Optimization of formulation processes using Design Expert® Software for preparation of polymeric blends-artesunate-amodiaquine HCl microparticles

Journal of Drug Delivery Science and Technology 39 (2017) 36e49

Contents lists available at ScienceDirect

Journal of Drug Delivery Science and Technology journal homepage: www.elsevier.com/locate/jddst

Optimization of formulation processes using Design Expert® Software for preparation of polymeric blends-artesunate-amodiaquine HCl microparticles John Dike N. Ogbonna a, b, *, Anthony A. Attama a, Kenneth C. Ofokansi a, Sanjay B. Patil b, Ganesh D. Basarkar b a b

Drug Delivery and Nanomedicines Research Group, Department of Pharmaceutics, University of Nigeria, Nsukka, 410001, Enugu State, Nigeria Department of Pharmaceutics, SNJB's ShrimanSureshdada Jain College of Pharmacy, Neminagar, Chandwad, Dist. Nashik, 423101, India

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 May 2016 Received in revised form 15 February 2017 Accepted 23 February 2017 Available online 24 February 2017

A 23 Factorial design of cissus-gelatin B polymer blends was developed to formulate amodiaquine HClartesunate (AQ-AS) microparticles by varying the polymer blend concentration (2 %w/v, 5 %w/v), crosslinking time (0.5 h, 1 h) and glutaraldehye volume (0.5 ml, 1 ml). The formulations were evaluated using drug entrapment efficiency (EE), particle size, polydispersity index, thermal behavior with differential scanning calorimetry, crystallinity with powder X-ray diffraction, morphology with scanning electron microscope and in vitro release using combination of simulated gastric fluid (SGF, pH ¼ 7.4, 95%) and methanol (5%). The expected responses, EE and in vitro release were fitted into Design Expert®. The polymer blends exhibited pseudoplastic behavior and there was no marked change in rheology behavior of 2% w/v dispersion at 55  C. The AQ-AS formulated microparticles were dark brownish discrete mass, physically stabilized, irregular shape, polydisperse, and partially crystalline system. An optimal formulation comprising polymer blend (5 %w/v), glutaraldehyde (1 ml) and cross-linking time (0.5 h) was identified to provide desired values for EE, amodaquine HCl (47.41%), artesunate (36.42%) and in vitro release. This study proposes the best opportunity for selection of factors required for optimum microparticles formulation using the polymer blends and the drugs. © 2017 Elsevier B.V. All rights reserved.

Keywords: Microparticles Crystallinity Entrapment efficiency Polymer blend In vitro release

1. Introduction Drug delivery systems offer numerous advantages such as improved efficacy, reduced toxicity, reduced frequency of doses, and convenience when compared with conventional formulations. Of the different drug delivery systems reported, drug-loaded nanoparticles and microparticles attained importance because of the possibility to achieve passive targeting when their sizes are in particular ranges. The special interest for polymeric micro- and nanoparticles for oral drug delivery especially poorly water soluble drugs is owing to their small size and large surface area which favour their absorption compared to larger carriers as drug particles in the nanometer size range will dissolve more rapidly than a

* Corresponding author. Drug Delivery and Nanomedicines Research Group, Department of Pharmaceutics, University of Nigeria, Nsukka, 410001, Enugu State, Nigeria. E-mail address: [email protected] (J.D.N. Ogbonna). http://dx.doi.org/10.1016/j.jddst.2017.02.011 1773-2247/© 2017 Elsevier B.V. All rights reserved.

conventional formulation [1]. Biodegradable microparticles most often used as controlled drug delivery system (injectable, oral, transdermal, etc) have widely been studied in the area of controlled release and works have been done in combining the microparticles with other polymers for novel drug delivery systems. Artesunate-amodiaquine incompatibility resulting in severe degradation of the drugs, short stability of the combination coupled with reported side effects caused by the high strengths of the drugs combination in Ghana in 2005 worsened the challenges in the AQAS formulations [2]. Therefore it is pertinent to formulate the drugs in microparticles for oral administration at reduced drugs amount but maintaining the same combination ratio using polymer blends of cissus populnea gum and gelatin. Cissus gum (C) is a natural, nonionic polysaccharide derived from the incised sliced root of Cissus polpunea Guill. and Perr. (Vitaceae). It is a natural, biosynthetic, edible substance consisting of sugars-galactose, xylose, glucose, mannose, and D-glucuronic acid, used locally as thickener in foods, attracting attention in many

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

works owing to a lot of its pharmaceutical uses; binder [3e5], medicinal uses [6,7]. Till date, no work has been reported on its use as a polymeric carrier matrix in novel drug delivery system in combination with another natural polymer. A polymer blend or polymer mixture is a member of a class of materials analogous to metal alloys, in which at least two polymers are blended together to create a new material with different physical properties having most common specific interactions such as hydrogen bonding, dipole edipole, and ionic interactions [8]. An important attribute of such hybrid polymers or polymer blends is that the new species either combine the qualities of the components in terms of functional and physicochemical properties or new functional properties different from the primary materials are obtained. This may result in superior functional and physicochemical properties when compared with the primary materials. Modification of naturally occurring polymers by formulations of polymer composites are popular methods by which new polymers and pharmaceutical excipients are produced for purpose of drug delivery as novel polymer biomaterials with effective multifunctional properties are continually being sought for drug delivery purposes [9e11]. In order to address the multifaceted oral delivery challenges, more sophisticated carrier systems are required, either as polymer-polymer or polymer-lipid hybrid systems, developed with the primary aim to combine the valuable features of both polymeric and lipid-based systems [12]. Therefore, there is always the need to develop an approach for determining the relationship between various process parameters and responses with the various desired criteria and searching the significance of these process parameters on the coupled responses [13]; after Box and Wilson [14] described the original basic theoretical and fundamental aspects of response surface methodology (RSM). Optimization using factorial designs is a powerful, efficient and systematic tool that shortens the time required for the development of pharmaceutical dosage forms, improves research and development work by reducing the number of experimental trials needed to evaluate multiple parameters and their interactions thereby making the process less laborious. Factorial designs which entails studying all the factors in all possible combinations, are considered to be the most efficient in estimating the influence of individual variables and their interactions using minimum experiments [15] and has played a key role in understanding the relationship between the independent variables and the responses to them in pharmaceutical formulations development [16]. The independent variables or parameters or factors are controllable, whereas responses are dependent. The contour plot gives a 2-D visual while the responses surface gives a 3-D visual of the representation values of responses and helps the process of optimization by providing an empirical model equation for the response as a function of the different variables [17e19]. Our research group, have been developing different novel microparticulate delivery systems especially lipid based formulations for antimalarial studies [20e23] but the microparticles production described here includes polymer blend systems (cissus and gelatin B at ratio 1:1) in which biodegradable microparticles containing drugs [amodiaquine HCl and artesunate (AQ-AS)] are embedded within the polymer blend matrix and the drug released through the porous polymeric matrices with desirable properties. Rapid opsonization by cells of the phagocytic system is a major limitation for achieving effective drug targeting to the site of action by gelatin B formulations. Thus, to maximize the therapeutic benefits of drug loaded microparticles, they should be able to evade the reticuloendothelial system (RES) through the use of various surface coatings of hydrophilic polymers, as opsonization of hydrophobic formulations may occur more quickly in comparison to hydrophilic formulations due to the enhanced adsorption of opsonins on their

37

surfaces [24]. This was collaborated by Parveen and Sahoo [25] who investigated chitosan and polyethylene glycol (PEG) modification in order to reduce uptake of encapsulated hydrophobic drug (paclitaxel), hence the introduction of another natural hydrophilic polymer, cissus polymeric gum which is also biodegradable. Development of really new anti-malarial drugs to market level is a very rare event as a lot of lead structures have already been screened and discarded, coupled with the administrative barriers that are increasingly high and costly [26]. Of the two major aspects to drug development for antiparasitic; pathogen-specific biochemical intervention strategy, and optimal formulation and application strategy, we focused on the later in our research using polymer blend of cissus and gelatin B after characterization of the polymer blends and the AQ-AS loaded formulations. The objective of this work is to prepare, and evaluate cissus gum-gelatin (polymer blend) generated by compatibilized reactive polymer blending of colloidal dispersions of cissus gum and gelatin B at controlled temperature conditions. The AQ-AS microparticles formulated from the polymer blends were characterized and evaluated in vitro for oral drug delivery.

2. Materials and methods The following materials were used as procured locally without further purification: gelatin B (Acofarma, Barcelona, Spain), Span 80 (Sigma Aldrich, Germany), sodium hydroxide (Merck, Germany), light paraffin oil, dilute hydrochloric acid, sodium dihydrogen phosphate monohydrate, disodium hydrogen phosphate absolute ethanol, sodium chloride, conc. hydrochloric acid, acetone, glutaraldehyde (BDH, England). Cissus populnea was sourced from Uvuru Town in Uzo-Uwani LGA of Enugu State, Nigeria, while artesunate and amodiaquine HCl were gift samples from Emzor Pharma (Lagos, Nigeria). All other reagents were of analytical grade and were used as such.

2.1. Experimental design The Design of experiment (DoE) was constructed in this study using Design Expert® Software (Version 9.0.3.1, Stat-Ease Inc, Minneapolis, MN) by adapting the 23 factorial design approaches to optimize the polymeric blended loaded microparticles. The independent parameters for optimization were; polymer blend concentration in the aqueous phase (A ¼ X1), volume of crosslinking agent, glutaraldehyde (B ¼ X2) and the crosslinking time (C ¼ X3). Preliminary studies also provided a setting of the levels for each formulation variable. In addition, the design is appropriate to study the quadratic response surfaces and to construct the second-order polynomial models. The selected responses were the encapsulation efficiency (Y1) and cumulative % of in vitro drug released after 7 h (Y2) of the study. Each independent variable was given a high and low level value as shown in Table 1 and a total of eight experimental runs of polymer blends AQ-AS microparticles formulations, as depicted in Table 1, were prepared.

Table 1 Factorial design parameters and experimental conditions. Factors

volume of glutaraldehyde (ml) crosslinking time (h) polymer blends concentration (%)

Levels used, actual (coded) Low (1)

High (þ1)

0.5 0.5 2

1 1 5

38

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

2.2. Extraction of gum Fresh roots of Cissus populnea were peeled and sliced into smaller pieces with the aid of knife and thereafter soaked in sufficient quantity of distilled water containing 1% sodium metabisulphite (antioxidant) and macerated for 48 h. The roots were removed from the water and the exudates (aqueous mucilage) were strained after which it was sieved with a triple folded muslin cloth. The gum was precipitated with about 10 L of chilled acetone at 28  C. The gum extract was dried under a current of air for 24 h and later transferred into a desiccator containing silica gel and left to dry for about 1 week. The dried gum extract was milled with the aid of blender, after which it was sieved using sieve mesh size 150 mm (ISO 3310-1 BODY 316L MESH, Germany). The fine cissus powder was stored in air-tight containers. The process of cissus gum extraction was repeated in triplicate. 2.3. Preparation of optimized polymer blends The polymer blends were prepared by combining each of 6 h hydrated 10 %w/v aqueous dispersions of cissus gum powder and gelatin B in the ratio 1:1 according to Attama [27] with little modifications. The admixture was incubated at a temperature of 37  C for 48 h, recovered by precipitating with chilled acetone and air-dried for 24 h after which the polymer blends were dried with a vacuum pump. Thereafter, the dry coarse material were pulverized using a blender and the polymer blends particles were screened through a 100 mm mesh sieve (US Standard sieve, USA) using a sieve shaker (Retsch, D 42781 Haan, Germany) and then stored in a desiccator for 48 h before storing in an airtight container. 2.4. Characterization of polymer blends 2.4.1. Rheological studies The measurement of viscosity of the prepared polymer blends was done with Brookfield RVT viscometer (Brookfield Engineering Lab Inc, MA 02346, Middleboro, USA). A 1, 2 and 3% w/v dispersions of the polymers blends (cissus:gelatin B; 1:1) were prepared in water and rotated for 2 min at different speeds from 0.5 to 100 rpm for spindles 2 to 7. Triplicate readings were determined. A 2% w/v aqueous dispersion of polymer blend was further prepared and the shear rate of the ThermoHaake® viscotester (Gebruder HAAKE GmbH, VT 550, Karlsruhe, Germany) with probe 222e0505, cup 807-0792, software (Rheowin 3 User Manager, Rheowin 3 Job Manager), was run from 1 to 1000/s at a temperature of 20  C. A fresh dispersion of polymer blend was heated at 55  C for 30 min. Triplicate studies were carried out for the unheated polymer blend and the one heated at 50  C for 30 min. Herschel- Bulkley Equation model was used to evaluate the results with Solver software as expressed in the equation.

t ¼ t0 þ Kyn

(1)

where t, t0, K, y and n are shear stress (Pa), yield stress, consistency index parameter, shear rate (s1) and behavior índex respectively. 2.4.2. Particle size analysis of polymer blends The average diameter (Z-av.) of polymeric blend powder was determined using dynamic light scattering (DLS) technique (Nano ZS, Malvern Instruments, Malvern UK) at 25  C. The polymeric blend was dispersed in distilled water and analyzed at an angle detection of 90 . Each value was the average of 10 measurements. 2.4.3. Formulation of hybrid polymeric microparticles A 23 Factorial design of the polymer blends developed by

changing the concentration of the polymer blends (2, 5 %w/v), crosslinking time (0.5 h, 1 h) and volume of the crosslinking agent, glutaraldehye (0.5 ml, 1 ml) after preliminary studies has shown that 2 %w/v of the polymer concentration is the minimum for control release microparticles development. A 2% concentration of the polymer hybrid was dispersed in water and added dropwise using a hypodermic syringe to light liquid paraffin containing 1 ml of Span 80, maintained at 42  C and the content homogenized at a speed of 2000 rpm for 30 min. Then glutaraldehyde (0.5 ml) was added to the mixture and the homogenization continued at the same speed for a further 30 min. The formulation was centrifuged for 10 min at 4000 rpm after which the microparticles were recovered by washing off the oil with analytical acetone and drying under a current of air at ambient temperature for 24 h. This formed the batch UnH2-ve which is unloaded drug hybrid microparticles (UnH2-ve). The same procedure was repeated for batch H2-ve except that 100 mg artesunate and 300 mg amodiaquine HCl were added to the light liquid paraffin and polymer blend dispersion respectively. The same procedure was also repeated for other AQ-AS loaded and unloaded batches (H5a, H2b, H2c, H5ab, H2bc, H5ac and H5abc) using parameters presented in Table 1. Triplicate formulations were made. 2.4.4. Calibration curve of amodiaquine and amodiaquine HCl A 10 ml solution containing 3.3 mg artesunate and 10 mg amodiaquine HCl was prepared using methanol in 10 ml volumetric flask. Serial dilution of the stock to 2.5, 5, 10, 20 and 40 mcg/ml was done after filtering the solution using Whatman No. 1 filter paper. The dilutions were scanned in a UV-Vis spectrophotometer (UV1800 240 V, Shimadzu Corp., Japan) with absorption maxima obtained at 342 nm using methanol as the blank and triplicate studies were made. According to Niesko et al., [28] with little modifications, a 10 ml solution containing 10 mg artesunate and 30 mg amodiaquine HCl was further prepared using methanol. A 2 ml of this stock was diluted to 10 ml using methanol: 0.2% NaOH (1:4) in a 25 ml volumetric flask and vortexed. The solution was heated in water bath at 50  C for 30 min after which the content was cooled and diluted further to 20 ml using 2 ml of methanol and 8 ml of 0.2 M acetic acid. The content of the 20 ml volume in the flask was filtered using Whatman No. 1 filter paper and serial dilutions of 1, 2, 3, 4, 5 and 10 mcg/ml of artesunate was made. The dilutions were scanned in a UV-Vis spectrophotometer (UV-1800 240 V, Shimadzu Corp., Japan) with absorption maxima obtained at 242 nm using mixture of methanol: 0.2% NaOH (1:4) and methanol: 0.2 M acetic acid (1:4) in the ratio of 1:1 respectively as the blank. Triplicate studies were made. The above procedures were repeated for in vitro release calibration except that methanol:PBS (5:95%) is used in place of methanol for the initial solubilization in 10 ml volumetric flask. 2.4.5. Determination of drug content and encapsulation efficiency A 50 mg each of H2-ve, H5a, H2b, H2c, H5ab, H2bc, H5ac and H5abc was crushed and soaked in 10 ml methanol for 24 h. A 0.5 ml of the filtrate in each batch was diluted to 5 ml using methanol. The dilutions were filtered through Whatman No. 1 filter paper and analyzed spectrophotometrically for amodiaquine HCl at a predetermined wavelength of 342 nm using methanol as blank. Also, 2 ml of the filtrate from each batch was diluted to 10 ml using 0.2% NaOH in a 25 ml volumetric flask. The solution was heated in water bath at 50  C for 30 min after which the content was cooled and diluted further to 20 ml using 2 ml of methanol and 8 ml of 0.2 M acetic acid. The content of the 20 ml volume in the flask was filtered using Whatman No. 1 filter paper and analyzed spectrophotometrically for artesunate using a UV-Vis spectrophotometer (UV-1800

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

240 V, Shimadzu Corp., Japan) at a predetermined wavelength of 242 nm using mixture of methanol: 0.2% NaOH (1:4) and methanol: 0.2 M acetic acid (1:4) in the ratio of 1:1 respectively as the blank to determine the drug content. Triplicate readings were taken.

EE ¼

actual drug content x100 theoretical drug content

(2)

2.4.6. In vitro drug release A 47.5 ml phosphate buffer saline (PBS, pH ¼ 6.8 ± 0.2) was poured into a 100 ml beaker and 2 ml methanol was dropped in streams into the PBS. A 100 mg each of batches H2-ve, H5a, H2b, H2c, H5ab, H2bc, H5ac and H5abc was soaked in 0.5 ml methanol. The dispersion is poured into the basket and fixed with basket spindle. Then 100 ml beaker containing dissolution medium is then placed in the 900 ml jar of USP type 2 apparatus and the basket spindle released into it. Each of the microparticles batches H2-ve to H5abc was subjected to the basket dissolution method using the PBS: methanol (95:5%) as the dissolution medium, placed in a dissolution apparatus set to rotate at 50 rpm and the temperature was set at 37 ± 1  C. At intervals of 15, 20, 30 min for the initial 1 h and subsequent 1 h for the remaining 7 h, 5 ml aliquots of the dissolution medium was collected and immediately replaced with 5 mL of fresh medium, phosphate buffer saline: methanol (95:5%). The withdrawn samples were analyzed using a UV-Vis spectrophotometer (UV-1800 240 V, Shimadzu Corp., Japan) at predetermined wavelength of 342 nm for amodiaquine HCl using mixture of PBS: methanol ratio of 95:5 as blank. A 1 ml of withdrawn sample was made up to 10 ml using 1 ml methanol and 8 ml 0.2% sodium hydroxide. This solution was heated in water bath at 50  C for 30 min after which the content was cooled and diluted further to 20 ml using 2 ml of methanol and 8 ml of 0.2 M acetic acid. The content of the 20 ml volume in the flask was filtered using Whatman No. 1 filter paper and analyzed in a UV-Vis spectrophotometer (UV-1800 240 V, Shimadzu Corp., Japan) at predetermined wavelength of 242 nm using mixture of methanol: 0.2% NaOH (1:4) and methanol: 0.2 M acetic acid (1:4) in the ratio of 1:1 respectively as the blank. Triplicate readings were taken. 2.4.7. Optimization of 23 factorial design using Design Expert® software version 9 The results obtained for the dependent variables, drug entrapment efficiency (Y1) and in vitro percentage cumulative drug release after 7 h (Y2) was introduced into the response columns of the experimental design software and the model was run for the eight different formulations in triplicate. Statistical analysis of the data was considered to be significant for any factor at p values < 0.05. The 23 full Factorial design was selected to study the effect of independent variables (polymer blend concentrations, volume of crosslinking agent and crosslinking time) on dependent variables (drug entrapment efficiency and in vitro percentage cumulative drug release after 7 h). The relationship between the response and independent variables can be directly visualized from the response surface (3-D) and Contour (2-D) plots

Y1 ¼ b0 þ b1A þ b2B þ b3C þ b12AB þ b13AC þ b23BC þ b123ABC

(3)

Y2 ¼ b0 þ b1A þ b2B þ b3C þ b12AB þ b13AC þ b23BC þ b123ABC

(4)

where Y1 is measured response (entrapment efficiency) or Y2 is measured response (in vitro percentage cumulative drug release) associated with each factor level combination; b0 is an intercept

39

representing the arithmetic average of all quantitative outcomes of eight runs; b1 to b123 are regression coefficients computed from the observed experimental values of Y1 or Y2 and A, B and C are the coded levels of independent variables. The terms AB, BC and AC represent the interaction terms. The main effects (A, B and C) represent the average result of changing one factor at a time from its low to high value. Table 1 showed the Factorial design parameters and experimental conditions. 2.5. Characterization of the microparticulates systems and polymer hybrid 2.5.1. Crystallography determination using powder X-ray diffraction (PXRD) The X-ray diffraction patterns of amodiaquine HCl, artesunate, polymer blend, the unloaded and drug loaded microparticles were obtained by X-ray diffraction analysis of the powder sample at 25  C using a Panalytical Xpert Pro Diffractometer (PANalytical, JB Eindhoven, The Netherlands). Measurement conditions were as follows: the target metals Cu, Ka filter, voltage 40 kV, current 40 mA, the analysis performed at 2 theta range of 5e80 . The sample was placed on the sample holder and leveled to prevent particle orientation during sample preparation. 2.5.2. Particle size and morphological study Particle size analysis was carried out on the AQ-AS loaded and unloaded microparticles after formulation. A 5 mg of the polymer from each batch was dispersed in distilled water and smeared on a microscope slide using a glass rod. The mixture was covered with a cover slip and viewed using a polarized photomicroscope (Hund Weltzlar, Germany), attached with a Motic image analyzer which is an Automated imaging system at a magnification of X100. Triplicate readings were taken. The morphology of polymer blend, the unloaded and AQ-AS HCl loaded microparticles were assessed using scanning electron microscope (SEM). The samples were prepared by gold-plating, while imaging was carried out on a scanning electron microscope (FEI Quanta 400, FEI Company, Oregun, USA). 2.5.3. Thermal analysis using differential scanning calorimetry Melting transitions and changes in heat capacity of the polymer blend and the optimized unloaded and drug loaded microparticles were determined using a differential scanning calorimeter (Netzsch DSC 204 F1, Germany) equipped with an autosampler. Approximately, 10 mg of each microparticles was weighed into aluminium pan, hermetically sealed and the thermal behavior determined in the range 20e240  C, under a 20 mL/min nitrogen flux at a heating rate of 10  C/min. The baselines were determined using an empty pan, and all the thermograms were baseline corrected. 2.5.4. Data and statistical analysis All experiments were performed in replicates for validity of statistical analysis. Results were expressed as mean ± SD. ANOVA and Student's t-test were performed on the data sets generated using Design Expert® software 9. Differences were considered significant for p-values < 0.05. 3. Results Polymeric particulate formulations have been widely used as they exhibit high structural integrity, stability during storage, controlled release capability, in addition, they are also easy to prepare and readily functionalized for active targeted delivery, making them very attractive as therapeutic delivery vehicles [29,30].

40

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

Murambiwa et al., [31] observed that formulation and evaluation of novel drug delivery systems is not only less expensive than developing new drugs, but may also improve delivery of antimalarials at the desired rates as patients' non-compliance due to inconvenient dosing schedules of conventional formulations (tablets, capsules, etc) which leads to variable bioavailability will be overcome. Therefore, we assessed the feasibility of developing novel formulations and delivery systems of cissus-gelatin loaded AQ-AS microparticles for oral delivery. The use of emulsification crosslinking technology to formulate microparticles of poorly water soluble compounds has been in application however there are no reports in literature on use of polymer blends of gelatin B and cissus polymer as drug delivery carrier. The research in the field of polymer combination with different properties present a tremendous opportunities for developing microparticles to suit various drug delivery needs especially for combination therapies through co-delivery of multiple agents of distinct characteristics. The multiple components in polymeric blend microparticles impart on them capability of loading multiple agents in one carrier and releasing the payloads at different times and intracellular locations making this idea a novel, exciting and interesting application in delivery of synergistic drug combinations [32]. Good knowledge of how biological systems process polymers and surfactants and about mechanisms of drug(s) action(s) could help optimal formulation design. As the temperature of the 1 %w/v and 3 %w/v dispersions of polymer blends increased from 25  C through 45  Ce60  C, the viscosity changed marginally which was not significant. Spindles 2, 3, 4 and 5 for 1 %w/v dispersion and spindles 2, 3, 4, 5, 6 and 7 for 3 %w/v dispersion gave noticeable readings at the varying temperature. Using the ThermoHaake® viscotester, polymer blends exhibited shear thinning behavior at a range of shear rate from 1 1/s to 1000 1/s and obeyed the Hershel-Bulkley non-Newtonian law as shown in Fig. 1 and Table 2. The polymer blends at 2 %w/v is pseudoplastic and the Herschel-Bulkley model showed its application on fluids with a non-linear behavior and yield stress. The to, K, n and apparent viscosity at 1000/s values for the polymer blend dispersion; at 25  C were 0.1 ± 0.0, 0.0 ± 0.0. 1.1 ± 0.2 and 9 ± 1 (Pa) while at 55  C were 0.1 ± 0.0, 0.0 ± 0.0, 1.1 ± 0.1 and 9 ± 1 (Pa) for 1 h respectively. The graph showed that there was no appreciable change in the rheological parameters except the yield stress which decrease at the temperature of 55  C. The PDI of the polymer blend was 0.696 while the Z-Average (d.nm) was 572.3 as shown in Fig. 2 of the particle size distribution analysis.

Fig. 1. Shear stress versus shear rate (1/s) for 2% dispersion of polymer hybrid after preparation and after warming at 55  C.

Entrapment efficiency of AQ in the formulation ranged from 21% to 50%, while EE for AS ranged from 18 ± 3% to 40 ± 5% as shown in Table 3. The 5 %w/v polymer blend formulations (H5a, H5ab, H5abc and H5ac) had higher entrapment efficiency of AQ than the 2 %w/v formulations (H2-ve, H2b, H2bc and H2c) while AS showed variability of EE in 2 %w/v and 5 %w/v formulations. The entrapment efficiency of AQ is higher than AS in almost all batches of the microparticles formulations except H2b. The batches that had appreciable entrapment efficiency of both AQ and AS are H5a [EE of AS (40 ± 5%), EE of AQ (47 ± 5%)] and H5ab [EE of AS (36 ± 4%), EE of AQ (47 ± 7%)]. The dissolution profiles of the drugs are shown in Figs. 3 and 4. The AQ-AS loaded microparticles showed that H2-ve, H5a, H2b, H5ab, H5ac and H5abc released more than 80.00% of amodiaquine HCl after 1 h respectively except H2c (79 ± 6%)and H2bc (57 ± 5%). Also all the 5% polymer blend formulations released up to 80% of amodiaquine HCl [H5a (98 ± 5%), H5ab (88 ± 7%), H5ac (83 ± 4%)] at the 7 h study except batch H5abc which had 73 ± 5% while the 2% polymer blend formulations had less than 77% amdiaquine HCl released at 7 h [H2-ve (74 ± 4%), H2c (77 ± 5%), H2bc (53 ± 8%)] except H2b that had 86 ± 10% of amodiaquine HCl released after 7 h. After 1 h, the batches H5a, H5ab, H5ac and H5abc released 58 ± 4%, 57 ± 3%, 83 ± 5% and 89 ± 7% of AS respectively while H2ve, H2b, H2bc and H2c released 50 ± 3%, 39% ± 6, 35 ± 4% and 58 ± 5% respectively. At the 7th h, H5a, H5ab, H5ac and H5abc released 46 ± 5%, 55 ± 8%, 71 ± 10% and 69 ± 7% of AS respectively while H2-ve, H2b, H2bc and H2c released 38 ± 3%, 37 ± 4%, 51 ± 5% and 33 ± 5% respectively. Except for batch H2c, which maintained more than 50% AS released throughout the experimental duration, all the 2% hybrid polymer formulations had lower level of dissolution. Therefore, higher concentration of polymer blends increases both the percentage and prolongs drug release of the active pharmaceutical ingredients, (AQ-AS) though H5ac and H5abc batches had dose dumping of 81 ± 7% and 82 ± 9% of drug released within 30 min of administration. Therefore, as a drug delivery platform, reduction in drug particles sizes through microparticles formulations have been shown to provide a number of advantages including enhanced solubility, enhanced bioavailability, improved stability and a delivery platform acceptable for the oral route of administration [33]. The standard equations are shown in equations (3) and (4) while the equations in terms of coded factors are shown in equations (5)e(8). The plots Figs. 5 and 6, Tables 4e7 showed that various combinations of independent variables A (polymer blends concentration), B (glutaraldehyde volume) and C (crosslinking time) may satisfy any specific requirement (ie increased or decreased, the drugs entrapment efficiency/percentage cumulative drug release or both drug(s) entrapment efficiency and percentage cumulative drug release) while taking into consideration various factors involved in dosage formulations. The results in Tables 4e7 revealed the estimated factor effects and their associated p values for the investigated responses deduced from ANOVA. The results showed that A factor has a synergistic effects, according to the positive value of the estimated response; on EE of AS with p values of 0.0682 (non-significant), EE of AQ with p values of 0.0004 (significant) and in vitro release of AQ with p values of 0.1040 (non-significant) for the responses. Also A factor has antagonistic effect on the in vitro release of AS with p values of 0.0523 (non-significant) for the response. The factor B showed a synergistic effect on EE (p values of 0.1506) and in vitro release of AS (p values of 0.2829) but an antagonistic effect on EE (p values of 0.0043) and in vitro release of AQ (p values of 0.2524). The factor C showed a synergistic effect on EE of both drugs, AS (p values of 0.0735), AQ (p values 0.0024) and in vitro release of AS (p values of 0.0893) but an antagonistic effect

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

41

Table 2 Rheological parameters of 2% hybrid at 20  C and 55  C. Batch

K

n

to

R2

App. Viscosity (1000/s) (Pa)

Hybrid at 20  C Hybrid at 55  C

0.01 ± 0.0 0.01 ± 0.0

1.1 ± 0.2 1.1 ± 0.1

0.07 ± 0.04 0.13 ± 0.05

0.9957 0.9951

8.7 ± 0.9 8.7 ± 0.7

Fig. 2. Particle size distribution of polymer blend.

Table 3 Physical parameters of the artesunate-amodiaquine-loaded polymeric multiparticulates drug delivery system. Batches

% Yield

H2-ve H5a H2b H2c H5ab H2bc H5ac H5abc UnH5ab

39 86 46 50 79 46 62 85 90

± ± ± ± ± ± ± ± ±

3 5 6 4 7 5 8 5 8

Mean particle size (mm)

Drug content AS (mg)

Drug content AQ (mg)

Encapsulation efficiency AS (%)

83 ± 5 77 ± 7 103 ± 7 117 ± 35 112 ± 4 200 ± 19 111 ± 14 79 ± 17 51 ± 6

8±1 35 ± 3 9±2 9±1 29 ± 5 13 ± 2 13 ± 1 21 ± 2

25 ± 4 121 ± 9 23 ± 3 39 ± 2 112 ± 7 41 ± 3 86 ± 10 128 ± 12

20 40 25 18 36 28 21 25

Fig. 3. Amodiaquine HCl in vitro release of hybrid polymer (C:G) microparticles in media [SGF (95%), methanol (5%) for 7 h.

on in vitro release of AQ (p values of 0.0742). The EE of AQ showed synergistic effect significantly by quadratic terms AB and BC with p

± ± ± ± ± ± ± ±

2 5 2 3 4 3 1 3

Encapsulation efficiency AQ (%) 21 47 21 26 47 30 46 50

± ± ± ± ± ± ± ±

3 5 2 2 7 2 2 5

Fig. 4. Artesunate in vitro release of hybrid polymer (C:G) microparticles in media [SGF (95%), methanol (5%)] for 7 h.

values of 0.0826 and 0.0048 respectively. In addition, the EE of AQ

42

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

Fig. 5. Contour plots for the effects of hybrid polymer concentration (X1), volume of crosslinking agent (X2), and cross-linking time (X3) on EE (Y1) and in vitro release (Y2) of artesunate and amodiaquine HCl in optimized batch (H5ab).

showed antagonistic effect significantly by quadratic terms AC with p values of 0.0032. The interaction terms AB, AC and BC have no significant effects on the investigated dependent variables (EE and in vitro release of AS). The equations that described the variables for the best optimized batch is ANOVA for Response Surface 2FI model for EE of both drugs and in vitro release of artesunate while ANOVA for Response Surface Linear model described in vitro release of amodiaquine HCl. Results of the multiple linear regression analysis for each dependent variables derived by the best fit method are given in Equations (9e12) that reflect the quantitative influence of formulation variables and their interactions on the responses. Table 8 showed a summary of results of regression analysis for responses Y1 and Y2. Equation in Terms of coded Factors:

EE of AQ ; 2FI ðinteractionÞ model : Y1 ¼ b1A þ b2B þ b3C þ b12AB þ b13AC þ b23BC

(6)

AS in vitro; 2FI interaction model : Y2 ¼ b1A þ b2B þ b3C þ b12AB þ b13AC þ b23BC

(7)

AQ in vitro; Linear model : Y2 ¼ b1A þ b2B þ b3C

(8)

Final Equation in Terms of Actual Factors:

EE of AS ¼ 1:10 þ 13:97A þ 2:62B þ 1:34C  4:49AB  10:51AC þ 27:16BC

(9)

EE of AQ ¼ 1:30 þ 10:58A  7:98B þ 9:90C þ 0:15AB EE of AS; 2FI ðinteractionÞ model : Y1 ¼ b1A þ b2B þ b3C þ b12AB þ b13AC þ b23BC

 3:95AC þ 15:84BC (5)

(10)

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

43

Fig. 6. Response surface plots for effects of hybrid polymer concentration (X1), volume of crosslinking agent (X2), and cross-linking time (X3) on EE (Y1) and in vitro release (Y2) of artesunate and amodiaquine HCl in optimized batch (H5ab).

AS in vitro release ¼ þ30:02  7:08A þ 6:16B þ 31:45C þ 8:46AB þ 10:11AC  57:16BC

(11)

AQ in vitro release ¼ þ100:95 þ 4:70A  17:95B  32:27C (12) where EE is entrapment efficiency, A is polymer blend concentration, B is volume of glutaraldehyde, C is crosslinking time. PXRD is one of the most sensitive techniques to acquire

information on the molecular arrangement within the crystal. The X-ray diffraction patterns of AQ, AS, polymer blend, H2-ve, H2b, H2bc, H2c, H5a, H5ab, H5abc, H5ac and UnH5ab showed different patterns as presented in Figs. 7e9. The reflection angle, 2q, along with the interplanar spacing d, and the relative intensity of the peaks differed. The PXRD patterns of amodiaquine HCl has very strong, sharp and intense crystalline diffraction peaks at 2q ¼ 19.68 , d ¼ 4.51 Å; 2q ¼ 25.53 , d ¼ 3.49 Å, Artesunate has very strong, sharp and intense crystalline diffraction peaks at 2q ¼ 10.16 , d ¼ 8.70 Å; 2q ¼ 17.50 , d ¼ 5.06 Å; 2q ¼ 20.67, d ¼ 4.29 Å; 2q ¼ 23.03 , d ¼ 3.86 Å; 2q ¼ 41.99 , d ¼ 2.15 Å; and the

44

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

Table 4 Analysis of variance table to predict EE of artesunate. Source

Sum of squares

df

Mean square

F value

p-value Prob > F

Model Polymer concentration (A) Glutaraldehyde (B) Crosslinking time (C) A (B) A (C) B (C) residual Cor Total

442.84 132.52 26.35 113.85 22.71 124.35 23.05 1.53 444.37

6 1 1 1 1 1 1 1 7

73.81 132.52 26.35 113.85 22.71 124.35 23.05 1.53

48.20 86.54 17.21 74.35 14.83 81.21 15.05

0.1098 0.0682 0.1506 0.0735 0.1617 0.0704 0.1606

not significant

*A is Polymer blend concentration, B is Volume of glutaraldehyde, C is crosslinking time.

Table 5 Analysis of variance table to predict EE of amodiaquineHCl Source

Sum of squares

df

Mean square

F value

p-value Prob > F

Model Polymer concentration (A) Glutaraldehyde (B) Crosslinking time (C) A (B) A (C) B (C) residual Cor Total

1144.42 1077.41 9.86 31.76 0.026 17.52 7.84 4.500E-004 1144.42

6 1 1 1 1 1 1 1 7

190.74 1077.41 9.86 31.76 0.026 17.52 7.84 4.500E-004

4.239Eþ005 2.394Eþ006 21904.00 70578.78 58.78 38940.44 17424.00

0.0012 0.0004 0.0043 0.0024 0.0826 0.0032 0.0048

significant

*A is Polymer concentration, B is Volume of glutaraldehyde, C is crosslinking time.

Table 6 Analysis of variance table to predict in vitroartesunate release. Source

Sum of squares

df

Mean square

F value

p-value Prob > F

Model Polymer concentration (A) Glutaraldehyde (B) Crosslinking time (C) A (B) A (C) B (C) residual Cor Total

1452.91 843.37 25.21 286.80 80.52 114.91 102.10 5.71 1458.62

6 1 1 1 1 1 1 1 7

242.15 843.37 25.21 286.80 80.52 114.91 102.10 5.71

42.39 147.64 4.41 50.21 14.10 20.12 17.87

0.1170 0.0523 0.2829 0.0893 0.1657 0.1397 0.1479

not significant

*A is Polymer concentration, B is Volume of glutaraldehyde, C is crosslinking time.

Table 7 Analysis of variance table to predict in vitroamodiaquineHCl release. Source

Sum of squares

df

Mean square

F value

p-value Prob > F

Model Polymer concentration (A) Glutaraldehyde (B) Crosslinking time (C) residual Cor Total

1078.55 396.77 161.10 520.68 360.90 1439.45

3 1 1 1 4 7

359.52 396.77 161.10 520.68 90.23

3.98 4.40 1.79 5.77

0.1075 0.1040 0.2524 0.0742

not significant

*A is Polymer concentration, B is Volume of glutaraldehyde, C is crosslinking time.

Table 8 Summary of results of regression analysis for responses Y1 (Encapsulation efficiency and Y2 (in vitro release). Models

R2

Adjusted R2

Predicted R2

SD

% CV

AS response (Y1) Interactive model AS response (Y2) Interactive model AQ response (Y1) Interactive model AQ Response (Y2) Linear model

0.9966

0.9759

0.7795

1.24

4.65

0.9961

0.9726

0.9694

2.39

4.78

1.0000

1.0000

1.0000

0.021

0.059

0.7493

0.5612

0.0029

9.50

11.92

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

Fig. 7. Powdered X-ray diffraction pattern of amodiaquine HCl, artesunate, polymer hybrid and drug unloaded microparticles.

Fig. 8. Powdered X-ray diffraction pattern of 2% hybrid polymer drug loaded microparticles.

Fig. 9. Powdered X-ray diffraction pattern of 5% hybrid polymer drug loaded microparticles.

polymer blend has very strong, sharp and intense crystalline diffraction peaks at 2q ¼ 26.49 , d ¼ 3.36 Å, strong peak at 2q ¼ 20.75 , d ¼ 4.28 Å. The drug loaded 2 %w/v polymer blend formulations: H2b has very prominent peaks at 2q ¼ 20.75 , d ¼ 4.28 Å; and 2q ¼ 26.77, d ¼ 3.33 Å; corresponding to the two principal peaks of AQ though with reduced intensity. H2bc has a very prominent peak at 2q ¼ 26.67, d ¼ 3.34 Å corresponding to one of the principal peaks of AQ or principal peak of hybrid polymer with a reduction in intensity. H2c has a very prominent peak at 2q ¼ 26.53 , d ¼ 3.36 Å; (corresponding to one of the principal peaks of AQ or principal peak of polymer blend with a reduction in intensity), 36.13 , d ¼ 2.48 Å while H2-ve has very prominent peak at 2q ¼ 26.67, d ¼ 3.34 Å;

45

corresponding to one of the principal peaks of AQ or principal peak of polymer blend with a reduction in intensity. The drug loaded 5 %w/v polymer blend formulations: H5a has very strong, sharp and intense crystalline diffraction peaks at 2q ¼ 19.24 , d ¼ 4.62 Å; 2q ¼ 26.55 , d ¼ 3.35 Å; corresponding to the principal peaks of AQ or principal peak of polymer blend with a reduction in intensity (26.55 ). H5ab has very prominent peaks at 2q ¼ 26.56 , d ¼ 3.35 Å, corresponding to one of the principal peaks of AQ or principal peak of polymer blend with a reduction in intensity. H5abc has very prominent peaks at 2q ¼ 19.12 , d ¼ 4.64 Å; 2q ¼ 26.11, d ¼ 3.41 Å corresponding to the principal peaks of AQ or principal peak of polymer blend with a reduction in intensity (26.55 ), 49.88 , d ¼ 1.83 Å; and the H5ac has very prominent peaks at 2q ¼ 26.64 , d ¼ 3.34 Å corresponding to one of the principal peaks of AQ or principal peak of polymer blend with a reduction in intensity. The drug unloaded formulation, UnH5ab has very prominent peaks at 2q ¼ 26.23 , d ¼ 3.40 Å. The SEM micrographs of the polymer blend, AQ, AS, drug loaded formulations and drug unloaded optimized formulation are shown in Fig. 10. The SEM micrographs of AQ revealed non-uniformed, flake-like and long rod-like crystal structure with a little degree of asperity; AS has discrete, non-uniformed, and short rod-like crystal structure with high degree of non sphericity and polymer blend has a little flake-like dense mass of highly microporous particles. The SEM images for the drug loaded formulations batches: H2-ve has a highly irregular compact dense mass and many hollows with minute flake-like particles embedded within the rough surface; H2b has irregular dense mass and many hollows with many discrete and flake-like particles at the surfaces; H2bc has irregular compact dense mass, many hollows and undulating surfaces while H2c has a highly irregular compact dense mass and many hollows with minute particles embedded within the rough surface. H5a has irregular shaped convoluted, densified structure with depressing hollows on the surface; H5ab has irregular dense mass showing microfine particles uniformly dispersed within the polymeric matrix with fine edges; H5abc has hollow irregular dense mass showing microfine particles uniformly dispersed within the polymeric matrix with convoluted fine edges and H5ac has hollow irregular dense mass showing microfine particles uniformly dispersed within the polymeric matrix with rough edges. The drug unloaded formulations (UnH5ab) micrograph has a fine flower-like mass with little hollows and undulating surface. The crystal transformation of the optimized drug unloaded and drug loaded formulations, polymer blend, AQ and AS were studied using DSC as shown in Fig. 11. The polymer blend has 159.4  C as onset of melting, and enthalpy of 177.3 J/g. Amodiaquine HCl has 163.3  C as onset of melting and enthalpy of 173.3 J/g, while artesunate has 82.9  C as onset of melting, crystallization temperature of 169.3  C and enthalpy of 31.47 J/g and 423.6 J/g respectively. The optimized batch (H5ab) has 163.2  C as onset of melting, and enthalpy of 96.16 J/g while the drug unloaded batch (UnH5ab) has 154.8  C as onset of melting and enthalpy of 120.3 J/ g. 4. Discussion The ultimate goal of research in drug delivery is to develop effective and safe therapy for clinical use with increased efficacy and reduced toxic side effects. All reported clinical product formulations made from polymers, have been used in approved pharmaceutical products. This mentality in the field is healthy and crucial for translating preclinical results to nanomedicines that can benefit millions of people suffering from milieu of diseases. Use of pharmaceutical and GRAS excipients would accelerate product development by removing one hurdle to regulatory approval of the

46

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

Fig. 10. Scanning electron microscopy of amodiaquine, artesunate, optimized hybrid microparticles, cissus, gelatin B and hybrid polymer.

Fig. 11. Thermogram of amodiaquine, artesunate, loaded and unloaded microparticles.

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

products. Successful design of formulations requires comprehensive consideration of basic properties of particulate drug carriers, which are; drug loading capacity, loading efficiency, particle size, and drug release and these properties can be optimized by rational selection of polymer, surfactant, and their relative ratios by theoretical calculations, physicochemical characterization, and optimization tools [34]. Therefore, understanding of material properties in relation to physicochemical properties of formulations is beneficial for successful design of product as physicochemical compatibility and affinity are essential for efficiently loading the drug(s) into formulations and precisely controlling drug release kinetics. The polymer blend showed a non-significant marginal change in apparent viscosity when the 2% w/v dispersion was heated at 55  C at a shear rate of 1000/s. According to Basheer et al., [35], there is not distinct change in viscosity of metallocene polymer as the temperature changed from 25  C to higher temperature. Our finding is in line with such shear thinning behavior of methallocene polymer as the viscosity does not show any distinct change at 55  C. This implied that heating does not change the rheological behavior of the polymer blend as consistency index parameter and behavior index showed negligible change in values. The significant improvement in solubility and dissolution rate of the APIs is probably due to the semi-crystalline character and reduced particle size at colloidal level and hence enhanced surface area, thereby increasing wettability by polymer blends. According to Kulthe and Chaudhari, [36] the polymer that exhibits such properties are mostly non-ionic which implies that the polymer blend (cissus: gelatin B; 1:1) developed might be a non-ionic polymer combination. Polymer concentration is also a key factor influencing the characteristics and release profiles of microparticles. The particle size of the microparticles fabricated using 2 %w/v and 5 %w/v of polymer dispersion (volume ratio of internal water to oil: 1: 5) ranged from 79 ± 6 mm to 200 ± 19 mm. Our findings on the particles sizes of the different percentage of polymer dispersions used in the formulations is in contrast with the work by Yang et al., [37], who stated that increase in particle size of microparticles at higher concentration arises from more viscous (concentrated) polymer dispersion leading to different surface porosity. The higher entrapment efficiency and initial burst release of the polymer blend batches formulated with 5 %w/v of the polymer especially in H5ac and H5abc may be due to factors like large porosity/hollows in the formulations and more hydrophilicity effect of the polymer and glutaraldehyde which were consistent with the issues raised by Cheow and Hadinoto; Plapied et al., Rao and Prestidge, [38e40] regarding inadequate encapsulation of water-soluble compounds and also burst drug release of the AQ being more than the burst release of AS in all the formulations. Also, as the percentage of polymer blend increases in the aqueous mixture, the porosity of the formed microparticles increases; hence the level pf dissolution of the drug is improved which is consistent with the work by Ahmed et al., [41] where they found out that as percentage of poly (ecaprolactone) increased in the poly (e-caprolactone)/Poly (D, Llactide co-glycolide) mixture, the porosity of the formed microspheres increased thereby improving the release rate of the drug but in contrast to their submission that increase in poly (e-caprolactone concentration) reduced both drug release rates and initial burst amounts. For optimum entrapment efficiency of AS in the polymer blend loaded microparticles; the response parameters (Equation (9)) requires, an increase in (a) the polymer blend concentration, (b) volume of glutaraldehyde, (c) crosslinking time, (d) combination of volume of glutaraldehyde and crosslinking time, or a decrease in (a) combination of polymer blend concentration and volume of glutaraldehyde, (b) polymer blend concentration and crosslinking

47

time. For optimum entrapment of AQ in the polymer blend loaded microparticles as shown in (Equation (10)) requires; increase in (a) the polymer blend concentration, (b) crosslinking time, (c) combination of polymer concentration and volume of glutaraldehyde, (d) combination of volume of glutaraldehyde and crosslinking time, or a reduction in (a) volume of glutaraldehyde, (b) combination of polymer concentration and crosslinking time. The response parameters showed that for optimum in vitro release of AS shown in (Equation (11)), increase in (a) volume of glutaraldehyde, (b) crosslinking time (c) combination of polymer blend concentration and volume of glutaraldehyde (d) combination of polymer blend concentration and crosslinking time or decrease in (a) the polymer blend concentration (b) combination of polymer blend concentration and crosslinking time while the response parameters showed that for optimum in vitro release of AQ (Equation (12)) the factors requires; increase in the polymer blend concentration, decrease in volume of glutaraldehyde or a decrease in crosslinking time. The results verified the validity of the models generated from the design of experiment study for all the responses. Volume of glutaraldehyde followed by polymer blend concentration has the greatest effect on the model prediction for synergistic response. The AQ and AS produced prominent diffraction peaks indicating high crystalline nature of the pure drug samples. The polymer blend exhibited weak peaks therefore it has least crystallinity compared to the two APIs and consequently changed the crystalline status of the APIs (AQ and AS) with most characteristics peaks reduced in intensity in all the formulations. The PXRD pattern demonstrate that the pure drugs exhibited crystalline characteristics and these peaks showed dramatic decline in intensity in all the microparticles formulations thereby revealing the semi-crystalline or less crystalline nature of the APIs in the formulations owing to presence of the polymer blend. This disappearance of some prominent peaks in both AQ and AS occurred at both 2% and 5% w/v without much difference. This none effect of increasing level of the polymer blend is in contrast with other polymers like PVP where it was observed that the degree of artemether crystallinity decreases with increasing PVP levels [42]. The drug unloaded formulation (UnH5ab) is the least crystalline showing that the polymer interaction with the APIs must have been the cause of reduction in crystallinity. The solid state characterization studies showed that following emulsification crosslinking technique, the crystal form of AQ-AS signified a loss of partial crystalline of the AQ and AS in the formulations owing to consequent reduction in peak intensities as also seen for refecoxib [43]. From the SEM micrograph it was evident that emulsification crosslinking technique of AQ-AS formulations, resulted in a significant particle size reduction of active pharmaceutical ingredients (APIs). The SEM images showed that the microparticles fabricated from the lower polymer concentration have less porous matrix compared with the microparticles fabricated from the higher concentration (Fig. 10). SEM confirmed the non-spherical nature and almost non-smooth surfaces of the microparticles. The study indicated a molecular level dispersion of AQ-AS in the polymer blend microparticles and the shapes were without sharp edges as obtainable in the pure APIs. Surface morphology of the formulations showing fine irregularity as crest, hollows and fold were also supposed to enhance apparent surface area of the formulations thereby improving solubility of the poor water soluble compound, artesunate as postulated by Kulthe and Chaudhari, [36]. The presence of short rod-like crystals indicated crystalline nature of AQ and AS while the polymer blend is the least crystalline of the starting materials. According to Gu et al., [44], the presence of glass transition indicates non homogenous mixture of active pharmaceutical

48

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49

ingredients and the excipients and this was in consistent with our findings when the solid state characteristics was analyzed with DSC for the optimized formulations. This finding collaborated the results of the SEM and PXRD studies where it was observed that owing to molecular level dispersion of the AQ and AS, the properties of the drugs were changed in the polymer blend formulations. Also, comparing the curves of pure drug samples, unloaded and drug loaded optimized formulations samples, no significant differences were found and no new peaks corresponding to glass transition or recrystallization appeared, implying that no amorphous forms were produced but instead semi-crystalline which were in consistent with research work of Peng et al., [45] who asserted that new peaks of glass transition or recrystallization confirms the presence of amorphous forms of a drug. Generally dissolution profile of these preparations agreed with the data of SEM, PXRD and DSC which indicate that there is improvement on the physicochemical properties of artesunate and amodiaquine HCl substantially.

[7]

[8]

[9]

[10]

[11]

[12]

[13]

5. Conclusion A polymer blend of cissus and gelatin was successfully developed. This paper focuses on the development of an effective methodology using Design Expert Software Version 9 to determine the optimum levels of three independent variables, polymer blend concentration (5%), glutaraldehyde volume (1 ml) and crosslinking time (0.5 h) to obtain optimal formulations with maximum EE of amodiaquine HCl (47.41%), artesunate (36.42%) and the drugs in vitro release. The study further illustrates the importance of formulation conditions in influencing product properties and the adaptability of the method. Hence, the results of the present study clearly indicated promising potentials of polymer blend microparticles for oral delivering of amodiaquine HCl and artesunate. Combining the results of DSC and XRPD, it was demonstrated that the emulsification crosslinking does not interfere significantly with the crystalline states of the drug compounds. Information on the stability of the two drugs in a combination micronized-dosage form is worth researching further. Conflict of interest The authors declare no conflict of interest.

[14] [15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

Acknowledgements The authors are grateful to The World Academy of Sciences (TWAS), Italy (Grant Number: 3240281996) and TETFUND, Abuja Nigeria for the funds to execute the research.

[24]

[25]

References [1] E. Merisko-Liversidge, G.G. Liversidge, Nanosizing for oral and parenteral drug delivery: a perspective on formulating poorly-water soluble compounds using wet media milling technology, Adv. Drug Deliv. Rev. 63 (2011) 427e440. [2] S. Danso e Mensah, Formulation and Monitoring the Stability of Artesunate in Artesunate - Amodiaquine Combination Tablets- an M.Sc. Thesis Submitted to the Department of Pharmaceutical Chemistry, Kwame Nkrumah University of Science and Technology, 2009. [3] A. Chukwu, P. Okpalaezinne, Preliminary evaluation of cissus root gum as a binder in sodium salicylate tablet formulations, Drug Dev. Ind. Pharm. 15 (2) (1989) 325e330. [4] O.A. Ademola, A.O. Michael, T.J. Kolawole, Evaluation of cissus gum as binder in a paracetamol tablet formulation, Farmacia 59 (1) (2011) 85e96. [5] A.A. Olutayo, N.F.-O. Mbang, A.O. Michael, The effect of processing variables on the mechanical and release properties of tramadol matrix tablets incorporating cissus populnea gum as controlled release excipient, Polim. Med. 44 (4) (2014) 209e220. [6] A.B. Ojekale, O.A. Lawal, A.K. Lasisi, T.I. Adeleke, Phytochemistry and spermatogenic potentials of extract of Cissuspopulnea(Guill and Per) stem bark,

[26]

[27]

[28]

[29]

[30]

[31]

[32]

TSW Holist. Health Med. 1 (2006) 176e182, http://dx.doi.org/10.1100/ tswhhm.2006.148. S.F. Akomolafe, G. Oboh, A. Afolabi, A.A. Akindahunsi, A.J. Akinyemi, O.G. Tade, Inhibitory Effect of Aqueous Extract of Stem Bark of Cissus Populnea on Ferrous Sulphate- and Sodium Nitroprusside-induced Oxidative Stress in Rat's Testes in Vitro ISRN Pharmacology, 2013. Article ID 130989, http://dx.doi.org/ 10.1155/2013/130989. S.I. Ofoefule, A.A. Attama, Biopolymer blends: applications, limitations, and future prospects, in: M.U. Adikwu, C.O. Esimone (Eds.), Biopolymers in Drug Delivery: Recent Advances and Challenges, Bentham Science Publishers Ltd., Nigeria, 2009, pp. 198e223. X. Wu, S. Huang, J. Zhang, R. Zhuo, Preparation and characterization of novel physically crosslinked hydrogels composed of poly (ring alcohol) and amine terminated polyamidoamine dendrimer macromol, Biol. Sci. 4 (2004) 71e75. P.F. Builders, N. Ibekwe, L.C. Okpako, A.A. Attama, O.O. Kunle, Preparation and characterization of mucinated cellulose microparticles for therapeutic and drug delivery purposes, Eur. J. Pharm. Biopharm. 72 (2009) 34e41. P.F. Builders, M.B. Agbo, T. Adelakun, C.L. Okpako, A.A. Attama, Novel multifunctional pharmaceutical excipients derived from microcrystalline cellulosestarch microparticulate composites prepared by compatibilized reactive polymer blending, Int. J. Pharm. 388 (2010) 159e167. K. Raemdonck, K. Braeckmans, J. Demeester, S.C. De Smedt, Merging the best of both worlds: hybrid lipid-enveloped matrix nanocomposites in drug delivery, Chem. Soc. Rev. 43 (2014) 444e472. R.H. Myers, D.H. Montgomery, Response Surface Methodology, John Wiley and Sons, USA, 1995. G.E.P. Box, K.B. Wilson, On the experimental attainment of optimum conditions, J. R. Stat. Soc. 13 (1951) 1e45. B. Singh, N. Ahuja, Development of controlled-release buccoadhesive hydrophilic matrices of diltiazem hydrochloride: optimization of bioadhesion, dissolution, and diffusion parameters, Drug Dev. Ind. Pharm. 28 (2002) 431e442. J. Vandervoort, A. Ludwig, Preparation factors affecting the properties of polylactide nanoparticles: a factorial design study, Pharmazie 56 (2001) 484e488. M. Kincl, S. Turk, F. Vrecer, Application of experimental design methodology in development and optimization of drug release method, Int. J. Pharm. 291 (2005) 39e49. A.K. Mehta, K.S. Yadav, K.K. Sawant, Nimodipine loaded PLGA nanoparticles: formulation optimization using factorial design, characterization and in vitro evaluation, Curr. Drug Del. 4 (2007) 185e193. S.B. Patil, K.K. Sawant, Development, optimization and in vitro evaluation of alginate mucoadhesive microspheres of carvedilol for nasal delivery, J. Microencap 26 (5) (2009) 432e443. E.E. Chinaeke, S.A. Chime, J.D.N. Ogbonna, A.A. Attama, C.C. Müller-Goymann, V.C. Okore, Evaluation of dika wax-soybean oil-based artesunate-loaded lipospheres: in vitroein vivo correlation studies, J. Microencapsul. Early Online (2014) 1e9, http://dx.doi.org/10.3109/02652048.2014.940008. J.D.N. Ogbonna, I.T. Nzekwe, F.C. Kenechukwu, C.S. Nwobi, J.I. Amah, A.A. Attama, Development and evaluation of chloroquine phosphate microparticles using solid lipid as a delivery carrier, J. Drug Discov. Dev. Deliv. 2 (1) (2015a) id1011. J.D.N. Ogbonna, F.C. Kenechukwu, C.S. Nwobi, O.S. Chibueze, A.A. Attama, Formulation, in vitro and in vivo evaluation of halofantrine-loaded solid lipid microparticles, Pharm. Dev. Technol. 20 (8) (2015b) 941e948. C. Agbo, C. Umeyor, F. Kenechukwu, J. Ogbonna, S. Chime, L. Charles, O. Agubata, K. Ofokansi, A. Attama, Formulation design, in vitro characterizations and anti-malarialinvestigations of artemether and LumefantrineEntrapped solid LipidMicroparticles, Drug Dev. Ind. Pharm. (2016), http:// dx.doi.org/10.3109/03639045.2016.1171331. R. Gref, A. Domb, P. Quellec, T. Blunk, R.H. Muller, J.M. Verbavatz, R. Langer, The controlled intravenous delivery of drugs using PEG-coated sterically stabilized nanospheres, Adv. Drug Deliv. Rev. 16 (1995) 215e233. S. Parveen, S.K. Sahoo, Long circulating chitosan/PEG blended PLGA nanoparticle for tumor drug delivery, Eur. J.Pharmaco 670 (2011) 372e383. O. Kayser, C. Olbrich, S.L. Croft, A.F. Kiderlen, Formulation and biopharmaceutical issues in the development of drug delivery systems for antiparasitic drugs, Parasit. Res. 90 (2) (2003) S63eS70. A.A. Attama, Polyelectrolyte complexes of eudragit L30 D-55 and gelatin: antinociceptive activity of entrapped piroxicam, Drug Deliv. 14 (3) (2007) 155e162. P. Niesko, F.V. Jan, B. Sieb, J.W. Herman, M.M. Theo, Laboratory selection of artemisia annuaL. for High artemisinin yielding types, Phytochem. Analy. 2 (1991) 80e83. D. Peer, J.M. Karp, S. Hong, O.C. Farokhzad, R. Margalit, R. Langer, Nanocarriersas an emerging platform for cancer therapy, Nat. Nanotechnol. 2 (2007) 751e760. K. Hadinoto, A. Sundaresan, W.S. Cheow, Lipid.polymer hybrid nanoparticles as a new generation therapeutic delivery platform: a review, Euro. J. Pharm. Biopharm. 85 (2013) 427e443. P. Murambiwa, B. Masola, T. Govender, S. Mukaratirwa, C.T. Musabayane, Anti-malarial drug formulations and novel delivery systems: a review, Acta Trop. 118 (2) (2011) 71e79. X.Y. Wu, Strategies for optimizing polymer-lipid hybrid nanoparticlemediated drug delivery, Exp.Opin. Drug Deliv. (2016), http://dx.doi.org/

J.D.N. Ogbonna et al. / Journal of Drug Delivery Science and Technology 39 (2017) 36e49 10.1517/17425247.2016.1165662. [33] G. Sachin, K.J. Gaurav, T. Renu, K.P. Swaroop, A. Mohammed, H.W. Musarrat, S. Mayank, K.K. Roop, J.A. Farhan, Enhanced Aantimalarial activity lumefantrinenanopowder prepared by wet-milling DYNO MILL technique, Coll.Surf. B Biointerf. 95 (2012) 16e22. [34] Y. Li, M.R. Abbaspour, P.V. Grootendorst, A.M. Rauth, X.Y. Wu, Optimization of controlled release nanoparticle formulation of verapamil hydrochloride using artificial neural networks with genetic algorithmand response surface methodology, Eur. J. Pharm. Biopharm. 94 (2015) 170e179. [35] Al-S. Basheer, Al-F. Tariq, Al-S. Fares, E. Rabeh, The effect of polymer concentration and temperature on the rheological behavior of metallocene linear low density polyethylene (mLLDPE) solutions, J. King Saud Univ. Eng. Sci. 23 (2011) 9e14. [36] V.V. Kulthe, P.D. Chaudhari, Effectiveness of spray congealing to obtain physically stabilized amorphous dispersions of a poorly soluble thermosensitive API, AAPS PharmSciTech 15 (6) (2014) 1370e1377. [37] Yi-Y. Yang, Tai-S. Chung, N.P. Ng, Morphology, drug distribution, and in vitro release profiles of biodegradable polymeric microspheres containing protein fabricated by double-emulsion solvent extraction/evaporation method, Biomaterials 22 (2001) 231e241. [38] W.S. Cheow, K. Hadinoto, Enhancing encapsulation efficiency of highly watersoluble antibiotic in poly(lactic-co-glycolic acid) nanoparticles: modifications of standard nanoparticle preparation methods, Colloid Surf. A 370 (2010)

49

79e86. at, Fate of polymeric nanocarriers for [39] L. Plapied, N. Duhem, A. des Rieux, V. Pre oral drug delivery, Curr.Opin. Colloid Interface Sci. 16 (2011) 228e237. [40] S. Rao, C.A. Prestidge, Polymer-lipid hybrid systems: merging the benefits of polymeric and lipid-based nanocarriers to improve oral drug delivery, Exp. Opin. Drug Deliv. (2016), http://dx.doi.org/10.1517/17425247.2016.1151872. [41] S.A. Ahmed, A.K. Hussein, F.M. Mady, Optimisation of microstructured biodegradable finasteride formulation for depot parenteral application, J. Microencap. (2016), http://dx.doi.org/10.3109/02652048.2016.1144821. [42] T.A. Muhammad, K. Shahid, M.R. Nazar, H.S. Nisar, S. Muhammad, Physicochemical characterization of artemether solid dispersions with hydrophilic carriers by freeze dried and melt methods, Arch. Pharm. Res. 33 (6) (2010) 901e910. [43] N. Ahuja, O.P. Katare, B. Singh, Studies on dissolution enhancement and mathematical modeling of drug release of a poorly water-soluble drug using water-soluble carriers, Eur. J. Pharm. Biopharm. 65 (2007) 26e38. [44] B. Gu, B. Linehan, Y.-C. Tseng, Optimization of the Büchi B-90 spray drying process using central composite design for preparation of solid dispersions, Int. J. Pharm. 491 (2015) 208e217. [45] L. Peng, R. Xinyu, L. Johanna, van V. Bert, K. Juha, H. Jouni, L. Timo, P. Leena, Nanosuspensions of poorly soluble drugs. Preparation and development by wetMilling, Int. J. Pharm. 411 (2011) 215e222.