Application of quality by design approach for

European Journal of Pharmaceutical Sciences 78 (2015) 54–66

Contents lists available at ScienceDirect

European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps

Application of quality by design approach for intranasal delivery of rivastigmine loaded solid lipid nanoparticles: Effect on formulation and characterization parameters Brijesh Shah a,1,2, Dignesh Khunt b,2, Himanshu Bhatt b,2, Manju Misra b,⇑, Harish Padh c a b c

Department of Pharmaceutics, B.V. Patel PERD Centre, Ahmedabad 380054, India Department of Pharmaceutics, NIPER-Gandhinagar, C/O B.V. Patel PERD Centre, Ahmedabad, India Sardar Patel University, Vallabh Vidyanagar, Advisor – NIPER-Gandhinagar, C/O B.V. Patel PERD Centre, Ahmedabad, India

a r t i c l e

i n f o

Article history: Received 5 March 2015 Received in revised form 18 June 2015 Accepted 1 July 2015 Available online 2 July 2015 Keywords: Intranasal route Brain targeting Rivastigmine Solid lipid nanoparticles Compritol 888 ATO Factorial design

a b s t r a c t In the present investigation, Quality by Design (QbD) approach was applied on the development and optimization of solid lipid nanoparticle (SLN) formulation of hydrophilic drug rivastigmine (RHT). RHT SLN were formulated by homogenization and ultrasonication method using Compritol 888 ATO, tween-80 and poloxamer-188 as lipid, surfactant and stabilizer respectively. The effect of independent variables (X1 – drug: lipid ratio, X2 – surfactant concentration and X3 – homogenization time) on quality attributes of SLN i.e. dependent variables (Y1 – size, Y2 – PDI and Y3 – %entrapment efficiency (%EE)) were investigated using 33 factorial design. Multiple linear regression analysis and ANOVA were employed to indentify and estimate the main effect, 2FI, quadratic and cubic effect. Optimized RHT SLN formula was derived from an overlay plot on which further effect of probe sonication was evaluated. Final RHT SLN showed narrow size distribution (PDI- 0.132 ± 0.016) with particle size of 82.5 ± 4.07 nm and %EE of 66.84 ± 2.49. DSC and XRD study showed incorporation of RHT into imperfect crystal lattice of Compritol 888 ATO. In comparison to RHT solution, RHT SLN showed higher in-vitro and ex-vivo diffusion. The diffusion followed Higuchi model indicating drug diffusion from the lipid matrix due to erosion. Histopathology study showed intact nasal mucosa with RHT SLN indicating safety of RHT SLN for intranasal administration. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction Abbreviations: BBB, blood brain barrier; CPP, critical process parameters; Compritol, Compritol 888 ATO; CQA, critical quality attribute; CV, coefficient of variation; D: L, drug: lipid; DOE, design of experiment; DSC, differential scanning calorimetry; %EE, % entrapment efficiency; EMEA, European Medicines Agency; GMS, glycerol monostearate; HLB, hydrophilic lipophilic balance; HPLC, high performance liquid chromatography; HSH, high speed homogenizer; HT, homogenization time; MHRA, medicines and healthcare products regulatory agency; PA, Precirol ATO 5; PBS, phosphate buffer saline; PDI, polydispersity index; Pol-188, Poloxamer 188; QbD, quality by design; RES, reticulo endothelial system; RHT, rivastigmine hydrogen tartrate; SLN, solid lipid nanoparticle; TEM, transmission electron microscopy; TPGS, alpha-tocopherol polyethylene glycol 1000 succinate; USFDA, United States food and drug administration; UV–Vis, UV–visible; XRD, X-ray diffraction. ⇑ Corresponding author at: Department of Pharmaceutics, NIPER-Gandhinagar, C/O B.V Patel PERD centre, S.G. Highway, Thaltej, Ahmedabad 380054, India. Communication Ref. No.: PERD310215. E-mail addresses: [email protected] (B. Shah), digneshkhunt80@gmail. com (D. Khunt), [email protected] (H. Bhatt), [email protected], mmisraniper@ yahoo.com (M. Misra), [email protected] (H. Padh). 1 Brijesh Shah is registered research scholar at Institute of Pharmacy, Nirma University. 2 All three authors have equal contribution for manuscript work and preparation. http://dx.doi.org/10.1016/j.ejps.2015.07.002 0928-0987/Ó 2015 Elsevier B.V. All rights reserved.

Alzheimer’s disease is a neurodegenerative disorder characterized by deficiency of acetyl choline in the brain resulting in loss of neurons, synapses, memory dysfunction and pathological changes like formation of abnormal protein aggregates known as senile plaques and neurofibrillary tangles (Lazenby, 2010; Serpell, 2000; Yates and McLoughlin, 2008; Williams et al., 2003). Rivastigmine hydrogen tartrate (RHT), a USFDA approved reversible cholinesterase inhibitor is a candidate of choice, used in the treatment of Alzheimer’s disease to treat mild to moderate dementia due to its favorable effect on patient’s cognitive and behavioral symptoms. RHT is a non-competitive dual inhibitor, which inhibits the metabolism of both acetyl cholinesterase and butyryl cholinesterase and helps in enhancing acetyl choline level to moderate Alzheimer’s disease by increasing central cholinergic function (Scarpini et al., 2003; Grossberg, 2003). RHT, a phenyl carbamate derivative undergoes extensive first-pass metabolism in the liver resulting in reduced absolute

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

bioavailability of only 36% after 3 mg dose, leading to restricted entry into brain and lesser concentration at the target site (Fazil et al., 2012). Owing to its hydrophilic nature, oral delivery of RHT necessitates frequent oral dosing, resulting into accumulation of severe cholinergic side effects in the systemic circulation (Wilson et al., 2008; Yang et al., 2013). Brain targeting of hydrophilic drugs like RHT is even more challenging, because the therapeutic molecules must be transported not only across the brain cell membrane, but also across the blood brain barrier (BBB). BBB an obstructive gatekeeper being lipophilic in nature, hinders the permeability of hydrophilic drugs and molecules above 500 Da, thereby making it difficult to treat many severe and life threatening neurological disorders including Alzheimer’s disease (Kaur and Bhandari, 2013; Eyal et al., 2009; Pardridge, 2012). Various strategies capable of delivering neurotherapeutics across BBB, includes invasive techniques requiring disruption of barrier integrity and noninvasive techniques. Noninvasive intranasal delivery has come to the forefront as an alternative to invasive delivery due to presence of direct connection between olfactory and trigeminal region in the upper nasal mucosa. This connection helps in delivering therapeutics into brain by circumventing BBB and minimizing systemic exposure thereby providing rapid absorption and enhancing drug influx at BBB (Alam et al., 2010; Illum, 2000). Nanotechnology based delivery systems hold great potential for delivery to brain when given intranasally. These systems are efficiently taken up by nasal mucosa thereby increasing their specificity, bioavailability, duration of therapeutic effect, besides offering higher drug loading and protection against enzymatic and/or chemical degradation (Kumar et al., 2014a; Mittal et al., 2014). Nanoparticulate carriers (lipid emulsions, polymeric nanoparticles, liposomes and micelles) within size range of 10– 400 nm, allow easy access across BBB by efficiently encapsulating drug molecules and increasing their diffusion through biological membranes (Bagwe et al., 2001; Chang et al., 2009). However, brain delivery with nanoparticulate system like polymeric nanoparticles shows carrier’s interaction with reticulo endothelial system (RES), resulting in rapid clearance from blood circulation (Costantino et al., 2009). In this regard, solid lipid nanoparticles (SLN) are considered as an attractive colloidal carrier system for brain targeting, since they showed higher ability to escape the RES, thereby prolonging the residence time (Brioschi et al., 2009). SLN are defined as lipidic nanocarriers generally spherical in shape with an average diameter between 10–1000 nm containing biocompatible solid lipid core matrix (mono-di and tri glycerides, fatty acids, steroids and waxes) stabilized by various classes of emulsifiers. They are preferred over other colloidal carriers because of their lipophilic nature and other versatile properties like high drug payload, controlled release, drug targeting and feasibility to incorporate both hydrophilic and lipophilic drugs, make SLN an efficacious carrier system for a wide range of therapeutics facing challenges in the area of brain targeted delivery system (Patel et al., 2011; Kaur and Bhandari, 2013). The mechanism behind SLN uptake by the brain is believed to be an interaction between plasma proteins adsorbed onto SLN surface and endothelial cells of BBB thereby facilitating adhesion and subsequently activating endocytotic process (Kreuter, 2001; Brioschi et al., 2009). In the past few years, regulatory agencies like USFDA, EMEA and MHRA etc. are emphasizing on the concept of quality by design (QbD) to develop a better quality product by understanding critical process and product parameters based on risk management (ICH, 2009; FDA, 2006). Design of experiment (DOE), a part of QbD has a very significant role in evaluating the effect of large number of critical process parameters (CPP) on critical quality attributes (CQA) of the product (Wechsler, 2008). DOE helps in the

55

development of quality product by minimizing the number of experiments which are often costly and time consuming. The present investigation was aimed at applying QbD approach for the development of hydrophilic drug loaded SLN for intranasal delivery. 33 factorial design was applied to optimize process parameters affecting quality attributes of SLN. RHT SLN were formulated and they were characterized for physicochemical, morphological, in-vitro, ex-vivo diffusion and histopathological parameters. 2. Material and methods 2.1. Materials RHT was received as a gift sample from Cadila Pharmaceuticals Ltd. (Ahmedabad, India). Apifil, Compritol 888 ATO (Compritol) and Precirol ATO 5 (PA) were gift samples received from Gattefosse Pvt. Ltd. (Mumbai, India). Stearic acid, Tween 80, Poloxamer-188 (Pol-188), D-alpha-Tocopherol polyethylene glycol 1000 succinate (TPGS) and Glycerol monostearate (GMS) were purchased from Sigma-Aldrich (Bangalore, India). All other chemicals and reagents were of analytical reagent grade and were used without further purification. 2.2. High-performance liquid chromatography analysis of RHT RHT was analyzed using high-performance liquid chromatography system (HPLC) LC-2010C HT (Shimadzu, Japan) which consisted of UV/VIS detector and Labsolutions chromatographic software. A reverse phase C18 column (250 x 4.6 mm, 5 l, kinetex, Phenomenex, USA) was used at room temperature. Mixture of acetonitrile and potassium dihydrogen orthophosphate buffer pH 6.0 (20:80 v/v) was used as a mobile phase at a flow rate of 1.0 ml/min. Injection volume was 10 ll and elute was analyzed at 215 nm. The R2 value of 0.9996 was found to be linear in the concentration range of 1–100 lg/ml. 2.3. Solubility study of RHT in different solid lipids It was not possible to determine the solubility of RHT by equilibrium method since different lipids taken into study were solid in nature. Hence an alternative method was followed, wherein drug and lipids were mixed in two different drug: lipid ratios (D: L) viz., 1:2 and 1:3 individually as shown in Table 1. Each test tube containing mixtures of drug and lipids were melted 5 °C above the melting point of lipid using water bath and mixed using a cyclo mixer (CM 101, REMI, Mumbai, India). Process of heating and mixing were continued for five minutes and test tubes were observed visually for miscibility and clarity (Das et al., 2011). 2.4. Formulation of RHT SLN SLN were formulated by homogenization and ultra sonication method (Mehnert and Mäder, 2012). Formulation procedure was divided into two parts in which one part contained lipid and drug while other part contained aqueous solution of surfactant and stabilizer. Drug and lipid mixture was melted 5 °C above the melting point of lipid. Aqueous part was heated at the same temperature. When both parts attain equilibrium, the aqueous phase was incorporated into lipid phase and emulsified using high speed homogenizer (HSH, Kinematica AG, Polytron PT 1600 E, Switzerland). Temperature was maintained constant throughout the emulsification process. Primary emulsification was followed by ultrasonication using a probe sonicator (SONICS, VibraCell, VC 505, USA) and temperature was kept constant (Das et al., 2011). Resulting lipidic dispersion was cooled down at room temperature for 15 min to obtain RHT SLN.

56

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

Table 1 Solubility study (n = 3). Lipids

Melting point (°C)

Apifil Compritol GMS PA Stearic acid

62–65 65–77 55–60 52–55 69–70

Drug: lipid ratio (D: L) 1:2

1:3

+ ++ ++ ++ +

++ +++ +++ +++ ++

+ Not clear, ++ Turbid, +++ Clear.

2.5. Quality target product profile (QTPP) and risk analysis of RHT SLN The QTPP is described as the quality properties that a drug product need to possess so as to fulfill the objectives set in target product profile as quantitative attributes. QTPP should furnish a quantitative surrogate to describe the aspects of clinical safety and efficacy by determining the CQA, CPP and control strategy (ICH, 2009). In case of RHT SLN, QTPP is a lower size and PDI with lipidic core is expected to facilitate transport of drug across the nasal mucosal barriers both into the cerebral tissues and systemic circulation. Lower PDI is to reduce aggregation of particle during long term stability. Higher entrapment efficiency is to achieve higher drug loading in lipid matrix (Vora et al., 2013). The crucial step in risk assessment is to gather the entire responsible factor systematically that could influence the desired product quality. These factors were categorized hierarchically using an Ishikawa diagram (Fig.1). The parameters summarized in Ishikawa diagram assisted in the identification of failure modes of SLN formulation. 2.6. Preliminary screening of lipid, stabilizer and HSH rpm In DOE, preliminary screening is a crucial step for the selection of CPP which affects CQA like size, PDI and % entrapment efficiency (%EE) of developed SLN. Preliminary screening of lipid, stabilizer and HSH rpm was carried out so as to obtain SLN of smaller size, narrow PDI and higher %EE. Based on solubility study, D: L and lipids were screened for optimization of final lipid. Further these lipids were taken and different batches of RHT SLN as shown in Table 2 were prepared by method described in Section 2.4 using tween 80 (2% w/w) as surfactant and HSH at 10000 rpm for homogenization time (HT) 10 min without probe sonication. To screen out lipid Batch nos. 1–3 (Table 2) were characterized for size, PDI and %EE. After selection of final lipid, D: L, HSH rpm and time were kept constant and the effect of stabilizer on CQA was verified by formulating RHT SLN using surfactant alone (2% w/w tween 80, Batch no. 3, Table 2) and combination of surfactant with stabilizer (Batch no. 4 and 5, Table 2). Subsequently the effect of higher HSH rpm (Batch no. 6) on CQA was determined as shown in Table 2. 2.7. 33 factorial design Based on preliminary experimental data, 33 factorial design was selected for the optimization of RHT SLN where the effect of three independent variables or CPP viz., D: L (X1), surfactant concentration (X2) and HSH time (X3) on CQA (Y1 = size, Y2 = PDI and Y3 = %EE) was determined at three different levels (Table 3 and 4). Compritol, tween 80 and Pol-188 were selected as lipid, surfactant and stabilizer respectively. HSH rpm (10000) and Pol-188 concentration (1% w/w) were set as fix levels. 33 Factorial design was analyzed using Design expert software (Version 8, Stat-ease. Inc, USA) and the polynomial equation was derived. The magnitude of coefficients in polynomial equation

Fig. 1. Ishikawa diagram illustrating CPP affecting on CQA of RHT SLN.

have either positive sign, indicating synergistic effect or negative sign, indicating antagonistic effect. Best fitting experimental model (linear, two factor interaction, quadratic and cubic model) was taken statistically on the basis of comparison of several statistical parameters like coefficient of variation (CV), multiple correlation coefficient (R2), adjusted multiple correlation coefficient (adjusted R2), predicted residual sum of square and graphically by 3D response surface plot provided by Design Expert software. The level of significance was considered at p-value < 0.05. The regression analysis, linear regression plots (observed versus predicted value) and Pareto chart of the dependent variables were plotted using MS-Excel. 2.8. Data optimization and model validation The effect of each independent CPP on CQA was analyzed for establishment of design space with the target of ensuring desired product quality. Hence, 33 factorial design was applied for establishment of design space to investigate the responses of process parameters on quality attributes of RHT SLN. The optimization was done on the basis of attaining lower particle size, PDI and higher %EE using overlay plot (graphical) and desirability (numerical) criteria. In order to establish the reliability of developed model, check-point analysis was performed by taking two validation batches viz., V1 and V2 (Table 4) whereby magnitude of error between observed and predicted values was evaluated. 2.9. Effect of probe sonication on optimized RHT SLN It is often sonication which brings about further reduction in size with narrow PDI. Sonication breaks coarse drops into nano droplets and hence, it is responsible in obtaining smaller particle size of SLN (Das et al., 2011). To evaluate the effect of probe sonication on size, PDI and %EE of RHT SLN, different trials were taken by varying amplitude and sonication time as shown in Table 5. The final optimized formula with composition of RHT SLN is shown in Fig.2. 2.10. Characterization of RHT SLN For characterization, three batches of optimized RHT SLN were formulated and they were characterized for physicochemical, morphological, diffusion and histopathological parameters as shown below. 2.10.1. Particle size, PDI and Zeta Potential Particle size, PDI and zeta potential measurements were performed by photon correlation spectroscopy using Zetasizer (Nano-ZS90, Malvern, Worcestershire, UK). Before measuring size,

57

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66 Table 2 Preliminary screening of lipid, stabilizer and HSH rpm (data represents mean ± SD, n = 3). Batch no.

Lipid (D: L, 1:3)

Surfactant (2% w/w)

Stabilizer (1% w/w)

HSH RPM 10 min

Size (nm)

PDI

% EE

1 2 3 4 5 6

GMS PA Compritol Compritol Compritol Compritol

Tween-80 Tween-80 Tween-80 Tween-80 Tween-80 Tween-80

TPGS Pol-188 Pol-188

10000 10000 10000 10000 10000 15000

587.50 ± 4.98 468.90 ± 3.53 345.00 ± 2.32 274.70 ± 5.28 189.20 ± 3.08 244.60 ± 4.19

0.589 ± 0.071 0.527 ± 0.048 0.320 ± 0.016 0.292 ± 0.091 0.237 ± 0.055 0.347 ± 0.039

47.86 ± 2.14 44.36 ± 2.63 51.51 ± 3.09 53.75 ± 3.33 57.60 ± 2.40 44.42 ± 3.29

Table 3 Selection of variable and their levels. Type of variables

Levels

Independent variables (CPP) X1 = Drug: lipid ratio (D: L) (%w/w) X2 = Surfactant concentration (S) (%w/w) X3 = HSH Time (HT) (min)

Low (1)

Medium (0)

High (+1)

1:3 2 5

1:4 3 10

1:5 4 15

Dependent variables (CQA) Y1 = Size (nm) Y2 = PDI Y3 = %EE

PDI and zeta potential, RHT SLN were diluted 10 times with double distilled water. Particle size and PDI measurements were performed by taking 1 ml of diluted formulation into polystyrene cuvettes and disposable folded capillary cell for zeta potential at 25 °C, respectively. The dynamic light scattering measurements were taken at the wavelength of 633 nm using helium-neon laser as a light source at the scattering angle of 90°, where diffusion of particle due to Brownian motion gets converted into particle size. In Table 4 Size, PDI and %EE of factorial runs and validation batches (data represents mean ± SD, n = 3).

a

Run

X1a

X2

X3

Y1

Y2

Y3

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 V1 V2

3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 4 4 4 3.5 4.5

2 2 2 3 3 3 4 4 4 2 2 2 3 3 3 4 4 4 2 2 2 3 3 3 4 4 4 3 3 3 2.5 3.5

5 5 5 5 5 5 5 5 5 10 10 10 10 10 10 10 10 10 15 15 15 15 15 15 15 15 15 10 10 10 7.5 12.5

262.5 ± 5.29 240.1 ± 4.87 289.7 ± 3.43 220.9 ± 6.55 201.6 ± 4.11 243.2 ± 5.01 280.2 ± 3.72 287.3 ± 4.36 307.8 ± 5.63 189.2 ± 3.08 167.3 ± 3.51 230.5 ± 4.13 176.4 ± 5.29 140.5 ± 2.05 200.4 ± 1.33 210.2 ± 2.58 224.1 ± 4.77 241.6 ± 6.63 195.1 ± 5.97 184.9 ± 6.28 247.7 ± 8.39 190.2 ± 3.99 172.4 ± 9.25 226 ± 4.58 250.7 ± 6.89 263.2 ± 7.15 253.1 ± 4.92 134.3 ± 7.50 149 ± 6.25 144.2 ± 4.10 168.6 ± 2.14 181.2 ± 4.98

0.32 ± 0.074 0.292 ± 0.022 0.329 ± 0.039 0.307 ± 0.051 0.259 ± 0.029 0.327 ± 0.043 0.363 ± 0.036 0.37 ± 0.064 0.395 ± 0.025 0.237 ± 0.055 0.212 ± 0.047 0.272 ± 0.059 0.251 ± 0.067 0.202 ± 0.018 0.279 ± 0.083 0.309 ± 0.077 0.321 ± 0.045 0.300 ± 0.089 0.427 ± 0.049 0.404 ± 0.098 0.455 ± 0.057 0.387 ± 0.032 0.362 ± 0.063 0.402 ± 0.087 0.442 ± 0.091 0.467 ± 0.056 0.447 ± 0.028 0.189 ± 0.075 0.201 ± 0.062 0.210 ± 0.070 0.225 ± 0.037 0.280 ± 0.051

52.2 ± 2.13 58.7 ± 1.89 63 ± 1.11 54.7 ± 2.56 64.1 ± 2.44 71.9 ± 1.21 47.2 ± 1.35 51.3 ± 2.47 61.7 ± 3.12 57.6 ± 2.40 67.3 ± 1.39 73.2 ± 2.78 65.1 ± 3.29 74.3 ± 2.87 79.1 ± 1.55 52.7 ± 2.31 58.8 ± 3.33 67.6 ± 3.59 43.2 ± 3.54 49.2 ± 2.65 59.8 ± 2.48 45.3 ± 1.66 53.6 ± 2.09 62.1 ± 2.98 39.81 ± 3.23 47.3 ± 2.29 55.5 ± 1.73 77.1 ± 1.19 73.5 ± 1.47 71.8 ± 1.56 65.5 ± 2.54 69.3 ± 2.91

Drug concentration was maintained constant (1% w/w) in all D: L.

case of zeta potential, due to application of an electric field, particles move with a velocity related to their zeta potential which is measured using a technique called phase analysis light scattering and gets converted to the zeta potential by inbuilt software. 2.10.2. % Entrapment efficiency (%EE) The %EE of formulated RHT SLN was determined by centrifugation method. Samples were taken in centrifuge tubes and centrifuged at 10000 rpm for 20 min at room temperature in order to obtain pellet of lipid nanoparticles. Supernatant was collected, suitably diluted with methanol and analyzed for free drug content by UV spectroscopy. %EE was calculated by following equation: (Rahman et al., 2010).

%EE ¼

Total amount of RHT  Amount of free RHT Total amount of RHT

2.10.3. Assay The formulation equivalent to 10 mg of RHT was diluted 10 times with methanol and final dilution was made with mobile phase. Samples were prepared and RHT content was determined by means of HPLC method (Venkateswarlu and Manjunath, 2004). 2.10.4. pH pH of the RHT SLN were determined by taking 10 ml of formulation in a beaker. pH was measured at room temperature using a calibrated digital pH meter (EUTECH pH Tutor, Singapore).

Table 5 Effect of probe sonication on optimized RHT SLN (data represents mean ± SD, n = 3). Batch. no.

Amplitude (%)

Time (min)

Size (nm)

PDI

%EE

0

1

0

148.97 ± 2.17

0.213 ± 0.054

72.97 ± 2.23

2 3 4

20

1 2.5 5

121.6 ± 3.87 82.5 ± 4.07 169.1 ± 5.68

0.187 ± 0.039 0.132 ± 0.016 0.290 ± 0.043

70.11 ± 1.87 66.84 ± 2.49 60.53 ± 2.71

5 6 7

30

1 2.5 5

107.4 ± 4.56 146.2 ± 5.65 232.7 ± 8.94

0.174 ± 0.067 0.233 ± 0.029 0.321 ± 0.079

68.19 ± 3.02 62.77 ± 2.30 55.86 ± 3.45

58

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

Mumbai, India) and analyzed for drug content by HPLC at 215 nm. Percentage drug diffused (mean values) was plotted versus time (h).

Fig. 2. Optimized formula for RHT SLN and characterization parameters (n = 3). (DDW- double distilled water, HSH- high speed homogenization, PS- probe sonication, ZP- zeta potential).

2.10.5. Differential Scanning Calorimetry (DSC) DSC analysis of RHT, Compritol, physical mixture, blank SLN and RHT SLN were performed using DSC Q20 (V24.9 build 121, TA instrument, USA). In case of blank SLN and RHT SLN, samples were lyophilized prior to DSC analysis. Samples (4–6 mg) were sealed in standard aluminum pans and analysis was performed at heating rate of 10 °C/min from 25 to 150 °C under a nitrogen atmosphere with a flow rate of 50 ml/min. An empty sealed aluminum pan was used as reference. 2.10.6. X-ray diffraction (XRD) study The XRD study of Compritol, RHT, lyophilized blank SLN and lyophilized RHT SLN were carried out using X-Ray diffractometer (X’PERT MPD, Philips, Holland) having Cu anode material (1.54 Å) at 30 mA, 40 kV. Analysis was performed in continuous mode with step size of 0.017° over an angular range (2h) of 3–50° with Xe filled counterate detector. Obtained diffractograms were analyzed using JCPDF database for diffractometry software. 2.10.7. Morphological characterization The morphology of optimized RHT SLN was observed under transmission electron microscopy (TEM; Philips, Tecnai 20, Holland) at an acceleration voltage of 200 kV and viewed at a magnification of 50000. The size of the RHT SLN was measured using AnalySISÒ software (Soft Imaging Systems, Reutlingen, Germany). To perform TEM observations, RHT SLN were diluted with double distilled water (1:10) and deposited on the carbon-coated copper grid, stained by 1% aqueous solution of phosphotungestic acid. RHT SLN sample were allowed to adsorb on the holey film grid and observed after drying (Shah et al., 2014). 2.10.8. In-vitro diffusion study of RHT SLN Franz diffusion cell (Hanson Research – Telemodul 40 S, Chatsworth, CA) with a receptor volume capacity of 12.5 ml was used to perform an in-vitro diffusion of RHT SLN and RHT solution using cellulose acetate membrane (MWCO- 12000–14000 Da, pore size- 2.4 nm, HIMEDIA, Mumbai, India) and PBS pH 6.4 as a dialyzing medium (Arumugam et al., 2008). Cellulose membranes were soaked in PBS for overnight and then mounted between the receptor and donor compartment. Comparative diffusion study was performed in triplicate by taking RHT SLN and aqueous solution equivalent to 10 mg/ml of RHT in the donor compartment. PBS in the receptor compartment was stirred continuously with magnetic bead in a manner that PBS touches the membrane surface. Aliquots of 1 ml were withdrawn at different time intervals (0.5, 1, 2, 4, 6 and 8 h) and replaced with an equal volume of PBS. Samples were filtered through syringe filter (Millex-GV, 0.22 mm, Millipore,

2.10.9. Ex-vivo diffusion study An ex-vivo diffusion was carried out using Franz diffusion cell with a receptor volume capacity of 12.5 ml using goat nasal mucosa as a dialyzing membrane. The freshly excised goat nasal mucosa was collected from the slaughter house and cleaned to remove adhered tissues. It was rinsed thoroughly with PBS pH 6.4 and allowed to equilibrate in PBS for 15–20 min. Nasal mucosa having thickness of 0.2 mm (measured using vernier caliper, CD-600 CSX digital, Mitutoyo Corp., Kanagawa, Japan) was mounted between the receptor and donor compartment. Donor compartment was filled with RHT SLN and aqueous solution equivalent to 10 mg/ml of RHT. Similar procedure as described in Section 2.10.8 was followed to determine the amount of drug diffused across excised nasal mucosa using HPLC method. Study was performed in triplicate and mean values for percentage drug diffused was plotted against time (h) (Florence et al., 2011). Flux (lg/cm2/h) and diffusion coefficients (cm2/h) values were calculated from the slope of plot obtained between amount of drug permeated/unit area (lg/cm2) of mucosal membrane versus time (h) (Jadhav et al., 2010). The data obtained from ex-vivo study was fitted to different kinetic models viz., zero order (cumulative percentage of drug release versus time), first order (log cumulative of drug remaining versus time) and Higuchi model (cumulative percentage of drug release versus square root of time) (Costa and Lobo, 2001). 2.10.10. Nasal histopathology Freshly excised goat nasal mucosa was collected from a slaughter house to perform nasal histopathology study (Seju et al., 2011; Kumar et al., 2009). Three pieces of nasal mucosa with even thickness (0.2 mm) were selected. First piece was treated with positive control (Isopropyl alcohol) for 1 hr. Second and third piece of mucosa were treated with negative control (PBS pH 6.4) and RHT SLN respectively for 1 h. After 1 h, mucosa were washed with PBS and immersed into 10% v/v formalin solution for overnight. All three pieces of mucosa were embedded in paraffin blocks and cut by a microtome into sections having thickness of 5 lm. Sections were stained with hematoxylin–eosin and observed under inverted microscope (Olympus-IX51, USA) to evaluate any damage to mucosa. 2.11. Statistical analysis All data were reported as mean ± standard deviation and the difference between the groups were tested using student’s t-test at the level of p < 0.05. More than two groups were compared using ANOVA, with p < 0.05 considered statistically significant. 3. Results and discussion 3.1. Solubility of RHT in solid lipids In SLN formulation, the amount of drug being entrapped into solid lipid is a limiting factor. Optimization of solid lipid plays a vital role, since amount of drug being solubilized has a major impact on %EE of drug. As per data shown in Table 1, five different solid lipids were taken to perform solubility study. Among them GMS, Compritol and PA showed higher solubility than Apifil and stearic acid at D: L (1:3) due to presence of mono, di, triacylglycerols and glycerides (Rowe et al., 2009; Rohit and Pal, 2013).

59

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66 Table 6 Lack of fit test. Response

Source

SS

df

MS

F Value

p Value Prob > F

Size (nm)

Linear 2FI Quadratic Cubic Pure error

48291.8 47715 3423.78 1070.12 115.38

23 20 17 10 3

2099.643 2385.75 201.3985 107.0124 38.46

54.59291 62.03199 5.236571 2.782433 –

0.0035 0.0029 0.0989 0.2163 –

PDI

Linear 2FI Quadratic Cubic Pure error

0.150 0.148 0.007 0.004 0.000

23 20 17 10 3

0.0065 0.00741 0.000437 0.000379 0.000075

86.67213 98.7955 5.827495 5.053037 –

0.0017 0.0014 0.0859 0.1046 –

% EE

Linear 2FI Quadratic Cubic Pure Error

1897.59 1892.20 81.24 55.74 14.67

23 20 17 10 3

82.50373 94.60975 4.778883 5.573548 4.889167

16.8748 19.3509 0.977443 1.139979 –

0.0194 0.0160 0.5929 0.5149 –

SS – Sum of Squares, df – Degree of freedom, MS – Mean Square.

Table 7 ANOVA of Size, PDI and% EE. Response

Source

df

SS

MS

F Value

Significance F

Size (nm)

Regression Residual Total

9 17 26

42359.84 3272.09 45631.94

4706.65 192.48 –

24.45 – –

5.12  1008 – –

PDI

Regression Residual Total

9 17 26

0.1427 0.0069 0.1497

0.0159 0.0004 –

38.93 – –

1.35  1009 – –

% EE

Regression Residual Total

9 17 26

2554.99 65.42 2620.41

283.89 3.85 –

73.77 – –

7.61  1012 – –

Therefore GMS, Compritol and PA were taken further for preliminary optimization.

3.2. Preliminary screening of lipid, stabilizer and HSH rpm From the result of solubility study three batches of SLN were formulated using three different solid lipids (Batch nos. 1–3, Table 2) in which Compritol in D: L (1: 3) showed lowest size and PDI of 345 ± 2.32 nm and 0.320 ± 0.016 respectively in comparison to GMS and PA. The %EE of 51.51 ± 3.09 was found to be highest with Compritol in comparison with other two lipids. This could be attributed to the fact that Compritol being a mixture of mono, di and triglycerides (glyceryl behenate) along with fatty acids of different chain lengths forms less ordered crystals with many lattice defects which helps in accommodating large amount of drug (Hippalgaonkar et al., 2013; Muller et al., 2000). Hence, Compritol was selected as an optimized lipid and further 1% w/w of TPGS and Pol-188 were taken as stabilizers and/or surfactants (Batch nos. 4 and 5, Table 2). It has been reported that SLN stabilized using mixture of surfactants have lower particle size and higher stability when compared to formulations having only one surfactant (Mehnert and Mäder, 2012). Table 2 clearly shows that size and PDI was lowest with tween 80 and Pol-188 in combination (Batch no. 5) when compared to surfactant alone (Batch no. 3) and in combination with TPGS (Batch no. 4). This data was in accordance with the results reported by Hippalgaonkar et al., 2013. They have reported lower size and PDI when combination of tween 80 and Pol-188 was used as emulsifiers and higher size and PDI when tween 80 was used alone. This was assigned to the role of stabilizer in presence of surfactant, which reduces the surface tension thereby facilitating the

particle partition during the size reduction process. In other words, combination of emulsifiers and their HLB values could lead to different surface absorption. In our case, presence of two surfactants viz., tween 80 and Pol-188 rapidly covered the new lipid surfaces generated during shearing process thereby, avoiding aggregation and increasing surface area (Liu et al., 2007; Hippalgaonkar et al., 2013). Higher %EE of 57.60 ± 2.40 was achieved with combination of tween 80 and Pol-188 (Batch no. 5), because as described by Liu et al. (2007) the addition of Pol-188 results in slight increase in the viscosity of the external phase thereby, reducing diffusion speed of drug towards external phase thus increasing the %EE of SLN. It was also observed that, when HSH RPM was increased from 10000 to 15000, size (189.2 ± 3.08–244.6 ± 4.19 nm) and PDI (0.237 ± 0.055–0.347 ± 0.039) were gradually increased and %EE (57.60 ± 2.40–44.42 ± 3.29) was decreased (Batch no. 6, Table 2). This could be ascribed to the fact that once stable lipid core with lower size and PDI is formed; higher kinetic energy will not play any further role on size and PDI. High kinetic energy results in poor drug loading capacity as the lipid matrix is ruptured at higher RPM and there are chances for drug molecule to leach out from lipid matrix into external phase. Therefore, HSH RPM (10000) and Pol-188 (1% w/w) were set as fix levels for further optimization of RHT SLN using 33 factorial design.

3.3. 33 factorial design 33 factorial design was performed to analyze the main effect, interaction effect and quadratic effect of three independent variables (CPP) on three responses (CQA). For this, thirty experimental runs were conducted and results were shown in Table 4. A probability value (p-value) by regression analysis was shown in Table 6

60

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

Fig. 3. 3D response surface plot of CQA; (A) Size, (B) PDI and (C) %EE.

and an F test value was shown in Table 7 for three CQA with 95% confident level. The large p value for lack of fit (>0.05) indicates that the lack of fit test is insignificant implying that significant model correlation existed between the CPP and CQA. The polynomial equations for size, PDI and %EE were as per following;

Size ¼ 152:07 þ 14:70  D : L þ 17:29  S  19:44  HT  4:97  D : L  S þ 1:14  D : LtimesHT þ 4:69  S  HT þ 25:14  ðD : LÞ2 þ 43:44  S2 þ 42:01 HT2

0.9786 respectively. Data shown in Table 6 indicated that the p-value for all three CQA was >0.05 for cubic and quadratic model. These results were in accordance with the 3D response surface plot (Fig. 3) and Pareto chart (Fig. 4). Table 7 revealed that the significance value for all three dependent variables was 0.9 for predicted versus observed value were found to be linear for all CQA responses indicating good correlation between them. Magnitude of error is useful to establish the reliability of generated equations and to express the relevant domain of model (Bhatt et al., 2014). The prediction error was found to vary between 1.32 and 5.33 which depicts the reliability of the optimization procedure in predicting the effect of CPP on CQA. 3.5. Effect of D: L on size, PDI and %EE

Fig. 5. Overlay plot showing location of optimized RHT SLN in design space.

The particle size, PDI and %EE of factorial runs varied between 134.3 to 307.8 nm, 0.189 to 0.467 and 39.81 to 79.10% respectively as shown in Fig.3. As shown in Eqs. 1–3 positive sign denotes synergistic effect, while negative sign denotes antagonistic effect. It was observed that D: L had a positive effect on particle size (Y1), PDI (Y2) and %EE (Y3). It was also observed that as lipid concentration was increasing, viscosity of the dispersed phase was

Table 8 Observed versus predicted value of CQA. Run

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 V1 V2

Size (nm)

PDI

% EE

O

P

R

O

P

R

O

P

R

262.5 240.1 289.7 220.9 201.6 243.2 280.2 287.3 307.8 189.2 167.3 230.5 176.4 140.5 200.4 210.2 224.1 241.6 195.1 184.9 247.7 190.2 172.4 226 250.7 263.2 253.1 168.6 181.2

251.0 244.4 288.1 225.1 213.5 252.2 286.1 269.6 303.3 183.7 178.2 223.0 162.5 152.1 191.9 228.2 212.8 247.7 200.4 196.1 242.1 183.9 174.6 215.6 254.3 240.1 276.1 170.82 183.69

11.51 4.28 1.65 4.22 11.93 9.03 5.93 17.73 4.50 5.50 10.93 7.45 13.88 11.57 8.48 18.03 11.29 6.08 5.33 11.20 5.63 6.26 2.24 10.37 3.64 23.13 22.98 2.22 2.49

0.320 0.292 0.329 0.307 0.259 0.327 0.363 0.370 0.395 0.237 0.212 0.272 0.251 0.202 0.279 0.309 0.321 0.300 0.427 0.404 0.455 0.387 0.362 0.402 0.442 0.467 0.447 0.225 0.280

0.303 0.291 0.331 0.298 0.282 0.318 0.382 0.362 0.395 0.250 0.236 0.275 0.234 0.217 0.253 0.309 0.288 0.320 0.417 0.403 0.441 0.392 0.374 0.408 0.457 0.435 0.466 0.213 0.294

0.017 0.001 0.002 0.009 0.023 0.009 0.019 0.008 0.000 0.013 0.024 0.003 0.017 0.015 0.026 0.000 0.033 0.020 0.010 0.001 0.014 0.005 0.012 0.006 0.015 0.032 0.019 0.012 0.014

52.2 58.7 63 54.7 64.1 71.9 47.2 51.3 61.7 57.6 67.3 73.2 65.1 74.3 79.1 52.7 58.8 67.6 43.2 49.2 59.8 45.3 53.6 62.1 39.81 47.3 55.5 65.5 69.3

51.73 58.42 65.40 56.32 63.19 70.34 45.99 53.04 60.36 58.71 65.96 73.48 63.65 71.07 78.77 53.66 61.26 69.13 42.28 50.07 58.15 47.56 55.53 63.78 37.91 46.06 54.48 66.79 68.18

0.47 0.28 2.40 1.62 0.91 1.56 1.21 1.74 1.34 1.11 1.34 0.28 1.45 3.23 0.33 0.96 2.46 1.53 0.92 0.87 1.65 2.26 1.93 1.68 1.90 1.24 1.02 1.29 1.12

O – Observed, P – Predicted, R – Residual.

62

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

also increasing which resulted in particle agglomeration with higher size and PDI and decreased efficiency of homogenization. Increase in size and PDI along with higher %EE could be due to presence of higher amount of lipid which provides additional space for drug molecule to embed in, thereby decreasing total surface area. This can lead to reduction in the diffusion rate of the solute molecule as viscosity of the lipidic phase is higher and thus showed higher %EE (Shah and Pathak, 2010; Subedi et al., 2009). From the result shown in Table 4, it was observed that D: L (1: 3) showed poor %EE of 39.81 ± 3.23 as compared to D: L (1: 5), which showed %EE more than 55 in all the experimental runs, since amount of lipid was higher. 3.6. Effect of surfactant concentration on size, PDI and %EE As per results shown in Table 4 for various experimental runs, as the concentration of tween 80 increased from 2% to 3% w/w, size and PDI were decreasing with increase in %EE. With increasing concentration of tween 80 from 2% to 3% w/w, it showed reduced interfacial tension between lipid and aqueous phase which may control the aggregation of lipid particle by facilitating the particle partition thereby resulting into lower size and PDI (Mehnert and Mäder, 2012). As per previous reports, higher surfactant concentration stabilizes the lipid matrix effectively by forming steric barrier on their surface, thereby avoiding aggregation (Reddy and Murthy, 2005). It was observed that when 4% w/w tween 80 was used, size and PDI were continuously increasing while %EE was decreased. This was because; during homogenization process alkyl chain of surfactant molecule covers the surface of lipid particle via hydrophobic interaction to form a stable lipid matrix. Once this stable matrix is formed, excess surfactant may lead to accumulation of surfactant particles on the surface of stable lipid matrix causing increase in size and PDI as was observed in our case (Thakkar et al., 2014). Decrease in %EE could be attributed to higher solubilization effect produced by higher concentration of surfactant on RHT. At higher concentration of surfactant, solubility of RHT in the external phase may increase due to diffusion of drug from lipid core into aqueous phase leading to reduced %EE. The findings were in accordance with the outcome of other group, which describes lower %EE could be due to solubilization effect of emulsifier on drug molecule in aqueous phase (Abdelbary and Fahmy, 2009). 3.7. Effect of HSH time on size, PDI and %EE Homogenization was performed at 10000 RPM for three different time intervals viz., 5, 10 and 15 min. As per Table 4, with increase in HT from 5 to 10 min, size and PDI were gradually decreasing while at 15 min both size and PDI were increasing. Homogenization speed and time for which it is applied is one of the important strategy of applying kinetic energy to achieve lower size and PDI. Applying high kinetic energy for longer time may lead to instability of formed lipidic structures thereby resulting into aggregation and formation of larger particles. Size and PDI were higher at 5 min when compared to 10 min of HT due to insufficient homogenization. Therefore an optimum duration of homogenization will result in formation of stable particles with uniform size distribution. It was observed that %EE was less in 15 min than at 10 min. This was owing to removal of surfactant particle from lipid surface, thereby causing lipid disruption and escape of entrapped drug into aqueous phase. 3.8. Effect of probe sonication on RHT SLN Further size reduction of optimized RHT SLN was performed by ultra sonication at 20% and 30% amplitude for the time period of 1, 2.5 and 5 min and the results are tabulated in Table 5. From the

result, it was observed that the size and PDI were found to be higher at 20% and 30% amplitude for 5 min. The possible explanation could be the higher sonication time, which disrupts the outer core of lipid, causing drug molecule to leach into external phase and form aggregates, resulting in higher size and PDI. This was further supported by other researchers who described increase in particle size beyond a given sonication energy threshold leading to formation of aggregates generally known as ‘‘sonication induced aggregate formation’’ (Taurozzi et al., 2011; Aoki et al., 1987). In similar way applying higher kinetic energy or amplitude on previously formed lipid particles may fetch out the entrapped drug into aqueous medium, resulting into reduced %EE and increased probability of probe to product contamination (Mehnert and Mäder, 2012). In fact in our case, formation of metal shade was observed when the amplitude of probe sonication was increased to 30%, as a result of product contamination. Size and PDI was found to be lowest at 20% amplitude for 2.5 min with %EE of 66.84 ± 2.49 which was slightly less in comparison to %EE of 70.11 ± 1.87 at 1 min, however with respect to size of 82.5 ± 4.07 nm at 2.5 min which was quite acceptable when compared to size of 121.6 ± 3.87 nm at 1 min (Table 5). Hence these optimized sonication parameters (20% amplitude at 2.5 min) were applied on CPP (X1–X3) described in Section 3.4 which resulted in final optimized RHT SLN. 3.9. Characterization of optimized RHT SLN Characterization parameters like size, PDI, zeta potential, %EE, assay and pH were performed for three individual formulations of RHT SLN and results were shown as mean ± SD in Fig. 2. Particle size of RHT SLN was found to be 82.50 ± 4.07 nm with PDI value of 0.132 ± 0.016 indicating narrow size distribution. Zeta potential is the electric charge on the particle surface which describes about the storage stability of the formulation. The zeta potential value of 3.20 ± 1.44 mV was found to be positive which could be attributed to drug molecule since the surfactants used were non-ionic in nature. Despite of the lower zeta potential value, SLN were found to be stable due to the presence of non-ionic surfactants which impart stability to the system by steric stabilization. Adsorption of these steric stabilizers decreases the zeta potential value and produces strong repulsion between particles thereby preventing aggregation during storage (Singh et al., 2010; Muller et al., 2000). Entrapping hydrophilic drug into solid lipid is one of the challenging task, since the type of lipid used in SLN critically affects %EE. Compritol being a mixture of mono, di and triglycerides along with fatty acids acts as an amphiphilic lipid and emulsifying agent which has tendency to form imperfect crystal lattice to encapsulate hydrophilic drugs like RHT, as was observed in our case with %EE of >60 (Rowe et al., 2009; Rohit and Pal, 2013). Assay value of 94.59 ± 3.12% indicated higher amount of RHT was present in total SLN system. RHT SLN showed pH value of 5.96 ± 0.35, which was within the normal pH range of 5–6.5 of human nasal mucosa and hence will not cause nasal irritation upon intranasal administration (Arora et al., 2002). 3.10. DSC analysis Fig. 6A shows DSC curves of Compritol, RHT, physical mixture, blank SLN and RHT SLN. Compritol and RHT showed sharp melting endotherm at 72.96 °C and 127.18 °C indicating their crystallinity. In case of blank SLN, the enthalpy of Compritol was reduced to 48.36 J/g from 91.38 J/g indicating change in the polymorphic state of lipid from crystalline to amorphous with more defects in the crystal lattice which can accommodate space for drug entrapment (Kumar et al., 2014b). With RHT SLN, enthalpy for Compritol was

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

63

Fig. 6. (A) Comparative DSC curves of Compritol, RHT, physical mixture, blank SLN and RHT SLN. (B) Comparative XRD of A – Compritol, B – RHT, C – Blank SLN and D – RHT SLN.

3.11. XRD analysis The XRD of Compritol showed sharp peaks at 2h scattered angles 21.18 and 23.38, while RHT showed sharp peak at 2h scattered angles 20.00, 22.20, 26.46, 33.63 and 36.36, proving their crystalline nature. However, reduction in intensity of Compritol and RHT characteristic peaks in blank SLN and RHT SLN indicates the reduction in crystallinity of RHT (Fig. 6B), which might be attributed to incorporation of drug in lipid matrix of Compritol. XRD results were in agreement with the DSC analysis. However, diffraction pattern of blank SLN and RHT SLN showed no significant difference in the peak pattern, indicating that the addition of RHT did not change the nature of SLN.

Fig. 7. TEM analysis of RHT SLN.

lowered to 22.58 J/g from 91.38 J/g and for RHT enthalpy was reduced to 0.516 J/g from 53.92 J/g indicating conversion of crystalline RHT to amorphous form in the SLN formulation due to incorporation of drug into melted lipid matrix. Presence of small endothermic peak at 124.73 °C in RHT SLN could be attributed to unentrapped RHT present in the aqueous phase shown in inset Fig. 6A.

3.12. TEM analysis Morphological examination of RHT SLN using TEM analysis revealed that SLN were spherical in shape (Fig. 7) and were in the size range of 50–200 nm which was in the further agreement with the size distribution performed using Zetasizer based on the principle of dynamic light scattering. From Fig. 7 it was observed that the SLN particles were having uniform size distribution which was confirmed by Zetasizer, showing PDI value below 0.150 indicating absence of aggregation.

Fig. 8. (A and B) In-vitro and ex-vivo diffusion of RHT SLN and RHT solution.

64

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

Table 9 Flux, diffusion coefficient and release kinetics of RHT SLN and RHT solution (RS) (data represents mean ± SD, n = 3). Formulation

Flux (lg/cm2/h)

Diffusion coefficient (cm2/h)

Release kinetics (R2) Zero order

First order

Higuchi

RHT SLN RS

9.135 ± 0.203 4.843 ± 1.04

(5.70 ± 0.1)  103 (3.02 ± 0.6)  103

0.9916 0.8989

0.9725 0.9309

0.9949 0.9670

Fig. 9. Microscopic images of nasal mucosa; (A) – Positive control, arrows indicate loss of epithelial cells with cell necrosis and internal tissue damage. (B) – Negative control treated mucosa indicates intact nasal mucosa, (C) – RHT SLN treated mucosa showing no nasociliary damage.

3.13. In-vitro diffusion study of RHT SLN Fig. 8A shows comparative in-vitro diffusion between RHT SLN and RHT aqueous solution as percentage drug diffused versus time (h). Percentage of RHT diffused up to 8 h was 56.89% with RHT SLN and 31.63% with RHT solution. During initial time points, slight increase in % drug diffusion in case of drug solution is due to higher aqueous solubility of RHT in aqueous medium, where as in case of RHT SLN, initial time required for drug to diffuse out from lipid core was higher in comparison to drug solution. The release of drug from SLN can be influenced by the nature of the lipid matrix and surfactant concentration. Higher release of RHT from SLN could be due to the amphiphilic nature of Compritol, which attributes good solubility to RHT followed by homogenous distribution of RHT within the lipid matrix. As shown in DSC and XRD data, lowering of enthalpy and intensity respectively with lipid in RHT SLN confirmed imperfection in the crystal lattice of Compritol. As already known, for less ordered and/or imperfect crystal lattice the melting process requires less energy than perfect crystal substance (Vivek et al., 2007; Westesen et al., 1997). This could be one of the reasons for higher diffusion in case of SLN since the lipid surface of RHT SLN was having imperfect crystal lattices through which drug diffused out.

Higuchi models were shown in Table 9. On the basis of ex-vivo data obtained after 4 h, maximum diffusion was seen with RHT SLN up to 8 h, indicating enhanced diffusion of RHT with SLN system compared to pure drug solution. RHT SLN showed highest values of flux and diffusion coefficient when compared to RHT solution indicating higher penetration enhancing property with SLN system. RHT being hydrophilic in nature cannot diffuse much across nasal membrane due to its lipophilic nature. The higher value of diffusion and diffusion coefficient for RHT SLN could be attributed to the lipidic nature of SLN system, which diffused across nasal mucosa in superior way compared to drug solution. Another reason for higher diffusion of RHT in case of SLN could be either formation of amorphous dispersion or solubilization of RHT in Compritol matrix as revealed from DSC and XRD data as shown in Fig. 6B. In general the solubility increases with increase in amorphicity and hence higher release was obtained with SLN, which was not in case of drug solution (Rohit and Pal, 2013). RHT SLN exhibited highest R2 value (0.9949) for Higuchi model. The possible mechanism for the drug release might be drug diffusion from the lipid matrix and degradation of lipids via erosion. Other probable factor could be, larger surface area of RHT SLN (Size < 100 nm) which could easily undergo partitioning between lipid and aqueous phase resulting into higher drug diffusion (Venkateswarlu and Manjunath, 2004; Dash et al., 2010).

3.14. Ex-vivo diffusion of RHT SLN 3.15. Nasal histopathology study The purpose behind performing an ex-vivo study using goat nasal mucosa is to assess drug diffusion through biological membrane simulating in-vivo barrier, since the cellulose acetate membranes are artificial membrane and hence the findings obtained from in-vitro study would not be suitable to rely on. Fig. 8B indicates the ex-vivo diffusion of 65.86% for RHT SLN and 37.82% for RHT solution. Flux and diffusion coefficients for RHT SLN and RHT solution along with R2 values for zero order, first order and

Histopathology study was performed with an aim to determine the toxic effects of excipients used in the SLN formulation. As shown in Fig. 9, the mucosa treated with positive control (Isopropyl alcohol) showed loss of epithelial cells, with an internal tissue damage indicating sloughing of epithelial cells of nasal mucosa. The epithelium layer was intact, neither nasociliary damage nor cell necrosis was observed when nasal mucosa was treated

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

with negative control (PBS pH 6.4) and RHT SLN. RHT SLN showed pH value of 5.96 ± 0.35, which was well within the limits of human nasal pH range (5–6.5). These observations revealed that the excipients used in RHT SLN formulation had no harmful effect on the nasal mucosa and seems to be safe for nasal administration. 4. Conclusion In the present work QbD approach was applied for the development of hydrophilic drug (RHT) loaded SLN as a novel concept. RHT SLN were prepared by homogenization and ultrasonication method. The individual effects of CPP on CQA were evaluated using 33 factorial design with the objective of achieving desired product quality. Optimized formula for RHT SLN was obtained from the overlay plot. %EE of >65 in SLN indicated higher incorporation of RHT. RHT SLN being lipidic in nature showed higher drug diffusion in comparison to drug solution having crystalline form of drug. This was further consolidated by DSC and XRD data which showed that RHT has lost its characteristic crystalline nature. RHT SLN did not show any nasociliary damage and/or cell necrosis indicating its safety for nasal administration. From the above findings it can be concluded that QbD could be successfully applied for the development of SLN based colloidal carrier system with fewer numbers of trials and better quality attributes. Conflict of interest The authors report no conflict of interest. The authors alone are responsible for the content and writing of paper. Acknowledgement The authors are grateful to Lady Tata Memorial Trust (Bombay, India), Department of Science and Technology (Grant No. IFA-LSBM-13, Delhi, India) and Department of Biotechnology (Grant No. BT/PR6541/GBD/27/438/2012, Delhi, India) for providing scholarship as financial assistance. The authors are thankful to Cadila Pharmaceutical Ltd. (Ahmedabad, India) for providing RHT as a gift sample and Gattefosse (Mumbai, India) for providing various excipients as gift samples. Authors are thankful to industries commissionerate (Govt. of Gujarat) for providing financial support for instrumentation facilities. References Abdelbary, G., Fahmy, R.H., 2009. Diazepam-loaded solid lipid nanoparticles: design and characterization. Aaps PharmSciTech 10, 211–219. Alam, M.I., Beg, S., Samad, A., Baboota, S., Kohli, K., Ali, J., Akbar, M., 2010. Strategy for effective brain drug delivery. Eur. J. Pharm. Sci. 40, 385–403. Aoki, M., Ring, T.A., Haggerty, J.S., 1987. Analysis and modeling of the ultrasonic dispersion technique. Adv. Ceram. Mater. 2, 209–212. Arora, P., Sharma, S., Garg, S., 2002. Permeability issues in nasal drug delivery. Drug Discov. Today 7, 967–975. Arumugam, K., Subramanian, G., Mallayasamy, S., Averineni, R., Reddy, M., Udupa, N., 2008. A study of rivastigmine liposomes for delivery into the brain through intranasal route. Acta Pharm. 58, 287–297. Bagwe, R.P., Kanicky, J.R., Palla, B.J., Patanjali, P.K., Shah, D.O., 2001. Improved drug delivery using microemulsions: rationale, recent progress, and new horizons. Crit. Rev. Ther. Drug. Carr. Syst. 18, 77–140. Bhatt, H., Naik, B., Dharamsi, A., 2014. Solubility enhancement of budesonide and statistical optimization of coating variables for targeted drug delivery. J. Pharm. http://dx.doi.org/10.1155/2014/262194. Brioschi, A.M., Calderoni, S., Zara, G.P., Priano, L., Gasco, M.R., Mauro, A., 2009. Solid lipid nanoparticles for brain tumors therapy: state of the art and novel challenges. In: Sharma, H.S. (Ed.), Progress in Brain Research, 180. Elsevier, London (Chapter 11). Chang, J., Jallouli, Y., Barras, A., Dupont, N., Betbeder, D., 2009. Drug delivery to brain using colloidal carriers. In: Sharma, H.S. (Ed.), Progress in Brain Research, 180. Elsevier, London (Chapter 1). Costa, P., Lobo, J.M.S., 2001. Modeling and comparison of dissolution profiles. Eur. J. Pharm. Sci. 13, 123–133.

65

Costantino, L., Tosi, G., Ruozi, B., Bondioli, L., Vandeli, M.A., Forni, F., 2009. Colloidal systems for CNS drug delivery. In: Sharma, H.S. (Ed.), Progress in Brain Research, 180. Elsevier, London (Chapter 3). Das, S., Ng, W.K., Kanaujia, P., Kim, S., Tan, R.B., 2011. Formulation design, preparation and physicochemical characterizations of solid lipid nanoparticles containing a hydrophobic drug: effects of process variables. Colloids Surf., B 88, 483–489. Dash, S., Murthy, P.N., Nath, L., Chowdhury, P., 2010. Kinetic modeling on drug release from controlled drug delivery systems. Acta Pol. Pharm. 67, 217–223. Eyal, S., Hsiao, P., Unadkat, J.D., 2009. Drug interactions at the blood–brain barrier: fact or fantasy? Pharmacol. Ther. 123, 80–104. Fazil, M., Md, S., Haque, S., Kumar, M., Baboota, S., Kaur Sahni, J., Ali, J., 2012. Development and evaluation of rivastigmine loaded chitosan nanoparticles for brain targeting. Eur. J. Pharm. Sci. 47, 6–15. Florence, K., Manisha, L., Kumar, B.A., Ankur, K., Kumar, M.A., Ambikanandan, M., 2011. Intranasal clobazam delivery in the treatment of status epilepticus. J. Pharm. Sci. 100, 692–703. Grossberg, G.T., 2003. Cholinesterase inhibitors for the treatment of Alzheimer’s disease: getting on and staying on. Curr. Ther. Res. 64, 216–235. Hippalgaonkar, K., Adelli, G.R., Hippalgaonkar, K., Repka, M.A., Majumdar, S., 2013. Indomethacin-loaded solid lipid nanoparticles for ocular delivery: development, characterization, and in vitro evaluation. J. Ocul. Pharmacol. Ther. 29, 216–228. ICH Q9, 2006. US Food and Drug Administration. Guidance for industry: Q9 quality risk management, June. Available from: . ICH Q8 (R2), 2009. Pharmaceutical development, August. Available from: . Illum, L., 2000. Transport of drugs from the nasal cavity to the central nervous system. Eur. J. Pharm. Sci. 11, 1–18. Jadhav, K.R., Shetye, S.L., Kadam, V.J., 2010. Design and evaluation of microemulsion based drug delivery system. Int. J. Adv. Pharm. Sci. 1, 156–166. Kaur, I.P., Bhandari, R., 2013. Solid lipid nanoparticles entrapping hydrophilic/ amphiphilic drug and a process for preparing the same. WO 2013/105101 A1. Kreuter, J., 2001. Nanoparticulate systems for brain delivery of drugs. Adv. Drug Deliv. Rev. 47, 65–81. Kumar, M., Pathak, K., Misra, A., 2009. Formulation and characterization of nanoemulsion-based drug delivery system of risperidone. Drug Dev. Ind. Pharm. 35, 387–395. Kumar, A., Pandey, A.N., Jain, S.K., 2014a. Nasal-nanotechnology: revolution for efficient therapeutics delivery. Drug Deliv. . Kumar, M., Kakkar, V., Mishra, A.K., Chuttani, K., Kaur, I.P., 2014b. Intranasal delivery of streptomycin sulfate (STRS) loaded solid lipid nanoparticles to brain and blood. Int. J. Pharm. 461, 223–233. Lazenby, R.B., 2010. Handbook of Pathophysiology, fourth ed. Lippincott Williams & Wilkins, Philadelphia. Liu, J., Gong, T., Wang, C., Zhong, Z., Zhang, Z., 2007. Solid lipid nanoparticles loaded with insulin by sodium cholate-phosphatidylcholine-based mixed micelles: preparation and characterization. Int. J. Pharm. 340, 153–162. Mehnert, W., Mäder, K., 2012. Solid lipid nanoparticles: production, characterization and applications. Adv. Drug Deliv. Rev. 47, 165–196. Mittal, D., MD, S., Hasan, Q., Fazil, M., Ali, A., Baboota, S., Ali J., 2014. Brain targeted nanoparticulate drug delivery system of rasagiline via intranasal route. Drug Deliv. . Muller, R.H., MaÈder, K., Gohla, S., 2000. Solid lipid nanoparticles (SLN) for controlled drug delivery – a review of the state of the art. Eur. J. Pharm. Biopharm. 50, 161–177. Pardridge, W.M., 2012. Drug transport across the blood–brain barrier. J. Cereb. Blood Flow Metab. 32, 1959–1972. Patel, S., Chavhan, S., Soni, H., Babbar, A.K., Mathur, R., Mishra, A.K., Sawant, K.K., 2011. Brain targeting of risperidone-loaded solid lipid nanoparticles by intranasal route. J. Drug Target. 19, 468–474. Rahman, Z., Zidan, A.S., Khan, M.A., 2010. Non-destructive methods of characterization of risperidone solid lipid nanoparticles. Eur. J. Pharm. Biopharm. 76, 127–137. Reddy, L.H., Murthy, R.S.R., 2005. Etoposide-loaded nanoparticles made from glyceride lipids: formulation, characterization, in vitro drug release, and stability evaluation. AAPs PharmSciTech 6, E158–E166. Rohit, B., Pal, K.I., 2013. A method to prepare solid lipid nanoparticles with improved entrapment efficiency of hydrophilic drugs. Curr. Nanosci. 9, 211– 220. Rowe, R.C., Sheskey, P.J., Quinn, M.E., 2009. Handbook of Pharmaceutical Excipients, sixth ed. Pharmaceutical Press and American Pharmacist Association. Scarpini, E., Schelterns, P., Feldman, H., 2003. Treatment of Alzheimer’s disease; current status and new perspectives. Lancet Neurol. 2, 539–547. Seju, U., Kumar, A., Sawant, K.K., 2011. Development and evaluation of olanzapineloaded PLGA nanoparticles for nose-to-brain delivery: in vitro and in vivo studies. Acta Biomater. 7, 4169–4176. Serpell, L.C., 2000. Alzheimer’s amyloid fibrils: structure and assembly. Biochim. Biophys. Acta 1502, 16–30. Shah, M., Pathak, K., 2010. Development and statistical optimization of solid lipid nanoparticles of simvastatin by using 23 full-factorial design. AAPS PharmSciTech 11, 489–496.

66

B. Shah, D. Khunt, H. Bhatt et al. / European Journal of Pharmaceutical Sciences 78 (2015) 54–66

Shah, B.M., Misra, M., Shishoo, C.J., Padh, H., 2014. Nose to brain microemulsionbased drug delivery system of rivastigmine: formulation and ex-vivo characterization. Drug Deliv. . Singh, S., Dobhal, A.K., Jain, A., Pandit, J.K., Chakraborty, S., 2010. Formulation and evaluation of solid lipid nanoparticles of a water soluble drug: zidovudine. Chem. Pharm. Bull. 58, 650–655. Subedi, R.K., Kang, K.W., Choi, H.K., 2009. Preparation and characterization of solid lipid nanoparticles loaded with doxorubicin. Eur. J. Pharm. Sci. 37, 508–513. Taurozzi, J.S., Hackley, V.A., Wiesner, M.R., 2011. Ultrasonic dispersion of nanoparticles for environmental, health and safety assessment-issues and recommendations. Nanotoxicology 5, 711–729. Thakkar, H.P., Desai, J.L., Parmar, M.P., 2014. Application of Box-Behnken design for optimization of formulation parameters for nanostructured lipid carriers of candesartan cilexetil. Asian J. Pharm. 8, 81–89. Venkateswarlu, V., Manjunath, K., 2004. Preparation, characterization and in vitro release kinetics of clozapine solid lipid nanoparticles. J. Control. Rel. 95, 627– 638. Vivek, K., Reddy, H., Murthy, R.S., 2007. Investigations of the effect of the lipid matrix on drug entrapment, in vitro release, and physical stability of olanzapine-loaded solid lipid nanoparticles. AAPS PharmSciTech 8, 16–24.

Vora, C., Patadia, R., Mittal, K., Mashru, R., 2013. Risk based approach for design and optimization of stomach specific delivery of rifampicin. Int. J. Pharm. 455, 169– 181. Wechsler, J., 2008. FDA proposes flexible oversight for quality systems. Pharm. Technol. Available from: . Westesen, K., Bunjes, H., Koch, M.H.J., 1997. Physicochemical characterization of lipid nanoparticles and evaluation of their drug loading capacity and sustained release potential. J. Control. Release 48, 223–236. Williams, B.R., Nazarians, A., Gill, M.A., 2003. A review of rivastigmine: a reversible cholinesterase inhibitor. Clin. Ther. 25, 1634–1653. Wilson, B., Samanta, M.K., Santhi, K., Kumar, K.P.S., Paramakrishnan, N., Suresh, B., 2008. Poly (n-butylcyanoacrylate) nanoparticles coated with polysorbate 80 for the targeted delivery of rivastigmine into the brain to treat Alzheimer’s disease. Brain Res. 1200, 159–168. Yang, Z.Z., Zhang, Y.Q., Wang, Z.Z., Wu, K., Lou, J.N., Qi, X.R., 2013. Enhanced brain distribution and pharmacodynamics of rivastigmine by liposomes following intranasal administration. Int. J. Pharm. 452, 344–354. Yates, D., McLoughlin, D.M., 2008. The molecular pathology of Alzheimer’s disease. Psychiatry 7, 1–5.