and pure DC were filled. The intestine was then immersed into a physiological ringer solution (composition NaCl, KCl, KH2PO4, CaCl2, glucose) as permeation media and placed over a magnetic stirrer. The system was fixed at 37 ± 0.5 ◦C with a regular supply of 95% O<sup>2</sup> and 5% CO2. At specific time, 2 mL sample was collected at a fixed time (0, 30, 60, 90, 120, and 180 min) and analyzed for drug permeation through previously developed HPLC [26]. HPLC system was run

using acetonitrile and methanol (7:3) with a flow rate of 0.75 mL/min, injection volume of 20 µL, and UV-detector at 276 nm. The apparent permeability and enhancement ratio was measured.

$$\text{Apparent permeability} = \frac{\text{Flux}}{\text{Area} \times \text{Initial drug concentration}} \tag{2}$$

$$\text{Enhancement ratio} = \frac{\text{Permeability coefficient of DC BLopt}}{\text{Permeability coefficient of the pure DC}} \tag{3}$$

*2.5. In Vivo Study*

2.5.1. Bioavailability Study

The study protocol was approved by institutional animal ethical committee Jouf University Sakaka, Al-Jouf, Saudi Arabia (Approval Number 04-02-43). The animals (Wistar Albino rats, 200–250 gm, either sex) were procured from an animal house. The animals were provided with free access to food and water and kept at 25 ◦C/50%RH. The study performed with three animal groups, each group having six rats. Group 1 was administered pure DC, Group 2 and Group 3 were administered with DC-LP and DC-BMs-opt. The samples of pure DC, DC-LP, and DC-BMs (equivalent to 2 mg/kg of DC) were administered orally to rats. At a definite time of 0, 0.5, 1, 2, 3, 6, 12, and 24 h, blood sample was collected from retro-orbital plexus into an EDTA tube. The plasma was separated by centrifuging blood sample at 5000 rpm for 15 min. The plasma was extracted by liquid-phase extraction method. The plasma was mixed with ethyl acetate and acetone (8:2, 0.5 mL), vortexed for 1 min, and then centrifuged to collect supernatant. The supernatant was dried under a stream of nitrogen and dried sample was reconstituted with acetonitrile and filtered through a 0.25 µm membrane filter. The sample (20 µL) was injected into HPLC system to calculate DC concentration in each animal.

#### 2.5.2. Pharmacodynamic Study

The rats were divided into four groups and each group containing six rats (n = 6). Group A was taken as normal control, Group B was used as disease control, Group C was administered with pure DC, group D was treated with DC-LP, and group E was treated with DC-BMs-opt. The carrageenan solution (1%, in saline) was administered to different groups by intra-plantar injection in a right hind paw to induce inflammation. The pure DC, DC-LP, and DC-BMs-opt were administered orally before 30 min of carrageenan injection. The paw volume of each rat was measured by plethysmometer (Ugo Basile, Varese, Italy), before and after carrageenan injection at different time intervals (0, 1, 2, 3, 6, 9, 12, 24 h). The degree of edema induced was assessed by following equation.

$$\% \text{ Edema inhibition} = \frac{\text{Vt} - \text{V0}}{\text{V0}} \times 100 \tag{4}$$

where Vt and V0, are volume of right hind paw after and before carrageenan treatment.

#### *2.6. Statistical Analysis*

Data are represented as average ± SD. Graph Pad software Inc., La Jolla, CA, USA was used for statistical analysis. *p* < 0.05 was taken as statistical significance.

#### **3. Result and Discussion**

#### *3.1. Optimization*

DC-BMs were optimized by using 4-factor at 3-levels Box–Behnken design. The formulation composition of prepared DC-BMs with their dependent variables VS (R1) and EE (R2)are shown in Table 1. The minimum and maximum vesicle size was found in range of 169.34 nm (F1)–380.14 nm (F4). The lowest EE was found for formulation (F1) as 50.23% and highest found for formulation (F12) as 94.03%. The experimental value of all

prepared DC-BMs was applied into experimental design model and best fit model was found linear for vesicle size and 2nd order (2F1) model for EE. The adequate precision for VS and EE found >4 and represents model as well fitted [27]. The predicted R<sup>2</sup> values were found to be closer to adjusted R<sup>2</sup> and statistical analysis expressed in Table 2. The ANOVA of both responses were analyzed and sum of square, mean square, F-value, and *p*-value of dependent variable is given in Table 3. 3D-plots were constructed (Figures 1 and 2), and effect of an independent variable over responses were interpreted. The polynomial equation of responses was given below and it explains direct relationship of independent variables to responses. The positive and negative signs denote favorable and unfavorable effect of formulation factors over response.


**Table 2.** Statistical summary of best fit model for vesicle size (R<sup>1</sup> ) and entrapment efficiency (R<sup>2</sup> ).

**Table 3.** ANOVA of best fitted designing model for vesicle size (Y<sup>1</sup> ) and entrapment efficiency (Y<sup>2</sup> ).


[33].

CD 49 49 4.97 0.0404 -- Residual 157.65 9.85 -- -- -- Lack of Fit 155.44 11.10 10.04 0.0942 NS Pure Error 2.21 1.105 -- -- -- Total 3757.54 -- -- -- --

> The vesicle size of prepared DC-BMs was found in range of 169.34 nm (F1)–380.14 nm (F4). The formulation (F1) prepared with composition cholesterol (A, 0.1%), lipid (B, 0.5%), surfactant (C, 0.5%), and bile salt (D, 1%) showed lowest size and formulation (F4) showed maximum size with composition cholesterol (0.5%), lipid (1.5%), surfactant (0.5%), and bile salt (1%). The difference in size found due to variation in used composition. From the result, it can be observed that used independent variables have shown a significant effect. 3D response surface plot (Figure 1a,b) and polynomial Equation (5) show that increasing a CHO (A) concentration led to an increase in vesicle size due to greater amount of CHO deposited into lipid bilayer [28]. The second factor lipid concentration (B) increases vesicle size due to enhancement in viscosity of dispersion and thickness of lipid bilayers. Similar types of findings reported in reported research of papain liposomes [29], and diclofenac liposome [30]. The surfactant concentration (PP123, C) and bile salt (D) depicted a negative effect on BNs size. The enhancement in surfactant (PP123) gave the reduction in vesicle size because at high concentration of surfactant, interfacial tension reduced between lipid phases and aqueous phase. The bile salt increases flexibility of liposomes by incorporating into lipid bilayer, thereby decreasing vesicle size [31].

> > Vesicle size (nm, R1) = 274.36 + 49.20 A + 50.23 B − 30.45 C − 18.69 D (5)

The F-value fitted to linear model and value found to be 646.04 revealed that model was significantly (*p* < 0.0001) fitted. The regression coefficient of best fit model was found to be 0.9915 and it indicates lesser variation between actual and predicted value (Table 2). The ANOVA value showed that model term cholesterol (A), lipid (B), surfactant (C), and bile salt (D) were found to be significant model term (*p* < 0.0001, Table 3). The adequate precision was 85.98 (<4), revealing the close relationship between actual and experimental value [32]. The *p*-value of lack of fit is >0.05 indicated insignificant which is good for model

*3.2. Effect of Independent Variables over Vesicle Size (R1)* 

**Figure 1.** 3D plot showing **e**ffect (**a**) of independent variables lipid (A), cholesterol, and bile salt (D) on vesicle size. Contour plot showing effect (**b**) of independent variables lipid (A), cholesterol, and bile salt (D) on vesicle size (R1). **Figure 1.** 3D plot showing effect (**a**) of independent variables lipid (A), cholesterol, and bile salt (D) on vesicle size. Contour plot showing effect (**b**) of independent variables lipid (A), cholesterol, and bile salt (D) on vesicle size (R<sup>1</sup> ). *Pharmaceutics* **2022**, *14*, x FOR PEER REVIEW 9 of 16

**Figure 2.** Effect of independent variables lipid (A), cholesterol (B), Pluronic F127 (C), and bile salt (D) on entrapment efficiency (R2). **Figure 2.** Effect of independent variables lipid (A), cholesterol (B), Pluronic F127 (C), and bile salt (D) on entrapment efficiency (R<sup>2</sup> ).

#### *3.4. Optimized Formulation (DC*-*BMs-opt) 3.2. Effect of Independent Variables over Vesicle Size (R1)*

The formulation (DC-BMs-opt) was selected from point prediction of software. The composition CHO (0.3% *w*/*v*), lipid (1% *w*/*v*), surfactant (0.5% *w*/*v*), and bile salt (1% *w*/*v*) depicted an experimental vesicle size of 270.21 ± 3.76 nm and EE of 79.01 ± 2.54%. The software showed a predicted value of vesicle size of 274.36 nm and EE of 76.56%. There was non-significant variation in result observed between experimental and predicted value. The closeness in result revealed that model is valid and reproducible. *3.5. Vesicle Evaluation* The vesicle size of prepared DC-BMs was found in range of 169.34 nm (F1)–380.14 nm (F4). The formulation (F1) prepared with composition cholesterol (A, 0.1%), lipid (B, 0.5%), surfactant (C, 0.5%), and bile salt (D, 1%) showed lowest size and formulation (F4) showed maximum size with composition cholesterol (0.5%), lipid (1.5%), surfactant (0.5%), and bile salt (1%). The difference in size found due to variation in used composition. From the result, it can be observed that used independent variables have shown a significant effect. 3D response surface plot (Figure 1a,b) and polynomial Equation (5) show that increasing

The size of prepared DC-BMs (F1-F27) was found between 169.34 nm and 380.14 nm. The optimized composition (DC-BMs-opt) showed VS of 270.21 ± 3.76 nm (Figure 3A).

was highly stable and in disaggregated form. The surface morphology exhibited spherical

shape vesicles with a smooth surface without any aggregation (Figure 3B).

a CHO (A) concentration led to an increase in vesicle size due to greater amount of CHO deposited into lipid bilayer [28]. The second factor lipid concentration (B) increases vesicle size due to enhancement in viscosity of dispersion and thickness of lipid bilayers. Similar types of findings reported in reported research of papain liposomes [29], and diclofenac liposome [30]. The surfactant concentration (PP123, C) and bile salt (D) depicted a negative effect on BNs size. The enhancement in surfactant (PP123) gave the reduction in vesicle size because at high concentration of surfactant, interfacial tension reduced between lipid phases and aqueous phase. The bile salt increases flexibility of liposomes by incorporating into lipid bilayer, thereby decreasing vesicle size [31].

$$\text{Viscide size (nm, R\_1)} = 274.36 + 49.20 \text{ A} + 50.23 \text{ B} - 30.45 \text{ C} - 18.69 \text{ D} \tag{5}$$

The F-value fitted to linear model and value found to be 646.04 revealed that model was significantly (*p* < 0.0001) fitted. The regression coefficient of best fit model was found to be 0.9915 and it indicates lesser variation between actual and predicted value (Table 2). The ANOVA value showed that model term cholesterol (A), lipid (B), surfactant (C), and bile salt (D) were found to be significant model term (*p* < 0.0001, Table 3). The adequate precision was 85.98 (<4), revealing the close relationship between actual and experimental value [32]. The *p*-value of lack of fit is >0.05 indicated insignificant which is good for model [33].
