Natamycin Niosomes As a Promising Ocular Nanosized Delivery System

Drug Development and Industrial Pharmacy

ISSN: 0363-9045 (Print) 1520-5762 (Online) Journal homepage: https://www.tandfonline.com/loi/iddi20

Natamycin niosomes as a promising ocular nanosized delivery system with ketorolac tromethamine for dual effects for treatment of candida rabbit keratitis; in vitro/in vivo and histopathological studies

Mohamed Ahmed El-Nabarawi, Randa Tag Abd El Rehem, Mahmoud Teaima, Mohammed Abary, Hala Mohammed El-Mofty, Mohamed M. Khafagy, Nancy M. Lotfy & Mona Salah

To cite this article: Mohamed Ahmed El-Nabarawi, Randa Tag Abd El Rehem, Mahmoud Teaima, Mohammed Abary, Hala Mohammed El-Mofty, Mohamed M. Khafagy, Nancy M. Lotfy & Mona Salah (2019) Natamycin niosomes as a promising ocular nanosized delivery system with ketorolac tromethamine for dual effects for treatment of candida rabbit keratitis; invitro/invivo and histopathological studies, Drug Development and Industrial Pharmacy, 45:6, 922-936, DOI: 10.1080/03639045.2019.1579827 To link to this article: https://doi.org/10.1080/03639045.2019.1579827

Accepted author version posted online: 12 Submit your article to this journal Feb 2019. Published online: 27 Feb 2019.

Article views: 68 View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=iddi20 DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 2019, VOL. 45, NO. 6, 922–936 https://doi.org/10.1080/03639045.2019.1579827

RESEARCH ARTICLE Natamycin niosomes as a promising ocular nanosized delivery system with ketorolac tromethamine for dual effects for treatment of candida rabbit keratitis; in vitro/in vivo and histopathological studies

Mohamed Ahmed El-Nabarawia, Randa Tag Abd El Rehema, Mahmoud Teaimaa, Mohammed Abarya, Hala Mohammed El-Moftyb, Mohamed M. Khafagyb, Nancy M. Lotfyb and Mona Salahc aDepartment of Pharmaceutics and Industrial Pharmacy, Faculty of Pharmacy, Cairo University, Cairo, Egypt; bDepartment of Ophthalmology, Faculty of Medicine, Cairo University, Cairo, Egypt; cDepartment of surgical pathology, Faculty of Medicine, Cairo University, Cairo, Egypt

ABSTRACT ARTICLE HISTORY Objectives: This study was aimed to develop dual-purpose natamycin (NAT)-loaded niosomes in ketorolac Received 13 October 2018 tromethamine (KT) gels topical ocular drug delivery system to improve the clinical efficacy of natamycin Revised 31 December 2018 through enhancing its penetration through corneal tissue and reducing inflammation associated with Accepted 2 February 2019 Fungal keratitis (FK). KEYWORDS Significance: Nanosized carrier systems, as niosomes would provide great potential for improving NAT Natamycin; niosomes; ocular bioavailability.NAT niosomal dispersion formulae were prepared and then incorporated in 0.5%KT ketorolac; first derivative gels using different mucoadhesive viscosifying polymers. method; fungal keratitis; Methods: Niosomes were prepared using the reverse-phase evaporation technique. In vitro experimental, histopathological and in vivo clinical evaluations for these formulations were done for assessment of their safety and effi- examination cacy for treatment of Candida Keratitis in Rabbits. In vitro release study was carried out by the dialysis method. In vivo and histopathological studies were performed on albino rabbits. Results: NAT niosomes exhibited high entrapment efficiency percentage (E.E%) up to96.43% and particle size diameter ranging from 181.75 ± 0.64 to 498.95 ± 0.64 nm, with negatively charged zeta potential (ZP). NAT niosomal dispersion exhibited prolonged in vitro drug release (40.96–77.49% over 24h). NAT-loaded niosomes/0.5%KT gel formulae revealed retardation in vitro release, compared to marketed-product (NATACYNVR ) and NAT-loaded niosomes up to57.32% (F8). In vivo experimental studies showed the superiority for F8 in treatment of candida keratitis and better results on corneal infiltration and hypopyon level. These results were consistent with histopathological examination in comparison with F5 and combined marketed products (NATACYNVR and KetorolineVR ). Conclusions: This study showed that F8 has the best results from all pharmaceutical in vitro evaluations and a better cure percent in experimental application and enhancing the prolonged delivery of NAT and penetrating the cornea tissues.

Introduction nonionic surfactant vesicles that are formed from self-assembly of nonionic surfactant in aqueous media resulting in a closed bilayer Infectious keratitis represents one of the serious causes of corneal structure [7]. The bilayered vesicular structure is an assembly of blindness worldwide. Also, it is considered as the second cause of cataract [1]. Natamycin was the first antifungal specifically devel- hydrophobic tail of surfactant monomer shielded away from the oped for topical ophthalmic use and is currently the only topical aqueous space or core located in the center and hydrophilic head ophthalmic antifungal compound approved by the Food and groups in contact with the same. Drug Administration of the United States [2] and commercially The advantages of niosomes over conventional dosage forms; available (NATACYNVR 5% ophthalmic suspension; Alcon). It is the niosomes particles can act as drug reservoirs, can carry both reported to have a broad spectrum activity against various fungi hydrophilic drugs or hydrophobic drugs either in an aqueous core [3]. Its solubility is 20–50 mg/L in water [4]. Natamycin (Pimaricin) or in vesicular membrane [8,9], the encapsulated drug can be has been considered as the drug of choice for filamentous released in controlled fashion [10] and the drug release rate can mycotic keratitis (MK). Current therapy with natamycin appears be adjusted [8]. In addition, niosomes have been reported to unsatisfactory for several reasons; such as high dosing frequency, decrease the side effects [11]. Niosomes are osmotically active and long treatment duration (4–6 weeks), and short residence time at are stable on their own, as well as increase the stability of the the ocular mucosa. This prolonged dosing schedule is difficult to entrapped drugs [12,13]. They are biodegradable, biocompatible attain; resulting in suboptimal concentration at the corneal site, and non-immunogenic. Niosomes improve the therapeutic per- treatment failure and increase resistance to MK [5,6]. formance of drug molecules by delayed clearance from the circu- Novel carrier systems, such as niosomes would provide great lation, protecting the drug from biological environment, and potential for improving drug ocular bioavailability. Niosomes are restricting its effects to target cells [14]. They are preferred to

CONTACT Mahmoud Teaima [email protected] Department of Pharmaceutics and Industrial Pharmacy, Faculty of Pharmacy, Cairo University, Cairo, 11562, Egypt ß 2019 Informa UK Limited, trading as Taylor & Francis Group DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 923

Table 1. Composition of lipids for preparation of NAT niosomes. NAT (% Conc.) Span 20 (molar ratio) Cholesterol (molar ratio) Dicetylphosphate (molar ratio) Formula Parts F1 0.2% 1 1 0.1 F2 0.2% 1 1 – F3 0.2% 1 0.5 0.1 F4 0.2% 1 0.5 – typical liposomes as they provide higher chemical stability, ease KetorolineVR Ophthalmic Solution Drops, Pharopharma, (Cairo, of production, less cost, and broad flexibility in type of surfactant Egypt) were purchased from private pharmacy in (Egypt). [15]. They have been studied through the most common routes of Hydroxypropyl methylcellulose-E4 (HPMC-E4) was purchased from administration; such as intramuscular, intravenous, subcutaneous, Tama, Tokyo, Japan. ocular, oral, transdermal, and vaginal [16]. However, in recent years, there were a few studies on ocular delivery of natamycin using novel approaches to overcome aforementioned problems Methods associated with treatment with marketed preparation such as Preparation of NAT niosomes Bhatta et al., who prepared natamycin encapsulated within leci- NAT niosomes were prepared using the reverse-phase evaporation thin-chitosan mucoadhesive nanoparticles demonstrated improved technique according to Szoka and Papahadjopoulos [25,26]. Lipid pharmacokinetic profiles in comparison to the marketed natamy- components consisting of Span 20, cholesterol and with or with- cin suspension. In another study by Chandasana et al. [5,17], who out DCP were dissolved in mixture of chloroform and diethyl prepared poly D-glucosamine functionalized polycaprolactone ether (1:1) in round-bottomed flask. The organic solvent system nanoparticles of natamycin were prepared for targeting corneal was slowly evaporated under reduced pressure using a rotary mycotic keratitis. Also in another study by Phan et al. [18], they evaporator (Rotavapor, Hidolph-Laborota4000, D-91126, Germany) prepared natamycin encapsulated within poly(D,L-lactide)-dextran nanoparticles (Dex-b-PLA NPs) from model contact lens materials to obtain a thin film of dry lipid on the inner wall of the rotating to evaluate the uptake and release of the antifungal agent nata- flask. The lipid film was re-dissolved in diethyl ether. The0.2% of mycin. In yet another evaluation by Paradkar et al. [19], natamycin NAT [7] was dissolved in methanol and phosphate buffered saline niosome-loaded insitu gel could sustain the natamycin release up (PBS) pH 7.4 [27,28] as aqueous phase and added to the lipid to 24 h in comparison to the marketed formulation with a signifi- solution. The resulting two-phase system was sonicated for 5 min cantly higher trans-corneal penetration. in bath-type sonicator. Then the organic solvents of the mixture Many reports have shown that increasing the viscosity of oph- were removed under reduced pressure using a rotary evaporator thalmic solutions using cellulose derivatives as viscosity imparting at 50 C and 60 rpm for 15 min. The niosomes dispersion was agent would increase the precorneal residence time, promotes allowed to equilibrate at room temperature, kept in the refriger- ocular drug absorption through decrease in the drainage rate [14] ator (4 C) to mature overnight [10,29]. and hence increases ocular bioavailability [20]. The addition of The niosomal components and NAT concentration are illus- anti-inflammatory drug as KT to the antifungal drug is very trated in Table 1. important to treat inflammation, ulcer, and edema associated with MK [21,22]. Candida albicans utilizes arachidonic acid (AA) released during the course of infection (Candidiasis) from phospholipids of Pharmaceutical evaluation of NAT niosomes dispersion infected host cell membranes and synthesizes extracellular prosta- Determination of NAT entrapment efficiency percent (%EE) and glandin(s) which play an important role in hyphae formation and drug loading percent (%DL) in niosomes. The entrapped NAT con- host cell damage. The main mechanism of action of KT is inhib- taining niosomes was separated from unentrapped NAT by cool- ition of cyclooxygenase enzyme, which transforms AA into prosta- ing centrifugation (Megafuge1.0/1.0R, Kendro Laboratory Products glandins [23,24]. (Fussex, UK)) of a known aliquot (1 ml) of the prepared niosomal The objective of this study was to design and evaluate suspension at 15000 rpm for 30 min at 4 C[29,30]. The super- NAT-loaded niosomes/0.5% KT gel formulations. Clinical and histo- natant was separated from the niosomal precipitate. The precipi- logical examination of the corneal tissues of rabbits receiving tated niosomes was washed by 1 ml of PBS, pH 7.4, and niosomal gel formulations and comparison with the combined recentrifuged for 30 min to remove excess unentrapped NAT, and VR VR marketed products (NATACYN and Ketoroline ) Ophthalmic sus- then the combined supernatant was diluted to 10 ml by PBS pH pension were carried out for assessment of ocular irritancy, clinical 7.4. The concentration of unentrapped natamycin was determined efficiency, bioavailability of niosomes and treatment inflammation, spectrophotometrically (Shimadzu U.V-1800-PC.UV–visible double ulcer, and edema associated with MK. beam spectrophotometer, Shimadzu Corp., Kyoto, Japan) at

kmax 304.5 nm. Materials and methods The %EE and %DL were calculated using the following equa- tions: Materials %EE ¼ ðÞCd Cf =Cd 100 (1) Natamycin and dicetylphosphate (DCP), Sodium carboxymethylcel- lulose (Na.CMC) with average Mw 90,000, Sorbitan monolaurate %DL ¼ ðÞCdCf =CTC 100 (2) (Span 20) and Cholesterol were purchased from Sigma Chemical Co. St. Louis, MO, USA. Ketorolac tromethamine USP (99.40% pur- where Cd is the total amount of natamycin added, Cf is the ity) (KT) was supplied by Amryia Pharm. Ind. (Egypt). NATACYNVR amount of unentrapped natamycin detected in the supernatant Ophthalmic Suspension, Alcon Laboratories, INC. (USA) and and CTC is the total amount of formulation components. 924 M. A. EL-NABARAWI ET AL.

Photomicroscopic analysis and transmission electron microscopy comparison, the in vitro release of an equivalent amount of VR (TEM). Samples of natamycin niosomal vesicles (F4) were exam- NATACYN suspension was carried out adopting the same proced- ined microscopically at magnification of 40 and 100 with a ure as previously described. binocular microscope (Leica-Queen 550IW, Germany) equipped Based on the above tests, selection of NAT niosomal dispersion with camera to study their size [29,31]. formulae was based on the highest EE%, with the highest in vitro The morphological of dispersed niosomes were also examined release at Q24hrs.and minimum Particle size. The selected formu- by Transmission Electron Microscope (JEOL, JEM-1400; Tokyo, lae were further investigated through Physical Stability Study. Japan) at the candidate magnification. A drop of dispersion was stratified onto a carbon-coated copper grid and left to adhere on Kinetic study of release profile of NAT-loaded niosomes. Data the carbon substrate for about 1 min. The dispersion in excess obtained from the release of the drug from different niosomal dis- was removed by a piece of filter paper. A drop of 2% phospho- persion formulae were kinetically analyzed using excel 2007VR tungestic acid solution was stratified and, again, the solution in (Microsoft, Software). Generally, zero order kinetics, first order excess was removed by a tip of filter paper. The sample (F4) was Kinetics and Higuchi diffusion models were used for the kinetic air-dried and photographed under TEM [32]. analysis of the release data. Korsmeyer–Peppas model was also taken in consideration (log Particle size distribution and zeta potential (n) measurements. cumulative percentage drug released vs. log time, Equation 4) [38] NAT niosomal suspension (100 mL) was diluted to 10 ml with dis- in order determine the mechanism of NAT release from niosomal ‘ ’ tilled water, and then subjected to photon correlation spectros- formulations. In this model, the release exponent n value is < < copy and laser Doppler anemometry [Zetasizer ZEN 3600 Nano ZS 0.45 for Fickian diffusion release and 0.45 n 0.89 for non- (Red badge) Malvern Instr., UK] for measurement of the particle Fickian release (anomalous) [39]: n size [29,33]. Mt=M1 ¼ Kt (3) The zeta potential of NAT niosomes was measured by diluting an aliquot of the niosomes suspension with a large amount of where Mt/M1 is the fraction of drug released at time t. K and n bidistilled water, and evaluated for surface charge at room tem- are kinetic constant and the diffusion exponent related to the perature using Zetasizer [29,34]. mechanism of the drug release, respectively. The values of K and n were estimated by linear regression of Log (Mt/M1) versus Log t where Log K is the intercept and n is the slope of the straight Differential scanning calorimetry (DSC) analysis. DSC experiments line. were performed with differential scanning calorimeter (Schimadzu ðÞ= ¼ þ DSC-TA-50 WSI, Schimadzu, Japan) calibrated with indium. Log Mt M1 Log K n Log t (4) Samples of natamycin, span 20, cholesterol, dried free drug-loaded niosomes and drug-loaded niosomes formula of (F4) were submit- Effect of storage of natamycin niosomes on the leakage of the ted to DSC analysis. Briefly, 2–4 mg of each sample was placed in natamycin and its release from niosomes (physical stability study). a standard aluminum pan, and heated from room to 400 Cat Physical stability study was performed to investigate the leak out constant scanning rate of 10 C/min. of the drug from niosomes during storage. Samples of NAT niosomal suspension formulation F4 (the optimized formula) which In vitro release study of NAT from niosomes dispersion. The in showed the highest drug release, minimum particle size, and high- vitro release of NAT from the prepared niosomal dispersion formu- est zeta potential, were sealed in 20 ml glass vials and stored in lations (F1, F2, F3, and F4) were carried out by the membrane dif- refrigerator at 4 C for 6 months. Samples were withdrawn at def- fusion technique [35,36]. inite time intervals 1, 2, 3, and 6 months and evaluated for its An accurately measured amount of niosomal dispersion formu- entrapment efficiency and for it’s in vitro release. lations, equivalent to 0.2% of natamycin was transferred to a glass The retention of entrapped drug was measured after periods cylinder having the length 8 cm and diameter 2.5 cm with surface of storage in the selected formulations as described previously in 2 area 4.91cm and fitted at its lower end with presoaked dialysis % EE determination [40]. Stability for the formulation was defined membrane on which the niosomal spread over. (Spectra/Pore in terms of retaining its initial entrapment efficiency for six dialysis membrane, 12,000–14,000 Mwt Cut off (Spectrum months duration as the following equation: Laboratories Inc., USA). The glass cylinder was attached to the % shaft of the dissolution apparatus and then was placed in vessels NAT retained in niosomes of the U.S.P. Dissolution Tester (Classic Version 6–Vextra-Model ¼ðEntrapped natamycin after storage (5) BLHMO15K-10 Oriental Motor, Co. Ltd., Japan), containing 120 ml = Entrapped natamycin before storageÞ100 PBS (pH 7.4) and 1% sodium lauryl sulfate (SLS) at 37 ± 0.5 C and Also, the effect of storage on drug release from niosomes was at rotation speed 75 rpm [26]. SLS was used as diffusion medium studied. Similarity factor (f ) was calculated to compare the release and to solubilize NAT [37]. Sink condition is fulfilled since 2 profile before and after storage [41]. Similarity factor (f ) was cal- the saturated solubility was previously determined to be 2 culated from the following equation [42]. 0.13 ± 0.002 mg/ml. hi Samples (4 ml) were collected at predetermined time intervals 0:5 f2 ¼ 50 log 10 ðÞ1 þ w=n 100 (6) till 24 h. Then, the collected samples were filtered using millipore filter (0.45 mm pore size) and immediately replaced with equal vol- where w is the sum of squares of differences in the cumulative ume of fresh buffer [35]. The absorbance of collected samples was percent dissolved between reference (release profile before stor- then analyzed spectrophotometrically at kmax 304.5 nm. The results age) and test (release profile after storage) and n is the number of the in vitro release experiments were repeated three times and sampling times till a drug released % of 85%. If (f2) value lies the mean ± SD was calculated. The results obtained from the in between 50 and 100, the two dissolution profiles are considered vitro were compared with the results from NATACYNVR release. For to be similar [42]. DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 925

Preparation of NAT-loaded niosomes/0.5%KT gel formulations followed as previously described in “In vitro release study of NAT In order to facilitate the administration of niosomal preparations from niosomes dispersion” section. The absorbance of the col- to patients and to control NAT release, the optimized niosomal lected samples were measured spectrophotometrically at kmax formula F4 was incorporated into different types of 0.5% KT gel. 323.5 nm for natamycin and at kmax 344 nm for KT by first deriva- NAT-loaded niosomes/0.5% KT gel formulations used two types tive method [45]. The results obtained from the in vitro were com- of polymers Na.CMC and HPMC-E4, each containing 0.5%KT at pared with the results of NATACYNVR . two different ratios for each (2% and 4%w/w) and were given The results of in vitro release of all investigated NAT-loaded codes F5 (2% HPMC-E4), F6 (4% HPMC-E4), F7 (2% Na.CMC), and niosomes/0.5%KT gels formulations were kinetically analyzed as F8 (4% Na.CMC). Combined formulation of NAT-loaded niosomes previously mentioned in “Kinetic study of release profile of NAT- and 0.5% KT gel were prepared by mixing the niosomal dispersion loaded niosomes” section. with the 0.5% of KT gels in order to have a final natamycin concentration 0.2% in 0.5% KT gels [10]. Effect of gamma (c) sterilization on the chosen NAT-loaded niosomes/0.5%KT gel formulae. The selected NAT niosomes-loaded/ 0.5%KT gel formulae (F5, F6, and F8) were subjected to c-steriliza- Evaluation of the prepared NAT-loaded niosomes/0.5%KT gels tion in the presence of dry ice to avoid the side effect that might Determination of in vitro mucoadhesion strength. The modified occur due to temperature elevation by means of c-irradiation [46]. two arms physical balance method with minor modifications was Samples were irradiated using a [47] Co irradiator (National Center used for studying the mucoadhesive force of the prepared oph- for Radiation Research and Technology, Nasr City, Cairo, Egypt) at thalmic gel [43,44]. The modified physical balance apparatus com- a dose 25 KGy and a dose rate of 1.774 KGy/h. The in vitro drug prised of a tared two-arm, one side of which contained release study of the selected NAT-loaded niosomes/0.5%KT gels, experimental tools and the other side contained a beaker. The before and after sterilization, were compared using the similarity apparatus was designed for measuring the minimum weight factor (f ) as previously described in “Effect of storage of natamy- required for detachment of two membranes from each other with 2 cin niosomes on the leakage of the natamycin and its release a film of polymer spread between them. from niosomes (physical stability study)” section. Briefly, a section of rabbit’s small intestine membrane was used as model mucosa for this study. The section of rabbit’s small intestine membrane was attached to the undersurfaces of 100 gm Long-term stability study of NAT-loaded niosomes/0.5%KT gel for- weight and on the inverted beaker using a cyanoacrylate adhe- mulae (physical stability study). Physical stability study was per- sive. An accurately weighed 0.5 gm of the gel prepared was added formed to investigate the leak out of the drug from NAT-loaded onto the rabbit’s small intestine membrane of the inverted beaker niosomes/0.5%KT gel formulae (F5, F6, and F8) during storage. at room temperature. A 100 gm weight was suspended by means The formulations were stored at room temperature for a period of of a thin steel wire to the left of the balance. The balance was 9 months. Samples of NAT-loaded niosomes/0.5%KT gel were with- equated. The 100 gm weight was removed from the steel wire of drawn at definite time intervals 1, 2, 3, 6, and 9 months and eval- the left side of the balance, and of the gel prepared just to come uated for its pH measurement, drug content and in-vitro release into contact with 100 gm weight. After 4 min of the contact, the of natamycin. 100 gm weight was reattached to the steel wire and the switch of The in vitro drug release study of formulations before and after the infusion apparatus was opened to allow the water drop into storage was compared using the similarity factor (f2) as previously the beaker at a constant flow rate of 13–15 drops per minute. The described in “Effect of storage of natamycin niosomes on the leak- weight of the water in the beaker was kept increasing until the age of the natamycin and its release from niosomes (physical sta- gel and rabbit’s small intestine membrane just detached from bility study)” section. each other. The minimal weight of water required for detachment of the glass plates was recorded and the mucoadhesive strength Antifungal activity study. The antifungal activity study of nio- of the applied gels was calculated as force of detachment. These somes formulae F5 and F8 in comparison to CMP was performed experiments were repeated with fresh rabbit’s small intestine and on Mueller–Hinton broth plates as previously described according gel samples in an identical manner (n ¼ 3). All detachment tests to National Committee for clinical laboratory Standards 1993; were carried out at room temperature. The mucoadhesive force, using the agar well diffusion technique [48–50]. 2 that is, the detachment stress (dyne/cm ) was determined using Agar wells were made with the help of a sterilized cork borer the following equation stated [43]. having an inner diameter of 6 mm. 100 lL aliquot of each of Detachment stress dyne=cm2 ¼ m g=A (7) niosomal gel formulations; F5 and F8, CMP and free drug niosomal gel (FDNG) as blank were filled to the cups with a sterile syr- where m: the weight of water in gm; g: acceleration due gravity inge. The plates were left for 30 min to allow the diffusion and 2 2 taken as 981 cm/s ; A: area of rabbit’s small intestine (cm ) (area incubated at 37 C for 24 h, after which the zones of inhibition of contact). around the wells were measured in millimeter. The minimum inhibition concentration (MIC) for F5 and F8 was Determination of viscosity. Viscosity of Combined formulations of performed as previously described [50]. The formulations were ini- NAT niosomes containing 0.5%KT gel was measured using tially dissolved in DMSO with different concentrations Brookfield viscometer (Brookfield DV-E Viscometer) at various (8–0.0078125 mg/ml) of the formulations and dispensed in wells. speeds using spindle 6 at 25 C[43]. The zones of inhibition around the wells were measured as previously described. In vitro release studies of NAT-loaded niosomes/0.5%KT gels. An accurately measured amounts of NAT niosomal 0.5% KT gel for- Statistical analysis:. Data of pharmaceutical evaluations were mulations, equivalent to 0.2% of NAT were carried out by the dia- expressed as mean ± standard deviation (SD). Statistical analysis lysis method as the same procedures and conditions were was accomplished using Data Analysis excel sheet. The 926 M. A. EL-NABARAWI ET AL. significance differences between the mean values of the pharma- procedure for documentation, from the start of infection to last ceutical evaluations were determined by One way analysis of vari- day of treatment. Two weeks later, rabbits were sacrificed; corneas ance (ANOVA) statistical test at significant level (p < 0.05). were dissected at the limbus and sent to department of Pathology Kasr Al-Aini Hospital, Cairo University, under complete aseptic conditions in 10% formaldehyde solution to evaluate the In vivo studies presence or absence of the fungus and extent of inflammation for Visual assessment of ocular irritation potential of the optimized each formula. As per REC-FOPCU recommendations, the bodies NAT-loaded niosomal/KT gel. Two groups, each of three albino and remains of the rabbits were frozen and transferred to be rabbits, were assigned for testing formulae. Within each group, incinerated at the Faculty of Veterinary Medicine, Cairo the right eye of the rabbits received (100 mg) of NAT-loaded nio- University, Egypt. somes/KT gel (0.2% w/v NAT and 0.5% w/v KT) F5 and F8 as the first and second group, whereas the left eye was kept as a control. Histopathological evaluation of corneal specimens. Two slides ’ Application of the tested formulation onto rabbit s cornea was were prepared for every specimen; one stained by hematoxylin repeated three times daily for a week [51]. According to the Draiz and eosin (H&E), and the other one stained by Gomori test [52], ocular irritation scores for every rabbit were calculated Methenamine silver (GMS). All H&E stained slides were examined by adding together the irritation scores for the cornea, the iris, under light microscope to evaluate the type and intensity of and the conjunctiva. The eye irritation score was obtained by inflammatory reaction as follows: dividing the total scores for all rabbits by the number of rabbits.

The presence of necrotizing inflammation with wide exudation In vivo antifungal evaluation study. It was conducted at the of neutrophils was scored as 3. Animal- House in Kasr Al-Aini Hospital, Cairo University, in the The presence of moderate inflammatory reaction with moder- period from September to December 2017. Forty-five male New ate exudation of neutrophils was scored as 2. Zealand albino rabbits weighing about 2–2.5 kg, without any eye The presence of mild inflammatory reaction with mild or min- diseases, were included in the experiment. The protocol of this imal exudation of neutrophils was scored as 1. study (S.No. PI 2102) was approved by the Research Ethics The absence of inflammatory reaction was scored as 0. Committee in the Faculty of Pharmacy, Cairo University (REC- FOPCU), Egypt. The use and treatment of rabbits in this study All GMS stained slides were examined under light microscope were conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research as well as the EU Directive 2010/ to evaluate the presence or absence of fungal spores as follows: 63/EU for animal experiments. In the operating room, all rabbits were sedated using intramus- The presence of large number of fungal spores was scored cular injection of 1 ml (50 mg/ml) Ketamine (KetalarVR ), moreover, as 2. topical anesthesia was used by applying Benoxinate eye drops The presence of small number of fungal spores was scored into the conjunctival sac of rabbits’ eyes. Right corneas of all rab- as 1. bits were marked using 7 mm corneal trephine, followed by scrap- The absence of fungal spores was scored as 0. ping of corneal epithelium, then corneas were inoculated with Candida. Needle 27gauge was used to inject (0.1 ml of Candida Statistical analysis. Statistical analysis was done by IBM SPSS species with a concentration of 5 105cells/ml) into posterior cor- v21.0 statistical software (IBM Corporation, New York, USA). neal stroma [21]. In either case, the left eyes were kept as control Descriptive statistics was calculated and the data were summar- and received equal volume of saline solution. ized as mean ± standard deviation (± SD), or frequencies (number The whole procedure was done under complete aseptic pre- of cases) and percentages as appropriate. For assessing the associ- cautions, with the aid of a binocular microscope and washing the ation between categorical data, Chi square (v2) test was per- eyes with Chloramphenicol eye drops. formed. Fisher’s exact test was used instead when the expected Forty-eight hours later, all corneas were examined for signs of frequency is less than 5. Comparison of numerical variables infection to confirm flourishing of the organism. Then, the treat- between the groups was done using one-way ANOVA test. The ment was started where, the rabbits were subdivided into 3 results will be considered statistically at significant level (p < 0.05). groups, and given codes and numbers, the right eyes of the rabbits in each group received a different antifungal formula. Groups A and B; each 15 rabbits received 300 mg of F5 and F8 (0.2%w/v Result and discussion m NAT 0.5%w/v KT). Group C: 15 rabbits received 2 drops (100 l) of Pharmaceutical evaluation of NAT niosomes dispersion CMP (2% NATACYNVR and 0.5% KT). In all cases, the left eyes were kept as control and received equal volume of saline solution. Preparation of NAT niosomes Application of the tested formulation onto rabbit’s cornea was The reverse-phase evaporation technique is considered an ideal repeated three times daily for 2 weeks. and widely method used for preparation of NAT niosomes. Also, The rabbits were examined every day. Follow-up with a written this technique is relatively simple and reproducible and gave satis- sheet was done for each rabbit according to system of enumer- fied results with respect to highly percent encapsulation efficiency ation. Treatment and comments were done by an ophthalmolo- and drug release extent [10,29]. gist, who did not know the difference among the two formulae (F5 and F8) and CMP and was totally blinded in order not to be biased. Percent entrapment efficiency (%EE) and percent drug load- The following clinical data were evaluated: extent and progress ing (%DL) of ciliary injection, size, and depth of corneal infiltration, and the The results of entrapment efficiency and drug loading of NAT in anterior chamber reaction. Photos were taken along the whole niosomes are illustrated in Table 2. The increasing in entrapment DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 927

Table 2. Physical parameters of NAT niosomal formulations and (n ¼ 3). % Entrapment Efficiency (% EE) % Drug Loading (% DL) Particle size (nm) Zeta Potential (mV) Niosomal Formula mean ± SD mean ± SD mean ± SD mean ± SD F1 (1:1:0.1) 92.38 ± 0.15 14.71 ± 0.008 498.95 ± 0.64 65.15 ± 3.89 F2 (1:1) 95.93 ± 0.65 14.74 ± 0.025 382.25 ± 1.34 52.90 ± 4.10 F3 (1:0.5:0.1) 90.99 ± 0.77 14.68 ± 0.011 367.85 ± 0.92 73.75 ± 4.17 F4 (1:0.5) 96.43 ± 0.24 14.75 ± 0.009 181.75 ± 0.64 58.95 ± 0.64 of NAT niosomal formulations may be due to the large encapsu- the particles size were 367.85 and 181.75 nm, respectively. These lated volume of unilamellar vesicles prepared by REV tech- results can be attributed to the inclusion of a charge inducer in nique [10]. niosomes, which increased the spacing between the adjacent The in neutral charged niosomes F2 and F4 (95.93% and bilayers, resulting in the formation of niosomes larger in size com- 96.43%, respectively) have higher entrapment efficiency than pared with the neutral ones [47]. Also another explanation is the negatively charged niosomes F1 and F3 (92.38% and 90.99%, incorporation of charged molecules into the bilayer increases the respectively) at all molar ratios of cholesterol. The lower entrap- volume of the aqueous compartment by separating adjacent ment efficiency of negatively charged niosomes could be attrib- bilayers due to charge repulsion, resulting in increase in the par- uted to that electrostatic charge repulsion forces may occur ticle size. The increase in the particle size of the negatively between the carboxylate anion of NAT molecules and the nega- charged niosomes prepared could be explained by the difference tively charged DCP, thus suppressing the loading effi- in chemical structures of Span 20 and DCP. The chemical structure ciency [53,54]. of DCP is composed of two cetyl chains, whereas, Span 20 com- The drug loading of NAT in niosomes ranged from 146.77 to prises of lauryl chain. During the formation of vesicle tightly 147.77 mg/mg. The drug loading of NAT in neutral charged nio- packed bilayers could not be formed resulting in reduction of somes F2 and F4 (14.743% and 14.745% respectively) have higher membrane curvature and obtaining the larger particle size [36]. drug loading than negatively charged niosomes F1 and F3 These results are in agreement with Junyaprasert et al, Hathout (14.71% and 14.677%, respectively) [55]. et al and Hosny et al., who studied the incorporation of charge Statistically, one way ANOVA test was applied to test for varia- inducing agent into salicylic acid niosomes [36], acetazolamide lip- – < tions all formulae F1 F4 at significant level (p 0.05). The results osomes [47], and ciprofloxacin liposomes [26], respectively. showed significant differences in entrapment efficiency where The results tabulated in Table 2 showed that the zeta potential ¼ (p 4.91E 07). Also, there was significant differences between for NAT-loaded neutral charged niosomes (F2 and F4) were ¼ negatively charged niosomes (F1 and F3) where (p 0.0065). This 52.90 and 58.95 mV, respectively. While the zeta potential for effect was due to the difference in molar ratios of cholesterol, that NAT-loaded negatively charged niosomes (F1 and F3) were is, increasing in cholesterol molar ratio of neutrally charged nio- 65.15 and 73.75 mV, respectively. This effect was attributed somes have higher entrapment efficiency than decreasing in chol- due to the fact that DCP introduced the negative charge onto the esterol molar ratio of negatively charged niosomes in presence of surface of niosomes and the extent of ionization increased at inducing negative charges. Also, there were no significant differ- higher pH (pK of DCP 4.5). At pH 7.4, the process of dissociation ences between neutrally charged niosomes (F4 and F2) where a reached completion, leading to large increase in zeta potential of (p ¼ 0.273) due to the absence of inducing negative charges at all NAT-loaded negative charged niosomes (F1 and F3) and leads to molar ratios of cholesterol. increase in particles size as previously explained as compared to that NAT-loaded neutral charged niosomes [36]. Photomicroscopic and TEM analysis The photomicrograph of NAT niosomes prepared by REV tech- DSC analysis nique of F4 as the best chosen formula is shown in Figure 1(a,b). DSC thermograms of NAT, NAT-loaded niosomal formulation (F4) It showed that the niosomes are spherical in shape [47]. and free drug niosomal formulation, as well as niosomal compo- Negative stain transmission electron micrographs of NAT nio- nents, is shown in Figure 2. Thermogram of NAT demonstrates a somes dispersion (F4) is shown in Figure 1(c,d). It demonstrated sharp endothermic peak at 188.86 C corresponding to its melting that the vesicles are well identified and present in a nearly perfect point. This is in agreement with the data reported for NAT [56]. sphere-like shape having a large internal aqueous space relative to the sphere diameter [47]. Thermograms of individual niosomal components including Span 20, and cholesterol showed endothermic peaks at each compo- nent’s transition temperature; 41.16 C and 149.19 C, respectively. Particle size distribution and zeta potential (n) measurements DSC thermogram of NAT-loaded niosomal formulation F4, showed The results of particle size analysis are illustrated in Table 2 and shifting broad endothermic peak from 188.86 C to 231.75 C due shown in Figure 1(e,f). They reveal that as the concentration of to interaction of all components forming the bilayers of niosomes cholesterol increases, the particle size also increases, which might [57] and can account for the enhanced entrapment of NAT into be due to the formation of rigid bilayer structure [26,29]. This these formulations [29,47], while endothermic peaks of span 20 effect is shown in the negativelu charged niosomes (F1 and F3), were shifted to from 41.16 C to 26.06 C. DSC thermogram of free and the particles size were 498.95 and 367.85 nm, respectively. drug-loaded niosomes showed the absence of melting endother- Also, this effect is shown in the neutrally charged niosomes (F2 mic peak of NAT and the endothermic peak of Span 20 and chol- and F4), and the particles size were 382.25 and 181.75 nm, esterol were shifted from 41.16 C and 149.19 C to 30.33 C and respectively. 122.56 C, respectively. This may suggest that NAT can fit into the Also as the presence of DCP, the particle size of different for- niosomal bilayer producing significant perturbation in the packing mulations also increases as shown in F1 and F2, the particles size characteristics of the niosomal membrane [57]. These results are were 498.95 and 382.25 nm, respectively, and also in F3 and F4, in accordance with those obtained by El-Ridy et al. [58] during the 928 M. A. EL-NABARAWI ET AL.

Figure 1. Photomicrograph of NAT-loaded niosomes, magnification power40 x (a) and 100 x (b), TEM (b and c) and particle size (e and f). investigation of silymarin niosomal encapsulation with different (CV%4.82), respectively, while from neutral charged niosomal for- nonionic surfactants. mulations of F2 and F4 were 65.66% (CV%2.37) and 77.49% (CV%1.71), respectively. In vitro release study of NAT from NAT-loaded nio- The results in Table 3 showed that the superiority for F4 over somes dispersion other formulae and the release at 24 h was 77.49%, which was V The release profile of NAT from niosomal formulations was illus- more controlled release in comparison with NATACYN R 85.52% trated in Table 3 and shown in Figure 3(a). The percentage of (CV% 3.93) after 8 h. NAT released after 24 h from negatively charged niosomal formu- Statistically, the results showed significant differences in the in lations of F1 and F3 were 40.96% (CV%3.58) and 48.00% vitro release between all formulae F1–F4 at significant level DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 929

Table 3. In vitro release and kinetics analysis data of NAT from niosomal suspension and niosomes-loaded/0.5% KT gel (n ¼ 3). Higuchi % NAT % NAT diffusion- released at released at controlled Korsmeyer Kinetic Release Niosomal/KT 8h (T8h) 24h (T24h) Zero order First order release Peppas Order constant exponent Mechanism gel formula mean ± SD mean ± SD (R2)(R2)(R2)(R2)of release (K) (n) of transport NATACYNVR 85.52 ± 3.36 100.66 ± 1.26 0.5875 0.4623 0.8263 0.9362 Diffusion 43.27 0.326 Fickian F1 27.05 ± 1.23 40.96 ± 1.47 0.8042 0.4991 0.9594 0.9405 Diffusion 6.88 0.650 Non-Fickian F2 38.94 ± 0.62 65.66 ± 1.56 0.9446 0.7808 0.9969 0.9844 Diffusion 17.96 0.379 Fickian F3 19.85 ± 0.62 48.00 ± 2.31 0.9821 0.7549 0.9569 0.9865 Zero 5.24 0.664 Non-Fickian F4 41.52 ± 0.86 77.49 ± 1.32 0.9568 0.7416 0.9919 0.9941 Diffusion 15.68 0.479 Non-Fickian F5 21.31 ± 2.47 41.07 ± 2.74 0.9471 0.6603 0.9955 0.9902 Diffusion 6.28 0.599 Non-Fickian F6 17.49 ± 0.86 37.32 ± 0.72 0.9774 0.777 0.9767 0.9872 Zero 5.82 0.547 Non-Fickian F7 30.12 ± 0.40 61.95 ± 1.60 0.9597 0.6632 0.9945 0.9981 Diffusion 6.94 0.708 Non-Fickian F8 25.02 ± 1.20 57.32 ± 1.47 0.9829 0.7083 0.9745 0.9982 Zero 5.47 0.734 Non-Fickian

Table 4. Physical stability of natamycin niosomal formula (F4) stored at refrigerated temperature (4 C), similarity factors of NAT entrapped and in vitro release profiles at different time intervals compared to the zero time profile (n ¼ 3). Percent of drug retained in Percent drug released after period of storage niosomes (%±S.D) (recovery) (%±S.D) % Drug released time of storage to Similarity factor of NAT Similarity factor of NAT Time Intervals (months) EE% %Drug retained % Drug released initial release entrapped (f2) released (f2) At (initial) 96.43 ± 0.24 100 ± 0.00 77.49 ± 1.32 100.00 ± 0.00 –– After 1 month 96.15 ± 1.63 99.71 ± 1.69 76.32±1.08 98.49 ± 1.40 99.93 61.76 After 2 months 95.26 ± 0.85 98.79 ± 0.88 75.84 ± 3.11 97.88 ± 4.01 98.83 54.95 After 3 months 94.93 ± 1.95 98.44 ± 2.05 74.84 ± 0.95 96.64 ± 1.22 98.13 51.07 After 6 months 92.20 ± 0.49 95.61 ± 0.51 73.93 ± 2.30 95.41 ± 2.97 90.06 55.41

medium followed by a slowing in drug release. An initial burst release is beneficial in terms of antifungal activity of therapeutic concentration of drug in minimal time followed by constant release to maintain sustained release of drug. The burst release phase was mainly due to desorption and diffusion of drug from the surface. It was supposed that smaller particles of niosomes possessed more specific surface area and the drug delivery process would be facili- tated [7]. This release pattern was in agreement with that reported in previous works [5,7,58]. Another explanation for initial fast release phase could be because the drug is mainly incorporated between the fatty acid chains in the lipid bilayers of niosomal vesicles which leads to rapid ionization and release upon dispersing niosomes in large buffer pH 7.4 volumes until reaching equilibrium. Also, it has been reported that, highly ordered lipid particles cannot accommodate large amounts of drug and is the reason for initial burst release [61]. Almost the entire amount of loaded drug was not released from the niosomes. This may be due to entrapment of Figure 2. DSC measurements of niosome components and NAT-loaded and the drug in the lipophilic region. unloaded niosomal formulae. Linear regression analysis for the release profiles data of natamycin from niosomes (F1–F4) is tabulated in Table 3. The data showed (p < 0.05), where (p ¼ 9.16E05) at 24 h. Also, there were signifi- that the natamycin is released following Higuchi diffusion controlled cant differences in the drug extent release at 8 h and 24 h mechanism except F3 showed zero order. Korsmeyer–Peppas was between formulae F1 and F2, where (p ¼ 0.0067) and (p ¼ 0.0037), used to further confirm the exact mechanism of drug release. The respectively. Also, there were significant differences in the extent data were treated with korsmeyer–Peppas model and the mechan- of drug release at 8 h and 24 h between F3 and F4 at 8 h and 24 h ism of drug release was identified according to the value of expo- where (p ¼ 0.0027) and (p ¼ 0.0041), respectively. This may be due nent n which was found between 0.45 and 0.89 for F1, F3, andF4 to the fact that charged lipids serve to tighten the molecular indicating non-Fickian (anomalous) release mechanism, including packaging of the vesicle bilayer, resulting in lower extent of drug both erosion and diffusion. Whereas, F2 and NATACYNVR the expo- release from negatively charged niosomes than from the neutrally nent n was 0.45 indicating Fickian release mechanism. charged niosomes [59,60]. The results revealed that the release of NAT from niosomes of different formulae was biphasic release processes. Rapid drug Effect of storage of natamycin niosomes on the leakage of the release was observed during the initial phase where about natamycin and its release from niosomes (physical stability study) 10–30% of the entrapped NAT was released from various formula- Physical stability study of NAT niosomal suspension formula (F4) tions in the first three hours of niosomal incubation in dissolution which gave the highest entrapped and in vitro release profiles was 930 M. A. EL-NABARAWI ET AL.

Table 5. NAT loaded niosomes/0.5% KT gel, maximum and minimum viscosity and mucoadhesive force (n ¼ 3). g Max. (cP) g Min. (cP) Detachment force (dyne/cm2102) Formula Polymer concentration mean ± SD mean ± SD mean ± SD F5 2% HPMC-E4 9306.00 ± 10.15 523.00 ± 5.29 147.37 ± 9.19 F6 4% HPMC-E4 18943.33 ± 11.06 747.33 ± 6.51 170.91 ± 11.8 F7 2% Na.CMC 2734.67 ± 7.80 78.00 ± 2.65 109.00 ± 19.98 F8 4% Na.CMC 4233.33 ± 3.04 492.67 ± 3.71 178.76 ± 19.98

(a) 110 120 100 90 100 80 70 80 60 60 50

Released 40 40 30 20

Percentage of Natamycin Percentage 20 10

0 Percentage of Natamycin Released 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (hr) Time (hr) NATACYN® F1 F2 F3 F4 NATACYN® F4 F5 F6 F7 F8 (b) 100 Figure 4. In vitro release study of NAT from NATACYNVR , F4 and NAT-loaded nio- 90 somes/0.5%KT gel formulations (F5–F8). 80 70 60 them are sufficiently stable under refrigerator storage and the 50 advantages of the lipid membrane were retained [47]. 40 Released 30 20 Pharmaceutical evaluation of the prepared NAT-loaded

Percentage of Natamycin Percentage 10 niosomes/0.5%KT gels 0 0 5 10 15 20 25 30 Evaluation of viscosity of NAT-loaded niosomes/0.5% KT gels The results of viscosity evaluation of gel formulae are illustrated in Time (hr.) Table 5. It could be observed that the viscosity of the formulations Zero time 1 month 2 month follows the order F6 > F5 > F8 > F7. The results revealed that 3 month 6 month increasing in the polymer concentration led to increasing in the Figure 3. (a) In vitro release study of NAT from NAT niosomal dispersion formula- viscosity of gel [29,62]. tions (F1–F4) and commercial eye drops NATACYNVR . (b) Effect of storage on in vitro release of NAT from the best chosen niosomal dispersion formulation F4.

ͦ Evaluation of in vitro mucoadhesion properties of gels conducted at 4 C for 1, 2, 3, and 6 months. Drug leakage from F4 The effect of polymer type and concentration on the mucoadhe- was evaluated at definite time intervals, and the results are dem- sive strength of different gels are summarized in Table 5. In gen- onstrated in Table 4. Also, the release from this formula was eral, increasing the concentration of each polymer in the gels studied after periods of storage as shown in Figure 3(b). significantly increases the mucoadhesive strength. These results The respective percentages of natamycin entrapped as retained are consistent with literature data [43,44]. and in vitro released in the niosomal formulation after 1, 2, 3, and Results revealed that F8 and F6 gels showed the highest values 6 months from the initial formulation (F4) were 99.71, 98.79, 98.44, of mucoadhesive strength. Na.CMC had higher mucoadhesive 95.61, and 98.49, 97.88, 96.64, 95.41, respectively. One way strength to the mucous compared with HPMC due to Na.CMC is ANOVA reveal that there are no significant differences in the per- considered an anionic polymer, which can be strongly adhered to centages of natamycin entrapped and in vitro release at level mucin more than nonionic HPMC. Similar evaluation found in the (p < 0.05) between formula (F4) at initial time and periods of stor- literatures [63,64]. age, where (p ¼ 0.098) and (p ¼ 0.492), respectively. Statistically, one way ANOVA test was applied on all formulae Similarity factors (f2) between NAT entrapped and the in vitro at significant level (p < 0.05). The results showed a significant dif- release profiles of the formulations and compared to the initial ferences in mucoadhesive force, where p ¼ 0.00283. time before and after storage periods were calculated and illustrated in Table 4. The highest similarity factor values of entrapped and in vitro release profiles were found 99.93 and 61.76, respect- In vitro release studies of NAT niosomal/0.5%KT gels ively. This indicates that the similarity of the entrapped and in The incorporation of NAT niosomes in gel resulted in further delay vitro release profiles in the acceptable range (between 50 and due to the formation of an additional diffusion barrier to drug 100) before and after the period of storage. release [65]. The release extent of natamycin from niosomal gels It is obvious from the results that despite the partial hydrolysis through the membrane was apparently dependent on polymer that would occur for the surfactants, the niosomes made from concentration used (HPMC-E4 and Na.CMC), where the increase in DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 931 the polymer concentration is associated with decrease in the Statistically, there were significant differences in the release release extent due to increase in viscosity of the prepared gel. extent profile between all formulae F5–F8 at significant level These results are in accordance with Jones et al. [66] and El- (p < 0.05) at 8 h and 24 h, where p ¼ 9.17E05 and p ¼ 3.16E05, Nabarawi et al. [29], who stated that the physical reason for respectively due to the difference in the viscosity of the prepared slower extent rate of release from viscous gel is most probably gel. Also, there were significant differences in the release extent due to formation of highly viscous diffusion layers of hydrated profile between niosomal gels formulations (F5–F8), F4 and polymer chain which entraps the excess of water and reduce NATACYNVR at 8 h and 24 h, where p ¼ 2.128E7 and p ¼ 1.534E7 migration of drug molecules. respectively. The results illustrated in Table 3 and shown in Figure 4 It is clear that the encapsulation of NAT within niosomes revealed that the extent released of NAT from Na.CMC was more extends the time required for releasing of 75% was 24 h, while than HPMC-E4 at 8 h and 24 h due to the difference in viscosity of the time required for releasing 75% NAT from NATACYNVR , was 4 h. polymers used (Table 5), where the viscosity HPMC-E4 was more Also, the time required for releasing of 75% of NAT from niosomal than the viscosity of Na.CMC. The percentage of NAT released gels was over 24 h. This may due to the viscosity of polymeric obtained from F5, F6, F7, and F8 after 24 h were 41.07% hydrogel used, which provide an extra barrier for NAT release [62]. (CV%6.67), 37.32% (CV%1.93), 61.95% (CV%2.59), and 57.32% Linear regression analysis for the release profiles data of nata- (CV%2.56), respectively. These results showed an increase in the mycin from niosomes (F5–F8) is tabulated in Table 3. The data release from F5 > F6 in HPMC-E4 and F7 > F8 in Na.CMC due to showed that the natamycin is released following Higuchi diffusion the low viscosity of these formulations F5 and F7, on the other controlled mechanism F5 and F7, whereas F6 and F8 showed zero hand, the decreasing NAT released obtained from F6 and F8 due order. The data were treated with korsmeyer-Peppas model and to the high viscosity exhibited by this formulation [67]. Another the mechanism of drug release was identified according to the reason for increasing the release extent of Na.CMC is due to the value of exponent ‘n’ which was found between 0.45 and 0.89 high solubility of Na.CMC at this pH 7.4 value. This polymer char- indicating non-Fickian (anomalous) release mechanism, including acteristic gives a quick gel erosion rate and a high erosion degree both erosion and diffusion. of the overall system [68]. The two formulae F5 and F8 were taken for in vitro antifungal The results of release profile of NAT from NAT-loaded nio- activity evaluation and in vivo study due to its satisfied in vitro somes/0.5%KT gel at 8 h and 24 h showed that the superiority of pharmaceutical evaluations. F8, and F5 over other formulae where the release at 24 h were 57.32% (CV%2.56) and 41.07% (CV%6.67), respectively, which were more controlled release in comparison with F4 and NATACYNVR Effect of sterilization on the NAT-loaded niosomes/0.5% of KT after 24 h (Figure 4). gels formulae Although the highest extent release was found for F7, its con- In order to protect the NAT niosomes-loaded/0.5% of KT gels from sistency was liquefied due to its low viscosity. This led to rejection the destruction by autoclaving, the formulae were sterilized by F7 from further investigation, whereas, the consistency of F8 was c-radiation in the presence of dry ice. The similarity factor was cal- gel with suitable viscosity, as illustrated in Table 5. culated to check the influence of sterilization on in vitro release

70 70

60 60

50 50

40 40 30 30 20 20

Percentage of NAT Released of NAT Percentage 10 10 Percentage of NAT Released of NAT Percentage 0 0 At Initial After 1 After 3 After 6 After 9 F5 F6 F8 month months months months Formula Time (Month) Before Sterilization After Sterilization F5 F6 F8 Figure 5. Effect of Sterilization on percentage of NAT released from NAT-loaded Figure 6. Effect of storage on in vitro release of NAT from NAT niosomes-loaded/ niosomes/0.5%KT gel formulations (F5, F6 and F8). 0.5%KT gel formulations (F5, F6 and F8).

Table 6. Long-Term Stability of NAT niosomes-loaded/0.5% KT gels formulae stored at room temperature, Similarity factors of NAT content and in vitro release profiles at different time intervals compared to the zero time profile (n ¼ 3). % Drug released time of storage to initial release Similarity factor of NAT Time Percent drug released after period of storage (%±S.D) (recovery) (%±S.D) released (f ) intervals 2 (month) F5 F6 F8 F5 F6 F8 F5 F6 F8 At (initial) 41.07 ± 2.74 37.32 ± 0.72 57.32 ± 1.47 100.00 ± 0.00 100.00 ± 0.00 100.00 ± 0.00 ––– After 1 month 40.01 ± 1.32 35.73 ± 0.95 56.20 ± 1.82 97.42 ± 3.23 95.74 ± 2.54 98.05 ± 3.17 84.16 92.69 71.78 After 3 months 39.18 ± 0.68 30.31 ± 0.91 55.79 ± 0.91 95.41 ± 1.67 81.23 ± 2.44 97.34 ± 1.59 80.29 77.01 81.54 After 6 months 40.31 ± 0.91 30.15 ± 0.68 54.18 ± 1.37 98.15 ± 2.22 80.80 ± 1.83 94.54 ± 2.39 85.79 75.46 87.54 After 9 months 39.99 ± 1.37 43.05 ± 1.60 53.21 ± 1.37 97.37 ± 3.33 115.36 ± 4.28 92.83 ± 2.39 67.65 74.92 70.81 932 M. A. EL-NABARAWI ET AL. profiles. The similarity factor values were 69.10, 87.21, and 64.37 Long-term stability study of NAT-loaded niosomes/0.5%KT for F5, F6, and F8, respectively. This indicates that the similarity of gels formulae the drug release profiles in the acceptable range (between 50 and Long-term stability study of NAT-loaded niosomes/0.5%KT gels for- 100) before and after sterilization as shown in Figure 5. mulae (F5, F6, and F8) was stored at room temperature for a period of storage. The drug content of the three formulae at zero (initial), 1, 3, 6, and 9 months were found to be within the acceptable range Table 7. In-vitro antifungal susceptibility by agar well diffusion method, each value represent mean ± SD (n ¼ 3). (90–102%), and the results of release extent of the three formulae Zone diameter (mm) are demonstrated in Table 6 and shown in Figure 6. Fungal strain Test item mean ± SD The respective percentages of natamycin released in the Candida albicans F5 21.00 ± 0.557 loaded niosomes/0.5%KT gels formulations after 1, 3, 6, and F8 19.00 ± 0.700 9 months from the initial formulations (F5, F6, and F8) were C.M.P. 20.00 ± 0.781 95.41–98.15%, 80.80–115.36%, and 92.83–98.05%, respectively. FDNG NA One way ANOVA reveal that there is no significant differences Control 20.00 ± 0.608 in the percentages of in vitro release at level (p < 0.05) between where F5 represents (NAT-loaded niosomal/0.5% KT 2% HPMC-E4 Ophthalmic Gel); F8 represents (NAT niosomal/0.5% KT 4% Na.CMC Ophthalmic Gel); C.M.P. initial formulae (F5 and F8) and periods of storage, where represents combined marketed product NATACYNVR and KetorolineVR Suspension p ¼ 0.564 and p ¼ 0.109, respectively and there is a significant Eye drops; FDNG represents free drug niosomal 4% Na.CMC Ophthalmic Gel; NA difference between formula (F6) and periods of storage, represents no activity; Control represents Ketoconazole as control. where p ¼ 0.0002.

Figure 7. Ocular irritancy study of NAT-loaded niosomes/0.5% KT gels; (a) before and (b) after testing.

Figure 8. (a) Rabbit’s cornea infected with Candida keratitis and hypopyon. (b) Rabbits cornea after treatment with F8 formula showing improvement of infiltration and hypopyon. DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 933

Figure 9. (a) Graph showing improvement of corneal infiltration in response to each formula. (b) Graph showing improvement of ciliary injection in response to each formula. (c) Graph showing level of hypopyon in response to treatment with each formula; Group (Type) A: received F5; Group (type) B: received F8 and Group (type) C: received CMP. Infd ¼ Corneal infiltration; Ci ¼ Ciliary injection; h ¼ Hypopyon.

Similarity factors (f2) for the release profiles of the formulations that Candida albicans are highly sensitive to NAT in formulae F5 compared to the formulae at initial time were calculated to check and F8 at low concentrations in comparison with Ketoconazole as the influence of storage on the in vitro release profiles as illustrated control (Table 7). in Table 6. The similarity factor values were 85.79, 77.01, and 87.54 Statistically, there were significant differences in the zone of for F5, F6, and F8, respectively. This indicates that the similarity of inhibition between F5, F8, and CMP at significant level (p < 0.05), the in vitro release profiles in the acceptable range (between 50 where p ¼ 0.0327. The results revealed the antifungal activity of and 100) before and after the period of storage. NAT niosomal-loaded/0.5% KT gel formulations. The MIC for NAT in F5, F8 against Candida albicans were 63 mg/ ml, and 8 mg/ml, respectively. The results revealed that MIC of F8 Evaluation of in vitro antifungal activity of the selected formulae was more effective than MIC of F5 in Candida albicans due to F5 was more viscosity than F8 led to decreasing in the diffusion of Evaluation of in vitro antifungal activity by agar well diffusion NAT around the agar wells [29,65]. methods. The results of antifungal activity evaluation revealed 934 M. A. EL-NABARAWI ET AL.

Figure 10. (a) Corneal stroma showing large number of fungal spores (GMS) from a specimen of the F5 group. (b) Corneal stroma negative for fungal spores (GMS) from a specimen of the F8 group.

In vivo studies from all pharmaceutical in vitro evaluations, and a better cure percent in experimental application. This new combination formula Visual assessment of ocular irritation potential increases the corneal permeability and ocular bioavailability which The in vivo results showed no signs of irritation or damaging decreases the need for frequent drug application and systemic effects in the cornea, conjunctiva, or iris. The scores for conjunc- side effects and also relieve the inflammation associated candida tiva swelling and discharge were always grade zero. Iris hyperemia keratitis and increase patient compliance. More working are stil- and corneal opacity scores were grade zero at all observations. ling on the same formulae to study its safety and efficacy on Therefore, the potential clinical interest of the combination nioso- other types of FK. mal hydrogel is supported because of the absence of irritant effects in vivo as shown in Figure 7(a,b). Acknowledgements In vivo study A rabbit model was used to evaluate the efficacy and safety of The authors would like to thank Faculty of Pharmacy NAT-loaded niosomes/0.5%KT gel formulae and compared with (Pharmaceutics and Industrial Pharmacy department) and Kasr Al- CMP antifungal ophthalmic suspension. The treatment was started Aini Hospital (Ophthalmology and Pathology Departments), Cairo 48 h after the inoculation; to allow fungal keratitis (FK) to manifest, University, Cairo, Egypt for them helping in success this study. and that was confirmed by clinical examination of the corneas This study did not receive any specific grant from funding agen- and photographs for documentation (Figure 8(a,b)). cies in the public, commercial, or not-for-profit sectors. Regarding the size and depth of corneal infiltration, rabbits which received F8 formula showed better improvement than rab- Disclosure statement bits treated by F5 and CMP one way ANOVA (repeated measures, p ¼ 0.043) (Figure 9(a)). No potential conflict of interest was reported by the authors. However, the follow up of the severity of ciliary injection and level of hypopyon revealed that both F5 and F8 had more or less equal results and both were better than CMP (Fisher’s Exact test, References p < 0.05) (Figure 9(b,c)). The clinical scoring of the ciliary injection [1] Whitcher JP, Srinivasan M, Upadhyay MP. Corneal blindness: and hypopyon levels used in this study was similar to those used a global perspective. Bulletin WHO. 2001;79:214–221. in Rabbit FK models previously [69,70]. [2] O’Brien TP. Therapy of ocular fungal infections. Ophthalmol These results were consistent with the findings of the path- Clin. 1999;12:33–50. ology of specimens collected from the infected cornea; F8 had the [3] Manger, T, T, Mauger E. Craig Havener’s Ocular best results regarding inflammation and fungi (Fisher’s Exact test, Pharmacology. 9th ed. St. Louis (MO): C. V. Mosby; 1994. p ¼ 0.016 & p ¼ 0.028, respectively). These results also agreed with [4] Raab W. Natamycin (pimaricin): its properties and possibil- the pathology of the excised rabbits’ corneas, as F8 formula had ities in medicine. Stuttgart (Germany): Georg Thieme the best results regarding inflammation and fungi. Figure 10 Publishers; 1972. shows the difference of the fungal spores load detected in a cor- [5] Chandasana H, Prasad YD, Chhonker YS, et al. Corneal tar- nea from the F5 group (Figure 10(a)), compared to a spore-free geted nanoparticles for sustained natamycin delivery and corneal specimen taken from the F8 group (Figure 10(b)). These their PK/PD indices: an approach to reduce dose and dos- results are attributed to the combined efficacy of NAT-loaded ing frequency. Int J Pharm. 2014;477:317–325. niosomes/KT gel; the efficacy of each drug preparation has been [6] Patil A, Lakhani P, Majumdar S. Current perspectives on proven independently in previous studies [19,71]. natamycin in ocular fungal infections. J Drug Deliv Sci Technol. 2017;41:206–212. Conclusion [7] Kumar GP, Rajeshwarrao P. Nonionic surfactant vesicular Accordingly, this study concluded that the combination of F8 systems for effective drug delivery—an overview. Acta (NAT-loaded niosomes/0.5% KT- 4%Na.CMC) has the best results Pharm Sin B. 2011;1:208–219. DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY 935

[8] Bendas ER, Abdullah H, El-Komy MH, et al. [27] Zeng X, Danquah MK, Jing K, et al. Solubility properties Hydroxychloroquine niosomes: a new trend in topical man- and diffusional extraction behavior of natamycin from agement of oral lichen planus. Int J Pharm. 2013;458: Streptomyces gilvosporeus biomass. Biotechnol Progress. 287–295. 2013;29:109–115. [9] Carafa M, Santucci E, Alhaique F, et al. Preparation and [28] Luo J, Z, Jin P, Cen M, et al. Determination and correlation properties of new unilamellar non-ionic/ionic surfactant of natamycin solubilities in water and methanol. J Chem vesicles. Int J Pharm. 1998;160:51–59. Ind Eng China. 2008;59:281. [10] Cortesi R, Esposito E, Corradini F, et al. Non-phospholipid [29] El-Nabarawi MA, Bendas ER, El Rehem RTA, et al. vesicles as carriers for peptides and proteins: production, Transdermal drug delivery of paroxetine through lipid- characterization and stability studies. Int J Pharm. 2007;339: vesicular formulation to augment its bioavailability. Int 52–60. J Pharm. 2013;443:307–317. [11] Balakrishnan P, Shanmugam S, Lee WS, et al. Formulation [30] Aggarwal D, Kaur IP. Improved pharmacodynamics of timo- and in vitro assessment of minoxidil niosomes for lol maleate from a mucoadhesive niosomal ophthalmic enhanced skin delivery. Int J Pharm. 2009;377:1–8. drug delivery system. Int J Pharm. 2005;290:155–159. [12] Abdelkader H, Wu Z, Al-Kassas R, et al. Niosomes and dis- [31] Jain RL, Shastri J. Study of ocular drug delivery system comes for ocular delivery of naltrexone hydrochloride: mor- using drug-loaded liposomes. Int J Pharma Investig. 2011;1: phological, rheological, spreading properties and photo- 35. – protective effects. Int J Pharm. 2012;433:142 148. [32] Pham TT, Jaafar-Maalej C, Charcosset C, et al. Liposome and [13] Palozza P, Muzzalupo R, Trombino S, et al. Solubilization niosome preparation using a membrane contactor for and stabilization of beta-carotene in niosomes: delivery to scale-up. Colloids Surf B Biointerfaces. 2012;94:15–21. – cultured cells . Chem Phys Lipids. 2006;139:32 42. [33] Wu ZH, Ping QN, Wei Y, et al. Hypoglycemic efficacy of chi- [14] Biju S, S, Talegaonkar P, Mishra R, et al. Vesicular systems: tosan-coated insulin liposomes after oral administration in an overview. Indian JPharm Sci. 2006;68. mice. Acta Pharmacol Sin. 2004;25:966–972. [15] Shirsand S, Para M, Nagendrakumar D, et al. Formulation [34] Marianecci C, Paolino D, Celia C, et al. Non-ionic surfactant and evaluation of Ketoconazole niosomal gel drug delivery vesicles in pulmonary glucocorticoid delivery: characteriza- system. Int J Pharma Investig. 2012;2:201–207. tion and interaction with human lung fibroblasts. J Control [16] Mohsen AM, AbouSamra MM, ElShebiney SA. Enhanced oral Release. 2010;147:127–135. bioavailability and sustained delivery of glimepiride via nio- [35] El-Samaligy M, Afifi N, Mahmoud E. Increasing bioavailabil- somal encapsulation: in-vitro characterization and in-vivo ity of silymarin using a buccal liposomal delivery system: evaluation. Drug Dev Ind Pharm. 2017;43:1254–1264. preparation and experimental design investigation. Int J [17] Bhatta R, Chandasana H, Chhonker Y, et al. Mucoadhesive Pharm. 2006;308:140–148. nanoparticles for prolonged ocular delivery of natamycin: [36] Junyaprasert VB, Teeranachaideekul V, Supaperm T. Effect in vitro and pharmacokinetics studies. Int J Pharm. 2012; of charged and non-ionic membrane additives on physico- 432:105–112. chemical properties and stability of niosomes. AAPS [18] Phan CM, Subbaraman L, Liu S, et al. In vitro uptake and release of natamycin Dex-b-PLA nanoparticles from model PharmSciTech. 2008;9:851. [37] Kakkar S, Karuppayil SM, Raut JS, et al. Lipid-polyethylene contact lens materials. J Biomater Sci Polym Ed. 2014;25: 18–31. glycol based nano-ocular formulation of ketoconazole. Int J – [19] Paradkar MU, Parmar M. Formulation development and Pharm. 2015;495:276 289. evaluation of Natamycin niosomal in-situ gel for ophthalmic [38] Korsmeyer RW, Gurny R, Doelker E, et al. Mechanisms of drug delivery. J Drug Deliv Sci Technol. 2017;39:113–122. solute release from porous hydrophilic polymers. Int J – [20] Lang JC, Roehrs RE, Rodeheaver DP, et al. Design and Pharm. 1983;15:25 35. evaluation of ophthalmic pharmaceutical products. Drug [39] Dash S, Murthy PN, Nath L, et al. Kinetic modeling on drug Pharm Sci. 2002;121:415–478. release from controlled drug delivery systems. Acta Pol – [21] El-Mofty HM, Abdelhakim MASE, El-Miligi MF, et al. A new Pharm. 2010;67:217 223. combination formula for treatment of fungal keratitis: an [40] Pardakhty A, Varshosaz J, Rouholamini A. In vitro study of experimental study. J Ophthalmol. 2014;2014:1–7. polyoxyethylene alkyl ether niosomes for delivery of insulin. [22] Martindale. 35th ed. London: PhP Pharmaceutical Press; Int J Pharm. 2007;328:130–141. ~ 2007. [41] Ocana J, Frutos G, Sanchez P. Using the similarity factor f2 in [23] Alem MA, Douglas LJ. Effects of aspirin and other nonster- practice: A critical revision and suggestions for its standard oidal anti-inflammatory drugs on biofilms and planktonic error estimation. Chemometrics Intel Lab Sys. 2009;99:49–56. cells of Candida albicans. Antimicrob Agents Chemoth. [42] Costa P. An alternative method to the evaluation of similarity 2004;48:41–47. factor in dissolution testing. Int J Pharm. 2001;220:77–83. [24] Bandara B, Sankaridurg P, Willcox M. Non-steroidal anti [43] Kaur H, Yadav S, Ahuja M, et al. Synthesis, characterization inflammatory agents decrease bacterial colonisation of con- and evaluation of thiolated tamarind seed polysaccharide tact lenses and prevent adhesion to human corneal epithe- as a mucoadhesive polymer. Carbohydr Polym. 2012;90: lial cells. Curr Eye Res. 2004;29:245–251. 1543–1549. [25] Szoka F, Papahadjopoulos D. Procedure for preparation of [44] Bansal K, Rawat M, Jain A, et al. Development of satranida- liposomes with large internal aqueous space and high cap- zole mucoadhesive gel for the treatment of periodontitis. ture by reverse-phase evaporation. Proc Natl Acad Sci U S AAPS PharmSciTech. 2009;10:716. A. 1978;75:4194–4198. [45] Kamal AH, El-Malla SF, Hammad SF. A review on UV spec- [26] Hosny KM. Ciprofloxacin as ocular liposomal hydrogel. trophotometric methods for simultaneous multicomponent AAPS PharmSciTech. 2010;11:241–246. analysis. Eur J Pharm Med Res. 2016;3:348–360. 936 M. A. EL-NABARAWI ET AL.

[46] Razem D, Katusin-Razem B. The effects of irradiation on [58] El-Ridy MS, Badawi AA, Safar MM, et al. Niosomes as a controlled drug delivery/controlled drug release systems. novel pharmaceutical formulation encapsulating the hepa- Rad Phys Chem. 2008;77:288–344. toprotective drug silymarin. Int J Pharm Pharm Sci. 2012;4: [47] Hathout RM, Mansour S, Mortada ND, et al. Liposomes as 549–559. an ocular delivery system for acetazolamide: in vitro and in [59] Hasan AA, Madkor H, Wageh S. Formulation and evaluation vivo studies. AAPS PharmSciTech. 2007;8:1–12. of metformin hydrochloride-loaded niosomes as controlled [48] Nesseem DI. Formulation and evaluation of itraconazole via release drug delivery system. Drug Deliv. 2013;20:120–126. liquid crystal for topical delivery system. J Pharm Biomed [60] El-Gazayerly ON, Hikal AH. Preparation and evaluation of Anal. 2001;26:387–399. acetazolamide liposomes as an ocular delivery system. Int J [49] Fetih G. Fluconazole-loaded niosomal gels as a topical ocu- Pharm. 1997;158:121–127. lar drug delivery system for corneal fungal infections. [61] Mokhtar M, Sammour OA, Hammad MA, et al. Effect of – J Drug Deliv Sci Technol. 2016;35:8 15. some formulation parameters on flurbiprofen encapsulation [50] Hindler J, Howard B, Keiser J. Antimicrobial agents and anti- and release rates of niosomes prepared from proniosomes. microbial susceptibility testing. In: Howard BJ. Clinical and Int J Pharm. 2008;361:104–111. Pathogenic Microbiology. 2nd ed. St. Louis (MO): Mosby; [62] Jones DS, Woolfson AD, Brown AF, et al. Design, character- 1994. isation and preliminary clinical evaluation of a novel [51] Tayel SA, El-Nabarawi MA, Tadros MI, et al. Promising ion- mucoadhesive topical formulation containing tetracycline sensitive in situ ocular nanoemulsion gels of terbinafine for the treatment of periodontal disease. J Control Release. hydrochloride: design, in vitro characterization and in vivo 2000;67:357–368. estimation of the ocular irritation and drug pharmacokinet- [63] Habib F, M, Abdel Azeem G, Fetih M, et al. Mucoadhesive ics in the aqueous humor of rabbits. Int J Pharm. 2013;443: buccal patches of lornoxicam: development and in-vitro 293–305. characterization. Bull Pharm Sci Assiut Univ. 2010;33:59–68. [52] Draize JH, Woodard G, Calvery HO. Methods for the study [64] Baloglu E, Ozyaz€ ıcı M, Yaprak Hızarcıoglu S, et al. of irritation and toxicity of substances applied topically to Bioadhesive controlled release systems of ornidazole for the skin and mucous membranes. J Pharmacol Exper Ther. vaginal delivery. Pharm Deve Technol. 2006;11:477–484. 1944;82:377–390. [65] Hoare TR, Kohane DS. Hydrogels in drug delivery: progress [53] Guinedi AS, Mortada ND, Mansour S, et al. Preparation and – evaluation of reverse-phase evaporation and multilamellar and challenges. Polymer. 2008;49:1993 2007. niosomes as ophthalmic carriers of acetazolamide. Int J [66] Jones DS, Woolfson AD, Brown AF. Textural, viscoelastic Pharm. 2005;306:71–82. and mucoadhesive properties of pharmaceutical gels com- [54] El-Ridy MS, Abdelbary A, Essam T, et al. Niosomes as a posed of cellulose polymers. Int J Pharm. 1997;151: – potential drug delivery system for increasing the efficacy 223 233. and safety of nystatin. Drug Dev Ind Pharm. 2011;37: [67] Attia DA. In vitro and in vivo evaluation of transdermal 1491–1508. absorption of naproxen sodium. Aust J Basic Appl Sci. [55] Kashi TSJ, Eskandarion S, Esfandyari-Manesh M, et al. 2009;3:2154–2165. Improved drug loading and antibacterial activity of mino- [68] Conti S, Maggi L, Segale L, et al. Matrices containing cycline-loaded PLGA nanoparticles prepared by solid/oil/ NaCMC and HPMC: 1. Dissolution performance characteriza- water ion pairing method. Int J Nanomedicine. 2012;7: tion. Int J Pharm. 2007;333:136–142. 221–234. [69] Ozturk F, Yavas GF, Kusbeci T, et al. Efficacy of topical [56] Cevher E, S¸ensoy D, Zloh M, et al. Preparation and charac- caspofungin in experimental Fusarium keratitis. Cornea. terisation of natamycin: c-cyclodextrin inclusion complex 2007;26:726–728. and its evaluation in vaginal mucoadhesive formulations. J [70] Kalkanci A, Ozdek S. Ocular fungal infections. Curr Eye Res. Pharm Sci. 2008;97:4319–4335. 2011;36:179–189. [57] Nasr M, Mansour S, Mortada ND, et al. Vesicular aceclofenac [71] Manjappa A, Nanjwade BK, Manvi F, et al. Sustained oph- systems: a comparative study between liposomes and nio- thalmic in situ gel of ketorolac tromethamine: rheology and somes. J Microencapsul. 2008;25:499–512. in vivo studies. Drug Dev Res. 2009;70:417–424.